CA2246431A1 - Recombinant expression of proteins from secretory cell lines - Google Patents

Abstract

The present invention provides methods for production of heterologous polypeptides, for example amylin, using recombinantly engineered cell lines.
Also described are methods engineering cells for high level expression, methods of large scale heterologous protein production, methods for treatment of disease in vivo using viral delivery systems and recombinant cell lines, and methods for isolating novel amylin receptors.

Description

RECOMBINANT EXPRESSION OF PROTEINS
FROM SECRETORY CELL LINES

WO 97126321 PCTrUS97/00761 I. BACKGROUNI) OF THE IIWENTION

The present application is a continu~ion-in-part of U.S. Patent Application Serial No.
Serial Number 08/589,028, filed January 19, 1996, and U.S. Provisional Patent Application Serial No. Serial Number 601028,279, filed October 11, 1996. The entire text of each of the above-referenced disclosures is specifically incorporated by reference herein without r1iCcl~impn A. Field of the Invention The present invention is related to the recombinant expression of proteins from 10 eukaryotic cells. More par~icularly, the invention relates to the production of recombinant proteins from gen~tir~lly eng...ce.ed secretory cells. Methods for use of the cells also are provided.

B. P~lst~l Art ~ mm~ n cells of neuroendocrine origin have been used extensively over the last fifteen years as systems for the study of pa~l~ays and ~r~ k~ mc of polypeptide secretion (Burgess and Kelly, 1987 and Chavez et al., 1994). FY~mples of cell lines in which such studies have ~ecn carried out include the mouse piluil~u~/ line AtT-20 (ATCC CCL 89), the rat pilui~
growth ho~ one secreting lines GH3 (ATCC CCL 82.1), the insulin secreting ~TC lines derived 20 from ~ r ~ic mice expressing SV40 T antigen (Efrat et al., 1988), radiation in~ recl, rat islet cell tumor derived RIN lines (Gazdar et al., 1980) and the rat adrenal pheoch~,.,locytoma cell line PC12 (ATCC CRL 1721). These cell lines ...~ rnany of their çnAogenous functions, inr~ g synthesis of peptide ho. . .~on~s destin~d for the regulated se.,letol ~ patll~ay. These cell lines also are transfectable, allowing e~plession of novel transgenes for studies of heterologous 25 protein ~ .t~s.

T~ee major areas have been studied using these heterologous systems. The first is the study of the sorting ~ m, whc~by a given protein, ~estin~od for secretion, is targeted to the regulated se~lo", pathway or the dcfault consLiLu~i~re secretory p~ way. The second relates to 30 ~ n~rling the cQ~T~rlex process of se.,,~to,y protein maturation This would include the W O 97126321 PCT~US97/00761 specific steps of protein folding, ~ lfi~o bond formation, glycosylation, endoproteolytic pluce~ and post-tr~n6~ Qn~l morlifir~ions of sperifir amino acids as well as under~tq-n~lin~
the erlL).~l~es involved in thcse processes. And the third relates to control of the regulated release of peptide h-)rmnn~.s from secretory Er~n~ s following physiological stimuli.

Ncuroçn~tornne cell lines havc been generated in which genes erlro~line specific peptide hc~ ol}cs have been stably inserted. These cnLyll~cs include insulin (Moore et al., 1983, Powell et al., 1988 and Gross et al.. 1989), somq-tost~tin (Sevarino et al., 1987), thylùl~u~ cing holl..ol~ (Sevarino et al., 1989), nc~lup~,~tide Y (Dic~ oll e~ al., 1987), insulin-like growth factor-l (Schn~idt and Moore, l994), ~Jluopio~ nocortin ~Thorne et al., 1989), glucagon ~Drucker et a~., 1986 and Rouille e~ al., 1994), panclealic polypeptide (TqL-~lrhi et al., 1991) and growth ho.. nr (Moore and Kelly, 1985). In general. heterologous expression of these l~lul~ills has ~ n~ r~d faithful sorting to the regul-ted secretory p alh~ ay, as well as ~ ul~ion of the protcms in thc se~ ul~r granules. However, the c~p~ssion levels of thc heterologous plot,h~s 15 have gçn~r~lly boen low when CO.l~ ,d to normal end~5c"ous e,~lession of the same p~teh~s in a ~ gous system Neuroenr~ornne cell lines e~ ssing the enzymes involved in the ploce~..g of peptide ho....t~ s in se~.~tol~ granules also have been generated. These include the endoplûtoases PC2 20 and PC3 (Ohagi et al., 1992, BenJann¢t et al., 1993, and Rouille et aL, 1995) and peptidylglycine alpha~ t~l---g mono~y~ nase (PAM) (Milgram et al., 1992 and Yun and Eipper, 1995).
O~ ssiQn of these ~lvcrc,~ c~Ly~les has hclped A~ssect thcir relative col.l.;k~ionc to pep~de h~ n~e ~lUce4c.~g as well as their intr~r~ r sites of action. These studies te the ~ iemir use ol l~ue!lAocrinr cells in studying the regulated Se~ ,tol~y 25 pathway.

A seIies of papers over the last five years has addressed the po~$i~ y of ~ludue~ion of heterûlogûus peptide hn..~ u~5 in n~u~ lorrin~- cells. Three ûf these repûrts (S~mh~ni.~ e~ al., 1990 and lg91, ~rnpp e~ al., 1992) use previously est~ hed AtT-20 lines e~ h~g either 30 insulin ~Moore e~ al., 1983) or growth hn....--~.r (Moore and K¢lly, 1985). The highest level of WO 97/26321 PCTnUSg7/00761 secretion of insulin under srim~ tP~ conAitior~ was in the range of 35 to 144 microunil~/million cells/hour (equiYalent to 1 to 5 ng insulin/million cellslhr). Growth hol...rn~ secretion under stimn~ted conditions was 130 to 340 ng/million cells~our. l'hese levels of production are well below those reported in the literature for growth hormone production from other l~co,..h;n~
S ~y~",s ~Pavlakis and Hamer, 1983 and Heartlein et al., 1994). Another study dealing with protein production from a neuro~n~ornne cell makes use of an inc~ om~ line e-n~...r-e.~,d to express plol~lin (Chen et al., 1995). ~hsolllte levels of l~roduclion of prolactin on a per cell basis are not reported. A n~,~oe~ oçt-ne cell-based system for either in vitro, biologically active peptide h.~ ,.c production or for in vivo, cell-based delivery of biologically active peptide 10 h~lmmn.-s has not been achieved in any of these earlier studies.

A ~ be~ of hl~ t fcd~ s must be addressed before a n_.l.oe~ ocr ne cell-based system for protein production can be developed. The firs~ feature is the ahsolute level of ~l~uclio~l of the ~ e~ide in q~ sti~-n A ~..rr.. ;Çhtly high level of production to make either 15 in vitro pl~rifir~tion or in wvo efficacy must be achieved. As stated above, while many groups have ,~,~o~ted expression of recombinant proteins in nc-l.oe'~ornnf~ lines, the proteins are plo~ d at very low levels.

A second feature is quantitative l~lOCCS'~g of the peptide to their biologically active 20 folms. Ne~ Ao~ r cell lines ...~ .in variable levels of the enL~-Ies ~._s~onsible for peptide hc,.. ~l-r ploc~c;~g and in many lines the C~IL~rll1C levds may be i.. ~.. rr.. ;~ to ensure sllffi~ i~nt prooessing. This is a critical ~ r, espe~ ly as ~ nc are made to e ~ Pcl high level n of specifi~ pcptide hu~ one tral-cgenes A tbird feature is the "~h;nt~ n-~e of a dynarnic l~spollse of the re~ ted secretory pathway. For both in vivo and in vitro use of a l~oendocrine cell-based system, the ability to quic}ly re~ease high concentrations of the b~ g~ lly active peptide by e~ e~ r stimuli is important. ln vivo moA~ on of peptide hn~mn~ release is requircd for ~itrating the biological efflcacy of the cell-based delivery. ~n vitro mn~lu~ n of peptide h-~....nl~ release est~li.ch~
30 Grr;~ p~lu~t;ol of _jghly ~ h~ fractions of starting m~t~ri~l fors~l.~e~ l purification.

CA 0224643l l998-08-l2 W O 97~6321 PCT~US97/00761 Yet another feature is the ability to further engin~er functions into m,~oc~-dorrin~ cells other than just the high-level production of a given polypeptide. This further engi--c ,h~g could invo}ve ~ the cells capa~ilities such tnat any of the three previous points are i~llpro~ed S or st~ ili7Pd (i.e., inc~ased protein levels, increased plucec~ g effici~.-rles or h~;,.,ascd dynamic regulated se~ t~ ei,yl~n~e).

A final en~ P~ ;.-g~ ,u~e. that may proYe cignifi~nt is the ability to reduce orc~., 'et~ly ablate the endogenous expression of an ullwdl~ted gene product. Reduction or 10 ~bl~tion may result in an improved capability to produce, process or release the heterologous po~ypeptide. This also may confer advantages by removing unwanted or cQnt~min~ting biological ~,lopc.~ies of the endogenous peptidc h-s....onF Endogenous pcptide pro~uction also II~ight Cuu~ aCt the biological activities of the exog~.lous peptide h~ r bcing prod~lce~l resulting in ull~ ted immllnolog~ical reactions, re~lnring the c~acity of the rn~ ,d lines to 15 ~uanlil~i,cly produce the exogenously en~..Pel~d protein or complicating pu~ifir~tion of the exog,~ ~o~ly ~ hlced protein. Because all of the ~yicting nculc~-ndcY,;~-e cell iines ploducc en~og~ o-.c secl~ted proteins, these concçrnc are .cig~.;fi~nt Thus, despite the benefits of developing a secretory cell line in which the protein 20 synthctic ~ hin-,/~r has been comInandeered for the production of a hetcrologous polypeptide, there appear to be ~iy,,.;r.r~-.t ~chni~ obs~ s that are not addressed by the art. As a result, thcrc ~ tly cxist no ~ ,d cells that address all of these problems.

II. SU~ARY OF THE INVENTION
The present invention ~c~uns to the engjl~r,F .ng of ~ n cells for production of he~erologous protcins, for ey~mr~ in t-hle ~ v~ of s~,e,ted peptide h.. ~ne5 In particular the p~:scnt inventioD describes nlt~th~c of er~p,~..r,~ g cells for high levcl ex~l~,ssion of a variety of proteins ;~ r amylin and mctl~ ~s of large scale protein pr~l~ ion Further, the present WO 97~6321 PCTrUS97/00761 ion describes methods of treatment of disease and also methods of isolating novel amylin ~C~IOl~.

There is provided, according to the present invention, a method of en~in~ ering a 5 m~mm~ n cell cc~ lising providing a starting cell, introducing into the starting cell an arnylin-enroriing gene operatively linked to a first plulllotel, and selectin~ a cell that exhibits incl~,ased amylin production as col~ ,d to the starting cell. In other embotlim~.nts, the method of C~r;~ g a ".~-I"~ n cell further cV---I--;ces introd~lrin~ into the s~l~ct~l cell an insulin-e~ orli.~ gene operatively linked to a second pfun~otel.
In particular embo~imrnts~ the starting cell produces amylin, and in other embo-iim~ ntc the starting cell does not produce arnylin either naturally or as a result of e~ . The star~ing cell may be human or non-human. The starting cell may be se~,tcl y or non-secn,t~y.
The start~ng cell may bc a nc~ucn~orrin~ cell, a beta cell, or a pilui~ cell. It also may be 15 se~- ~ti~go~;u~ ,,pol~;v~, glucose~ oni.ive or non-glucose-l~sponsi~,.,. In pl~,f~ d e.. kP1;.. 1~ of the present invention the starting cell is derived from a ,BTC, RD~, HIT, BHC, CM, TRM, TRM6 AtT20, PC 12 or HAP5 cell.

The arnylin produced by the el~ r~ l~,d ~ n cells may be proteolytically 20 I,,oce~ mifl~tç~, andJor ~Iycos~lated. Glycosylated amylin species of the present invention may be O-~l~c~ylated or N-glycosylated. When the amylin species of the present invention are os~l~t~,d, such ~ly,~ lation may c.,...~ c an oligosaccllaride linkcd to ll~cc~ .c-9 of an amylin of SEQ ID NO:~l or SEQ ID NO:53. Alternatively, the glycosylated amylin co...l.~;ces an Q~ ?~ch~. ,de linked to lL.~or,ine-6 of an arnylin of SEQ ID NO:5 1 or SEQ ID NO:53.
ln certain ç.~.hor~;.. l~ the ~lyco~.ylated amylin of the present il~ve~l~ion cq.. l.. ;.ces an oligosacchande ~ u~lu~'C having '~t~.~n about 3 and about l0 s~c~b~ e units. The s~( ch5 ;d-s may be s~ 1 from the group CQ~Q~ .g of ribose, arabinose, xylose, Iyxose, allose, altrose, gluoose, mannose, rluct~se, gulose, idose, galactose, talose, ribulose, sorl~ose, tagatose, &1V~Q~;C
30 acid, gl~iulo~lic acid, glucaric ;U~ ronic acid"I.~....-ose, fucose, N-acetyl g1.lcos~...;..e, N-W O 97126321 PCT~US97/00761 acet-yl galactos~Tnin~7 N-acetyl neuraminic acid, sialic acid, amino glycal and a substituted amino glycal or any other s~ ch~ri~e unit commonly present in glycopivteins. In ~l.,fc.ll,d e~ho~ the glycosylated amylin species of the present invention C~J~ ;Ce ~NeuAc,HexNAc2]Gal(,B 1 -3)GalN~c.

In particular embo~ e~ , the first ~ ot~l is selected from the group con~cicting of CMV E, SV40 E, RSV LTR, RIP, m~ified RIP, POMC and GH. Ln other embo~ the second ~lolnotel is sele~ted from the group co~-c~ g of CMV E, SV40 E, RSV LTR, RIP, ",o~l;fi~ RIP, POMC and GH.
The amylin el~ ing gene that is eng;,~e.,.~d into the ",~ cells may be a hDn amylin e ~cos~ g gene, and may be linked to a sçlc~t~ble markcr. The sekct?ble marker may be selPct~-d from a group concisting of hyg~ cin resistance, neomycin lr,;c~ e, I)Ul'Olll~il l~-;c~ rG. zcocin, gpt, DH~:R and hict~inol. Likewise, the insulin gene in other e...ho-l;...~ nl-j 1~ may be linked to a scle~ le marker sele~te~l from a group cQnCicting of h~ ul~ ci~ e, ne~,~l,y-,iL r~ e, puromycin reSi~t~nre~ zeocin, gpt~ DHFR and hi~t~iinol The amyliD of the present invention may be an analog of human amylin and may also be a non-amyloidogenic analog. ~IteTn~tively, the amylin gene used in the present invention is a rat 20 amylin e~ g gene or a rat amylin analog en~o~i..g gene.

Also provided is a method of providing amylin to a .~ 1 oolllylising providing astar~ng cell, i~ n~ into the starting cell an amylin ~nro~ g genc opcratively linked to a first promotcr; selecting a cell that exhibits increased amylin pro~ ction as Co~ d to the 2~ starting cell, and administering the srk~t~ cell to a ...~ h 1II cc~ain e~ it is envisioned th~lt the 1"l- "-~.~1 to which the amylin }s provided e~chihitc at least one pathologic condition sel~ctP~l from the group consisting of angiog~nPci~
gastric ~ .~g, anore~ia, obesity, hypertension, hyperc~ mi~ Pagets discase and O~'~pOluiiS.

wo s7n6321 PCT/US97/00761 The a~rninictP~ing of the SP1PCt~ cell may be achieved by (i) ~1-c~ ul~ing the selected cell in a bioco pa~ihle coating or (ii) placing the cells into a selectively pe...lcable nle~ c in a protective housing. ~ItPrn~tively, the cell is ~rlmini~tered int-ay~,litoneally ~ ~cously or 5 via the CNS In ~mini~t~ring the selectP~l cell to the m~mm~l, the cell may be contained within a selcctively semi-p~ ,eable device, the device being connected to the vasculature of the ",~..,...~1.

Other embo-limPnts provide a method of procl~lcing m~mm~ n amylin compricing 10 providing a starting cell. introducing into the starting cell an amylin-encoding gene operatively linked to a first ~ro.l.ol~, selPc~ a cell that exhibits i-.cl~,ased production of amylin as cc -.~ ,d to the starting cell and c~ ing the s~ t~ cell. The m~-~Q~I of pro~uoing amylin rnay further co.. ~ e thc step of purifying the arnylin.

Also c<n~e .platcd in the presient invention is an amylin produced acco. lil,g to a process com~ricing the steps of providing a starting cell, intro~ cing into the starting cell an amylin-e.co-l; P gene Op~a~ ,ly linked to a first promoter, sel~cting a cell that eYhi~it~ increased ~J~vdu~,lion of amylin as co~ d to the starting oell and c~lltllnng the select~d cell The amylin produce may be secreted and the process may also comprise the step of purifying the amylin The process may further cornpri~e the step of intro~ ring into the cell an insulin-encoding gene opcratively linlced to a second ~ -ot,r. In ~..,f~ d embo~ nts, where amylin and insu}in are both c~ ~cd within the cell, the amylin-to-insulin content of the s~lcc~,d cell is b~,t~ about 0.002 to about 10 0 In other cll-boA;~ . the amylin-to insulin ratio may be ~t~ ~.. about 0 005 to about 9 in yct other e~ tne ratio may be b~t~ cen about 0.01 and ~, in other e hG~; - r -ts the ratio may be b~tw~n about 00~ to about 7 in otber the ratio may bc b.,t~ ~ about 1 and about 6 in furtber e-lnho~ the ratio rnay be b~ ~n about 2 and about 5 or ~h ~e,- about 3 and about 4.

Also provided is a method of regulating blood glucose levels in a m~-nm_l comprising p~ovidil,g a star~ing cell, introducing into the star~ing cell an amylin-en(~-oding gene oyc~d~ ly linkcd to a first promoter; selecting a cell that exhibit incn,~sed amylin secretion as co,llp~h~ed to the star~ing cell, and ~Aministering the selectPd cell to the m~rnrn_l, whe,~y the secreted amylin 5 regulates blood glucose levels of thc ~ 'AI In ~,ef~ d embo~lim~nt~ this m~tho~l may further comprise providing the cell with an insulin-enro~ing gene oyclalively linked to a second pl~lUVt~.

~ still other embo-lïm~iAt~, there is provided a method for m~lAting the circ.li-ting 10 levels of insulin in a m~nAm~l COlllyli~illg the steps of providing a starting cell. introducing into the cell an amylin-encoding gene o~alivcly linked to a promoter; selecting a cell that exhibits illcl~ased amylin scc~,lion as ccn~,a..,d to the starting cell and ~Amini~tcring the seJectP~l cell to the . .~ 1 wh~ the secreted amylin modulates ~ucose-stinA,~ i insulin secretion in the .tl,..,, ,,..l 1~
Also provided is a method ior d~c,~,asillg ~l~o~,_n S~ lCSiS in a ~ 1 comprising the steps of providing a starting cell, i~ll~hn,;ng into the cell with an amylin e~coli~g gene op~ ly linked to a first ~ , selecl ;..g a cell that exhi~it~ in~l~,ased arnylin secretion as Co~ d to the slarting cell and ~ clering the selecte~l cell to thc ...~..n.~l, whereby the 20 se~ t~d amylin reduces glycogen synthesis in the 1-..,....~h In other embo~ , the secretory cell further is tr~n-cfect~d with an insulin-enrorlin~ gene opc.~ ly linked to a plOlll~t~l.

Tbere also is provided a method of SC,l~U~g for an amylin receptor compricin~ obtaining amylin from a l~o..~hi1~q1-~ amylin e..y.,ss~g se~ toly cell; ~.1.,,;~;~,g the amylin with a 2~ con~l~r~ cotnrricing a putative amylin l~,CG~IOl, and det~ an amylin receptor bound to the amylin. In ~ fe.l,,d embo l;... ~ , the co...~i1;on comrristng a lJu~i~e amylin lec.,l,~or is a composition CQ...~ g a pop~ tion of reco-.-hi~ t cells ~r~ ~r~ed with por~ions of a DNA
library. ln ~l~,f,.l~,d ~ ~bo~ tc the m.~tho~l funher co...~ es obt~init~ a DNA sc~;,..~..t from the I:~NA library that eA~ ie S an ;unylin l~cc~.l.

In certain emborlimPntc there is provided an amylin receptor gene ~ c~. d by theprocess of obtai~hlg amylin from a r~ col"bil~ant amylin e~ ssing secretory cell ~(~mixing the arnylin with a co",lJosilion comprising a putative arnylin receptor the composition co...l. ;.c;,lg a population of recombinant cells ~ cr~e(l with portions of a DNA library and vbtai~ hlg a DNA
5 segm~nt from the DNA library that e~ sses an amylin r~ccl)tol.

Other embo~ i of the present innovation provide an arnylin l-ce~tor-like gene Wh~ at least a portion of the gene hybn~li7oc to at least a portion of the amylin receptor gene i~lentifi~l under low string~n~y hybridization co~;tio~C. In still other embo~ e .tc. the present 10 invention provides a purified amylin col,lposition cG",l"ising an SlTnir1;3te(~ and glycosylated amylin polypeptide.

In other aspects the present invention pertains to the engine~ .ing of m~mm~ n cells for prodncti- n of hct~clogo~s proteins for eYAmr}e in the pro~ c~io~ of secreted peptide o....l~ ~rs These m~mmqli~n cells also may be e~gin~ered such that production of at least one en~C.ge~ous gene is blocked by -~os~ r e~ Pe.;i~g i.e., pennitting the usurping of the h;.~P~r for the pro~ucLion of the heterologous protein.

Thelervl~ in one elllbo~l;lllrlll there is provided a metnod for producing a polypeptide 20 co..~ g providing a SCCl~tu~ host cell~ blocLing the production of an end~ .o--~ se~ l~d pd~ e cont~ with the bost cell an cxogenous polyml~leoti-~e cornrn~ing a gene c .co~l;ag an ~YO~ U~ pol~ Wh~ ill tbe gene is under the control of a promoter active in euhryotic cdls and c~ ring the secretory host cell under conditions such that the exog.---ol~s poly.~ o~ e~ ;.scs the exogenous polyl G~ f.
Ln particular embo ~ .L.~ the promoter is sel~ted from the group co~ ;ng of CMV
SV40 E RSV LTR, G~PHD and R~l. The eYo~nous polyr~cleoride may further co...~ e an a~..o.;lus tripar~te 5' leader se~u~,~ce and intron, and the intron may ~ e the 5' donor site of thc ad~ ~,.u~ maJor late L~ans~ JI and the 3' splice site of an i........ oglobulin gene. The e~rfiC,,-~vl~s polyn~ eoti-le may further comrnee a polyadenylation signal.

W O 97~6321 PCT~US97/00761 The secl~tuly host cell m;ay be a n~o n~loçrinP cell, such as an in~lllinoma, more par~icularly, a rat in~lllinomA cell or a human inclllinoma cell. It also may be ghlcose lespon~ive or non-glucose l~s~onsive.

l~e exogenous polypeptide may be secreted, Ami~lAte~ or a fusion protein. ~mirlAted polypeptides include c~lritorin cAl~itQnin gene related peptide (CC~ c~Alritnnin gene related pepti~le, hyperc~k~PmiA of TnAIi~nAn~y factor (140) (PTH-rP~, parathyroid hormone-related protein ~107-139) (PI'H-rP), p~lhylOid l~llllolle-related protein (107-111) (PTH-rP), ~ho~ okinin (27-33) (CCK), galanin m~ss~e associated peptide, ~ .g~lAnin (65-105), gastrin I, gastrin ~. 1PA~jn~ peptide, glucagon-like peptide (GLP-l), panclc~ pancreatic peptide, peptide YY, PHM, secretin, vasoactive int~stinAl peptide (VIP), oxytocin, va30pl~ssin (AVP), ~ACOtoc;~., e~ hAl.n~, on~PphAlin~m;~, metul~ e (a~ ol~hin), alpha mP~ e stimlllAtin~ hormone (alpha-MSH), atrial n~iul~,lic factor (~-28) (ANF), amylin, 15 amyloid P colll~ollc,lt (SAP-l), cortico~l~in l,~A ;~g ho....~ e (CRH), growth ho.,..-l~
. lF~ factor (GHRH), hlt~ ;";~ g horrnore-l~lPA~ g hol.llone (LHRH), n~,~opcl~lide Y, s~ re K (neurokinin A), sul~s~ e P and Ihylullù~in l~ g hollln~ne (TRH).

The exogenous polypcptide may be a hollllûl~c, such as growth hormone, prolactin, 20 plAArent~ l~togr~ te 1~i7i~g ho, ...~ e, folliclc-r~iml-lAtin~ holl.,onc. chorionic gonadotropin, thyroid-stimulating hol~ c~ lcptin, adrcnocortic~ .,pin (ACTH), an~ioten~ I, angiotensin II, ~ ç.~ yk~ ~-m~-lA ~Ie stiml~la~ing l-.o....n~ -MSH), chrl- J -'-inin, ~ Qth~lin I, galanin, gastric inhi~it~ry pcpticle (GlP), gll~ol, in~lllin, lipo~c~uls, n~.lluph~ins and somatostatin. In the case of insulin, r~,co~h;~ t cells having an insulin contcnt of at last about 1000,1250, 1500 and 2500 ng per lo6 cells are provided. Recombinant cells producinp 200, 300, 400, 500 and 1000 ng of insulin per 106 cclls per hour also are provided. Recolllbinal~t cells secre~ng at least 25 ~g of hurnan grûwth h..,...ol~e per 106 cells per hour, at least 50 ~g of human growth i o....-~,.c per 106 cells per hour and about 200 llg of human growth ho...u ne per 1 o6 cells per hour are provided.

W O 97126321 rCTrUS97tO0761 The exogenous polypeptide may be a growth factor, such as epirl~rm~1 growth factor, platelet-derived growth factor, fibroblast growth factor, hepatocyte growth factor and insulin-like growth factor l.

In a particular embodiment, the endogenous, secreted polypeptide and the exogenous polypeptide are the same, for example, where both the endogenous. secreted polypeptide and the exogellous polypeptide are insulin.

In another emho-1im~ont the exogenous polypeptide enh~nces the production and/orsecre~ion of at least one polypeptide produced by said cell, for ex~ le, a protein pl~esci~
e~ le, a l._ce~tol and a ~ .on factor. F.~mples include hexokin~ce, g111ccl-in~c~o GLUT-2, GLP-l, IPFl, PC2~ PC3, PAM, g1~ago~-like peptide I receptor, ~lucose-~epçn-lent inQ~1inntropic polypeptide lecel)tor, BIR, SUR, GHRFR and GHRHR.

1~ Other elements that may be inrl1l~le~ in the construct are a selec~ lc marker and an intern~l ribosome entry site.

Methods for blocking of produclion of an en~logel-ol~c, secreted polypeptide include e~ ssion of an RNA ~n~ nce to the DNA or mRNA c~ll.,sponding to the endogenous, 20 sec,~d polypepti~e, r~nction of libo~yIlle specific for the mRNA of the endo~ ovs, secreted pol~ e, int~,.lu~" of the gene .,ncoding said endogenous, secreted polypeptide by hf~ gr~ lcc~ ion genomi~ site di~ected Tm1t~g~ ci.c or random il~te~tion. . As used he~in, genomic site directed mnt~g~nesic may employ RNA:DNA oligG"l~de.Jt;-le,c or DNA:DNA oligon11cleo~i~es.

Also &o~t~ pla~ed by the present hnr."ltivu are large scale production ...~ c incluAing stimng a s~ ..cior of the s~;l~tuly host cell, gas stream agitation of a s ~cncion of the SCC~ y host cell, i~ .d.~;on of thc secretory host cell in a non-~ lu~ tt:-~h~-1 cdl con~,.;.-f.~
or a ~ scd ~ h~ cell cont~in~r~ culture on micloc~lie.s. ll~iclve~ tion of the W O 97/26321 PCTnUS97/00761 secleto, ~ host cell, followed by cell culture and inc~ tion of the secretory host cell in a ~c.ruscd packed bed reactor.

- Also provided is a method of preventing type I rli~ etes comprising identifying a subject 5 at risk of type I r~ et~s and providing to the subject a polynucleotide ComrricinE a human insulin ~-chain gene, wh.,.~ln the: ~-chain gene is under the control of a promoter active in eukaryotic cclls. The providing may comrrice introducing the polyl,~lcolide to a cell of the subject in vivo. Alternatively, the providing co...l..ice,C cont~cting with a SC.~I~tGI~ host cell ex vivo and ~lminict~.ring the secretory host cell to the subject. Further, the e~l.,ssion of the 10 ~,ndog~ ous insulin ~-chain in said Se~;lGIOI,~ host cell may be blocked. An advantageous vehicle for providing of the L~Glyl.~cleolide is in a p~c~ge~ le, ~ iQn defective adenoviral e..~l~ssion COIl~lluCl.

A further embor~ t includes a method for treating a subject ~fflirte(l with ~ b~,t~s 15 cc~ g idcnliîyil.g a subject ~mtrted with diabetes and providing to the subject a se.,let~..y hos~ cell, Wll~ iA (j) the l.lc,duclion of an en~o~nous, secreted polypeptide has been blocked and ~ii) WllC~C;il1 the sec.etu.~ host cell comrr~ c an exogenous poly~~uclcolide comrrising a gene en~o~ling insulin, ~L~"~iill the gene is under the control of a ~luluiotcr active in e~ rolic cells.
In yet another Pmho~1im~nt, there is provided a method for providing a polypeptide to an animal ce....~ e the step of providing to the animal a sc~,to.y host cell, wl-c,~;n (i) the n of an e~ oO,..~ s, se~,~t~d pol~lidc in the secl~toly host cell has been blocked and (ii) wl~lei~ the secretory ho,st cell conl~ rs an exGgel ~ s polyr~lrleo~i~e CQ...~ .g a 2~ gene ~ ~co~ -g the polypeptide, wh~lein the gene is under the control of a pr~ te. active in e~ ic cells.

Other obJects, feahlres and adv~nt~s of the present invention will becGll-e ap~ .t from the following de~iled de,sc.i~on. It should be ....~ ood, ho ~ ., that the ~et~ d 30 d~ ;oll and the specific eY~ ~n'~s, while ir~Aic~ g ~ f~ ,d embo~; ... ~lx of the invention, W O 97~6321 PCTrUS97/00761 are given by way of illustration only, since various ch~n~s and m~ tlifi~ ionc within the spirit and scope of the invention uill become ~pa~nt to those skilled in the art from this ~le~ pd des~rirtion.

5 III. BRIEF DESCRIPIION OF THE DR~YINGS
The following drawings form part of the present spe~ifici~tion and are inc.llldeA to further rl~ ..O1~ ~e certain aspects of the present invention. The invention rnay be better understood by efe..,nce to one or more of these drawings in colnbini~tion with the Itet~i1ed description of S~ lC embo~ tc presented herein:
FIG. l: Map of wild-ty,pe HKI allele. vector for re~ rn~P~t~ and disru~ted HKI allele.
Arrows h~ the direction of transc.i~lion of hexokinase 1 (El for exon I shown), neomy.,in 1Lci~ . c (positive selection gene) and the hsv-tk (negative selection gene).
Oligos 1, 2, 3 and 4 used in PCRlM analysis are in~iic lted Capital bold letters in~ ç
,~ on enzy~.c sites h~ luced by the knock-out vector and lower case letters indicate sites in the en~log~nol-c gene. b, B = ~ n~l; e = EcoRI; k = KpnI; N = NotI; X = XhoI.
The 16 kB KpnI fr5~lP.nt cloDed from RIN 1046-38 gen- mic DNA is il-~;rated as well as thc probe used in ~.~o..~ Southems (FIG. 2).

~fG. 2: Ce1~ol.. ic Southern col-fi.. i~p hexokinase I ~ene disruption. The probe (h~ hPd rect~ , Fig. 1) is a 1 kB Pst I Lra~l~llt u~s~ of the .cch.,.hirtation site. ~çnomic DNA was digested with NotI and Eco~I. The DNA in each lanc is as follows: first lane, RIN 104~38; second lane, RIN-52/17 c4r~ a randomly intc~ ed HKI ~l~r~
vector; and lane 3, RIN-52/17 crnl~inil~g a disrupted allele of tne HKI gene (clone 2~ 861X4).

~lG. 3: Rat insulin 1 gene ~nockout strate~eY. Map of wild-typc RIN insulin I (RINS-l) allde, Ycctor for repl~ , and dislu~ d RINS-I allele. I~es~ir~ion cnz~luc sites are shown. Capital bold letters indicate sites ~atl~,d~-ced by the repl~c~ vector and lower case letters indica~e sites in the .,ndGg~ ous gene. b = ~ ; bg = Bgm; N = Nod; P =

WO 97t26321 PCT/US97100761 PacI; s = SpeI; x, X = X}toI. The coding region for RINS-l gene is in~ir~te~l by the rect~n~le with an arrow showing the direction of Ll~ ~.q.~ ion. The h~t.~hç~l rect~ngle in~ tes the seq~lenre used as a probe in genomic Sol~th~rnc The arrows, 1 and 2, show the loc~tio~ of the primers used to amplify genornic DNA specifically l~co.l.~i..ed at the RINS- 1 gene.

~lG- 4A: Insulin content in en~,j..cc,l~d cell lines. ~Illlll"~OlGa~ ivc insulin was ~let(~ ed from acid extr~cts prepared from the following cell lines: RIN 1046-38, R5C.I-17, RSC.I-17 chronically treated with 1.0 mM butyrate, and 1113Eg. Values are reported as ~lg of insu}in per million cells.

~IG. 4B: Basal and stimulated insulin secre~on from cell lines en~ e~ ,d to ~roduce human insulin. Se~ ted i~ o~eactive insulin was ~te....;..~tl from the following cell lines: R~ 1046-38, R5C.I 17, R5C.I-17 chronically treated with 1.0 mM ~ul-ly~d~e, and llJ3E9. Basal samples are from a one hour inrub~tion in med}a lacking glucose and CQ.~ g 100 ~lM ~ 7s~cide Stim~ t~l s~ s are from cells incubated for one hour in media co.~ S mM ~lurose, 100 )uM c~lua~ ol, 100 ~lM 113MX and amino acids.
Values are l~.~d as ng of insulin per million cells per hour.

FIG. ~A, ~IG. 5B, ~IG. 5C: Human proinsulin is errlc.ellt~l~ Plocessed to matureinsulin. T.. -n,~a~ , insulin was d~P~ rci from HPLC ~actionated acid/ethanol prepared from R~ 1046-38 (FIG. SA), RSC.I-17 (FIG. SB) and EPll/3E9 (FIG.
~C~. Arrows in~ir3te positi~n~ where the following standards elute: mature rat-and human insulin (RI and HI), rat and human proinclllin (RPI and HPl), and rat and human El~uceCci~g ;.~ p~ res des-31.32- and des-64,65-split proinc~llin (R 3132, R 6465, H
3132~ and H 6465).

FIG. 6A and FIG. 6B: Blood ~lucose levels of nude rats injected with hurnan insulin-p.~ cells. Nude rats were illjcct~d with either 3 million R5C.I-17 cells (NR1~, FIG. 6A) or EP1113E~ cells (NR21-24, ~;IG. 6B) on day 0. NR5 is an u~ d control W O 97/26321 PCTrUS97/00761 animal. Blood glucose was ~,t~ d on the in~ tçtl days. NRl, NR2 and NR23 died prematurely from severe hy~,~,lycemia.

~IG- 7: Insulin ll~f.55~pe analysis from tumors exPlanted from nude rats iniected with RSC.I-17 cells (see NR1-4. FIG. 6). Primer extension analysis of endog.,n~us rat insulin produces a 91 base ~.xtrn-~ed product (lower band) while the human insulin tr~ncgen~
produces a 101 base e~ten-led product (uppcr band~. Analysis of in vitro ..~ -t~ A RIN
1046-38 is shown in the first lane and in vitro l"~ lr,d RSC.I-17 is shown in the second and last lanes. The day of tumor explant is inrlic~t~d for each in vivo s~mrle.
FIG. 8: In vivo potency of ~nrh~re~d RIN cell lines. The in vitro stim~ te~ insulin secretion values of RIN 1046-38, R5C.I-17 and EPl 1/3E9 (see I;IG. 4B) are co.l.l.~ed to the çxrl~nt~l turnor mass at initial onset of hypoglycernia in nude rats (see FIG. 6).
Individua} twnor masses are indicated.
~IG. 9: Gene L~ ,ssion of manv en~lo~r-T~us ~enes is stable in vitro versus in vivo with the noted exceptilon of GLUT-2. Northcrn analysis of RNA from in vitro ~ A cellsversus day 25 in vivo tumors (R5C.I-17 cells). Signals on NUlLL~,~11S are running at cor,rect sizes relative to pllb!t~hed m~es~s- islet GK - 2.8 kB (Hughes et al., 1991);
GAPDH - 1.3 kB (Fort et al., 1985); amylin - 0.9 kB (Leffert et al., 1989); IPF1 - 1.4 ~B
(Leonard et al., 1993 and Miller et al., 1994); Sulfonylurea reccptor - 5.1 kB (Aguilar-Bryan et al., 1995); EKl - 3.7 kB (Schwab and Wilson, 1989); GLUT-2 - 2.6 kB
~horens, et al., 1988). human insulin t~ P - 0.7 kB (this study); and Neo tr~ns~n~
1.6 kB (this study).
EIG. 10A: GLUT-2 tr~nc~ i,ssiop as driven by the C~V promoter is stable in vitro and in vivo. Northem analysis of GLUT-2 ll~S~ -r eApl~,ssion of a cell line eA~lcssi~g high levels of GLUT-2 (49/206) is m~ t~ d in vivo following a 16 or 34 day passage of the incl~linom~ in a nude rat model.

W O 97t26321 PCT~US97/00761 ~lG. lOB: Low level of endo~ ol~s GLUT-2 ex~ression seen in the parcntal RIN cells t~ P~ in vitro (Lane ]. Panels A and B) is 1os2 followin~ a 24 day v&SS~t of thecells in vivo. Thc message i or GAPDH serves as a loading control.

~lG. 11: Inclcased insulin content resu1tinp from expression p1~cmi~ls containin,~
internal ribosol-~ entn~ sites (IRES). Tmmlmo .~,a.,live insulin was det~ ined from acid/cthanol l.~lr~ from 29 in~e~ ,nt G418 lesis~.t ctones (EP18/3 clones) g~neptf~d from pCMV8/INS/IRES/NEO. Values are reported as a percentage of the insulin content in R5C.I- 17 cells.
~IG. 12A: Hi~herhuman ~msulin-~ro~ cin~clones ~ t,dbvi~ldtiV~ en.~ f~f hl~ of RIN clones witb IRES-co--t~n;n~ insulin e~"cssion pl~mi-lc. Northern analysis ofEP1813El (FIG. I l), a clone eA~lessing a human insulin/IRES/NEO tr~n~ne (first lane) and clones of EP18/3E1 e~ples~ g a second transgene ~ ~co~;~.g human insulinJIRES/Pu.u~lycl.~ (EP111/205, 206, 227, and 230). The n~llly~in col~lA;I~;.. g m~Ss~e iS 1.9 kB while the puromycin co~ ;n~ message is 1.7 kB. l~lPsca~s were ~ietecte(l with a probe specif;c for human insulin.

~IG. 12B: In,.~ase in insulin content followin~ iterative çnr;D~ of RIN clones.
Insulin content was ~lct~ ~1 from acid/cthanol e~l.~ls of 1813E1 cells and 5 clones derived from 1813E1 expressing a second human insulin transgene (EP111/205, 206, 220, 228 and 230). Cell counts were ~t~.~.ined as values are ~~ d as ng insulin per million cells.

EIG. 13: Northçrn blot analysis of p.~ otcl- activity in stably 4~.,~r~.l~d ~IN lines.
Dirr.,.~.~t promoters were driving eAi)lession of the cc.lll,llon tra~sgene, INS/IRES/NEO, were co~u~;~d. For RlP/RIPi, the 5' gencric intron from INS/IRES/NEO was reFI~edwith the rat insulin 1 gene intron (RIPi). All lanes co.~lAii~d 10 Illicrogl~l s of total cellular RNA. The lane labele~ RIN38 c~ ;nc RNA from untra~sr~ted cclls. The lane W O97126321 PCT~US97100761 labeled PC (PolyClone) contains RNA from a pool of RIN38 clones transfected withp~'Y~3lRIp8/INS/~REs/NEo .

~IG. 14: Human ~rowth hormone production in RIN cells Secreted growth hormor was determinPd from six i~d~f ~ nt RlN clones. Conditioned media s~mples were collected from each following a one hour i~ ion in mc~dia lacking glucose and colltAi~ .g 100 ~M fli~7nyiAe (Basal/l~r), a one hour i~ ibAl;on in media co~ ;ng 5 mM gll~cose, 100 ~lM c~l,&chol, 100 IlM IBMX and amino acids (Stim~ te~1/hr), and a 24 hr collection in standard tissue culture media cont~ g 1 I mM glucose and 5% fetal calf serum. Cell counts wcre ~lete .. il-rcl as described and values are reported as ,ug gro~,vth hoImone per million cells.

EIG. 15A: Co~,A,~le,s:"on of PAM and amvlin in cell lines. Endogenous levels of c~lcssion of PAM and amylin in a series of cell lines was ~f~ by Northern analysis. Cell lines e~t~min~ were RIN 1046-38, AtT-20, RIN 1027-B2 and RIN 1046-44 (Phillipe et al., 1987), EP18/3G8 and EP53/114 (this study). Pam mPss~ge nuns at 3.5 to 4.0 kB (Stoffers et al.. 1989) while amylin message is 0.9 kB (Leffert et al., 1989) ~IG. lSB: Northem analvsis of R~ 1046-38 cells stablv transfected with an amvline~yl~,SSiOn pl~cmi~ t~,s hi~h }evel e~ ;ssion of tl~e ~ s~cnc. Amylin is c~l".,ssed as a amylinl~ES/NEO bic,~l,ollic mess~e of 2.1 kB in the polyclone.
OI~Se~lCSSiOU of annylin is present in ~he polyclonc as well as hUI~ 1046-38 andR5C-I. 17.

FIG. 16: Insulin P~ otel Factor 1 (IPF-1) tr~ ~ eA~"~ssion in RIN cells. Levels of stably~ c~d IPF-l mRNA e~.cssed in RIN 38 polyclones and m-.noclona} cel lines were ~ ;r~rd by NGILI1C.11 blot analysis. All lanes co,.1; ;~ d 10 llg of tota~
cellular RNA. The lane l~ ed RIN 38 co~ in~ RNA from lmt~ rc~ted cells. The lane labeled INSl conatins RNA from another untr~ncfeet~l stable b cell line called INSl. IPF-1 tr~ncgPn~ rnRNAs is .lenote-d by IPF-ltIREStNEO. Also shown are levels of endogenous IPF- 1 in diffi rent RIN lines.

~IG. 17: Stim~ .cl Secretio4 of Reco~mhins~nt Rat A~nylin from F,n~ j~.f eled RIN Lines.
S ~mml-nb~ c~Clive amylin species were d~,t~ .rd for basal and st~ S~tecl media sSlmrlf s from parental RIN 38- 1046 cells and five r~o~ Slnt rat amylin pro~lf in~ RIN clone.

~IG. 18: Human Proinsulin is EfficientlY Proteolvtically Pl~cesscd in the Rat T~nljnoms Tmm-lnoreacti~e insulin was dcte~ ed from HPLC frr~tion~tPfl acid extracts ~l~pa,.,d from E~IN 1046-38 cells (FIG. 18A) and BG18t3El (FIG. 18B). Arrows in~lir~tç positions where standards elute: mature rat and human insulin (RI and HI), rat and human proinr~llin (RPI and HPI), and rat and human proce.csi~ ;,lt ~."~-~Lates des-31~32- and des-64,65-split proin~lllin (R3132, R 6465, H 3132, aI~d H6465).

EIG. 19: Secreted and ~haLl~ble Amvlin Species from F.n~i~.rl.~,d RIN Lines. A
stimulated one hour sccretion sample from either R~ 1046-38 ~parcntal), EP97/134 (rat amylin), or EP182/13 (human amylin) is c~ ~cd to the e~h~ xl intra~rell~ ms~tPri~l from the same cell lines. Ccll e~ s were prepared following either the S cetic acid or the triflouroacetic acid/3~etQr~itrile (TFA/ACN) extraction protocols ~les~rihed in ~riAlC and Methods. T.,.. ~o~ ive amylin is reported in ng/million cells/hour for the se~ ed s~n~rles and ng/million cells for the e~bdClCd m~t-risl ~IG. 20A and ~lG. 20B: Rat and H~2rnan Amylin are F~Fici~ntly Proteolyticallv P~oc~ssc~l in the Rat In~ulin~A PIG.20A. Tm mlmoreactive rat amylin was dete ...;- ~.d from HPLC f.,cl.QI.Al~ ar~id e~ plc~ ,d from rat arnylin producing BG 183/20 cells with co.~c.;~on to fractionated rat amylin standard. Shown is a plot of the W
absorbance versus time in ...;.\ ~I~ s and the amylin immlmolc&Lilivily data in the indicated r.~ c overlaid on the plot. FIG. 20B. Llll~lul~ ive arny}in was ~e~ .cd from HPLC rl~ ;O,~ TFA/A('N e~ pl~p~d from either hurn n amylin p~ Jc;l-g BG

182/12 cells or parental RIN 38 cells with comp~ orl to fractionated human and rat amylin standards.

~IG. 21A and ~IG. 21B: TmmnlloplG~iuil~tion of RIN Produced Human Amvlin S Species. F~G. 21A. ~nmunopfcc;~ ,5 were prepared from metabolically labeled eA~ s ~,pared from parental RIN 1046-38 cel}s or human amylin pro~hlring BG
182112 cells. Extracts were p,~ d using a NP~0 Iysis buffer (1) or RIPA Iysis buffer (2) as ~esçribe~l FIG. 21B. Pulse/ chase experiment d~ .L~S that BG 182/12 human amylin procluoing cells secrete .~ ole~;livc amylin species. Cclls were either labeled for four hours or labeled for three hours followed by a one hour chase by il.. V~ in the stin~ ted secretion buffer described above. The stim~ te~ media was collected and cell e~ s prepared using the NP40 Iysis buffer followed by imn~nno~,c;r.il~1iorl from the three samples. MW ~ la, ls in kD are i~ r~te(l and arrows inrlir~te the relative llugration of ~y~t~lcLiC human and rat ~mir~ rl amylin.~5 EIG. 22: Cc~ ,ssion of ~mylin Tr~ qr ~- with Endo~enous Insulin. PAM and Amvlin in F.nr,;~..erl RIN Ce~l Lines. Northern analysis of RNA prepared from four recon.h,;..~-t rat amylin ~ RIN clones and RIN 1046-38 cells. Parallel filters were probed with rat amylin, PAM or insulin specific probes.
~IG. 23: Tn~luctj-)n of Hy~ cG~llia in N~de Rats Bearin~ Amv~ o~ Tumors.
Daily blood glucose levels are plotted for an uninjected control and five nude rats injectçd with a rat amylin ~ RIN clone (BO97/134), five animals i~je~ led with a human amylin profi~ing RIN clone (BG182112) and three ~nim~ ljech ~1 with RIN 1046-38 cells.

FIG. 24: Amvlin Tr~t~s~.ene I;~yl~ssiol1 is Stable in vivo. Northern analysis co...l-A ;..
RNA i~ t~ cl from cells ~ in tissue culture (in vitro) or tumors following the indicated time of in vivo passage (days) using a rat arnylin S~;re probe ~:le~.~o~s~ .,s no .l;rf ~G in eithcr the low lewls of endG~nou~ amylin eAp~ssion in either RIN 1046-CA 0224643l l998-08-l2 38 or BG97/134 cells or high levels of the rat arnylin transgene iD eng;. ef ~ed BG97/134 cells .

PIG. 25: RIN Cells F.~in~çred for Ar,nylin or Insulin Expres~cion EfficientlY Reduce S Blood Glucose Levels in Streptozotocin-Tn~ eed Diabetic Nude Rats. Daily blood glucose levels are plotted for an Iminj~tecl control and thrce strepto7,0toein (SIZ) ect~ nude r~. Fifteen days following STZ injection, ~nim~15 were i~.jG~ with a rat arnylin proA~ g RIN clone (BG97/134), a human insulin ~Jlu~ ei~g R~ clone (BGl 11/228), or RIN 1046-38 cells.
EIG. 26: Blood Gl-leose ~vera~es of Nud~e Rats Bearin~ F.r~ cl~d Rat Tn.culino~qc~
Average daily blood glucosc levels are plotted for l~l,;.jCet. (~ controls (3 q~nimqlc), and rats injected with an insulin producing RIN clone (EP18/3E1, 4 ~nim,qlc), an amylin pr~lue~g RIN clone (EP97/134, 4 AnimAqlc), an insulin and arnylin l~r~d.~ RlN clone (EPl81159, 4 Anim~lc), and parental RIN 1046-3~ cells (3 qnimqlc).

~lG. 27: Serum Arn~lin ~evcls in Nude R~ats ~e-q~ri~ r~ P..)~d Rat Tn.cl~lin~ c,Average serum amylin lev~,ls from rats d~nbeA in FIG. 26 legend. Amylin values were Q~e~u~,d against a rat arnylin standard as descrikd in Materials and I~e-tho~lc FIG. 28: Serum Insulin Levels in Nude Rats Bearin~e Fn~ineered Rat Tr~clllinomas.
Average serum insulin levels from rats de cribed in FIG. 26 leg~nd. Insulin values were ll,c,z~u,-,d against a human ,aIIylin standard as ~1~seribe~l in ~ tr '~~ c and ~I.~c.

FIG. 29: Sermn A,mvlin~!lnsulin Rati,os in Nude Rats R~Rrin~ F,n~,jnee,~,d Rat ~nc~ll;..o..,Ac. Molar ratios of serum amylin to serum insulin were d~te~uul~cd using values f~om FIG. 27 and FIG. 28.

wo 97n6321 PCT/US97/00761 IV. DETAILED DESCRIPIION OF THE PREFh~Rl~Fn El 1B(~DIMENl'S

Secretory cells, especlally n~ oel~docrine cells, have several endogenous functions that make them uniquely suited for production of a wide range of l~i'u~i~lS, in~ in~ secreoed peptide S h~ s These speci~li7ed ~lnrti~mc include the reg~ ted secl~,to~ way. The regulated se~ ,tu ~ pathway e...bod;es the sec.~o-y granules of neuroendocrine cells which serve as the site of ~ wa~ion and storage of a large class of peptide ho."lol~es with pfof~,u..d biological f~m~ nc Proper biological function of the pepli~s is due both to their secretion in a regulated and titratable Illa~ El as well as a complex set of post-translational motlifi~fio~c res~l1ting in the 10 final biologically active product As a result, these cells can be used in vitro to produce large i1 no! -~c of proteins, in vivo to supply theraF~eutic proteins, or in vivo to ;--------ni7~ hosts, for e~ nrl~, in the pio~ ;on of monoclon~l ~nti~i~c The present invention is dP.Ci~f'~l to take advantage of this s~l~tol~ n~ y for the 15 ~ ose of ~.od~le;l~g hetcrologous proteins. A variety of dirre..,.ll mo-lir.c~ c may be made to incl~zse, the effi~ien~y of the cell, one eY~mple is the blocking of produetion of an endogenous protein in the host cell. This will, in ess ~ e, "rnalce room" for the hcterologous protein and, hence, avoid co...l.~t;lion b~,~., ~n the endG~e~ s and heterologous proteins during synthesis The coll~o~ ts for such a system~ and m~.-horlc of ~ lucing proteins th~.e~ , are set forth in 20 detail below.

~ n a particular çmlDo~ nt, the present invention relates to the p,o~h~ ;Qn of amy}in species in a mammalian host cell system These amylin species possess all the post-translational ~;r~C~ioîtc of natural amylins The av~ hility of such amylins allows for the çluc~ tion of 25 thc nahlral amylin l~c~ , tne role of amylin s~,e~,on and the con ~c~ ces of its absence Further, a ~ nt~orr~ne cell-based system for either in vitro biologically active amylin ;Qll alone or in ,-o "~ tiol with insulin, or for in vivo cell-based delivery of biologically active amylin alone or in col~.h;n~~ion with insulin, would provide an trr~Li~e therapy in the t~nent of ~ het.'.S, h~Gcl~rc~ ia and the ,eslord~,on of ~-cell function f WO 97/26321 PCl'tUS97100761 The present invention describes mrthods and col"~,ositio~c for g~ atillg amylin species from çng.~P,. /~,d eukaryotic cells. ;Fusther, the present invcntion details mçth~3.c of el~lr~ ting the role of amylin secretion and amylin ~ lOl(S). The methods and c(u.,lposilions described in this invention will be useful in the ~ of pathogenic states res.lltin~ from aberrations in 5 amylin D~ hcsis or secrction. The present invention for the first timc ~sr~ es methods of oblail~1g "natural" amylin species that is to say, arnylins that have bcen post-translationally ...o~l;f.~d in ways that result in natural amylin instead of a ~yllllletic amylin D~ ut~.

A. Host Cells Fn~;.. P., .i~g of secretory cells to synthesi7e proteins for either in vilro large scale pro~uction~ or for in vivo cell-based d~ ely, will adY~nt~geously make usc of many attributes of these cells. Regulated secretory cel]s present a natural bioreactor co,.l~inil~e speci~li7~l e~ .,es involved in the ~ocecci~-g and maturation of secreted proteins. Thcse plufecs;ng e~nz~ les include cndop,ot~ases (Steiner et al., 1992) and carboxypepti~ces (Fricker, 1988) for the 15 cleavâge Of p~uho~..nn~.s to hormones and PAM, an e.,~ylllc catalyzing the amidation of a n-,lll~r of peptide ho~..-m.Ps (Eipper et al., 1992a). Similarly, ~ ul~ion and folding of peptide h~..l...~l ~s is perforrned in a controlled, stepwise rnanner with defined ~ ,t~:t inrhlrling pH, c~lri~m and redox states.

Cc~ pr~c~cc;llp [c~uil.,S snffirient levels of the ~roce~ e cl,L~llles as well as ~-rr.- ;~ retçr~ti~)n of thc .~ .g pcptides. In this way, physiological signals leading to the release of the contents of the sP~ r granules e~ s relcase of fillly l~lucc~ssc~l, active ~lUt.,~S. This is illlp~ t for both m~Yim~ v~ c~iotl for in vi~ro ~u~tJoses and for the possible use of cells for in vivo y~O~S.
All cdls secrete proteins through a con~ e, non-regulated sccl~,tul~ p.~ .ay.
A subset of cells are able to secrrte ~rotcins through a speci?li7ed rPg~ ' see~ "y palll~a,~. Proteins d~s in~.~ for secretion by either ".Frh~ m are targeted to the en~o~ ;r re-tirlllllm and pass through the golgi a~p~dlls. ~Q~ ly secreted p,~)t~",s pass directly from the golgi to the plasma lll.,.l,l,l~ in vesicles, fusing and ~lcs~.;ng the contPntc col~cl.t~,lively without the need for extPrr~l stimuli. In cells with a regulated ~alhw~y~ proteins leave the golgi and conre-~l.ate in storage vesicles or secl~t~"y granules. Release of the proteins from se~ ,tol~ gr~nl~lps is re~ tecl r~lu;. ;i.g an e~tPrn~l stimuli. This extPrn~i stimuli, defin~l as a seclet~gogue, can vary dc~..ding on cell type, 5 optimal con~e~ ;on of secretagogue, and d~ll~l~iCS of secretion. Proteins can be stored in secl~,to,y granules in their final p~ocessed form for long periods of time. ~ this way a large intr~ce~ r pool of manlre sccrelul~ product ex~sts which can be released quickly upon se~togogue st-mnl~tio~l A cell speci~lized for secreting proteins via a regulated p~lh~r can also secrete proteins via the con~ e scc[~t~ly p~ y. Many cell types secrete plOt ins by the cQl~c~ e ~ ay with little or no S~,liOll through a regulated pathway. As used herein, "se~ loly cell" defines cells speci~li7p~d for regulated secretion, and eYcl~ ec cells that are not spe~ i~li7P~l for regulated secretion. The regulated secretory p~ hn~ is found in lS sCCI~,toly cell types such as endocrine~ exocrine, neuronal, some gasLlu;l~esl;n~l tract cells and other cells of the diffuse endoc..l.c system.

(i) Glucose Rcs~ sive Cells.
For delivery of some peptidc hol.~.o,.~s or factors, it may be desirable to cause the 20 polypeptide to be released from cells in ~ ce tO ~h~ es in the circ~ tinE glucose C~OI~ n. The most obvious PY-vnr}e of a sc~i~el~n~ cell type that is regl~lAted in this fashion is the ,B-cell of the ~ f ~ ;r islets of Langerhans, which l~lcases insulin in lc~nse to chs-ng.-s in the blood glucose co-~re..l.ttion. r;~g~ J~ ;..g of pli~ cells for ~ h~ n of products other than insulin is not rr,9A~ . Tn~ a p-cfe-,-,d vehicle may be one of the several céll lines 25 derived from islet ~-cells that have e.lltr~,cd over the past two dec9dec. While early lines were derived from rvrliAtion- or virus-in~ ed tumors (Gazdar et al., 1980, Santerre et al., lg81), more recent work has ccn~,~d on the applic-s-~ion of ~ c~e~ technology (Efrat et al., 1988, Miyazaki et al., 1990). A general ~ ach taken with the latter tçchniqne is to express an ol ro~ --r., most often SV40 T-antigen, under control of the insulin promoter in hA~ce~ ~ic 9nim91c, thereby 30 generating ~cell nlmors that can be used for propagating in~lllinnma cell lines (Efrat et al., 1988, ., Miyazaki et al., 1990). While lncllliin-~lnq lines provide an advantage in that they can be grown in esse~ q-lly lmlimited quantity at relatively low cost, most exhibit di~f~ ces in their glucose-stimulated insulin sec,~to.y responsc relative to normal islets. These difr.,lcl~ces can be quite ~r~fouud, such as in the case of R~mSF cells, which were deriYed from a r~ nn-induced S inculinnmq. and which in their c-uTent form are completely lacking in any acute glucose-s~imlllqted insulin secretion response (Halban et al., 1983, Shim-l7~l et al., 1988). RIN 1046-38 cells are also derived from a radiation-in~ e~ inomq- but can be shown to be glucose n~i-/e when studied at low passage ~ he,~ (Clark el al., 1990~. This ~,sl>onse is m~yim~l at sul,phy~iological glucose concen~tions and is lost entirely when these cells are cultured for more than 40 pqcsq~gçs (Clark et Cll., 1990). GLUT-2 and glll~okin~c~ are e~ ,ssed in low p~C~age R~ 1046-38 cells but are gradually Aimil~i.ched with time in culture in s~l,cl..~..y with the loss of glucose-stimulated insulin release (Ferber et al., 1994). Restor~q-tion of GLUT-2 and glucokinase e,~ ;.sion in RIN 1046-38 cells by stable transfection ~ .GS glucose-stim~ qtPd insulin secretion (Ferber ct al., 1994), and the use of these genes as a general tool for l~n~ rA - ;ilg lS of glucose sensing has been rlesc,il~fl in a previously issued patent (Ne~.~al~, US Patent S,427,940). RIN 1046-38 cells tr-q-r~ef~cteA with the GLUT-2 gene alone are mqximq~ly glucose ,,i,~nsive at low co~ tions of the sugar (~ Ai~ tely S0 ~uM), but the threshold for c can be shifted by ~,ei~,~ u~ g the cells with 2-deo~y~ ose. which when coh~,.t~,d to 2-deo~yglucose-~1)hos~ r inside the cell serves as an inhibitor of low Km hexokin-q-ce, but not 20 glucose activity (Ferber et af., 1 994)l.

Recently, Asafari e~ aL have l~u~t~,d on the i~ol-q-tio~ of a new incn~innmq cell line called INS-l that retains many of the c~ s of the A~rfe.c,.liated ,B~ell, most notably a relatively high insulin content and a ~l~)cQsc-stim-llqtecl insulin secretion rcspo~-~e that occurs 25 over the l~h~D,clGgical range (Asafari et al., 1992). This line was isolated by ~lupaga~ g cells fieshly Ai~l".~rl from an X-ray in~ur~ed im~llin--m~ tumor in media cc~ ;,.g 2-mercaptoethanol. ConsiDtent with l~e fincling of physiological eh~cose ~ onsi~ ess, a recent report ;~A~ es that INS-1 cells e~press GLUT-2 and glucokinase as their ~dC~ -t glucose tra~ and glucose ~ .hn,ylating e.~ e, r~pe~ ly (Maric et al., 1993). INS-1 cells30 grow very slowly and require 2-n~.~ lh~ ~ol. It remains to be detr ~ d whether glucose WO 97/26321 PCT/US971~0761 responsiveness and expression of GLUT-2 and glllf,Qkin~ce are retained with prolonged cult lnn~
of these cells.

Cell lines derived by lr-~c~eli5 eA~ ,ssion of T-antigen in ~-cells (generally termed ,B TC
5 cells) also exhibit variable phenotypes (Efrat et al., 1988, Miyazaki et al., 1990, Whitesell et al., 1991 and Efrat et al., 1993). Some lines have little glucose-stim~ te~l insulin release or exhibit n~ OIlSeS at subphysiological glucose c~..re-.t-~tions (Efrat et al., 1988, Miyazaki et al., 1990, Whitesell et al., 1991), while others respond to ~h~cose co..~ ions over the physiological range (Miyazaki ct al., I990 and Efrat et al., 1993). It ~ that the near-normal 10 ~,s~onsi~reness of the latter cell lines is not pcrn qn~nt, since further time in culture results in a shift in glucose dose l~syol~ce such that the cells secrete insulin at ~ubl)hyaiological glucose co~e--l.dtions (Efrat et af., 1993). ln some cases, these ch~ll~s have been col.~laled with changes in the eA~i~,s~ion of ~;lue~ose transporters and glucGsc-~h~ hr~l~rlating ell~ylllcs.
Mi~i et al. i~ol~tPd two classes of clones from transgenic ~nim~lc eA},l.,ssing an insulin 1~; ~c",lote./T-antigen COnal~uCl. (~h~cose-u~ ,s~n~,ve }ines such as M~-7 were found to express GLUT-l rather than GLUT-2 as their major glucose llanspolt~l isofollll, while MrN-6 cells were found to express GLUT-2 and to exhibit normal glucose-stimulated insulin secretion (Miyazaki e~ al., 1990). More recently, Efrat alld co~ci~ a ~ ..nr~l.ated that their cell line bTC-6, which eYhihitc a glucose-s~ t~d insulin se~,.e~ion l~ pQ~e that I~SC ~bl~ s that of the islet in 20 .. ~,~;~v~ and col-te,~ dtion depen~nre~ e~ ssed GLUT-2 and col~l;.;t-fd a glucokinsce:h~,Yokinase activity ratio sirnilar to that of the normal islet (Efrat et al., 1993). With time in cul~re, E;lucGse-stimulated insulin rclease bc-ns~ ..,tl at low, aul~ll~iological glucose co~ce~ ;onc~ GLUT-2 e~ ,ssion did not change with time in culture, and glucolcinase activity actually incl~ased slightly, but the major change was a large (approximabely 25 6-fold) illcl~aSe in he~cokin~e exl,lcssion (Efrat et aL, 1993). Furlh~ , ov."~ ssion of h~rl~insce I, but not GLUT-l, in well-dirr~,.e~ s-t~A M~-6 cells results in both incl.,ased glucose m~ ic~n and insulin release at ~ulJ~L~iological gh~coce co~-r<-~ tions. Similar resultc _ave becn obtained upon ov_c,A~essio~. of hexokinase I in nom~al rat islets (Becker et al., 1994b). These resultc are all consicte~t with the observations of Ferber, et al. ~escribed above in showing that a high hexokin~ce:glllrokin~ce ratio will cause insulin-secreting cells to respond to glucose con~e~.l.ations less than those required to stimlll~e the norrnal ~-cell.

(ii) Non-~lucose Responsive Cells An ~l~c.~ ive to insulinoma cell lines are non-islet cell lines of neuroendocrine origin that are erginPP~red for insulin e~r~,SSiOn. The foremost example of this is the AtT-20 cell, which is derived from ACTH secreting cells of the anterior piluikuy. A decade ago, Moore et al.
fle .~ cl~ated that stable transfection of AtT-20 cells with a construct in which a viral plulllûtcr is used to direct e~ ession of the human proinculin cDNA resulted in cell lines that secreted the 10 cvllc~,lly processed and mature insulin pol-ypeptide (Moore e~ al., 1983). Insulin secretion from such lines (generally tcrmed AtT-20ins) can be stim~ t~ by agents such as forskolin or dibutyryl cAMP, with the major secreted product in the form of mature insulin. This s?l~gests that these cells contain a regulated secretory yath-. ay- that is similar to that operative in the islet ~-cell (Moore et al., 19~3, Gross et al., 1989). More recently, it has b~ulllc clear that the 15 r~A~o~idases that process proinsulin to insulin in the islet ,B-cell, termed PC2 and PC3, are also e~"es~ed in AtT-20ins cells (Sm~eL~-nc and Steiner., 1990. Hakes etal., 1991). AtT-20ins cells do not respond to glucose ac a secretagogue (Hughes et al., 1991). Int, cslingly, AtT-20 cells express the glucûkinase gene (Hughes et al., 1991, Liang et al., 1991) and at least in some lines, low levels of gll~cokin~ce activity (Hughes et al., 1991 and 1992, Quaade et al., 1991), but are 20 c~ k~ y lacking in GLUT-2 ~A~l~ssion (Hughes et al., 1991 and 1992). Stable transfection of these cells with GLUT-2, but not the related transporter GLUT-l, confers gll~cosc stimulated insulin secl~lion, albeit with m~irn~l lcS~)On~ CSS at s.lb~hy~iological glucoce levds, probably because of a non-optimal h~xokin~ glYcQkin~ce ratio (Hughes et al., 1992, 1993).

The studies with AtT-20ins cells are h~li)oll~t because they d~ .ale that r,e.u~nA~ p cell lines that normally lack glucose-stimulated peptide release may be A for this r--rl;O~ Othcr cell lines that are characterized as l~,~V~ ~llo~rinP., but laclcing in e,..dvg~.lo~ls glucose ~cs~ol~se include PC12, a l~vnal cell line (ATCC CRL 1721) and GH3, an anterior ~ilui~ cell linc that se. .~t~,s growth huln.ol~e (ATCC CCL82.1). It is not 30 possible to ~te-. ;..r ~h~ller such cell lines will gain glucose ~s~onsi~eness by ~

WO 97n6321 PCT/US97100761 sirnilar to that described for the AtT-20ins cell system without pe-rull~ling the c~ ,;"-~n~c However, these lines do exhibit other ylo~ ies illlyOl~t for this invention such as a regulated scw~,to.~r pathway, expression of endopeptidases required for procescing of prohorrnnnes to their mature hormone products, and post-translational moAifi~ ~tion e&zy~l,es. In sum, all S ne~ Aocrine cell lines are useful for the essenti~l aspect of this invention, which is the pr~ u~1;on of heterologous products in a cell line in which the natural product (insulin, growth h.. n.~f-., ACI~I, e-c.) has h-een e~ r~~-~ Some or all of these lines will also be useful for glucose-regulated product delivery, using the rn.~hoAc. d~s~ribed in U.S. Patent 5,427,940 to gen~ le such ,eiyollsi~ne (iii) Methods for Blockin~ E~n,do~enous Protein Production Blocking eA~ ion of an endogc,nous gene product may be an illlyOl~lt mo~1ifir~io~ of host cdls acco,Lng to the prescnt invention. The targeted endogenous gene erl~odes a protein norrn~ly sccl~t~,d by the host cell. Rloc~ing eA~ ssion of this endogenous gene y~ cl, while 15 en~ re/ ;i.g high level c~y~ssion of genes of intetest, ~ sellls a unique way of ~IPciening cells for protein pro~ln~ n Cells 6,ene.~d by this two-step process express heterologous proteins, in~lllAine a variety of natural or en~in~oered proteins (fusions, chim~r~c~ protein r~ ,t-, etc.). Cell lines 20 developed in this way are uni~uely suited for in vivo cell-based delivery or in vitro large-scale yr~ n of defined peptide ho,~llol~es wi~ little or no cO~-tY....;~ in~ or ul~ t~,d endogenous protein ~ n Five basic appl~aches are c(.-t~ ted for blocking of ~Ayres~ion of an endogenous25 gene in host cells. Fir t, consLIucts are desigr~d to h- mologously leCC)~.i.;~C into particular ~ O~,Cr~O~C gene loci, l~nrl.";n~ the endogenous gene nonflmrti~n~l Second, constructs are designed to randomly integrate throll~h~u~ the gl~nnrn~ res~lting in loss of eA~I~,ssion of the e-~o~ o~IS gene. Third, constructs are de~i~Pd to introduce nucleic acids c~ ler~nl;-~r to a target c~ ,v-~ s gene. Expression of RNAs cc.ll~ Jo~ g to these comple.,lellt~ nucleic 30 acids will hlt~,fe.~ with the IlAnC~ ion and/or tr~nC~ n of the target se~ en~e~ Fourth, WO 97/26321 PCTrUS97/00761 cons~u~ are deci~rd to introduce nucleic acids encoding ribozymes - RNA-cleaving ell2y-~es - that will specifir~lly cleave a targct mRNA cc..~onding to the endogenous gene. Fifth, endogenous gene can be rendered dysfunctional by &erlornic site directed mllt~g,on~cic S Antisense. A~ cn.ce methodology ta~es advantage of the fact that nucleic acids tend to pair with "co...l le-"~ " seqllenre~s By complcmentary, it is meant that polynucleotides are those which are capable of base-pairing acco.ding to the standard Watson-Crick complc...~ y rules. That is, the larger purines will base pair with the smaller pyrirni-lin-s to form col~lb;~ ;QI c of gu~ninr paired with cytosine (G:C) and ~f~.onine paired with either thymine 10 (A:T) in the case of DNA, or ~eninp paired with uracil (A:U) in the casc of RNA. Inclusion of less commnn bases such as inosin~ 5-methylcytosine, 6-methyl~ onin~ hypox~nthine and othcrs in hyhri~li7.in~ se~ .res does not i~.~f.,.G with pairing.

Targeting double-stranded (ds) DNA with polynucleotides leads to triple-helix form~tirJn;
15 t~5_Lil~g RNA will lead to douUe-helix form~~ion ~nticçnce po~ nrlPA~ $~ when ill~lUcc~
into a target cell, spe~ifir~lly bind to their target pc,~ le,ot;~le and int~,fe.~ with !-~'S~ lion, RNA pl~ccsc;..~ ,s~ll, tr~ncl~tiorl and/or stability. Antic~r~ce E~NA cor~ ,ct~, or DNA
c ~ro~ e such ~ e RNA's, may be employed to inhibit gene !~ cc-,~ ol or translation or both within a host cell, either in vitro or in vivo, such as within a host animal, inrlu-lin~ a human 20 subject.

~ r.t;~..,ce consl~u~ may be d~si~p~l to bind to the plollloecl and other control regions, e~cons, introns or even exon-intron bo~ -;es of a gcne. It is cc~ -pl~ted that the most er~ antisense con,l.u~l~ will include regions complementary to intron/exon splice j~l~..;!;o--c 25 Thus, it is p~ that a pl~f~,lGd çmhorlimPnt inr~ s an antisense COllS~luCt with complementarity to regions within 50-200 bases of an intron-exon splice junction. It has been observed that some exon s~ res can be inrhld~A in the construct wilhout s~ ;o~cl~ ~fr~;..g the target selectivity the~eof. The ~molmt of exonic m~tçri~l inc~ ed will vary ~Fp~n~l;"~ on the particular e~on and int~ SGyu~res used. One can rea~ily test wlh.lllcr too much exon DNA is 30 ;.~ ~ simply by testing the corlsllu-;~ in vitro to d~ r whether normal cellular function WO 97t26321 PCT/US97100761 is affected or whether the expression of related genes having co~ ..f~t;~ sequences is ct~A

As stated above, "complem~ ry" or "~nticen.ce" means pol,vnucleotide sequences that 5 are ~lL.sl ;~nti~lly complc.~ lt~y over their entire length and have very few base ~ tohP s For çx~rle. seqnenre!s of fifteen bases in length may be termed comple...~ y when they have complc~ tzu~ nucleotides at lhil~n or fuulL~D pocitinnc. Naturally, sequences which are completely complt~ .L~y will be sç~JI.~ n~es which are entirely co~ ..rl.l; y throughout their entire length and have no base mi.~m~trh~s. Other seq-lcnccs with lower degrees of homology 10 also are co~t~ .lated. For ex~mple, an ~nti~n~e construct which has lirr~ited regions of high hornology~ but also co~t~inC a non-homologous region (e.g., ribozyme) could be cl~si~d These molec~ c~ though having less than 50% homolog~y, would bind to target sG~ cl-ces under a~,o~"ate col~-litinr~.c It may be advantageous to c.,.. bi~ portions of ~nO~ DNA with cDNA or synthetic seq ~ CC~ to gc~ te spccific co..~ . For ~rle, where an intron is desired }n the Illtim~te consllu~;l, a E~cnt!mi~ cione will nced to be used. The cDNA or a synth~i7ed polynucleotide may provide more conv~,ni~"~l r~,sl~ ;or sites for the .~ g portion of the cc~n~ cl and, thc.c,fo~c, would be used for the rest of the se.l~f ~-~e Ribozymes. .Altho~l~h plot~ s traditionally have been used for catalysis of nucleic acids, ~oll-~l class of mac~ ulps has e.-~c.~cd as uscful in this e,ndc~or. Ribo~ylllcs arc RNA-protcin complexes that cleavc nucleic acids in a sitc-spe~ifir fashion. Ribozymes have specifir catalytic ~lom~in.~ that possess endonuclease activity (Kim and Cech, 1987; (-~rl~h et al.~ 1987;
25 Forster and Symons, 1987). Por example, a large nu~ er of ~ o~yl~-es accclerate IJho~hocst~
transfer rearti- n~ with a high dcgree of s~c~-if. -ily, often cleaving only one of several ~h~sl~hoc~ in an oii~o~ el~otide ~ubs~ Ccch e~al.~ 1981; Michel and West_of, 1990;
Rcinhnld-Hurek and Shub, 1992). This spc~ificity h~s been ~ ed to the ~t~luh~"llcnt that the substrate bind via spe~ifi~ base-pairing interactions to the int~ l guide se~u~,nce ("IGS") of the lib~L~ C priorto ChP-~ lr~ ~;O~

W O 97126321 PCTrUS97/00761 Ribozyrne catalysis has primarily been observed as part of seql~enre-specific cleavage/ligation re~ctionc involving nucleic acids (Joyce, 1989; Cook et al., 1981). For e~mple, U.S. Patent No. 5,354,855 reports that certain ribozymes can act as enrlo~clç~es with S a seyu~ ce specifirity greater than that of known ribon~cle?ees and a~rc ~-hing that of the DNA
restliçtion ell~yllles. Thus, seq~.~ncc-specific ribo~yllle-mç~ te(l inhibition of gene expression may be particularly suited to th.,~a~tic applic~tinn.c (Scanlon e~ al., 1991; Sarver eJ al., 1990;
Sioud et al., 1992). Recently, it was reported that .ibo~yllles elicited gcnetic ch~E~,s in some cells lines to which they were applied; the altered genes included the oncogenes H-ras, c-fos and 10 genes of HIV. Most of this wor:k involved the mo~lifi~ation of a target rnRNA, based on a sperifi~ mutant codon that is cleaved by a specific ribo;Gylllc.

Homologous Recombination. Another approach for blocking of endogenous protein plo~lncl;on involves the use of homologous lecGlllbilla~ion. ~lomt)logous recombination relies, 15 like ~ icçn~e~ on the tendcncy of nucleic acids to base pair with colll~le~ se.lu~ ~cs. In this instance, the base pairing serves to f~cilit~tç the inter.~lion of two S-~p.~dte nucleic . cid cl~les so that strand bl.i~'~~e and repair can take place. In other words, the "homologous"
aspect of the m~.th~ relies on s~.l~nre hom~l~gy to bring two c~ ,lc~ sey~ es into close proxi-m-ity~ while the "recombination" aspect provides for one cc,.~.ple....~ y seyu~ ce to 20 replace the other by virtue of the brealcing of certain bonds and the form~tio~ of others.

Put into pracri~e hnmologous l.~co .~k; ,ation is used as follows. First, a target gene is s~ te~l within the host cell. .Sequ ~ec hr~molo~ous to the target gene are then in'']-.ld~ in a genetic col~..t~u;l, along with some m~tstion that will render the target gene b~a~Ll~e (stop codon, 25 int~.lu~ion, etc.). l~e homologous se~uen~c flsnking the hlacLivdLibg mutation are said to "flank" the mutation. Flsnking, in this c~- ext simply means that target homologous seq~lçn~es are located botb upstream (5') and d~ lSI~ l (3') of the m~ on These sequ~,~.ces should coll~ond to some se~lu~ es U~x~ l and duu~sL~ of the target gene. The cûllsllucl is then intlod~ed into the cell, thus p~ g l~co~bination ~ .e~,n the cellular s~.~e..~es and 30 the co~sllu~

W 097/26321 PCTnUS97/00761 As a practical matter. the genetic construct will normally act as far more than a vehicle to h~t. ~lu~)l the gene. For e~m~le, it is i~ t to be able to select for l~coll.bil,ants and, the~c,folt, it is co~ llon to include within the construct a select~l-le marker gene. This gene 5 permits se!ectiQn of cells that have integrated the construct into their genomic DNA by co.~r~,.;,~g rçci~t~n~e to various biostatic and biocidal drugs. ~ addition, a heterologous gene that is to be e~iessed in the cell also may adv~ ously be inrludcrl within the Cons~luC~. The n~.,..~. .-t might be as follows:

... vectorS'-fl~nking seque-nce~hetelologous gene- selectable marker gene-fl~nkin~
seyu~nce-3 '-vector. . .

Thus, using this kind of construct, it is possible, in a single recombinatorial event, to (i) "knock out" an endogenous gene, (ii) provide a select~'~le marker for identifying such an event and (iii) 15 intrvduce a heterologous gene for e~pl~,s~ioll.

Another r~r~ t of the homologous ~col..hil-~tion approach involves the use of a "ne~;~tive" scl~ lc marker. This marker, unlike the selcct~ble marker, causes death of cells which express the marker. Thus, it is used to identify undesirable l~coll,billation events. When 20 se~lrine to select homologous ,eco...~ using a select~ole marker, it is tliffir~llt in the initial SCl~nlllg step to identify proper homnlogous l~col.lbin~.t~ from ~co.,.l~ ant.c gelle.illed from random, nor. se~ re spcr;r.~ cvents. These fccu~ n~c also may contain the select~
marker gene and may e~press the he~rologous protein of interest, but will, in all lik~lihood, not have the desired "knock. out" ~h~"o~ Je. By ~ , chi~g a negative selectable marker to the 25 co~ , but outside of the fl~nking regions, one can select against many random Iccolllbination events that will incol~,oldte the nc,ga~i~, sele~ le marker. Homologous recombination should not ~l~uce the ~eg~i~_ s~lPrt~llle m~rker, as it is outside of the fl~rlking sequences.

W O97126321 PCTrUS97/00761 Ln a particular aspect of this embollim~nt~ the negative select~ le maker is GLUT-2. It is also col-re.l.pl~r~t that GLUT-5 would function in a similar manner to GLUT-2. Th~refo,c, the selection protocols described are inlrnrlPd to refer to the use of both GLUT-2 and GLUT-5.

In a first embo-lim~t a target gene within a GLUT-2- host cell is selected as the location into which a srlecte~l gene is to be lr~ls~l~,d. Sequences homologous to the target gene are inrh~led in the expression vector, and the selecre~l gene is insertcd into the vector such that target gene homologous se.lu~,~ces are int~lu~ted by thc selecte~l gene or, put another way, such the target gene homologous sequences "flank" the s~olecte~ gene. In preferred embo-lim-ntc, a drug selectable marker gene also is inserted into the target gene homologous se4~ es. Given this po,c.cihili~y, it should be a~ t that the term "flank" is used broadly herein, namely, as sr ihing target homologous seq-~e~ces that are both uysl~ (5') and downstream (3') of the selerted gene andlor the drug s-,kc~ le marker gene. In effect, the fl~nking sey~ es need not directly abut the genes they "flank."
The construct for use in this ~Yrnho-iim~nt is further ch~cte~ ;~e~l as having a functional GLUT-2 gene ~tt: ~hr~i thereto. Thus, one l~ossible arrangement of se lu~,~ces would be:

... S'-GLUT-2-fl~nking target s~4u~nces-sck~e~ g.,~e~ g-sclccl- ble marker gene-fl~n~ing target seqU~nres-3~

Of course, the GLUT-2 could come at the 3'-end of the CO~ U~:I and the select~rl gene and drug-scle~-l .kle rnarker genes could eYrh~nge ~oc;l ;OnC

Application of a drug to such cells will perm~t isolation of rcco.. ~ tc, but further application of Strepto7vtocil- (E;luco~ ~lose, 2-deoxy-2-[3-methyl-e-nitroso~ri~o-D]; STZ) to such cdls will result in killing of non-homologous recol.,h;~ t~ because tne incol~ola~d GLUT-2 gene will p~luce GLIJT-2 transporter, ren~oring the cells susceptible to STZ t~ rnt (the original cell was GLUT-2-).

W O97126321 PCT~US97100761 On the other hand, site-specific reco~lbi~ ~ion~ relying on the homology bel~,en the vector and the target gene, will result in incorporation of the sele~te~ gene and the drug selP~t~ble marker gene only; GLUT-2 seqn~nres will not be introduced in the homologous recombination event because they lie outside the fl~nking se~l~enres. These cells will be drug S resistant and but not acquire the GLUT-2 sequences and, thus, remain in.~c~ " to STZ. This double-selection procedure (drug~/STZres) should yidd reco~llbinauts that lack the target gene and express the s~l~ct~d gene. Further screens for these ~h~,nol~es, either functional or immllnologic, may be applied.

A mo~ific~tion of this ,.,locedu~c is one where no selected gene is inrlll~ecl i.e., only the scl~ ble marker is inserted into the target gene homologous sequences. Use of this kind of col,s~ c~ will result in the "knock-out" of the target gcne only. Again, proper l.,coll~billanl~ are sclcened by drug resistance and STZ re-cict~n~e (the original cell was GLUT-2-).
Genomic Site-D;.~ l Mutagenesis with Oligonucleotides. Through analysis of r~ ti~n-sensitive ,,,..~ i of Ustila~o may~is, genes have been ch~ that participate in DNA repair ( Tsukuda et a~., 1989; B~ cklwi~ and Holloman, lg90). One such gene, REC2, en~ o~l~s a protein that catalyzes homologous paising b~ ,cn co~ enl. ~)ts~ ~ nucleic acids and is Ic.~ ,d for a functional recombir~ o~l repair ~lway (Kmiec et al., 1994; Rubin et al., 1994).
20 ~n vitro ch~t~..;7A~ion of the REC2 protein showed that homologous pairing was more e~ri~nt bel~.e~ RNA-DNA hybrids than the cu,.c;.pol~ding DNA dl~plPxes (Kmiec et al, 1994;
PCT, WO 96/22364). However, çrr;~ in pairing ~I DNA:DNA d~ple.~çs could beenhanced by in~ ~g the length of the DNA oligolmr3eQtides (Kmiec et al., 1994). These observations led inves-ig~tors to test the use of rhimPric RNA-DNA oligo~ ot~ s (RDOs) in the ~,~d m~ifi~tior~ of genes in ~.z~ .51i~n cell lines (Yoon et al., 1996; Cole-Strauss er al., 1996; PCT W095/15972). Thc RNA-DNA oligonucleotides that were used to test this application CC,~t~ d seif-~nn~ ng .SG 1~ -ees such that do-lblc-hairpin capped ends are formed.
T~iis feature is tho lgh~ to in~,.,ase thc in vivo half-life of the RDO by dccie~i~g degradation by h~ Ps and .os~.r~u~ -. Further, the RDOs coQ~ -d a single base pair that differs from the target s~q.~ e and othenvise aligns in perfect register. It is believed that the single mismatch W O 97/26321 PCT~US97/00761 will be recognized the DNA repair enzymes. And the RDOs contained RNA residues modified by 2'-O-methylation of the ribose sugar. Such modific~tion makes the RDOIesis~lt to degradation by ribon-lrl~Ace activi~y (Monia et al., 1993).

S Two separate e~ t~l systems have been used to test the use of RDOs for targe~ed gene disruption in IllA-lllll~iiAn cell lines. In one system RDOs were used to target and correct an lin~ l)ho!,~,k-~ce cDNA in that was m~;n~A;.~rd in the L~.so,.lal DNA of Chin~se h~mcter ovary cells. An in~li~,~ forrn of Al~ ine phosphatase was converted to a wild-type form with an effiniency of about 30% (Yoon e~ al., 1996). In a second system, a genetic mutation within chromosomal DNA was targeted and collc~;lcd. A lymphoid blast cell line was derived from a patient with sickle cell disease who was homozygous for a point mutation in the ~globin gene.
Here again the overall frequcncy of gene conversion from the mutant to the wild-type forrn was very high and was found to ~e dose~ n~c ~~ on the co~ d~ion of the RDOs (Cole-Strauss et al., 1996).
Lf the use of RDOs or DNA oligonucleotides for the ~u.~oses of targeted gene conversion is broadly applicable to various ~ n.n~ n cell lines, then it offers several advantages to current t:~lu~vl~gi~s that have been used to A~co..~l?l;~h gene disruption such as homologous ,cc-J...h;..~~ion. Pirst, if gene com~e.~.on by RDO or DNA oligonucleotides occurs in various cell 20 lines at an effiri~n~y of 30% then this will r~ .,3ent a much higher rate than has been reported for targeted gene disruption via homologous recombination. Secondly, only short sequcnces are required for gene disruption by RDOs or DNAoligor~ oti~s(typically 60-mers to 70-mers);
wl~c~as hul~ologous l~o.,l~h~ation l.,qu,.~,s very long ~ ,tch~s of compl~ ~y se.lu~,nces.
~omnlogous se,l~cnfes from 9 to 15 kilobases are typically ,~co.. ~ in the co~ ion of 25 t~E,_th~ vectors. As a result, cc,~ u~lion of DNA vectors for homologous leco"lbination usually involves e~t~,n~ ,e gene mapping studies and time col,sul,~ g cfforts in the isolation of ~-~n.. ;r, DNA s~~ Y5 Such efforts are .. ~rces~ if RDOs are used for ~_t~,d gene co.,~ ,ons. Thirdly, assays for gene con~ion by RDOs can be ~.ru"l,ed 4 to 6 hours following i~llu~h ~ n of the RDOs or DNA oligonucleotides into thc cell. ~ co~ , gene W O 97/26321 PCT~US97/00761 conversion by homologous l~coln~ ation requires a relatively long period of time (days to weeks) b~ .ce,n the time of introducing the targeting vector DNA and assaying for l~co...bin~

R~ndom Integration. Though lacking the specificity of homologous ~ceol.lbillation, 5 there may be situations where random integration will be used as a method of ~ o~L ing out a particular endogenous gene. Unlikc homologous rcco.l.bination, the l~iwlllbil-~tori~l event here is completely random, ie., not reliant upon base-paiIing of co...l~lef ,r~ , y nucleic acid seq~ c. Random i,lh~dtion is like homologous l.,conlbil~a~ion, however, in that a gene co~lstlucl, often cr.~ a heterologous gene and a select~hle marker, integrates into the target 10 cell ~nn-mic DNA via strand breakage and reformation.

nc~ ce of the lack of s~ e-rc s~ ;r~ y, the chances of any given r~coln~
integra~ing into the target gene are greatly reduccd. ~Iso possible is int~ ion into a second loci, re~slllting in the loss of expression of the gene of interest. l'his second locus could encode a 15 tr~-~ ;on factor needed for e~,ession of the first gene, a locus control region needed for tne eA~ ,ssion of the first gene, e~c. As a result, it may be nrCCc~ to "brute force" the sel~ctiQ~
process. In other words, it may be l~ececc~ y to screen hLhd~Gds of thousands of drug-resistant ~o .l~ c before a desired mutant is found. Screening can be fa~ilit~tç(l for e-~n~rle, by e~rninin~ l~conlbil,fmts for e~ ssion of the target gene using immllnologic or even fimc-iQn~l 20 tests; eA~ ssion of the target gene in-lir~t,~ lcrolllhin~ ion elsewhcre and, thus, lack of suitability.

(iv) ~fe~hods for Increasing Pro~uctinn of ~ecol..bif ~ ~t Pcptides from Sec,~ Cells The present invention also contemplates 'M~;JI~ or increasing the C~ ~ tiçs of cells 25 to 1"~1 ce biolQg~c~lly active polypeptilips~ This can be ~Co~ hpd in some .~ es, by ,;.sing the ~JIu~s involved in protein p~cessi.-g, such as the endoploteases PC2 and PC3 (Steiner et al., 1992) or the peptide ~mirt~ting enzyme, PAM (Eipper et al., 1992a) in the case of amidated peptidc h.,...~ s W O 97/26321 PCT~US97100761 Expression of proteins involved in m~int~ining the speciA1i7~d phenotype of host cells, espe~islly their sG~.~tol~r capacity, is i~ u~ t. Fn~ e~ g the ovc~A~ssiûn of a cell type-specific ~ c~ /tion factor such as the Insulin Promoter Factor 1 (IPF1) found in ~ cleatic h-cells (Ohlsson et al., 1993) could increase or stabilize the capabilities of e~ e~,.c;d S n.,~oe ~~iocrine cells. Insulin promoter factor 1 (IPF-1; also referred to as STF-1, IDX-l, PDX-1 and bTF-l) is a homeodomain-containing tr~ncrnrtion factor ~vl)osed to play an in~l~oll~t role in both ~ lclc~ic development and insulin gene expression in rnature b cells (Ohlsson et al., 1993, Leonard et al., 1993, Miller et al., 1994, Kruse et al., 1993). In embryos, IPF-l is cA~lGDsed prior to islet cell hollllol1e gene expression and is restricted to positions within the 10 primitive foregut where pal~c~as will later form. Indeed, mice in which the IPF-l gene is disrupted by targeted knockout do not form a pancreas (Jonsson et al., 1994). Later in pancl~,z~ic develG~ cl1t, as the dirr~ t cell types of ~he panc~as start to emcrge, lPF-l expression becG...~s lG~ d preAo...;.~ ly to b cells. IPF-l binds to TAAT con~rC-ls motifs contA;nrd within the FLAT E and Pl ele ..r~ls of the insulin enh~nre~/~,lul~oter, wlle,. .~pon, it i 15 with other tr~nccrirtion factors to activate insulin gene transcription (Peers et al., 1994).

Stable ove.GA~l~,;,sion of IPF-l in n~ oc~ jn~ ~3 cell lines will serve two ~ulyoses~
First, it will incl.,ase L~ g. ~~r GA~SSioll under the control of the insulin en~ e /I,r~ ût~r.
Seron~l, be~nce IPF-l appears to be critically involved in ~ cell ~l~luldtion, stable 20 o~ ,x~s~ion of IPF-l in ~ cell lines should cause these mostly de~lirf~le~ d ~-cells to regain the more difr,.~ d function of a nonnal animal ,B cell. If so, then these~ f~ iated ~ cell }incs couid yot~ y function as a more ~rfc~ e nc~u~ c celltype for cell-based delivery of fully processed bioactive peptide hormon~c Also, further ~.n~;.. r~ g of cells to generate a more physiologically-relevant regulated secl.,tolylf~;>yQn~ is~ im~od F~ l~les would include en~ , ;ng the ratios of glucokinase to he-Aoiunase in rat in~lllinom~ cells that also o-elcAyl~ss the Type 1~ glucose l,~s~l~l (GLUT-2) such that a physiologically-reievant glucose-stimll~ I secretion of peptide hormf)nçs is achieved. Other çys~pl~5 include c.~ re-;ng ov~"cA~ siûn of other ~ lling yl~t~ins 30 icnown to play a role in the regulated sccl~loly Ic;~yOl'lSe of l~ oçrinr cells. These include W O 97126321 PCTrUS97/00761 cell surface proteins such as the ~-cell-specific inwardly rectifying pot~csillm channel (BI~;
Inagaki et al., 1995), involved in release of the secretory granule contents upon glucose stim~ ion the sulro,lylu,ea receptor (SUR), and ATP sensitive chAnn~l Other cell surface sigr ~lling l~,ce~lols which help pot~nti~te the glucose-stim~lAt~l de~ tion of ,B-cells S inrlurling the ~lucagol~-like peptide I receptor (Thorens, 1992) and the glllcose-dependent inc~ ol~u~ic polypeptide ~cceptol (also known as gastric inhibitory peptide receptor) (Usdin, 1993) can be ~nejne~red into neuroendocrine cells. These ~-cell-s~irlc ci~nqlin~ l~;cept~"~, as well as GLllT-2 and glucokinase, are involved in secr~to~y granule release in response to glucose. In this way, glucose stirn~ te~l releasc of any hetcrologous peptide targeted to the 10 secl~tol~ granule can be engl,-~ .,d. ~It~rn~tively, other cell surface signaling proteins involved in non-glucose-stimulated release of secretory granule col~t~ can be engil-~,..,d into neUloe~ or~rine cells. Examples would include rele~ci~g factor rec~,~>lw~ such as Growth ~cnnnn. Rele~cing Factor ~ceptor (Lin e~ aL, 1992) and SoTn~tsstAtin or Growth Hormone cing ~o~ nc Reccptor(Mayo. 1992).
One potential target for genetic es~ P~-;-.g to improve cell chaln(~ ;.ctics for protein ~u~ isl~ is hexokinace I. It now has been d~t~ ~ined that intelr~ing with hexokinase I
fu.,c~ion r~d~ces the growth rate of cells. The following is a ~jccllccion of çn~ ;"~ of h.-Y-~kin~cçs according to the present invention.
Mitochondrial Bindillg. Low Kn, h~ol ;.-~ces _re flictinguich~d from gl~c~ .Ace in that tney are allosterically regulated by glucose~hos~tç and by binding to mi~orh~ ris~
(Wilson, 1968; 1973; 1985; 1995). Micromolar conr,~ 1;ons of glucose-6-~ho~pl~le inhibit the ~clivilies of hexokinases I, II, and m, but appreciable inhibition of glucokinase lC~lu~lcS
25 glucose-6-phos~hS~e con~e ~l~alions in excess of 10 mM. Binding of hPYc'-in~es I and II to mit,~l~nnr1.;s~ alters their kinetic plO~, ies (Wilson, 1968; 1985; 1995), while glucokinase does not appcar to bc capable of binding to mi~ochol-~. ;a at all (Beckcr e~ al. 1996).

When bound to ~it~h-~Ddns hexokinase I undergoes an inc,~asc in affinity (a d~,~.ase 30 in K",) for its ~ c ATP (Wilson. 1985). In addition, the C~ llC t!eCr~ S far less inhibitable by g}ucose-6-phos~hate, as inAic~re~l by a several-fold increase in Kj for this ligand (Wilson, 1985). Studies with hcxokin~ce 1 have reveaied the existence of two types of mitochorltlriAl binding sites (Kabeer and Wilson, 1994). Glucose-6-phosphate causes rli~ rç."~ of a pro~.lion of mitochondrially-bound h~xokin~ce from one type of site. The enzyl~le that remains S bound to mitochonrlriA after glucose-6-phosrhAte ll~a~ t is considered to occupy the second site, from wnich it can be removed by IIC~A~ t with 0.5 M KSCN.

It has been known for some dme that limited digestion of hexokinase I with chymotrypsin yields an enzyme fragment that retains catalytic activity but that loses its capacity for 10 ~---~ocho~ q-l binding, and that C.IZylllC treated in this Illanner is lacking in a portion of its N-terrninAI domain (Polakis and Wilson, 1985). The N t~,---in~l sequences of both hpyo-kin~ces I
and Il are relatively hydrophobic, and it has been shown that the hy~ ph~bic N te....;~ of hey~ qce I is capable of insertion into the lipid bilayer of the mitochnn~n~ ne (Xie and Wilson, 1988).
~ lb5~ ly, Gelb et al., 1 i992) ~ 0~ ,ated that a chimeric protein concicting of the N-terminal 15 amino acids of hexo~inase I fused to chlolA ~.ph~ icol ac. lyl~ r~ -~ce was L~p7~e of binding to rat liver mitochor~AriA and that this binAing was c~ -er;~ , with ;~-II...-~I;r h~Yql~in~ce I (Gelb et al. 1992). ~lthon~h Gelb etal. (1992) have 5n~gest~cl that the first 15 20 amino acids of h~y~kin~ce are S -rr~ to larget such a c-k.-..---;C protein to mitocho~r1~ these studies were not ~3esi~ to attempt to alter m~t~l~olic regulation in target cell lines. Thus, the e~ required to effect ~licpl1,e.,.. ~n~ of ~do~'ollc h~Yrl-in~ce from its ,.. ito~h~.r bi~ g site were not unequivocally i(lcntifie~l in the study of Gelb and eo ~ ul:~ as ~licc!.csc,~
below.
While the results of Gelb et al. (1992) argue for the in,~u~ ce ûf this small N-terminql se~ in ~ g of hexûkinase to mito~hQn~ others haYe sugge~ that other regions of the ~ may also be hllyû~ t in stAhili7in~e the interac~on (Polakis and Wilson, 1985;
Felgner and Wilson~ 1977; Smith and Wilson, 1991). This is based on studies showing that 30 hexokinase I binding to ~ ochol~.Lia is st ' jli7ed by Mg2~, an effect likely ~.n~c~ g elecllusld~ic interactions between the enzyme and the outer mitochondrial membrane (i.e., not involving the N-termin-q-l 15 arnino acids that are intercalated into the l,lc.,lblane~. Therefore, the mi~ochol~rlri~l binding regions of HK have not been clearly i~lentified to date, and t_ere is even less i~ lion available on the issue of HK ~
s At least part of hexokma~se binding to mitoch~ is via interactiûns with 11.~ ,c,~ of a family of p~uttillS known as voltage-dcy~ndellt anion çhqnn~lc (VDAC) or porins (Fiek et al., 19~2; Linden el al., l 982). These porins form a chqnn~l through which metabolites such as ATP
and various anions t~ e the outer mitorhon~lris1 nI~m~ C. Binding of hexokinqces to porin thus may ensure a supply of i1.~.,.. loc1.onl1riq11y-gc,l~ ted ATP ac substrqte Co~sl-u-;ls of the present invcntion may comrrise the N tcl~ al l~ amino acids of a he~cokinase el~y..le, ~f~,.ably hexokinase I or ~, since this se~ t should be easily e~ esscd in cells and l~tah~cd as a stable peptide. COnSllU~,lS comprising the entire N-terrninal ~lom~in of 15 either hexokinase I or hcxokinase II, ûr the intact, full-length hexokinase I or II pr~t~,ins that have been l~;nd~,d inactive by site-directed m11tag~-n~cic of amino acids that are illl~t for th enzyme's catalytic function are also co.~ ..plated. CO11~IUCLS based upon hexokinase I will be particularly, or even exclusively, ~lc~.,.l~d in certain ~.. hoA;.. 1~;

The reason for pl~,f~,.,.. ~ the N te .. ;.. ~1 dnm~in CO1J~ Cl is that this elem~-n- seerns to C~ C a complcte ~ UI~1 t1n~sin based upon studies in which this domain can bee,~ ui in bacteria and shown to bind glucose-~l-hosph~te (Wilson, 1994; Arora et aL, 1993;
Whitc and Wilson, 1987; White and Wilson, 1990). This sn~stc that the intact N-~domain should fold and form a ~ c~ , analogous to its ~llu~;lul~ in the full-length hexokinase I
25 or II protein. As the present inventors cont~mrlate that this ~LIu~lul~ m~dist~s ~tt~5h",. l-t of the intact hexolcinase protein to mitorh~ .ia, the intact, col.~clly fo}ded N-terrnin~l domain is a ~,~f~ ,d c- .bo~ of this invention.

Por ç~ involving the N-t~rrnins1 domsin a SC~J--- ~1 CC~ amino acids l-30 455 is ~l~f~ d because of a naturally oc~ NcoI l~ n ~zyllle site in the DNA

W O 9~/26321 PCT~US97/00761 se~ nrc co..~ ,onding to amino acid 482. This NcoI site allows the fragment encoding the N-te....;..~l domain to be easily isolated and subcloned, and also allows direct fusion of the N-terminal dom~in of hexokinase to the intact fi~nction~l sequence of glucokinase via an NcoI site located at the AUG start codon of this gene.

Of course, it will be understood that peptides, polypeptides and protein ~om~inc of any te length bet~ about 15 amino acids and about 455 amino acids, and longer t~,ins, may be used in displacing endogenous hexokinase from the ~~-;tocho~
Accordingly, corls~ c~s co~ ;c;l~g about 20, about 50, about 100, about lS0, about 200, about 300 or about 400 arnino acids in length may be used for these purposes. lt is also col~tf.~ tPd that an intact hexokinase protein that is ~cl~d~ d catalytically inactive will interact with ...;~xh~ A ia in a manner ide-nt~ to the active proteins. Expression of such a HK variant is fo.~, another method for ;l~hi~it;rlg endo~ o~ls HK (Baijal and Wilson, 1992). In~tivated, hexokinase proteins include those that have been subjected to ch~tnir~l m~t~g~nçsiC and also 15 those y~oduced using molp-c~ r biologjr~l techu.i~u~s and l~col..h;~ t protein pro~ltction The ifi~ntifi~tion of -~yl(J~iate pol~ dc regions and/or particular amino acid sites that may be ~ted in order to ina~,~iv~te h~.;okirsce will be known to those of skill in the art.
The crystal structure of certaln hP~in~c~ er~y~s is available. Coupling the crystal ~ c 20 illfO~ ';OIl with a c~ul~A..~ol of the primary sc~ -re info,...-~;ol- for various distinct hexokinases will allow one to identify those regions and sites that are hllpol~lt for hexokinase acdvity, such as the binding sites for ATP, glucose and glucose-6-~,hasp~te. This has been scd in detail in various publications, such as Printz et al. (1993), h~cu.yc~;ated herein by .~,f~,nce, which inforrnation can lbe used in co~ on with pl~y&ing ..~ and variants for 25 usc herewith. Deletion of cerhin ~Imino acids or peptide sc~,J..r-~-~c, as may be achieved by molecular biolog~l m-~-.ir~ ion is another col~t~ t~d m~tho-1 for preparing inactive he~cokinases.

The ~ ~ glycerol kinase is another protein thought to bind to mi~.)ci ~m~ via porins 30 or VDACs (Adams et ~ll., 1991). Glycerol kinase catalyzes fo~ion of glycerol phosph~t~

CA 0224643l l998-08-l2 W O 9~/26321 PCTAUS97/00761 from glycerol, using ATP as phosphate donor. Thus, expression of glycerol kinase in cell lines ~y~es~ts an ~ rn~tive to ~plussion of inactive hexokin~ce proteins or fr~ ntc thereof which is also con~ ~ed for use in the rlicpl~remPnt of endogenous low-Km hexokin~ces from their normal mitochondrial binding site.

A particularly po~.~.rul m~.tho~l of inhihiting hexokinace within a ~ n cell invol~,s the ~iicpl~e.... nl of hexnl~; .n~e from the mitocllon~l~t~ and the col-co~ provision of active gll~c~ ce. This is advantageously achie~,~d by providing to the cell a hexokinase-glucokinase chhllcla or fusion protein, in which the h~s~'-in~ce portion is capable of binding to 10 the .~ ol--h~ and yet does not exhibit h~xo~in~ce catalytic activity, and in which the glucokinase portion is catalytically active. ~h~rn;-~lly-fused polypcptides are a possibility, but ~COI~k;l'~-~t ~lute;lls a.~e ~ r~l~y most ç,.~fe,l~d for use in this manncr. The identifi~ ion of a~,.upliate hexokinase fr~g~lltc for use in such a Chi~ ,la has becn llescri~ed herein above.

In terms of the glLco~ ase ponions of these fusion ~lU~ inc, any gluc~l-in~ce-derived sG~lu- - re that cont~inc enough ~ r se~ e information to confer gJIleokin~ce catalytic activity to the cL,~,e.a will be useful in this corlt~Y~ However, it will often bc ~r~fe..~d to use the entire glucokinase ~.llc as this is more s~i~hlrol~r.al~l in terms ûf m~thoclQlogy. Again, one may look to the extensive infonnation availa~le in various puhliched lef~_~cnccs in order to 20 assist with the i~lentifi~ ion of ay~.op,iate glucokinase el~ylllcs or frae~ntc thereof.

At this point, a ~ c~c~ion of the kinetic ~,op~.~ies of hexokillase and glucokinase is relevant. It will be ~ e.~lood that in providing a filnr~ions~ equivalcnt of a hexokinase or glucokinase u~ .,e, one would desiTe to lu,c~vide a protein that has substantially the same kinetic 25 pa~ as the native C~1L~n11C. Equally, in providing a hexokinase mutant that is devoid of cat_lytic activity, one would provide an cl,~ e that is either comrle~ely inactivated or whose kinetic pa~ have been shifted so that it is, in fact, distinct from the native e.,L~e.

wo 97/26321 PCT/US97/00761 Table l, below, sets forth a co~ uison of glucolcinase with hexokin~es I-m. Thisinformation may be used in order to d~t~ .. -e whether any particular variant is "equivalent", and also, to co~finn that any inactive Illl~ have indeed been properly ~ bl~-1 A C~Q~ ~y-~ ~r n of Glucokinase With ~lf~ irs~es Km glucose 5-12 rnM 0.02-0.13 mM
Km ATP 0.5 mM 0.2-0.5 mM
Ki C~-6-P 60 mM 0.2-0.9 mM
Molecular weight ~2 kd 100 kd S~ cf~e~ce ~ ose I~ mose 0.8 1-1.2 2-Deoxyglucose 0.4 1-1.4 ~,u~ose 0.2 1.1-1.3 The activity of glucose as a ~ul)~ e is taken as 1. The other ~ f ~.~ are c~ynessed in re}ation to 5 the activity of glucose as a substrate.

Treh-'~se-~ _l.osphate Metabolism. Ln Baker's yeas.t, glucose phos~h~j~ylation is also catalyzcd by a farnily of hexokin~ces that are related in sequP-nre and function to the ~ n hexo~inase gene family. Yeast Cf.,llS, however, contain other genes involved in carbohydrate 10 mPt~holi~m for which there are no l~ n co~ntc~y~. The trehalose-6-phosphate synt~c/~l~,halose~ n "~h~ pho5~h~ ce col~ x is an examp~e of such an activity.

The trehalose-6-phosF!h~le synthasel~ho~h~lpce complex catalyzes t_e formation of ~halose, a ~i;c~c~h~ id~ of tvvo glucose molec~ s (a-D-glucopyranosyl (1-1) a-D-1~ ~h~v~l -~o~;~ie~ by first forn~ing trehalose-6-phosl~h~ by conri~ncqtion of two m~lecllles of glucose-6-pl~o~pi~ nd then using its phosphatase activity to remove the l~nsi?h~le groups to W O 97t26321 PCTnUS97/00761 g~ e.d~e free trehalose (Bell et al., 1992). Trehalose is thought to lep.escl)t a form of storage poly,~ k~ P in yeast, bacteria and other lower org~nicmc, but neither the trehalose-6-ph~sph~te synthase enz~.l,e, comp]ex nor its products treh~lose-6-phosphate or free trehalose are known to be present in m~mm~ n cells.
s Blazquez et al. have d~.llc,n~Llaled that trehalose-6-phosphate can inhibit the activity of hexol~inases from a variety of ditr~-,n~ Gl~ ..c, in~ ing rat brain, which eAplcssc5 p,~do...;. ~ .tly hPYokin~ce I (Blasquez e- al., 1993). This has led to the s..ggeslion that trehalose-6-1~hosph~te may be an i~ )olL~lt regulator of ~Iycoly~tc flux in yeast cells. ('or-sicte~ with this 10 notion, the yeast gene known ac ci,f-l wa~s originally cloned from yeast that are unable to grow in glucose (Blasquez et a~., 1993) and s ~hse~lv~ ~ly shown to be iA-ontir~l to the .cm~lPst subunit (56 kD) of the trehalose phosphate~ hasc/trehalose-6-phosphate ~hos~ ce cor ~rlex (Bell et al., 1992). Cells lacking in the Cl[F-l gene product exhibit rapid depletion of ATP, presumably becallse they are ur~able to produce treh~lose-6-~ nsyhn~p that norm~lly scrves to ~Od~tr yeast 15 hexokinase activity. It is bclieved that the 56 kDa CIF-1 gcne product enro~lrs the trehalose pho~ t. ~ se actilrity (Bell et a~., 1992).

One of the three gencral mrtho~ls. desc~il~d in this applir?ti~n for inhibiting low E~", h~ in~ce activity in ~"a"~ n cells is to express an enzyme, such as yeast trehalose-6-20 pht~sphst~ ~ylllhase, that will allow treh~lQse-6-phos~h~t~ to accl.~nt~ te. Tbis will have two effects. l;irst, the ~rcl~mll~tPd trehqlose-6-ph~ k~ will serve to ~IlQste.ric-s-lly inhibit e~ sg .~ c low Km h~ in~ce activity. ~S~Qn~ where treh~losc-~hos~hAr~ synthase is used, this c~."e will divert glucose 6 ph~sF~hate into trehalose-6-~hos~hAt~ at low, non-stimulatory glucose c~-~r<..l. ,t;o~.c where low K", hexokinases but not glucokinases are active, thereby 25 ''short~J.~;uit~g" metabolic ci~sllirlg for insulin secle~,on, which is lLougl~L to require ATP
y~ uced via further glucose metabolism (Newgard and McGarry, 1995~.

A ~;u~l~n~ f~ ,d gene for use in t_ese aspects is the S. cerevisiae gene c-~oAi~g tr~hslvse-6-1-ho~hs~e sJ..lhgse (ll?Sl). Genes from several other org~ni~rnc f ~~ro~ g treholose-30 ~yh~At - synthase have ~cen i~olatPd and the amino acid sc~luences dedo~e~l These include W O 97/26321 PCT~US97/00761 E. coli (ACCeSS;On # X69160~, S. pombe (# Z29971), Mycobacterium laprae (# U15187) and Asperg~llus n~er (# U07184). It is contel..pl~e~l that any of the foregoing or other biological filn~tion~l equivalents thereof may be used in the context to the present invention.

S Hexokinase Inhibition at Nucleic Acid Level Several different ribozyme motifs have been ~ies~ribe~ with RNA cleavage activity (revicwed in Symons, 1992). F.Y~rnples that would be eYrecte~ to function equivalently for the down reg~ on of low Km h~xcl-in~ces include sey~ rec from the Group I self splicing introns including Tobaco Ringspot Virus (Prody et al., 1986), Advocado Sunblotch Viroid (P~111k~;~jC et al., 1979 and Symons, 1981), and Lucerne T~ siclll Streak Virus (Forster and Symons, 1987). Se~u~ cs from these and related viruses are ~er~ d to as h~.. ~.l.~d h:.. ~.l.~a~l riboLyll,c based on a p.~i~;lcd folded secondary ~llu~ c.

Other ~uiLable ribozymes include se~ res from RNase P with RNA cleavage activity(Yuan et al., 1992, Yuan and Altrman, 1994), hairpin riboz~lllc ~ s (Berzal-Herranz et al., 19g2 and Chowrira er al., 1993) and TI~AI;I;S Dclta virus based ribozymes (Pc~lvtla and Been, 1990). The general design and o~ ion of ribozyme directed RNA cleavage activity has been rl;r ~ ~ ,s~i in detail (Haseloff and C~erlr-h 1988, Symons, 1992, ChO~ a et al., 1994, and Thc....l~sol- et al., 1995).
Tlhe other variable on libo~",e design is the SCIeC~iO11 of a cleavage site on a given target RNA. Ri~;~."cs a~e ~d to a given se~uence by virtue of a~n~ ng to a site by co..~ r~ basepairblt~ c~ nc. Two~llct~h~sofhomologyare~ dforthisl~ tillg.
These ~ ,t~hcs of hnrnolQgous se~ es flank the catalytic ribo;~"le slluctulci defined above.
25 Each stretch of hnmollq3ou~ sG~IJOl ~ can vary in length from 7 to 15 ~nçleoti~ s The only nl for ~efinin~ t_e homologous se~u~ A~es is that, on the target RNA, they are s~ "te~
by a spçifi1c Se~lu~ G which is the cleavage site. For h~.. ,bcad lilxiLyll-e, the cleavage site is a ~ lw!;-le se<~ ce on the target RNA is a uracil (U) followed by either an ~ienin~., r~n~ or uracil (A,C or U) (PeITiman et al., 1992 and Tho...l~~4~- et al., 199~). The L~4u~ y 30 of this ~ oLide OC~ in any given RNA is st~ticri~ y 3 out of 16. Th~ rul~, for a W O 97~26321 PCT~US97/00761 given target ..lcsc~ RNA of ]000 bases 187 dinucleotide cleavage sites are st~ticti~ y possible. The m~Ss?ge for low K,r, hexo~in~ces targeted here are greater than 3500 bases long witn greater than SOO possible cleavage sites.

S The large number of possible cleavage sites in the low Km hexokin~ces coupled with the growing number of seyhences with demonstrated catalytic RNA cleavage activity in-ljc~tç$ that a large ,~ of ribozymes that have the potential to dow...~ ~llate the low Km hcxokinases are available. Decigning and testing ribo~yl"es for effi~ t cleavage of a target RNA is a process well known to those skilled in the ;~t. Examples of sci~ntific m~tho~s for tl~ci~ing and testing 10 riboL~.,es are ~escn~ed by Chowrira et al., (1994) and Lieber and Strauss (~995) each incol~,ated ~y l~f~.~..ce. The i~ntifi~ ~tion of operative and plcÇc~lcd se41lerlces for use in hexokinase-targeted riboL~ll.es is simply a matter of ~p~illg and testing a given sequence and is a l.Ju~ ely practiced '~ enil~g mcthod known to those of skill in the art.

Combination of ~nhibitory Methods. Any of the three gcneral m~thods of HK
inhibition fies~nbed above (~ oci~ l HK ~1icpl~çrn~nt, trehalose~phosph~tç ~ , dlion and anti-HK rihG~,lles) may be c~ kined with one an~th~r and/or with other ~ r. .; ~g c It is particularly co.lrf-. plated that these mf.th~c could be used in co..-k;.-~-iQn with glucokinase overproduction. Glucokinase ov~ c~J~Jction alone is even ~hought to be a useful 20 method of il~hi~ g h~ in~c~, as set forth ~elow.

Low K,l~ hexokinaces in~ hexokin~Qes I and n that are prcsent at high levels in mamma}ian cell lines7 are inhil~itP~l by ~ cosG-6-phosphate. Thus, this invention also relates to h~-ds for ~ glucose-6-~>ho~ph ~r- at high levels in cell lines. The ~l~fe.l~d method 25 for achieving cnncicte~ntly high levels of glucose-6-phosp~ in oells is to OV~ 5S
glucokinase in such lines.

ssi~n of gll~crl-in~ce is co~ci~lered advantageous for two distinct reasons. First, as dc~ in U.S. Patent S 427,940 c~ .,sion of g~Y~cl-in~ce is part of an adv~nt~ollc method 30 for engin~ering of ~lucose-stim~ d insulin sccretion in cell lines. Glucokinase eA~l~ s~ion is herein shown to have the added benefit of ~ h~ ing high levels of glucose-6-phosphate to keep low K,l, h~Y~kin~ce~ in an inhibited state. This advantage would bceGlllc particularly ~elt,~ at glucose concentlations in the physiological range (4-9 mM), bec~ e glucQkin~cP is âctive at these levels. Also, while glucokinase is a ~ ,n,ber of the hexokinase gene family, it is S not itself inhibited by glucose-6-phos~h~te Advantages of Hexokinase Inhibition in Mamn~lhn Cells. The various aspects of this invention focus spe~if~lly on red~lcing the levels of low Km hexo~inase activity in m~mm~ n cells. A particular type of target cell is a neuroer~locrine cell. There are at least two 10 sigrific~nt achievements ~r,co.~ ;hed by the h~x-'-in~e inhibition of the present invention, as set forth below.

~ addition to the regulation of insulin se~.~,~io n by ghlcose, the h~xokin~ce gene family may also be i~l~o,l~t in the regul~i(?n of cell growth and proliferation. As tl~-crril~ed above, 15 in~ s in low Km h~o~in~ce acl~ivity usually correlate with the t~i~ On of cells from â
no~mal to CallCelOu~phC~Ol~C. However, the correl~tior- has not been proven to exist as a cause and effect rel~tion~hiI~. In addition, increases in mitotic activity are not llui-el~ally linked to in~l~C~ion of low Km h~-~cin,q$es The activity of these cnL~ e,s did not inc ~asc in ,op~ cli~ mouse beta cell lines over-e~y,cs~ing sirnian virus 40 large T antigen (Tag) 20 (Radvanyi et al., 1993); nor are tl~ey universally elcv~t, d in fully ~u~srol,l,cd mouse ,B cells (Efratetal., 1993).

The ~ nr~ion of hexokinase a~livily in a ccll line by any ~ klc method, inr~ ing any of thc noYel l-- Ll~ c disclosed he~ein, is CQ~ ed to be of use in inhibiting cell growth.
25 He~co~ se I was di~co~ to be a regulator of cell growth during the i"~ntol~' studies in which a RlN cell line ~861X4) that c~ ,.;...c a dis~upted allele of the HKI gene was ~ iisingly found to grow about one-half as fast as clones c~--t~;l.;.lg the nonnal cQmrlim~nt of two HKI
wild-type genes.

W O 97126321 PCT~US97/00761 A relationship between law Km hexokinase activity and cellular growth rates has three ilnpo, l~1t implications relative to the application of cell-based therapies. First, from the perspective of iterative genetic en~in~ering, an untimely or unregulated decrease of hexokinase activity will potentially hinder the growth and selection of clones possessing desired genotypes 5 and traits. A cell line that over-expresses hexokinase I from a regulatable promoter may provide the optimal genetic background for eng"-f~ g of gene targets. For example, a RIN cell line could be developed that transgenically expresses hexokinase under the control of the tetracycline (Tet}-recict~n~ e operon regulatory systern (Gossen and Bujard, 1992). This expression system allows powerful transcription of gene products and permits the ablation of gene expression in the 10 presence of Tet and Tet derivatives. ~frat et ~l. (1995) have demonstrated the feasibility of using this expression system to regulate large Tag gene expression. The expression of Tag caused transformation and expansion of rnouse beta cells. A decrease of Tag expression, by the in vitro or in vivo admint~tration of Tet, led to an inhibition of cellular proliferation.

A RIN or neuroendocrine cell line that expresses HKI from a repressible promoter could be further engh~e~.~,d to express high levels of human insulin, glucokinase, and GLUT-2. In addition, such a cell line would be an ideal host for the ablation or down regulation of low Km hex--lrin~cec Such en~"1ee.,ng could be pursued without the hindering complication of slowed growth. Following a series of desired genetic manipulations, the growth of the cells and the 20 glucose sensing ability could be mo~lul~t~d by down regulating hexokinase expression.

A second implication of low Km hexokinase as a regulator of cellular growth concerns the use of engineered cells for in vivo ~l,el~;es. It is envisioned that cell-based delivery will be con~uc~e~l by m~int~ ce of the cells in vivo in a perm-selective device. It is contemplated that 25 cells with reduced levels of low Km hexokinase activity will survive for longer periods of time in devices or capsules as a consequence of their reduced growth rates.

A ~hird implication of low Km hexokin~ses as regulators of cellular growth involves the creation of novel ,13-cell lines. The over-expression of HKI by introduction of exogenous DNA
30 into a p~ ~y beta cell could be an essenti~l ingredient of the transfolTnation process. NIH-3T3 WO 97n6321 PCT~US97/00761 cells, an irnmortalized cell line, showed in~lcases in glycolysis and growth rates following transfection with low Km heyn'-inq~e (Fanciulli et al., 1994). In a ple~.l~d embo-iim~n~
hexokin~ce I would need to be under the control of a promoter that can be down regulated. Such tl~lsc~ onal regulation would allow the s~!~sc~ .t modulation of growth and glucose s~ ncing s A second illlyOl~lt reason for re~uring h~xQkin~ce activity is that it will contribute to the de..lor~.I.. In of e~ d cells that exhibit glucose-re~ trblc protein secl~,~ioll, the most G~ l aspect of which is L~l~scntly the physiologically regulsted release of insulin. ~sulin release from the ~-cells of the islets of ~ ~ng~rh~c in the pal,cl.,as is lJlo...;--f l~lly regtll-s-te.l by 10 the circtlls~ting glucose con~ .d~ion. Glucose stim~ tes insulin releace over the physiological range of glucose col-fe~t.nlionc (approxima~ely 4-9 mM), with the ~m~unt of insulin secreted being yl~olLioual to the rate of gll~c~e metabolism (Newgard and McGarry, 1995).
C'Tlncose ph~ ,ho, ~lation appears to play an i~ ol~t role in regulating glucose15 metabolism and insulin l.,sponsi~,~,. c,ss (Megls-csr~n and ~gtc~hinc~y, 1986). Thus, while islet ~I.,.rl~: contain apy~o~ ly equal ~ ~-o ~l~ of high Km gl~ in~ce and low Km hPyo~insce iCs (Megi~cso~ and l~ ~tr~in!~ky, 1986; Hughes et al.. 1992), the hexokinases appear to be inhibit~d in intact islets, ~ ,sul-,ably by gJIlr,ose~ sph ~.t~" allowing the glucokinase activity to be yl~d~llunant. Since glucokinase has a Km for glucose (appro~im~tely 6-8 mM) that is within 20 the physiologjr~l range, it is ideally suited for reg~l~tin~ glycoly~ic flwc and insulin release in pluyullion to the ~;A~ Pll~ r ~ se c~ tl -~

The co~cc,~t of a regulator~r rûle for glucokinase, which has been d~ ,lopcd over seve~alyears (.~fç~3~cso~ and Ma~crhinc~y~ 1986; ~atchtthin~y~ 1990), is ~uy~lled by reccnt genetic 25 and molecular s~l~ies, in which reduced eA~les~ion of ~ in~ce was shown to result in less robust glucose-ctimlll~ted insulin s~x~etion (Frogucl e~ al., 1993; Efrat et al., 1994). Islet ~-cells are also e~uipped with a spe~ i7~1 glucose ~l~S~lt~, GLUT-2, which like ~lucokin~ce is the high K,~ ..he, of its gene family W O 97/26321 PCT~US97/00761 One of the present inventors has shown that GLUT-2 and glllc~-ki~ce work in tandem as the "glucose sensing a~y~a;us" of lhe ~-cell (U.S. Patent 5,427,940; N-,wgar~ et al., 1990). U.S.
Patent 5,427,940, incol~olated herein by ~efclence, describes methods for co~ g glucose sensing in ~ of nrlocrine cells arld cell lines by ~ sr~ ion of such cells with one or more 5 genes selcc~ed from the insulin gelle, the glucokinase gene and the GLUT-2 glucose transporter gene, so as to provide an en~;~.P.,!~,d cell having all three of these genes.

The OVe1G~Y1~SjjOI1 of low Km hexolin~c~c is known to exert a ~o...;,~ t effect on the ~lucose co~ nt.ation threshold for insulin se~le~.on. O~ ,ssion of a low Km hexokinase 10 from yeast in islet ~-cells of tr~n~g~ic ~nim~lc results in increased rates of low Km glucose m~t-hclicm and enh~nced insulin release at :~ub~ iological glucose co~ tions (Epstein et al., 1992, Voss-McGowan et al., 1994). Similar c~ g~.s were noted upon ov~ yl~ssion of hexolcinase I in icol~t~d rat islets (Bec~er et al., 1994a) or in an ill--lirf~ t~A inclllinoma cell line called M~-6 (Ishihara et a~., 1994).
1~
It has bcen shown that the ll~,ulo~ o(rine cell lines that are col.t~ ted for use in ~n~ g artificial ~-oells generally have siE,--~r.~ A ~ higher low Km hcxokinase activity than normal islet ,B-cells (~ughes et al., 1992; Efrat et aL, 1993; Hughes et al., 1993; Ferber e~ al., 1994; Kn ~rL et ~1., 1994), and that gll~cose metabolism in such cells is highly active at low 20 glucose COI-C~ a~ions. As the gluco}cinase:htoyo~in~ce activity ratio is a critical ~3e~ u of the glucose rejpol-.ce threshold in insulin se~ g lleu.~fn~oçrin~o- cells, and as an imbalance in favor of hexokinase can cause insulin secretion to occur at ~Incose co~ n dliOnS that are bdow the physiological ~hlGshold, it is evident that the most ~IGfcll~d artificial ~ cells should be further ~...2~,;..~ -~,d to reduce h~Y~'-in~e activity. The application of the mP.tho~s of the present 25 in~ ioD to the develo~nt of improved insulin se.,le~illg cells thus ~ ,s~nts a .cignifir--lt advance.

Il~hibition Le~els As ~efin~(i herein, the degree of inhihition of hpyokin~ce that is cf~ is that necessary to achi~ a ~hlcose l~,s~onsJ~G insulin secretion in the physiologic .

range of 1.0 to 20 mM ~ c~ce. It will be understood by those working in this field that the absolute level of inhibition is fiiffc~llt to predict. Mea~u~c~ nl~ of hexokinase and ~hlf c'-inqce in freshly i~n~ islets as well as cell lines varies dr~m~fif~lly. Ratios of HK to GK can vary from 2.8 (Burch etal., 1981) to 0 8 (Liang etal., 1990) to 0.5 (Hosokawa etaL, 1995) in fresh 5 islets all with "normal" glucose stim~ t~l insulin secretion. Reports of inccn~,o~atcd herein by lef~,..".ce cell lines with "normal" secretion shows an HK to GK ratio of 0.6 (Efrat et al., 1993) in the range of the frcsh islets. These disclc~ -f -f. s illustrate the f1iffif'11ltiee in specifying a~soh-t~ ..he.l~ of glucokinase and hflrokin~ce activities, hence the p~cf~ nce for using glucoee respo~;vc insulin secretion ranges as a mf ~nin~ful p~ha~ er in this c~ t. ;~ the 10 cells invention.

(v) Methods for Rc-f ~p;,~rf - ;q~ Fn~ r~l~d Cells In many situations, mnltirle rounds of iterative ~n~.-eel;.~ will be undertaken in g~ g the final cell lines. The events that may be con~ t~ as separate consl.uclion events~5 include blockin~ c~ s~ion of endogenous gene pr~-lucls by molecular mPthorle (in~hlrlin~
g of both copies of the e~ G~ ~o~le gene), intro~ cing a heterologous gene, and further ;fi~ ;Qn of the host cell to achieve high level e~ ssion. The particular t~iffiGlllty in p~r~JI~ulg m-lltiple steps like this is the nced for distinct selectable llla,~.~. This is a }imit~tion in tha~ only a few s~lP~ le nlalAcl~ are available for use in m~mm~ n cells and not all of these 20 work sllffir~ lly well for the p~ )ses of this invention.

The present invention th.~,Çol~ co t~S the use of the Cre/Lox site-specifi~
l~CO~ 1 system (Sauer, 1993, available through GibcolBRL, Inc., Ga~ , Md.) to rescue sperific genes out of a ~nomr, most notably drug s~le~tio~ t is c~ "rA as a 25 way ~f ;~ e the number of rounds of en~ ;ng. Briefly, the system involves the use of a b~ l nUcleQti~e se~ e kno~ws as a LoxP site, which is recognized by the b~c~ ti~l Cre protein. The Cre protein C ~lyLCS a site-spccific .cco- .~h;n~l ;on event. Tbis event is bidireclion~i ie., Cre will catalyze the insertion of sequenres at a LoxP site or excise se~u~nces that lie betwecn two LoxP sites. Thus, if a cons~uct co~ ;n;~ a selectable marker also has 30 k~xP sites fl~nlring the selectable marlcer, h~llu~ !;on of the Cre protein, or a polynt-cl~otirl~

encoding the Cre protein, into the cell will cataly~e the removal of the sel~-ct~hle marker. If s~lcceccfillly r . c~J.nplished, this will rnake the sele~hle marker again available for use in fu~her genetic çng;. e~ g of the celI. This technology is explained in detail in U.S. Patent No.
4,959,317, which is hereby h~cOl~o~aled by ~efe.~nce in its entirety.

lt also is co~ pls~ that a series of ~r~,~nt IllaLll~Cl't may be employed in some ~ihl~ nc, These T~ i are ~ sseA in greater detail, below.

B. Proteins A variety of different proteins can be e~,p~ssed according to the present invention.
~oteins can be ~;~uu~e~ g~nt-r~sllly into two C~gOl~S - secreted and non ~ ,~d - liiecllccio~c of each are ~let~ below. There are some general ~lv~.lies of pl~tc,~ls that are worthy of ~;c~-"_c;Qn at this junelu-c.

First, it is co~ .ls~t~d that many proteins will not have a single seq~en~e but, rather, will e~cists in many forms. Thcse fv~ms may l~pl~.sent allelic v~ sttion or, rather, mu~ant forms of a given protein. ~je~n~l it is cont--..pl~ted that various ~uteins may be e~ csscd a~lv~ c,tou~ly as "fusion" plutCi~l~.. Fusions are generated by linking together the coding regions for two proteins, or parts of two ploteins. This g_~lel~-tes a new, single coding region that 20 gives rise to the fusion protein. Fusions may be useful in p.oJ,~ g se~ d forms of proteins that are not nonnsllly sec.~te~ or pl~h-- ;..g mrle,~ c that are ;~ .ologically tagged. Tagged proteins rnay be more e~sily ~ ;fie~l or .... ~ oi~d using ~ibo~lies to ~e tag. A third v~ri~tiQn contemp~ated by the present invention involves thc e"~lession of protein fr~grn~ntc. It may not be l-~C~S~ to express an entire protein and, in some cases, it may be desirable to express a 25 palticlllar f~mCti~n~l ~Inn~in, for e~mpl~. where tne protein fi~ remains fimrtion~l but is more stable, or iess A.~ .;c, or both.

(i) Secreted Proteins .,ssion of several proteins that are normally secreted can be ~n~ r~.~,d into ocrinp cells. The cDNA's enrolling a ~ b~,. of useful human l~lot~,ins are available.
F~ ,lcs would include soluble CD~ Factor VlII, Factor ~, von W~ br~n~l Factor, TPA, 5 urokinase, hirudin, int.,.~rolls, TNF, interle~lkinc, hematopoietic growth factors, antibodies, mtn leptin, transferin and nerv,e growth factors.

Peptide Hormones. Peptide hG~.-lones çl~im~l herein for enginç~ring in neu~u~ n~loc~in~
cells are grouped into three classes with specifi~ ex~mples given for each. These classes are 10 defined by the complexity of their post-tr~ncl~tior-~l procec~in~ Class I is l~ se.lted by Growth ~orm-ne, Prolactin and Parathyroid hn....m~r A more e~t~ e list of human pep~i~ec that are ;~ lu~cd in Class I is given in Table 2. These require relatively limited proteolytic ~)n)c~CS~in~ followed by storage and stim~ ted release from se~,..,to.~ granules. Class ~ is l~,y~,S~,~t~,~ by ~sulin and Glucagon. A more extensive list of human peptide ho....~ s that are 15 in~ in Class II are given in l'able 3. Further proteolytic p.occs~ g is required, with both ~ndo~ s and c~lo~y~.,ytidases ploces~ g of larger ~ or m~ s occ~rin~ in the ~se.,-~,tu~ nl~s Class m is l~p~ re~ by Amylin, Glucagon-like Peptide I and Calcitonin.
Again, a more e~tul~si ~e list of Clas;s m human peptide h.~"~ r~ iS given in Table 4. In ~drlitinn to the proteolytic pr~e~C~ g found in the Class II peptides, ~mi~1~tif~n of the C te.",i",~.~ iS
20 ~c.~ ,d for proper biolo~c~l fim~tiQn FY, , l~s Of eng~r. - ;'~c eA~,.c~ion of all three of these clas;ses of peptide hG~lu~es in a neL.l~ I~Aor-- ;--1~ cell are found in this patent.

CA 0224643l l998-08-l2 W O 97/26321 PCT~US97/00761 TABLE 2 - Class I Human Peptide Hormones Growth Hormone Prolactin pl~rent~l Lactogen T ntçini7ing Hormone Follicle-stim~ ting Hormone Chorionic Gonadotropin Thyroid-stim~ ting Hormnne Leptin TABLE 3 - Human Peptide Hormones ~ oc~ by Endoproteases from Larger Precursors Adrenocorticotropin ~ACT~I) ~n~io~ I and II
do~
,B Melallocyte Stim~ ting 3~Iormone (~-MSH) ~ f -y~lQ}rinin Fnrloth~lin I
n~l~nin Gastric Tnhibitnry Peptide (GIP) n lnsulin Lipotropins Ne~ phy~i~s Sorn~ost~tin TABLE 4 - A~nidated Human Peptide Hormones ~lrillrn Metabdism Pepti~les.
jtQni~
C~lcjtonin Gene related Peptide ~CGRP) ~-C~l~itcnin Gene Rclated Peptide IIy~rcalo~-mi~ of ~ nr~y Factor (1 40) (PTH-rP) rd~ ylo d Hc,lll,GI,c-related protein (107-139) (PTH-rP) r~dlhyl~ Ho...~ u~-related protein (107-111) (PTH-rP) Gastrointestinal Peptides:
Cholecystolcinin (27-33) (CCK) nin ~ a~e Acs~i~ Peptide, Preprogalanin (65-105) Gastrin I
Gastrin R~k-~ Peptide Glucagon~ ce Peptide (GLP-1) Pan~ ,asLalin Pancreatic Peptide Peptide YY
S PHM
Secretin Vasoactive Tntes~in~ l Peptide (VIP) PiluiL~y Peptides:
Oxytocin Vaop~essin (AVP) Vasotocin F.nk~ph~linc:
1 5 Enkeph~lin~mi~
Me~o,~h;~ e (Adrenorphin) Alpha Melanocyle Stimnl~ing HG1111O1-C (alpha-MSH) Atrial N.~ etic Factor (5-28) (ANP) Amylin Amyloid P Cf~ on~l-t (SAP-l) Col~icol~o~ll R~k~cine Hollllolle (CRH) Growth Horm- ne ~le~in~ Factor (GHRH) o~ n~ R~ cin~Hormone (LHRH) Ne~l,c~1i.3e Y
Sul.~ e K (Neurokinin A ) Suh~ ~ P
Th~rlotlu~in Rel~scing Hormone (TRH) W 097/26321 PCT~US97100761 (i) Non-Secreted Proteins Expression of non-secreted proteins can be es~ P~.~d into n~ oc~ine cells. Two general classes of such proteins can be defined. The first are IJIoteins that, once tA~,~ssed in cells, stay associated with the cells in a variety of desrin~tions. These destin~tionc include the 5 cytoplasm, nllrlf~llc, mitoc~londria, endopl~tcmic reticulum, golgi, membrane of secretory granules and plasma llle~ e. Non-secreted proteins arc both soluble and llle~ e ~ssoci~ed. The second class of plot~,hls are ones that are no~nslly ~ccoci~rfd with the cell, but have been mo~lified such that they are now secreted by the cell. Mo~ifir~tions would include site-directed ,sis or expression of truncations of eng.~ d proteins res~llting in their secretion as 10 well as creating novel fusion plol.,h~s that result in secretion of a normally non-secreted protein.

Cells el~,i,lf f ~cJ to produce such ~ cins could be used for either in vitro production of the protein or for in vivo, cell-based therapies. In vitro pl~du~;~ion would entail purification of the .,Ap.ess_~ protein from either the cell pellet for protc.lls r~m~inin~ ~CSoc;~ d with the cell or 15 from the con~iitil~ned media from cclls secreting the en~ d protein. In vivo, cell-based C would either be based on secretion of the en~h~ d protein or be-n~firi:~l effects of the cells e"~..,ssi.lg a non-secreted protein.

The cDNA's enrotlin~ a nt~ bcl of Ihe~ ;r~lly useful human proteins are available.
20 These include cell surface receptors, ~ and ch~nn~lc such as GLUT2, CFl'R, leptin ie~ Jt~, sulfonylurea receptor, ~-cell inward rectifying rh~nn~l~, etc. Other p~O~CillS include protein pl~c~cci~ c~ues such as PC2 and PC3, and PAM, transcli~ n factors such as IPFl, and ~&~ oli~ C~ s such as ~denQci~ dc~ se, phenyl~l~nin~ h~L~yl~, blosidase.
P.~h~P.,;"g mllt~t~, truncated or fusion proteins into neuro~-n~ocrin~ cells also is contemplated. Fy~n~ri~s of each type of r.nE~h~.. ;.~gr~.clllting in secretion of a protein are given (Stoller and Mains, 1989; Ferber et al., 1991). Reviews on the use of such proteins for studying ~he reg~ sd secretion l,~h. ~ are also cited (rlu~ ss and Kelly, 1987; Chavez et al., 1994).

wo 97126321 (ii) Amylin an ~mi-l~t~l, secreted peptide ho~ ol~e Amylin is the ma~or co"~ulh~nt of islet amyloid in the pancreas of p~ti~ntC with non-insulin d~pen~Pn~ rli~ktes nn~llitll~ (NIDDM). Also known as d~ et~s associated peptide (DAP) or islet amyloid polypeptide (IAP), amylin is a 37 amino acid peptide with a 50%
S homology to c~lcitonin gene relat~ed peptide (CGRP). Amylin also has been found in normal islets in insulin~ont~ining se~ ,tuly granules of the ~ cells.

The role of amylin in diabetes remains unknown. However, since ~-cells are proficient amylin secretory cells, and the loss of ~-cell function accompanies insulin ~ iPbctes, it 10 is evident that there is a loss of am~ylin secretion as well as a loss of insulin secretion as a result of II~DM. Arnyiin also has been imrlic~t~d in many other biological procesccs, however, its true roJe in metabolism has been ~iiffirll~t to define.

Two related f~tors have po~nti~lly cnn~rihll~ecl to the confusion ~u"uu-.A;.-g amylin's 15 role in metabolism and participation in disease p~ocesaes. The first of these factors relates to the i~entity of the amylin species available to r~s~,~ch~ for study. The amylin mo~ ule that has becn used for most in vitro st~ es~ and currently is in phase m clinical tnals for the tl~ of NIDDM, is chPmi~lly ~y~1h~si~ rl and ~m~ te~l~ but it is not glycosylated. However, it has been shown from analysis of parlel~alic extracts ~at about 50% of naturaliy occ~ g amylin is 20 O-Jinl~ .,o~ylated (RinPnho -~e, et al., 1996), a m~-rlific~ion that cannot be re-created with ch~ yl~tLesis .~

The second fac~or that has contributed to the mysteries ~lV~ 1;ng amylin's physiolo~c~l role COI.CC...c the failurc to isolate and ch~ c a ,ec~plor that binds amylin 25 with s~ciflrity and high affinity. l~e potential i"t~ JÇ .d~- ~ry of these two factors is evident, i.e., it is possible that ic~ tion of an amylin l~,CptOl has been h&llE~,ed by the failurc tO use its physiologIcaI ligand C~.llly, ~miflq~A amylin is assayed by its hlt~ ons with l~ptOl~ for which it has 30 relatively low affinity. These include the c~lritonin gene-related peptide (CGRP~ receptor and the calcitollin receptors (Pittner e~ al., 1994). .An~ ted amylin binds these receptors with 100 -1000 times lower affinity than their true ligands. The naturally oCcurrin~, glycosylated form of amylin does not bind these receptors (E~ittP.nhouse, et al., 1996). In arlAition, most of the roles that have been ~cribed to arnylin have been r~ e-l from studies in which the doses of amylin 5 giYen were ph~nn~rologic, far excee-~ing norrnal circ~ ting levels of the peptide (McGarry, 1994).

Given the indirect evidence for the ç~i~ten~e of an amylin receptor, several groups have ed to isolate a protein from membranes that has been described as an arnylin-specifi~
10 l~ccp~l. A repon in lg92 ir4..~if.Pd several ~)IOI~;nS in lung m~ es that bound amylin with high affinity (Bhogal et al., 1992). Kinetic studies by ~oth~,l group imr~ tP~ a spe~ific amylin l~icc~ as being prescnt on CHO-Kl Chinese h~ t~.l ovary cell line (D'Santos e~ al., 1992).
High affinity amylin binding sites in rat brain wcre des-,libed in 1993 (Ref ....o~t et aL, 1993).
None of these studies which cl~impA to identify high affinity, spccific amylin r~,ct~lv, ~ have 15 come to fruition as evi~n~ed by the failure to isolate such ~ce~

The procil~cti~n of amylin i~rom ",~.,,,,.~li~ cell lines, in particular those that are derived from pancreatic beta cells, will allow for the large sca}e production of amylin species that f~ithfillly mirnic the ch. ~,.in~l iden~ity of naturally oc.-.~ -g amylins.
Due to the clear correlation ~l-.~e,n a potential role for amylin snd the pathogel~esis of diabetcs, it is clear that there is a need to s~ e amylins from a malnm~ system which will posscss all the post~ ncl~tiQn~l mO~lifir~tioDs of natural amylins. The availability of such arnylins will allow for the elucidation of the natural amylin I~CG~ the role of smylin 2~ s~ n. and the consc~ul,~ces o:f its ~hsf ~ce Further a l~.,~oc~ ocrin~ cell-based system for either in v~tro, biologically active amylin proAuction~ alone or in cornhir~-io~ with insulin or for in vivo, cGII-based deli~.rery of biologically active amylin alone or in cumb.nation with insulin would provide an effective amylin therapy in the L,G-~m~nt of diabetes, hypoglycemia and the restoration of ~cell Ç ~ n In the context of the presen.t invention, the term "amylin" is inren~lt d to refer to a peptide or pcpi~ieS having biological and/or ;.. ~lological identity of the hwmor~e arnylin, or other peptides derived from the arnylin. cDNA for example, as exemrlified by natural}y oCc~ ng amylins such as those found in human rat, or other m~mm~ n speciPs Amylin is a 37 amino acid polypeptide hc.~ one (SEQ ID NO:5 1 ) processed from a larger precursor polypeptide by proteolytic l,locessi~.~ tSanke, et al., 1988) . Amylin is norm~lly co-produced and co-secreted with insulin by beta cells at fairly constant ratios, acting as a hnnnon.o to regulate carbohydrate metabolism (Hoppel1el, Oosterwijk et al., 1994) . Any variation in the se4--~.re of amylin depicted in SEQ ID NO:54) that allows for the biological andlor ~ cl~gical integrity of the armylin peptide to be m~int~in~d is incol~o.~led as part of the present i.l~ ion.

Arnylin is slluclul~lly similar to c~l~ito~in gene related peptide (CGRP) having an identity with human CGRP-2 (U,S Patent No. 5,124,314 inco~yu,dted herein by l~,fcrenc.,).
Cooper et al., showed that arnylin iis a major co~ on~l-t of islet arnyloid (1987). Soon thclcal~.
it was deu~m~ dted the plC.SC.lCe, of amylin in ~-cells, thereby sn~ .g the co-secretion of amylin and insulin for such cells.

The biological effects of amylin are ul~cl~. ~l~,;7~orl, however, a ~ lllber of theories abound as to the biologi~ action of amylin. It has been post~ ted that amyloid deposits result in thc dc~ u,tion of ,B-cells in the Islcts of Langerhans. As these cells are involved in type I
diabetes, it is likely that this type of diabetes is associated with a d~ fi~ c~ in the arnylin se~l~,lion, as well as insulin sec~,tion. r~~ ..nre, amylin mod~ es the rate of glucose stimulated insulin secretion from islet ,B cells (Ahren e~ al.~ Diabetologia 1987; 30:354-359) and amylin reduces the rate of glycogen s~ thesis in both basal and insulin stim~ t.~ modes (US
Pa~tent No. 5,124,314).

~ence, it appears that amylin exerts a po..~,.rul force on insulin-il.lluc~l storage of 30 glucose as gl~cogell. This may well be a ~ c~ icm by which type-2 ~ hetes related insulin W O 97/26321 PCTrUS97/00761 ~~s;~ e occurs. Thus amylin mod~l?~rs and reduces the hypoglycemic effect of insulin both by rednring the release of insulin in re.;.yonse to a glucose stimulus and by red~lcing the rate of glucose storage.

S Several reports have ~1en-nn~t~t~A the down-regulation of insulin secretion by amylin.
This result has been seen in i~ol~çd islets (Ohsawa, et al., 1989; Wang, et al., 1993; Wagoner, et al., 1993), pcrfused whole panc~ascs, and with a RIN cell line (Murakarni, et aL, 1990).
~Ithsl~gh in many of these reports, l~h~....~rological levels of amylin were needed (Ohsawa, et al., 1989; Wang? et al., 1993; Mu~ i, et al., 1990), while in others physiologic levels of amylin (5 - 20 x 10-9 M~ had signifir~nr effects on glucose-stim~ tP~ insulin secretion (Wagoner, e~ al., 1993; Degano, et al., 1993).

A physiologic role for amylin's ability to inhibit or m~ tç insulin secretion was ~v~ ted by in v~vo ~lminictration of amylin antagonists (Young, et al., 1992; Gedulin, et aL, 199~). An enh~nr,eA insulin l~s~ol~e following a glucose çh~ n~e was obs~ ,d. These reports would s~l~st that e~ P~.,;. g higher lcco...hin~ aTnylin levels into an insulin prod~ ng cell linc would result in down-regulation of insulin secretion.

As stated above, amylin is a 37 amino acid polypeptide hol ...one processed from a larger 20 p~ecu~or ~ol~ de by proteolytic ploccc~;-.g (Sanke, Bell et al., 1988~. ATnylin is noml~lly co-~luccd and co se~et~,d with insulin by beta cells at fairly consl~t ratios, acting as a h~ r to regulate carbohydratc metabolism (Ilo~ ,r, Oost~" wijk et al., 1994).

The present in~ on providés the ability to e~neer a range of amylin-to-insulin ratios.
2~ This is an ~ .c~t~d result since thcre are scveral reports d~ lg the ability of amylin to down re~ te insulin in beta cells and beta cell lines, in~ hlAing RIN lines.

Co_~I...,;,sion of insulin and amylin at, colu~ ~ly, any ratio, and at ~ lly ~fr,. ;.~ ~1 levels, has several b~ .- r.l~ in the conte~ct of cell based co-administration. Also novel is ..

the use of a novel e~l ~ssion pl:~mi~1 leading to the effirient co-expression of insulin and amyJin.

F.ngin~ed neuroendocrine cells lcpl~sel~t a novel source for ~ n cell produced 5 amylin and amylin related species lhat are similar to naturally-oc.-~ I- . ;ng form-s seen in plasma of rats and hllmsn~ By producing amylin in lle~l~'Ocn~OCrin~ cells, illli~Gll~t cell-sperific ce,~;.lg steps are provided il~cll~ g dibasic amino acid proteolytic ~,locesci~.g from larger pro~l~rlin p~.,.,UI~o[, proteolytic ~ ;ng by mono-basic amino acid endol.lctc--Es, ~irlA~ion of the gl~hle-extçn~ied amylin pl~;u~or and O-linked glycosylation of amylin at T~h~ol~ c 10 residues in positiQn~ 6 and 9 of SEQ ID NOs: S9 and 62. RIN1046-38 cells have been dem~ ~ed to produce proteins with all of the above moAifi~-io~ FY~rnples cited in this application include insulin (both ra~ and human), glucagon, glucagon-like peptide-I (human wild type GLP-l and site di~d ~ ;), and arnylin.

Glucagon, a 29 amino acid, peptide ho.. ~ e involved in the re.~ tio~ of glucose and fatty acid metabolism (Unger amd Orci et al., 1981), is proteolytically ploC~sse~ from pl~_p ogh~aeoll, a large polypepdde ~l~ulaOl. Expression of the mRNA e~eQtlirg PI~PI'O~;ILI~ ~g~n iS found in a number of cell types, most notably alpha cells of the pa.l.,l. as and L
cells of the in~stine Preproglucagon pOa~ ionql procescing differs in these cell types, 20 giving rise to pl~do~antly glllr~gnn from the aJpha cells and glucagon-like peptides I and II
(GLP-I and Il) from L cells (Mojsov, Heinrirh e~ al., 1986). The reason for this dirr~
c~ in alpha cells and L cells is due to difr~ntial levels of e~ ,ssion of the ~ ,.dop~)teases PC2 and PC3 (Rouille, et al., 1995). The ~,ession of these clldop~t.,ases is known to vaIy in other cell types as well (Day, Schafer et aL. 1992), giving rise to cell specifi~
25 posttranslational ~e~c;-~g of POMC into distinct ho....~ f pe~i~les.

Amylin and GLP- 1 are two peptide h~ s that are ~midq~d in vivo. Alpha-q~ q-tionis now apprcciated as a critical d~t~ n~ for biological activity of a large ~ i~, of pcptide h.. ~ The c,.L~le involvcd in alpha-amidation, peptidylglycine qlrh~ ";~l~ti~g 30 ~ oG~ ~se (PAM) has been well ch~ d at the m~)k ~ r level (reviewed in tEipper W O g7/26321 PCT~US97100761 et al., lg92). ~ltho~lgh there ls only one gene in "~.,.",~lc encoding PAM (Ouafik, et al., 1992), there arc several forms of PAM due to alternative splicing and endoproteolytic ~lOCC5~
~Stoffers, et al., 1989; Stoffers, e~ al., 1991; Eipper, et al., 1992) leading to both IllC~ l~lC
bound and secrcted forms of this enzyme. PAM also is known to be developlll~.ltally re~l~t~d 5 and dirr~.clltially eAylcssed in vivo (Ouafik et al., 1989).

The ill~yoll~ce of alpha-~m~ tion of peptide h~ nes is such that the Incs~"lcc of the co..e- u$..c amidation s~ e (glycine followed by two basic arnino acids, Iysine and/or alg~e) is usually prcdictive of that peptide functioning as a yl~or to a bioactive 10 polypeptide (Cuttitta 1993). Attempt~s at ",""""~ cell ~ro.lu~ion of any ~mi~lr--~ hollllolles ~.lu-~es endo~ ,teolytic cleavage f~om larger p~ ol~, c~l,uAyy~tidase l~.,...";,~g and alpha-ztmi~ n For instance, GLP-1 is a peptide h~ ,r sccreted from the il.tc5~ L cells in ~ s~ .ce to meals with po.l~.rul in~--linotropic effccts (Kl~ullalln and Williams et al., 1987). It is yloccssed from a larger polypeptide pl~u,~ through steps that are very similar to the 15 I~lu~c~ g of amylin.

P~uc~c~ involves the e,l~dloy.ùteases PC2 and PC3 and carboxypeptida~se from the same ~ ~u~ tha~ gluc~,on is })~oduccd from (Mojsov, ~çinri~ et al., 1986) (Rûuille, et al., 1995).
The final biologically active peptide is a mixture of GLP-l 7-37 and GLP-1 7-36 amide, a 20 di~r~ ncc rcsnltirtg from the ~ ,re ~locccc;l~ of the glycine at position 37 to an alpha-a~dated forïn by peptidylglycine alpha-~miA~~in& monooA~ _,.ase (PAM) (Orskov, Bersani et aL, 1989) (Mojsov, Kop~;L~, ski et al., 1990). Both GLP-l 7-37 and GLP-I 7-36 amide are both ~i~lv~,icdlly active in h~lm~nc (Orskov, We,lt~,.~.,n et al., 1993). The rat ;~ u~ cell line used here, RIN 1046-38 has already been sho~,vn to express sllffiri~nt levels of PC2, PC3 and 2~ ca~ t~ ce for c~ P~ plocec~ g of highly e~.~l.,ssed human insulin.

~ OÇ~C~: .g of insulin, GLP~ lr~on, and amylin by PC2, PC3 and c~l~ ces is d~ -~sd. rrr~ RIN 104~38 ~Tn~ on of GLP-l using two specific assays complete}ys~ ;r~rforamidatedGLP-lornon~ ;da~i~1GLP-lis~e~ l-~ed.

W O 97/26321 PCT~US97/00761 RIN cells also are known lo synthPsi7e and secrete glycosylated proteins. Examples of glycosylation inrhl~ps N-linked as well as O-linked glycosylated proteins. Amylin is Icnown to have O-linked glycosylation of specific residues leading to its heterogeneity in vivo (l~itt~ho~se, etal., 1996).

Co-~minictration of amylin and amylin-related species with insulin to ~nim~lc results in novel physiologic effects, in~ in~r enh~n''ed blood glucose lowering effects. Co ~~minic-t~ation can be by injection of p-l-ifiÇ~ recoml~ina-nt amylin and amylin related species formlllatP,d with or in conjunction with insulin. .AI~ ively, co-~mini~tration can be by in vivo cell-based 10 delivery of amylin and arnylin-related species with insulin. Because cell-based delivery of amylin provides the full ~pe~ of post-translational mo-lifir~tio~c of the peptide, biologic effects may be distinct from those achic~led with injection of ~y~-lL~ic mat~ h Co-~ h~ dlion of amylin and amylin related species with insulin can be a-~cQ.~ ichPA over a large range of ratios (arnylin/insulin ratios of .002 to 0.9).
The present invention allows for the production of m~mm~ n cell IJlu~lluc~d recomhinq~lt amylin and amylin-related species. In further embo~imf~ntc~ these arnylin species are used as a l.,~lgelllS for in vitro and in vivo drug testing, bi-~lQgi~l screcns and a reagent for idçntifir~tiQn and i.col~tion of novel receptors for amylin and amylin related peptides. The drugs and arnyiins 20 p-~l~lccd by the present in~_lltio.l may be u~ced in the ~ t of ~ ctes mPllihlc, hy~ c.l ia~ os~ol,orosis, Pagets ~licP~ce, I.~ lcçmi~, obesity, hy~.l~,nsiol~ or any other diSol~r~ arnylin regulation. The mPthn~c of tr~."~."~l-t of such diSOl'~le.~ can be found in US Patent ~be.:, 5527771; 5508260; 5405831; 5376638; 5367052; 5364841; 5321008;
5281581; 5280014; 5266561 the entîre text of each pa~ent being specifically il.colluG.~ted herein 2~ by ~f~ nce.

The present i.. ~.. ,nlioll allows for ~u~ ;on of ".~-.. ~li~n cell ~-oduced r.. CG---~
amylin and amylin relatcd species post-translationally ...ocl;fird Thesc ml)Aifi~a~ionc include dibasic-amino acid ~..o~l~sis, calbo~ cP l.;...,,,;l~g, ~mi~1~tion and glycosyla~ion. There plC,S.,~ iS no alternative ~ecu.--bin~nl produc~ion system for amylin that provides all of these W O 97/26321 ~CT~US97/00761 moAific~tions. Glycosylation. and specific~lly O-linked glycosylation as has been described for naturally occ~ n~ forms of amylin (~ittenhollcc, et al., 1996), is a ~ Aifi~ztion that is not possible by any existing methm1s of producing amylin, including synthetic mPthorlc for prodllring amylin.

O-linked glycosylation of amylin is an expected mollif1~tion, as exempliffed in this patent. Tl~or~ .s at positions 6 and 9 would be the eAyected site for 61~_0sylation, as these are the residues msrlifiP~ in naturally oCcllrring arnylin. O-linked glycosylation can be ht;l~.u~ eous for naturally occuning recombinant proteins as well as recombinantly produced 10proteins (Jenkins, et al., 1996). ]n fact, hete.u6_. eily of leco,l~b.nant proteins can be in part controlled by selection of a~ylu~llate eApl~ssion systems, cllltll~ing Cor~Airion~ etc. (Jenkins et al., 1996). O-linked 61~co~ylation generally refers to the moAifir~tiol~ of serine or l~ueoluuc residues by rdAition of N-acetyl Galactose linked to galactose and r,e.llu~ ic acid resiAu~s 15The present invention further provides n--~tho-lc for the co-exy~ssion of insulin and amylin at conceivably any ratio and at the,~ zlly effi~ient levels. In light of the earlier AiC~ c:on l~ding the role of amylin in dia~tes these m~th9As undcul~ledly have several b~ r~ and possibly novel f~mctior~c in the context of cell based co ~lminictration~ The present invention further provides fûr the use of a novel eAylession pl9cmi~ leading to the efficient co-20 eA~,~,ssion of insulin and amylin.

(iii) Leptin - r."~ ...~.,.. ;i-g Leptin Expression in Cells In another &...h~l;...~.nt of the present invention the eng;n~ d cells may express and o~c~cXpleSS the obesity-~c,soc~ A protein known as leptin. Leptin is a peptide h~ that 25 controls body co..~ ;.~ and is ~lie~,~,d to do so, at least in part, via illt"~lion with vll~zlzn ic receytGl:~ that regulale food intake and body weight. The various isoforms of lcptin ~,C~,~tOl (O~R), i~rl~Ai.~g the long isoform (OB-Rb), are widely ex~ ssed in various tissues, sll~g.~ g that leptin may play an iul~ull~lt role in actions on eA~ l tissues as wcll.

Additional evidence that leptin has non-neural function comes from a report that~xl~aordillaty cl-~nges in body fat are seen in rats made chronically h~ irp~ ir by L~
with an adenuvi~us vector e~.essing the leptin cDNA. Chen et al., Proc. Nat'l Acad. Sci. l~SA
93:14795 (1996). In this report, rats lost all ~liccernihle body fat within 7 days of adellovil.ls 5 inruaioll7 while gnim,glc that were "pair-fed" at the same low rate of food intake as the ,e~lc~ti~ ..ic ,gnim,glc retain more of their body fat. The magnit~ e and rapidity of the lipid depl~ l.ol- sug~st~d the possibility of a direct "ho~...n~.c-to-cell" action by leptin, in ~dditio~ to effects cause through the 5y~ h~ ~;c n~l v- us system.

Chen et al. (1996) also ex~min~d the effects of leptin ove.ux~r~ssion on plasma glll~ose, insulin, plasma triglyccrides and free fatty acid levels. While glucose did not change, both plasma triglycerides and free fatty acids dropped by about 50% in adenoviral-leptin treated snimslc, when compared to controls (Ad-~-gal or saline). These studies now have been cQ~f,..-~d and e,~t~ td with respect to phospholipids. No clear cut Ç~ ,.,5 in phospholipid 15 con~ tion was ob3~ ,d. Howcver, using an in vitro system, it was est-s-~lich~l that reductirnc in triglyceride levels could be acl~,e~red in the ~bsç~e of s~ l.el;l nervous system effects. Studies ~.rol,l,ed to ~L! t .... i-e what pathways are involved in the triglyceride dçpk I ;o~
indic ted that leptin-ilnl.,~ed triglyceride ~pletios~ involves a novel ...~ch~.-ic...c by which triglyccride dis~e~a through l~nlqnred in~~ç~ r triglyceride metabolism~ rather than 20 tluougll more tr, ~lition~) free fatty acid export palhv~ays.

Insulin levels in adenovirus-leptin inf~ted rats d~o~ed even more L~ l;r~lly than the fatty acids, being only about 1/3 of the q~ mt seen in controls. As s~ated above, the glucose levels of these q~nimql$ was nnrrnsl, ho~ ,r. These i-"--lh~ are Co-~ t with e~h~rG~
2~ iDSlllin s~n3iliYIl~ in treated ~nirrl~lc P~l~lca~a were icn~ from ~ ,pli.~ .~ic rats and e~camined for ~-cell fi~nction and ~ holoE;~. The most stnkin~ finding was the complcte absence of insulin s~l~ion in l~ln)nse to either gll~os,e or a~ ne. The morphology a~c Gd normal, and it was ~l~t ....;..Pcl that insulin se~ oll could be reestrb~ following pe.ru~ion of pancreatic tissue in the ~lescllce of free fatty acids. thereby est7l lichin~ an illl~ol~lt role for 30 these molec~ s in ~-cell filn~tjon These studies also inAicate th~t leptin-...FAi~le~ CtiQn of elevated tissue lipid levels will improve ~-cell function, reduce insulin r~C~ nre and help restore rrn~l glucose ho.~,f.o~l~cic in obese individuals.

A further connp~rion bet~een r1i~hetfs and leptin comes from studies with genetir~lly S obese ZDF rats, which contain mutant OB-R genes. The islets of these ~nim~l~ become overlo~ with fat at the time that l~pelglycesI~ia begins. Rec~llse ~ eu~ that reduce islet fat content prevent rli~hetes in Zl:~F rats, it has been proposed that the ~C~ laticln of triglycerides in islets plays a causal role in ~-cell dycfilnction. Thus, the p~dis~osition to ~ hetes in homozygous ZDF rats may reflect the fact that their tissue have been completely 10 '~llnk,~ i7~d" throughout their life and therefore have ~ "~ te~l high levels of TG. In normal rats, this ~rcllmlll~ior~ is ~ tc;d by the action of leptin. It is e~cct~d that any Ihc~ that reduces triglycerides in islets and in the target tissues of insulin will improve ~-cell filnl~tion and reducc insulin reCict?-~re-In h~ .leptine~llic rats, every tissue that was ex~n~in~l was li~op~" ic. Thus, it is spec~ Pd that norrnal non-adipocytes carry a rninute ~uantity of triglyccride, ~ p5 to serve as a rescrve source of fuel in adipocytcs that are depleted of fat by starvation and becollle unable to meet the fuel needs of certain tissues. It is ~u~pcc~e~1 that this triglyceride storage function is closely regulated by leptin. In the obese ZDF rats, this regulatory control is absent, and these 20 puta~ive in~r~e~ r triglycerides r~SEi.~'eS soar to levels of over 1000-times that of h~ ..;c rats.

In light of these observations. the present ~ppli~io~ ,~fol., cnco..~ ses various ,d cells which express leptin in ~ 5 in excess of normal. The m~th~c by which 25 leptin genes may be manipulated and introduced are much the sarne ac for other genes inc Illd~l herein, such as amylin. A ~lcf~ "llbo~ t would involve the use of a viral vector to deliver a le~in ~ ~co~ n~ gene, :for ~x~rnple, an adenoviral vector. This a~l.7-ll may be o~ l in at least two ways. First, in the engj.r~ g of cells to plo.luce certain polypepti~-s in vi~ro, it may be dcsi~able to ~precc high levels of leptin in order to dc,~ lale various 30 cdlular ~ c, i..~l~..lit.g synthesis of certain proteins. Similarly, leptin u~ "sion may W O 97126321 PCT~US97/00761 synergize with cellular functions, resulting in the increased expression of an endogenous or exogenous polypeptide of interest.

Seco~d, it may be desirable to use a leptin-o~,elc~ ssing cell, or a leptin expression 5 construct, such as a leptin-eA~.cssing adenovirus, in an in vivo context. This in~hlc}e~ various "col.lbin~ion" ap~o~chcs to the t~r~ t of disease states such as obesity, hyy~ Jiderni~ and het~c For example, leptin eA~J~,s~ g cell lines may provide for prolonged e~ cssion of leptin in vivo and for high level e~y,~ssion. Prelil..;n~.y results indir~t~ that injection of l~co...hins~ntly produced leptin is less efficacious at achieving weight loss and reduction of lipids.
10 Induction of h~ pl;... ,..i~ using cells lines or eA~ ssion COC~lluCL~ also may find use in rerluring fat content in livcstock just prior to sl~ng)~t~r. Mo,cw~,r, b~ c leptin-in~uce~l weight loss may act through diff~ nt mechanisrns than those ~ ,ntly employed, it may be possible to avoid related side effe ts such as diet-in~ ce~ tc.sic, heart attack and other diet-relatcd sym~tomc. These r~,g;...~-c may in~volve co-~ tion~ of other en~ineered cells, cells eng;..P~d 15 with leptin and at least one other gene or genetic construct (knock-out, ~nti~ence ~ibo~,"lc, etc), co...hi--s-~ion gene therapy or co.~ ;o~ with a drug. The methotlc of delivering such ph~.. 7r~ 1 preparations arc ~escrihed els.~ in this D. Express on of Polypeptides.
l~e amylin gene can be ins.,~t,d into an a~y~Oy~idte e~ ssio.l system. The gene can be ehy~ ~ in any nurnber of diCf~ cnt rcco-~ ,ant DNA e~ sion systems to g~nc.,..e large amounts of the l)ol~ Jti~4 product, which can then be ~ ;fi~d and used to v ~rc~ nimqlc to t antisera with which further studies may be c~
In one ,,.. ho l;.. ,t, amino acid se-~ pnre varia,nts of a polypeptide can be ~ d. These may, for instance, be minor se~l~h- ~e variants of a poly~lXide that arise due to natural ~ tion witbin the population or they rnay be h-)mt>logllPs found in other speri~-s They also may be r~,s tha~ do not occur naturally but that are ~.rl ;~ ly sirnilar that they ru-.-, ;0., similarly 30 and/or elicit an ;~ that cross-reacts with na~ural forms of the polypep~i~e. .Se~e-~re wo s7n6321 PCT/US97/00761 variants can be ~lGp~u~d by standard m~o.tho lc of site-directed mllta~nPcic such as those ~lesç~ibed below in the following section.

Amino ~id sequence variants of a polypeptide can be substitutional, insertional or ~ele~ir.n 5 variants. Deletion variants lack one or more residues of the native protein which are not esse~
for runeLioll or i..~ ogrnir activit~, and are eY~mrlifi~(l by the variants lacking a !~ kl~e seq~ re described above. Another cQmmf)~ type of deletion variant is onc l~king sec.etuly signal se~v~ res or signal se~ .re,c direc~ing a protein to bind to a particular part of a cell. An exarnple of the latter se~l!Je.~ce is the SH2 dom~in~ which induces protein binding to pho~!holyrosine 10 residues.

S~ ;on~l variants typically contain the ~Yf h-~e of one amino acid for another at ûne or more sites within the protein, and may be ~Irc;~ A to modulate one or more plOpC.~liCS of the pol~lide such as .~ilil~ against ~loteolytic cle..~ ,. St~ -lC preferably are C0113~
15 that is, one amino acid is replaced wi~ one of similar shape and charge. Conservative ~ JS~ .C
aIe wdl known in the art and inrl~ule, for eY~mple, the chA~g.-s of: alanine to serine; ~linc to lysine; Al'~p'~aZ5h to gh ";l~r or hic1irline; aspartate to glutamate; ~;r~t~-"C to serine; gh~ c to aspa~gine; glutamate to .~.~c; glycine to proline; hi~itline to ~r~ or ~h~
isoleucine to leucine or valine; leucine to valine or t.cc'eurin~; Iysine to a.~uline; m~thioninr to 20 leucine or isoleucine; phenylalanine to tyrosine~ leucine or ,...,~ o~ ; serine to llu~onille;
llu~onine to serine; ~l~ph~l to tyrosine; l~ illC to ~ tophan or phenyl~lq-nin~; . nd v. line to isoleucine or leucine.

Insertional va~iants include Pusion pr~x~ s such as those used to allow rapid p!- ;r.~ u- of 25 the pol~ ide and also can include hgbrid proteins co-.l~ equ~nre~ from other pl~t~s and l)ol~ s which are hrmr)}o~ues of the pol~ ide. For eYq~nple, an insertional vari nt could include ~lLiolls of the amino acid sequence of the polypeptide ~om one species, t-g~th~or with p~liO~S of the homologous polypeptide from ano~er spec ies Other iuls~lional i~,~ can include those in which r~liti~nql amino acids are intluduced within the coding s~.~ .re of the polypeptide. These typically are smaller insertions than the fusion proteins dese~ above and are e~l forex~mrle intoa~"ot~asecleavagesite.

In one embo~iimPnt, major ~ ,~nir rle~ of the polypeptide are irlentifiçd by an 5 emriri~l ayyl~ach in which portions of the gene el-~o~ g the polypeptide are e~y~ sed in a .c~o---~;--q~lt host, and the resulting proteins tested for their ability to elicit an .. ~ne Icsponse.
For example, PCR can be used to prepare a range of cDNAs enrorlin~ ~p~;~cs lacking ~ccccsi~,~ly longer r.. ~GJ.. ,n~ of the C t~,.. ~.. -c of the protein. The ;.. ~ otL~ re activity of each of these pclJ~idcs then irlpntifips those ~ ,GIIl ~lc or dc.m~ins of the polypeptide that are ccsçnti~l for this 10 activity. Further e~T- ,;..~ nrc in which only a small n~ l~. of amino acids are removed at each itPr~ti~n then allows the lor~ion of the ~nri~nir d~t -...;..~ ~tc of the polypeptide.

Another P.~,horl;"~ for the p~paration of polypeptides accch~ling to the invention is the use of peptide ll~ P.ti~C ~SimPtirs are peptide~ ,;ng mrJ~ lPc that mimic el~ of15 protcin se~o~ sL~uclll,c. See, for r~ rle Joh..col- et al., "Peptide Turn ~;~ I;r~" in BIOTECHNO~OGY AND PHJ4RMACY, rt~u-O ct al., E~s., Char~ . and Hall, New York (1993). The underlying ~ n~ behind the use of pepdde .~.;.... t;~s is that ~e pepdde b~L1.,~1.r of proteins exists chiefly to orient amino acid side chains in such a way as to ~Ci1its~tP molecular interqruQr-~, such as those of antibody and antigen. A peptide mim~tir, iS ç~ tecl to pennit 20 m~ c~lar ;.-t~ jo~.c similar to the natural m~lo~llf~.

S~c~es~rlll spp1i~ ~;nl~c of the peptdde .--;n.. I;c cc~lle~L have thus far focused on mimPti~s of ,B t~ns within proteins, which are known to be highly ~ lt;~ ikely ~-turn aLIu~,lul~ within an pol~tide can be p.~ -~tl d by e~ ,r-based 5~1e~ h~ as ~1;cc!~s~c~ above. Once the 25 c~ q~ t amino acids of the tum are sl- ~ ;nrA, peptide ~-~ s can bc COI~a~lu-,~d to achieve a similar spatial ~ n~ ~;nr of the c~ l f1~ .n. .-1~ of the ~ino acid side chains.

~ l;r~ ;o~ and ch~ s may be made in the ak~ lul~, of a gene md s~ill obtain a fi-~ ' molecu}e ths~ e~ fc a protein or ~ ide with desirable characte,.~Llca. The 30 following is a tiicrt~scjon based upon chS~ ln~ the amino acids of a protein to create an e~uivalent, W O 97J26321 PCT~US97/00761 or even an improved, second~ e.d~ioll mo}ec~3lP The amino acid c~ ~s may be achieved by change the codons of the DNA se~..enre. accc"~ g to the following data.

For eY~mple, certain amino acids may be su~stit-lt~d for other amino acids in a protein 5 sLluelu~_ without appreciable loss of illre,~ e binding capacity with ~LIu~;lu.es such as, for eY~rlP, antigen-binding regions of antibodies or binding sites on substrate mr hPC~llPs Since it is the i~te~d:tive capacity and nanlre of a protein that defines that protein's biological Çu.~ ;o~
activity, certain amino acid ~ .S~ c can be made in a protein sequence, and its underlying DNA coding sec~uen~e, and r..,~ eless obtain a protein with like p,~,~.~ies. It is thus 10 cont~ ...p~ by the inventors that various ~h~ne~Ps may be made in the DNA seq~lenres of genes without a~le~iable loss of their biological utility or activity.

In making such changes, the h~o~dllic index of amino acids may be conci~lPred. The hllpOll;.nCe of the h~ hic amino acid index in c~- ~f~ tel~Li~re biologic function on a 15 protein is generally nntlpr~too~ in the art (Kyte & Doolit~le, 1982).

W O 97/26321 PCT~US97/~0761 TABLE

Amino Acids Codons Alanine Ala A GCA GCC GCG GCU
Cysteine Cys C UGC UGU
Aspar~ic a¢id Asp D GAC GAU
Glutamic acid Glu E GAA GAG
Phenyl~l~nin.o Phe F WC UUU
Glyc~ne Gly G GGA GGC GGG GGU
~ictitlinP. His H CAC CAU
Tcs)~ ine Ile I AUA AUC AW
Lysine Lys K AAA AAG
T P,l~inP Leu L WA WG CUA CUC CUG CW
MedlioI~ine Met M AUG
r Asn N AAC AAU
P~oline Pro P CCA CCC CCG CC13 Gln Q CAA CAG
Arginine Arg R AGA AGG CGA CGC CGG CGU
SeIine Ser S AGCAGU UCA UCC UCG UCU
Tl~ .e Thr T ACA ACC ACG ACU
Valine Val V GUA GUC GUG GW
T~ JIU~ I TIP W UGG
Tyrosine Tyr Y UAC UAU

wo 97/26321 Pcr/uss7/00761 It is accepted that the relative hy~o~athic cl,alac~el of the ~mino acid colltlibL~tes to the secon~ y structure of the res~ nt protein, which in turn defines the il~t~,a~;lion of the protein 5 with other m~)lecllles, for example, e~ yllRs~ ates~ receptors, DNA, antibodies, ~nti~nc, and the like.

Each amino acid has been ~ccigned a l~y~hu~Jathic index on the basis of their hydroi)hobicity and charge cho~a~ ;cs (Kyte & Doolittle, 1982), these ate: Isoleuc ine (+4.5);
10valine (+4.2); leucine (+3.8); phenyl~l~nin~ (+2.8); ~ eine/cystine f+2.5); methionine (+1.9);
alanine (+1.8); glycine (-0.4); ~ co~ c (-0.7); serine (-0.8~; tryptophan (-0.9); tyrosine (-1.3);
proline ~-1.6); hictitlin~ (-3.2); ~ e (-3.5); gl~ .c (-3.5); ~ e (-3.5); asparagine (-3.5); lysine (-3.9); and arginine (-4.S).

15It is known in the art that c:ertain arnino acids may be su~sliluled by other arnino acids having a similar hydropathic index or score and still result in a protein with similar biological activity, l.e., still obtain a biological functionally equivalent protein. In maicing such rh~ngeS~ the n of amino ~ids whose h~u~dll~ic indices are within +2 is p..,ft,ll~,d, those which are within +l are particularly ~l~.lcd, and those within +0.5 are even more particularly preferred.
It is also ~ tood in the art that the ~ .on of like amino acids can be made er~ cly on the basis of h~lfo~hilicity. U.S. Patent 4,554,101, if~co~ dted herein by ".,uce, states that the greatest local average hydroFhilirity of a ylo~ein~ as ~,ov~ d by the u~ ity of its 7l~jA~ nl amino acids~ correlates with a biological IJf ~ ,ly of the protein.
As det~iled in U.S. Patent 4,554,101, the following hydrophilicity values have been z~cci~.od to amino acid reci.l~l~c: arginine (+3.0); lysine (+3.0); as~e (+3.0 + 1); gl~ f.
(+3.0 + 1); senne (+0.3); ab~ ag~t (+0.2); g]~ --n~ (+0.2); glycine (0); llll~ol~ine (~.4);
proline (-O.S + 1); alaninc (-0.5); hi~tiAine *-0.5); .;~ ~e (-1.0); methionine (-1.3); valine (-1.5);
30 leucine (-1.8); isoleucine (-1.8); lylos~e (-2.3); pheny~ ninp~ (-2.5); tryptûphan (-3.4).

W O 97/26321 PCTnUS97/00761 It is understood that an amino acid can be substituted for another having a simila hydrophilicity value and still obtain a biologically equivalent and immlmologically equivalent protein. In such rhQngeC~ the s~ ion of amino acids whose hydrophilicity values are within +~ is plefel,~d, those that are within +1 are particularly preferred, and those within :~0.5 are even S more particularly plef,,.lcid.

As outlined above, am~no acid s~)bstitutions are generally based on the relative similarity of the arnino acid side-chin ~bs~ tc, for e~cample, their hy~llvphobicity, llydl~hilicity, charge, size, and the like. F~mrlQry sllbstitllrions that take various of the foregoing 10 ch~acl~listics into consideration are well known to those of skill in the art and include: a~ c and lysine; ~ tr and asp~late; s¢rine and ~ co~ e; ~IIltQmir~e and asparagine; and valine, leucine and icol~

E. Site-Specific Mutagenesi~s 1~
Site-sperifir, mnt~L~I~F~jc is a techni~ e uscful in the p~a~dtion of individual peptides, or biologically fimrtic~nQl equivalent ~luteills or peptirl~s, through specific mllt~l~n~cis of the underlying DNA. The t~chnit1ue further provides a ready ability to prepare and test sequence ~a i~lt~ lcci~olat~ng one or more of the fo,~go;,~g considerations, by intro~lrin~ one or more 20 n~.rl~ot;de sul"~ .-ce chQng~s into the DNA. Site-specific mllt~.n~sis allows the production of mutants ~ ou~L the use of S~ r oligo~ l&~ e se~u~ ces which encode the DNA se~ nce of the desired mlltQti- n, as well as a s~lffiri~nt .~u,~cr of adjacent nucleotides, to provide a primer se~ ~ee of sl~ nt size and se~e-~e co~leAily to form a s~able duplex on both sides of the del.,o-tion ju~ ioll being lla~ ~d. Typically, a primer of about 17 to 25 ..~cl~oti~es in 25 length is pl~f~,~,d, with about 5 to 10 residues on both sides of the junction of the se~uence being altered.

ln g~n~l, the t~ ,Jc of site-speçifi- mut~gen~cic is well known in the art. As will be appreciatcd, the tCC~ i4u~, typically employs a bacteriophage vector that cxists in both a single 30 st~andW and double strallded form. Typical vectors useful in site di~ted m~lta~ ;C include vectors such as the M13 phage. These phage vcctors are c.. ~--.c;ally available and thcir use is W O 97126321 PCTrUS97/00761 generally well known to those skilled in the art. Double stranded pl~s~i(lc are also routinely employed in site directed mnt~q~n~ic, which e~ s the step of tr~qncferring the gene of interest from a phage to a plasmid.

S In general, site-directed mtltqg~qn~ci~ is performed by first obtaining a single-stranded vector, or m~ltin~ of two strands of a double stranded vector which includes within its seqllenre a DNA se.lue~ce erl~o~1in~ the desired protein. An oli~onllrleoti~e primer bearing the desired mutated sc~luenre is synthetir~lly ~"~,pal~d. This primer is then annea~ed with the single-stranded DNA plC~ on, and subjected lo DNA pol)~ ;..g e~ s such as E coli poly ncrase I
10 Klenow fr~m~nt, in order to complete the synthesis of the mutation-bearing strand. Thus, a he~.o~lu~lex is formed wl~.ein one strand em~o~les the original non-m~lt~e~l sequence and the second strand bears the desired mutation. This heteroduplex vector is then used to transforrn ap~lo~ e cells, such as E. coli cells, and clones are sPIecte~ that include .~cc,~..k;..q~.l vectors bearing the mntqtPA se~lu~ ~f e arrqn~m~nt The p~ Jau~lion of sequence variants of the selected gene using site-directed mlltqg~nP~ic is provided as a means of producing potentially useful species and is not meant to be limiting, as there are other ways in which s~- n~e variants of genes may be obtained. For exarnple, l~co~ ;n~ vectors e~rotlin~ the desired gene may be treated with mu~nic agents, such as 20 hydroAylarrune, to obtain sequenre variants.

F. Genetic Constr~cts and Thelr Deli~ery to Cells I. Forma~on of Geoetic Constructs Also clqim~l in this patent are eYqm~l~s of DNA e~ ession plq~rnirlc deci~Pd to 25 ot)~i,.-,Le ~ inn of the heterologous ~l~t~,ns such as amylin. Th~se include a n~ f. of enh~ ~f~ r~ o~l~ from both viral and .~h~ q1iqn sources that drive c~ ,ssion of the genes of interest in ~ r~inr cells. Flo n.~ kci~Pd to ol,li,.uLe mf~S~nger l?NA stahility and translatahility in ~ oc~ r cells are ~f finçd The conditions for the use of a ~ ber of domillant dn~g sF~ ;nl- markers ~For est-qbli-chin~ l~f ~ nf.l~t, stable r,~ f n~f~ . ;.-r, cell clones W O 97126321 PCTrUS97/00761 ~y~,ssillg the peptide hormones are also provided, as is an element that links expression of the drug selection 1l~~ to e~.cssion of the heterologous polypeptide.

~i) Vector B~cl~h~
Throughout this ~plir~tion, the terrn ''~ .. ,ssion construct" is meant to include any type of genetic construct col~ta;~ a nucleic acid coding for a gene product in which part or all of the nucleic acid cncoding scq.le~re is c~ble of being transcribed. The transcript may bc tr~ncl~tçd into a protein, but it need not be. In certain emhoflimPntC~ e~ ssion incl~des both ~ sclil~lion of a gene and tr~ncl~tiol~ of mRNA into a gene product. In other embo~ t~ e/~p,.,ssion only 10 inrl~des Ll~ c~ ion of the nucleic acid enco-lin~ a gene of interest.

.,f~ cd ~-~ho~ .f-ntC, the nucleic acid enr~oflin~ a gene "lu~;l is under c. ~;rtion~l control of a promoter. A "~ ot.,l" refers to a DNA sequ~ ~re lcco~ .Lcd by the ~yllthetic m~rhinçry of the cell, or introduced synthetic m~ h~c"~, l~uil~,d to initiate the 1~ specific transcription of a gene. The phrase "under transcriptional control" means that the pl~Jlllote~ is in the correct loc~tion and orientation in relation to the nucleic acid to control RNA
polylllc~sc h~ l;on and e~ s~ion of the gene.

The terrn ~lOllll)U, will be used here to refcr to a group of transcrirtior~l control modlllPc 20 that arc clust~,~d around the initiation site for RNA polylllelase II. Much of the thinkin~ about how l~v~not~ are OI~Led derives from analys¢s of severai viral promoters, inr~ in~ those for the HSV thymidine kinase (tk) and SV40 early tr~-C~ on units. Thcse snl~iPs, ~ d by more recent work, have shown that l~r~lllot~ l~ are c~ osed of discIete fu ~ ,l ;o~ modules, each consisting of ~ ly 7-20 bp of DNA, and con~ one or more recognition sites 25 for l.,..~c~ o~ activator or r~ ,cssul proteins.

At least one module in each promoter f. nrl;~ns to position the start site for RNA
~"~csis. The best known example of this is the TATA box, but in some ~rul,lot_l~ lacking a TATA box, such as the y~ulllo~r for the ~ m 1..."l;~ .,.clcolidyl tra~,sfe.ase gene CA 0224643l l998-08-l2 W O 97~26321 ~C~rUS97/00761 and the promoter for the SV40 late genes. a discrete ct~ t overlying thc start site itself helps to fix the place of initiation.

A~ tion ~l promoter elemen-c regulate the frequency of Ll~sc;lil~lional initi ~ion.
S Typically, these ate located in the region 30-110 bp u~h.,anl of the start site, although a nùlllbcl of p~olnotcls have recently been shown to contain functional elçm~ntc dbwll~Llcalll of the start site as well. The sp~i-l~ b.,~ cn prulllot~. el~ frequently is fle~ihle, so that pr~lllo~.
fullci~on is preserved when elemcnts are inverted or moved relative to one another. In the tk ~lulllolcr~ the sp~cing b~t~. ~en promoter cle~ t~i can be increased to 50 bp apart before activity 10 begirls to ~leclin. DepçnAing on the promoter, it appears that individual elP-..- .~s can function eithet co-operatively or indepen~en~ly to activate tr~ncrnption.

The particular promoter that is employed to control the cA~iession of a nucleic acid e ~4rl ~g a particular gene is not believed to be hllpcslL~Il, so long as it is capable of e~ ,ssillg 15 the nucleic acid in the targetcd cell. Thus, where a human cell is targcted, it is ~l~,f~lablc to position the nucleic acid coding region ~dj~ent to and under the control of a pl~l.lol.,r that is capabie of being eA~lc5!~l in a human cell. Generally ~ kin~, such a plullloter might include either a human or viral plolllotel.

In various e,l.ho.1;1~ ts, l~he human cy~cll.~g~lovirus (CMV) i.. lll.PAi~t~ early gene plu~l~r~ the SV40 early plulll~r, the ~ous salcollla virus long tt~rrnin~l repeat, rat insulin pl~l and glyceraldehyde-3-phosphate d~,h~u~cP can be used to obtain high-level c~p~t Cc Q~ of the gcne of interest. The use of other viral or ~ n cellular or h~
phage ~ull~vl~l~ which are well-hlown in the art to achieve .,A},l~,ssion of a gcne of interest is 25 cvntcmplated as well, provided that the levels ûf e~t,lession are s~lffirient for a given p~l,ose.

By employing a ~lVllloter with well-known pro~.Lies, the level and pattern of eA~icssion of the gene plCNluCl following tr~ncrc~-~;on can be opl;--l;,~d Further, selection of a plvlll~tel that is reg~ t~ ir~ e to SpG-;rl~ physiologic signals can permit inAllrihle eA~l~ssion of the 30 gene plvd~ll,l. Tables ~ and 7 list several rle~ /promoters which may he employed, in the W O 97126321 PCTnUS97/00761 context of the present invention, to regulate the expression of the gene of interest. This list is not 3ed to be e~h~crive of all the possible elçln~ontc involved in the promotion of gene e~L~, ssion but, merely, to be exern~rl~ry thereof.

S F.nh~nr~rs were origin~lly aletecte(l as genetic elem~ntc that increased transcription from a ,lo.ll~tel located at a distant position on the same molecule of DNA. This ability to act over a large ~ e had little ~ ce~ in classic studies of proka~yotic ~ ti~n~l regulation.
S~bse~ .,t work showed that regions of DNA with ç~h~nrer activity are c'~ 7~ much like ~lO~ t~ ~. That is, they are coll,~osed of rnany individual cle-..-- -ts, each of which binds to one 10 or more llaus~ tional proteins.

The basic rlis~inction bet~een enh~nt~r~ and promoters is operational. An enh~nrer region as a whole must be able to stim~ t~ .c~ Lion at a distance: this need not be true of a ~lOI~lOt~ region or its component cl,.-~ . On the other hand, a IJ1U1~1OI~I must have one or 15 more el,. ~-~t~ that clircct initiation of RNA sy,lll.esis at a particular site and in a particular o,;.~ t~ , wl~.~s enh~ l~-e.~ lack these spc~ s. I'~..;,l,ot~,r~ and e-nh~uc~.~"~ are often ov~ l,g and contigllsll~ often SG.~ g to have a ver,v similar ~ lar O~gd~ ; I;o~
Below is a list of vira~ ~r~lllote.s, cellular promoters/enhancers and inducible20 ~lo---~Jt~ le~ e.~ that could be used in combination with the nucleic acid en~oAin~ a gene of interest in an ~ cGn;~l-uct (l~able 6 and Table 7). Additionally, any ~mote,/e.-h~nr,~r combination (as per the Eul~aryotic ~,~ote. Data Base EPDB) could also be used to drive F~~~c~ n of ~e gene. ~ otic cells can support c~to~ r transcription from certainb~rh-~ promoters if the a~ ,pliatc bAct~ l pol~ 3e is provided, either as part of the 25 dc,l~ c.~ or as an a~tlitirJn:~l genetic e~l,iession con .I-ucl.

E~HANCER
T..,....~..o~lobulin Heavy Chain ~mm~noglobuliIl Iight Chain T-Ce}l R~ptor HLA DQ a and DQ ,B
,t~lf~,roll Tnt~.rlP~ -2 T--t~ -2 Rcceptor MHC Class II 5 MHC Class Il HLA-DRa ,B-Actin Muscle Crea~ine Kinase Prealbum~n (T~ yl~ n) F.1~.ctace 1 t51ll~t~1;C~
('.o!la~fnase hllmin Gene a-Fetop~ooein ~ lobin ~Globin e-fos c-HA- ras ~sulin Neural Cell ~rlhP5ion M~ lle (NCAM) a~ ti H2B (l~B) Hi~ton~

ENH~MOER
Mouse or Type I Collagen Glucose-Regulated ~u~ins ~GRP94 and GRP78) Rat Growth IIo~ Q~
Human Semm Amyloid A (SAA) Tl~ol~in ][ (IN I) Pla~elet-Denved Growth ~actor Du~k....-~- M--C~ Dystrophy Polyoma Retroviluses P~rillQrn~l Vinls H~p~titi~ B virus Human ~mmllno~ firien~y Virus Cytom~ lo~ims Gibbon Ape ~ o ~ Virus Elem~t Inducer MT II Phorbol Ester (TPA) Heavy metals MMTV (mouse 1.1~.1111.~. y tunnor ~l~lcocol~icoids vin~s) t~,~f~.~un poly(rI)X
poly(rc) Adenovirus S E2 Ela cjun Phorbol Ester (TPA), H202 g~n~cP Phorbol Ester (lPA) Stroqnelysin PL~.l,ol E~ter (TPA), IL,l SV40 Phorbol E~ter (I~A) Munne MX Gene I~te~f~uil, N ~._&.lle Disease Virus GRP78 Gene A23187 a-2-Mac,u~lob.,lin ~6 V~nentin Serum MHC Class I Gene H-2kB I~tu~ft,~uil HSP70 Ela, SV40 Large T Antigen F~ol~f~"l Phorbol Ester-TPA
Tumor Necrosis Factor ~lA
Thy~id S~nulating I~ a Thyroid Iloll.. f~l-Gene Insulin E Box ~ 1JCOSe ~ ~..,f~,.,.,d em~l;--~ of the invention, tne e~lnes5ion constmct comrrices a virus or engineered co~llu~ derived from a viral ee~- ...lr. The ability of certain viruses to enter cells via ù,-mcdiated elldoc~tbsis and to integrate into hûst cell ge~..l.lr and e~ ss viral genes stab}y and l rr.G.e "ly have made them attractive candidates for t~e l~f~. of foreign genes into W O 97/26321 PCT~US97/00761 m~mm~ n cells (Ridgeway, 1988; Nicolas and Rube~ctçin~ 1988; Baichwal and Sugden, 1986;
Temin, 1986). The first viruses used as gene vectors were DNA viruses inr.l~ ing the papovaviruses (simian virus 40, bovine p~rillom~ virus, and polyoma) (Ridgeway, lg88;
Baichwal and Sugden, 1986) and adenoviruses (Ridgeway, 1988; Baichwal and Sugden, 1986).
5 These have a relatively low ca,va~ for foreign DNA sequences and have a rest~cte~ host s,vecllulll. FurthPrm~)re, their oncog,enic potential and cytopathic effects in 1~ .~.iC~ive cells raise safety CO~ . They can accornmo~1~t~ only up to 8 kB of foreign genetic material but can be readily introduced in a variety of cell lines and laboratory ~nim~lc (Nicolas and Rl~henc~ein, 1988; Temin, 1986).
(ii) Other Regulatory cl~ ,..f .-ts Where a cDNA insert is er~ployed, one will typically desire to include a polyadenylation signal to effect proper ,uGlyddellylation of the gene transcript. The naturc of tne pol~adenylation signal is not believed to be crucial to the succ~s~rul practice of the invention, and any such 15 sc~lu~ -re may be employed. The hlve,-tols have employed the human Growth ~nmnrlp and SV40 polyadenylation signals in that they were c~ ient and kno~,vn to filn~tiorl well in the target cells elu~lo~_d. Also col~t~ 1 as an elç~ nt of the e~,v~ssion c~cs~l~e is a terminator. These c~ can servc to Pnh~n~e .~ ss~- levels ~nd to ~ i7~- read through from the c~csettç into other sey~J( n~ ~ c, (iii) Select~hle M~
certain e ..ho.l .". -1s of the invention, the delivery of a nucleic acid in a cell rnay be d in vitro or in vivo by i"~ A;"e a marlcer in the e~ s~ion co.,~llu~;l. The rnarl~er would result in an illentifi~ e change to the ~cf~cte~l cell ~e ~ e eacy i~entification of 25 e~ ,SSiOl-. Usually the in~hlcion of a dlug s~kc1;0l marker aids in cloning and in the sele~,voll of transformants, for ~Y~rle, neomycin, ,ULUO y~, hy~;~uulychl, DHFR, GPT, zeocin and h.~ l AltG~natively, c~ ,es such as herpes simplex virus thymidine kinase (~k) (eulcaryotic) or chlolA..~l~h~ ol ~ cr~ c (CAT) (prokaryotic) may be employed.
c markers also can be employcd. The s~1~Ahle mar~er e.ll~l3~,d is not believed to 30 be illllJVl~, so long as it is capablc of being e~Gssed ~iml~ r4-lsly with the nucleic acid CA 0224643l l998-08-l2 W O 97t26321 PCTrUSg7/00761 enr~or~ g a gene product. Further e~mrles of selectable markers are well known to one of skill in the art.

(iv) ~vlultigert~ constructs and IRES
S In ccrtain embo~ t~ of the inven~ion, the use of internal ribosome binding sites (IRES) elcm~nts are uscd to creale nl~lt-~en~, or polycistronic. m~ssagt~s IRES elem~nt~ are ablc to bypass the ribosome sc~ ;n~ model of 5~nethylated Cap .4~ A...,I translation and begin translation at intcrnal sites (Pelletier and Sonellbe.g, 1988). IRES cle l.FI~Ic from two of the picanovirus family (polio and ellreph~lomyocarditis) have been described 10 (Pelletier and Sonf---bf.~, 1988), as well an IRES from a m~rnm~ n message (Macejak and Sarnow, 1991). IRES elçmf~ntc- can be linked to heterologous open reading frames. Multiple open reading frames can be ll senbe~l together, each iep~ed by an IRES, creatingl~ol~i~ lic .~ C~ s. By virtue of the ~ES cl~ t each open reading frame is ~c~scil~le to ril,oso.,.es for err.~ir.-t tpn~ on Multiple genes can be efflciently c~pl~ssed using a sing}e 15 promoter~enl-~r.rer to tr~nscribe a sin~le message.

Any heterologous open reading frame can be linked to IR~S rh .. .l~i. This incluA.~c gencs for secrcted proteins, multi-subunit proteins, çnroAe~ by inde,p~ t gcnes, intracellular or lu~,u~ bound proteins and sc~ hlc markers. In this way, eA~ i~sion of several IJlOteL~S
20 can be sim~ ro!~sly en~r~c} into a cell with a single constluct and a single selectq-ble marker.

G. In Vivo Delirery snd l'reatment F'~ cds It is p.upose,d that ~ rqcd cells of the present invention may be hlllu-luced into 25 a~mals with certain noeds, such as qni~l~ with insulin~ ppenfipnt diabetes. In the ~ eti~
t~eatment ~pe~tC, ideally cells are en~ r~,~d to achieve glucose dose lei.~oi~ ncss closely .hl;..~ that of islets. However, other cells will also achie~, advantages in accul~l~cc with the in~,~,uLiull. It should be pointed out that the eA~ c of Madsen and co~o~ have shown that irnpl~tation of poorly .I;Lre.~liated rat in~lllinornq cells into animals results in a 30 retmn to a more ~lir~-.,;~ed state, m~-l~Pd by e-nh-q-n~ed insulin ~cl.,Lion in response to W O97/26321 PCT~US97/00761 metabolic fuels (Madsen et al., 1988). These studies suggest that exposure of en~in~çred cell lines to the in vivo milieu mav have some effects on their ~spo~se(s) to secretagogues.

The ~.~;f~llGd methods of ~tlrnini.~tration involve the encapsulation of the erl~ ccl~,d cells 5 in a bioco...p~tihle coqting In this apl"~,ach, the cells are e~ ped in a c~rs.~ coating that p.otec~ the contents from imrr~unological l.,S~OI~ S. One ~f ..l,d ~ p~.~lqtion technique involves el-ra~ qtiQn with qlpj~ polylysine-alginate. C'-qps~ s made employing this tççhni~ e generally have a ~iqme~l~r of ~.o~ ately 1 mm and should contain several hundred cells.
Cells thus may be i~nplqnted using the q,lginql~-polylysine enc~s ~ ion technique of O'Shea and Sun (1986), with ml ~lifir~~ionc~ as later ~iescrihe~ by Frit$chy et al. (1991). The d cells are ~ ed in 1.3% sodium ql~in~ and ~.~r ~ ted by extrusion of drops of the cell/alginate s~ ion ~ 1- a syringe into CaCl2. After several washing steps, the ~lr~lcls are sll~pen~ed in polylysine and r~ cd. The ql~nqt~ witbin the c~q-psllles is then reli~ e~1 by ~ e~cinn in 1 mM EGTA and then rewashed with Krebs hqlqnned salt buffer.

An ~ ,f ~ spproach is ~o seed Amicon fibers with cells of the present invention. The cells becorn~ enm~c~ d in the fibcrs, which are se.~n~, -~le, and are thus protected in a manner similar to the micro enr~ ~ es (Altmall et al., 1986). After successful erl~srsll1ation or fibcr seeAing~ the cells may be i...~ d i~ ~e,iloneally, usually by injection into the f ~1 cavity through a large gauge needle (23 gauge).

A variety of other cl~c~ ation t~chnQlogies have been developed that are applirable to 25 the practicc of the prcsent invcntion (see, e.g., Lacy et al., 1991; Sullivan et al., lg91;
WO 91/10470; WO 91110425; WC) 90115637; WO 90/02580; U.S. Patent 5,011,472; U.S. Patcnt 4,892,538; and WO 89/01967; eacll of the fcJlGgoil2g being incorporated by lef~,;cDce).

Lacy et a~ (1991) cncapsulated rat islets in hollow acrylic fibers and i.l"l,obilized these in 30 alginate h~Lu~l. Pollowing ~ r~l tr~-s~ t~l;on of the encapsulatcd islets into W O 97/26321 PCTrUS97100761 ~i~hetic mice, normoglycemia was l~,~,o.~edly restored. Similar results were also obtained using s~ o~lc imrl~ntc that had an d~y~o~'iately constructed outer surface on the fibers. It is th~,~cfole cnl~t~ . ..pl~tPd that en~ ,d cells of the present invention may also be straightforwardly "tr~ncpl~nt~" into a .. -.. ~1 by similar sl)~e~llA~-eous injection.

Sullivan etal. (1991) reported the devclopment of a biohybrid pc.Çu~cd "artificial pà~c~as~ which e ~ ul~ c islet tissue in a sel~ y po-...~,91 le ll.e.,ll..~c. In these studies, a tubular semi-pe-...f~l~le ~-.~ blane was coiled inside a ~l~ot~li~re holcing to provide a col~ t for the islet cells. Each end of the ,lle.nbl~le was then co..i-e~t~ A to an arterial 10 pol~ t~nllQroethylene ~PI'FE) graft that eYten~ed beyond the housing and joined the device to the vascular system as an arteriovenous shunt. The imrl~ntqrion of such a device contAi~ islet aUografts into ~ e~ i7~;1 dogs was reported to result in the control of fasting glucose levds in 6/10 ~.limql~ Grafts of this type encapsulating çn~ r-cd cells could also be used in ~cordance with the present invention.
~S
The colllpany Cytotlle.,.l~e~l~ics has de-eloped enca~ulation technologies that are now co. --..- .;ially available that will likely be of use in the application of thc present invention. A
~, ~el~lqr device has also been developed by Biohybrid, of Sh~ sl,u.~, Mass., that may have ~pli~t;on to the technology of the present invention.
Impl~ ;ol- ~loying such an Pn~arS~ ion techni~ue are ylcfe~ d for a variety of reasons. For example, ~ansplantation of islets into animal modcls of ~ tes by tnis m~thod has been shown to significantly inC14a31C the period of normal gl~ccnlic control, by prolonging xenograft survival cornpared to u .F~r~y~lated islets (O'Shea and Sun, 1986; r.i~c~ et al., 25 199I). Also, en~~rslllation will prcvent uncontrolled proliferation of clonal cells. ~'~rsnies co~ cells are ;..~ ..t.~d (ayp~ ely 1,000 10,000/animal) ~ y~ onP~lly and blood ks tai~cn daily for ...~u.;~.. ;..p of blood glucose and insulin.

An alternate ap~u~h to enrsr~ r~iQ~l is to simply inject glucose-sensing cells into the 30 scapu~ar region or ~ l cavity of ~ hetic rnice or rats, where these cells are lcpu~l~,d to form tumors (Sato et al., 1962). lmrl~nt~tion by this approach may circumvent ~roble.lls with viability or function, at least for the short term, that may be en~ountPred with the encaps~ tion strategy. This a~r~ach will allow testing of the function of the cells in e~ fn~ nim~lc but obviously is not applicable as a stral:egy for treating human ~ be~s F.n~ g of primary cells isolated from p~tiP~tS is also con~emrl~P(1 as described by Dr. ~chard Mllllig~n and collP~ ,s using l~tlo~ilal vectors for the IJ~scS of introducing foreign genes into bone lua~ w cel:ls (see, e.g., Cone et al., 1984; Danos et al., 1988). The cells of the bone l-~.vw sre derived fr~m a co...~ protS_nitor, known as pluli~otent stem cells, 10 which give rise to a varicty of blood borne cells i~ riin~ e.yLl~yLes~ platelets, lymphocytes, h~",~, and granulocytes. L~ ..tillgly, some of these cells, particularly the macrophages, are capable of se~l~hllg peptides such as tumor nocrosis factor and in-er~ in 1 in l~ once to ~er;r.~ stimuli. There is also evidence that these cells contain gr~nlll~s similar in ~l~u;lu~e to the selletol~ granules of B-cells, although there is no clear evidence that such granules sre 15 coll~tPd and stored inside l.l~r~ha~es as they are in B-cells (Stossel, 1987).

It may ultimately be possible to use the present invention in colnhin~tion with that pn~viously A~sc~ihed by the one of the present in~nt~l~ (U.S. Patent S,427,940, h~oll,ol~ted herein by reference) in a ~ll~m~,r described for clonal cells to e~ .~r IJlh~ cells that ~e,r~,lll, 20 gluco3e-stimulated insulin secretion. This a~luach would co~ ,letely ci~culll~e~t the need for enca~s~ on of cells, since the patient's own bone ~llallU~ cells would be used for the e..c..~ d then re-implanted. These cells would then develop into their difr.,,~ folm (i.e., the ~ropha~,_) and circulate in the blood where they would be able to sense changes in glucose by scc~L~ng insuliD.
2~
Alternatively, it may be ~sir~ 'lle to hlLIoduce genetic constructs to cells in vivo. There are a number of way in which nucleic acids may illllud~ced into cells. Several m~thoAs are u ~ rdbelow.

CA 0224643l l998-08-l2 W 097~26321 PCTrUS97/00761 (i) Adenovirus One of the ylcfc.l~d metho~,s for in vivo delive~ involves the use of an adenovirus eA~lcssion vector. "Adenovirus c~ .,ssion vector" is meant to include those constructs cont~ .g adenovirus scquences s~ffirient to (a) support p~ck~ging of the co~ ucl and (b) to 5 express an ~ntic~n~e polynucleotide that has been cloned therein. In this context, e~ ,ssion does not require that the gene product be synth~ci7P-d The ex~,.,ssion vector comprises a ge~tir~lly ~n~ re,~d form of adenovirus.
Knowledge of the genetic or~ani7~tion or adenv~i,u~, a 36 kB, linear, double-stranded DNA
10 virus, allows sl~l,s~ .on of large pieces of adeno~iial DNA with foreign sequences up to 7 IcB
(Gmnhaus and Horwit_, 19g2). In contr~ to r~t~o~,il.ls, the adenoviral infection of host cells does not result in cl~ so. . .~l integration because, ~ - r ~/ilal DNA can ~ licate in an el),so~
manner ..ilhoul ~ot~ loto~ y. Also, a~lc,no~ uses are st~ u~ y stable, and no g~ rÇ~ ng~m~nt has been ~let~ed after extensive ~mrlific~tif~n. Adello~ s can infect 15 virtually all epith~ cells regardlcss of their cell cycle stage. So far, adenoviral i"r~-i appears to bc linked only to rnild discase such as acute r~s~ to, ~ disease in hllm~nc Adcllo~hus is particularly suitable for use as a gene l,~.sfer vector because of its mid-sized ~ " ease of manipulation. high titer, wide target-cell range and high infectivity. Both 20 ends of the viral ~ v~ contain 100-200 base pair inverted repeats (lTRs), which are cis e~ ts n.~ c~ ~ for viral DNA ~ inn and p~k ~ing. The early (E) and late (L) regions of the ~ Q~ - conta~n d;~f~.~nt trar~crrirtion units that are divided by the onset of viral DNA
replication. The El region (ElA and ElB) el~rode~ ~loteins respon~i~le for the re~ tinn of cr.;ldion of the viral ~ oll~c a~d a few cellular genes. The eA.~lession of the E2 region (E2A
2~ and E2B) results in the ~ I,e~is of the ~lut-inc for viral DNA replir~~ These ~ te;lls are involved in DNA replication, late gene tA~,s,ion and host cell shut-off (Rcnan, 1990). The ~u.l~ls of the late genes, inrl~djng ~e ll~jo,il~ of the viral ca,psid proteins, are eA~rc,ssed only after ~;~,J~;fif~ processing of a sin~le pli~/ ~n~r,rirt issued by the major late promoter ~MLP). The MLP, (located at 1~.8 m.u.) is pslrticularly effirien~ dunng the late phasc of W O 97/26321 PCT~US97100761 infection, and all the mRNAs issued from this promoter possess a 5'-tripartite leader (TP~) seq~ nre which makes them ~l~f~ cd mRNA's for tr~n~l~tion.

In a current system, recombinant adenovirus is genel~led from homologous S l~c~...hi--~tion bet-.een shuttle vector and ~lO~ s vector. Due to the possible l~co...hi,.~;on b~ two proviral vectors, wild-type adenovirus may be g~,n~"dtcd from this process.
Thc~fol~" it is critical to isolate a s:ingle clone of virus from an individual plaque and e~minP its g~ ;t' S~u~lulc.

('.en~r~tion and propagation of the current adenovirus vectors, which are replic~tion .l~r~r,cl~t~ depend on a unique helper cell line, fi~ ted 293, which was l.~u,~ ed from human e.ll~"~onic kidney cells by Ad5 DNA r.~ and COJ1'I;~L~ Y e~.,ss_s El ~ nls (~'Tr~ n et al., 1977). Since the E3 region is ~ csble from the adenovirus &c~o.~.P (Jones and Shenk, 1978), the current adenovirus vectors, with the help of 293 oells, calTy foreign DNA
in dther the El, the D3 or both regions (Graham and Prevec, 1991). In nature, adcnovi,us can pac~age approximately 105% of the wild-type ~ (Ghosh-Choudhury et al., 1987), providing capacity for about 2 extra kB of DNA. Co...hi.~cd with the a~l"u~ ely 5.5 k33 of DNA that is repl~ in the P,l and E3 regions, the m~i.. , c~pacily of the currentadeno~ilus vector is undcr 7.5 kB, or about 15% of the total length of the vector. More than 80%
20 of the ad~"u~i us viral ~ ains in the vector b~L~ e and is the source of vector-borne CylOIO~; -ily. Also, the re~ on d-~f. ~ cy of the E1-delcted virus is i ~D. r ~t~P For eY~mrle, leal~age of viral genc c~p~;SSiOll has been o~s~ ~ with the ~ tly available vectors at high ml~hir~ Ps of ;.~r~-,;..,. (MOI) (~~ ig~n, 1993).

Hclper cell lines may be derived from human cells such as human embryonic kidneycells, muscle cells, hematopoictic cells or other human embryonic -senr!.~ al or epith~
cells. Altematively, the hdper cells may be derived from the cells of other l,~l.llalian spechs that are p~ ;C~;~re for human zd~ ovh,l~. Such cells in~ dP.~ e.g., Vero cells or other monkey e~b,~ ic ...~ ,l,al or epi~hP~ cells. As stated above, the ~lefe"~d helper cell line is 293.

CA 0224643l l998-08-l2 w 097n6321 PCT~US97/00761 Recently, Racher et ~1., (1995) disclosed improved methods for c~ ring 2~3 cells and g adenovirus. In one format, natural cell ag~,ates are grown by inoc~ tin~
individual cells into 1 liter siliconized spinner flasks (Techne, Cambridge, UK) co~ ;t-;ng 100-200 ml of m~rlillm Following stirring at 40 rpm, the cell viability is e;,~ with trypan blue.
S In another format, Fibra-Cel rnicrocarriers (Bibby Sterlin. Stone, UK) (5 g/l) is employed as follows. A cell inoclllllm, fe jvs~ n~1l cl in S ml of ~er~ , is added to the carrier (50 ml) in a 250 ml F.rk.~ r~ flask and left s~ti~n~ry, with occasional agitation. for 1 to 4 h. The ...~A;u~..
is then replaced with 50 ml of fresh ..~ -.. and sh~L-tng initiated. For virus production, cells are allowed to grow to about 80% co~fl-len~e. aftcr which time the m~ m is replaced (to 25%
10 of the final volume) and adenovirus added at an MOI of 0.05. Cultures are left s-~tirn~ry ov~rnight, following which the volume is in.,,~,ased to 100% and sh~lring col- -.~ c~i for another 72h.

Other than the l~uh~ t that the ad-,nv~ s vector be re~ tir~ defective, or at least 15 con~ ionally defective, the nature of the ad~..ovi.,ls vector is not bclie~,_d to be crucial to the svcc~s.r~ r~tice of the invention. The adeno~,u~-s may be of any of the 42 dirr~ t known serotypes or sub~.ou~s A-F. Adenovirus type S of subgr~- l, C is the ~.~,f~ d starting m~t~ti~l in order to obtain the conditional reE~lir~tion-defective adcnovhus vector for use in the prescnt invention. This is because Adenovirus type 5 is a human adenovirus about which a great deal of 20 bio~hr~ r~l and genetic information is known, and it has hi~toric~lly been used for most COuSt~uCIiGllS e."~loy,llg ade.lovi..ls as a vector.

As stated above, the typical vector acc~l~li"g to the prescnt invention is n~
d~fecli~, and will not have an a~ovi.us Ei region. Thus, it will be most convenient to 25 introduce the pol~. ~cle~ e-nro~ ~ the gcne of interest at the pOsilioll from which the E1-coding se-~-,c~-ccs have been l~,.lloved. However, the position of insertion of the consllu~;l within the ad~,~o~ ,s se~ e~ s is not critical to the invention. The polr~lrl~oti~le c ~co~ g the gene of interest may also be inserted in lieu of the dele~ed E3 region in E3 ~ '~emPnt vectors as -.hc~ by E~arlsson et al., (1986) or in the E4 region where a helper cell line or helper vims 30 c~ lc--.~f nl~ the E4 defect.

W O 97126321 PCT~US97/00761 Adenovirus is easy to grow and manipulate ar d exhibits broad host raoge in vi-ro and in vivo. This group of viruses can be ob~aincd in high titers, e.g., 109-10ll plaque-forming units per rnl, and they are highly infective. The life cycle of adenovirus does not require il~t.,~l~tion into 5 the host cell genom~ The forei~ genes delivered by adenovirus vectors are episomal and, the~, have low genotoxicity tc~ host cells. No side effects have been reported in studies of v~ on with wild-type adenovilus (Couch et al., 1963; Top et al., 1971), de~Q~ g their safety and ~ al~uLiC potential as i,n vivo gene ~r vectors.

Adenovirus vectors have been used in eukaryotic gene expression (Levrero et al., 1991;
Gomez-Foix et al., 1992) and vaccine development (Grunhaus and Horwitz, 1992; Graham and Prevec, 1992~. Recently, animal snldies s~l~g~-s~l that recombinant adenovirus could be used for gene therapy (Stratford-Perricaudet and Peni~ ldet. 1991; Stratford-~ .ic~"det et al., 1990;
Rich et al., lg93). Snudies in a~ ;r~"~ ~ing recQ~u~ t adenovirus to ~r~,e~t tissues include trachea in~till~tiorl (Ros~nf~l~l et al., 1991; Ros~nf~lA et al., 1992), muscle injection (Ragot et al., 1993), ~ .hF,.Al intla~,nous inJ~~tion.C (Hen and Gerard, 1993) and stuleot~l;c inoculation into the brain (Le Gal La Salle ct al., 19g3).

(ii) Retroviruses l~e retroviruses are a group of single-str~n~fl RNA viruses ch~ ,~ 1 by an ability to convert their RNA to double-str~n~led DNA in ~nfectf d cells by a process of reverse~ s,~ ion (Coffin, 1990). The re~ tin~ DNA then stably integrates into cellular ch~ oso~.~Ps as a p~ Us and directs ~ thcsis of viral ~l~t~ S. The i~tee;l~ion results in the ret~ ;ol- of the Yiral genc sG~ r-~ces in the lec;f~ r cell and its dGSCc~ . Ihe retroviral ~ r conlalns three genes, gag, pol, and env tnat code for capsid l~lut~,lns, ~ .se C,~ylllC, and envelope co---~ s, ~ ,ly. A sc.~ ce found U~ 1G~11 from the gag gene CO~ ;U~ a signal forp~~~ng of the ~5e~ e into virions. Two long ~f~.l-,;-,al repeat (LTR) s~~ .ccs are present at the 5' and 3' ends of the viral L" ~O~ e These contain strong promoter and enhancer sequ~ es and are also r~u.~d for j~t ~ ;On in the host CGI1 gen(~ (COff1II~ 1990).

In order to construct a retroviral vector, a nucleic acid encoding a gene of interest is inscrted into the viral g~ P in the place of certain viral seq~enres to produce a virus that is replication-defective. In order to produce virions, a pc~~eing cell line co-~t~ ing the gag, pol, and env genes but without the LTR and p~rL-a8ing CO~ )OrCntS is co~stlucted (Mann et al., 5 1983). When a recol..hin~-t plasmid co,.t;~ a cDNA, together with the retroviral LTR and p~rlr~in~ Se4~ CeS is introduceld into this cell line (by c~lci~lm phosph~te ~ ip.l~tion for example), the p~rL-~ing SeqllÇnre allows the RNA ~ el.l-t of the lcco..~h;nP~t pl~Cmi~l to be p~a~d into viral particles, which are then secreted into the culture media (Nicolas and Rul,cn-t~,in, 1988; Temin, 1986; Mann et al., 1983). The media c~ ;.;..;..g the ~.,co,l,billant 10 let~ovi,uses is then collected, optionally col~ccl~aLed~ and used for gcne ~ ls~l. Retroviral vectors are able to infect a bro;ad variety of cell types. However, ink,E;Ialion and stable e~ ssion require the division of host cells (~as~ d et al., 1975).

A novel appr~,ach ~esi~rd to allow speçifi~ ta.~getillg of retrovirus vectors was recently~5 d~ ,lo~,cd based on the clu~ c~l m~ifi~tior of a retrovirus by the ch~ ir~ lition of la ctose cs to the viral envelope. This modification could permit the specific infection of h..p~ yt~s via sialoglyco~,luteiu reccE?tors.

A ~lif~.c"t approach to ~aut,_~lg of lccGl"binant fe~ vi~uses was de~ d in which20 biotinylated ~ntil~o~ s against a retroviral c--v-,lo~ protein and agains~ a speçific cell l~,C~)IOl were used. The antibodies were cwl~lçd via the biotin col,l~one"ls by using streptavidin (Roux et al., 1989). Using allti~oAips against mayor his~oco-..p~tihility complex class I and class II
antigens, they demonstratcd the inf~tion of a variety of human cells that bore those surface A~ , ..c with an ecotropic virus in l~itro (Roux et al., 1989).
Thcre are certain limi~atiorls to the use of le~vvhL~s vectors in all aspects of the present invention. For e~cample, ~e~o~ ls voctors usually integrate into ~ dvlll sites in the cell e,~--n---r This can lead to ills_.lional rn..t~genesis lhl~ the in~c~ Liùn of host genes or tl~oùgh the liO.~ of viral reguhtory sc~ ces that can int~.,~.~, with the fil~tion of fl~nlrin~ genes 30 (Varmus et al., 1981). Another eQ"~ with the use of defective ~~ ilUS vectors is the potential ap~e~,ce of wild-type replication-co.~.yet~lt virus in the F~c~q-~ng cells. This can result from recombination even~:s in which the intact- sequence from the lccol..l-;..~nt virus inserts UpS~hll from the gag, pol, env se~uen~e integrated in the host cell ellulllc. However, new p,q~ging cell lines are now available that should greatly dec,ease the likr~ihood of I~,c~.. ~.in~tion (Mau~.it;~ et al., 1988; Hers~l~rffrr et al., 1990).

(iii) Other Viral Vectors ~ls~ s~ion Constructs Other viral vectors may be ¢mployed as e~Lyll~ssion consllucl~ in the present invention.
Vectors derived from viruses such as vaccinia virus ~Ridgeway, 1988; Baichwal and Sugden, 1986; Coupar et al.. 1988) ade,n~ccociqtetl virus (AAV) (Ridgeway, 1988; Baichwal and Sugden, 1986; ~P.rmr~n~t and Muzycska, 1984) and herpes viruses may be employed. They offer several attractive fedlul~,s for vanous ..,hn....~ n cells (~.ier~ , 1989; Ridgeway, 1988;
Baichwal and Sugden, 1986; Coupar et al., 1988; Horwich et al., 1990).

With the recent recognition of defective h~pqtiti~ B viruses, new insight was gained into the ~l~u~ function relationship of different viral se4u~nces. In vitro studies showed that the virus could retain the ability for helper~ n~ p~r~q~jng and reverse ~ c~ .t;on despite the dele~ion of up to 80% of its ~ (Horwich et al., 1990). This sngges~d that large portions of the ~ o...~ could be replaced with foreign genetic mqtPn~l The hepatotropism . nd persi~te-nre 20 (integration) were par~icularly attr_ctive ~,o~c.~,es for liver-directed gene transfer. Chang et al., ~tly introduced the chlul ~ hc.)irQl accLyl~ sferase (CAT) gene into duck h~.pq,titi~ B virus .e in the place of the poly~ AiG. surface, and pre-surface coding seq~ res It was co-sfeeted with wild-type virus into an avian hepatoma cell Line. Culture media co~ ini~g high titers of the reco---h~ t virus were used to infect ~i~y ~1lJC~ hepatocytes. Stable CAT
2~ gene eAyl~ssio~l was ~te~te~l for at least 24 days after ~ ul;On (Chang et al., 1991).

(iv) Non-viral ~rectors ln order to effect ~ ssion of sense or anticence gene cO~lsLlu.;ls, the e"~l.,ssion Co~sh~ must be dcli._.~d into a cell. This delivery may be ~ l;ch~d in vitro, as in 30 laboratory p~ccdul~s for tran~foll~ g cells lincs, or in vivo or ex vivo, as in the tl~ t of W O 97/26321 PCT~US97/00761 certain disease states. As àescribed above, the preferred m~.h~nicm for delivery is via viral infection where the expression construct is ~n~rs~ ted in an infectious viral particle.

Several non-viral mrthol1s for the transfer of e~ ,ssion constructs into cultured 5 ~ n cells also are co~ ted by the present invention. These include c~lrinm ~hGs~.k~le p.~ ipil;~lion (Graham ;and Van Der Eb, 1973; Chen and Okayama, 1987; Rippe et al., 1990) DEAE-dextran (Gopal, 1985), cle~ opul~lion (Tur-Kaspa et al., 1986; Potter et al., 1984), direct microinjection (Harland and Weintraub, 1985), DNA-loaded liposomes (Nicolau and Scne, 1982; Fraley et al., 1979) and li~oft~ ne-DNA complexes, cell sonication (FechhPim~r et al., 1987), gene boh,ba.d.. ~.lt using high velocity ll,iclù~.ujecliles (Yang et al., 1990), and lec~ ol-mediated !.~ .cr~lio~ (Wu and Wu, 1987; Wu and Wu, 1988). Some of these techni~ues may be s. ~ ~c~ s ~fi~lly adapted for in vivo Ol ex vivo use.

Once the eA~.~ssion consuu~l has been delivered into the cell the nucleic acid en~otlin~
15 thc gene of interest may be positioned and e~ ssed at dir~,.,, t sites. In certain embo~ lc, the nucleic acid cnCo~ g the genc may be stably i.ltcg.~t~d into the g~ ~u...P of the cell. This ,n~zy~;on rnay be in the cognate loc~tion and ~i~n-~~ion via homologous ,ecoi..hin~ion (gene replacement) or it may be i~ ed in a random, non-specific location (gene ~ 1~ .t~tion). In yet further embo~ c, the nucleic acid may be stably ~ A;~ ~d in the cell as a sep~r~te~
20 e~ o~ l se~ t o~ DNA. Such nucleic acid se~l~,nts or "episomes" encode seq-l~nces s ~rl;~k-~ to permit ..~ e and repli~tion in-l~ penA~ of or in s~..clllun~ with the host cell cycle. How the eA~ ssion co~.shu~:l is deli~.,d to a cell and where in the cell the nucleic acid le ~ nsisd~,~ndc.lt on the type of eAl,.ession cons~uct employed.

~ one c.. l)o~ t of the in~,~,,lLion, the expression construct may simply consist of naked lGcc...hi~ant DNA or pl~crni~1c. Transfer of the colls~ cL may be ~,~lll.ed by any of the h~S ~ 'io~f.~ above which physically or cllrmir~lly p-~nn~-~bili7e the cell lll~,.lll~l~e. This is particularly applicable for h~urlsç~,r in vitro but it may be applied to in vivo use ~c well.
Dubensky e~ al., (1984) succes~fi~lly inje~ted pol~u~ ril~l~ DNA in the form of c~lrilm~
30 phosphate ~ s into liver and spleen of adult and ne~. l~ll, mice d~ ~O~ ating active viral replication and acute infection. E,envenisty and Neshif (1986) also demonstrated that direct inl.~,pc,liloneal injection of calcium phosphate-precipitated plasmids results in expression of the transfected genes. It is envisioned ~:hat DNA encoding a gene of interest may also be transferred in a similar manner in vivo and express the gene product.
s Another embodiment of the invention for transferring a naked DNA expression construct into cells may involve particle bomb~d~ ll. This method depends on the abi}ity to accelerate DNA-coated microprojectiles to a high velocity allowing them to pierce cell membranes and enter cells without killing them (Klein et al., 1987). Several devices for accelerating small 10 particles have been developed. One such device relies on a high voltage discharge to generate an electrical current, which in turn provides the motive force (Yang et al., 1990). The microprojectiles used have consisted of biologically inert substances such as t~lng~ten or gold beads.

Sele-cted organs including the liver, skin, and muscle tissue of rats and mice have been bombarded in vivo (Yang et al., 1990; 7~.1enin et al., 1991). This may rcquire surgical exposure of the tissue or cells, to elimin~te any intervening tissue bet~eell the gun and the target organ, i.e., ex vivo treatrn~nt Again, DNA encoding a particular gene may be delivered via this method and still be incol~o~ated by the present invention.
In a further embodiment of the invention, the expression construct may be entrapped in a liposome. ~iposomes are vesicular structures characterized by a phospholipid bilayer membrane and an inner aqueous m~hlm Mult;l~mPIl~r liposomes have multiple }ipid layers s~ ted by aqueous me~ lm They fornn spon,taneously when phospholipids are suspended in an excess of 25 aqueous solution. The lipid col-lpone~ undergo self-rearrangement before the formation of closed structures and entrap water and dissolved solutes bel~weell the lipid bilayers (Ghosh and R~chh~wat, 1991). Also contempl~ted are lipofect~rnine-DNA complexes.

Liposome-mPdi~tPd nucleic acid delivery and expression of foreign DNA in vitro has 30 been very successful. Wong et al., (1980) demonstrated the feasibility of liposome-m~ ted W O 97/26321 PCT~US97/00761 delivery and expression of foreign DNA in culturéd chick embryo, HeLa and hepatoma cells.
Nicolau et al., (1987) accomplished successful liposome-m.odi~ted gene transfer in rats after intravenous in~ection.

S In certain embo~lim~nti of the invention, the liposome may be complexed with a h~m~ggllltin~ting virus (HVJ). This has been shown to facilitate fusion with the cell membrane and promote cell entry of liposome-ent~ps~ t~d DNA (K~n~da et aL, 1989). In other embo-lim.ont.c, the liposome may be complexed or employed in conjunction with nuclear non-histone chromosomal proteins (HMG-l) (Kato et al., 1991). In yet further embo~im~nt~. the liposome may be complexed or employed in conJunction with both HVJ and HMG-l. In that such expression constructs have been ~~1ccessfully employed in ~ sfer and expression of nucleic acid in vilro and in vivo, then they are applicable for the present invention. Where a b~cteri~l ~u~-loter is employed in the DNA construct, it also will be desirablc to include within the liposome an a~propliate bacterial polymerase.
Other expression constructx which can be employed to deliver a nucleic acid encoding a particular gene into cells are receptor-m~ ted delivery vehicles. These take advantage of the selective uptake of macromolecules by receptor-m~ endocytosis in almost all eukaryotic cells. Reca~se of the cell type-specific distribution of various receptors, the delivery can be highly specific (Wu and Wu, l9g3~.

Receptor-m~ ~d gene targeting vehicles generally consist of two co.lll,ol.enl~: a cell ecel)lol-specific ligand and a DN A-bin~1ing agent. Several ligands have been used for receptor-,..ed;~od gene transfer. The most extensively char~teri7çcl ligands are asialoorosomucoid (ASOR) (Wu and Wu, 1987) and transferrin (Wagner et al., 1990). Recently, a synthetic neoglycoprotein, which recognizes the same receptor as ASOR, has been used as a gene delivery vehicle (Ferkol et aL, 1993; Perales et al., 1994) and epi~rm~l growth factor (EGF) has also been used to deliver genes to squamous carcinoma cells (Myers, EPO 0273085).

9~

In other emboriim~ntc, the delivery vehicle may compTice a ligand and a liposome. For e~mrlr, Nicolau et al., (1987) employed lactosyl-cerarnide, a ~ tose-teTmin~l ~ci~lg~n~lioside, incGl~ ted into liposomes and observed an increase in the uptake of the insulin gene by hPp~tocytes Thus, it is feasible that a nucleic acid enroriing a particular gene also may be sperif - ~lly delivered into a cell type such as lung, epith~ l or tumor cells, by any ulllber of l~c~lor-ligand systems with or wilhoul liposomes. For e~? nrle, epider~ l growth factor (EGF) may be used as the l~ccptor for mediated delivery of a nucleic acid e l~ro~;"g a gene in rnany tumor cells that exhibit upregulation of EGF receptor. ~nnose can be used to target the mannose rt,c.,~lor on liver cells. Also, ~ntibofii~c to CD5 (CLL), CD22 (l~- l.hnl-.~), CD25 10 (T-cell lellk~mi~) and MAA (mr~l~Tlom~) can sirnilarly be used as tar~,~Llilg moicties~

In certain e l~bfyi;~f ~c~ gene l,~.sr.,l may more easily be perforrned under ex vivv confiitinn~ Ex vivo gene therapy refcrs to the isolation of cells from an animal, the delivery of a nucleic acid into the cells in vitro, and then the return of the mofiifierl cells back into an animal.
15 This rnay involve the surgical removal of tissue/organs from an animal or the primary culture of cells and tissues. ~ Ol~ et al., U.s. Patent 5,399,346, and Llc~ ted herein in its entirety, osp ex vtvo lLc.~"lic m~tho~i$

H. Bioreactors and Lar8e Scale Cultures The ability to ~oducc biologically active polypeptides is inc~essingly i~ t to the ph~ l industry. Over tihe last decade, advances in biotechnology have led to the~loJ~IiQl- of i..,~ l protcins and factors from b~rtçri~ yeast, insect cells and from rnammalian cell culture. ~ mslian cultures have advantagcs over cultures derived from the less advanced li~u uls in their ability to post-tr~cl~tiQn~lly process co~nrlex protein ~l1U~U1~S
2~ such as ~ A~,It folding and glyco~ylation. NeuroPnriocnnl~ cell types have added unique capacities of e~dc~lut~,olytic cleaving, C-terrnin~l arnidation and regulated secretion.
h~deed, m~nm~lian cell culture is IIOW the ~ f~ d source of a llulll~el of illl~o~ ploleins for use in hurnan and animal m~lirint~" esperi~lly those which are relatively large, cc~ 7~ or W O 97/26321 PCT~US97/00761 Development of m~mm~lizln cell culture for production of ph~nn~euti~lc has been greatly aided by the development in molecular biology of techniques for design and construction of vector systems highly efficient in m~mm~ n cell cultures, a battery of useful selection markers, gene ~tnplifi~tion schemes and a more comprehensive underststn-lin~ of the S biochen~ l and cellular m~h~ni~m~ involved in procuring the final biologically-active molecule from the introduced vector.

However, the traditional selection of cell types for expressing heterologous proteins has generally been limited to the more "common" cell types such as CHO cells, BHK cells, C127 10 cells and myeloma cells. In many cases, these cell types were selected because there was a great deal of preexisting lilelal-lre on the cell type (e.g., "cookbook" methods for transfection of the cells) or the cell was simply being carried in the laboratory at the time the effort was made to express a peptide product. ~re~luelltly, factors which affect the downstream (in this case, beyond the T-75 flask) side of m~nllfstct-trin~ scale-up were not considered before selecting the cell line 15 as the host for the expression system. Also, development of bioreactor systems capable of s~t~t~ining very high density cultures for prolonged periods of time have not lived up to the increasing l1em~nd for in.,l~ased production at lower costs.

The present invention will take advantage of the biochernical and cellular capacities of 20 secretory cells as well as of recently available bioreactor technology. Growing cells accol.lillg to the present invention in a bioreactor allows for large scale production and secretion of complex, fully biologically-active polypeptides into the growth media. By ~l~si~irl~ a defined media with low co~t~nt~ of complex proteins and using a scheme of timed-stin~ tion of the secretion into the media for increased titer, the purification strategy can be greatly simplified, thus lowering 2~ production cost.

(i) Ancho~ cdePendentve~susnon-a~uh~la~ee-dependentcultures.
Animal and human cells can be propagated in vitro in two modes: as non-anchoragedependent cells growing freely in s~ ion throughout the bulk of the culture; or as anchorage-W O 97/26321 PCT~US97/00761 dependent cells requiring ~tt~chm~nt to a solid substrate for their propagation (i.e., a monolayer type of cell growth).

Non-anchorage dependent or sucpencion cultures from contin~louc established cell lines S are the most widely used means of large scale production of cells and cell products. Large scale suspencion culture based on microbial (b~t~ri~l and yeast) ferment~tio~ technology has clear advantages for the manufacturing of mzlmm~ n cell products. The processes are relatively simple to operate and straightforward to scale up. Homogeneous conditions can be provided in the reactor which allows for precise monitoring and control of temperature, dissolved oxygen, 10 and pH, and ensure that represen~tive s~mplçs of the culture can be taken.

However, suspension cultured cells cannot always be used in the pro~ c~ion of biologicals. Suspension cultures aLre still considered to have tumorigenic potential and thus their use as substrates for production put limits on the use of the res~llting products in human and veterinary applications (Pe!ric~ rli~ 1985; Larsson and Litwin, 1987). Viruses propagated in suspe~cion cultures as opposed ltO anchorage-~lepe~ent cultures can so~ s cause rapid changes in viral markers, leading to reduced immllnogenicity (B~ P~ n) 1980). Finally, so,~ s even recombinant cel.[ lines can secrete considerably higher amounts of products when ~n)p~&~r~l as anchorage-dep~-n~lent cultures as compared with the sarne cell line in 20 snepçn.cion (Nilsson and Mosbach, 1987). For these reasons, different types of anchorage-depçn-l~nt cells are used extensively in the production of ~lirrtlel,t biological products.

The current invention includes cells which are anchorage--lepe-n~lent of nature. R~ cells, e.g., are strictly anchc,lage-deper~-1P~t and when grown in sllcpencion~ the cells will attach to 25 each other and grow in chlmrc, eventually suffocating cells in the inner core of each clump as they reach a size that leaves the cc,re cells nncllc~in~hle by the culture conditions. Therefore, an effiriçlt means of large-scale culture of anchorage-dependent cells is needed in order to effectively take advantage of these cells' capacity to secrete heterologous proteins.

W 0 97/26321 PCT~US97/00761 (ii) Reactors and ~,locesses for s~l5~en~
Large scale s~ ion culture of ~ n cultures in stirred tank was undertaken.
The in~ n-~tion and controls for biolea-;Lcs,~ adapted, along with the design of the f~.... n~QrS~ from rela~ed rnicrobial ~pplir~tionc However, acknowledging the inclea~ed dPm~n~
S for col~ n~ or~ control in the slower growing m~mm~ n cultures, illl~o~cd aseptic designs were quickly implementPd. improving ~eperlA~hility of these reartors. Il.s~ t~l;on and controls are basically the same as found in other fe....f lltc,l~ and include ~git~tion~ te...~ ...e, dissolved oxygen, and pH controls. More advallc~d probes and ~lto~n~lyzers for on-line and off-line Illc~uc~ s of turbidity (a function of pa.~icles present), Cal)a~;iLal~Cf (a function of viable 10 cells present), glucoset~actate, carbonate/bicd-l.ollate and car~on dioxide are available.
... cell de-~c;t;rs obt~in~ble in s~pe~ ;on cultures are relatively low at about 24 x 106 cells/ml of ~ ;. ." (which is lesx than 1 mg dry cell weight per ml), well below the n acL~_d in I~ bial f~ ,r..~

Two ~ cion culture reactor designs are most widely used in the industry due to their .c;..~ y and l~..c~ cc of o~e.~on - the stirred reactor and the airlift reactor. The stirred rcactor design has su~c~c~rully been used on a scale of 8000 liter cap~cily for the production of interferon (Phillips et a~., 1985; Mizrahi, 1983). Cells are grown in a ct~inl~-C.C stcel tank with a height-to~ --- te~ ratio of 1:1 to 3:1. The culture is usually mixed with one or more ~git~torC~
20 based on bladed disks or marine propeller p~Jcrnc. Agitator systems offering less shear forces than blades have been r~s~;~d ~git~tion may be driven either directly or indirectly by m~tirqlly coupled drives. Indirect drives reduce the risk of ~ Iic,obial co~ l;nn ~ ugl, seals on stirrer shafts.

The airlift reactor, also initially ~escribe~i for l,liclobial f~ on and later ~ çri for m~mmqli~n culture, relies on a gas strearn to both mix and oxygenate the culture. The gas stream enters a riscr section of the reactor and drives circ~llqtion Gas rlicçngagcs at the culture surface, causing denser liquid ~ce of gas bubbles to travel du~ v~d in the do..~lcoL.I~,. scction of the rcactor. The ~in advantage of this design is the s;~ y and lack of need for ...-~r.

mixing. Typically, the height-to-~ mrser ratio is 10:1. The airlift reactor scales up relatively easy, has good mass l.arlsL. of gasses and ~nr~1teS relatively low shear forces.
Most large-scale sllepencion cultures are operated as batch or fed-batch processes bccau~e 5 they are the most straiglllru.w~d to operate and scale up. However, con~inuous processes based on c-hr-..osl~l or perfusion ~ riple,s are available.

A batch process is a closed system in which a typical growth profile is seen. A ~ag phase is followed by eA~on~n-;~l, s~;~t;on~ y and decline phases. In such a system, the en~ilull~ t is 10 col~ o~lcly ch~ngjng as nutrien~s are depleted and metabolites ~rCllrnlllgte. This makes analysis of factors ;..fl~ ncing~ celLl growth and plud~ivily, and hence cp~ ,ization of the process, a co~ le~ task. Produ~;livil~ of a batch process may be increased by controlled feeding of key ~ to prolong the growth cycle. Such a fed-batch process is still a closed system I~GC~ Se cells, products and waste products are not l~_.l,o~,~,d.
In what is still a closed systcm, perfusion of fresh ,..I~A;,..,, thrwgh the culture can be achic,~cd by re-~ining the cells with a fine mesh spin filter and spil---ing to prevent clogging.
Spin filter ~;ul~s can plu~ce cell ~len~ities of app~ t~ ly S x 107 cells~ml. A true open system and the r lF~-St perfusion process is the che~osts-t in which there is an inflow of ~"1- l;l-."
and an outflow of cells and produc:ts. Culture .. ~i;.. is fed to the re~tor at a p~clet~.. ;.. ~l and cor~; nt rate which ~ t;~ the ~lihltin~ rate of the culn~re at a value less than the .. spe~ifir grow~h rate of the cells (to p..,~e..t w~sh~ul of the cells mass from the reactor). Culture fluid col~ nillE cells and cell products and byproducts is ~ o-~,d at the same rate. These ~c.rused systems are not in cu,~ .cial use for production from l~ n cell 25 culturc.

~ui) Non-perfused at~ s.
Trs~ liti~r~slly, nchoragc-dependent cell cultures are propa~.s-tPd on the bottom of small glass or plastic vessels. The ~csl~ic~ Ulr~e to-volumc ratio offered by rl~CSi~ ~l and tr~ liti~ns~l 30 tc~k..; lucs, suitable for the labolat~l 1r scale, has cIeated a bottl~n~r~ in the pro l~lction of cells W O 97/26321 PCT~USg7~0761 and cell products on a large scale. In an attempt to provide systems that offer large ~ cescible ~ulrLces for cell growth in small culture volume, a number of techniques have been plo~osed;
the roller bottle system, the stack plates propagator, the spiral film bottles, the hollow fiber system, the packed bed, the plate ex~h~n~er system, and the "~ bl~le tubing reel. Since these 5 systems are non-homogeneous in their nature, and are so..,. ti~ s based on multiple pl~cesses, they suffer from the following shollcc.~ - limited potential for scale-up, ~~iff~ tiçs in taking cell s~ ,}es, limited potential for ~ u,ing and controlling the system and difficulty in homoge.~col~ envi~r,...- nt~l con~-1itionc throughout the culture.

Despite these ~I,a~l,acks, a c~ mmonly used process of these systems is the roller bottle.
Being little more than a large, dirr~ ly shaped T-flask, ~.,..l.li~i~y of the system makes it very d~ A~hle and, hence, attractive. Fully ~ . r-~lçd robots are available that can han~lle thousands of ro}ler bottles per day, thus f~ .;n~tir~e the risk of c~ in-~;on and inconci~t~n~y ~cso~i~ted with the otherwise ,~quil~d intense human h~ntlling With frequent media f;h~s~
roller bottle cultures can achieve cell ~ ;es of close to 0.5 x 106 cells/cm2 (uJll~;aponding to 109 cellslbottle or 107 cells/rnl of culture media).

(iv) Cultures on microcarners In an effort to OV~ OIllC the shortcornings of the traditional anchorage-d~ .1c..~ culture 20 ~locesses7 van Wezel (1967) developed the con~ept of tne lllicloc~li~,l cult~ ng systems. ~n this system, cells are propagated on ~e surface of small solid particles s~l,c-~ cd in the growth . -~ 3;----- hy slow agitation. Cells auach to the llli.,~oca";~ and grow gradually to c01~n.~ y of the mic~ surface. In fact, this large scale culture systcm upgrades the ~l;'k-- r~
d~ ~-A~ culture from a single disc plocess to a unit process in which both rn(~-)layer and 25 ,~ ;on culture have becn brought together. Thus, col..hi~ g the n~cess~ry surface for a the cells grow with the advantages of the homog_ncolls ~ e l~ion culture i~ ases production.

The advantages of microcarrier cultures over most other ~lcholL.~_-depen~ent large-scale cultivation ~--- Ihn~l$ are sevcral fold. First, microcarrier cultures offer a high ~ulr~e-to-volume 30 ratio (~ by ch---.E...g the camer col.re~ Lion) which leads to high cell density yields and a potential for obtaining highly col~c~"~la~ed cell products. Cell yields are up to 1-2 x 107 cells/ml when cultures are pr~ ~Ir~l in a perfused reactor mode. Second, cells can be ~u~ e~ in one unit process vessels instead of using many small low-productivity vessels (i.e., flasks or dishes).
This results in far better lltili7~tion and a col~c~ rable saving of culture ~ A;I..~ Moreover, S propag~tior in a single reactor leads to redur~ n in need for facility space and in the nu~l~be~ of handling steps le~luiled per cell, thus reAIlring labor cost and risk of con~ ;on. Third, the well-mixed and homogeneo~ls u~icloc~ r suspension culture makes it possible to mol-itor and control envilon-.-- ~ conditions (e.g, pH, PO2, and cul~cent~lion of ",.,.1;.,", cvlll~ûllcnts)~ thus leading to more reproducible cell propagation and prûd~ recovery. Fourth, it is possible to take 10 a ~ l.,sen~ati~e sample for miclvsco~ic obse.v.,tion, chrmir~l testing, or ellulll.,~alion. Fifth, since UliCIo~ settle out of ~ui~ellsion easily, use of a fed-batch process or harvesting of cdls can be done relatively easily. Sixth, the mode of the ~nrh~r..~ de~n~lPnt culture propagation on the ~~fi.;luc~li~, makes it possible to use this system for other cellular m~ iplll~tions~ such as cell transfer without the use of proteolytic enzymes, cocultivation of cells, 15 transplantation into ~nimq1c, and perfusion of the culture using decanters, cQlllmnc, fl~ i7.~l beds, or hollow fibers for microcarrier ~cl~;nl~- nt Seventh, I.lic.oca.l.e~ cultures are relatively easily scaled up using convel~-ion~l e~ ~;p,-- n~ used for cultivation of Il~iclub;al and animal cells in s..cp~.,.;""

(v) Microenr~r~sulation of m~m~ cells One method which has shown to be particularly useful for c~ ing ~ n cells is ulic~loencapsulation. The ~ .l cells are retained inside a S~.~U~J~ ne?~le lly~ûg~,l membrane. A porous membrane is formed around the cells ~ g the e~h~nge of ~ nt~,gases, and m-.t~bolic ~udu~ls with the bulk medium :iu lu~ding the capsule. Several m~th~s 2~ have been dcv~ ,~d that are genlle, rapid and non-toxic and where the resl~lting ule~ u,e is s -rr- ~ ly porous and strong to sustain the g,u~ g cell mass throughout the term of the culture.
These ~~ ~s are all based on soluble ~lgin~J~ gelled by droplet contact with a c~lr~
cc~ g soll~tir~n Iim (1982) d~s~~ibes cells con~e~ ~ed in an ~f~iu~ately 1% solution of sodillm alginate which are forced through a small orifice, fom~ing dr~Jlc~, and brcaking ~e into an a~t~.u~i,uately 1% c~lr,i~ chloride soh-tion. The d~u~ are then cast in a laycr of W O 97126321 PCTrUS97/00761 polyamino acid that ionically bonds to the surface ~Igin~tP Finally the Algin~te is reliquefied by treating the droplet in a rhPl ting agent to remove the c-~lci-lm ions. Other m~tho-ls use cells in a c~lcillm solution to be dropped into a ~lgin~te solution, thus creating a hollow ~Igin~t~ sphere. A
similar appro~h involves cells in a cllitQs~n solution dropped into ~Igin~te. also creating hollow 5 spheres.

Mi~loe~ .s~ tçd cells are easily propagated in stirred tank reactors and, with beads si~es in the range of 15~1500 ~Lm in ~ mt~-tPr~ are easily l~,t~ined in a perfused reactor using a fine-m~chPd screen. The ratio of capsule volume to total media volume can kept from as dense as 1:2 to 1:10. With int~ r~ r cell densities of up to 108, the effective cell density in the culture is 1-5 x 107.

The advantages of microenr~s~ tion over other p~ucesses include the protection from the deletl,lious effects of shear stresses which occur from sparging and ~t~tion~ the ability to 15 easily retain beads for the ~ulpuse of using ~e.ru~ed systems, scale up is relatively strai~ trulw~ l and the ability to use the beads for i...~ t~ i~n (vi) ~ellused ~ s~,rstems Perfusion refers to continuons flow at a steady rate, through or over a population of cells 20 (of a physiological ...~ tsohltio~). It implies the retention of the cells within the culture unit as opposed to co~t;~ .,c-flow culture which washes the cells out with the wilhdlawll media (e.g., chemostat). The idea of ~.r, Ur)n has been ~cnown since the be~ ...;ng of the centu~y, and has been applied to keep small picces of tissue viable for k~ t~ d~ sc~;cobsc~v--~ol~. The t~hni~ e was initi~ ,d to mimic the cells milieu in vivo where cells are co~ ously suppli4d 25 with blood, lymph, or other body fluids. Without perfusion, cells in culture go through alternating phases of being fed and starved, thus limiting full c~pl~,ssion of their growth and metabolic ~t~ 1 The current u~e of p~- ~.J,ad culture is in lespo~e to the çh~ n~e of ~wing cells at high ~ nciti~s (i.e., 0.1-S x 108 cellslml). In order to i&~sse -lencities beyond 2~ x 106 cells/ml (or 2 x 105 cell.s/cm2), the ".el~..." has to be co~Cl~nlly rep~ e(l with a fresh 30 supply in order to make up for nntlition~l de rclc~ .es and to remove toxic pf~l~Ct~. Pe,ruSiOll CA 0224643l l998-08-l2 WO 97/26321 PCTrus97/00761 allows for a far better control of the culture en~/ho~ t (pH, PO2. nutricnt levels, etc.) and is a means of cig.,;r,. A~-ly i-cl~,asing the utili7~ion of the surface area within a culture for cell i~tt~k.. 1 Microcarrier and microenri~rs~ t~i cultures are readily adapted to perfused reactors but, as noted above, these culture methods lack the cal~acily to meet the ~hPrni~nd for cell dP~ncitips above 108 cells/ml. Such dPr~iti~oc will provide for the advantage of high product titer in the "iilllll (f~t~iliti~ting do~lla~ ll p~es~ ), a smaller culture system (lowering facility needs), and a bctter ",~"1;1,." utiii7~sjon (yielding savings in serum and other ex~llsi~re additives).
S~ o.ling cells at high density requires extremely efficies!t perfusion techniques to prevent the develop...enl of non-homogeneity. This means the use of highly sophictic~tpd ,v.ocedu.~s and ~pa,d~i and has, until rccently, been confined to a relatively small scale.

(vii) CelliGenrM bi(,..,~lor system The developm.sn~ of a p~l~d packed-bed reactor using a bed matrix of a non-wovenfabric has provided a means for ...~;nt~ ing a ~.r..c;O" culture at ~len~ s exceeding 108 cells/ml of the bed volume ('CelliGenTM, New ~ s~ick Scientifir, Edison, NJ; Wang e~ al., 1992; Wang et al., 1993; Wang et al., 1994). Briefly Aesr, ;becl this reactor comprises an ,d reactor for cnltllring of both allchol..ge- and non-al~cho~ e-depende~t cells. The 20 reactor is ~lc~i~-PA as a paclced bed with a means to provide int~l recirc~ ion P~cf~ bly, a fiber matrix carrier is placed in a basket within the reactor vessel. A top and bottom portion of the basket has holes, allowing the ~..rJ;~ tO flow through the basket. A speci~lly dÇsi~f~d intrçll~r provides lc~ ion of the ~,r.1;~ , lhl~ugh the sp~e occupied by the fiber matrix for ~c.clmn~ a ullirullll supply of ..u~ nt and thc removal of wastes. This cimll~ u~ cly assures 2~ that a negligible amount of the total cell mass is s~lcpen~leA in the ~-~PA; ~ The co,..hi~. l ;OI of the basket and the recirc ll~ticn also provides a l,uWle-free flow of 0~ tyl 1..~..l;...,, 11,,~,~, the fiber ma~ix. The fiber matrix is a non-woven fabric having a "poreN ~ m~ter of from 10 ~n to 100 ~m, providing for a high i~-e-n~l volume with pore volumes cull~s~onding to 1 to 20 times the volumes of individual cells.

CA 0224643l l998-08-l2 W O 97/26321 PCTnUS97/00761 In col~lyalison to other culhlring systems, this approach offers several si~ific~nt advantages. With a fiber matrix carrier, the cells are ~lotected against ~ rh~...c~1 stress from agitation and foalrung. The free medium flow through the basket provides the cells with O~)tilllUlll regulated levels of oxygen, pH, and n~ ; Products can be continuously removed S from the culture and the harvested products are free of cells and produced in low-protein m~rli11m which fa~iliti~tec subse.lu~,nl pllrifi-~~tinn steps. Also, the unique design of this reactor system offers an easier way to scale up the reactor. Currently, sizes up to 30 liter are available. One hundred liter and 300 liter versions are in development and theoretical c~ tiorlc support up to a 1000 liter reactor. This technology is expl~in~(i in detail in WO 94117178 (August 4, 1994, o r.ee~ et al.), which is hereby h~l~lated by reference in its entirety.

A number of clllhlring y~;~ t,~ . ~, used in conjunction the CelliGenTM system, have been ~e~ u-cl~dted to play a role in increased pro-h1ctiQr For c~arnr1~, the CelliGenTM Plus reactor system, in~ ing the use of non-woven polyester fiber matrix (~"~f~.ably, Fibra-CelTM) and 15 centrifugal lift impell~r (yl~fe~ably~ Fibra-CelTM) are system colllyol~ lls that give hl~ ved yields. Also, several media form~ tions have been employed with illlp.oved t)~ro~ nl~e. For example, use of serum free ~r.1;~ is ~l~r~ d, as is the use of chol~st~ol rich lipid extract (0.01% to 0.10%, volurne to volumc), ascorbic acid (from bct~ een about 0.001 to 0.100 mM), ~ J~ t~ (rather than 2 mM ~ ...n~) at 2 to 20 mM, ~re~.ably 4 mM, alpha keto~lut~ate (rather than 2 mM gl~ .e) at 2 to 20 mM, preferably 4 mM, and the ~se~re of growth ollloting factors.

viii) CellCubeTM bioreact ~r sYstem The Cell~ubelM (cqrr~ing-costar) module provides a large styrenic surface area for the immobi1i7~tio~ and growth of substrate ~ ~h~ cells. It is an integplly en~a~ ted sterile single-use device that has a series of parallel culture plates joined to create thin, sealed laminar flow spaces ~t~ __n ~d,~ce~t plates. The Ce11r~lM module has inlet and outlet ports that are ~!i~ol~ally opposile each other and help ~l;c~ibut~ the flow of media to the parallel plates. The ~I~r l;----- iS CO~ r recirculated from the module through an oxygena~or and back to the cube.
The eYt~ oAy~natoi provides a bubble free stream of ~~ ~l "w ~ and allows for the W O 97/26321 PCTnUS97/00761 ~ditinn~l control of the pH of the l,.r.1....,1 With concurrent addition of fresh me~ m, mP~linm with secfet._d p-udu~;l and wastes can be harvested continnoucly, ret~inin~ the cell population in the cube.

During the first few days of growth the culture is gçn-~r~lly s~ticfie~l by the media co~ lt~d within the system after initial see-iine The ~ml-unt of time b~,c~ the initial seeding and the start of the media perfusion is dependent on the density of cells in the see~1ing inoculum and the cell growth rate. The mea~ "lcnt of n~ nt conc~.ll-~tion in the circ~ rine media is a good inrljr~tor of the status of the culture. When est~lichin~ a procedure it may be ~-ecec~ f to morlitor the nl~trientc composition at a variety of different perfusion ratcs to det.,~ ne the most eco~t....ir~l andproductiveop.,,~ g~ "~t.~

Cells within the system reach a higher density of solution (cells/ml) than in tr~iti~n~l culture systems. Many typically used basal media are ~3eCi~ç~l to support 1-2 x 106 cells/mllday.
A typical CellCubeTM run with an 21 000 cm2 surface, cont~-ls approximately 1.2 liters of media within the m~dll4 The final cell density can çyree(lc 2.5 x 106 cell~cm2 or ~ x 107 cells/ml in the culture vessel. At co,.fl~ re depending on the cell line used, media le4ui~d can vary a~ wl,c.e form 4-16 module volumes per day.

The advantage of the CellCubeTM system is that it to a large extent replir~t~s the con~itior~ the cells ~,;r -ce in T flask culture. This allows for very linear scale up of any culture that is ~ cce~ lly grown in flask culture without severe loss in per-cell ~r~ e.

1. Purification of Prote}ns Protcin purification te~hni~ s are well known to those of skill in the art. These t~ ves tcnd to involvc the f~ ;ol~A1;~n of the cellular milieu to se~leA the amylin form other co~ .o~ of the ~ ulc. Having separated amylin from the other plasma cn~.pu~u ~
the amylin sample may be p~ifi~l using cl~ 0f~raphic and elec~ ,h(~le~ic te~-hni~lues to achieve complete purification. Analytical m~h ~5 particularly suited to tbe preparation of a pure peptide are ion~cchange chromatography, P~elucion cbromatography; polyacrylamide gel W O 97/26321 PCT~US97/00761 ele~lluphulesis; isolectric focusing. A particularly efficient method of ~ui~ g peptides is fast protein liquid chrolnatography or even HPLC.

The present invention isolates arnylin from cells c~ t~ amylin by pl~ing S acidlethanol extracts of whole cells or conditioned media and analy_ing the extracts by HPLC as described (~qlhqn et al., 1986, Si70rP~ko and ~ql~qn, 1991). Solvent systems, gradients and flow rates used were as des.;~ cd by Halban, et al., (1986) ho~ _~re~ it is well within the skill of the O~di~ person ion the art to adapt the c~ull~ography conditions to suit individual need.
S~dalds may be used to obtain O~ io~ of chromatography co~ tion~ and mPtho~$
Cert. in aspects of the present invention col rel 1I the pnrifirqtion~ and in particular e,~hO~ t~, the.,ub.,l~ ~t;~l purifilcation,ofanPnrocleclproteinorpeptide. Theterm"p~l-fie~l protein or peptide " as used herein, is il-t~ -~ed to refer to a cG~Ilpoc;l;nn~ isolatable from other co.npol f -t~, whe,cin the protein or peptide is punifiecl to any degree relative to its naturally-15 obti~in~ e state, i.e., in this case, relative to its purity within a k. t,~ e or ~-cell extract. A
plmfiç~ protein or peptide ~h~"~fo~c also refers to a protein or pepti~P~ free from the cn~irol,lllellt in which it may naturally occur.

~enPr~ily, ~purified" wil] refer to a protein or peptide colllposilion that has been 20 s~ll,je~t~d to rlac~iona~ion to remove various other colll~nents, and which colllposilion ~ul~ .t;~lly retains its CAP,~SSCd biological activity. Whcre the term "subst~nti~lly purified" is used, this designation will refer tû a co...l.os;l;nn in which the protein or peptide forms the major co...pQn~nt of the co~ Qs.l;on, such as cc~ .g about 50% or morc of the p~ote;.ls in the cn..~ ;l;on.
Various mptho~lc for 4ual~ifj~llg the degree of purifir~~iorl of the protein or peptide will be known to those of skill in the art in light of the present riirrlG,-~-e~ These inclu~lp~ for P~ nrle, ~l~,t~ g ~he specific activity of an active fraction, or ~ses~ the llulu~r of ~o~ îitlec within a fraction by SDSIPAGE analysis. A pl~,fe,l~,d method for ~essin~ the 30 purity of a fraction is to c~ t~ the sperifir, activity of the fraction, to COlllp&c it to the specific W O 97t26321 PCT~US97/00761 activity of the initial extract, and to thus calculate the degree of purity, herein AACcecce~ by a "-fold pnrifiration number". The actual units used to lep,c,scnt the amount of activit~y will, of course, be dependent upon the parricular assay technique chosen to follow the pllnfic~tion and whether or not the eA~lcssed protein or peptide e~chibits a ~letectAhle activity.

Various techniqnes suitable for use in protcin pllrifir~tion will be well known to those of skill in the art. These inrh~, for exAAmple. pl~ei~ tion with ,mm~nillm s~lphq~te7 PEG, antibodies and the like or by heat denaturation, followed by cenhirugation; ch~ û~a~lly steps such as ion exch~ ngP, gel filtration, reverse phase, hydroxylapatite and affinity ch.ol~lalo~la~hy;
10 isoelecrric focusing; gel ele~ uph~l~,sis; and comhinAtions of such and other t~hni~ os As is generally known in the art, itiS believed that the order of con~h~cting the various pnrifir,ltion steps may be ch ~ng~ ~ or that certain steps may be omitted, and still result in a suitable method for the ~ on of a a~ ly purified protein or peptide.

There is no general ~qu"., Lcnt that the protein or peptide always be provided in their most purified state. Indeed, it is col~t~ that less ~..b~ t~ y purified l~,udu~ls will have utility in certain e-l.ho~;...r,ltc Partial pllrifir~tio~ may be accû.l.l~lich~l by using fewer pu~ification steps in co...hi~ ion, or by ~tili7ing ~lif~nt forms of the same general p~rific~tion seh~m~ For ex~n-rle it is appreciated that a cation-exchange column cl~ulllalography 20 ~ .,~ed lltili7ine an HPLC ~pal~U5 will generally result in a greater -fold punfic~Ation than the same technique ntili7ing a low p,~ssu,., chromatography system. Methods exhibiting a lower degree of relative p...;rl~ ~io~ may have advantages in total ~eco~ of protein ~ducl, or in the activit,v of an e~ s~d protein.

It is known that the migrAti~ n of a polypeptide can vary, somPtim~ s si~nific~ntly, with di~f~ L conditions of SDS/PAGE IC~PA1-1j et ~11., Biochem. Biophys. Res. Conun., 76:42~, 1977). It wiIl ~ fo~ be appreciated that under differing tie~ u~holesis cQn~1itirm~ the p~t ~cn~ ,;~ of ~ d or parti~lly purified e~-,.,SiOn pl~lu~ may vary.

High Pelro~ nre Liquidi Chromatography (HPLC) is chal~lized by a very rapid separation with e~ ao,.lmary r~cohltiQn of peaks. This is achieved by the use of very fine p~licles and high ~ s~.~le to m~int~in and ~eql~te flow rate. Separation can be acco"ll)lished in a matter of ...;..~ s, or a most an hour. Moreover, only a ve~y small volurne of the sample is S needed because the particles are so small and close-packed that the void volume is a very small Çl~.ùn of the bed volume. Also, the cG~ n of the sample need not be very great bcc ~e the bands are so narrow that there is very little dilution of the sample.

Gel chromatography, or molec~ r sieve chromatography, is a special type of partition 10 c~ulll~tography that is based on molecular size. The theory behind gel chromatography is that the colnmn, which is prepared with tiny particles of an inert ~.lbs~ e that contain small pores, separates larger mole.c~les ~rom smaller molçclllee as they pass through or around the pores, de~..~ g on their size. As long as the m~t~ri~l of which the particles are made does not adsorb the ~Ir~ eS~ the sole factor ~let~ g rate of flow is the size. Hence, n~nl,xulçs are eluted 15 from the column in dccl~asing size, so long as the shape is relatively conctS-lt Gel chromatography is ull;.ull~assed for se~d~illg mole~ s of dirr,~nt size because separation is h~nd~,nt of all other factors such as pH, ionic strength, temperature, etc. There also is virtually no adsol~ion, less zone s~.~ling and the elution volume is related in a simple matter to~nole lll~rweight.
Affinity Chromatography is a c~umato~l~hic plocedule that relies on the specificaffinity b~ .~Q a substance to be isolated and a mol~rl-le that it can spr~-;r.~lly bind to. This is a ~cptùr-ligand type illt~ ion. The column material is s~ c~cd by covalently co~ .g one of the binding y~l~ to an insoluble matrix. The colurnn m~t~ l iS then able to 25 spc~;f.~ ~lly adsorb the s ~b~ e from the solution. Elution occurs by ch~ g;ng the cor.rlitiortc to those in wbich binding will not occur (alter pH, ionic strength, t~,lllpe~a~Ule, e~c.).

A particular type of affmity chrom ~tography useful in the pllrifi~tion of carbohydrate cont~ ;n~ c~ oullds is lectin affinity chromatography. Lectins are a class of :, ~bsl~ es that 30 bind to a variety of polys~crh~ s and gl~GI)lot~llls. Lectins are usually coupled to agalose by cyanogen bromide. Conconavalin A coupled to Scpharose was the first m~ eri~l of this sort to be used and has been widely used in the isolation of poly.~-cl-~ ides and gl~;ol,-oLeins other lectins that have been include lentil lectin, wheat germ ~g~lntinin which has been useful in the p-l ifi~qtion of N-acetyl ~II.cos,..l.".yl residues and ~elix pomatia lectin. ~ectins th~mcelves are 5 purified using affinity chromatography with carbohydrate li~f~c I ~-tose has been used to purify lectins from castor bean and ~c~ ; m~l~se has been useful in ei-~ac~illg lectins from lentils and jack bean; N-acetyl-D g~l~rtos~min~ is used for purifying lectins from soyl~ ; N-acetyl gllleos~minyl binds to lectins from wheat germ; D-galactosamine has been used in oblai~ g lectins from clams and L-fucose will bind to lectins from lotus.
The matrix should be a ~l,s~ re that itself does not adsorb rnol~_u~Ps to any signifir~nt extent and that has a broad range of chpmic~l~ physical and thermal stability. The ligand should be collple~ in such a way as to not affect its binding lnc~cllies. The ljgand should also provide re}atively tight binding. And it should be possible to elute the ~ re without destroying the 15 sample or the ligand. One of the most common forms of affinity chl~".alography is ;.I.... ~ ~o~ffinity chromatG~.I,hy. The g~ Lion of antibodies that would be sllh~le for use in accord with the present invention is ~ lsce~i below.

J. I-ol~ Receptors That Bind Amylin Species With High Aff~nity And Specificity.

The present hl~_~on also OllllilleS the use of amylin species of the present invention in isolating amylin-.ccc~to~. The n~th~c for iCo~ n~ the endogenous amylin ~ nol~ will g~nP~lly involve:
1. ~-;r.~ .n of arnylin species from ccll-source and develop assays that /1i~tin~ich among various &~;~vi~ics or chPJnir~l id~P-n~iti~s These m~th~s are set out else~ in the ~CC ;rr~;n.~
2. Radio-1 ~li-tg amylin species of interest and using the l~hP.ll~d species to isolate or screen for a sperifir l~lJt~

P~,.)t~in purification made feasible by amylin-receptor interactions in cell Iysates:
Protein p~lrifir~ion techniques are well within the skill of the oldin~y person skilled in the art and are described elsG~ ., in the speciflr~ion These techniques may well involve ç~
cross-linking and protein pn~ific~tion~ cross-linking or non-covalent il-t~ on~ of amylin and S its receptor followed by co-immunG~ itdlion of the amylin bound-receptor c~ Once a protein has been pulified, its amino acid sequence is ~etel...;,led and the cDNA isolated. This will far~ilit~te the ploduclion of cDNA libraries for eAp,~,ssion cloning and homology SC~,.,l)lng.

Expression cloning: In eA~ ssion cloning it will be desirable to identify a cell type that 10 binds amylin species of interest. This will generally be a l~c~plor-rich cell. cDNA from rccc~lol-rich cell type is then made by lec-h~ )es well known to those of skill in the art. the cDNA is then ~ s~lcd into a ~ n ctll type that does not bind amylin or ~ srollll E. coli. In either case, the use of DNA vector sy~lcllls d~s~ .~hc,~l above that provide for protein expression will allow sc~eening for cells that switch from a non-binding to a amylin-binding phenotype.
15 These types of SCl~. lul.g studies will be useful in isolating the cDNA that confers the phel~ c switch.

Homology scre~~ing: Amylin belongs to a family of related peptides and pl~ hly so does its l'~,CCy)lOl. In order to perforrn homology s¢çecning studies of the amylin binding l~,ccpl~r 20 it is necessary to identify a cell type that binds amylin species of interest with high affinity. ~t will then be possible to make cDNA from these cells and screen the cDNA pop~ tion with low-B~ .Y PCR that employs various oligo~ le pairs for ~mrlifir~ on of DNA that iscons~ d across family ~...,..k.:i. The DNA amplified by this pl~lule is sey~f~e~l and translated into protein for detection of novel types of rec~ptols. Altcrnatively, cons~ d regions 2~ of .ece~ r family 1l.- l~h . ~ could be used in low ~ y~ hybri-li7~iqn screens with cDNA or O~ DNA from cells that contain amylin ~CCG~t~ls.

K ~ ~I ~ration of Antibodies Specific for Amylin ~ 3t~inC
The protein species of the present i~.~.,n~ , for examplc amylin, can be used to l~,o~luce 30 novel a~ odies for use in ~~.--"; no~;ays.

W O97/26321 PCT~Us97/00761 For some u~ c, it will be desircd to produce antibodies that bind with high specificity to the protein product(s) generated by the present invention. Means for pl~c~a.ing and rk- -~t~ g antibodies are well known in the art (See, e.g., Antibodies: A Laboldtol~ Manual, 5 Cold Spring Harbor Laboratory, 1988; incol~,ol~ted herein by l~,f~,lcnce).

Methods for g~cl~ling polyclonal antiko lies are well known in the art. Briefly, a polyclonal antibody is l~lcp~cd by il..~l.ll..i7.il.g an animal with an antigenic CGll~Silion and collPsting antisera from that il.ll.l~ ed animal. A wide range of animal species can be used for 10 the production of antiscra. Typically the animal used for production of autise~a is a rabbit, a mouse, a rat, a h~m~ter, a guinea pig or a goat. 17~ef~lce of the relatively large blood volume of rabbits, a rabbit is a ~l~f~ ,d choice for proA~ tion of polyclonal antibodies.

As is well known in the art~ a given colll~osition may var,v in its imm~nQgeni.;ily. It is 15 often nr~eS~ y lll~,.efo~e to boost the host ill-l-.,J -r system, as may bc achi~,~,c1 by coupling a peptide or polypeptide imml-nogen to a carrier. F~emrlary and yl~:~e~lcd carriers are keyhole limpet hemocyanin (KI,H) and bovine serum albumin (BSA). Other ~lhl~minc such asov~lhnmin, mouse serum albumin or rabbit serum ~lbllmin can also be used as c~ rs. Means for conjugating a polypeptide to a carrier protein are well known in the art and include 20 ~ t~r~ hydc, m~ k; l~ obenzoyl-N-l~o~y~cc;~ id~ ester. carbo-liimi~e and bis- bi~7~tl7~d ben7i~lin. .

As is also well known in the art, the ill~ og~ ieily of a particular il~.. l.og~n c~ q~ ;Qr can be c~h~ ~r-ed by the use of non-specific stim~ rs of tne i~ c;.yollse, 2~ known as adj~Y~I~. P~ .y and p.~,f.,.l.,d zd~uv~ts include complete Freund's adjuvant (a non-spe~-ific stimlll~tor of the i~ ~-r l-,S~Ol~C con~ killed Mycobnr~rium tuberculosis), incomplete Freund's adjuv lts and ~ll..l.i.l..l., llylhu~ide adjuvant.

The ~mmlnt of immllnogr-n culllposilion used in the Inoducl~on of polyclonal ~ntiho lir~
vanes upon the nature of the i.. ~-ogen as well as the animal used for ;~ ;7~tion. A variety of routes can be used to ~lminist~pr the im~lmQgen (subculalleous, intr~m--sc~ r, intra~l~Prm~l, intravenous and inLI~c.iloneal). 'l'he production of polyclonal antibodies may be monitored by 5*~ .g blood of the immlmi7ed anirnal at various points following i ~ tion. A second, booster in~ection, may also be given. The process of boosting and titering is repe~te~ until a sui~lc titer is achieved. When a desired level of immllnogenicity is obtained, the i.. ~ i7.rd animal can be bled and the serum isolated and stored, andlor in some cases the animal can be used to gcnc.dtc MAbs. For p~ ;on of rabbit polyclonal antibodies, the animal can be bled through an ear vein or alternatively by cardiac pul,clulc. The rcmoved blood is allowed to co~ te and then centrifuged to scpa~ate serum co~ )onents from whole cells and blood clots.
10 The serum may be used as is for various applications or the desired antibody fraction may be purified by well-known mPth~ such as affinity ch~ atography USiDg another an~ibody or a peptide bound to a solid matrix.

Monoclon~l ~nril~ol1içs (MAbs) may be readily ~,pa~d through use of well-known tCc~ es, such as those exelnplifi~ in U.S. Patent 4,196,265, inco,l.u,~ted herein by ~ef~ ,Ke.
Typically, this t~chl-iq.le involves i~ i7.ing a suitable animal with a select~P~l immnnogen CO-~-i os;~io~, e.g., a p~lrifie~l or partially purified c~ ssed protein, polypeptide or peptide. The immnni7ing CC~ yOSi.iOII is ~ in~inict~red in a manner that erL~ti~,cly stimlll~Ps an~ibod~
pro~lncing cells.
The m~tho-lc for g~ ra~ onoclonal antibodies (MAbs) ~r,n~riqlly begin along the same lines as those for preparing polyclonal antibodies. Rodents such as mice and rats are p.cfcl,~d ~nim~lc, ho.._~.,r, thc use of rabbit, sheep or frog cells is also possible. The use of rats may provide certain advantages (t'~lin~, 1986, pp. 60-61), but n~icc are p~ d, with the 25 BALB/c mouse being most yl~,f~l.,d as this is most routinely used and generally gives a higher ~elce"~age of stable fusions.

The ~rlim~lc arc iniect~ with antigen as ~lçscnkecl above. The an{igen may be coupled to carrier m~ c~les such as keyhole limpet h.,~ anin if nPc~cc--y. The antigen would typically CA 0224643l l998-08-l2 W O 97/26321 PCTnUS97/00761 be rni~ced with adjuvant, such as Freund's complete or incomplete adjuvant. Booster injections with the same antigen would occur at approxirnately two-week intervals.

Following immlmi7~tion, som~~ic cells with the potential for pro~ ng antibodies,5 speçifi~lly B l~ ho~ytcs (B cells), are selected for use in the MAb g~n~laLillg protocol. These cells may be oblA.I~ed from biopsied spleens, tonsils or lymph nodes, or from a ~ )hr,~1 blood sample. Spleen cells and ~ yh~"al blood cells are ~ r~ d, the former because they are a rich source of antibody-L)lod~cing cells that are in the dividing p!~Cm~hl~ct stage, and the latter bec~ase ~el.~}l~,lal blood is easily ~ces~ e. Often, a panel of ~nim~c will have been 0 i~ ;7eA and the spleen of animal with the highest antibody titer will be removed and the spleen lymphocytes obt~in~d by homog~.ni,.;ng the spleen with a syringe. Typically, a spleen from an i-.. ~ i7ed mouse cont~inS ~lu~ t~ly S X 107 to 2 X 108 Iymphocytes.

The antibody~ B Iy,..~hocytes from the i..~ .i7e~1 animal are then fused with 15 cel}s of an illJlllUllal myeloma cell, generally one of the same species as the animal that was i--.. ---;~.'fl Myeloma cell lines suited for use in hybri-l~m~-producing fusion ~ XelluleS
preferably are non-antibody-pro~uc in~, have high fusion erLcica,c~, and have c.,Lrl,le d~r.r;.--fics that render them ;~ lc of growing in certain selective media that support the growth of only the desired fused cells (hybridomas).
Any one of a number of mycloma cells may be used, as are known tû those of skill in the art (~in~, pp. 65-66, 19861. For example, wh¢re the ;.,.. ~ ~;Yerl animal is a mouse, one may use P3-X63/Ag8, X63-Ag8.653, NSl/l.Ag 4 1, Sp21~Agl4, FO, NSO/U, MPC-l 1, MPCll-X45-GTG 1.7 and S194/SXX0 Bul; for rats, one may use R210.RCY3, Y3-Ag 1.2.3, IR983F and 4B210; and U-266, GM150~GRG2, LICR-I,ON-HMy2 and UC72g-6 are all useful in co~ çcl io~ with human cell fusions.

One ~efe..~,d murine mydoma cell is the NS-l myeloma cell line (also termed P3-NS-1-Ag~ll, which is readily available ~om the NIGMS Human Genetic Mutant Cell Rcposi~ r by I~q.le,l;r-g cell line I~Osi~u,~ number GM3573. Another mouse myeloma cell line that may be used is the ~-azaguanine-resistant ~ouse murine myeloma SP2/0 non-producer cell line.

Methods for g~ne~ ng hybrids of antibody-pro~uring spleen or Iymph node cells and S I~ lollla cells usually comprice r~uxing somatic cells with myeloma cells in a 2:1 plu~oll~on~
though the pfopoI~ion may vary from about 20:1 to about 1~ s~cliYely, in the pI~,~ence of an agent or agents (chPrnic~I or çIe~ Al) that promote the fusion of cell membranes. Fusiûn mPthofl$ using Sendai virus have been described by Kohler and Milstein (1975; 1976), and those using polyethylene glycol (PEG), such as 37% (v/v) PEG. The use of el~ctric~IIy in~ e(l fusion m~.tho~c is also aplJIo~ te (Goding pp. 71-74, 1986).

Fusion pIvcedu~.,s usually p,c,duce viable hybrids at low frequPn~ies, about 1 X lob to 1 X 10-8. However, this low fre~uency does not pose a ploble.l., as the viable, fused hybrids are Lrf~ d from the parental, unfused cells (pa~icularly the I~ Iced myeloma cells that would 15 norrn~IIy co~I;....r to divide il~dcr~ y) by c~Iltn~ing in a selective ~ A;~...,. The sele.,~i~_ ~r,.1;..~,~ iS ~n~r~lly one that co~ c an agent that blocks the de novo ~ thesis of nucleotides in the tissue culture media Exemplary and ~ cd agents are aminopterin, ~ .st..,~ate, and ";..P ~minspt-rin a~d ~~-- tIn~I~c~ate block de novo ~ylltLesis of both purines and pyrimillin~s, whereas z~7~ce~ Ç blocks only purine synthesis. Where ~,lmo~t~,.in or 20 methotrcxate is used, the media is supphP~m~ntp~d with hypox~nthin~ and thyn~idine as a source of ~ leot;~es (HAT I"rl1;,...,) Where ~,~c~.;..ç is used. the mcdia is ~u~pl~ t d with hypoxanthine.

The pl~ d s~lecl~ol~ is HAT. Only cells . l~a~le of Op.. ahllg nn~leoti~3P25 salvage pathways are able to survive in HAT "~e~ The myeloma cclls are dcf~;~ in key llle,S of the salvage p~Iway, e.g., hypox~r thin~ phosphoribosyl t~ansferase (HPRT), and thus they cannot swive. The B cells can operate this p~lIway, but they have a limited life span in culture and g~nPplly dic within about two weeks. Thc~foIc, the only cells that can survive in the ~cl~ , media are those hybrids formed from ~llyel~lla and B cclls.

W O 97/26321 PCTnUS97100761 This c~lltllring provides a population of hybridom~c from which specific hybridomas are sehPctP~ ~ Typically, selection of hybri-lom~c is phfu~ ed by clllhlring the cells by single-clone dilution in microtiter plates. followed by testing the individual clonal supernatants (after about two to three weeks) for the desired ~ ity. The assay should be sensilive, simple and rapid such as radio.. ~no~s?ys, enzyrne ;.. ------o~c~ys, cytotoxicity assays, plaque assays, dot i.~.. -oblllding assays, and the like.

The selecte~ hybridomas would then be serially diluted and cloned into individual antibody-pro~uçing cell lines. which can then be pr~p~t~-d in~.~finitçly to provide MAbs. The 10 cell lines may be exploited for MAb production in two basic ways. A sar~ple of the hybridoma can be in.jected (often into the peritoneal cavity) into a hictoco~ ble animal of the type that was used to provide the somatic and myeloma cells for the ori~in~l fusion. The injectPd animal develops tumors secl~,tillg the spe ;r.c mon~lons~l ~ntiho-ly produced by the fused cell hybrid.
The body fluids of the animal, such as serum or ascites fluid, can then be tapped to provide 15 MAbs in high con.,.,.ll.~tion. The individual cell lines could also be e~ ,d in vitro, where the MAbs are naturally sec~ted into the culture ~ from which they can be readily obtained in high c~ . alions. MAbs produced by either means may be further pll~ifie~l if desired, using filtr~-ion, ce.,~irugation and various C}h~ tc~hic met~odc such as HPLC or affinity C~ulll..t~ y~
Large ~ o!~..tC of the mnnoclon~l ~ntibo~içc of the present invention may also be obtained by multiplying hyhri~lom~ cells in vivo. Cell clones are injectcd into ~ -."-~lc that are l~slueu...y~ ,le with the parent cells, e.g., syngeneic mice, to cause growth of antibody-p~ ;n~ tumors. Optionally, the ~nim~lc are primed with a h~bon, especi~lly oils such as 25 pristane (te~ yl~t~decane) prior to injc~l;on In accordance with the present invention, L~ entS of the ~ ~oclon~l antibody of the invention can be obtained from the .~ no~ l antibody ~oduced as ~lesc Y~ above, by m~thod$ which include ~ sl;on with e~ es such as pepsin or papain and/or cleavage of 30 tliclllfi-le bonds by c~ re~ cti~m Alt~,..~ively, monoclonal antibody fr~grn~ntc W O 97/26321 PCT~US97/00761 e ..ro...l~cse-l by the present invention can be synthF si7e~1 using an ~ltom~tçcl peptide :jy~lhF c; ~f 1, or by e~ ssion of full-length gene or of gene fr~gm-F ntc in E. coli.

The monoclonal conjugates of the present invention are p~c~ ,d by mPtho~c known in S the art, e.g., by reacting a monoclonal antibody p~ d as described above with, for in.ctqn~e, an enzyme in the yl~is~nc- of a coupling agent such as glutaraldehyde or penoAq~P. Conjugates with fluol~,sc_.n ulcul~ are yucl~ed in the ~n,scllce of these coupling agents or by reaction with an isothiQcyanate. Conjugates with mctal çht l-q-~F~s are similarly proAuGe~ Other moieties to which ~n~ihoAies may be conjugated include radionnclitlec such as 3H, l25I, 131132P, 35S, 14C, s~Cr, 36Cl, 10 5'Co, 58Co, 59Fe, 75Se, 's2Eu, and 99~c, are other useful labels that can be conjugated to ~ntihoAi~s R~lio-q-rtively labeled rnonoclonal antibodies of the present invention are produced acco~ g to well-known methtxls in the art. For in.ct-nre, monorlon~l antibodies can be io~ t~d by contact with sodium or potq-ccillm iodide and a ck ~ l oxi~i7in~ agent such as sodium hypochlorite, or an enz~J~ic oxiAi7in~ agent, such as la.;lo~c.~xidase. Morl~,i~nsl 15 ~ntihoAi~,c according to the invcntion may be labeled with tccht.~ 99 by ligand exrhqnge process, for çYq-rnrle, by re~lnring ~.t-,cl~ t~ with stannous solution, rh~lqting the ,~,.h-ced .t;~ onto a Sephadex column and applying the ~n-ibody to this column or by d~rect Iq~lin~ ~çchni~llFs, e.g., by incubating y~t~,c~ ate, a l~J~c;.-g agent such as SNCI2, a buffer sol~1tinn such as sodium-potassium phth~l~tF sol~ or, arld the antibody.
It will be appreciated by those of skill in the art that monoclonal or polyclonal antibodies ~-;rlr for the amylin and amylin-related c~ .o.c;~;ol-.c of the present invention will have utilities in seYcral typcs of ~ tionc These applications will include the production of ~ii~ o~lic kits for use in rle~ g amylin or ~lia~osing der~iF~ie~c in amylin se~;,elion. The 25 skilled pr~titinnpr will realize that such uses are within the scope of the present in~ tion.

L. Immunodetection Assays The immnno~ ;QI- mf~thnds of the present invention have evident utilit~Y in the 30 det~'ing of co~ ;Qns such as amylin and diagnosing ~ s associated the.. ~ . Here, a W O 97J26321 PCT~US97/00761 biological or clinical sample suspectçd of containing either the er~roded protein or peptide or corresponding antibody is used. However, these embo~ t~ also have ~pplir~ti~m~ to non-clinical s~mrlPs such as in the titering of antigen or antibody s~mpl~os, in the selection of hybndom~c, and the like.

Those of skill in the art are very f~m;li~r with differ~nti~tin~ between signifit~nt e"~ s~ioll of a protein, which l.,~l~sents a positive icl~ntifir~tion~ and low level or background e~yl~ssion of such a protein. Indeed, background e~lession levels are often used to form a "cut-off" above which increased st~ining will be s¢ored as si~ifi~nt or positive. Si~ifie~rt 10 expression may be re~l~s~,nted by high levels of ~nti~cnc in tissues or within body fluids, or t.. ~ ly, by a high ~upollion of cells from within a tissue that each give a posilive signal.

1. ~mmzulod~ctrQn Methods In still further emho~ -ts, the prescnt invcntion co,~ c i.l.. -odPtection m~thorls 15 for ~inAing, purifying, removing, ~u~lliîying or otherwise ~en~r~lly ~cl;..g biological CO-upQ~ . The en~oded ~rot~ s or pepti~es of the present invention may be employed to detect antibodies having reactivity lL~.e~ l, or, alh,~ ivcly, antibodies pl~,p~,d in accordance with the present invention, may be employed to detect the e~ ode~ proteins or pepti~eS. The steps of various useful immllnofleP~tion m~th')~g have been rl~scribed in the sc.~utific li~ c 20 and are well known to those of s~ill in the art.

~ ~en~r~l, the immnnobinding ~-- !hOrlC include obtaining a sample su~e~ted of CQ~t~ g a protein, peptide or ~.t;bo.~y, and contacting the sample with an antibody or protein or peptide in accordance with the present invention, as the case may be, under co~-iition~
25 t;rr~ to allow the f~ .OI~ of jmmnnocol~ s The ................ ob~dillgm~tho~sincludemethodsforclct~t;~gorquantifyingtheamountof a lea~ , CO-..i.Q~ in a sample, which meth~1c re~uire the ~letectio~ or ql~ntit~iQn of any h.~ ...r complexes formed during the binding process. Hcre, one would obtain a sample 30 sui",cct~,~ of con~ an amylin, an amylin-related peptide or a COI.,Sl~Ql. I;.~g antibody, and WO 9~126321 PCT/US97/00761 contact the samp}e with an antibody or çn~or1ecl protein or peptide, as the case may be, and then detect or ~lualltiry the amount of l...~ comr~eYes formed under the specific con.lition~

In terms of antigen detection, the biological sample analyzed rnay be any sample that is 5 ~ of col,l~ining an amylin ~ntigen, such as a pancreatic ~-cell, a hc"llogc~.7~d tissue extract, an icol~~d cell, a cell .l.e~ .a~e ~ ion, sc~ at~l d or purified forms of any of the above protein-con~ g co,~l~osi~ , or even any biological fluid that comes into contact with di~etic tissue, inrlu-~ing blood.

C(jn~cl.ng the chosen biological sample with the protein, peptide or antibody under con~itiQnc effective and for a period of time sllffiri~nt to allow the forTn~tion of ;""""'~f ~r. l-x~s (prirnary ~ e col~ ,.ec) iS g~n~ ,tily a matter of simply adding the composition to the sarnp}e and ir~ b~ g the ~ , for a period of time long enough for the antibodies to forrn i~ comrl~Y~s with, i.e., to bind to, any ~nti~nc prescnt. After this time, the sample-15 antibody co~ osition, such as a tissue section, ELISA plate, dot blot or western blot, will g~ner~lly be washed to remove any non-specift~ ly bound antibody sp~,jes, allowing only those antibodies specific~lly bound within the ~;lllaly; .. ç complexes to bc .1e~t~l In general, the detection of i.~ oco~ lc~ formation is well known in the art and may 20 be acl~ ed through the application of nulll~,luL~s ap~uroacl~es. These m~thor1c are generally based upon the det~c~ion of a label Ol rnarker, such as any ~ Acl.vc. nuolescc.lt, biological or c~zyl,~atic tags or labels of standard use in the art. U.S. Patents COl~f~ g the use of such labels include 3,817,837; 3,8S0,752; 3,939,350; 3,996,345; 4,277,437; 4,275,149 and 4,366,241, each inc.,l~ul~te~ herein by ~c,f~,le.~e. Of course, one may find additional advantages through the use 25 of a secord~ry binding ligand such as a second ~ ibQdy or a biotinlaYidin ligand binding a~l ag~ t, as is known in the art.

The e~eo~1ed protein, peptidc or co.l~,;,po~ lg antibody employed in the ~l~tection may itself be linked to a dctectable label, wLe~ one would then simI~ly detect this label, thereby 30 allowing the amount of the primary ;..~ e complexes in the co...l ocil;~n to be dct~..;..ed W O 97J26321 P ~ rUS97100761 ~ l4 ~ ely~ the first added co~ on~ that becomes bound within the primary immnn.o col1.~ Lçs may be detected by means of a second binding ligand that has binding affinity for the erlroded protein, peptide or collc;,~l.ding antibody. In these cases, the second bin-ling ligand 5 may be linked to a ~etect~le label. The second hin-ling ligand is itself often an antibody, which may thus be termed a "s~con~l~ry" antibody. The primary j,}",~ r complexes are cont~rte~l with the l~ rl, seCon~3~ry binding ligand, or antibody, under con~1ition~ effective and for a period of time 5~fflrjerlt to allow the form~tion of seConrl~ry imm1me complexes. The secon~y i~
comrl~Yes are then generally washed to remove any non-sperifi~11y bound labeled secondary 10 ~ o~ies or 1iE~n-ls, and the rernsining label in the secondaly imm11nP co~rleYes is then rr~c~l Further mrthoAc include the cletection of primary i.. ~ r com, l~~s by a two step a~p,~.ach. A second binding ligand, such ac an antibody, that has binding affinity for the c-1r,o~1~
15 protein, peptide or collc~onding antibody is used to fo~rn secQ~ c comrleYes, as described above. After washing, the secon~ y j~ r, cu~ e~ec arc CQ~t~ C~l with a third binding ligand or ~~ oly that ha-s binding affinity for the second antibody, again under conrlitiol-~ effective and for a period of time s~1ffirirnt to allow the formation of imm1m-o complexes (tertiary ;~ ..r compl~r~s). Thc third ligand or antibody is linked to a detectable 20 label, allowing rkre~ of the tcrtiary i.~ complexes thus formed. This system may provide for signal ~mrlifir~tion if desired.

2. Imm~/nohistochemistry The ~ntibo~ s of the present invention may also be used in co~ .on with both fresh-25 frozen and fonn~iin-fLxed~ p~ rr.~ o dd~ tissue blocks ~.c~_~,d for study by imm1mQhictorh....,i~1.y (IHC). For ex~mple, each tissue block consists of 50 mg of residual "pulvenzed" rli~betir tissue. The m~thn-l of plC~J~ing tissue blocks from these particulate ;"-r~c has been s~eces~r~llly used in previous IHC studies of vanous P1~;J~OS!;C factors, and is well known to those of skill in the art.

Briefly, frozen-sections may be ~ d by rehydrating 50 ng of frozen "pulverized"
diabetic tissue at room telllp~a~ in l,hosph~e b~lrf~cd saline (PBS) in small plastic c~ps~ s;
pellP-ting the particles by centrifugation; ~c;~n~ nAing them in a viscous emhed~ing mP~ m (OCI); inverting the capsule and pelleting again by c.,.lt,ifugation; snap-freezing in -70~C
5 iSo~e~ln~e~ cutting the plastic capsule and removing the frozen cylinder of tissue; seClring the tissue cylinder on a ~ o~Lat micloto"le chuck; and cutting 25-50 serial sectionc P~.,.,anc.lt-sections may be ~ ed by a similar m~thod involving l~hy-llalion of the 50 mg sample in a plastic microfuge tube; pell~inE; resllcpenfiing in 10~o forrn~lin for 4 hours 10 filc~tiorl; washing/pellPring; ~,u~ n~ g in warm 2.5% agar; pÇlleting; cooling in ice water to harden the agar; removing the tissue/agar block from the tube; infiltra~ing and embedding the block in l~.,.rfi,l; and cutting up to 50 scrial permanent sectior c 3. ELISA
As noted, it is cont~ .. pl~ted that the çnco~ed proteins or peptides of the invention will find udlity as immlmogcns, c.g., in cQnl~ ion with vaccine development, in imml~nohictoçhf,..i~ and in ELISA assays. One evident uti}ity of the enrocleA ~ntigen.c and cG.l~,sp.~ i.. g antibodies is in i.. oo~s~ys for the detection of amylin and amylin-related pepti-l~s, as needed in diagnosis and pl'O~llOSLiC mr nitoring of various rli~oced states.
.nn~cs~, in their most simple and direct sense, are binding assays. Certain ,h,.l~,d i-.-."--.. Oo~c~ are the various types of ~"~Zyl-lC linked i.. ~ osorbent assays (F~ ~SA) and r.~ noocc~ (RIA) known in the art. ~ h.cloc~ -;r~ tectioll using tissue seet;r~nC is also par~icularly useful. However, it will be readily ap~eciated that detection is not 25 limited to such ~eçhniTles, and westem blotting, dot blotting, FACS analyses, and the like may also be used.

In one eYr ..~ y ELISA. antibodies binding to the enro-l~d plot~ s of the invention are i.. ~i~i7ed onto a st~lect,ed surfacc e~thibiting protein affinity, such as a well in a poly~lylc,le 30 ll~.elY~Lit~r plate. Then, a test cu.llposilion s~ ed of con~;ni~g the amylin, such as a clinical s~mrle. is added to the wells. After binding and washing to remove non-specifir~lly bound i.. ~-~-.~ c"--, 'o~s, the bound ?~ bGA~r may be ~tectefl Detcction is generally achi~,d by the addition of a second antibody speçifi~ for the target protein, that is linked to a ~lPtect~ '-le label.
This type of ELISA is a simple "sandwich E~ISA". Detection may also be achieved by the 5 addition of a second antibody, followed by the addition of a third antibody that has binding affinity for the second antibody, with the third antibody being linked to a det~ct~hle label.

In another e~ pl, ~ El ~SA, the s~ ..ples ~llsl.ccted of cont~ining the arnylin antigen are h~ obilized onto the well surface and then col tP-t~l with the antibodies of the invention. After binding and washing to remove non-~rerifir~lly bound il.. ~r, compl~l~es, the bound antigen is ~lete~t~ Where the initial antibodies are linked to a detectable label, the ;~ nr co...~ es rnay be ~1~,ct~d directly. Again, the i.. ~ co~l-le~es may be APtected using a second ~aly that has binding affinity for the first andbody, with the second antibody being linked to a ~otect~hle label.
Another FLIsA in which the proteins or peptides are immobilized, involves the use of til~dy co..~l~tilion in the d~t~cti~tn In this ELISA, labeled antibodies are added to the wells, allowcd to bind to the amylin protein, and detçcted by means of their label. The amount of marker antigen in an unknown sample is then ~t t~,.. rd by mixing the sample with the labeled 20 antibodies before or during ~ ;QI1 with coatcd wells. The plCSellCC of marker antigen in the sample acts to reduce the ~mollnt of an~,bod~ available for binAirl~ to the wdl and thus reduces thc llltim~tJ~ signal. 1 his is a~ ,c,~liate for d ~ g ~ntihoAies in an unknown sqmrl~, where the unlabeled ~ ~l;ho~ s bind to the antigen-coated wells and also ~educes the ~m~lmt of antigen aYailable to bind the labeled antibodies.
Ll~s~ecLi~e of the format employed, ELlSAs have certain features in CO~ , such as coa~ng, incllh~ting or bin~ling, washing to remove non-specifr~lly bound speri~s, and det~ctin~
the bound ;~ r complexes. These are ~les~ cd as follows:

WO 9712C321 PCTnUS97/00761 In coating a plate with either antigen or antibody, one will generally in~ub~te the wells of the plate with a solution of the antigen or antibody, either overnight or for a specifi~d period of hours. The wells of the plate will then be washed to remove incompletely adsorbed m~t~
Any r~ g available surfaces of the wells are then "coated" with a nons~ccirlc protein that is S ~ntiE~ ly neutral with regard to the test antisera. These include bovine serum albumin (BSA), casein and solutions of miLk powder. The coating of nonspecific adsorption sites on the W O 97t26321 PCT~US97/00761 solution such as PBS~TweenrM~ or borate buffer. Following the f~ tion of specific imml-n~o co...~ es between the test sample and the origin711y bound ITI ~t~ l, and sul~se.lv~ washing, the oc~ul~mce of even minute ~llOWIts of ;".~I."~r. complexes may be det~!-...;..ed S To provide a detecl;l~g means, the second or third antibody will have an acsoci~t~ label to allow detection. ~l~r~,ably, this label will be an ell~ylllc that will ge~ color developmcnt upon incubating with an ap~ liate chrom~genir ~ st~t~. Thus, for ç.~c~mrle, one will desire to contact and inr,~lb~te the first or second ;1~ c complex with a urease, glucose oxidase, ~lk~line phosphatase or hydrogen peroxi~ e-conjugated antibody for a period of time and under 10 con~itio~c that favor the development of further ;~ r complex formation (e.g., inr~tb~tion for 2 hours at room t~ el~lUI~ in a PBS-col-r~ g solution such ac PBS-TweenTM).

After inrllb~tion with the labeled alJ~ib~ly7 and subse~lvent to washing to remove u~lbo~l,d material, the amount of label is qu~ntifiP(l e.g., by inr~-lh~tjon with a chromogenic 1~ ~ul~ te such as urea and bromocresol purple or 2,2'-azido-di-(3-ethy~ 7~ oline-6-sulfonic acid [ABTS] and H202, in the case of peroxidase as the C~ r111C label. Qu~ntit~tion is then aclLeved by ...~ g the degree of color gel~tioll, e.g., using a visible spectra o~boto~ t~

4. Use of Anrr~odres for Radioimaging The ~ntiborli~s of this invention will be used to (lual~tify and localize the e~ s~ion of ~nti~nc such as amylin, ;~ 1;Qg receptor-bound amylin. The antibody, for e~ will be labeled by any one of a variety of rn~tho~lC and used to vi.c-~1i7.e the loc~li7~d conrcr~-~ion of the cells ~ .l".;"g the rl~od~l protein. Such an assay also will reveal the s~lbce localization of the protein, which can have diagnostic and the.a~"lic ~rplir~tinnc ~ accold~ce with this invcntion, the monoclonal antibody or fia~lll.,nt thereof may be labeled by any of several t~chniq~s known to the art. The metho-~c of the present invention may also use pa~ isot~,~s for p~oses of in vivo detection. F.l.. ~-lc particularly useful in Magnetic RrCo~ ce ~ma~ing ("MRI") include l5~Gd, 55Mn, I62Dy, 52Cr, and 56Fe.

W O97126321 PCT~US97100761 alion of the labeled antibody may be local or systemic and accomplished h~ ously, h1tl~L~.ially, via the spinal fluid or the like. A(1minictration may also be h.l,ade.lllal or hltlàcavil~, depending upon the body site under e~min~tion After a suffirie~t 5 time has lapsed for the monoclonal antibody or fr~6lll~,nt thereof to bind with the ~lice~ced tissue, for example 30 ~ IC~S to 48 hours, the area of the subject under investig~tion is eY~fnined by routine imi gjng techni~lu~s such as MRI, SPECT, planar scintillation imaging or ncwly c~ ,hlg imagjnf~ teçhniqlle-c The exact protocol will npcplcs;~rily vary ~epentling upon factors specifir to the patient, as noted above, and ~lepen~in~ upon the body site under e~ri mi~tion, method of 10 ~ ni~ lion and type of label used; the clete ...in~tion of specific proc~dul~s would be routine to the skilled artisan. The distribution of the bound r~io~çtive isotope and its increase or dcclease with time is then mo~ ol~d and l~cor~e~. By cv...p-- ;.-g the results with data obld~t,d from studies of clini~ ly normal individuals, the ~l~s~ ce and extent of the t~ice-ce~ tissue can be ~I~ r~ - ...;n~d It will be a~l)àrc~ll to those of skill in the art that a similar ~pp~oach may be used to radio-image the production of the enrodet~ amylin or amylin-related proteins in human p~tientc~ The present invention provides rn~.tho-lc for the in vivo det ction of amylin or arnylin-related peptide with a view to correlating such detection to riio~rlsi~ hetes in a patient. Such metnods 20 generally co~ ..i.ce ;1rlminictpring to a patient an effective ;1mount of an amylin antibody, to which antibody is conjugated a marker, such as a radio~ tive isotope or a spin-labeled molecu}e, that is ~l~,t~e~ le by non-invâsive ~th~dc. The antibody-marier cohju~5ate is allowed sllffiri~n~
time to come into contact with reactive; ntigen.s that are present within the tissues of the patient, and the patient is then exposed to a l1et~ction device to identify the d~tectahlP marlcer.
5. Kits In still further emho~limP.~tc, the present invention co~ mm~modet~ction kits for use with the i~ ode~ecliQn m~tho~l~ desc;l.bed above. As the ellcoded proteins or pep-ides may be employcd to detect a~tibodics and the coll~spc~ n~ i-ntibo~ s m~ be employed to detect 30 c ~-~o~le~ role.ns or peptides either or both of su¢h cu~ onlnls may be provided in the kit. The W O 97126321 PCTrUS97/00761 immnnoAetPction kits will thus comprise, in suitable container means, an encoded protein or peptiAP, or a first antibody that binds to an enr,odçd protein or peptide. and an i~ ....odet~Pction reagent.

In certain embo~limp~nt~ the enroAPA protein or peptide, or the fir.ct ~ntibody that binds to the e<nroA-~p~ protein or peptide, may be bound to a solid support, such as a colllmn matrix or well of a ~ olit~r plate.

The immllno-~Ptection reagents of the kit may take any one of a variety of forrns, 0 ;~rl.~ those dFteclAh3e la~els that are associated with or linked to the given antibody or ~nti~n, and de~ hle labels that are ~csori~tpd with or ~tt~rhP~ to a sccond~r binding ligand.
F.YP1T~rY SGCO~ ligands are those seco~A~y antibodies that have binding affinity for the first ~ntiho~ly or ~nti~n, and sec~ ~A~ y antibodies that have binding affinity for a human an~ibody.
1~
Further suitable i.. -o-kte~.l;on rea~.lts for use in the present kits include the two-c~ -o~ t reagent that co.nl.~;seC a seCon~y antibody that has binding affinity for the first antibody or ~nti~n, aiong with a third ~nti1~ociy that has binding affinity for the second ~nti the third Antibody being linked to a A~tect~hle label.
The ~its may further co...l.. ;ce a suitably ~ o~ç~l co~roxition of the en coAP~t protein or pol~lide ~ntigen, ~LclL~l labeled or .~ Ahel~tl as may be used to prepare a standard curve for a ~k~tjr~n assay.

2~ The kits may contain antibody-label conjugates either in fully conjugated form, in the form of il~t~ rs, or as sel,~ate moieties to be conjugated by the user of the kit. The co ..~ r.tX of the kits may be p~f~ ge~l either in ~ueous media or in Iyophili7Pd fonn.

The CQ~ '-f rneans of the kits will genP~lly include at least one vial, test tube, flask, 30 bottle, sy~inge or other col~A.I ,r meu~s, into which tne antibody or antigen may be placed, and W O97126321 PCT~US97/00761 preferably, suitably aliquoted. WheK a second or third binding ligand or additiona} co~ )onellt is p~ovidcd~ the kit will also gener~lly contain a second, third or other additional CO~ f~ into which this ligand or co~ )on~ t may be placed. The kits of the present invention will also typically include a means for co.~ i..g the antibody, ~ntigP,~, and any other reagent co~t~iners 5 in close co~r.i.~ .nt for co~ ,lcial sale. Such containers may include injection or blow-molded plastic containers into which the desired vials are ret~inr M. Pharmaceutical Composition~c Where clinical applications are contemrl~tefl, it will be necess~ry to prepare ph,.. ~ ir,~l colllpositions - either gene delivery vectors or Pngi~ d cells - in a form appropl,ate for the i~ ed application. Generally, this will entail ~ ,p~uhlg col..po~ ions tnat are eSs~-nt~ y free of pyrogens, as well as other hll~ ies that could be harmful to h~ nc or C

One will g~nP~Plly desire to employ ayylulJliate salts and buffers to render delivery vectors stable and al}ow for uptalce by target ce}}s. Buffcrs also wi}l be employed when cco..~ nt cel}s are introduced into a patient. Aqueous colllpositions of the present invention cornprice an effective amount of the vector to cel}s, dissolved or dispersed in a ph~nn~reotic~lly 20 ~r~e~ le carrier or aqueous ".r~ , Such compositions also are referred to as inoc~ The phrase ''ph~mace )tirolly or ~h~ -rologically acceptable ' refer to moleclll~r entities and co...l)oc;~ c that do not pl(~ Ce adverse, allergic, or other IYnt~ r1 ;nnc when d to an animal or a human. As used herein, "phar~ -rc-~l;cAlly ~rcept~ble camer"
ir.. ~ es any and a}l so}vents, .l;~ ion media~ co~tin~s, ~ ;h~te ;5~l and antifungal agents, 25 icotonir, and absol~n.o.l delaying agents and the like. The use of such media and agents for pha~ lir~lly active su'titl5--~es iS well know in the art. Except insofar as any conventional media or agent is incv~y~t~ye with the vectots or ce}}s of the present invention, its use in Ih~ .~ l;c colllpos;l;~ c is c~ d Su~lJl - . f ~ y active h~ ,di~ s a}so can be illcol~ ted into the cu...l~o~
~0 Solutions of the active ingredients as free base or ph~ ologically ~c~pt;~hle salts can be prepared in water suitably rnixed with surfactant, such as hydroxypropylc~ lQse.
Dispersions also can be l~le~d in glycerol, liquid polyethylene glycols, mixtures thereof and in oils. Under ordinary conditions of storage and use, these ~l~,pa~ations contain a In~5c.v~livc to S prevent growth of lllicrool&~ ....c.

The e~icssion vectors and delivery vehicles of the present invention may include classic yh~ lcp~d~ions. ~rlmini~tration of these compositions accoldil~g to the present invention will be via any common route so long as the target tissue is available via that route.
10 T}iis in~ludec oral, nasal, buccal, rectal, vaginal or topical. ~ltern~tively, ~1minictration may be by o,ll.ot~,l.ic, intr~ nn~l~ slll,cu~n~4~lc, intramuscular, inl,a~ oneal or intravenous injection Such colllyo~ ions would nonn~lly be ~-lmini~t~red as yh~ uli~ally accc~ t co...~ onc, described supra.

The vectors and cells of the present invention are advantageously a~minictPred in the form of iujeclable co...l-os;lions either as liquid snivtionC or ~ iol~c; solid forms s~lit~hle for solution in, or s~c~ ;on in, liquid prior to injection also may be prepared. These pl~,l,alaliolls also may be ernlllcified A typical co~ ,osiL-on for such ~u,~oses con~rrices a 50 mg or up to about 100 mg of human selum albumin per rnilliliter of phosphate-buffered saline. Other 20 ~h~ t~ iiy ~cc~t~hle carriers inr~lvd~ 3.quc~QuC solutions, non-to~cic ex.,ipi_nls, in~h~ ng salts, pl.,s~.~atives, buffers and the like. F-~mrles of non-aqueous solvents are propylene glycol, pol~_lh~.l.,..e glycol, vegeta~le oil and injectable organic esters, such as ethyloleate.
~'lU~QUS carriers include water, ~ oholi~/z~lucous soh~tinnc, saline s~hlti~nc p~nt~ hicles such as sodium chlori~!, Ringer's d~,Allose, e-c. Lllla~.~nOUS vehicles include fluid and ~ t 25 repl~ ,s include ~ntimirrobjal agents, anti-oxi~l~ntc~ ch~l~tine agents and inert gases. The pH and exact cun~e~ d~ion of the various co~ oncnl~ in the ~ -t.l;f~l are e1 according to well-known p.~. ~. . .- te. ~.

.~tiflitinn~l formulations are s~litahle for oral ~.l....ni~l.ation. Oral form~ tinnc include 30 such typical excipients ac, for ex~ le, p~ tir~l grades of m~nnitol, lactose, starch, 12~

m~n~cium stearate, sodium saccharine, cellulose, m~ntosium carbonate and the like. The cu,~ os;lions take the for n of solutions, s~ cior~, tablets, pills, cars~ C, ~ r~ release form~ tions or powders. When the route is topical, the forrn may be a cream, o;~ salve or spray.

An effective amount of the th~ lic agent is deterrnined based on the int~n~ l goal.
The terrn "unit dose" rcfers to a physically discrete unit suitable for use in a ~ubje t, each unit COI~t~ g a pre~etf ...;.led quantity of the lll~la~ulic ccn")osition c~lclll~tecl to produce the desired response in ~csoc~ on with its ~Amini~tation~ i.e., the a~plu~.iate route and ~
10 n,~-~ The quantity to be ~lmin jcte.red, both accol~ling to nurnber of l~ F~ and unit dose, depen~lc on the subject to be treated, the state of the subject, and the protection desired. Precise 7....n...~ of the the.a~ lic co~ ;on also depcnd on the jU~l~n~nt of the pr~itio~p~ and are pec~ r to each individual.

N. Examples The following PY~mllçs are in~luded to de...ol-s~,~te preferred emb~lim~ntc of the inveDtion. It should be appreciated by those of skill in the art that the techni~lues llic-~losed in the eY~ml~lec which follow r~lJr~sent techniques discovef~d by the inventor to function well in the 20 p.aLhce of the invention, and thus can be corcidered to constitute plefell~,d modes for its pl"~ e However, those of skill in the art should, in light of the present disclosure, a~.eciale that many c~ Ees can be made in the spe~-if~r embo~;...~-nl~ which are rlic~lose-l and still obtain a like or sirnilar result without de~&ling from the spirit and scope of the invention.

h'.XAMPLE 1 Hexokinase I T~ te.l Disruption Metbods Co~ u~ on of ~ene r~t vector. A 15 kB clone, contA;~ g a portion of the rat hexokinase I (HKI) gene e'~yqc~;ing exon 1, about 0.2 kB of intron 1 and about 14.8 kB of 30 ~e~ e u~ of exon 1, wa_ employed. Seq~ nre and maps of this clone aided in the mapping of the HKI gene and in the isolation of homologous isogenic sequences from RIN
gc .v...i~ DNA. The novel 1082 base sequence of the non-transcribed rat HKI genQmi.~ DNA as well as the first 170 bases of HKI tr~ncc~ihed DNA (Schwab and Wilson, 1989) is given as SEQ
ID N0:13. A plasrnid vector plo~ddillg positive and negative selection, pPol~short-neobPA-S HSV-tk, is derived from the pGEM3Zf(+) bac~one and contains a neomycin phospho ,~lnsferase gene (po~ e selection) and two tandem copies of herpes simplex virus thymidine kinase gene (HSV-tk) that provide negative sel~-ti--n in the ~ ,scn~e of ganciclovir (Ishibashi et al., 1993).
pPolIIshort-neobPA-HSV-tk was mo~lified to create pAT9 by creating a uni~ue Nod site 5' of the Neo c~Cs~e ~FIG. 1). A 873 base p~ur fragment was ~rnrlifiç~ from RIN genor~nic DNA using 10 oligos (l~TCCCCTCGAGCACCGCCCGGAACAGTACC, SEQ ID NO: 16 and GTTGCGCCTCGAGCATGCTGACGGTGGGGG, SEQ ID NO:17) to provide a short arm of hc".lol~r to the HKI gene. The se~Pnre e~t~n~ 5' from the first m. ~I ioril.e of exon 1 and is flanked by ~ G.~,d XhoI sites.

In addition, a 1121 base fia~l,ent was ~mrlifiP~l frorn RIN genomic DNA using oligos (GTTGGACTCGAGAGTCACCTAAGGGCCTATG, SEQ D~ NO:18 and GTTGCGCCTCGAGCATGCTGACGGTGGGGG, SEQ ID NO:17), providing a longer short arm to serve as a pGSiliV~; control for sc,e~,ning for homologous l~c~ h;t.n-.lc by PCRrM. The 873 and 1121 base pair PCR~M fr~ tc were le;,~ ed with XhoI and subcloned into pAT9 at a 20 unique XhoI site which is flanked by the Neo c~C~ctte and the copies of HSV-tk (FIG. 1), generating pAT21 and pAT22, l~,s~,ce~ ly.

Soll~hPrn blot analysls in RlN 1046-38 genomic DNA with a probe within intron 1 ~ ealed a 16 kB KpnI fra~ t This fr~nPnt was ~ d by sucrose density 25 ~ .;rl~g~;on, ...~if~Pd with aday~ to create fl~nking Not ~ sites, and s.lbclor.~ into lambda Dash II (S~ f, La Jolla, CA). Rcco,~ t phages col t~ g the fragment were isolaled by pL%que SC~ ~ug. The 16 kB NotI r. ~ was cloned into the unique Not I site of pAT22 to ~r~ide a long a~n of homology to the HKI gene (FIG. 1), geacrating pAT23, the HKl replaccment vector.

WO 97/26321 PCTrUS97/00761 Cell culture, ele.,llol~oration. and dru~ selection. Various cell lines derived from the rat in~nlino!n~ RIN 1046-38 line (Clark et al., 1990) were grown in Mediull- 199 with Earle's salts, cc..n~ 11 rnM glucose and 5% fetal bovine serum. Exogenous DNA was i-lL~ ced into the cells by clecLI~po~ion. RIN cell lines were grown to 50% to 75% confl~enre, harvested by S ll~y~ i7~1ion~ washed once with pho5~h~-buffered saline (PBS), and lcs~pc~ d in PBS for counting. For each electroporation, 1 X 107 cells were pelleted by centrifugation at 1000 rpm for 2 miml~s and lcsu~nded in 0.4 rnl cold Ele~ opolàtion Buffer (137 rnM NaCl, 6.1 mM
glllcose7 5 rnM KCl, 0.7 rnM Na;!HPO4, 20 mM Hepes, pH 7.0). DNA was added to the cell .ciol- to achieve a final con~ tion of 30-50 rnicrograrns per rnl. DNA was clccl-v~.dted into cells in a 2 rnm cuvette at 170 volts and 510 rnicroFaradies using an Electro Cell Manipulator 600 (BTX, Inc., San Diego, CA) Cells were plated in non-selc.,live "~r~i;u...
and cultured for 2 to 3 days. ~r~ lm cor~t;~;nin~ G418 at a final concc,rn~ion of 500 O~ ~llS per rnl was used for 14 days to select for clones integrated with the nc~,l."~
~s~ rc markcr. Following pO5~ , selestioll in G418, Ea~riclovir (Syntex Inc., Palo Alto, 15 CA) at a final co.~ç~ ion of 6 11M was used to se}ectively kill clones e~ .,ssing HSV-tk.
Ganciclovir was applied for 3 days; cclls were then ~--~ t~ erl in ~l~cliulllco~ g G418.

PCRTM assav for tar~eted l~collll~lnants. Following positive selçctiQn in G418 and negative selection in ganciclovir, clones were grown until visible by the naked eye. Individual 20 colonies were picked, dispersed in trypsin, and divided bc~- ~" duplicate cultures into 96-wel}
plates. Following 10 to 15 days in culture, cells of one ~lplir~tP- were rinsed in PBS and lysed by ;"~ t;o~ at 37~C for 8 to 12 hours in fifty microliters of Lysis Buffer (16.6 mM A~
sulfate, ~7 mM Tris-HCI, 6.7 mM MgCI2, 5.0 mM 2-1.l~ ptoelh~nol, 6.7 ~lM EDTA, 1.7 ~lM
SDS, 50 ~gfml ~ tei-.s -, K, pH 8.8), (Willnow and Herz, 1994). Five microliters of Iysate were~5 used as a t~nrl~tP in a twenty-five microliter poly...e.~se chain reaction (PCR~M) in 16 .6 mM
-.. sulfate, 67 mM Tris-HCl, 6.7 mM MgC12, 5.0 mM 2~ e.c,aptoethanol, 6.7 ~IM
EDTA, 1 mM each dNTP, 80 ~g/ml BSA, 0.8 llgfml of each primer, and 2.5 units Taq DNA
poly,~ .~e. The ~ ;r-~tiQn plo~.ul- cc,..c;~d of g2~C, 40 s~p~onr~c~ 57~C 40 secon-lc, 75~C, 1 minute (40 cycles) and a final e1ttension for 5 ~ s at 75~C. The oligor.nl~leotides used to amplify disrupted HKI included a primer in the 3' end of the Neo e~c~ettP
(5'GATTGGGAAGACAATAGCAGGCATGC3' SEQ ID NO:l9, primer 1, FIG. 1 ~hih~chi e~
al., lg93) and a primer in the HKI gene ulJal~ of the putative recombination site (5'AGTCGCCTCTGCATGTCTGAGTTC3' SEQ ID NO:20, primer 3, FIG. 1). The pl~cmi-l S pAT22, cont~ir ing the longer short ~nn of homology, served as a positive control in this PCR~
re~rrion A second control PCRTM reaction was also in~ de~1 using primer 1 and a primer in the HKl gene dowllsllea,l, of the l~co.llbination site (5'CTTGAG(~ lACATGGTGTCACG3' SEQ ID NO:21, primer 2, FIG. 1). This control PCRlM reaction should detect both homologous and random integrants of the HKl l~,p~ ".~ ,~t vector. Recombinants det~ct~l in the first screen 10 were COI firrnP~ in a second PCRTM reaction for which no positive con~rol plasmid exists. The ~hsence of such a control negates the poscibility of a false ~osilive due to co~ ;on. The primers in this secoIlA~ry screen were primer 1 and primer 4 ~5'TCCCCAGGCGTGGGGTAGAAG3' SEQ ID NO:22), an oliE~ cko!;~e ~ ,~,l of the l~O ~h;~tio~ site in the HKI gene (E;IG. 1). PCRTM l"u~ e~ analyzed either by gcl 15 cl~;LI~pliul.,SiS or a slot blot assay. For cl~;~ul~ho,~sis, reaction products were fr?~tior~rPd in 1% a~arose gels in Tris-borate/EDTA buffer (9 rnM Tris-borate, 0.2 mM EDTA). DNA was vi~ cl by st~inin~ in eth~ m~ bromide. For slot-blûts, re~~~ion plu~hlel5 were de.~ d in 0.5 N NaOH, 1.5 M NaCl, neutralized in 1.0 M Tris-HCl, pH 7.5, 1.5 M NaCl, and l~ s~l~d to a nylon ,l,.,.l,~,~e using a 96-well blot apparatus (ScllPil~hllP,r and Schllell~ Keene, N.H.). DNA
20 was cross-linked to the ll~ e and HKI ~n~rlified products were ~ cl by hybri-l;7~
with 32p ~ p" oli~nn~ ooti~les c~ ,ko~ to HKI and int~rnql to ~Jlilll~S used in the ","l,l;fir~ion ~ ;on Positivc clones were replatcd in 96-well dishes to obtain tl'n~iti~S of one cell pr well. These clones wcre allowed to grow and assayed by PCRlM with the plilll.,.~
fi~.~r ikcl above. This cycle of dilution cloning was ~ ri unti} all clones of a plating were 25 positive in the assay.

~ omic Southern analysis. I~IN clones that were positi~,~ by PCRnd for a disrupted allele of HKI were assayed by ger~mic South~rn C~nnmir DNA was i~olq~ed using ~
and protocols of the QIAamp Blood Kit (catalog llulllb.,l 29104, Qiagen, Inc., Chal~.ullll CA) CA 0224643l l998-08-l2 Five to ten micrograms of DNA were digested with e~ cs as inAir~ted and fractionated through 0.8% agarose gels using Tl~AN buffer (0.04 M Tris-HCI, 0.025 M sodium acetate, 0.018 M NaCl, 25 mM EDTA, pH 8.15). EleCl~ hOI~SiS was con~llct~ (1 for 12 to 16 hours at 25 to 35 volts with recirculation of the buffi~,r. DNA was vi.c~ i7~d by st~ining with ethi~ m bromide.
DNA in the gel was dellanll~d for 30 min~-ttos in 0.5 N NaOH, 1.5 M NaCl. Following n~utr~li7~tion in 1 M Tris-HCl, pH 7.5, 1 M NaCl for 30 minl-tes, DNA was transferred to a nylon ~ue~ e (Hybond-N+, ,A~ h~ ) in 10x SSC (lX: 0.15 M NaCl, 0.015 M sodium citrate) and cross-linked to the ~ ,.llbl~le by ultraviolet radiation (W Str~link~r 2400, Stratagene, Inc.). Radiolabeled probes (32p) for hybridization to and detection of genomic 10 fr~g~nt~ were synth~ci7~d as directcd using the rediprim~ Random Primer l ~hellin~ Kit (RPN
1633, ~mer~h~n~ Life Sciences) Mcmbranes were ~leh~ i7~ ~ and hybri~1i7ed in Rapid-hyb Buffer (N~;939, A...~ .. Life Sci~n~es). All inrl~b~tions and washes were ~ d in a Micro~ ~ ion Oven (Hy~aid l.imi~l). Mc~llbl~les were exposed to X-OMAT, ARSfilm (Kodak) to obtain ~loladiogr;lphic signals.
R~sults:
P~or to construction of a gene rep~ e ..f ~t vector, a co...l~P .cor was made of the copy t~ l~r of HKI alleles in rat versus RIN ~r~nlni~ DNA. DNA was fligestP(l with XbaL Southem bloned, and probed with a ra~o~ fragment from intron 1 of the HKI gene.
20 ~ ;ol!;.~hy l~-.ealcd equivale:nt signals dcrived from the rat and RIN HKI gene fr~m.onrc.
~s~l,ably, these signals cGll~a~ d to diploidy of the HKI gene in both the rat and RIN
gel.o...fs. This cQI~ is ~u~ulled by dala that show RIN-derived cell lines to have ~"~-~.t~;~ eA a diploid state in their chul..- SQl..~5. K~ul~ analysis of RIN 1046-38 showed a Aictri~ ion of 35 to 40 cl~ull~osomes with the normal rat co...pll."F nt being 42 ch.c,l..osom~s.
The HKl repl~çm~nt veCI:Or (FIG. 1) was transfected into RIN cells in three separate el~~ pola~ions (EP): EP81, EP8~5, EP95. These cle~,l.opGlalions differ from each other in their ~ol di~ ionc, the identit~r of the parental cell line, and the number of clones s~l~,elled from each ~Table 7). EP81 was derived from a low passage RIN 1046-38 cell line. Of the 500 col ~ s sc~ ed, none were positive for disruption of an HKI allele. RIN-52117, a RIN 1046-WO 97n6321 PCT~US97/00761 38 derived clone, was the parental line in EP86. One positive clone was ~etçctçd in a screen of about 970 colonies. RlN-52/9, a cell line engineered to express high levels of rat glucokinase with pcb71GK was used as a parental line in EP95. About 3200 clones were sc~ned by PCR~
for the ~iesence of a disrupted HKI allele. None were positive.

Potentially, the loss of an HKI allele could result in a growth disadvantage and thereby lead to a lowcr frequency of detçcting HKI gene repl~c~ nl events. To negate a potential m~t~holi~ disadvantage conferred by loss of HKI activity, efforts were made to create parental cell lines that ov~ tyless rat glu~'~in~ce. Such parental lines could potentially serve two lo r....~ .c - first, tO prevent metabolic stress should phosFhorylation of glucose became rate-limitinE~ in tran~Ço,l.led cell lines with rliminich~cl HKI activity; and second, to restore a high Km glucose-phos~Qho~ylating ~tivity to the ~IN lines to shift glncQse-r,,5ponsi~re insulin sec.~lioll towards a more physiological range. RIN-52117, the parental cell line in EP86, had previously been clectlopor~tcd with a pl~cmid con~.ling h~ y~in res;~l~nre and co.~ g a copy of 1~ the rat ~ cokin~ce (GK) cDNA. RIN-~2J17 was h~u~ rcin resistant and was thought to express n.ode.~le levels of g~llc~ n~ce from the tr~n~g~-ne. S~' se,llent data conf.. ed resi~t~n~e to hygromycin, but dis~.o~,d e~ ssion of GK from the transgene (Table 7). About 1000 individual clones were s~.,ncd from EP86. From this screen one clone, 86/X4, was posi~i~,~, by PCRTM. Clone 86/X4 was initially idçntifi~d by ~mplifi~ation with primer 1 and 20 primer 3. The molec~ r weight of the amplified prc>.luc~ was equal to that derived from the pl~cmi-l control. ~o..l~ ion of this clone as cc ~1~in~g a disrupted HICI allele was obt~ined by amplification with primer 1 and primer 4. No p~cmirl control exists for this PCRTM re~ction tl~cfolc, the plOdùCliS not the result of cc.ll~,.ination.

W O 97/26321 PCTnUS97/00761 Electroporation (EP) of R~N Cell Lin~s with a HKl Replacement Vector EP Parental line D~ugR,Parental Tr~n~ene Clones ~~ + by PCRTM
screened 81RIN 1047-38 ~-) (-) 500 o 8~RIN 52-17 HygroR (-) 970 95RIN 52-9 HygroR rat GK 3200 0 Targeted disruption of HKl was att~mrle(i in various RIN lines, in the ~senre of~l~s~ ce of high levels of e~ ssion of rat ghlco~in~ce (GK) from a tr~ncgen~ Cells eAplcssillg the transgene were first selected for rçsi$l~nre to hygromycin (HygroR) and then assayed by Western blotting for expression of exogenous rat GK.

The original positive culturc of 86/X4 was p~cs~ several times prior to dilutional plating for ~c~s5.,~g the purity of the clonal population. 197 individual colonies were cultured in 9~well plates, aJlowed to grow to 50-70% corfll)en~e, trypsini7~d, and split into ~llpli~-ç
cultures. Cells from one set of cultures were lysed and sc~"led by PC~lM using ~Ihl~ 1 and 3 (FIG. 1) and then Icac~ioll products were analyzed by a slot assay. Two clones were co~.r.. ~ as c4l-r~ a dislu~ted allele of HKI. This result delllon~llates two things. First, the original 15 culture that was iA~ntified as 86/X4 was a polyclonal rather than a monoclc!n~l popul~tion Seco~ the clone col~r~;n;,.~ the di:~lul~ted allele of HKI seems to have a growth disadvantage colllpal~d to other cells in the population. This latter possibility is suppollcd by obsel~ions of the growth rates of thc purified HKI repl~r~m~nt clone. The pure 86JX4 grows signific~ntly slower ~about one-half as fast) than clones randomly inte~ilat~-d with the l.,~ e,..~ vector.
Additional data verifying the identity of clone 86/X4 were derived by analysis of gellGll~C
DNA by Sout~ern blotting (FIG. 2). DNA was tligected with EcoRI and Nod, blotted, and hybri~1i7p(l with a probe ~l~s~ of the recomhin~tion site (h~tch~d re~t~n~1e, ~G. 1). DNA
from RIN 1046-38 cells (lane 1) and from RIN-~2/17 randomly integrated with pAT23 (lane 2) produce a predicted signal of about 5.5kB in the autoradiograph. T~is signal CO11CS~Jn~1C to a hom~L~,.Ju~, wild-typc HKI gene. Clone 861X4 pl~duces two autul~liogl~l~ic signals in the W O97126321 PCTrUS97/00761 genornic Southern (lane 3): a 5.5 kB signal c~ . s~,onding to a wild-type allele and an additional signal (about 4.6 kB), indicative of a HKI allele that has homologously ~ n~ rA with the repl~ ~e~ r vector.

Insulin Knockout Methods:
Construction of gene rerl~c~ P.~t vector. The rat insulin I gene ((~nh~nk ~çcescion mullbe. J00747) provided a template from which to create primers for arnplifying seq~lenres from RlN ~nomic DNA. A 590 base pair Cla~ .cnt 3' of the rat insulin gene and col~ onding to positions 4830 to 5420 was ~mrljfi~rl by polymerase chain reaction (PCRTM), subrlor~ and used as radiol~heled probe. RlN ~enomic So~ using this probe revealed a Bgm fragmen~
of about 12 kB that extends three pTime from position 1176. This r~g,..~, I was enrit~he~l by sucrose density ~ .l-;fllgation and sllbc~on~A into BamHI sites of lambda Dash lI vector 1~ (St~ ec, ~~. Recomhin~nt phages c~nt~ining the fr~grn~nt were isolated by plaque sc~e~ g. A
portion of this L~ ,.,nt e~tenr1in~ from an int~rn~l SpeI site to a NotI site provided by the larnbda Dash vector was used to provide a long arrn of hl mology to RIN DNA in the context of a e "~ vector (FIG. 3). A short arrn of ht m~logy to RIN DNA (fivc prirne of the rat insulin I gene) was derived by ~rnrlifi~a~ion of a fragrnent cc,~l~,syollding to nucleotides 1822 to 2860.
20 This r ~. -- ~.t, flanked by XhoI sites, was cloned into the repl~eem~-nt voctor (PIG. 3).

Ihe pl~cmi~ b~Ll.ol-e (pSL9), used for creating a rat insulin I (RINS-1) rep~ f~vector, provided several features ~si~Pcl to enh~n~e and co...l,lcm~nt dislu~tion of the rat insulin I gene. First, positive s~ ;ol- for i~ r~l;on of exogenous DNA into the RIN ~enom~
25 was provided by the gene en~or1i~g neomycin phosphotr~ncf~r~ce. The ex~l~,s~ion of this gene is linked to the c;~ s~ion of human insulin by an i~)t~ern~l ribosGllle entry site (IRES). This allows disruption of the rat insulin gene to be col r l~t' to eA~l~,s~ion of human insulin cDNA. SeconAly, negative selecl ;on, to allow çnri~hm~nt of t~l5~ted over l~dolll integration events, was provided by the eA~l~ssion of the type 2 rat gllJcose ~ t~,r (GLUl'-2). The ~ ,nce of a ru~ In.
30 GLI~-2 renders cells ~llcc~ t;blc to streptozotocin (STZ) toxicity ~-Schne~l et al., 1994).

W O 97126321 PCTrUS97/00761 Thirdly, a unique PacI site at the distal end of the long ann of homology was used to lin~ri7e the vector prior to el~ ~o.ation into RIN cells (FIG. 3).

Cell culture. e~ )o~a~ion~ and dru~ se~ection. Culture conrlition~ are as desrriheA
above except that following ~ositi~, sele~tio~ in G418, STZ (1 mM for 2.5 h) was used to sele~ti~ely kill clones eAI".,s~.g a fi~nrtion~l Glut-2 !~ polt~r.

PCRlM assay for t~c,t~,d lYco~ ants. Following positive selection in G41~ and negative sele~tinn in SlZ, clones were cultured for about 3~ weeks. Cells in each well were 0 ~ A in trypsin and divided bct~ e cultures into 96-well plates. Following 10 to 15 days in culture, cells of one duplicate were rinsed in PBS and lysed by inrllb~~ion at 37~C for 8 to 12 hours in 50 ~11 of Lysis Buffer (16.6 rnM ~mmQnillm sulfate, 67 mM Tris-HCl, 6.7 rnM
MgCl2~ 5.0 rnM 2-merca~loe~ c~l, 6.7 ~lM EDTA, 1.7 ~lM SDS, 50 ~lg/rnl plote;n~cc, K, pH
8.8) (Willnow and Herz, 1994). l,~ive ~ olit~,~ of lysate were used as a t~ e in a 25 ~11 PCRlM in 16.6 mM ~.. ol.;.,.. sulfate, 67 niM Tris-HCl, 6.7 mM MgCl2, 5.0 rnM 2-.e"cal~lo.,!~lanol, 6.7 ~lM EDTA, 1 rnM each dNTP, 80 ~Lglrnl BSA, 0.8 llg/rnl of each primer, and 2.5 units Taq DNA poly...r-~c The ~Tnr1ifi~tion plU~aUl co~ r~ of 40 cycles at 92~C, 40 sc~ s ~7~C 40 sernn~lc~ 75~C, 1 n~inute and a final extension for 5 .. ;~ s at 75~C, and was pc~rul,lled in a 96-well the~locycler (HB-96V, MJ Research, Inc., Watertown, MA) The 20 oli~onllcl~oJi-les used to ~l~ dis~u~d RINS-l in~ e~ a primer in the 3'-end of the Neo c~cre~e ~5'-CAACCG(3TGGGACATTTGAGTTGC-3' SEQ ID NO:23, primer 1, I;IG. 3) and a primer in the RINS-l gene ~ of the putative recu~ ation site (5'-CCAAGI'CATTATAGAATCATAGTC-3' SEQ ID NO:24, primer 2, EIG. 3). The pl~cmid pRDl was created to serve ac a positive control in the PC~ ioll. The b7sL1..u..ç of pSL9 25 wac ligated to an inser~ er~eo~ C~ all of the short arm of homology and eYtçn~ling an ~ 200 bace patrs 5'. PCR~ ~lo.l~ were analyzed using a slot-blot apparatus (part uu~ 27560, S~hti~fr and Schuell). Reaction p~ Cts were denaLul~d in 0.5 N NaOH, 1.5 M NaCl, ~ Al;~ d in 1.0 M Tris-HCl, pH 7.5, 1.5 M NaCl, and tr~f~ d to a nylon WO 97/26321 PCT~US97/00761 C. DNA was cross-linked to the lllc~ldne and RINS-l ~ntrlifiçd l"u~ were ~,t~t~d by hybri~i7~tion with 32P-labelled oli~ fleotides complc....,.ltary to R~S-1 and interr ~I to primers used in the amplification re~ction Positive clones were replated in 96-well dishes to obtain den~iti~s of one cell per well. These clones were allowed to gtow and assayed 5 by PCR~' with the primets described above. This cycle of .lil.lt~ cloning was tepeated until all clones of a plating were positive in thc assay.

RIN clones that were positive by PCRlM fot a dis, .l~led allele of RINS- 1 wete assayed by gr .. ;c Southern. Genomic DNA was icol~t~d using leag~ t~ and ptotocols of the QIAamp Blood Kit (catalog nullll,er 29104, Qiagen, Inc., Ch~ullh, CA). Five to ~en ~ U~ S of DNA was Ai~çsted with c~ ll,es as ;n.l;c~t~d and fi--rJ~o~ r,d through 0.8% ~ose gels using a TEAN buffer ( 0.04 M Tris-HCl, 0.025 M sodium ~-~t~te 0.018 M NaCl, 25 mM EDTA, pH
8.15). Ele~;tlo~ho,~;c was conductcd for 12 to 16 hours at 25 to 35 volts with recir~ on of the buffer from the ~osilivG to the ne~ati~, electrode. DNA was vi~qti7~A by staining with c.t~ bromide. DNA in thc gel was de,"~u,~,d for 30 ,,,;,,,vt~ S in 0.5 N NaOH, 1.5 M NaCI.
Following n~utr~li7~rion in I M Tris-HCl. pH 7.5, 1 M NaCl for 30 ...; ~s, DNA was fc .~d to a nylon ll~,~l~le (Hybond-N+. Arnersham, Chicago, L) in 10X SSC (IX: 0.15 M NaCl, 0.015 M sodium citrate) and cross-linlced to the ,,.r.,~ .,G by ultraviolet r~ ion (UV
Straealinker 2400, SL~ f Inc.). Radiol~h~l~d probes (32p) were synth~ci7Pd as directed 20 using the ~ Random Prirner ~ çllin~ Kit (RPN 1633, ~ ., Life Sci~n~es).
Mcmbranes were prel.~ çd and hybri~1i7çd in Rapid-hyb Buffer (I~:939, ~...c,~ ... Life ~ri~.nt~ l;t.~ n.c and washe~c were ~J~,.ru~cd in a Micro 4 Hybridisation Ovcn (Hybaid ;mited)~ Mc.,lb.anes were c~posed to X-OMAT, ARS filrn ~Kodalc) to obeain autorr~iographic sigllaL.

~ uman Insulin Expression Me~hods:
r~ $~ la~itl cc"~ ,uclion. general dcsi~n. Initial c~p~ ion p1~--";tls were based 30 on pCB6 and pCB7 (Brewer. 1994). These plasmids utili_e ehe strong p~ tu,h_"hancer of the W O 97/26321 PCTnUS97/00761 human CylG,I.eg~lovirus (CMV) imm.oAiqte-e"rly regulatory sequenre to express inserted genes of intcrcst. F.ffitj~nt polyadenyl~tion of l-i .c~ c~ ss~ r.g, ~ RNA is directed by the hum. n growth ho.,l.olle poly~denylation ~ e-nre pCB6 e~-r,odes the Nto.~ reCict-qnre gene co~f .;..~ ~s~ e to the neomycin analog G418, while pCB7 ençod~-,s the h~
S resistance gene. 130th resistant m~*~r~ are tr~ncrribed by the SV40 early promotcr.

A second r~ s~ion Fl~Cmi~l was collsllu~led with many of the same cl~ ~" ~ a pCB6.
The opcn reading frame of the neomycin resistance gene wac amplified with the polymerace chain reaction from pCB6 (Brewer, 1994) using oligos ~CCGGATCCCATGATTGAAC~AGAT, SEQ ~ NO:25 ~d CCAAGATCTCGCTCAGAAGAACTC,SEQ ~ NO:26). The re~lting 816 bp q.~rlifiPd plodu~l w s rçst~ir~.ed with BamHI and Bgm and s~lbclonP~ll into the BamHI site of pCMV8, generaling pCMV81NE0JhGH PolyA. pCMV8 was derived from pCMV4 (Ande.~oll et al., 1989) following removal of the alpha mosaic virus 4 RNAtrncl-q~ion~l çnhqnr,er and repla~ing it with the ~' leader se~.Je~-re of thc adenovirus tri-partite leader (+14 to +154 of major late ll~nC~ t) fused to a hybrid intron cc..llposed of the adenovirus major late tlailS~llpt S'-donor site and a 3'-splice site from a variable region i.~ -oglobulin gene on a 409 bp EcoRllPstI
rl ~c~" ~t (SEQ ID N0: 14, K~llfm~n and Sharp, 1982). SecQnr31y~ a portion of the gene enCor~in~
tne 5'-~ bccl leader of the human Glucose R~ t~l Protein 78 (GRP?8) was ~nrlifi.-d 20 using the pol~.ll.,.~ chain reaction from pThu6.5 (Cu~ g to bases 372 to 594, Ting and ~ee, 1988) using oligos (CCC;GATCCAGGTCGACGCCGGCCAA, SEQ ID N0:27 and CGAGA'~ ~CCAGCCAG~I'GG, SEQ D~ N0:28), generating SEQ ID N0:11. The 5'-le~der of human GRP 78 has been shown to diKct irt~rn~l initi ~ioll of translation allowing for consl~u~lioll of filn~ti~n~l poly,l~ ic genes in ...~ n cells (~ej~ and Sarnow, 1991).
2S The 235 bp ~!ified ~.. xhlct (SEQ ID N0:11) was restrict~cl with BamHI and Bgm and subcloned into the BamHI site of pCMV8/NE0/hGH PolyA genera~ing pCMV8/IRES/NEOlhGH
PolyA (l:~G. 4B). Unique ~ ;c~;on er~lo~ e sibes exist (5'-SalI/XbaI/BamHI-3') for o~ g ~ S into this e~ sion pl~C~i~ n the CMV ~lu~os~Jil~hull and the inte~Dal ribdso.,.~ entlg site~NEC) elements. cDNA's or other open reading frames cloned into W O 97/26321 PCTnUS97/00761 these sites are l~alls~;l,bed from the CMV ~lv~lluter into a bicistronic .. ~-cs~e cc,~ t-ing the cDNA as the U~ G~1I open reading frame and ne~.lly~in resistance (NEO) as the downstream open ~ ~ iin~ frame. Both open reading frames are trancl~t~ effi~iently, linking ncolll~in drug es cli~ Ye and e"~ssion of the uyaLle 1l gene of interest.

A final e~ssion pi~cmifl was dçci~e~ for eAplession of genes of interest. The 5'el~ found in pCMV8 co~ ose~ of the 5' leader seq..~.-re of the adew~i,us tri-partite leader ~14 to ~154 of major late ~.. .c~. ;pt) fused to a hybrid intron cvl~Gsed of the adenovirus major late ~ lscli~L 5' donor site and a 3' splice site from a variable region imm-~nQglo~ulin gene (SEQ
ID NO:14, Ks~ n and Sharp, 1982) was removed by tn~o~ ele~ce restriction by SnaB1 and BamHl and ligatcd into SnaB1 and BBm 1~ 5~ te~ pCB6 (Brewer, 1994), gel~e.~ing pC~B61intron (~:IG. 4A). SnaBl cuts uniquely in both pl~cmitlc at i(~ntie~l pocitiQn~ in the CMV p~ ot~.
seq~ r~ pCB6fintron has several unique ~n~o~n~lFs~e r~ ;on sites for subcloning r."L; ~ C dos ..s~ of the intron sequ~ce and U~ ,~1 of the hGH PolyA se~lur~c (5 -15 XbaI~Rpn~lMI~VClaIlBspDVX~aVBan~-3'). The neolll~i~ "ce gcne is tr~nc~ribed using the SV40 promoter from an i~ ,~..flf -t trar~nr io~1 unit e1-~oded on the pl~cmi(l (Brewer, 1994).

Human insulin e~ ss.on plasmid. A human insulin cDNA cuf t~;1-e~1 on a 515 base EcoRI rr~ t (SEQ ID NO: l, Bell er al., 1979) er~ ;.. g human p1~ in (SEQ ID NO:2) was ligated into the EcoRI site of rRI1ues..;l.l (Stratagcne, Inc., La Jolla, CA), genera~ing pBSllNS. pBS~lNS was ~ s~ with HinDm, loc~ted 5' of the insulin open ,, ~1ing frame, and BamH~ located 3' of the Insulin opcn reading ~ame. The reS~llting S42 base r.a6,..f .1 was ligated into pCB6 that had been l~ h d wi~ HinDm and BamHL ~ g pCB6/~NS.
pCB61~S was ~1i~st~d with Bgm and BamHl and the r~s1~1ting 549 base Ç~11e1lt co~ g the human insulin cDNA (SEQ ID NO:l) was ligated into the BamHl site pCMV8ARESJNEO~hGH PolyA generating pCMV811NS/IRESlNEO. The CMV IJlulllot~, drives 1~ r~ ~J ;~ of a bi~;~hol~ic .. ~ ~tl-E,_-RNA with human insulin ~ odcd in the ~L.~,~
open ~ading f~ame and the ne~ ~ resistance gcne c~-rodc~1 in the du. us~ , open reading W O 97~6321 PCTfUSg7100761 frame. Stable transfectants from this pl~cmid are selçct~-d in G418. The same 542 base Hu,Dm/BamHI fia ~ t was also ligated into HinDIIIlBatn~ ~ligested pCB7 generating pCB7/lNS. Stable ~ t~ from this pl~cmi-l are se~ctul in hy~lul~;in.

A third insulin e~,ession plgcmirlc was also cor~ ,d. pCB6/lNS was ~i~sted with Bgm and BamHI and the res~llting 549 base rl~Ln~ co.~ the human insulin cDNA (SEQ
ID NO:I) was ligated into the BamHI site of pCMV8/IRES/PURO~hGH PolyA, ~ ;..g pCMV8ANSlIRESlPURO. The C'MV ~lOlllOt~,~ drives h~s~ ~ion of a bicystrvnic RNA with human insulin enro~3f~d in the Up~ l open reaL~ing frame and the yulvlnyucill 10 ~.s~ e gene encorl~d in the dowlJ~ ull open reading frame. Stable ~ fe.~ from this pl~id are s~lrct~,d in ~,lllyoci.l.

Al~ , p~u~ote~ tili7~cl in hurnan in~ in e~vl~Ssioll pl~sn~ . The ratinsulin l ~ t~r L,~ t was isolated from pAClRIP (a d~,,ivd~,~rc of pACCMV.pI~A in 15 which the rat insulin 1 p,~.~o~. was s~,b~ ~d for the CMV ~o~,oler, Becker et al., 1994) as â
Kpnll~inDm r.r.l.. " (SEQ ID NO:12) coIl.~pQ~ i~ to bases 112 to +l relative to the start sitc of ~ ";l-~;Qr~ This fragment was ligatcd into KpnIl~inDm ~i~stcd pRl~lesc~irt (Stratagene, Inc.), genc~ating pBS/R~. pBSJRIP was ~iie~sted with KpnI, treated with Klenow r, ~ t to blunt the end, then ~lieçs~ed with EcoRl, g"..--,.1;,~g a 450 base pair r.~;,......
20 e~ t~ the rat insulin 1 ylOll~t~. This fragment was ligated into pCMV8/INS/IRESlNEO
that h~ been previously Aigçs~d with SpeI, treated with Klenow and then ~ligest~.~ with Eco~, generating pRIP8/INS/~ES/MEO.

The rat insulin l promotcr rl,~ .nt (441 base pair KpnI/HinDm fr~-nt, SEQ ID
25 NO:12) was also ligated into both KpnI and HinDm (1igrsted pCB6tl~S and pCB711NS
g pCB6/RIP.INS and p('B7/RIP.INS"~ ely. The CMV pr~ o~. fr~ .nt of bûth of these pl~cmitlc was ~ u~,ed by ~ligestin~ with SpeI ~d Bgm (l~l"o~,ng bases -585 to ~1 of the CMV promoter), treating with Klenow L ~t and ligating to close, ge~."~
pRlP61~S and pRIP711NS. Stable tr~ cr.~ .n~ of pRlP61~S are s~ t d in G418 while stable 30 tra~sr~ c of pRIP7JINS are ~~ 1 in l.~ cin.

W O 97/26321 PCT~US97/00761 The rat insulin 1 gene p~ l.otcr ~l~."n (R~) was also m(y~;fi~d in an attempt toslle~ n its ~ -.ccl;~n;~ activity. The p.;n~ modification involved the atta~lu~ t of varying nulllb~, ~ of mutant Far-FLAT miniP.nhq~lrMs (~:FE mini~r-h~nr~r) (Gçrrn~n, et al., 1992) to ~rf~.~nt posiri~)nC within an intact RIP or to a RIP ~ A at -205 (-205RlP). FFE
minienhancers were co~ lucted by gellc aling oligol.ucle~tides cGll~ g to the region of RIP ~t~.~n -247 and -196 (top s~nd, 5'-GATCCCTTCATCAGGCCATCTGGCCC~ l A ATAATC~ACTGACCCTAGGTCTAA-3' SEQ ID NO:29; bottom strand, 5'-GATC~ AGACCTAGGGTCA
10 GTCGAlTAl-rAACAAGGGGCCAGATGGCCTGATGAAG_-3', SEQ ID NO:30). The Imderline~ seq~-&'-''&s, at the ends of the oligor~cl~otides are BamHI and Bgm ~ccogni~ion sites.
The oligonucle~Lides were ~rlne~l~od and ligated in the p~ ,el.ce of l~cl.;c~iol c.~rllles Barr~
and B8m. Since BamHl and Bgm ~ e compatible DNA ends but can no longer be digested by BamHI or Bgm, the only mnltiTnPrs that cscaped BamHI and Bgm ~ stior were ligated 15 head-to-tail. ~ h~rcr dimers, trimers, etc. were separated by polywl~lamide gel el~:llvyhor~,~is and blunt~nd cloned into the t~nt ~ f-~c!;on vcctor, pBSlR~/hGH, at either a XhoI site i.. r.~ y u~l~,~ll of 415 of the intact RIP, into an AvrII site at -206 of an intact RlP, or into an ApaI site ;-~ ly upstreun of -205RIP. The ~ n~l and olic~ ;on of hq~rc, repcats were ve.;rled by l)NA se~lu~ ;..g. The stable l.- ~-f~l;on vector, 20 pFFE31R~8JlNSlIRES~NEO co..l~ g three copies of ~-~-~ minienhancers (F:~;E3, SEQ ID
NO: 15), was gcnerated by in~erting a blunt-ended KpnI/HinDm ~-~31RIP into pCMV8/INS/IRES/NEO in which the CMV ~ t~,l was ~o~d with SpeI and SacI.
pF~6/RIP8/INS/IRES~NEO was cG..sll~u;~d by inser~ing an ApaVblunt-endeWinDm FFE6/RIP r.~L,....,.t into pRlP81hGH polyA in which RIP was l~ lu~d by ApaIlEcoRV. A
25 BgmlSt~ NS/IRES/NEO ~ ;J~ t was then ins¢rted into p};FE6/RIP811hGH polyA to co...
p~ 6/RIP8/INS/IRES/NEO~

The rat insulin 1 gene intron (RIPi) was obl~l.cd by ~Ul,~lnC-dsc chain reaction from rat g~.nnmiC DNA using oliE.J~--e~ ;des ~ ;AAGC~AAGTGACCAGCIACAA. SEQ ID
30 NO:3 1 and GGGCAACCTAGGTACTGC3AC~ l ATC. SEQ ID NO:32. I'hese oligos .

~lo~lu~ed a 185 bp product CO-~Ainine the 119 base pair RIPi (Cordell e~ al., 1979) and a Hindm site on the 5'-cnd and a BamF~ site on the 3'-end. The PCRTM product was Ai~sted w~h HinDm and BamHI and ligated into pNoTA~7, ~ (JQ it was rcllw.~,d with XbaI blunt-ended with Klenow, treated with HinDm and inserted into EcoRVI~inDm ~1ige~A
S pR~P8~Sl~ESlNEO to generate pRIP8/R~i/INS/IRESJNEO.
pF~;E6/RlP8/RIPi/lNS/IRESINEO was consLI~ted by repl~cing the 5' adeno~us-i... oglobulin hybrid intronllNS/IRES of p~6tRIP81INS/IRESlNEO with RlPi/lNS/IRES
from pRlP8/RIPi/lNS/lRESJNEO. p~RIE)3/-8SRIP/~i/lNS/IRESJNEO cun~ l three copiesof the rat insulin 1 gene enh~ ,r (RIE~ fused to RIP tluncated at -85. This pi~cmirl was lO cons~ ted by reFl~rin~ a BsgRV~inDm RIP fragrnent from pRIP81RIPiJlNS~ES/NEO with an ApaI/~inDm (RE)3/-85RIP r.~5,-~ Both the BsgRI and ApaI ~ ;nn sites were blunt-endcd by Klenow t~~l~

The 2,000 base pair Class m human insulin-linked polymorphic region (IL PR), a region 15 d~ ncl. t-nA to enh~n~ C~ ;Q'-ql activity of the human insulin ~ otcr (~nn~ly et al., 1995), was ob~ ed from the phage lambda clone l-Hl-3 (O~ l.ach and Aagard, 1984). A
PstIlNcoI r~ "- co--lP~ g the :ILPR was treated with Klenow and inserted into a blunt-ended XhoI site ;~ ly uy~ ,~ll of RIP to rnake pILPR/RIP8/INSllRES/NEO. On~.nt~tion of the 14 bp repeats in the ILP~ with resE~ect to RIP was ~k-t..,...;~ l by DNA sequenring.
The human glycerald~h~de-3-~hosph~'e dah~llo~,_nase ~IOlllOtC~ (GAPDH) was isolated by tne ~ e cbain leaction from hllm~n e~ -.;c DNA using oligos (GGGTCTAGAGGA~ ~CCACCG, SEQ ~ NO:33 and GCCGAATTCGAGGAGCAGAGAGCGAAGC, SEQ ~ NO:34). These oligos gene,~d a 1143 base product cc~ ,s~om:ling to bases -1121 to +22 of the publich~l se~u~ e (Ercolani e~
a~., 1988) with the inl.c~ I;on of a unique XbaI site at the 5' end and a unique EcoRI site at the 3' end. The PCR~M p~ was ~lig~'St~l with Psd (located at position -73S relative to start site of transcription), trea~ed with Klenow, and then ~lig~stPd with EcoRI. The resllltin~ 757 base r,.~ .t was ligated into pCMV8~NS/lRES/NEO that had been prcviously ~i~sted with SpeL
treated with Klenow and then digested vith EcoRI, g~ ling pGAPDH8/lNS~ESlNEO.

CA 0224643l l998-08-l2 W O 97/26321 PCTrUS97/00761 The ~ous Sarcoma Virus Long Terminal Repeat (RSV) was isolated from pREP4 (In~ ogen, Inc., San Diego, CA). A 637 base pair SalVPvurI r~al5~t c0..~ g RSV was ;~Q1~t~d treated with Klenow to blunt the ends and ligated into pCMV8/lNS/IRES~NEO that had 5 been previously f~igested with Spel and EcoRI and treated with Klenow, generating pRSV8/INSJIRES/NE0 .

Bic,.~ or Inoc~ on and Culture. EP18/3El cells were grown, split, and .~.a~ rA in RPMI-1640 I~ ;.. with 2 rnM gl~ (JRH Biosci~nce, T ~nf~Y~ KS) suppl~rn~nted with5% fc~al calf serum (JRH) and 0.125 ~glml G418 (Gibco BRL, t~ h.. -~,.. g, MD) in T75 culture flasks as describcd previously. A large scalc bioier~tor (('çllig~n PlusTM, New ~lul~snrick .Sci~n-ifir (NBS), FAison~ NJ) with dissolved oxygen electrode, pH elc~,hode (both Ingold), and 4-gas p.o~ ional-integral-derivative (P-I-D) controller is set up for pe~ r~;o~ culture with a pacl~d bed of polyester discs (Pibra-Cel~, Sterilin, F.nglr- ~l) and a cenhifugal lift im r (Cell 15 Iiftm, NBS). The reactor has a woriing volume of 1.25 liters and a packed bed volume of 0.7 liters cQI~t~ 70 grams of polyester discs. Cells are l~y~S -~; fd and seeded into tne rcactor c~nt~ the same media co~ oC;Iio~. as the ,.~ n~ e media at a density of ~>loAimately 106 cells per ml of working ~l~lc. After transfer, the cells are allowed to seed onto the bed material for 8 h with a low impell~r speed of 50 rpm and no media perfusion. After see~in~, the 20 ;~ e~ speed is l~ou~ L up to 80 rpm and the culture is ...a;..~ rd with no perf~cion for a~~ ately 75 hours. Media p~ ~ fi,~:ol- is started and tne flow rate is bluugllL from 0 working 1ICS per day (WV/d) to 4 WV/d over the co~se of the following 500 hours. The ~- fi ~
ra~e is ll~rc~L. I,,A~ col-~ t at 4 WV/d. The pc.rusio~ media is RPMI-1640 with 2 mM
~J..t~ ..;..~ which is then supplemented with 2 gll ~ cose (final con~ tratiûn of 4 gA), 0.10%
2S fraction V bovine seram ~lbllmin 10 ~lg/ml human apo-L, ~ ~fc-~ , 50 ~M each of et~ ol~m~
and o-phosp}..l~rleth~nn1~ " and 0.10% cholesterol rich lipids from adult bovine serum (Clark and Cbick, 1990) (all Sigma Ch~mir~ S~. Louis, MO). The p- ~ r..~,O.~ media co~ .c no fetal calf scmm or other full sera. At a~plv~ ately 600 hours of cultu~e, the media was further e .. i~rd wi~h glucose to a final cc~ t,ation of 6 g~l. The i~rell~r specd was h..,~ to 100 30 rpm after 200 bours of culture. to 120 Ipm after 3~0 hou~s, and to 150 rpm after 700 hours. The ~44 W O 97/26321 PCTnUS97100761 cultu~e t.,l.lp~ , is ln~;n~.nr(l at 37~C, the dissolved oxygen level at 80% (inA~Y~ relative to saturation of air in 37~C pho~ tf~ r~,~d saline), and the pH at t.4. t}lucose levels in the reactor are Tn~ .;n~ in the ranKe of 1-3 gll by ~Iju~ g the },e,fusion rate and the glucose con~ t~ation in the frcsh perfusion media. Cultures have been ... ;~ d ~u~ fi~lly for as 5 long as 2000 hours in the bioreactor under similar contli~io~c.

Media c-.nl~le5 were collec~d once daily and ~ q~ ely analyzed for insulin se~ ted into the media by ELISA as previously described. Selec~d ~~--. ' s were clualitatively analyzed for insulin pl~ce~ by HPLC analysis as pl~,~io~,slr des~ ed ~m~n-~ni- and lactate levels 10 are .-.-~--;t~ d in the daily si mrles and analyzed using an ~ O~ t~ analyzer (lBI Biolyzer, Jo'nnson & lohncnn~ New 13.ul.~wick, NJ).

At the end of the culture, Ihe reactor is opened and a l~,ple~ e l.~- ..hf r of polyester discs are s~?le~ for ~ nl~ n of DNA and insulin cont~nt Cvclicallv Stirnul~ted Sec~ in tne Bioreactor. At a point during the culture when the oxygen contloller output has st~hili7~d around 60t the culture is cyclically stim~ tçd with ~ ;onofa10Xcor~f~ tedsc ~ n-s~tim~ ncocLt~ilonceevery24hours. The-I-lit~
of the coc~t~il yields final m~Aillm cor~ ;on~ of 10 mM each of leurinç, ~gh~ne, and 20 glutamine, 100 ,uM IBMX, and 100 }lM c&l,achol (all from Sigma). At the bc2~jnnil~ of every cycle, d~p~o~ utely oDe-tenth of ~e worlcing volume is replaced with the lOX co~ il while the p..r.-.,..~ of fresh media is left ~ k--.g~l At 4 l/d of ~.ru~On, e.g., the ~ inil~
con~ ol of coc~t jil after 24 h i6 less than 2% relative to the initial c~u~ Lion due to the continuous rlihltion by the pe.r~-cio-~ Six samples were talcen every 30 ~ s, then four 25 sa~les every hour.

StaUe ~ sr~-lio~ of cell lines. Cells were ~.~sç~lcd by clecl,u~ ion as ~les~-~i~d above for d~e Hexol~n~se 1 knockout ele~opolations.

W O 97/26321 rCTrUS97/00761 Insulin messa~e primer e~ Ic n~ioll analysis. Total RNA from RlN cell lines grown in vitro was isolated using RNAzol B RNA Isolation Reagent (CinnalBiotex ~boratories Int.). Total RNA from RIN cell lines grown in vivo as tumors was isolated using Tri~eagent (Mole~ r Research Center, Inc.). Ten ~g total RNA was hyb~i-1i7ed to a 5' ~li~xig~nin-labeled oligo (GCCAGCAGGGGCAGGAGGCGCATCCACAG&GCCAT. SEQ ~ NO:3~, Genosys Biot~hnologies, Inc.) in 0.25 M KCI at 68~C for 15 min. This oligo hybri~li7~c to the first35 bases of the e~ o~ .. uc rat insulin I as well as the human insulin open reading frames. Primer e~ ;o~ nc were then ca~T ed out with 2.5units AMV Reverse Trar c~ P in the d buffer (P~V~ ga, IrlC.) suppl-m~ntPtl with 0.8 mM dNTP's (~h~ Inc.) and 100 10 llg/ml A~ w~lly~ D (Sigma Ch~mir~l Co.) at 42~C for one hour. F~tçncirn ylv~hlClS were ated,res~lspçn~e~in~% w~/60%~o~ e, heated to 1~~C for ~ min and ~n on a 89rO acrylamide/7M ureallX TBE ~ gel. El~ll~yhol~ed ~,o~ were Ir~ .~r~ d to Qiabrane Uncharged Nylon Me~l.lu~c (Qiagen, Inc.) using a Tl~ .hol Unit, TESOX (Hoefer, Inc., San ~4isco~ CA). Digoyi~in-labeled products were ~ete~tcd using the Genius 7 Non 15 l~lio~ e Detection System (Boehnr~r ~nnh~im) followed by e~p~j~c; to Xornat-AR auto radio~hy film (Kodal~). Primer l~Yt~nci~r of endogenous rat insulin I ~ ,n~-~.tes a 91 basc p~lu~,t (Cordell et aL1 1979), the human insulin tl ~ - e~pl~cd from pCB6 generates a 101 base ~ludl.lcl~ and thc human insulin tr~n~ n~ e~l"~ ssed from pRIP7 gene.a~s a 68 basc 1,.~1~1. Primer eyten~iorl of the human insulin l.~n~ e eAy~scd from 20 pCMV81INSlIRES/NEO generates a primary signal of 280 b~ces with three other minor prema~ure ~ t;on signals of a~ ~ly 190, 130 and 115 bases.

N ~ analYsis. To~ RNA wac icol~tçd as d~r~rihed above for the primer e~ t~ ,~r;~ n protocol. Ten llg total RNA was recolved on me~yl 1.l~ /1.5% ~,_u3e gels as ~1~s~;k1 25 (Bailcy and Davidson, 1976). Gels were subs~ 1y stained with eth~ m bromide (1 ~glml in 0.5 M a~ .. carbonate) to vis~ i7e RNA for ~l~t~ and loading conC ~t n.~r RNA was electro ~r~ ~f~ d to nylon ~r.~k~ rs as described for the primer e~t~n~i~n protocol.
Mcrnl~ranes w~e hy rd with digo~rige-nin-labeled RNA probes using the Genius Non-Radioactive NL~clc,ic Acid 1 ~kling and Det~c~inn System for filter hybridi_ations as f~eS4' ;h~

(Ro~hrin~er ~snnhrim). Full-length digoxigenin-labeled ~nti~ence probes co.~ ollding to human insulin, rat glyceraldehyde 3-pho5l.k:~t~ de}~l.og~,nase (GAPDH) (c~,~ to bases 21 to 1162 of p~lhlichP~1 scquence, Fort et ~l., 1985) and the neomycin resistancc gcne (control template supplied in Genius 4 Kit) were made using Gcnius 4 RNA 191 P.lin~ Kit (Boek.;.~G~.
5 Mannheim) using either T7 or T3 PO~ C. EAYO~S of chemil~..;..rsoent ~ietPc~d membranes wcre p~-r~ using Xomat-AR autoradiography film (~r19~ In some cases, blots were hybridized with a 32P-labelcd cRNA probe for human insulin.

Sti~ insulin secretion assav. Four million RIN cells were seeded in 9 rnl media in 25 cm2 flasks (butyrate-treated cells were secded at 106 cells). Cells were then cultured with daily media cll~ng~s for one week with or wilho~ll 1 mM bu~ te until cells l~ached 70-80%
CQ~ r~f. Prior to assay, cells were incubated 2 times for twenty l.. ~ ~t~s at 37~C in RPMI
media lac~ing glucose and su~yle~ t.n-cl with 0.1% BSA and 20 mM HEPES, pH 7.2. The basal incubation of cells was for 1 h at 37~C in 4 ml RPMI cor~t~ining O rnM glucose, 0.1% BSA, 20 15 rnM HEPES and 100 IlM diazoxide (Sigma rhPmir~l Co.). Basal sec n ss~npl~s we~e collP~ted and aliquotted for insulin ;,.. ~o~cc~ys and HPLC analysis of insulin speeies This was followed by the stimulated incubation of cells for 1 hour at 37~C in 4 ml RPMI with 5 mM
glurose., 0.1% BSA, 20 mM HEPES, 10 mM each IP~C jnP ~gininc and gl."~ 100 ~M
cdul~cLol (Sigma ~hPtnir~l Co.) and 100 ~lM IBMX. Stim~li ted se.,lc~io.l s~..ples were then 20 colh~ct~P~ and ~ ottp~ Cells were l~ d to a basal ;~ ion for 1 hour at 37~C in 4 ml RPMI CQ~ 'g 0 mM Plllcose, 0.1% BSA, 20 mM HEPES and 100 ,u~ 7.o~ ;A~.

Cells were then col~ct~ or ~i~ t. ~ ;on of insulin content and cell ~ l~r by ~Ar1ition of ~DTA to ~e media to a f~al cn~ .t.a~ion of 2 mM and pipc~ ,g up and down to remove 25 cells. Twenty percent of the cell s~ .c;on was taken for de~ OI~ of DNA contPnt l~e Ic- ~ er of the sample was cc- .~ ged at 220 x g for 5 I~-;-- ~res to pellet ~e cells. The cell pellet was ~ ~ p. .l~f d in 0.5 ml cold 0.1 M acetic acid/0. 1% BSA and s5)nir -~1 on ice (Setting 2, Sonic D~ kJ~xor 50, Fishcr .Sc~ ntifir. Pittsburgh, PA). The sonicatc was ~ Qtl~d for insulin ;.-..-.---.nqCc~ and HPLC analysis of insulin speries W O 97/26321 PCT~US97/00761 Deti~ l;on of DNA content and cell n~ ber. RlN cells are pe~ te~l and PBS
removed. 0.5 ml of DNA e~ctraction buffer (2.0 M NaCl, 2.0 mM EDTA, 40 mM pl~os~h~e buffer, pH 7.4) is added to RIN c .~ s and the RIN cells are sonir~~ on ice, for 30 secon~ls at 5 -30% power (Fisher 50 watt soni~ . Four mi~olit~.~ of sonic~tP are then diluted into 1 ml fresh DNA assay dye solutio~ 10 mM Tris, 1 mM EDTA, 0.1 M NaCI, pH 7.4, co..~;n;l~ 0.1 ~lg/ml Hoechst dye 33258 (Pol~ .ces or Me1~ r Probes), with calf thymus DNA as a ~dald (Clontech ~c.). S~mrles are read using a DNA n.~..;n.. t~. ffloeff~r Sciçntific Llslr.. l-~c, Model TKO100~. 6 ~lg genomic DNA per 106 cells was used for the 10 co..~ ion from DNA content to cell ~ er values.

HPLC analvsis of insulin ulocessi~ ;ates. Acid/ethanol eA~ t~ of whole cells or conditioned media was prepared a~d analyzed ~y high ~Ç~" .-. re liquid cl"on~- to~;. aphy as d~ l (Halban et al., 1986, Si7nn~nko and ~J~lh~n, 1991). I zclive insulin (lR~
species were ~Iv--~l;l; t~ by r? 1;~ cc~y as ~5- ~ ibed (Halban et al., 1986).

Tumor formation of ~ r~ c~ R~ ccll lines in nude rats. Si~c to 8 week old athymic Fisher nude rats (Strain P344~Ncr-mu from N;Lonal Cancer T..e~;l.,t., rl~icL MD) were housed in a stcrile isolation facility with free acccss to sterile standa~d laboratory chow and 20 water. Three million cells were injff~ u~ ~ cously at two di~f.,.ent 1QC ~iQn~ in 100 IU1 PBS
using a 1.0 cc U-100 insulin syringe (Becton D~ n). Tumors were e~ci~çd once they were palpable, excess fat and associated dssue ~l;C~cn~rl away. S~ s frozen prior to ~ S~
Body wdght and bleeds for blood gl~ ose ~ ation were taken prior to ;.-je~ cells and ~U~6}~ the course of the e~ Blood ~ 4SG was ~ d using an IBI Biolyzer25 (Kod~ Pq~tmqn Chemical Co.).

Results:
Rat ~ no...s cells have been e-~ ,..,d to ~oduce hi~Jh levels of human insulin. The RIN cell line was derived from a radiation~ u~cl tumor (Gazdar e~ al., lg80). The insulin 30 se~,l.,tu~r cl~ s of the pq~tql cells uscd in thcse st~ s, RIN 1046-38, have been W 0 97/26321 PCT~US97100761 ~esr~ihcd and shown to exhibit abnormal sensitivity to gl~l~ose (Clark et aL, 1990 and Ferber et aL, 1994). These cells secrete insulin at ~ rose Co~CC-.I.aliOnS of 50 ,ul~f, s~ ~ti- ~ 2-10 ng rat insulinlmillion cellslhour. This level of insulin is weU below levels p~ oed by primary rat or human islets (Rhodes and Halban, 1988 and Marchetti e~ al., 1994) or other reported rodent inCIllinQm~ lines (Asafari et al., 1992, K~ et al., 1994). RIN 1046-38 cells were stably transr~tcd with an e~l-~ssiol~ pl~mi~ CC~ Ai~i"g a human insulin cDNA driven by the human .;yl~ )ovirus promoter (pCB6/INS). One clone, RSC.I-17, was S~lGCP'~ based on high insulin s~;l.,t;on and further cha~ I FIG. 4A. shows the total ;---- - ~ , insulin content as measured by RIA of RSC.I-17 versus the parental RIN cell line. RSC.I-17 has a total insulin 10 content of 450 ng per million cells, 3-fold above parental RIN.

Chronic culture of rat inclllinl rn~ cells in sodium l~u~ldte has been shown to incl~e endogenous insulin ~ ae~ contcnt and sec.~ l (Swarovsky et al., 1994). To d~ if simiiar i..e~ascs would result from a hurnan insulin IlAn~;g~ in rat ins~t~ oells, R5C.I-17 15 cells were cultured for 7 days in 1.0 mM sodium butyrate. Cell growth was .eld~d but c~ ,ed over the course of the week. Insulin content was Act~ rd at the end of the week and showed a 3-fold in ~c~e per cell above the untreated cells (~IG. 4A), CO~ t. ~1 with data on the ~,I.,dse in content of endo6~n~h~e insulin (S. ~.;.~ et al., 1994). This higher level of human insulin content ~..g~ that the RIN 1046-38 cells are capable of s~mth-~ei7.ing and~0 stonng more human insulin. Sodium bul~ treatment is also used to ~n~i~ntly induce insulin n in the large scale bi~,c~lor.

To i~ ,ase the level of ~ cî;~n of human inQ~llin R5C.I-17 cells were stably transfected a second time with pRIE'7/lNS. This cAy,~ssion pl~cmiA utilizes the rat insulin 1 25 ~ Ot~,~ dnving ~ ssioll of hum-n inQIl~in EPl 1 3E9 was ir~ u;f.t~l based on an in.,~cd insulin y,lod~ above R5C.I-17 and ~ t ;~ed further. The insulin content of EPl 1.3E9 is 1400 ng per million cclls, four times higher than its parent, R~C.I-17 or RIN (~;IG. 4A).

Human insulin, like e-~ot. ~ rat insulin~ is scc~ ,d via the regulsted pad way in the 30 çn~ t-- ~,d RIN cell lines. Insulin sccretion values for a one hour pre-i~cl~b. ~ in buffer alone WO 97126321 PCT~US97/00761 -(basal vâlues) followed by a one hour inrubqtion in a cor~tqil of IB~, glucose and amino acids (stimulated levels) are shown in FIG. 4B~ Low basal insulin secretion is seen from R5C.I-17 and EP1113E9, even with the human insulin ~ cc....~ .Yl;~,ely e~ ,sed by the CMV
promoter. A higher bas l secretion is seen from the b~ ale-treated R5C.I-17 cells. However, in .11 Iines, insulin release wa,c significs~ltly inc~eased following stimlllqtion to levels of 150, 425 and 4~0 ng per million cells per hour from R5C.I-17, ~.lt~.dte-treated R5C.I-17 and EPl lJ3E9, .pc.,~ ly. Stimulated insulin relcase per hour ranges from 25 to 35% of t.he intr"- e.~ 9r stores for all four R~ lines, a value concictent with primary islet data (Cwry, 1986 and Li et al., 1994). RSC.I-17 has ..,~;..r~;. ed its insulin output through more than 100 population ~ol~blingc 10 willluul drug sPlectiol (a~pro~ ly one year in culture).

Results from Lr ~ n~ir ,qnimqls (Scl..u~ et al., 1994) and from cell lines (Halban and WollhPim 1980) have ~ut>~oll~d the idea of a physinlogi~ -q-l set point for insulin pro~ tion in ,B-cells. However, a threshold or upper limit on insulin prodl~cti~m in the current Pn~
15 RIN cells has not been obs~. ~ed.

Human proin.c~lir is e~r~ y ~,.o.~ss~d to mature insulin bv ~at i~lculinomq cells.
Intracellular insulin species were isol~ted from parental RIN, R5C.I-17 and EP11/3E9 cells by acid e~traction. Sep - ~ ;on by HPLC of the insulin species ~.od.lccd by these oells was done as 20 ~es~,~ccl(Halban etal., 1986tSi7nl~-nl~andHalban,1991~. Theanalysisi~ r~tcsthathuman insulin produced by the rat in.cnlinf)ms is effiriently plucccced to ma~re insulin with very low detectable levels of pro-insulin or othcr ~ç~cc.~ eliates (FIG. SA, FlG. SB, FIG. 5C).
~ e of human insulin is slightly less effi~-ient compared to the p,~c~ g of rat insulin at these levels of ~ Whilc the ~~ of intracellular rat pr~i~c l1in ~md in~ tes 25 is 3 to 4% of total rat insulin in all ¢ell lines e~ , the ~c~ ge of intraccllular human proinC~llin and ;nt. ~ ~ L tes is 8% in R5C.I-17 and a~plO~ CS 18% in EPl 1/3E9. The ability ûf RIN cell lines to efficiently l).oce,ss prrinQ~llin to ma~re insulin in these e~ e.~d lincs - s~dtes the ".~ t~ re of the high levels of e,~ssion of the cndo~.ut~ases PC2 and PC3, known to be ~ y.:~hle for insulin pl~C~ (Vollc~-.cidcl e~ al., 1995). This is an 30 impor~ant feature of thc RIN cell line~s being de~ ,l~cd.

CA 0224643l l998-08-l2 W O 97~6321 PCT~US97/00761 E~cpression of human insulin tr~n~gen~ is stahle in vivo. RIN cells injert~d ~ubc~ rollsly into a nude rat will form tumors. These tumors can then be Çxrice~ and analyzed for gene e~pression. As has been seen previously, the majority of the tumor rnass is R~ cell in 5 origin with only small ~ ..ke i of cells being host-derived in the form of endoth~ m, fil;loblas~, etc. (Schnedl et al., 1994). This allows for a co~ ie.~t model for analysis of both ~ndGgen~us and tr-n~g~n~ ex~lessio~ in RIN cclls in vivo. In the ~b~n~e of any r_i,l.ainl of tumor growth, time points are n;~icled to onc to two months because RIN cells secrete inc~,~ing ~molmtC of rat, and in our studies human insulin, leading to hypoglyc~uia. Blood glucose levels wcre .. c,.. ;lofed ~ UE~houl the course of the e~ ";.. t One and a half rnillion RSC.I-17 cells were inject~A s~ cly at two sites per animal. The ~nimqlc quickly bcco-l~ h~l,o~l~.,~c in 8 to 10 days following this dose of cells (FIG. 6). Following the onset oif h)rpoglycemia, ~nimqlc are ~ ;ntA.~rA on glucosc in their ri~ ng water as well as IP glucose injc~ionc prior to surgery. One animal ~ ed a blood glucose rebound following removal of both tumors. This rebound after two days of e~Gg:Q~ c insulin-in~leed hypo~iyL~l~a iS fo~owed by the rapid removal of the exogenous insulin source.

Nine tumor masses in all were isolated from four ~nim~l$ ~L~ ell days 13 and 31. These tumors ranged in size ftom 40 to 2~0 mg in wet weigh~. RNA was ~ ted from tumors and tA~Ssiol~ of sevcral geoe plU lucl~ was analyzed and compared to ~ t~d cells "~ r~ in tissue culture (in vitro). Primer e~-Q~;Qn analysis was used to compare the rat and human insuLin signals in the same samples. The same primer hyhri~li7es erfi~ to both m~scs~c~
but upon primer eYtenciQn gives two ~3irr~ size products which can easily be resolved and quantitated. The results of this analysis are shown in F~G. 7. The same srn~-lmt of starting RNA
was used in each rcaction from cither cells maintained in vitro or from tumors, but cQnt~
of non-RIN cells may cause the rat and human insulin signals to be unde,~ ,s~-t~ A in the tumor samples. No attcmpt was made to cor~ect for this. The signal for the human insulin l .~
driven by the CMV ~lU~llOt~r w.as very con~,l throughout the time points eY~min~rl No 1~1 W O 97/26321 PCTrUS97/00761 ;nn of signal was a~p~enl, suggesting the in vivo en~ c~llcnt had no ~ t~ iULIS effects over the cour.e of this e~ S~ ~l.seq~ e~ s have analyzed tumors at time points of 48 days ~versus 31 days here) with no .~ ;mll;ol~ in CMV driven/hwnan insulin trqncg~np e,~ ,ssion.

To further test the stability of the CMV driven tr~n~nPs in vivo, e~ Pe.~,d RIN cells werc in~rl~nted into nude rats and transgene C-.plc~Siull A~sesse~ with tirne. Several i~ ~ nrlent cell lines were ir~rlqnrPd into mlllti~ imqlc cA~.e,ssing at least three lirf~ t tr-q-ncg~n~s The use of ;l~d~.c- ~tlen~ cell lines with ~lirr~.e.,- hltl~;ldtiOn sites should give an ~nbiqce~ answer 10 to the issue of CMV promoter stability in RIN cells in this particular modcl. Longer time points of 48 days have been analyzed with no redl~cti~-n in CMV driven e~L~l. s~ion. The in vivo model of ;~ 't;~B RIN cells into nude rats is limited by the u~lcont olled growth of the RIN cells as a nlmor. All the RIN lines used here make endogenous rat inclllin, with some also ma~ g human inculin so that the q-nitr~qls quickly ~ h~ol~,ly~ ic. Efforts were made to r~ t~ the 15 ~nimql$ blood glucose by ~ ~in~-~ J;i~ glucose in the ~lrin~ing water or by i.p. injectiQn~, allowhg analysis of tumors at longer time points. ~ ly, lower doses of cells can be injc~l initially (3 x 105 rather than 3 x 106) which is how a 48-day tumor was gencrated and analyzed for ...~in~ e of gene ~,A}~ession. U~f(~llullalely, this lower dose of inject~d cells leads to a more sporadic tumor growth, m~'-ing it harder to generate s~mples for analysis.
Sul~lisi~gly, el-do~-lvLIs insulin ~A~1e~;,ion h.cl~,ased in all nine in vivo c~"plr5 exam~ned. Tbis was ~ .r-~ t~A since all nine tumors were eY~ ~e~ following pcriods of eAIl~,..e I~_~i~ co~ known to down regulate ~&~C..,dliC ~-cell insulin message ~i~ ng~ e~
al., 1982 and Brunstedt nd Chan, 1982). Compadson of the ratio of rat to human insulin signals 25 changed from 0.73 +1- 0.6 for several in vitro samples to 1.87 +/- 0.17 for the nine tumor samples. The 0.73 in vitro ratio col~ tes very woll with the ratio of rat to human insulin (1 part rat to 1.~ parts human) obs~.~ for the RSC.I-17 cell line (see FIG. S for mass ratios). The incrcased rat to hwnan .~.~sca~,~ sigl~al in tumors is parallelcd by an ~cl~,~d rat to human insulin content in tumors ~ul~ t~d to acid e~ctraction and HPLC separation of the insulin speoiles WO 97n6321 PCTAUS97/00761 A sirnilar result was obt~ r~l following injection of the high human insulin prodltrin~
cell line, EPll/3E9, into nude ra~s. Aoimals became hypoglycernic while ,..~ t~ ,g body weight over the course of the eA~ ~lt (FIG. 6). Nine t,umors wcre isolated b~,t~.CCIl days 1~
and 28 following inJecti~n Prirner ~ on analysis of RNA isolated from the tumors allows S for s~ ~ ion of three insulin messag~s in the EPl 1/3E9 line. Fl~t~ncjon of the en~ogeno~lc rat insulin ~ 5'~ aod the human insulin ~ ss~f driven by the CMV pro.llote~ produced the i~lP-nti~s~ pattem as seen in RSC.I-17, the parent cell line to EPl 1/3E9. A third eYt~nrion y~u~lu~;L
results from e~y.~;ssion of the human insulin transgenc ~ Cc~ee by the rat insulin 1 ylol~te,.
Primer e~ten~iQn analysis on the tumor r~ ~~. ~f:S as well as the cell lines ~.9il~ ~ in vitro show 10 hurnan insulin driven by the CM[V ~Ion oter is stable throughout the course of the in vivo e,.~ -- --t Again, elld~sgcnous rat insulin is upregulated in the in vivo e.~vi~ even in the face of hypoglycemia. The human insulin transgene driven by the rat insulin pl~ OtC. is e,.~ ssed ~ro~l~ho~lt tbe course of' the in vivo ~i~ r~ ~-t ~n vivo notency of en ~e~"~,d RIN cell lines. When the rng;~ ed RIN lines are growing as turnors in the nude rats, thcir secreted insulin cventually irnpacts on the blood glucose levels of the hcalthy anirnals causing l,~ c~ ia (l;IG. 6). While parental RI~ cells have enl10f ;,~ l~o~.c rat insulin outputs that e~ ually Icad to hypoglycerr~ia, RIN cells çl~ d to o~..,A~Ss hurnan insulin should induce hypo,gl~_~,~ia either faster with the sarne ~ b~_r of cells or require a srnaller tumor rnass. The in~ t~ d the twnor rnass needed to induce the initial hypoglycernia in a nude rat as an in-lirstor of in vivo potency of the engineered RIN cells.

Tumors were removed frorn the nude rats i~je~ with either the R5C.I-17, EPl l/3E9 or parental RIN 1046-38 at the first s:ign of h~ ly~,~ia. The time het~ ;t-jeel;on of a constant n~ber of cells to hypoglycernia varied from 12 to 13 days for RSC.I-17 and EP1 l/3E9 (~;IG. 6) to 28 days for the parental cells. All of the lines grow at the sarne rates in vitro. A plot of the tumor ~nass versus the in vitro stimlllated insulin secretion values for these lines (l:IG. 4B) is shown in FIG. 8. The higher the in vitro insulin output (both rat and human insulin), the smaller the tumor mass needed to cause l.,~ c~., ia.

W O 97/26321 PCT~US97/00761 Endo,eenous GLUT-2 eA~ s~ioll in RIN cells is lost in vivo. Thc eA~lessiu~ of several othcr gcnes in the tumor samples was analyzed and co~p&~d to RIN cells ..~ t~ ed in vitro.
The results of the analysis of both en~o~ o!~c genes and introduced l~ .C~ f S iS shown in FIG.
9. RNA from two ;.~ tumor s~ les from day 24 and day 25 were cv.-~ d (in vivo 5 sample) and compared to RNA from R5C.I-17 cells l"~ tA;nF~ in tissue culture (in vitro sample). lUesc~ee levels of endogenous o~ c~, h~r~ '-in~c~ I, a nylin, GAPDH, sul~llyl~.ia l~et)t~, and IPFl, as well as m~ssa~e levels of human insulin and the i~e~""~ein resistance tr~ncgPn~c~ were relatively !- r~ d following 24 to 25 days of in vivo ~assage of R5C.I-t7 cells. In cont~ct, the ~..csc~e~. level of r~ ,g~, ~uus GLUT-2 d~bv te~l in RSC.I-17 cells ... ~ .f~i in vitro is comI letely abscnt in tumors at the 24125 day time point. This result was duplicated on a sepa,a~ analysis for col.fi....~ n (GLUT-2 a and b, FIG. 9). Analysis of individual tumors from day 13 through day 31 ~ o~ ted c"~,essioll of GLUT-2 was alrcady abscnt at the earliest time point analyzed and ~ 9 I~i absent throughout the r~m~in~er of the ~ v.;---..~t Loss of in vivo GLUT-2 CA~l~Ssioll could have serious cQI-.cG~ P-s for the ~c.rv~ n~e of cdl lines ~eci~d for insulin delivery to ~nimslc or ~; nl~ with diabetes. Stable tr;h.~r~,cl;QI-of insulin ~ hlc;--g cell lines with the GLUT-2 cDNA has been shown to confer glucose-stimulated insulin serrerio~ (GSIS), while transfection of the same cells with a related L,~~ r, GLUT-1, has no such effect (~1l~h~5 et al., 1992, 1993; Fcrber et al., 1994; U.S.
Patent 5,427,940). ~ullhl~-...ore, loss of GLUT-2 c~ s~ion has been ~o~ d in a large u~l of rodent models of type II diabetes (N~DM) in which ~-cell failure involving loss of GSIS is a cause of h~ cemia (Johnson et al., 1990; Orci ct aL, 1990; Thorens et al., 1990;
Unger, 1991). Endogenous GLUT-2 eA~,c,ssion is apparencly down-r¢gulated or e~tin2~ h~
2~ under diverse l~h~r~ ClGe~lCal co~ c. In ~ lrlition to the studies cited above, in which ~nim~lc were h~t,o~lyc~i~,ic, irn~ ;U~ of noImal islets from db/- mice into diabetic, h.r~ ;..c.~lin~.mis db/db mice or db/- m~ce ,~ ,d diabetic and h~l,v;..~ ic by ;-.j~ of the ,B-cell ozo~ , resulted in loss of GLUT-2 t~-yl~csioll in the transplanted islets (Tho~ens, 1992b). These results suggest that loss of GLUT-2 may also be responsible for 30 . I,ahcd glucose fesl)on~;..re insulin release in human islets t~n~pl~nt~l into ~t;~ ,~tc with Type I

CA 0224643l l998-08-l2 WO 97/26321 PCT~US97/00761 or Il diabetes (Scharp et al., 1994). Rçdll~efl GLUT-2 su~ ssion in the face of ll~ ;lycemia is ~ g in light of reccnt studies demo~ at~l,g that the GLUT-2 IAU~llot~,. is activated by glucose in hepatocytes or the inclllinorn~ cell line INS-1 (Waeber et al., 1994). Overall, these studies strongly imply that the Gl,IJT-2 plclllot~l is advtl~ely effected by various metabolic 5 pernlrbations in vivo, and that this promoter cl~ r~t is not ap~liate for use in directing stable ~;.sion of heterologous genes in cdl lines.

FIG. 10 illllctr~t~.s that GI,UT-2 expression can, in fact, be ~ int7~inr~1 in RIN cells implanted into nude rats for relatively prolonged periods of time if the gene is stably transfected under the control of a viral plu-llotel such as CMV. A RIN 1046-38 clone cApn~ssing high levels of rat GI,UT-2 driven by the CMV plollloter was gw.c.~tcd using pCB71GLUT-2 (clone EP49/206) as previously des~ ~d (Perber e- aL, 1994). Animals injectcd with RlN EP491206 fonn solid tumors and ~co~--P h~Gglycc~ic, much as reported for animals ICCe;~ g cells co.~ only the c~dcg~ ol~s GI,UT-2 gene. Unlike the ~ rected cells, ho~ r, GLUT-15 2 mRNA levels are ...~;n~l~;. ~ at a high, cons~t lcvel over the two time points sampled, 16 and 34 days (FIG. 10A). These particular cells also were stably L~ r~:led with pl~c~niAc cQI~Ail~;"~
the human insulin and glucokinase cDNAs under control of CMV, _nd transcript levels for these other ~ncErn~s were ,. ~ ts; .rd in a stable fs~lion An~logo,~$ to GLUT-2. These results in~ir~e that cell lines tr~ncfecte~ with mtlltiple genes under control of a strong virS~l p~OIllOt~,l 20 like CMV are able to ...ti-~t~ stable e~plession of all tr~nc~c for p~ n~rd periods of time when tranc~lant~l into ~.~imsl~. These results are sull~lising and would not have been çx~l-d in light of previous studies from otber groups who have l~,pu.~d that strong viral ~,u ..-,t.,.~ such as CMV or RSV are often down-Iegulatcd in the in vivo ~,.. vi.~.. t (Palmer e~ al., 1989 and l9g1, ~Srl - r...~ et al., 1991, Chs~ ts~ and Kohn, 1994).

L~cledsed insulin pro~uction usin,e e~v~ssion plSlcn~As contSlinul~ an jnterr~l ribosollle binding site. A new insulin c~ylession plasmid was deci~d that linlcs the eA~l~,ssion of the drug scl~ n marker to the e~"~ssiùn of insulin. The pl~cmiA, pC,MV8/INSJIRES/NEO, utilizcs the CMV pro~r to drive a bicistronic m~scqgç containing the human insulin open 30 reading frame U~ &ll of the neomycin ~ "C open reading frame. Placed bctween the two W O 97n6321 PCT~US97/00761 reading frames is a portion of the S'-tr~n.c~rihed leader of the gene en~ot1in~ hurnan Glucose RG "l~t.~ Protein 78 (GRP78; Ting and Lee, 1980. The 5'-leader of human GRP 78 has been shown to direct intPrn~l initi:~ion of t~ncl~tion (IntPrn~ ibosol..r Entry Site, IRES) allowing for construction of functional pol~ nic genes in ~ n cells (Macejak and Sarnow, 5 1991). ln order to ge~ Neo-resistant cells with this pl~cmitl, the human insulin ...~c~,., must also be present, h.. ~ g the nu~c. of RlN clones that express human insulin protciQ.
Since int~ l initiation of tr~n~l~ti~-n by ~ES e~ is less çffi~ nt than norrnal 5' cap-~ep~n~nt initiation (Macejak and Sarnow, 1991), cells must express high levels of the bici~LIullic tr~ncgçnp in order tO survive drug sel~rtion. In this way, it should be possible to 10 directly select with G418 clones eA~ ssing high levels of human insulin.

Twenty-nine i~ n~ G418 l~~;C~ t clones from an clecllopv,~tion of parental RlN
1046 38 cells with pCMV8/INSJIRESlNEO were sc,~ cd for insulin content following acid t;A~ n as tlCSC.ihe(l The results are shown in FIG. 11 with the insulin content of RSC.I-17 (450 ng/million cells, FIG. 4A) used for C<51~ 'CI~. Twenty-nine out of 29 clones e~ ss~
~ct~l~le levels of human insulin with at least 10 out of 29 of the clones (34%) eA~lcssing Ievels of hum n insulin more than 2-times that of RSC.I-17. RNA was isolated from the S
highest insulin p~o~ g clones and human insulin '~'S$9'gc, ana1yzed using primer eYt~nci~n Star~ing inputs of 10 and 3 llg of RNA from these 5 clones, as well as from a polyclone from ~s 20 ele.,~,u~o..~tion, were cu~lpzl~,d to 10 ~lg of RNA from R5C.I-17. In the S monoclones as well as the polyclone, bigh levels of buman insulin m~cCaf~e were ~et~te~l at tbe e~ l size of 280 base p ~rs with tbree other minor premature t~ s~tjon sign ls of a~o~il lately 190, 130 and 115 bases. Even with 3 ~lg of input RNA, the human insulin signal is still comparable to the signal from 10 llg of RNA from R5C.I-17, a levd of human insulin .~ c~e~ in these clones in 25 line with the higher levels of insulin protein.

One clone, EPl813El has been further cha~ t~ d The insulin contcnt of EPl8/3El is 1300 ng per million ceUs with a sti~ insulin secretion rate of 500 n~/million cells/hour.
These levels of insulin are comparable to those acLe~,d in EPll/3E9, our highcst insulin 1~6 wo s7n6321 PCT/US97/00761 p,o~ ;.,g clone to date (FIG. 4). However, in co.~ ( to previous msulin produrin~ clones, EP18/3El and other high insulin prod~lc in~ clones (FIG. 11) were generated from one round of cle~;L,o~uldlion using a single e~ ssion pl~cnjrl The utility of the c~,ssion plasmid pCMV8/lNSlIRESJNEO is in both the high l.l,...kc,~s of positive clones and the higher insulin 5 outputs of individual clones. Also, only one drug selection rnarker was used as opposed to two in E,~ g EPl 113E9.

Introd~Gtior of a second human insulin ll -~e~.le into R5C.I-17 cells produced 1113E9, a cell line with higher insulin prodnction Sirnilarly, a second insulin corellu~ l was e~ .,ssed in 10 1813El cells to produce clones with h~ ascd insulin output. The cons~uct col.c~.cl~i of the human insulin gene linked to the ~JUlU~ in rcsistance gene and the ll~scli~ion of the bicistronic In~s~q~ plo~ced is controlled by the CMV ~ llo~,. Colonies of cells that grew after selection in 2 ug/ml of puromycin were s~ncd for in.,l.,ascd insulin output. FIG. 12A
d~ t~,s the e~.~.e"sion of hum~n insulin RNA of both bici;,l.unic transgenes, and the iL~.lCL~sC~ insulin content for 5 sele~ clones. The cell line EP111/220 exhibited tbe bighest cellular insulin content (FIG. 12B) and se~ t~d the most insulin. The EPl 111220 clone when incubated with the stirnulation cor~il of mi~ced ..~ and secretagogues (as for FIG. SB) seel~d O.g9 llg insulin/106 cells-hour. Cu~-_~ly, EPl 111220 l~l~lc~l~ the highest ~ -,nted insulin secretion of our cells eng;n~c~d with human insulin.
The insulin content and sec.ct~l ~ output of human islets may be cstimated from reports in the }iteralure . The a~ ~e human pancrcas contains about 0.9 g of islcts (K. Saito et al., 1978) wbicb equals 9 x 108 cdls (ri.~good et al., 1995), and the ~l,c.age hum~n pancrcas co~ .c 200 U of insulin (with a 3-fold range; Wrenshall e al., 19S2). Thus, when in situ, the insulin contcnt 25 of the a~ ge human islets ~t~ uately 0.22 U/106 cells, or 8 ~lg/106 cclls. Freshly isolated human islets are .~o~d (Eizirik et al., 1992 and 1994) to contain 8-10 ~lg/6 ~lg DNA (6 ~ug =106 cells). The same authors report ~at after one week of culture human islets contain 4-5 ~lg/6 ~g DNA and with stim~ tion secrete 459 ng/6 llg DNA/h. Freshly isolated rat islets, for comparison, are l~,~rt~,d to contain ~8 ~lg insulinl6 llg DNA, and with stiml~1qtion secrete 0.2 to 2 ~lg insulinl6 llg DNA (Tokuya~lla et al., 1995; Rhodes and ~qlhqn, 1988; Nielsen, 1985~. The f~ rl;onslly-norrnal mouse ~-cell lines secrcte 400-800 ng insulin/h upon 5tim~ ion~ and contain 3-10 llg insulinllO6 cells ~Miyazaki et al., 1990; Radvanyi et al., 1993; Knaack et al., 1994). The values yl~s~nted for human islet insulin content and secretion are ç~pect~l to S l.,ple~ellt the higher end of the range, ~c~lce human islets are Icnown to be less potent than rodent islets, both in vitro (Smith and Wilson, 1991) and in vivo (Jansson et al., 1995). The cell line EP1111220 has an insulin content that appears to be 60-75% of the value prcsented for culturcd hurnan islets, while insulin secretion appcars to surpass that of cultured hurnan islets.

The hnm~ni7ç~ ,B-cell lines g._n~,~ted in these studies exhibit a ll~llber of unique cha~ t ;!~I;cs First they express only one of the two rodcnt insulin genes (Ficdorek et al., 1990 and the inventors' data), which will be adv- ~ tq~ollc in l...o -ko~l d~ loy.l,~,n~ of co~..pl~te insulin~ -cell lines. Se~n~, the pre~sent ~n~2;..P~,,~d lines have the capability to i~c.~ase insulin Se~.;leL~OI~ 10- to 20-fold in l~,SpQ~I~e to 5timll1i This chal~t~-.;clir is similar to 15 that of ~-cell lines derivcd from SV40-T antigen t~ncg~nic mice such as MlN6 (Miyazaki et al., 1990), and ,BHC (Radvanyi et al., 1993) cell lines as wdl as normal ~-cells (Curry, 1986). Third, these cells ~ A;~ cs~ntiqlly nolmal ylocec~ g of hurnan p~ in, even though the exogenous protein is in excess of çndog~ ouc rat protein. Normal plocesC;~g is not present in INS-I (N'~ermsrl-Arbez et al., 1993) and ~TC cells (Nagarnatsu and Steiner, 1992) two ~-cell 20 lines that have been eY~ninerl for this ~lo~,ly. Pinally, the present cell lines ~ te that itera~ive in~v~hl 1ior~ of thc insulin gene provides an ~ploach ~.h~,.el)y hurnan insulin output caD
be stably achie~_d which (at ,,~h~ -,m) rnatchcs that of cultured hurnan islets.

Analvsis of other ~ l~Jmote,/4,~kA~ , eh ~t~ for drivin~ ins~lin e~l~sion. Several 25 other ~hs ~c~ lllo~ers were co~d to the CMV enhar.c~./~..,~t~r for their ability to dircct Ir~ r of the same bicistronic ~~ a~ (5'-intron/h~lScDNlUIRES/NEO/hGH/3'-polyA~ in stably ~ e>~ RIN38 cells. These promoters include the rat insulin 1 gene ~lUllA~ (RIP), ~ ';Pd RIP (FPE/RIP), R~ lin~ced with the rat insulin 1 gene intron (R~JR~i~ in place of the hybrid ~nu~ ;lobulin 5'-intron, the Rous Sarcoma Virus W O 97/26321 PCTrUS97/00761 Long Tf~rrnin~l Repeat (RSV), the human glyceraldehyde-3-phosFh~te dehydi~genasc promoter (GAPDH), and the mouse m~pllQthion~in ~ u,oter (MT). F~p~ceion pl~cmi~s wcre co~ u.;led by Icluuvillg the CMV promoter found in pCMV8/~S/IRESlNEO and replacing it with the plolllolcr to be tested. In this way, mPs~a~ levels and insulin outputs from the RIN
S clones cousl,u~;led with the vaIious ~luluote~ can be colllyaled directly.

RIP activity is approximately 30- to 50-fold lower than that of the CMV prolllotel in .c~ tly transfected RIN38 cells. However, in stably L~ srecled RIN38 cells, RlP activity is much closer to the activity of the CMV promot¢r. The level of human insulin (h~S) mRNA
10 derived from pR~8/hINS/IRES/NEO is, on average, apylo~ tely only 3- to S-fold lower than levels ol~taJl~d from stable RIN38 lines cc,..~ pCMV81hINS/IRES/NEO. The Northern blot depicted in ~G. 13 d~ r..o~ dtes this result as the level of hlNS mRNA from two p.RlP8~hINS/lRES/NEO RIN lines, 2.18 and 2.38, is only 3-fold lower tllan the level of h~
rnRNA from the pCMV81hINS/IRES/NEO RlN .ine, EP18/3E1. As s~ated earlier, the EP1813EI
1~ line has a very high insulin con~n~ a~ ely equivalent to that of a norrnal hurr.an ,B-cell.
T~ ,iole, in addition to the CMV ,ululllotcr~ RlP offers anothcr choice as a strong ~ lscli~iûnal activator.

RIP a.so was ,.~-l;r,~d in an ;.~ t to make it an even stronger .~ s.;li~t;.ona. activator.
20 The ~ l ...o-l;r.r,~;ol- rnade to RIP was the ~ h.... n~ of Far-FLAT mir.i-enh.~lcers (FF
rnini e -k~ e.). The FF min.-~nhqr~r is located ~e,l~.~" -247 and -198 of RIP and contains several cis-.~cting ~ to.y el~ c cruci.ql for RIP aciivity in b cells (Karlsson et al., 1987;
Karlsson er al., 1989). The FP mir.i~nb.ancer region co~lAil~c both the Far box (-239 ~o -230) and the FLAT el. ~--- -t (-222 to -208) which further c(~n~i~tc of two adjacent re~ tory motifs, 25 l;LAT F and FLAT E. When isolated from the ~t insulin 1 gene l~v~oter and ~ ;...e- ;7~l to yield 5 linked copies, the F~ mini~nhq~rer is alrnost as active a s an intA~t RIP in tr, n.ci~.ntly f~ -c~ ~cells (~-~A n ct al., 1992). Threc base changes in the FL~T ~ motif at posjtin~c 209, -211, and -213 can further ;~ 3~:C the ~~ivity of the PE; mini~nhancer (now called ~
mini~nh_nrçr) ,L~ nt~ly 3-folt in ~ l" transfected ~-cclls (German et al., 1992). A
30 I.,..~nt ~ cr~ on system with RlN38 cells WaS set up for iDithlSC~ of ~--o~l;he~
}59 promoterlenhancers. Results from the transient ~ r~ nc lltili7in~ a human growth h~
(hGH) l~poll~. gene ~ ated that two mo-lifi~d RIP enha~ce./~lo,.lot~ were 5-fold more active than RIP. The two Tnn~ifi~.tl RIP erh~n~c-./~lolllot~.~ concictçcl of an intact RIP ( 115 to +l) to which either threc or six copies of FFE mini~nhancers had ~n attached just U~ Ga11~ of -5 415 of RIP (the FFE six-mer is in thc reverse o ;~t ~ion with respect to RIP). Coe"ylcssion of the RIP Ll;~c~ t;on factor. IPF-l, along with either pF~E31RIP/hGH or p~:FE6/RIP/hGH
y~od~ced an 8-fold inelease in activity over that of RIP alone.

To test whether or not the ~ -mo-lifiçd RIP enh~ lv~ would iu~lcase RIP
10 acdvity in stably tr~ncfçct~(l RIN38 cells to the same extent as was d~nor~ d in l cie~ y f~ A RIN38 cells, FF~3/R~ was placed into the 5'-intron/h~lScDNA/lRESJNEO/hGH/3'-polyA stable-~r~ .~r~cliol~ vector. A large nu.llbc,. of RIN38 cell lines CQ~ .; .g pE5~1'RIP8/lNS/lRES/NEO were analyzed for ~ 3/Rlp activity. A uu~l of clonal lines e"l"~,ssed higher human insulin mRNA than was obs~ ,d for the best pRIP811NSlIRESlNEO
15 lines. phos~h~ g~c~ analysis of the Northern blot shown in FIG. 13 d ~ r-~.Ated t~t PF~3/RIP clones 4.17 and 4.32 ~luced approxirnately 2-fold more hlNS than the highcst-plod~ RIP lines 2.18 and 2.38. T~ ,fol~" these data d~ .~ that RIP activity was ~nhq~ce~l in stable RIN lines by the ~ tion of 3 FFE mini en~ f~ tho~lgh not to the same e%tcnt as was shown in the l~ Si~,ut ~ cr~;OI~ system. p~-~k;6/R~811~S/IRESJNEO is 20 e~~ tl~ being introduced stably into RIN38 cells. ~tt~mrtC to stably coe~-yl~;~s IPF-l . re also . a,~r and are clicc~d below.

A second moAifir~~iorl to RIP o~ul,~d by placing the rat insulin 1 gene intron (RlPi) immediately d~ ,am of the tru-c ;l~l;n~ql start site. lt was previously noted that RIF
2~ activity was ci2~.;r.~ 1y hlcleâsed h ll~!n~ ie mice and, to â les_er e~tent, in cultured ~-cells when c~ h;..~d with RIPi. A large n~r of stable RIN38 lines transfected with pRIP8/R~iJlNS/lRESlNEO were esJqhlich~A and e~qmin~d for hINS mRNA levcls. As was obs_. ~_d for the ~l~k3 " .;~ nk ~ , on a ~ " the ?~Aition of RIPi to RIP yielded a modest but ~ r~ ~ -1 ill,l.,aSC in h~S mRNA levels. Thc RIP/RIPi line, 2.6~, eA~ 5 d a level of 30 h~S mR~A e.luiv~ to the CMV l"c,.l~ot~,. linc, 1813El, and tbIee times more hINS mRNA

W O 97/26321 rCTrUS97/00761 than the 2.18 and 2.38 RIP lines (~:IG. 13). Since the ~d~lition of either RIPi or t_e ~
e~h~ "~ enll~ ces RIP activity, thcn cu~fti;nillg both RIPi and ~ mini-cnhancers with RIP
could result in an additive hlcl~se of overall RIP sll~n~ll. To test this idea, p~FE6/R~81RIPi/lNS/IRES/NEO has been co..~ ed and stably ~ncfected into RIN38 cells.
S p(RE)3/-85RIPlR~i/lNSlIRESlNEO, a ~l~cmi~l which con~ls both RIPi and three full-length rat insulin 1 gene enhancers instead of mini-enh~ , has also been co~ .;ted and transfected into RIN38 cells. FPE6/RIPlRIPi did act as a strong ll~s~;,ip~ional activator but was only slightly stronger than either ~61RIP or RIP/RIP} alone. I,~t~,.~..Lillgly, the t~nAemly linked full-length RIP ~Q~.A.~r~... were very weak IIA ~CÇ~ ional activators when p(RE)31-10 8~RIPlRlPi/lNS/IRES/NEO was stably inte~l~d into RIN38 cells. This had not been the casein ~ ly tr-An~fected RIN38 cells in which the three linked RIP enh~n~ers ~o-luced high-level c,.~l~,..sion of a linlced 1~ ~. gene.

Another RIP d_,ivative, pILPR/RIPB/INSi~IRES/NEO has also been con..l.uc~ed in an 1~ attcmpt to gcnerate a more potcnt insulin ~lllote.. The human Class m insulin-linked pol~ll.o~h,c region (~PR) is cu~q~sed Of 139 t-An~emly-rcpeated 14 bp s~u~nccs and }ies immediatcly u~ of the hurnan insulin gene promot~ h~ r~r (O~ .1,ach and Aagaard, l9B4). lt has ~c,~ly been ~,/.n ~ t~A that the ~l~nce of the Class m ILPR cigrifi~ tly i~,nle~c the h~ sclip~,onal activity of the human insulin ~ lllO~./~ ~h~Qre~ nnPdy et aL, 20 1995). I,iL,~. lse, fwing the Class m lLPR to RIP may also ~ ce RlP activity.p~PRlRIP8JINSlNEO has been co..~u,_t~,d and stably inll~du~d into RIN38 cells. Analysis of poly~lo.~al and ...~ n~l lines c~ plLWR~81~S/lRES/NEO ~ ate that the human Class m ILPR bad no ci~,r; fi ~ ~t effect on RIP activity.

pRIP8(02)7 is a mntlifi~d RIP that has been altered by inserting seven copies of tne operator site [(O2)7] from the E. coli t.,t~~~lh~c (tet)-~iS;C~ re opron between the RIP
enhancer and ~lOlll~tC. at ~i~Oll -85. The t~hd~ lle-resistance operon regulatory system (Gossen and Bujard, 1992) is a binary system in which a transactivator protein is a}so ~
T~e tr~sac~ivator is a col~ulation of the tet l_~.,s;-or (tetR), wnich binds very tightly to tet operator sites, fuscd to the L~ ;o~ actival~on df~m~in of virion protein 16 (VP16) from W O 97126321 PCTrUSg7/00761 herpes ci . leY virus. Both pRlP8(02)7JRIPi/lNS/lRESJNE0 and an e~pl~,ssioll plasmid co.~t~ g the tetR-VP16 transactivator will be stably Ll~ r~te~l into RIN38 cells. Pl~,ccde ~ce for this type of scheme was l~c~nll~ (lem..l.~ t.,d when the activity of the already potent CMV
loter was inc,~ased nother l~fold by inser~ing seven tet o~.~lor sites bet~.,n tne 5 ~nhqn~er and yç~o~r followed by cotr-qn~r~l;on with the tetR-VP16 Llal~sa~Liv~or (Liang e~a~., 1995).

The tr~qn~ o~ activity of pr~ r~ other th n CMV, RIP, and RIP de,i~ es also has bcen analyzed. Shble RlN38 lines were cs~P~lichp~d which cont~in~f1 the ~lc~lllOte~ from the 10 Rous Sarcoma Virus Long Terminal Repeat (RSV) driving the standard hlNS/IRESlNE0 stable f~l;on vector. In ~en~l, the RSV ~r~lllot~ o.luced h~S mRNA levcls roughly equi~ralent to those produced by RIP. TL~,.ef,l~" the RSV l~lol-l~h, ~, }ilce the CMV promoter, RIP. and RIP del;v~es, acts as a strong ~ sclitJ~ional activator in RIN cells in culture. The human glyccraldehyde-3-FhosFh~te J,_h~Lop.r-~eG ~luillot~l (GAPDH) was also tested in stably 15 tr~-~f~ ,d R~138 cell lines and foulld to be a weak tr~scli~tional activator. In most GAPDH
~lC~l~lVI~,l }ines, hINS mRNA was either barely or not ~et~ct~~lo by Northern b}ot analysis.

~ Ol~l~t~l stabilitv in vivo. As ~P~r ibe~l carlier for the CMV and RIP promoters, the activity for some of the RlP d~,.ivu~i~es, RSV, and GAPDH ~.olllote.~ was analyzed in vivo by 20 subcutaneous ;~.je~l;o~ of e ~~;t~eelcd RIN 1046-38 lines into athymic Fisher nude rats. ~n vivo activity of RIP was aLso reanalyzed, but this time without the ple~ce of a CMV driven transgene -s-s was the case for RlN }~ne EP1 113E9. Time points were again ~ ;e~-~ to one to two months as most of the ~ sls devdoped h~ ia by two weeks after ;,~ n The ~a from these e~ .-r~ is :,.. n~- ;,~1 be}ow.
~ n vivo RIP activity was ~Y~min~ for two in~lepc..~ RIN lines col~ il-g the pRIP8JlNSlIRES/NEO transgene. Each line was injected into two individual nude rats. Animals containing cither line became hypoglycemic bct~.~n one to two wee~s after i..je~ Tumors werc e~ccised at ~1;rf~nt intervals, hG.~n~ l. and analyzed for bINS rn~NA levels by 30 I~olthc,... blotting- The amount of blNS mRNA n l~ c;n~ constant out to the longest e~z-..; ~ul W O 97126321 PCT~US97J00761 time points, 31 days for line 2.18 and 36 days for line 2.38. Therefore, RIP activity rem~in. A
stable LluC~u~Oul the length of the rjJ,~ 1 The same results were o~ for the mo-lified RIP promoter/çnh~ncers, RIP~RIPi and ~;FE3/R~. Both RIP/RIPi and FFE3/RIP l~roduccd concr~nt levels of hINS mRNA out to the longest time point of 49 days.

The activity of the RSV l)roll,oter appears to be attenuated in vivo. Despite the formation of .. c.1;.. to large tumors, neither animal ;~jC~tn~l with the 3.4 line became Ly~ogl~ell~ic even after 36 days. P~ ly~ if analyzed at later time points, these animals would becoll~e .c due to the endogenous e~ ssion of rat insulin from the e ~p;~u. ~d R~l lines.10 Both animals injected with the 3.34 line eventually did bcco...~ hyyo~ ic but it took much longcr (20 to 30 days) than it did forthe RIP and ~-.o~l;fi~l RIP lines (10 to 15 days). These data suggest that ~lthollgh the RSV enh ~ lloler is a strong L,.u~ onal ~tivator in cultured RIN cells, it may be ~ to direct the c~.~6~ion of a linked h~ gt~-n~ in R~ cells in an in vivo situation. Purther in vivo testing of RSV p~ulllotC. activity lltili7ir~g a transgcne other than 15 the human insulin cDNA is pl-,S_nll~' unde~

GAPDH l~r~ll.oter activity lc ..~ rd stable in vivo out to the longest time point of 22 days. Both ~nim~l~; .jce~ d with the 4.5 line (the GAPDH line that produced the highest level of insulin mRNA) started to becolne hypoglycemic by 13 to 15 days. This result was s~ .e~ Lal 20 sult,l;~i,.g based on the relatively low ~hlmfl~nre of hINS rn~NA e~ t~d in this line.

A co~re-.. with the use of thc viral ~.~ is their long-te~m stability of eAI~r~ssion in vivo. Thcre are rl~.ous reports co~r~ e loss of ~ c~,v.~e e~pl~,ssion in vivo, either following h~LI~u~i!;on of genes in vivo wi~h r~io~l~ t viTuses or il~ud~ n of genes into 25 cells ex vivo followed by i.~ tation of the cells in vivo (Palmer et al., 1989 and 1991, Sch~nn et al., 1991, Chqllit~ and Kohn, 1994). This second scenano is analogous to the p..r c se ~ use of the cell lines bcing developed here for Ih. . ~ use.

I.~t~G;.Lil.gly, the RSV ~ driving transgenes in RIN clones appcar to bc a~enuated 30 in vivo. The mcchanism for this attenua~ion is not clcar. Evidence ~ g~ that some of the W O 97/26321 PCT~US97/00761 pro~lems with long term stability of c~ es~.ion of tr~n~geneS driven by viral prom~ t-Pr.C is due tO
recognition and ~lltim~tely ~ ion of the en~ )t d cells (Dai ct al., 1995, Y~ng et al., 1994). ~ c l~cGEIliLion could 'oe directed against tne tr~ncg~nP product itself or against other ~nti~en~ e~ sscd following hltlvl~ tion of t'ne tl~s~les (i.e., low }evel viral protein 5 expression from ~,co...'~ t viral tranC~ tinns). However, in these studies using nude rats, there is no illl.~..l.-f. l~je~1;0.~ of the in~l~nte~l cells.

Cell ~erowth. insulin content. and ~locc.c~iq~ in bior~a~,lvr~.. The oxygen gas controller output is ~..nr.;iol~d throu~hollt the run. It is an indilc~l inflir~tion of the cells' oxygen co~ ;oll rate. It riscs steadily from around 40 at 0 hours to around 60 at appro~im~~Ply 5"0 hours whcre it st7hili7Pc for the rest of the run. The rate of inc,~,ase of the controller output cvlld~t~,s with an eYre~tPfl growth rate of the culture, and m~ .... Ievel of 60 is consistent with Jchieving a cell density of 1.1-2.3 x 108 cells per ml of bed volume. The cell !lencities are co r. . ~ i at the end of the culnlre. With a s~rfa-e~ to-volume ratio of 120 cm21cm3, the polyester 15 disc bed yields a surface cell density cu,~ ble to that ~bla; ~le in n~o ~ -.f!ncion~l T flask culture. It is hlJlJvllal t to note that the growth and the :~..C~ ~rd ~Pn.Cj~iP~C in the reactor are d using a serum free media. High density ~ es have been l~ problem free for up to 2000 hours in serum free ...~ This obscrvation is novel and vcry useful in the design of abuLlc process for p~ liQI~ of biological ph ....~n."~ lc Cells harvested from the reactor at the end of culture by trypsini7~tion, plated onto 1"75 culture flas~, and assa~d for insulin S~,l;O.~ .r~ e after 24 hours of culture, show no ~iu.;rlc~ dirr~ ,nce relative to sister cells maintuned in 175 flask cul~re, ~ that the bioreactor rnilieu is not changing the cells' ~h~not~ in any ~l~tc~hlc fashion and that the cells 25 quic~ly teadapt to cultute in tissue culture flasks.

HPLC separation of samples cc-llPxt~-~ mid-run at around 550 hours of culture showed err~li~,., insulin ~l~oce~;..e. Thc ratio of rnature human insulin to human proin~lin was 92:8.
~his e~r.~;~..t pl~CC~ iS O~ r~ from a ~ture that has .~ a steady state of oxygen 30 uptake, ;~ .g no o~erall growth. and that is su~tained in a serum free ...~ li~....

W O 97126321 PCTrUS97100761 The bio~a.;lor data indicates that the steady state enYi.un~ in the reactor allows for growth of up to approximately 2 x 108 cells per ml bed, while m~int~ining tlalh~s clucial for complete ~luce.s~ g and storage of insulin.
s Human insulin disulfide mutant production Methods:
Human insulin ~ nJfide mutant eA~ sio,l pl~c~i~l The hurnaII insulin open reading 10 frame was ~ d with the pol~ Osc chain reaction f~vm a human insulin cDNA using oligos 1 and 2 (CCGGGGAT~ }CCATGGCCC, SEQ ID NO:38 and GGGCTAGATCTAGTTGCTGTA~ CAGCTGGTAGAGGGAGCAGATGCTAGTACT
GCAl~ CAC, SEQ ID NO:39) generating a 359 base ~l~lucl (SEQ ID NO:3). Oligo 1 o~ces a BamHI site 7 bases u~ &~l of the initi-tor m-~thio~in~. of insulin. Oligo 2 15 inhudU~,eS a Bgm site just do. l-s~&lll of the insulin stop codon and illLluduces two point mutations into the insulin coding region. These mutations change c~jt~,lllc at position 96 and .;~it~;llC at position 109 to serines (SEQ ID NO:4~. Both of these amino acid ~b~ ul;onc are in the insulin A chain and disrupt the two ~liC llfi~lç bonds nonnqlly formed between the A and B
chains. The mutated insulin protein should be e~ 5se(l~ targeted to the reg~ t~ se~,~,to 20 ~dlll~.ay and proteolytically pl~Jcec~cl to human insulin A, B and C chain. Upon stimulated s~l~,on, the three peptide chains would be 1~ e~ by the cell will~ul the normal ~liCnlfid~
bonds bct~ tne A and B chain. As a control, the wild-type human insulin open reading frame was a~ .l;fie~ with thc pol~ "~c chain ~ from a human insulin cDNA using oligos 1 and 3 (CCGGGGA~ l ~CCATC;GCCC, SEQ ID NO:38 and 25 GGGCTAGATCTAGTTGCAGTAC~ , SEQ II) NO:40). Again, Oligo 1 inhu~luces a BamHI site 7 bases upst~eam of the initi~tor ~ Ih;Qr~ F of insulin. Oligo 2 inhod~c~s a Bgm site just do~ of the insulin stop codon without intro~1vcing any changes into the insulin coding se~ e The res~llting 358 base pair PCR~ pl~l~l.;ts were cloned directly into pNoTArI7 (Prime PCRTM Cloner ~loning System, 5 P~ime to 3 Prime, ~C.) generating30 pNoTArI7~mutlNS and pNoTArI7/wtINS. These pl~mi-i~ were s~ r4~ 1y ~ with CA 0224643l l998-08-l2 W O 97126321 PCTnUS97/00761 BamHI and Bgm e~nr~ le~ces and ligated into BamHI tligeste~ pCMV8/IRES/NEO/hGH
PolyA, ~ 'E pCMV8/mutINS/IRES/NEO and pC~IV8/wtINS/IRES~NEO, l~s~i~,ly.

A variation of pCMV81mut~SlIRES/NEO was crea~ed by restoring the normaI 3'-5 untranslated region of the insulin cDNA to its correct position following the insulin ~icl11fi~e mutant open reading frame. An HgaI cleavage site is located 9 bases 3' of the insulin stop codon, base 364 of SEQ ID NO: 1. pBS/INS w~ rlig~ste~l with HgaI, trea~ed with Klenow r~ t, and then ~igo~st~ with Hinrlm The resn1tin~ 198 base pair fragment was ligated into pNoTA/1"7/mutlNS that had been ~ ,. slrA with Bgm, treated with Klenow f.~ , and then 10 digested with H~nDm. The res-lltin~ pl~cmiA, pNoTArr7/mut~S+lNS3', cO,IIA;I~c an e~c~n~ y restored human insulin cDNA exccpt for the two point ~ inl~c i~lolluced into the coding region and a S base deletion at the BgmlHgaI cloning j.l~l. !;nll. This 198 base pair fr~
co~ln;ns 64 bases of the insulin 3'-un~anslated region, a 41 base pair poly A tract, a 16 base pair poly C tract and 77 base pairs of polylinker sc~ u~-e from the s~b~loning vectors.
15 pNoTArI~71mut~S+lNS3' was ~ l with R~l~r~, generadng a 512 base fr~m~nt CO~
the mutant insulin and teeo~ ucled insulin 3' se~lu~-~re, which was ligated into the BamHI site of pCMV8~1RESlNEOlhGHPolyA, ~ -~ pCbIV8/mut~S+3'/IRES/NEO.

Cell culture and stable t~sfection of cell lines. As des~bed above for insulin 20 p~ J~ cdls.

w~hicloch~ A~I staini~ for hllm~J insulin C-pcpti~. I~L~,i~al G418-~ ~t RlN clones generated by cl~ll~polation using p(~MV81mutINS+3'1IRESlNEO were sc.e."~cd by ;.. ~ ~1; ;.-;.. ~ for hwIlan C-pepdde. Cells were plated on multiwell slides one or more days 25 before ,~ . Slides with spread cells were rinsed with PBS, then fLlced 15-30 mim1tes in 4%
pal~r~ hJdC. Fi~ation was followed by a PBS rinse and r ~ i7~iQn by p, sssgt ou~h an ethanol series of 50%-70~a-50% (5 ..v-. ~,s each). Permeab~ n was followed by a PBS rinse and a 30 minute incubation in 50 mM Tris, pH 7.4, with 1% goat semm, 0.05%
Triton and 0.1% azide. Slides we~c incubated with l:lO,000 ~iltltion of rabbit anti-human C-W O 97126321 PCT~US97/00761 peptide (Linco Inc.) for 24 hours. Excess pliUI~ antibody was removed wit_ seq!lPnti~l washes (3 ~ s each) with PBS-Triton (0.05%), PBS alone, and 50 mM Tris, pH 8Ø The slides were then in~ AtPcl withl an ~ line ~ h lAce-labeled second antibody (goat anti-rabbit IgG, Sigrna Ch~mic~lc) in 50 mM Tris with 1% BSA and 1 mM mAyl~rsiu~ chlori-le (Tris-BSA-Mg) for 30 minlltes. Excess second antibody was ~ .ed with 3 washes of Tris-BSA-Mg. Alkaline phos~-k.1Acc activity was then vicu~li7~d by in~vb~in~ 5 minu~es in an ~ inP pho~h~ace ~ub~ t~, sollltion (BCIP/NBT).

Northern analvsis. Northern analysis of mutant insulin Lli-nc~ in cell lines was10 ~.Ço~ c,d as ~escril~ed above for human insulin m~Ss~e dcte~tion using a full-length ~ig~ nin-labeled ~ ~l;c~nce probe cGll~spQn(1hlg to the lle~ cin resistance gene (control ~emrl~'e :,U~liui in Gcnius 4 Kit).

Results:
Recent reports suggest that ;~ omt)d~ tory tre~ ti with insulin can dclay or prevent the onset of hJ~I~51ycemia in NOD mic¢ (Sh~h~ 1eh et al., l9g4; .S~ in et al., 1990;
Muir cr al., 1993). Clinical trials eV~ ting the prophylactic nature of insulin in l.~l..q~-s at high risk for dle de~ ~t of qpe I diabetes are u~ ay (Keller et al., 1993). Recently, 7AtiQI~ with metabolica}ly inactive insulin B-chain also pru~,nt d the onset of 20 h~yo~lyLci~a in NOD mice, sl~,ee6';1~g an active induction of; ...~----o~ Jl~ti~n by insulin.
lo~ t of an in vivo cdl-based dcli~re,~ system of insulin or metabolically inz~ forms of insulin could be uscd prophyl~-tic~lly in humans at high risk of de~,_lG~ing type I ~ hct~s Cell lines y~O~ g and sc~ ng high levels of mature human insulin have already been ~I~,s- ~ ikd here. This would be done in the context of the ex~ sion of ~duced endGg, l,ous rat 25 insulin. Neuro~ndoc~n~ cells y~ g an inactive, mutant human insulin, in the context of ~l~ed e,lldog~",ous rat insulin ~ ;on, would offer a safer, and possibly more efr~rious app~oach. The use of metabolica}ly inactive insulin would negate the pocci~ility of insulin J~d h~G~ a Higher amounts of a ~t~ y inactive insulin could ~ ,fole be safely a~h~ d in v~vo, pcis~ asil~g the efficacy of the 1~ t W 0 97/26321 PCT~US97/00761 To this end, RlN cells have been en~ c~ to produce a mutant form of human insulin.
Insulin is initially ~ludu~d in the cell as proinclllin a larger pcptide l)lwu.~or co~ of the linear ~~nge~n~n~ of insulin B-chain C-chain A-chain. The maturation of proin~11lin to mature insulin is well understood (Halban, l99l) with three major steps in the process. The first is 5 fold~n~ of the proinClllin into a native collfo~ ioil in the irnmature sccl~,tul~ granules. The second step involves the formation of three di~ fi~le bonds, one i~ in the A-chain and two intr~rnolec~ r ~eh._~n the A-chain and the B-chain. The final step is the endol,,ùt~,olytic ~,oces~ g by PC2 and PC3 followed by carboxypeptidase y~e~ g in the mature SeC~,[~ granule. The mature granules contain an c~l..;...ol~ mix of C-chain (C-peptide) 10 and mature insulin co~C;~1.ng of a A-chain/B-chain heterûdimer covalcntly linlced hy the two ~ .A...r,k~ r t~is~11fi~e bonds. A mutant form of insulin was cullsl~u~,t~,d from the human insulin cDNA in which the two codons el~co ~ ,5 cysteins in the insulin A-chain have been mutated to codons c ~;ng serines (SEQ 10 NO:3). E~ ion of this mutant opcn rcading frame should ~ùduce a mutant insulin peptide (SEQ ID NO:4) that still folds nnnn~lly, the 15 intrachain ~lict~1fide bond in the A~hain can still form, and endoproteolytic l"~Jces~;~-g and ca~buA~ e cle~v~,~, can still occur. The rna~ure ~ran~1f s should now contain an mix of C-chain (C-pcptide) and free B-cbsin and A-chain. The B~hain is irl~ntir~l in seq~ e to the wild-type human insulin B-chain used in studics showing the p,~ ~o~. of the onset of hypogl~ ia in NOD mice (Muir et al., 1995). Stin~ d release of the cc.~n~.l . of 20 the SC~ ,tol~r ~nllles would release all three pepddes. Engineering of these R~ cdls in the conte~t of .~uc~d rat insulin ~ o~- would ensure no insulin biologic activity.

FX~MPLE S
25Hum~n Growth Hormon~ Production Methods:
H~man ~owth l~ n~ n pl?~ The gene cn~ human growth hc,~ m~
was isolated on a 2086 base Bam~/AgeI 1"~l, ;el; ~ o~ cl~se ~ from pOGH (Nichols T~ ul ~ Di~nos~if s~ Inc., San Capistrano. CA). This fi~Jnf'nt co~esponds to bases 498 to 2579 30 of the p~ h~ gene sequence (SEQ ID NO:9, Sccbu,~" 1982). The BarnHI site is located at the W O 97f26321 PCTfUS97/00761 normal site of Ll~s.;liplion of the m~ecqge, 61 bases 5' of the initi~or m~thinnin~ The AgeI site is located 3' of the transcribed se4.,~ .~es of the growth h~ .ol~e gene. This r.~C~ t was ligated into pCB6 (Brewer, 1994) that had been ~ ted with Bgll~ and AgeI, gtnerating pCB6/hGH.
The Bgm site places the hGH gene just downstream of the CMV plv~ tc.. The AgeI site in 5 pCB6 is located in the human growth ho--~ n~ polyadenylation r~e~ t co~ ed in that plq~ The poly~dcn~/ldlion ek- ,.r~t is .~slol~d by clo~ing the entire human growth hol...ont gene into pCB6. Stable h~sçu..llants of pCB6/hGH are selectç~l in G418.

Cell culture and stable ~ .r~ulion of cell lines. These studies were ~.rv....ed as 10 desc ~ il.CA above for insulin prod~cin~ cells.

Sc~cnin~ and characte~ization of human e~owth ~ lol~ prod~ clones. Individual G418 resistant clones g~ F~ by cl~ Dporation using pCB61hG~ were s~ rd for hGH in the conditioned media using an hGH r~ioi~otQpic assay kit (Nichols I~.c~ .le Di-aE.~-Js~ies).
Stim~ q-ted ~rowth h-.,...Ol~C secretion assav and d~ t~ ation of DNA contcnt and ccll llUIll~. Done as ~les. . ;hed for insu}in secretion assay and cell ~ er ~h ~ t;nn ResulLc:
20 Mqn~mq~iqn cell ~ lu~ion of human ~rowth ho.. ~vr Growth h~-.. o~-e has been shown to be thc major regulator of growth in chi~dren as well as ~ A;~ g or .~ " ;..p various metabo}ic r,....,!;o.~c which can decrease with age (~al~cson et aL, 1985 and ~nml-r~ 1994).
icd rcc( ~ human growth l~....-...~ e is DOW being ~ luccd from ~ n cells in bioreactors for clinical use (F.~ch~nl, 1992). C'QI~ ., cell-ba ed delivery of growth h~ nP
from ex v~vo ~n~ e~,d ~ r fibroblasts (Selden et aL, 1987 and IIc lleh~ et aL~ 1994) or obla~ls (Dhawan et al., 1991 and Barr and I ~id,Pn, 1991) is also being ~ t~
Fully ~.. )cesse~ bioactive growth hn.. ~l~r, is ~ ced in all of these s~ "~ to ç~ r r.~,.u~-ndoc~ ne cells to ~u~hlce ,~co ~hil~An~ human growth hnrrnonP offers two advantages. The first is the ability to ~ high levcls of growth k~ - e into a stable cell 30 line with the various m~,tho~ llCd here to ~ z-i-..;,.-, ~C~ ;on lewls. This en~..~ 8 is W O 97/26321 PCTrUS97/0076 being done in a cell line in which production of an endogenous secreted protein has been bl~ P~A The second advantage is th~t the growth hrlnnnne ~loduced in these cells is ~r~-~d into SCCI~tul~ ~r~ eS where regulated release of growth hr ~ f is posci~lc Normally, growth h~ nnnn~ is not se,~ ,ted co..c~ ely, but is secreted in a p llc~tile ll~er as reg~ tPd by S Growth IT~ e R~le?cing Factor and So.~ i,slatin (Arimura, 1994). Growth h.~
}nuduced ~cu~ z~ y in n~uc-nd~ ;.-r cells is known to be secreted through the regulated S~l~,tul,y ~ath~y where its release from the cells can be regulated (Moûre and Kelly, 1985). In ~-cells, growth hor~nrn~ plu~luced from a ll ~.~sg~, is also secreted via the re~ t~cl se.,l~tol~t .~ and secretion can be cocl;....~l~t~d along with the endogcnous insulin (Welsh e~ al., 1986).

RIN 1046-38 clones p~ hi~eh levels of lecn...hinal~t human ~rowth ho..~ f, S~ t~,cn il~ t clones derived from cl~l,o~u,~tion of RIN 1046-38 cells withpCB61hGH were SClC~ for scu~tion of human growth h~ r (hGH). No ~t~ r~ 'e hGHwas detectable from con~liti~n~cl metia from parental RIN 1046-38. ro~t~.~ of the 17 clones t;Ayl~ d si~ific~t levels of hGH. Six clones wcre ~ P~ and characterized further.

hGH is ç~ eclc~l to be se.,l~t~ via the regulated se~;retol~ p~ y in these clones. Cells were cultured for 24 hours in fresh tissue culture media cont~ining 11 mM glncose and 5% fetal 20 calf serum. This con~;r;ol~r~d media was coll~t~A and ;-....- ~..~".,active hGH was ~t~ d (6 samplestclonc were analyzed, 24 hour co~ ction). Cells were washed and either incubated for one hour in media lacl~ing glucose and cc...~ ;9~r~Yi~ (basal, 2 C~ eS per clone) or incubated for one hour in media CO,~I.i,,;,,~ S mM ~h~COSG, 100 ~M
ca~b~hol, 100 ~lM IBMX and aminû acids (stim~ t~-~ 4 sampl~s per clone). Cell .~ for 25 e~h samplc was A~t~ and all hGH values are n-l-mqli?~,A to ~g of se~ ted p.v-luct per rnillion cells. The values are ~ v~d in FIG. 14.

Over a 24 hour collPction~ the six clones secreted b~t~.~n 25 and 229 ~g hGH permillion cells per 24 hours. Clone EP111131 has con~ CI ~ been the highcst hGH

W O 97/26321 PCTrUS97/00761 clone in both the initial screens and in these studies. 229 ~lg hGH per rnillion cells per 24 hours is higher than any value of hGH production by a ~ n cell. Previous reported values are in the range of 7-20 ~lg/million cells/24 hours (Pavlakis and Hamer, 1983) and the highest value c?oltLd is 40 llglmillion cells/24 hours (lIc~lleill et al., 1994).

hGH secretion by these six clones is also eY~ ely regulated. Basal secretion values were all less than 100 ng/million cellslhour, easlly ~e,t~cted in the assay, but barely visible in FIG. 14. Basal values are in the range of 0.1% to 1.0% of the stimulated values for each clone.
Stimulated secretion ranged from 6 to 40 llg hGHlmillion cells/hour. The one hour output of 10 EPll}/31 of 40 tlglmillion cclls is equivalent to the best 24 hour output lepolt~d to date ~Heartlein et al., 1994).

The '~solute outputs of hGH by RIN clones, as well as the fact that it is se~.let~,d via the regulated sec,~to,~ .ay~ are hll~l~lt for both in vitro p~ ul~t;~ and in vivo cell-based 15 deli~ery. For in vitro p~uJ~ iQn~ these cells are producing more hGH in no~mal tissue culture per 24 hours than previously ~lesc~ cells. Cyclical stimulation of thtse cells in â bio~
setting, as previously dcscribed for insulin p~ u~l ;Qn cab be used for ~ ur prodt~tioIl In vivo cell-based delivery of hGH could use the cells in their present form where sccl~LtOll of hGH
would be fairly coT-~t~nt ~It~m~tively, further en~ .g of the cells could produce a more 20 physiological p~llc~tile release ûf hGH in vivo by co,lfe.lil~g regll~tinn of growth ho.l~lonc ,5~,~,",.hQI~ to growth hnl...-.l e-l~ k~;l'g factor andlor som~tostatln, or other re~ nr~ of somaIotropes (~rimllr~ 1994).

~3X~qPLE 6 A. Rat Insulin Promoter Factor 1 Methods:
Rat IPFl ~,ul~ssion pl~$~ 1s A pl~cm;~l cClt~ n;~g the rat IPFl cDNA was ~ nrd from Chris Wright (XB-pdxl). This p~ co..1~ ;nc the opcn reading frame of rat IPFl (SEQ
ID NO:5, bases 7 to 861) cloned into pXBm (Krieg and Mclton, 1984), placing Xcnopus ,B glob~n CA 0224643l l998-08-l2 W O 97~6321 PCTAUS97/00761 5'- and 3'-tr~ncr ibed but untr~nCl~tcA seSl~Je~es 5' and 3' of the rat ~Fl sc~lu~,uce. This cu~ et was made to help st~t~ili7e the IPPl I~P.C~E~, allowing for higher steady-state ~ s.c~f.
levels and protein producti~ n A HinDm~BamHI r~ CQntPinin~ the IPFl and globin se~ es was ligated into the HinDm and Ban~ll sites of pC~6 (Brewer, 1994), genc.~t~g pCB6/IPF1. P.l~ ,1y, the IPFl and globin se.lv~.-res of pCB6/IPFl was removed byrli~ctic!n with Bgm and BamHI and cloned into the BamHI site of pCMV8/IRES/NEOlhGHPolyA, generating pCMV81IPF1/IRES/NEO. Stable l,~sf~ts of both of these ~ ,ssion pl~cmi~1c are sel~ct~ using G418.

It was not clear that the Xenopus ,3 globin sc.ll.. ,nces would st~bili7~ the IPFl tr~ncgene in RIN cclls. For this reason, the IPF1 open reading *ame was ~Tnrlifiecl with the poly~ se chain reaction from pCB61IPFl using two oligos (GGATCCATGAACAGTGAGGAGCAG, SEQ ID
NO:41 and AGA~ ;ACCGGG~~ ;~l~CGG, SEQ ~ NO:42). The r~slllting 867 base ud~l (SEQ ID NO:5) was cloned into pNoTArI7 (5 Prime to 3 Prime, ~nc., Boulder, CO) 5 ~,~...r.,..~ pNoTA~T7/IPFl. The IPFl open reading frame was rernoved from pNoTA/1"7/IPFl by ~i~s~io~ with BamF~ and was l~gated into BamF~ gested pCB6, generating pCB61IPFl(-Bg). ~ ,ly, the same Il?Pl Bam~ f~ t was ligated into BamHI ~ligest~d pCMV8/IRESJNEO/hGHPolyA, generating pCMV8/IPFl(-Bg)/IRES/NEO. A final e,.~l~ssion pl~ was rnade, ligating the IPF1 BamHI fragment into BamHI ~ .st~d 20 pCMV8/~s3'/IRESlNEO, ~c~ tiug pCMV8/IPFl9-Bg)J~ns3'11RES/NEO. The Ins3' nontranslated region in these plasmids was i:s~rihed earlier for the insulin ~1iclllfitle mutant e~cample and is ~ on a 198 base pair HgaVHinDm r~ This r,~nt was ligatcd into pCMV811RESlNEOlhGHPolyA ~ r~ fi i~ pCMV8/Ins3~/IRES/NEO. Stable ~ r~ctants of all of these eAy~ sio~ c~ are selP-c~ using G418.
Cell culture and stable transfection of cell lines. These studies were pe.r..lllled as hed above for insulin pl~h~ cells.

S.;~ r and ~h~ra~j~,,;~ 'i~ of IPFl prQ~ ~nV clones. Northern analysis of individnal 30 G418 recict~nt clones g~ Jt~'d from the vadous IPFl e~ylcssion Fl~cmide was done as .

W O 97126321 PCTrUS97100761 ~es~ d above for the human insulin northern analysis. Blots were hybn-ii7e~1 with a 32p_ labeled cRNA probe co~ s~,ond ug to the rat IPFl open reading frame ~SEQ ID NO:5).

Results:
S O _.~A~ ssion of IPF-I in RJN 1046-38 adls. IPF-l fiml~tiollc both in the sperifi~ ~tiQn of a region of the prim~tive gut to form pancreas in the maturation of the pancreatic ~ cells.
Because RIN 1046-38 cells retain only some of the Lfrl,.e~iated ftdl~,s of a normal b cell, o~ c~ of IPF-l in these cells could cause them to func~ion more lilce mature b cells.
Thus l~r~ ed RIN cells may serve as a more errec~.e bi~ oï for the pl~~ ;Qn of biologically r~,leval~t se~ ,ted ~lol~lns.

In initial r~ c~ stable tral~cfectinn of RIN 1046-38 cells with either pCMV81IPF-l11RES/NEO or pCB61IPF-l resulted in a low number of NEO .r,s~e~ col~ s. None of these colonies eA~ sed the IPF-l transgene as demonstrated by Northern blot analysis. A second round of stable 1.~ r~ ne were ~.~u~llcd with lPF-l c~ u~;~s in which the Xen~ ~ S' a~d 3t betaglobin nn~nei~ted ~q~h 1~r~s (UTR) were removed ~lPF-l~-Bg)]. Also, in some CO~SIIU~ the pOtf ~ 1y st~hili7in~ II1S3' UTR was fused immPdi~tPIy duwu~ of the ~'F-l cDNA. A ll.od&~te n.~ ber of NEO-l~Q;~ I colonies were obtained from RlN cells ~., .
with either PCMV81IPF-l(-Bg)/lRES/NEO or pCMV81IPF-l(-Bg)/Ins3'/IRESJNEO. 1'IG1~analysis of RNA from a ~ed popula~ion of cQ'onies co~ g dther construct ~ .c1rdted that tbe IPF-l transgene RNA was indeed o~ at,l~~cd related to endogenous IPF-l (FIG. 16, lanes labcled ~I~,clonc #} and #2). The ~ tiot~ of the 3'Ins Ul~ to the IPF-l cDNA did not aprar to have a s~ifirsnt effect on IPF-l transgene e,.~.~ssiou.

Al~so shown in FIG. 16 are several clonal RIN lines .,.~ lcss.ng IPF-l mRNA. As would bc e~E.c~ ~1, some of the clonal lines espress more IPF-l rnRNA than the polyclone and some less since the ~l~_lonc rep~ents an ~e IPF-l e~ ,ssion levcl from many dmg-resistant colc~ni~c. ~hhml~h not shown here, the pol~lonal cells were sna~y~ed for the p.~cc of lPF- 1 protein by Western blotting. A slight O..~ ;,s:~n of ~F- } protein wa~c d~ t~-t~ ~1 o~rer 30 and ~bove cnAo~ s~cly c~ ;ssed IPF-l protein in untr~.~ l RIN 1046 38 cells. Clonal }i21es co.-~;n~ IPF-I tr~n~r-nes are ~ lentl~ bcing analyzed for h~ ,ased levels of IPF-l protein.

The IPF- I col~t~ ng polyclonal lines were also ch~ cd for i~ ea~d levels of S ~ ndog~ v~c insulin, gltlcol-in~ and GLUT-2. L~ ased levels of any one or all threc of these ~lUt~s could pote-nti~l~y be in~ tive of more di~r~e~lliated RIN cells. Northern analysis ,.,.~,a}c~ that neither cn~,gr ~~ insulin nor GLUT-2 mRNA was crr~A by slight o~,~,..,Ap.~ssion of IPF-l protein in tho polyclonal RIN lines. Howcver, gluco~inase mRNA was slightly elevated in the IPF-1 tr~n~gr!ne col.~ .;n~ lines. This rnight be ~Ypert~l since it has 10 been recently de...ol.~l.ated that IPF-1 h~t,l~L~ with the ~-cclls glucokinase plolllote. to play a role in thc glucokinase gcne re~ tton (Watada et al., 1996~. It is also well proven that IPF-l is ~ lr nr in insulin gene re~ tion (Peers et al., 1994), but as statcd above, there was not an elevatcd level of insulin mRNA in the IPF-l polyclones. Whether or not slight elevation in glucokinase has any pl,~iological .ci~llifir~nce is currently under investigation. ,A~ itiQ~slly, 15 some of the clonal lines ~L~ ting a higher level of IPF-l mRNA (FIG. 16) than the pol~_lol~al lines are being analyzed in d~e sarne manner as the po}~,io~al RIN lines.

B. Alternative Drug Sdection Markers Methods:
I;~ io.~ tl i~ls with ~ ive selection 1~12~ . To f~ilit~t~ en~ e~ g Of mu~tiple genes into the same cell lille or to .,~ e cA~l~ss~on of a given gene, altemative C"~ Sioul~35~ s cQ~ 8 othcr drug s~l~c~;~n- marlcers werc desiglled. The drug scl~c1;0~
markers utilized include the h~v~ resistance gene (HYGRO), the puromycin resistance gene (PURO), the dihydrofolate ,~d~ ce gene (DHE~) conf~..;..g resistance to ~ w~ate~
25 the ~lA..ll.;.~F guanine phos~h~nk~ cf~ gene (GPI) CG~ft-,;..g resistance to _v~ph---.oli~ acid, the Zeocin resistance gene (20), and the hi~ti-linol ~l~l;Q~ gene (HISD).
All of the dIug s~lectiQn genes were tested for their ability to confer drug resistance to RIN cells in two CQ~ . The first was by ~ "';"E the new drug se}r~l;o~ gene for the nco~"~resistanoe gene in pCMV8/lRES~NEO. ~n this cnntext the drug resistance genc is tr~-~- . ;h~ off 30 of ~e CMV ~v~ote~ as the downstream open reading frame of a h:ci~vnic message. Tne W O 97126321 PCTrUS97/00761 second is by suL.s~ g the new drug selection gene for the ncomycin resistance gene in pCB6 (Brewer, 1994) such that the new dmg sÇlecti~n gene is driven by the SV40 promoter. pCB7 (Brewer et aL, 1994) was constructed this way with the hyg~ res;clAnce gene repl~ing the ~e~ l r~ -ce gene.

The open reading frame of the hy~ulll~ rÇsist~nre gene was ~mplifiPd using the pol~ se chain lCaCIiOll from pCB7 using oligos (GGGGATCCGATATGAAAAAGCCTG, SEQ ID NO:43 and CGAGATCTACTCTATTC(~ l I~C, SEQ ID NO:44). The res~ ng 1048 base pr~h~ wae ~ligee~pd with BamHI and Bgm and ligated into the BamHI site of pCMV8 10 genera~ing pCMV81HYGRO. In a second step, the IRES cle.ll., t (SEQ ID NO: l l ) co~ ed on a 235 base BamHllBgm r.~..,.,-, was ligatedl into the BamHI site of pCMV81HYGRO
g~ ,g pCMV8/IR~;SlHYGRO. Stable trar~r~ i of pCB7 and pCMV8/~ESlHYGRO
are s~ cte-l using 300 llgJml hygromycin (Bo~hr~l~ger l~nnh~im) for 14 days without media changes.
1~
T_e E col~ open reading frame en~o~lin~ XGPRT was ~mrlifi--d with the polymcrasechain reaction from pSV3/GPr (ATCC#37144. Ml~lligan and Berg, 1980 and 1981) using oligos (CCGGATCCCATGAGCGAAAAAT, SEQ ID NO:45 and GGAGA~ lAGCGACCGGAGAT, SEQ ID NO:46). The res-lltin~ 476 base pair ~ .l;r;cd 20 ylc~ was restnrted with BamHI and Bgm and C~klo~tA into the Bam~ site of pCMV8, gencrating pCMV8/GPI . Next, the ~ES rl~ .. n~ (SEQ ID NO: 1 1 ) was ligated into tne B~
sitc of pCMV8JGPI, gcneratmg pCMV8/lRES/GPI'. The GPT open reading frame was isolated from pCMV8/GPI' by t~ s~icm with BamHJ, and SmaI and the res~lltin~ 482 base pair r.~..~...
was ligated into pCB6rmtron (see above) that had previously been di~ested with NllrI, trcated 25 wi~ Klenow fr~ nt and then c~;g~st~ ~1 with Bcn, g~nc dling, pCB8. Stable Llan~fu~ snte of pCMV81IRES1GPI' and pCB8 sre s~le~t~d using 2.5 to 3.0 ~ u~k~ nlic acid (Sigma ChelI~icsl Co.) in medis wi~ l e~<~g~ . C rs~*t~in~ added for 14 daye. Media was ~h~n~d everS~ 3 to 4 days.

The open reading frame of the mouse dihydrofolate redl~t~ce cDNA was ~mplif ~d with the PO~ 3e chain reaction from pSV3~ (ATCC#37147, Subramani e- al., 1981) using oligos (CCGGATCCATGGITCGACCAlTG, SEQ lD NO:47 and GGAGA~ il lA~l~l~ , SEQ ID NO:48). The reSl~lting 581 base pair ~ ;fird 5 ~o-luet was l~slliel~d with BamHI and Bgm and subcloned into the BamFIl site of pCMV8, gene~ing pCMV81DHFR. Next, the IRES cle~ (SEQ I[) NO:11) was ligated into the BamB sitc of pCMV8/DHPR, g.,~ .g pCMV8~rlRES/DHFR. The D~ open reading frame was isolated from pCMV81D~ by digestion with BamHI and SmaI and the reslllring 582 base pair fi~gJn~ t was ligated into pCB~ tlon (see above) that had previously been ~igestç~ with 10 lVarI, treated with Klenow ~l..6ll~ t and then ~gested with Bc~, &~ pCB9. Stable transformantsi of pCMV81IRES/DH3;R and pCB9 are sele~te~l using 1 to 10 ~g/ml m~thotre~cate t~,.ill, Sigma ~ mi.~l Co.) for 14 days with media ch~ s every 3 to 4 days.

The open reading frame of the HisD gene was ~ .ed with the polyrnerase cham 15 re~ion from pREiP8 (~vitrogen, Inc.~ using oligos (CCGGATCCAT~AGCTTCAATAC,SEQ~ NO:4g and CCAGATCTGCTCATGCTTGCTCC,SEQ ~ NO:50). The resllltin~ 1063 base pair ~ ; r~rd pfod~ t was r~ e(~with BamHI and Bgl~ and s- ~ d into thc BamHI site of pCMV8, B~ 8 pCMV8/HISD. Next, the IRES rle...- ~l (SEQ ID NO:ll) was ligated into the BamHI site of pCMV8/HISD, ge.le,dhng pCMV81IRESlHISD. Stable ~ f~ of 20 pCMV8/IRES/HlSD are sele~ d in ~nedia with 0.8 to 1.0 mg/ml hic-i~lirol for 14 days. Metia was changed every 3 1 days.

The ~ hl r~sic~ e gene was isola~ed from pPUR ~ ~c.) by ~li~st~
widl PstI and XbaI. Thc ~eY~hir~g 792 bw pair f~n- was trea~ed with Klenow r.~E....~ and 25 ligated into the SmaI site of pCMV8, ~ g pCMV8JPUR0. Next, the IRES c~ (SEQ
ID N0:11) was ligated into the BamHI site of pCMV8/PUR0, ~,~,.I.,,~g pCMV811RES/PUR0.
The PURO open reading frame was isolated f~om pCMV8/PUR0 by ~i~stion with NcoL
treated with Klenow r.~;,.. --t, and then ~1i~st~ l with BamHI. The reS~lltin~ 723 base f.,~
was ligated into pCB6fimtron (sce above) that had previously 'oeen ~ çcte~ witn NarI, treatcd 30 with Klenow f.,.g,....l~, and then ~lig~s~efl with BclI, g~ ,.dting pCB10. Stable ~ r~ nlc of W O97/26321 PCT~US97/00761 pCMV8~IRES/PUR0 and pCB10 are select~ using 1.75 to 2.0 ~Lglml puromycin (Sigma C: hPmir~l Co.) for 10 days with media ehqn~es evcry 3 to 4 days.

Tbe zeocin l~,c ~1A-~ee gene was icol ttPd from pZeoSV (Invitloge,~, Inc.) by ~~i~stion with S NcoI and Acd. The res~lltin~ 430 base r ~ was treated with Nenow r. ~, .~r...t and ligated into the SmaI site of pCMV8, generating pC~V8QE0. Ncxt, the IRES çl.-"- "l (SEQ II) N0:11) was ligated into the BamHI site of pCMV81ZE0, ge~ ;ing pCMV8/~ESJZE0. TheZE0 open reading frame was isolated from pCMV8/ZE0 by digestion with Rsrrf, treated with Klenow r~ t, and then Aigrst~3 with ~nrr~ The reSltltin~ 406 b~se fi~lllcnt was ligated 10 into pCB6Antron (see above'~ that had previously been ~ige,~st~d with ~VarI, treated with Klenow Ll..~.o " and then ~i~sted with Bc~, generating pCB11. Stable ll~sfull,lants of pCMV81IRESlZE0 and pCB11 are selc~e~ using 400 ~lg/rnl Zeocin (lllvil,~gen, ~c.) for 14 days with mcdia ~h~.~e~c every 3 to 4 days.

Rat Amylin Production Methods:
A~nylin e~u.~,ssion plPcmi~ic~ A ~nDIIlJ Xbal rl~.,lc..t COll ~Qnf~ing to bases ~6 to 2061 l of the pu~ich~-~l rat amylin cDNA 5'~ e (SEQ ID N0:7, (Leffert, et a~., 1989)) e ~ro~ G
rat ,Ul~)lU..lll~ l (SEQ II) N0:8) was treated with Klenow E.~~ to blunt the ends. This ~lunt ended r. .G...f..-~ was ligated into the Klenow treated Xbal site of pCMV81IRES/NEOlhGH
PolyA generating pCMV8/r.Amylid~ES/NE0. The CMV ylu~ul~r drives ~ ;on of a bil"~s~U~C II-~SS~ ~ r ~NA with rat amylin e-~4dGA in the upstream open reading frame and the 2~ nco,..~_in t~C ~ re gene el co~kcl in the du....st,~" open reading fr~me. Stable transfcc~ts from this p!qcmi~l are c~l~cte~ in G418.

Thc human amylin coding region wac isolated by usc of the ~ol~ .,,~G chain fCs~iOn .lt;li~i~e two o~gos ~TC CTGATATTGCTGAC (SEQ ~ NO:62) ~d~0 TGGGACCTTAGTTAGTAC(SEQ ~ N0:63~3 ~d hun~n p~cn~c cDNA as a tcmplate lM

W O 97126321 PCT~US97/0076 (Human ~ ,ase QUICK-Clone cD~A, 7115-1). The reSulting 491 base pair fragment (SEQ
II) N0:52) encodillg human preproamylin (SEQ Il~ N0:53) was ligated into the PCR cloning vector, pNoTArI'7 (5 Prime to 3 Pr~ne Inc., Ro~ r Co.), genera~ing pNoTAT7/h.Amylin.
pNoTAT71h.Amylin was r~ctrict~d with Xbal and the rçsl~1ting 523 bace r. ,~,,,,~..t C~ .;..g the 5 human amylin open reating frame was ligated into the Xbal site of pCB7/intron generating pCB7/intron/h.Amylin. The CMV p~ oter d~ives tra~ .t;nn of tne human amylin coding se~,w~re while the hy~ m iec;~ e gene is tr~lccnbe~ using the SV40 ~,ol.lot~r from an ~ p..~e1~t~o~s~ io~ unit ~nco~ d on the plasmid (Brewer 1994).

A final e~ ssion plasrnid c~-ab?~ of CO~A~1eSSi~lg human insulin, rat amylin and the neomycin le~ e marker was c~sl~ t~l. l'he human GRP 78 ~ re e ~ro.lil-g the internal ri~cs~me entry site (lRES, SEQ ID N0:11) on a BamHl/ BglII f..~,5,... ~~t was ligated into the BamHl site of pBS/lNS, a pl~crri~l co~ ing the human insulin cDNA (SEQ ID NO:l), generating pBSfINSflRES. pBSflNSflRES was ~igeste~ with Xhol and Xbal, treated with 15 Klenow and the res~llting fi~ co,.~ ~S and the IRES SC~ re was ligated into pCMV8fr.Amylin/IRES/NEO that had been digest~d with Xbal and Klenow treated, g~nelOtillg pCMV8fINSfIRESJr.Amylin/IRES~NE0. The C~V plullloter drives e~l,.ession of a l.icis~ul~ic .co~ g human insulin, rat amylin and the neGlllyehl resistance ma~er. Stable ~f~ t~ from this pl~r~ arc sel~ct~d in G418.
Cell culnlre and stable ~ r~ ion of cell lines. RIN 1046-38 cells were cultured and trar~ c~ ~1 as d~s~- ikd above for insulin ~ 5 cells.

n~t~xl~.n~ Stainins for Rat ~ Individual RlN clones generated by 25 elh;~lu~lLlion using amylin e~ *C:on l-t~ rl$ were s~ using an anti-rat amylin pol~ al antibody (Pcnin~ Labs, IHC 7323). Cells were pla~ed on coverslips or in "Ck~---~
slidcs~' one or more days before i~.. ~G~ s, Mcdiulll is ,~ cd from cells, rinse cells with PBS, then fLlc cells with 4% paraformal~h.~.k, 15-30 minutes. Rinse with PBS, then S~lCCCS';~, washes in 50%, 70% and 50~o cthanol for 5 I-~ s cach. Rinse with PBS and then block free 30 aldehydes and non-s~er~r~l~ sites us~ng 1% scrum (same species as sccond ~tibody), 0,059ro W O g7126321 PCTrUS97/00761 Triton, 0.1% azide in 50 mM Tris, pH 7.4 for 15 to 30 . ~ s. Fixed cells were ;..~ e~ with anti-rat amylin primary antibody at 1:2,000 and 1:600 ~lih~tiQnc for 80 ...;.--~t~s at room temperature. Wash in PBS-Triton (0.05%) wash, 3 min~ltes, PBS alone for 3 ...;. ~l- s, then rinse in 50 mM Tris, pH 8Ø Inellb~t~ cells with Alkaline Phosph~t~ce labeled second antibody for 0.5 hr in 0.5%BSA, 1.0 mM MgCl2, 50 mM Tris, pH 8Ø Cells are washed 2-3 times in 50 mM
Tns, pH 8.0t 1 mM MgCl. Finally, inrllb~t~ cells with fresh BCIPINBT solutiQn (0.3 mg/ml nitroblue terazolium, 0.15 mglml bromo-chloryl~indoyl phosl~hate, dis~;~. salt (both from Sigrna ~h~mi~lc) in 50 mM Tris, pH 8.0, 1 mM MgCl). Reaction is stopped by washing in water and then air drying.
Northem analvsis. Northem analysis of rat amylin Ll~ ~r ~ t~i in cell lines was done as dFr~-i~d above for human insulin message d~t~Pio~ Filters were hyb1i~i7e~3 with a full-length ~lig~ labeled antisense probe cC~ ,C~o~ g to the rat amylin cDNA (SEQ ID NO:7) made using Genius 4 RNA l~~ling Kit (Boe~ ~A~ h~.i...) and T7 pol~ .dse. Northern 1~ analysis of rat peptidylglycine alpha-~mi~l~ting ~ . ooAygenase (PAM) in cell lines was done as descnbed using a ~1i~xig~nin-labeled ~r.t;.~ e probe co~ ollding to the bases 240 to 829 of the rat PAM cDNA (Stoffers et al., 1989) made using Genius 4 RNA Labeling Kit (Boen.~e Mannheim) and T7 pol~-- ~- -- A~e.

lnsulin and Arnvlin Assays. Tmmllno reactive insulin (IRI) species were cletected by "~ .~assay as ~1es~ hed (~ et al., 1986) or using a cc".. r-,~ ;~lly available insulin r~ o-assay (Coat-a-count, D;a~rQSt;C ~ UelS COrP., Los Angelcs). L~ .G,.,active amyJin species were ~ct~t,~d by r~ o-ass~y as describcd (Pieber, et aL, 1994) or using a c~ ,cially available rat arnylin immlmo~cc~y (Pt-~ Laboratories~ EIAH-7323).
Alternatively, ;.u.. ~-~,ea~ e amylin was ~eter~d with the following m~ifi~ n~
The same anti-amylin polyclonal serum (T~86~6, Pieber, et al., 1993) was used at a final dilution of 1:20,000 in RIA buffer (r~ Labs, Cat# RIK-BUF). l25I labelcd rat amylin and rat amylin peptide standard for the RIA were purchased from E~ Labs (Cat#Y-7323 and 30 #7323 ~ ,ly). Human amylin ~pecies were q.~ ~t1 using the rat amylin imm~lnoassays read against a human amylin standard (Rsrh~m, Inc., PCPE60). Standard ~ tion's ranged from 0.19 to 25 ng/rnl with 0.0125 ~lci trace labeled rat amylin per assay tube. Bound and free amylin were separated by shsking at 200 rpm overnight at room ~c~ ~e using a goat anti-rabbit IgGlbiotin conjugate (1:10,000 final dilution, Sigma ~h~ miral CO., St. Louis, Cat.# B-8895) and an avidin coated bead (Nichols ~n.ctitnt~ Diaei.oslir,s, San Juan Capistrano, Ca., Cat#30-0591).
r were count~d on a (~!s-mms C-12 gamma counter (Di~nostir ~ud~ucls Corp., Los, Angeles) and values ~e~ d using the enrlose~l Log-Logit sofl.. ~c.

Results:
PePtid~ vcine alpha-smi~s,tin~ monoo~r~enase e~ ssion in cell lines. Alpha-amidation is DOW appreciated as a critical d~t~ ;uant for biological activity of a largc "u~e~ of peptide k.,....n..~s Table 4 rcylc~l~ts a sample of hurnan peptide hls....n~s that are ~nown to be amidated in vivo. The e~ e involved in alpha-ls-mi~stion, peptidy}gly,~e alpha-amidating o~ e (PAM), has been well cha~ ~ ., d at the m~ lar levcl (~ ,d in Eiper et al., 1992a). ~l~ho~lgh there is only ane genc in 1~9-.. ~ c cnrot~ g PAM (Ouafik et al., 1992), thcre are scYeral forms of PAM due to alterna~ive splicin~ and end~plotcolytic yloccsc ~F
(Stoff~r~c et al., 1989 and 1991, Eipcr et al., 1992b) leading to both mcmbranc-bound and s~.~t~ forms of PAM. PAM is also known to bc d~,~. lo~ nt~lly regulated and dirr~ tially eAy~,SSed in vivo (Ouafik e~ al., 1989). The iln?o~cc of alpha-~mi~tio~ of peptide hf.. es 20 is such that the p~,sencc of the cor~ensus glycine followed by two basic amino acids (lysine andfor &~ .C) in a novel amino acid sc~ n-,e can be predictive of its bcing a yl~ul~or to a bioactiYe y~l)ry~,~Lide (Cuttita, 1993).

Amylin and GLP- 1 are ~wo peptide hormones that are amidated in vivo. A more complete list of amidated human polypeptide hntmnn-oc is found in Tablc 4. ~ c at ,.. , .. ~ n cell inn of any of these h.,, ~n.~f s l~Uil 5 ¢.,do~r~tcolytic clcavage of larger yl~ul ,ul~, e~utJvA~pl;A~C~ g and ~~ amidation. For ;,,c~ e f;ltt~g~n-Iike Peptide 1 (GLP-1) is a pcptide hn- ~-~ with y~ nc~ J~ ,.c effects secrctcd from the ;~hrS~ 51 L cells in ~ c to mcals (Kreymann and Williams, lg87). It plùcGscr~d from a largcr pol~ ide 30 ~ or ~hlvu~l~ steps ~at are vcry similar to the pl~ce~ g of amylin. P~v~e~ g of GLP-1 WO 97t26321 PCTrUS97~761 involves the action of the endoprot~ases PC2 and PC3 and carbox~ ce on the same ;,~ul~or that ~lu~agon (Mojsov et al., 1986 and ~ouille et al., 1995). The final biologically active peptide is a ,l~iAIwe of GLP-1 7-37 and GLP-1 7-36 amide, a difr~cnce res-llting from the vccs~ g of the glycine at position 37 to an alpha-amidated form by peptidylglycine a~pha-~mid~ting monoo~yE,_nace (PAM) (Orskov et al., 1989 and Mojsov et al., 1990). Both GLP-1 7-37 and GLP-l 7-36 amide are both biologically active in h...~ c (Orskov et aL, 1993).
The rat inclllinom~ cell line used hcre, RIN 1046-38 has already bcen shown to exprcss ~..rfi~-f ,.t levels of PC2, PC3 and c~uboAy~p~ ~ for complcte ~JloceC~ g of nighly eAl,.~d human insulin.
Amylin is a 37 amino acid polypeptide h~l ,..nne a8ain procçsse~l firom a larger pn,c~or polypeptide by the p.o~ol~tic ~r~ccs~ (Sanke et al., 1988). Amylin is nnnn~lly Co-~ uced and co-sc~ ted with insulin by b~ells, acting as a h.. ~l-.' to regulate c,uboh~h~e m~t~bolicm (~lo~pe.... etal., 1994). However, unlike insulin, amylin is alpha-ami~ted by PAM in thc b-15 cells (Sa~ke et al., 1988). O~ ,ssion of amylin in RIN 104~38 cells will serve as a L~n of the ability of these cells to produ¢e amidated peptide hormones.

NGl~ analysis was used to address the ~ ~Ac,g~ o!~ levels of PAM in various celllines. E~cpression of PAM in RIN 104~38 is co,~ d to AtT-20 and two related RIN lines, RIN 1027-B2 and RlN 1~44 (Philippe et al., 1487). Fn~logerlol~C eA~ssion of a single PAM
message of app~ t~ ly 3.5 kB is casily ~c~t~3 in aU three RIN lines (i;IG. 15A, Lanes 1, 3 and 4). Lower e~ ion of two PAM messages of at"~ ately 4.0 and 3.5 kB is found in AtT-20 cells (E;IG. 15A, ~ane 2). PAM ,~r.C~e sizes of 3.5 to 4.0 kB is conc;Lt~ with the larger spliced variants of PAM message known to encode active PAM protein (Stoffers et al., 1989). ~ ssiol- of eudog_llo~s PAM was compared with eA~l~ssion of e~lo6.-no~s amylin in these same cell lines. The three RIN lines with high levds of PAM also showed high levels of t .~Ane,e-~ s amylin ~ tl.~ssion (PIG. 15A, I ~l-es 1, 3 and 4). AtT20 cclls, a pituitary cell line does not have any ~ lo~ amylin CA}JI~SSiO~ t~ ,slhlgly, two RlN 1046-38 derived cloncs (EPI8~3G8 .,A~.,iSlng large amounts of human insulin (l:lG. 11) and EP~3/114 o~ A~ s,lng 30 rat glucokinase) that no longer e~cpress e~3Og~ amylin show lower levels of c~,~s~ion of W O 97/26321 PCTrUS97/0~761 çn~og~o.~c PAM ~;IG. l5A, Lanes S and 6). The majority of RIN 104~38 derived clones collti.~l~c to express both endogenous amylin and PAM, sllg~ting that RIN 1046-38 derived clones will m~int~in the ability to çrr;~if .~tly amidate peptide hn.~ f s The high level of PAM e~ s~ion in RIN 1046-38 collly&~d to AtT-20 is very ~ 'O...A~;t~g. Co.l,l.~ ;cn~- of PAM ca~l~;ssion in other cell types has shown that AtT-20 cells eApress very high enzylllc levcls (Takcuchi e~ al., 1990). This inr~lldes higher levels than PC12 cells and RIN5-f cells, a rat mc~llinomo line that is fairly de, I;rr~".,ntiated when collly&~,d to RlN
1046-38. M~int~ining high PAM activity in RlN 1046-38, sim,ilar to m~int~ining high levels of 10 PC2 and PC3 activity, ~U~ S ove.~,Ayl~,ssion of tr~n~genPs for ~mitl~Pd peptide hv".,~ es such as a nylin will result in their err~ y~Jv !.nn The rat amylin ~,Ay,~,s~ion p!~c~rti~, pCMV81r.AMY~lI~ES/NEO was cleeLIoyGrdted into RIN 104~38 cells and stable cloncs se)-rb d in G418. Analysis of poly.,lones by Nor~hern 15 aual~..is (1~ .cl.dtes ~rl;~ e~ ssion of the AMYI~/IRESINEO bie~ omc message (FIG. 15B~. Individual stable clones wcre scl~cned for arnylin c~)lc,ssion using an in situ ;... ~o~ protocol using two dilutions of the primary amylin antibody. At the lower rlih~tiQn (1:200) all the cells are ~osili~, due to the levels of endGgel~ous arnylin. At the higher ti~n (1:1000), only a subset of clones co..~ ed to show st~inin~. presurnably due to 20 ~ sion of the arnylin l~n~ rr. Five such clones from this series (BG97 clones) were picked and s~l~d amylin i~ reactive species (ktr ...;~d, using cclls in the basal state or s~imulated for sc~l~,Lioll. Regulated secretion of amylin is ~1~ "~ dted in FIG. 17.

The human amylin eAI,Iession plasrnid, pCB7fintronlh.AMYL~ was el~:lru~olated into RIN 104~38 cells and stable clones s~ A in ~uIn~ . Clones f~om this series (BG182 clones) wcre s-,l~.lcd using the amylin im~~ o assay. Once individu~ clones had been establisbed in 24 or 6 well tissue cultnre plates, a 24 hour media sample was co~ ed and assayed. Cell nulllbc- was c.l;---~t~d from visual irC~ectin~ of the culnlre plates. BG182 clones with ~,lous illcI~scd amylin im.,luno.~Lli~ y per million cells were ~ ;r~d and chara~t~,.~1 further.

CA 0224643l l998-08-l2 W O 97~6321 PCT~us97/00761 Several intinpcf~ pl'o~l~s for coçngjn~er amylin and insulin overproduction in R~J
104~38 cells were ~ ted. Thcse inollldPd P~-~;nt~ .g human insulin e~pl~ssion into a rat amylin yrod~lcing RIN clone (pCMV8/INS/IRESlPURO into BG97/134, g_nc.~ g the EP183 5 and EP196 series), c~ g human amylin into a human insulin producing RIN clone (pCB7/intronlh.AMY~N into BG1813E1, genera~ing the EP184 and EP197 series), and finally engj~ g human insulin and rat amylin co P~ s~iol~ into RIN 1046-38 using a single pl~cmi~
cO~ (pCMV8/INS/IRESlr.Amylin/~RES/N~O into RIN 1046-38, ~~ d~g the EP181 senes). ~ all cases, individual drug recict~ -t clones were screened by amylin and/ or insulin 10 imm.-no~cs~y of a 24 hour sample coll~cteA from cells cultured in fresh tissue culture media with an e,~ AIe of cell n~ ,be,. Individual clones from each series were id~ntifie~l and ch~,.et~.., furtber.

Table 9 gives er~-nrles of several RIN cell lines ov~ .,s~ing insulin and /or amylin at 15 ratios varying over three orders of n~a~inlde In addition to the ratios, ~hsohlte ~ ..o - .1~ of insulin and amylin are given in ng per million cells per hour during a stim..l~tt~cl secletion ,tnt Levels of human amylin were d ~ ...;-.rd using the c~ ;ially available rat amylin assay and running both a rat amylin and human amylin standard curve. Di~,nces in the two cu~ves were used for a co~ ,Liol~ factor for cross~ ivily in the assay.

WO 97t26321 P ~ ~US97/00761 Table 9 Ratios of Amylin to Insuli~ in various transfected cells.

CeD li~le (Trsnsgene) Amylin InsuUn Ratio (n&le6/hr) (n~/e6~hr) EP18/3EI (h.Ins into RIN 1046-38) 1.4 750 0.002 RIN1046 38(none) 1.4 50 0.028 Em series (Rat Amylin into RIN 1046-38) 1134 " 17.4 100 0.17 /138 " 6.4 nd 1117 " 5.4 nd /137 " 4.9 40 0.12 1109 " 3.6 nd Erl82 series (Human amylin into RIN I046-38) Il " 23 50 0.46 /5 ~' 46 50 0.92 n ~ 30 50 0.6 Il 1 " 6 50 0.12 /12 " 40 50 0.8 113 ~ 37 50 0,74 W O 97/26321 PCTAUS97/00~61 Several infl~ lf nt ay~oaches for coenE-.~ g arnylin and insulin in RIN 104~38 cells were used. These in~lud~-l en~ "~g human insulin expression into a rat amylin p~ -g RIN clone (pCMV81INS/IREStP~JRO into BG971134, g."l~aLi~lg the EP183 and 5 EP196 series), eng;.~ g hurn;~n amylin in human insulin Inuducillg RIN clone (pGB7/intron~h.AMYI~ into BG1813EI, ~,n~.d1illg the EP184 and 197 series) and finally Pn~ r~ g human insulin and rat amylin co-e~ ion into I~IN 1046-38 using a single pl~cm cunsllu~;l (pCMV8/INSllRES/r.AMYLIN/IRES/NEO into RIN 1046-38, g.,n.,ldLing EP181 series). In all ca_es individual drug resict~-~t clones were screened by amylin and/or insulin 10 i... ~ " of a 24 hour sarnple collected from cells cultured in fresh tissue culture media with an ectims~te of cell ~l~r. Colldili~ned media Ss nples from the vanous clones were collected and amylin and insulin ;....~ eacLi.~e species quantitated on the same samples. Ratios of arnylin to insulin for these s~ ,ks are shown in Table 10).

CA 0224643l l998-08-l2 W O 97/26321 PCTrUS97/00761 Table 10 Ce~ ~ne ~rransgene) A~ny~n lnsl~n ~a~o (np/n~) (n~ln~) EP181 se~es ~LIns~r~my unto PU~ 1046-38) 0.23 2.37 120.3 0.020 8 2.18 91.0 0.024 4.2 171 9 0.024 4.39 155.9 0.028 3.~3 130.2 0.024 Parenta~ R~N 1046-38 0.028 EP183 se~es (h.INS unto EP97/134) 1.8g 9.3 0.20 /10 ~ 20.3 0 0921 " 10.0 0.19 '~ 14.1 0.134 /22 9.2 0.20 /25 " 12.3 0.15 " 14.5 0.13 27 " 10.2 0.19 " 15.61 0.12 3 " 15.56 0.12 15.8 0.12 " 14.39 0.13 ~ 13.8 0.14 36 8.1 0.23 Paren~ EP97/134 0.17 E~P197 series~hl~M[lY into EP1813E1) 2.03 228 0.009 7.05 620 0.011 /9 ~.15 97 0.012 110 1.05 1 10 0.010 6 1.63 210 0.008 7 2.40 125 0.019 8 3.95 596 0.007 1 5.66 615 0.009 r22 4.55 725 0.006 123 4.90 500 0.010 24 4.22 650 0.006 4.85 535 0.009 W O g7/26321 PCTrUS97/00761 /26 4.48 64$ 0.007 127 5.02 640 0.008 /29 6.10 610 0.010 Parental EP18/3El Q002 W O 97~6321 PCT~US97/00761 Bioreactor production of amylin ~nd amylin-related species from engineered RIN cells.

S Bioreactor Inoculation and Culture. BG971134 cells were grown, split, and .. ~ d in RP~ 1640 l~r~ with 2 rnM ~ JI~ .f. (JRH Biosri~nre) sul~pl~ d with 4% fetal calfserum (JRH) and 0.125 ~Ig/ml G418 ~Gibco BRL) in T75 culture flasks as ~srribed previously in this clo~ A pilot scale bior~tQr (Celigen Plus, New Brunswick .Srientifir (NBS), Edison, NJ) with dissolved oxygen electrode, pH electrode (both lngold), and 4-gas PII) 10 controller is set up for perfusion culture with a packed bed of polyester discs (FibraCell, NBS).
The reactor has a working volume of 1.~5 liters and a packed bed volurnc of 0.7 liters co,.~ g 70 grams of polyester discs.

Cells are ~ s;l~ d and seeded into the reactor con-~ining the same media coll.i,o~ .n as the ",~ t~ n~c media at a density of approx. 2.5 X 105 cells per ml of working volume. After "~r~, the cells allowed to seed onto the bed material with an im~ller speed of 75 rpm and no media ~.fusiol~.

After see~ing, the impeller speed is kept at 75 rpm and the culture is ~ .rd with no 20 pe~r~tCi~n for approx. 90 hours. Media perfusion is started and the flow rate is linearly l~lVU~lt from 0 ~ c,~ing volumes per day (WV/d) to 3.54 WV/d over the course of the following 500 hours. Thc perfusion ra~e is ll~cl~arler ~ t~ Pd c~ C~A~Il at 3.54 WV/d. The yc - r~ cion media is RPMI 1640 with 2 mM ~ r ~ul/pl~ ..e-~t~l with 2 g/l g1~JCOSC (final cQI~rc-~ ll of 4 g/l), 0.10% r.~O~ v bovine semm ~lb~lmin, 10 ~lg/ml human apo-tra~cferrin 50 uM each of 25 eths-lol~mi~e and o-lhnsyh~ ethanols~in~ and 0.10% c~ rt--rol rich lipids from adult bovine serum (all Sigma ~hf~mi~slC, Saint Louis, MO) (a l..ofl;f.r~;orl from Clark, et al., 1990).
The perfilci~ media c~n~A:n~ no fetal calf semm.

The cul~re ~c,~u-e is .--~int~ f-1 at 37~C, the dissolved oxygen level at 60%
30 ~in~Y~ relative to san~ration of air in phosphate l~urr.,.ed saline). and the pH at 7.4. ('.ln~nse .

levels in the reactor is nl-in~ined at 2-3 gJl by adjusting the perfusion rate and the glucose conre~ .on in the fresh p~"ru~iOll mcdia. Similar RIN bio..,a.;lor culturcs have becn m~int~ined sl~cces~îully for as long as Z000 hours in the bioreactor under sirnilar co~ tionc.

S Media s~ les were c~ ,c~(l once daily and ~ ;vely analyzed for insulin and amylin s~l~te;i into the media by EL[SA as pievi~usly (1çs~ ni~ and lactate levels are ....~ .cd in the daily s~.-q,les and analyzed using an ~t.~ t~d analyzer (IBI Biolyzer, Johnson & Johncon, New Brunswick, NJ).

Bioreactor l~r~ ion of rat amvlin (Cell growth and densitY~. The PID controller's oxygen gas controllcr output is r ~rd throughout the run. It rises steadily from around ~0 at O hours to around 60 at approx. 4ao hours where it s~Ahili7es for the rest of the run. The rate of increase of the controller output is c~.-ci~t~ t with an c "p;,;~lly ~ cd growth rate, and the .. level of 60 is co~c.~ t with achieving a cdl density of 1.1-2.3 x 108 cells per ml of 15 bed volume. It should be noted that this growth and s-~ .;..r~ density is achieved using a serum free media This observation is u~-~A~ d Under no~nal culture,; ..-.~ levels a~ t~ d in the range of 2-5 mM which doesn't ;..11.~ ~re cell viability of most cell types. However, in that range ~ oni~ has an 20 inhibitory effect on insulin s~."e~l~ from RIN cer~s. Lactate are nnnn~lly ~ d b~ ..xn 2 and 10 mM, but have in some runs built up to cQ~e~ ~ion close to 30 mM. In the n- ~qlly observcd range of lacl~ate, both cell viability and secl~,tion are not significantly ~rr~t A
However, at the higher concentrations lactate can ilave an cffcct on both filn~ nc 2~ Peak levels of amylin l,le,&s~-,d in the ~~,fused media varies from 25 to 30 nglml, cu.~s~,o~ to 61.5 to 1.8p&/106 cells/hour. Cc~ d to secre~on of cultures on two .1;..,.~l..c;~!n~1 tissue culture flask ~r~es of approx. 17.4 ng/106 cellslhour, the bioreactor se.,.~lion ~.~b~ re~ent a drama~ic drop in ~Çu~ ance. Several factors can be infln~nri~
the s~ potentia~ of the cdls in the bio~actor ~,n~r~.~ nl Cell~ell contact ;nh;~ n 30 b.~ l about by the high d~r.c;l;e~ ~Ichieved in a 3 11;".. ~.c;..--el culture bed; media av~ilq~ ility W O 97/26321 PCT~US97100761 (fresh media available pcr cell per day); built-up of inhibitory levels of ~rnm~ni9~, lactate, and other metabolites or factor.

RlN Cell Lines E~l...... ~ing Illsulin And Amylin Demonstrate Effcienl Proteolytic Processing Of The Prehormo~ To Hormones. Demonstration Of Purification Of Prooe~ed Amylill By HPLC.

Methods HPLC Sc~d.dtion of ~nsulin. Amylin. and Am~lin Related Pe"tidcs. Acid/ethanol exh~rtc of whole cells or con~litiQr~ media was prepared and analyzed by high ~ro,l"ance li~uid cl~.m~.atvgraphy as described (H~ n, et al., 1986, S;~v~ d ~ n, 1991).
.AIh~ .ly, extracts of whole cells were prepared by sonication in 1% Triflou.~ ic acid/
50* accto .; l~ ;le, followed by high spced c~ ~; r~ tion. The HPLC system used was a I~e ~ ... .
15 System Gold with Gold Noveau Software c~ with a modd 126 solvent module and a model 166 ultraviolet absol~iol ~ktD~tor (ncr~ ,s~ , Inc.) and a FC204 ~ I;on collector (Gilson Inst~ r~t~, Inc.). A Merck LiChroCART 250~ with I irh~ ; 100 RP-18 column (25 cm x 4.6 mm) in cG.,.l~h~tion with a ~ ;rhOS~ f r 100 RP-18 guard colD was used for chlo~ ographic separation. ~1~ ti~,ely, a Vydac 214 TP C4 colD (25 cm x 4.6 mm) 20 (Alltcch, Inc., ~rfi~ IL) in cc,~b~ation with a Mac~ , 300 C4 guard colD (Alltcch, Inc.) was uscd for chromatographic scparation. Solvent systems, gr~ C and flow rates were uscd as d~sc.i~l (Halban, et al~r 1986) as was ~ cc~c;ng of individual fractions including n~ lt~r~li7~tion and iyopholizadon. S~dards used in~ d~l hum~ proinsulin and insulin, rat amylin (Pe ~ Labo,atolics, Inc., Cat. #7323), and human amylin (~ ~h~m Inc., PCPE60).
o~,. cipilation of Amylin and Amvlin-Related PeD~idPs RlN cell lines (parental RlN 1046-38, rat amyl~ pro~nring BG97/134 and human amylin pl.J~ BG182112) were rnctabolically labeled in 6-well tissue culture dishes using 35S by culh.ri~ cells initially for 20 minutes in DMEM laclcing m~thioni- ~ and cystine (ICN Bio~ alc~ Inc., Costa Mesa, Ca) 30 followed by a lwo to four hour incub tion in the samc media with 17 uCi/ rnl. Trans-Label (lCN

CA 0224643l l998-08-l2 W O 97~6321 PCT~US97/00761 Phalmace~lticalc~ Inc., ~vine, Ca). For pulse chase/ eA~ .t~, cells were metabolically labeled for 3 hours followed by a one hour chase by incubating the cells at 37 C in RPMI with S
mM gll~eose, 0.1 % BSA, 20 mM HEPES, 10 mM each le~lcin~, &E;inine and ~;lu~;.r~ f, 100 uM
carbachol and lC10 uM IBMX. Media was coll~cte~l after the 1 hour i.~ b ~;~n, Triton X lC10 S added to a final c01-~ P ~lration of 2% and ;...~ n.,cil,i~tions done as dcs~ ~ ;b~ below for the cell Iysates without di~ nn into the antibody binding buffer. Cl~ifiPA cell Iysates were ~ ,d as dcs~ ;bed using two dilr~,cnt lysis bllff~-rs. NP40 and RIPA lysis buffer (from "A"~odies. A
Laboratory Manual", E. Harlow and D. Lane, Eds., Cold Spnng Harbor Lab~ tul.es, 1988, pg.447). Lysates were diluted 10-fold into antibody binding buffer (2% Triton-X, 150 mM NaCl, 1 mM EDTA, 50 rnM Tris, pH 8.0) followed by the ~ Iditio~ of 5 ~lg amylin-speeific IgG and b~ overnight at 4 C. Amylin antisera used in~ c~1 rabbit anti-amylin amide (rat) ~gG, ra~bit anti-amylin 25-37 amide (human) IgG (both from p,!..;..-..l~ La~c,lies, Inc., Belmnnt, Ca), and rabbit anti-amylin amide Irat) IgG (Pieber, et a~., 1994). TmTmmo cc ~1-YÇS were isolated by bit.~ g for ooe hour at 4 C to 20 rnicroliters of a 20~o slurry of Protein A Scph&~sc 15 beads (Sigma (~h~ Inc.) rehydrated in 10 mM NaBorate, p~I 8Ø Protcin A Scpharosc beads were p~ ted by c~s-tnf~lgation and washed ~ lly in wash buffer #l (150 mM NaCl, 0.5%SDS, 10 mM Na Borate, pH 8.0) and wash buffcr #2 (150 mM NaCI, 0.1%SDS, 10 mM Na Borate, pH 8.0). Tmm~ o comple~es werc ~ .o._d from the Protein A Sepharose beads by heating for 30 t~ S at 42 C in 3() microliters sample buffer (2% SDS, 12 % glyccrol, 2% beta-20 ~ .C~'tG~ .nl, .01% blu~ ol bhe (all w/v), 50 n~ Tns, pH 6.8).

T~ .... ~.oplY~ ita~es were sep;~ated using a gl~,,ol cor~ g ~p~ gel (16.5 %T,6 %C) as ~ il~d (Schagger and Von Jagow, 1987). Five ~ u~amS of ~ th~c human amida~ed amylin (BACHEM Califomia, Inc., Torrance, Ca.), rat amidated amylin (P~,..;l.~..l~
25 LaboraIories, Inc.) or 30 microliters SeeBluc pre-stained mr)l~r~ r wcight ...~ (Novex, Inc., San Diego) wcre run as standards in parallel lanes. Following e~ o~ c~ gels were divided and ~ ce~ T ~nes co.-~ ~~ itates and one ~ol~ -'ar weight marker lane were ~,oe~ l for nuu u~5l~hy usi~g r"t~ ~~A;ry autoradiography enhal~cer as ~k~;hed (NEN
Resear~h P~ , Bo~ton) and r~l~5~ to Xomat-AR autoradiogra~hy film (Kodak, Inc.).30 L,ancs col~A I.;..g mol~ltl~- wei~ht markers and the amylin standards were stained Brilliant Blue W O 97t26321 PCTrUS97/00761 R (Sigma Ch~mic~l~, 25 milligrams per 100 mls. 10 % glacial acetic acid) for one hour followed by llf 51~ in 10 % glacial acetic acid to vi~u-q-li7~ relatiYe mi~ation of the amylin standards.

Results:
S Novel posttranslational pr~ '.. 5~ has been achieved inrh~ n~ dibasic amino acid p.otLase pl~ces~ from proamylin, ~mi-lqtion, and glycosylation. Naturally oc~-....;..g hum n amylin extracted from blood or human ~ as has been shown be very hct~r~ge .P~!-c (Pittncr, et a~., 1994 and Percy, et al.. 1996). This h~,t~"~a~n~,ity is largely due to O-linked gl~;Gs~rlation of amylin in one of two l)ossiblc sitcs in the protein (Percy, et al., 1996 and l~ nho~e, et al., 10 1996). The degree of ~t~r~ City of RIN produced rat and human amylin is being d~ d at Re~r~n~ by HPLC in c--mhin-qtio~ with a sensitiYe ELISA for amylin as well as amylin ;... ~ ~op.~;pi1~l;or~ and SDSIPAGE analysis.

Human Proin~ iD Is Frr~ occssed To Mature !nc~llin BY Rat ~cn~inn~nq Cells.15 lntr~cllular insulin species wcre isolated from parcntal RIN 1046-38 ~nd EP 18/3El cells by acid~ ol e~ctracdon. Separadon by HPLC of the insulin species l)lu.l~ced by these cells was done as ~rs~ibed (~-q-lb~-~. et al., 1986, Si70n~nl~-~ and Halban, 1991). The analysis indicates that human insulin produced by the rat inclllinnmq is effiriently ~ ced to rnature insulin with very low detectable levds of pro-insulin or othcr ~ cci,~ .It~ r~iqtcs (i;IG. 18). P1VCC4~
20 of human insulin is slightly less err~ t compared to the p-oc~ of rat insulin at these levels of ~ ;O~ While the ~ r~ c of intracellular rat l~n~ lin and i~ liates is 3 to 4%of tot. l rat insulin in all cell lines examined, thc ~.,;- ht~ge! of intr,~çlllllar human proinC~lIin . nd ;"~ tcs is 18% in EP18/3E1.

The ability of RIN cell lines to efr~ :e~ S5 ~l~J.~ to mature insulin in these r~ -.,d lines d~ ~~o~.C!rates the ..~ tc.lance of the high levels of e~ ssion of the cndo~l~)teases PC2 and PC3, known to be le;,~s;blc for insulin ~ ,ss~ (Vollen .idel, e~
al., 1995). This is a cntical feature of the R~ cell lines being de~ )}~d. HPLC analysis of se~ d fonns of insulin following stimulation ~ tes the human insulin l~,lca~s~d is 30 ~l~du~antly ma~re insulin with ~cry low detectable pro-insulin or ~IOCJ~ ates .

W O 97/26321 PCTrUS97100761 appearing in the media. These values are lower than mez~ul~,d in vivo circ~ ing levels of proinsulin (Rhodes and Alarcon, 1994).

Stim~ t~l secretion s~mrlcs of parental RIN 1046-38 cells, the rat arnylin produr in~
S clone EP971134, and thc human amylin ~rv~ g clone EP182/12 were ~ ,~ed as desçribed in F-~le 7. Following this i~rl~b~ion~ cell ~All~t~ were prepared of thc cells to ~lct.~ ...;n~ the intracellular co~ of the thrce cell lines. Extracts wcre prepared by two m~th~c. The first was an extraction of the col~t~ts into 1.0 M acetic acid as pc.~ulll.ed for insulin (Halban, et al., 1986~ t~rn~ively~ cxtracts were pl~l,ar~d using 1% triflou.~xt,lic acidl 50% ~elol.i~
10 (I~AIACNextract). This~lt~ tiveextractionwas~ .teAton.;l.ii.i7eamyloidfo~nationof amylin species which would be lost from thc C%tf~CtS by c. ~ ;rug~tio~ T.. --- o.~live amylin species in ng per million cells from tne above samplcs are shown in FIG. lg. Assays wcre done using thc rat amylin assay as describod in Example 7.

As e~ c~-~, parental RIN 104~38 cells secretc a low amount of mqte I with low amounts of eA~I~ble intr3eelh~ r amylin spe~iec~ The rat and human arnylin en~j~u~ lc,d lines both ~l~ntonct~e highcr 5~."o~...t~; of secrcted and in~7re~ r arnylin ~eri~s Thc rat amylin ~",~h.. ;..g clone sec.~s about ten times more rat amylin than the parental cclls and has intracellular ~ -.o~ 20 to 28 times more than parcntal cclls. The ~raction of ;~ O~ !;vc arny~n s~cl~d in one hour is 14 to 20% of the intracellular cont,ent Ap},..J ..l"atcly 40% more i . . . - -.~OI ~~tive ma~erial was extracted using the TFAIACN
procedure. The hurnan amylin producing clone h d lower values overaU than ~ d Even though the samplc:s were read against a human amylin st;uldard curve, the 25 crossIeactivity of the ssay may still be a probkm and give lower v~llues. The amylin assay utilizes rat a~nylin as the co...~ species. The human amylin p~ c;--g clones hadap~u~ ately three times more ;~u ~ d material compared to parental cells.
Intracellular content was two to six times higher than parental cells. IntL.~ gl~, the acetic acid e~ctract gave the higher values as compared to the TFAIACN e~tract. The ratio of SCCI~ to 30 ir~~ r mqt~n~1 was 17%, a value very similar to the rat amylin plo.J ~ clone.

W O97126321 rCTrUS97/00761 Rat AmYlin HPLC DemonsLl4Lin~ ~locess~-~ And ~lriff ~rion (I~G. 20A). FYtr: ~tc of a rat amylin ~lo(~cin~ RIN EP183/20 (see Table 10) were prepared and separated by the same HPLC system and con~ as used for separation of the insulin speri~s S~"1.L~lic :~mi~5lt~
5 rat amylin was also run to ~t~t~ ...;..c its retention time. Plots of the W absull,auces versus time for the HPLC runs of the RlN c~ctract are shown in FIG. 20A. A major W peak at app. 21.5 ";"..t~s was easily ~ rl~ d as well as a minor W peak at app. 20 ...:I .~,s. Individual f.~
assayed for amylin immlln.~ca~Liv~, species across these W peaks ~ tcd a peak of ..--o~ vily that again coinrid~d with the major UV peak at 21.5 ..~ ,s (FIG. 20A).
10 Again the minor peak at 20 ..-;..~"~s from the RIN cell extract showed no imm~nQlea~;Li~dLy~
Only the indicated ~ ,LiOnS were assayed for amylin i.. ~--ol~activity. There is a strong Foss;kility that other W peaks could consist of amylin-related sp~~ies There are several W
peaks, such as those at 7.5~ 16.3, 17.6, 26, and 31 ...;....l~c, that are as yet .~ c.~;r~
Uni~ ;r.c-l amylin-related species are e~rected to be present in a RlN cell extract, inr~ in~ o linked ~;ly~;o~/lated amylin spc~ . It is ~ se.~ not known wllcthcr the rat amylin ;.. ~ -o assay bcing uscd is capable of l~CG~ g alternatively ~, ~ess~ fo~ms of amylin such as Gsrlated species, or any other amylin related ~iperiPs. The W pealc at 26 ~ t~ S does not contain any amylin i.~ ,active material. but this W peak does c~_ide in ret~nti~n time to p~cc~ssed human insulin (FIG. 18).
HPLC Fractionation of E~tracts from Human Am~rlin P~ JC;~ RIN Cells. F~tr~~tc of a human amylin ~i~lu~ g R~ BG 182/12 cdls and parental RIN 38 cells were prepared using t~e TFA~ACN extraction ~vlucol and frac~ t~d using a C4 c-~h~nn as ~rse-ibe~ in Materials and ds (E:IG. 20B). This çytr;~~tirln ylOtOCOI and HPLC column are i~ Al to those used in 25 Pittner, ct al., 1994, where the h~ ti ~g~ -e~v~ nature of n~ttl~lly oc.; .,.;.~ human amylin was first described. HPLC fractions were assa~_d using the anti-amylin pol)relollal serum (T48~6, Pieber, elal., 1993)as d~nk~ inMaterialsandM~!I.~c TheBG 182/12~ kswereread against a human amylin standard in the RLA while the RIN 38 samples werc read against a rat amylin standard. The rcsiti~nc of the fractionated human and rat standards are also indicated.

W O 97/26321 PCT~US97/00761 The l~iD 38 extract shows vcry little amylin i~ o~ ;Livily in any of the fr~ s (E;IG. 20B). A minor pcak in the initial void volume fractions (fraction 4) and in the fractions where the rat amylin s~d&,d fr~ion~tes (r.~ ~ C 11-13) are all that is detected. This most likcly ~ ts the low endog~nous levels of rat amylin found in parental RIN 38 cells. In ~i C(il~h~-;, sevcral large peaks of; .~ on,a~ re material are ~ in fi actions from the hurnan amylin ~,lo~hl~ g cells. The largest peaks are found in fractions 27 through 32, with possibly two major species rcsolved. Human ~ ted amylin standard runs very close to these peaks tiorl 33). Other minor peaks are found in the initial void volume (f.~ ~ 4) and fractions 11 through 17, again close to where the rat st~ld~.l fractionates.
The identity of all the fr;~tion~d ~ re species in the human amylin plud~ g cells is presently not kn~wn. Based on t'he lack of ;--. -- ~ .ol~active material found in the RlN 38 extract, all of these peaks are lilcely to be dcrived from various }~l~e-cC~
;..t ~ iates of human amylin. b: is interesting that the h~ ,.o&e~ c nature of human amylin 15 pl~nlwed by RIN cells is similar to that ~es~ 1 for human arnylin e~ ted from plasma (Pitt~ler, e~ al., 1994). This report shows threc major spccies of human amylin, one ident~ in HPLC fractiol~tinn to the synthctic hurnan amylin pepti~, with two othcrs fractionating just prior to the al~dard. The e~t~tion and HPLC coll~mn used hcre were the samc as reported in Pittner, et al. with the only dir~ ce being thc ~~etollitrile ~ dicnt CQnflitionc used. It is 20 ~.cs~-,lg to spec~ t~ that the het~.og~--e.r~c human amylin produced in RIN cells is at least similar to naturally occl~nng human amylin. Of note is that there is no cvidence that rat amylin .c..l ~ced in RIN cells is hete~2.-~le~,u~ (I;IG. 20A). This agrees with thc results of Picber, et al.
(1993), where rat amyl~n c~tractcd from plasma and fractionatcd runs as a singlc species at the idcndcal rePntinn tirnc as synthetic rat amylin.
Recomhi-~t Human Amylin is IIet~ eous with Apparent Molecul~- Wei~hts of 4 to 8 Kd. RIN 1046-38 cells or human amylin ~ u ;..~ BG 182J12 cells were metabolically labcled using 3535 and clarified Iysates prcpared using two di~f.,.~t Iysis buffcrs.
;o~ of amylin spccies from these e~ls were r~ .t~d using three amylin 30 ~ ;f s Two coll~,cially available ~ltiboAi~ raised against citha amidated rat amylin or a fragment of ~mi~~~Pd human amylin failed to i~ y~ J-P, any labeled spPci~s. A third ~ tiho(~y raised against arnidated rat amylin (Piel~er, et al., 1994) i~ iLated labeled species specific~lly from the both eYt-r~tC ~ alCd from the human amylin prodllring R~ clone ac compared to parental RIN 1046-38 cells (FIG. 21A). One major species lUlll illg at an ~palc.lt I ~lf,.11151r weight of 5.5 kD and a IIunor species running at S and 7 to 8 kD are seen.
There is a very faint band at the ;~p&~,nt mn~PcllI~r weight of 4.0 kD where the human ~nni~1qt~d amylin ~d~d mi~q-teC (FIG. 21A). IIG~ , this is also close to where the rat amidated standard migrq-t~s~ so this 4.0 kD band could lcp~e~ L either human or endG~I~o.ls rat amylin.

A pulse chase e~pe,i~ t with a long pulse followed by a chase in media that stimulates s~,~tion of amylin immllnQreactive mqte~iql was ~ r1 Results of imm~moyiccipit~ions and SDS PAGE frac~ nqtiort from the cell Iysate will.uul a chase, ccll Iysates following the chase and from the cQllect~d media are shown in FIG. 21B. The four ;~ cipilated species of with ~ c~lt m~leclllsr weights of 4.0, 5.0, 5.5 and 7.0 to 8.0 are easily seen in the cell lysate without a chase. All four species are l~luced in intensity in the cell lysate following the one hour chase with a cul~c;.~ollding inc.~ase in the four species in the media. Tnis ~I,o~ly sugg.-c~c that the h.,t~ eo~ irmnunG.~a.,1i~e amylin m~te~ d-lced by the hurnan amylin e~ e~,d RIN cells is Crfi~ ly secreted and remains ~t~,.o5~nro~. The ratio of the 5.0 versus the 5.5 kD band is dirf.,lcnt ~el. ~n the intr~elllll~r rnqte~ from cell Iysates is compared to the se~,-e~d m~t~ri~l, possibly ~ s~ further Ixoces~ e e~ents.

R~o--.k;.~ human amylin species of an spparent molecular wdght of 5.5 and 7 to 8 IcD
are ~ ntic~l to amylin species extracted from human plasma (Percy, et al., 1996). It is now ood that the h~,t ,.o~ nP~.c nature of natur~lly oc~ - . ;..g human amylin is due to O-linked 25 g~ ;G~lation(~i~t~nhnlJce~ etal.. 1996). F~ toallalrLethe hc~ oge~r~...c l~cc-.~h;n~~~
human amylin ~l~Jdu~d from RIN cells is underway, p~;r.~lly looking for O-linkedgl~ co~lated species EXAMPLE 10RlN Cell Lines Expressing Preproglucapn Demonstrate Efficient An~idation Of A
Secreted, ~r~c~ Polypeptide.

S Methods:

)ro~luca~on cDNA isola~ion and e~ sion pl~crni~l Total rat ~/allC~tlC RNA was reverse ~ cwibell into total cDNA using AMV Reverse Tl~1s~ tase as ~co....l.- n~led by the supl~liel (Promega, Inc.). A rat glucagon cl:)NA corresponding to bases 10 to 904 of the 10 published sequence (Heinrich, Gros e~ al., 1984) WâS ~mp~ified with the polymaase chain reaction from the l~al~elealic cONA using oligos (CCAC(~ lACACCTCCTCTC (SEQ ID
NO:36) and GTAATCCAGGTGTCGTGACTGC (SEQ ID NO:37)). The res~ltinE 895 base PCR
~o.h,~l (SEQ ~D NO:55) e~ o~ g rat ~,lcpfo~ ~on (SEQ ~) NO:56) was lîgated into pNoTArI7 as l.,c.JI.u.. ~1-.1ecl by supplier (5 Prime to 3 Prime, Inc.), ,~;~,ne~g pNoTAr7/r.Glucagon.

A human ple~,oglucagon cDNA was generated by the pOIyl~lC-~;C chain l~:Lioll using oligos (CAGCACACTACG4,GAAGACAGC (SEQ ID NO:64) and~0 CACCACTGTGGCTACCA~l-l~;l (SEQ ~ NO:65)) ~d hum~ p~creatic cDNA ~ a 't~, (Human I'~ulcl~se QUICK-Clone cDNA,7115-1). The res~lltin~ 955 base pair r.~,.. ~ (SEQID No:57)encoA~ human prepro~lucagon (SEQ ID NO:58) was ligated into the PCR cloning vector, pNoTl~I7 (5 Prime to 3 Prime Inc., Ro~ld~r Co.), gencrating pNoTAT7~gluca~on.pNoTA~'~ ~onw~fc~ t~ with BamHl and the r~s~llting g84 base r. .6...l~nt co..~ g the human arnylin open reading frame was lîgated into the BamHl site of pCMV8/IRES/NEOmGHPoly.A" gc,~e.~ g pCMV8/hGlucagon/IRESlNEO. The CMV
~lU ..ot~, drives !~ ~C~ ;Oll of abi~ vllic ~ C~.~r ~NA with human ~ ~or ~.~co~ in the u~ open reading frame and the nconl~in r~J;~ e gene cllrodc~ in the d~wllslleam open reading frame. Stable trar~cf~ctPntc from this pl~ ed in G418.

Con~llu~;lion Of Mutated Hu~nan ~lc~Jluducagon cDNA And ~xpression Plasmid. A
~"~ t;t~l was i~ duced into the human ~l~plvgll~r~gon cDNA by the polymerase chain reaction using sets of ûligos for site directed mllt~gen~sic with a plasmid CO.~ ,i,.g the human preprogltlr~gon cDNA as a teln~ te (pNoTAT7/h.glucagon). The first set of oligos5 (CAGCACACTACCAGAAGACGC (SEQ ID NO:66 and CTGTAGTCACTAGTGAATGTGCCCTGTGAATGGGCCTTGTGGTCG (SEQ ID NO:67)) generated a 215 base pair fragment with a single base pair change of G to C at position 181. This base change results in an amino acid ~lb~ ;û~ of ~6illh.e to alanine at amino acid residue 52 of ~ vgl-~cagon. This fragment was ligated into the PCE~ cloning vector, pNoTA/T7 (5 Prime 10 to 3 Prime Inc. Boulder Co.), ~cneldlillg pNoTAT?/5' mut.gluc.

A second set of oligos (CAGCACACTACCAGAAGACAGC (SEQ ID NO:64) and ACTAGTGAATGTGCCCTG (SEQ ID NO:68)) was used to introduce a second mutation usingpNoTAT7/5'mnt glll as a template. The reS~llting 207 base pair f~n.~nt in alAitiol~ to having 15 the origjn~ G to c base pair change at position 181, had a second change of C to T at posi~;ol-204. This resulted in the creation of a unique SpeI ~ ion en~l nl-rle~ce recognition site, with no change to the dPd~lce~l arnino acid se~lucllce. This 207 base pair rr~l~le.lt was liga~ed into PCR cloning vector, pNoTA/T7, g~n~ g pNoTAT7/5'mut.gluc.Spe. A final set of oligos (ACIAGTGACTACAGCAAG ~SEQ ID NO:69) and cAccAcTGTGGcTAccA(~ (SEQ
20 Il} NO:65)) was used to introduce the Cv~ SpO~ g C to T base pair change and the crea~ion of a unique SpeI restriction en~io~ sce l~cG~ilion site in the 3' end of p~ nvglucagon. The lting 755 base pair C~,6~ nl was ligated into the PCR çll~n;ng vector, pNoTArr7~ g~ ~g pNoTAT13'mllt ~lllr ~~pe. pNoTAT715'mllt ~ r ~pe was digested with Spel and HinDm and the L~ nt ligated into pNoTAT7/3'mllt~ r.Spe that had been also ~1ig~sted with Spel 2~ and HulDm. The res~lltin~ pl~cmid pMTAT7/mut.glucdgon, contains a ,~consl,u.,~d vgl~ oncDNA(SEQII)NO:60)with2basepairch~ng s inthese~rl~reatposition 181 (G to C) and 204 (C to T) relative to SEQ ID NO:57. The rÇs~lltin~ amino ~cid se~ven.~e (SEQ
ID NO:61 has a single amino acid s~bs~ n of a~ e to alal~ine at amino acid residue 52 ~lative to SEQ ID NO:~8. pNoTAT7/ml~t gl.~cs~ was ~ with Xbal and BamHl and30 the ~eSllltine 976 base r.,~ L~t cu~ the mutant l.rt~.uel~ eon open reading frame was W O 97126321 PCTrUS97/00761 ligated into the Xbal and Ba]mHl site of pCMV8JIRES/NEO/hGHPolyA, gene~ g pCMV8/ml)t gltlca~onlIREslNEo~ The CMV promoter drives tl~scll~ion of a bis~ onic sse~g,- RNA with the mutant preproglucagon çnro~ed in the up~e~ll open reading frame and the neomycin reCic~nre gene ~nroded in the downstream open reading frame. Stable S ~ lsf~,clants from this pl:~cmi~ are sel~ct~(l in G418.

Cell Culture- R~ 1046-38 ~Gazdar, Chick et ul., 1980; Clark, Burnham et al., 1990), Rin 1027-B2 and Rin 104644 (Philippe, Chick et al., 1987) were grown in M~iiu~- 199 with Earle's salts, eQ~t~ining 11 mM glucose and sull,lP .~ Pd with 5% fetal calf serum (~ip,r~ch 10W~chingt~ D.C.), 100 millillnits/ ml penirillin and 100 ~lg/ rn~ e~tonl~cin. AtT-20 derived cell lines were cultured as des~ (Hughes, Johnson et al.. 1992). Cells wcre pqc~e~i once a week using 0.05% trypsin- EDTA solution and kept under an ~l...nsl~h~le of 95% air and 5% CO2 at37C.

15Northern analvsis- Total RNA ~rom RIN cell lines grown in vitro was ico~ted using RNAzol B RNA Isolation Reagent (Cinnal Biotex Laboratories ~nt.). 10 ~lg total RNA was resolved on methyl llle-~.;UI~ 1.5% agarose gels as described (Bailey and Davidson 1976). Gels were sl~bse.~enlly stained with eth~ m brornide (l~g/rnl in 0.~ M NH4CH3CO2) to visualize RNA for integrity and loading con~ictçn~y. RNA was electro transferred to Qiabrane Uncharged 20 Nylon Me~h.~ (Qiagen, Inc.) using a Transphor Unit, TE50X (~oefer, Inc.). Me.lll~ es were hybri~li7eA with rligoYi~nin-labeled RNA probes using the Genius Non ra~lio~cti~re Nucleic Acid T~lelin~ and l~eter~ion System for filter hybri~ as d~Y ;hcd (Roeh ;~ .r M~n~h~-im). Full-length digo~ri~T~in-labeled ,~nl;c ~se probes c~ ,s~ gto rat insulin (SEQ
ID NO:59), rat amylin (SEQ n~ NO:7), and human glt~e~Qn (SEQ ID NO:55) wcre made using 25 Genius 4 RNA I~I~P1jng Kit (Boe-hrin~er ~ h~;",) using either T7 or T3 Polymerase.
Northern analysis of rat peptidylgl~.ycine alpha-~rni~ in~ monooxygenase (PAM) in cell lines was done as dcs~ ;1~d using a digoxigenin-labeled ~ c probe cG~ onAil~g to the bases 240 to 82g of the rat PAM cDNA (StoiFfers, el al., 1989) made using Genius 4 RNA ~ ~klin~ Kit ~Ro~k~ e~ l Maonheim) and T7 ~,oly~c..,se. r~l Gs~-_s of ~h~ scrn~ de~cted 30 membranes was done using Xomat-AR ~ OI~ography film (Kodal~).
Ig9 W O 97126321 PCT~US97/00761 lmmllnoaSsarS For Gluca~on~ GLP-I (Non-~miA~te(l) and GLP-I (~ ted).
Tmm-lnoreactive species of glucagon, glucagon-like peptide 1 (7-37, non ~mi~i~tf d) and ~luc~go~rlike peptide (7-36, amide) were ~le~ d as descri'oed by the ~u~lie,~ of the lespecllvc col.lllle,cial kits (all ~,l.chased from Pellin~ Laboldto~ies ~c. Cat ~s RIK-7165, RIK-7123 and RIK-7168"~spe~ rely) Results:
Northern analysis of several of the rat amylin prod~lring clones (BG97 series) was done 10 probing for amylin ~ sc~s ~endogenous and tr~n~gen~), PAM and endogenous insulin as shown in FIG. 22. Amylin tr~n~nc ~. ~cc~ge~ and e~lo~nous insulin mpss~es varied bet~
clones with n~ ss~ge levels being con~ict~nt with d~ t~rct. ;l imm-lnoreactive horrnnl~ levels (Table 9 shows insulin and arnylin hQrrnnne levels for these clones). Endogenous PAM m~S~pge in the amylin procl~lring clones co.,l~&ed to parental RlN 1046-38 is very stable. In fact, the rlJajolil~ of RIN 1046-38 derived clones cv~ Je to express both endogenous amylin and PAM, s~lg~sting that RIN 104~38 derivcd clones will m~int~in the ability to e~ri-~-ntly arnidate peptide h~,"~

The high level of PAM t~ lcssion in RIN 1046-38 colll~r~d to AtT-20 is very 20 e~ n~. Comp~ncol~ of PAMi c~lcs~ion In other cell types has shown that AtT-20 cells express vely high ~ ylllc levels (T~ chi, et al., 1991). This inrhldPs higher levels than PC12 cells and RlN5-f cells, a rat inc~llinom~q line that is fairly de-l; rr." ~ ~I-Ptecl when colll~d to RIN
1046-38. ~AintZ;~,;ng high PAM activity in RIN 1046-38. similar to m~ t7.;..i.~g high levels of PC2 and PC3 activity, sllg~estc o~ ,Aylession of ll~c~s for amidated peptide hnrmnn~s such as amylin will result in their elfi- i ~ --t ~du~ l;m~

Post tr~nclA~inn~) ploces~ of ~lc~ gl~,oll into glucagon is depenAcn~ upon the spe~iq~17~d fimcti~)n~ found in cells with a regulated se~l~,toly pathway. This is true in t_e c~nAG~n~ s cells that ntn~nqlly make gl~ g~ (p~l~lC~t~C alpha cells) and, as ~1e~n~ t.~d in this patent, is true for R~ 104~38 cells. E~ sioll of pl~,l)rc,~ c~on tr~n~g~n~s in a variety W O 97126321 PCT~US97/00761 of cell lines has d~monc~rated cell-specific dirr~ ces in procescing (Drucker, Mojsov e~ al., 1986). RIN 104~38 cells have the capacity to produce, process, store and secrete hurnan insulin as ~emor~ctrated in the above ~ mple. This inrllldes high endogenous .,A~I-,ssion of PC2 and PC3, endoproteases involved in ,~)roces~ g both insulin and glucagon (Vollenweider and Kaufmann et al., 1995) (Rouille, et al.. 1995). RlN 1046-38 clones o~,.e,.p1essing pl~o~ on tr~ncgençs should effi~iently process the ~1.,cul~or into mature glucagon, store it i~ s~l~tvl~r gr~ lll.s and release glucagon upon stim~ tion to the media. This would of course be useful for both large scale production of glucagon as well as in vivo cell based delivery of ~,lucagoll.
RIN 1046-38 cells should also l,rocess In~ vglucagon into glucagon, GLP-I and ~1. Final rnaturation of GLP- 1 involves IC-terminal amidation by ~,.,pLidylglycine alpha-~miA~ing mOllOO~g~laSe (PAM). F.n~ c. - ;ng RIN cells to ~ -;n~ntly produce glucagon or GLP-l is ~os ~1~ by mCle~ r e~ cc-;.-g. ~Vcesci~g of l,l~ro~lucagon to glucago11 is ~e.1c....;..:~ ~tly by the action of PC2 while ~loce~ to GLP-l is pre~ln~ y by PC3 (Rouille, et al., 1995).
Ov~A~,1es~ion of eitber a PC2 or PC3 tr~nsg~ could result in predc,1..;~ e~1ession one peptide h--rrnone over another. ~ ely, the glucagon tr~ne~-o.nt could be .1.--~-~- d such that the dibasic amino acid residues reco~i7ed by PC2 and PC3 are altered such that only ~ ca~o~
or GLP-l is capable of being plocei~d to the mature, biologically active pol~ le.
Human Prepro~luca~on E~cpr~ssion in RIN 1046-38 Cell Lines. Rat insl~linoma cells have been shown to exprcss the ~luca~o~ ess a~c (Philippe, et al., 1987). A series of RIN cell d~.ivdi~_S all t)riyj~ g from the same origin~l in~lllinom~ (Gazdar, Chick et al., 1980) wcre s~ cd for e~ ,ssion of endGge~us ~ eagon m~Ssage While RIN 1046-38 cells do not 25 express e~dogcn.~c glucagon, other RIN cells do (Philippe et al., 1987), sugge~ g e~p~ ssion of preproglucago1l in RIN 1046-38 cells is pcs~

pCMV81h.Glucagon/IRES/NEO was cle~;l,u~olaled into RIN 1046-38 cells and stable clones sel~rt~d in G418. Individual coloni~s were sc1~,e~cd using the non-~ t~d GLP-l 30 im~--n~esay. Once individual clones had been established in 24 or 6 wcll tissue culture plates, a 24 hour media sample was collected and assayed. Cell llulllbel was estim~e~ from visual inspcc~n of the culture plates. Clones with obvious iuc.._ased GLP~ ..o-eactivity were iAentified and char~ ;7F~ further.

S Three such individual clones (BG175/5, 174119 and 174/45) were eY~mined for regulated se~ ion of ;..~.n...~o.eac~ive species of GLP-l and glucagon. r'nltl~rin~ of cells under either a one hour basal condition (B), a one hour stimulated condition (S) and a 24 hour condition (24) were collectP"l and the same sample assayed for non-~mi~ ed GLP-l, ~nnitl~t~-l GLP-l and gll~e~nn Each i~ oassay is spe~ifl~ with no clossreactivity according to the specific~tinns 10 supplied with the kits. Isnmunoreactive hormone levels for these three wild type human ogl-~a~ol~ e~rcssi-lg clones as C~jlllp ~,d to Im~n~ e,~,d RIN 1046-38 cells are giYen in Table 11. Parcntal RIN 1046-38 cells have .. ~ tc~;lOble or very low levels of GLP-1 or ~]n( ~on spe~ies All three wild type ~ rugh~ On e~ncs~ g clones had easily dct~ blç levels of all three h... n~ cs. Comr~ri~on of stim~ tP~l to basal s~m, l~s de..lo~.sl~alcd a good fold inrh~c~tic!n 15 in se.,l~lion for all ho~ nl~s Of interest is the pc.~ ge of non-~mi~tpA GLP-1 to total GLP-1 (non-amidated and ~miri~~cl). This ~.CC.m~, under a range of c~ in~ con-litionc in all thrce clones is 49% +9%. This pe.~ tae, of RIN secreted GLP-l species in vitro is very similar to t_e pc.c~t~EP, of GLP-l species secreted in vivo by the elldo~.l~Jus L cells of the ;.~t~
(Mojsov,eta~., 1990).

CA 0224643l l998-08-l2 W O 97/26321 PCT~US97/00761 o ~ ~ Z

;;~ O ~ d~ i U~ X
C ~ ~

O Z

-- V~ , U'~ _ ~ ~ O
* *
c~ u~ x X ~ ~3 ~ ui X ~ ~ ~ ~ ~ '-- ~ --o ~,j _ o ~ o VlU~ o ~ ~o oV~ ~ --* , -v ~ ~ v~ ~ v o v~ o Z o~ ~j ~ ~ -- X o ui _ ~ i ~ Z

~; m C~ C~ ~ m ;:
C~

e 191J26 B 1.74 0.58 75 1.45 S 10.37 5.96 6.25 10.87 62 l6.5 24 31.33 16.64 65 0.54 ~
191/30 B 0.6 0.28 68 0.54 '~
S 2.59 4.28 4.95 17.67 34 6.1 24 16.45 13.99 54 51.49 AVERACE 3.76+1.5 11.06+4.0 59+13 RIN 1046- B 0.08 not na notdetected - 38 detected S 0.09 1.09 0.23 NA 27 not detected r 24 0.08 not NA 0.19 r detected .p * values are in ng/ml O

WO 97n6321 PCTrUS97/00761 Mutatcd ~ ,,o~luca~on Ex~ression in RIN Cell I ines.
pCMV8/mllt~ntGlucagonlIRES/~O was elec~ orated into RIN 1046 38 cells and stableclones sel~cte(l in G418 Individual colonies were screened using the non-~mi~ od GLP~1 i ....~o~s~y as done before for the wild type preproglucagon e~p~e5sing clones. Clones with S obvious il,creased GLP-l were again i~l~ntifi.o.d and cha,~tf ;7~ further Pour such lndividual clones (BG 191 series Table 9) were çx~min~l for regulated s~,~,o" of imml~noreactive species of GLP-1 and glucagon. (~ hlnn~ and assays for non-~mid~-cd GLP-1 ~mi-i~tt-d GLP- L and glucagon were done as before AlLI four mutant 10 p~progl~c~on ~ ss~g clones had easily ~1etect~ble levels of all th~ree hormon~s (-ol 1 7 ;~ol~ of stimulated to basal sz~ lrs again d ~ ted a good fold inti-.rtion in secretion for all horm~n~s The ~.centage of non-~mi~ GLP-l to total GLP-I (non-~m~ l and amidated is 59% +13% again very similar to the wild type e~ essi~g cells or in vivo levels Gluc&g~n imm~lno~ ive species wcre easily r~ el~A, with levels cc,.ll})~dble to wild type 15 e~ g cells Whether the glusagon il..l.. ~o~c~ay dis~ rS bctween fully proce~sed ~slucago" and the major unprocesse~1 form of glucagon glic~ntin, iS presently not known ConforTr~ ion that the mllt~d ~J.. ~l~o~l..- ~gO.I can only produce non-biologically active glicc~ and GLP- 1 will be ~eded EX~WPLE 11 No~el In vh~o Acth~ity From Tumors Expressing Amylin And Inwlin Methods Tumor l~u,ll~ion Of Transfected RlN Cell Lines In Nude Rats 6 to 8 week old athyIruc Fisher nude rats (200-300 grams Strain F3441Ncr-rnu from National Cancer ~ rl~ick MD.) were housed in a stcrile isolation facility with free ~cess to sterile standard laboratory chow and water. Blood glucose waL ~ d using an IBI Biolyzer (Koclak Fqctm~n C~h~-m~
Co.) or a ck - ~l ;p ~l,.ec,....,~er (Beta-~lucose Ph- to~r ~emocll~ Inc. Sweden). A pre-transplant body weight and blood gh-rose profile was ~"".;..r.rl 3 days before transplant with WO 97n6321 PCT/US97100761 daily tail vein bleeds bell,.eell 8:00 and 9:00 AM. Post-tr~nspl~nt body wcight was recorded and blood glucose profile was determined throughout the course of the c~ by daily tail vein bleeds b~ .ccn 8:00 and 9:00 AM. Pre-~ t serum insulin and amy}in levels were det~ .Pcl on the morning of bleed (0.5 rnl bleed). Post~ n~p~ l insulin and amylin levels S were ~l~,tc,....nf~l 7 days after transplant and again at various tirnes as blood glucose levels begin~
to ~lr~linf~

Amylin and Insulin Blood Serum Sample E~cpalalion - 500 blood s~mplec are f~"cd to an Eppendorf tube co~ ;ui~-g 5 ul of a 100x protease inhibitor cocl~t~il (lOOx~0 stock:, 320 mM EDTA,l m~ml ~ c~i~ ). Samples are stored on ice until clot form~tion, then fuged for S min. Serum is ll_~src.,~d to fresh tube and stored at -20 C until assayed.
Serum insulin levels were detç ~ r~ as described by the snMlirr of the insulin r~lioi~
assay (Coat-a-count, D ~;~ostic l'roducts Corp., Los Angeles). Senlm arnylin levels wcre A using a rat arnylin tadio-i.... -~,Os~c,qy as ~escrihe~l by the snrFli~or (Pc.-inc~
Laboratories. Inc., Cat. # RIK-7323).

RIN cell ~repatation and miection. 24 h before i...~ n-~jon, media was ch~ to serum ftee media to minimi7e tt~ncrl~nt~tion of foreign ~ntig~n.c in serum. Cells were washed with PBS-CMP, then trypsinize~ 5 volumes of serum free media was added and cells pelleted 20 by low speed cc~ irug.,. Remove media and wash cells 2 X PBS, (c~lrium and m~,.~i.....
free). Add freezing so (5ml!.I KH2PO4 25 mM KOH, 30 mM NaCl, 0.5 rnM MgCk, 20mM L lactic acid, 30 mM L-glul~mine, 5 mM glllcose, 167 mM myo-ino~ l, and 200 mM
sorbitol (pH 7.2-7.4; fi}ter sterilize) to pellet, adjusting cell co.~ tion to 4 million cellstml.
Ali~uot 4 million cells to a 1.5 ml. r},~"lb,r tube and compare packed cell volumes after 25 ~ uru~ g at low specd, as a check on cell nul,lbel, then rewcpen-i cells in 2~00 ul freezing solution. Nude rats are ~ i.. t;7Pd with i.p. injection of 0.1 ml N~ LIIO1 (50 mg/ml) per 100 g body weight. Swab dorsal area at kidney level -with 70% EtOH, then inject 20,000 cells per gm.
body wdght s~ u~o~,ly above animals right kidneys.

W O 97~6321 PCTAUS97/00761 Nude Rat dissection and anal~sis. Nude rats with palpable turnors were ~nestheti7P~l as before. A dorsal mi~C~gjt~l incision through slcin of back was made. the skin pealed back to reveal the tumor laterally at level of the kidney. Photogl~hs were taken and then the tumor was 5 ~elu~ ,d using forceps and iris scissors. The tumor was blotted on Whatman paper to remove excess liquid and then weighed. Tumors were d;~6~rte~1 in half with one half frozen in liquid nih~g_n and stored at -80 C while tlhe other half was fixed in Bouin's Fixative for histology (85 parts saturated picric acid solution, 10 parts 37% form~ldehyde, S parts glacial acetic acid).
Tumors were placed in tissue c~csetl~ and fixed in a beaker cont~ir..~g 3 X volume of Bouin's for 10 24 hours at room t~ l.pc~ e Next day, wash in cold running tap water for 24 h. Transfer tumor for storage to a beaker co~ ..i..g 3 X volume of 10% formalin.

In ~ tio!~ to isoiation of palpable tumors, ~anclc,aces wcre iCol~c~i from the ~nim~lc A
mi~lcsgi~l laperotomy was ~ ed to expose ~bdomin~l COI~tC~ ,e~ O~;~g the p~ c,as in 15 area of spleen. A portion of the ~al~c.~,zs in juxt~position to spleen was icolP~d, t~iCcec~1 in half, and one half frozen in liquid lullog~n and stored at -80 C while the other half fixed in Bouin's Fi~cative for histology.

Stre~tozolochl-ind~lced Dia~*tic Nude Rats. Diabetes was indl~ceci in nude rats by a one 20 time IV a l~lu~ ion of 65mg/l~g body weight sL~.ptozoLcjcnl.

Results:
Three RIN lines were inj~Lcd into rats, RIN lQ46-38, rat amylin ~ ;n~led EP97/134, and human amylin en~ e~ d EP182/7. These three cell lines pl~duccd identie~ u~t~ of 25 insulin but exhibited a 12 fold (EP97/134) and 36 fold (EP182/7) dir~ ,lce in amylin output (Table 9). RIN 1046-38 cells do secrete amylin and insulin at a physiologic ratio of .028, EP97/134 se~ us amylin and insulin at a ratio of 0.17, and EP18217 secrete hwnan (and low levds of rat arnylin) at an arnylin/ir~sulin ratio of 0.6.

W O 97t26321 PCTnUS97100761 Tr-n~pl~nt~tion into nude rats of parental, rat amylin and. to a lesser extent, human amylin en~ r- -~,d RIN lines shows, that there is an additive effect of RIN produced amylin with insulin in helping lower blood glucose. Daily blood glucose levels of five nude rats injected with 5 rat arnylin producing cells and five injecte~l with human amylin proAl~cing cells are co~ ,d to three nude rats injected with parcntal RIN 1046-38 cells and an umniec~d control. As shown in FIG. 22, the rat amylin pro~lucing cells intlllred hypoglycernia rapidly with an average onset time of 8.4 days +1.2 days. Human amylin producing cells had an average onset time of 14.8 +1.8 days. Parental RIN 1046-38 cells had an average onset time of 17.3 +3.4 days. The simplest 10 explanation is amylin is doing exactly what it has been described to do, slowing gastric emptying which results in lower glucose levels in the blood (Young, et al., 1995; Brown, e~ al., 1994). Rat amylin in rats is very effective at ~his while human amylin in rats has may have an ~ttenll~t~l effect. Anirnals with human amylin ~ h~g cells did in fact have fii-rrh~a so sc"..- Ih;l.g gastric was going on. With a co.,~ /c source of insulin present (eladGgenous insulin from 15 tumor cells), blood glucose falls as amylin incl~,a3es.

Conf;"..alion that amylin expression is mq;nt~in~d in vivo in thc en~ re.~d RIN cells comes from Northern analysis of RI~A isolated from ~umors at the end of the e~l)e.h,lellt. RNA
was icol~te~ from rat tumors arising from either parental or rat amylin producing RlN cells 20 following in vivo passage and was co~ ~ed to the same RIN cells ~ .rA in vitro. Northern analysis using a rat amylin spe~ific probe dc .~.nr.~l. ol~s no dirf~.~llce in either the low levels of ~rl~g~nQl~s amylin e~ ,ssion in either RIN 1û4~38 or BG97/134 cells or high levels of the rat amyiin tr~n~rnr in e~ r ed BG971134 cells (FIG. 24).

A second animal model, the strepto70tocin-ind-lced diabetic nude rats, was also used.
Diabctes was indl~ ed in rats two weeks prior to injection of either RIN 38, EP97/134, or EPl ~ lJ228 ~our highest insulin prollucing RIN cell with only low endGg~,l,o~.s amylin ~xl~ n). As shown in FIG. 25, all strep-injccted ~nim~l~ were d~ ti-~ with blood glucose's WO 97/26321 PCTrUS97/00761 of 400 mg/dl. Once again, EP 9J/134 cells in-1ure~ hypoglycemia more rapidly than RIN 38 cells. EPI l l/228, with a 15 fold e]evated insulin production col~lp~d to RIN 38 and EP97/134, ce~l hypoglycemia more rapidly than any other cell line but only twice as rapidly as the amylin overproducing line, BG97/] 34.
s To directly test the blood gll-cQse lowering effects of a RIN line that coyloduccs elevated levels of amylin and insulin, four ;..-~epe ~ t RIN lines were injccted into nude rats. These lines were parental RlN 1046-38 cells, EP18/3E1 cells ovcr.,Apl.,ssing human insulin, EP97/134 cells o~ s~ g rat amylin, and EP 181/59 cells ov~ ;ssing hurnan insulin and rat 10 arnylin. Secreted amylinlinsulin r~ltios as Illeasu,~d in vitro are .03, .002, .17 and .024 as shown in Tables 9 and 10. EP181/59 cells wcre chosen bec~ ~e of the high ol,_.eA~ ,ssion of both rat arnylin and human insulin, reS~ ng in an almost norrnal amylinlinsulin ratio. The absolute l~u~lu~l;on of rat amylin and hum.m amylin are both elevated five fold co~ ~.,d to the parental RlN 1046 38. These absolute ou~puts, on a per cell basis, is actually about half the output of either human insulin from EP18/31,1 cells or rat amylin from EP97/134 cells.

Cells were injected into nude rats, with four ~nim~lc used per group ii-je~ n group.
Three ~ llinje~,l. d rats were used as normal controls. Blood glucose was llloni~ol~id throughout the c~l-e~;...~nl The average blood ~ cQse values of each group are shown in nG. 26. The 20 normal rats Il~ f(l blood glu~ose levels of 110 to 120 mg/dl throughout the e~ P ~t All four inject~d animal groups ~ a drop in blood glucose levels as the tumors grew. As in ~he previous e~i."; .~ .t nude rats bearing the rat amylin p.u.l~ clone showed a quicker drop in blood glucose than the ~ ;t-~ et parental cells. The drop in blood glucose in the rat amylin Incrd~le;~-g clones was actually earlier than the human insulin Inu~ch~g clone, EP 18/3E1.
2~ Surprisingly, the amylin/insulin cocAp,~ss~ng cells, EP181159, were very effective at lowering blood gl-)cosç, even with lower a~solute outputs of either arnylin or insulin than EP1813E1 or EP
g71134.

W O 97/26321 P ~ rUS97~0761 Serum amylin and insulin levels are given in FlG. 27 and FIG. 28 with the molar ratio of amylin to insulin given in FIG. 29. It should be noted that insulin values were f~ ...i..rd using a human insulin stanclards. This undere~ s the rat insulin levels by about three fold as seen in the nonnal rat control serum values of between 1.5 and 2 ng1ml rather than 4 to 6 nglrnl. This S also skews the amylinlinsulin ratios l~polled in FIG. 29, such that nonnal ratios are ~l~.~,en .08 and .I rather than the expeC~1 ratio of .02 to .03. As e,xpected, serum amylin levels rose dr~m~ti~lly in the rats bearing the rat amylin OV~ CSSillg cells.

Serum amylin rose mod~,lat~ly in the insulinlamylin coexpressing lines, p~,.l.a~s 3 to 4 10 fold above normal physiologic levels. Rats bearing the insulin ~ nhl~ or un~ng...~e ed parental cells oniy had a Illod~,.ate incl.,ase in serurn amylin. Serum insulin values rose fairly high in the insulin ovelprodllcin~ line and mnder~çly high in the insulin/amylin ploducillg line.
Rats bearing the amylin o~ ,o~ g cells or un~n~ r~r_d parental cells had mn~ler~tr il~w~ases in serum insulin. As shown in FIG. 29, the ratios of amy~in to insulin throughout the 5 ~;Al~f ~ ;...r.nt in the various anirnal groups is infcllllative. Animals bearing the insulin ovelprodncing clones result in lower amylin/insulin ratios than normal ~nimqlc Animals bearing the amylin o~ od~ g clones result in higher amylin/insulin ratios than normal ~nim~lc, Animals bearing the un~n~inl e~ed parental cells or the amylinlinsulin coen~i~.ee.ed cells, both having in vitro amylin ratios in the normal range, mqint,qin a near normal ratio throughout the 20 course of the ¢s~ t This result dc--~ es aL bçn~-firi-ql effect of lowering blood glucose by Co~n~ F~ -;ng amylin and insulin production in RIN cells. Lower blood gl~ Qce is achieved with co~ ely lower outputs of either amylin or insulin o~ luced alone. Also, serum levels of both amylin 25 and in ulin needed for this effect are only m-l~3ept~ly incl~,a ed above nolmal ~nimqlc by about four fold while ~ a near normal ratio of tne two.

W O 97126321 PCT~US97/00761 The inventors studies show that amylin related species act additively with insulin in lowering blood glllcose A cell O~ y~ Sil~g insulin and amylin will lower blood glucose more effectively than a cell overexpressing either one by itself. RIN clones coeng.n~el~,d for amylin and insulin over-production have alrc~dy been isolated and ch&~ ntu.;7ed The inventors have further ~lr~ n~ ted that the rat amylin prod~lcin~ RIN cell line used lowers blood glucose quicker than human arnylin producing RIN cell line or parental RIN cell.
This effect could be clone specific, or it rl~monctrates an advantage that rat amylin confers the ability of a given tumor to deliver ho....o1~rs to the serum. A very possible advantage is incçcascd v~cc~ ri7~tion stim~ ted by rat amylin in rats. E~yc~ el~ls desi~d to directly check this are being pl~nn~d Llw~~~e~ v~cclll~ri7~~ion in the context of devices is an obvious benefit.

Increased Expression of either Human Amylin or Mrt~tp~l Progl~--a~ ~n by Generation of Fusion Proteills with Human Growth Hormone Methods:
Fusion protein expression l~la~mi~ls. pCB6/hGH was digeste~ with TthIIIi, treated with 20 Klenow, then ~li~ste,cl with Spel ger.c~d~ g a 2060 base pair fragment cQl~Ai-.ig the majority of the human growth h~.~..onr 5' coding se.lu~,uce. pNoTAT7/h.Arnylin was Aigested with BsiHKA1, trcatcd with Klcnow, thcn ~igest~ with Xbal g~ g a 418 base pair fragment col~t~;ning the human ~ coding sc~l,, nfe. These two r~AE~ f~t~ were ligated into pCB6 that had becn tiigestrd with Spel and Xbal. The res~ltin~ pl~cmi~l~ pCB6/hGH.hAMYfusion cr~ the insert SEQ ~ NO:70 where the first 1466 base pairs are derived from the human growth ho....o~.~ gene ~bases 498 to 1964 of pl~liched se~lJen~e. See~, 1982) fused to 418 base pairs derived from the human amylin cDNA. The res~llting fusion protein, S~Q ID NO:7 1, W O 97/26321 PCT~US97/00761 COI'lSiSl~ of the first 198 amino acids (out of 217 amino acids) of hurnan growth h-.. I~C fused in frame to the final 76 amtno acids (amino acids 14 through 89) of human amylin.

pNoTAT7/rnllt~ ragon w;ls ~i~sted with PvuII and Xbal g.,.~,.a~illg a 891 base pair S fragment cont~ining the mllt-qJed proglucagon coding sequence. This fragment was ligated with the above mentioned TthIIIi, Klenow treated, Spel L~ lcnt of pCB6/hGH into pCB6 that had been ~li~s~ecl with Spel and Xbal. The res~llting plqcmi-l, pCB61ht'JH mllt~3lllr~gollrusion contains the insert SEQ ID NO:72, where the first 1466 base pairs are dcrived from the human growth hormone gene (bases 498 to 1964 of pllhli~hed sequence, Secbu~, 1982) fused to 891 base pairs dcrived from the ml)t~ted progll~cqgorl cDNA. The res~llting fusion protein, SEQ ID
NO:73, cc"~ of the first 198 arr~tno acids (out of 217 amtno acids) of human growth hc,~ ol.e fused in frame to the final 162 ami:lo acids (arnino acids 19 to 180) of ml~-qted progl~lcq~n.

Results:
pCB6/hGH.hAMYfusion and pcB6JhG~lm~lt~lucagonfusion were elecLro~olaLed into RIN1046-38 cells as well as RIN1046~4 cells. The latter cell line was generated from the same original rat inmlinomq as the RIN 1046-3~ line, but e~p,esses very low levels of endogenous insulin (Philippe, et al., 1987). Stable clones of each were selecte~l using G418.

The ~ I.e~l -ion is that both fusion proteins will be em~iently targeted to the regulated SC~l~,tu~y ~ way, as has been ~sc~ e~l for similar growth hormone fusion co~ C~ (Moore and Kelly, 1985). Once ~ d to the s~.~t~"r granules, post-translational p,occ~s:~g events .g proteolytic clca~tâge at dibasic amino acid residues and amidation should occur. The 1000 fold dif~,..,,lce in e~ ,ssion of wild type growth he....e.~ (app. 100 ugJmillion cells/24 25 hours) versus either amylin or GLP-l (âpp. lOOnglmillion cellsl24 hours) appears inherent to tne growth hmmo~ s~ Pnre.s Genomic Site-D;. ~ Mutagenesis with Oligonucleotides The inventors have previously d~monctrated that derivative cell lines of the RIN 1046-38 cell line are capable of ~rOlll~illg homologous lccolllbil,ation by disrupting an allele of the S hexokinase I gene. Feasibility studies are ~;u~ tly underway to detçrmin~ if RDOs or DNA
olig~nllrlcoti~s can be used for the ~u~ose of targeted gene disruption in RIN and other cell lines. Two test ~y~te~ls have been decigr~ed for testing oligonl)rleoti~les: the disruption of the ncol.l~in phosphotransferase tr~n~ne, and the disruption of the glucose tld~lspoller, type 2 (GLUT-2~. As a prelimin~rv eXpe~ fnt to testing RDOs or DNA oligonucleotides, protocols for 10 ef~lcierlt delivery of DNA into RIN cell lines by ele~ o~ation have been optimized A. Optin~i~ation of tr~fection of RIN cell lines.

A llulllber of transfection protocols were tested on R~ 1046-38 cell lines inrl~ ing a 15 variety of ele-:t~ oldtion c~ndi~ior ~ and multiple kinds of liposGme-m~~ t~d tr~r.s~clion. All protoco~, except one set of clc~ Joldtion conditions, failed to produce tl~ls~,~ion efficie-nci~s of greater than 5%. Protocols werc optimi7~d for delivery of exogenous DNA to RIN cells by cle~llo~ ~on using two types of DNA: a pl~cmirl vector cnCo~1ing beta-g~l~~tosiADce (~gal) that is ~r~ ;be~l from the CMV promoter, and a DNA oligonucleotide (62mer) that had been 20 r~dil~lA'~e~ with 32P-dCTP. One set of ele~tlo~oldtion condidons resulted in 25 - 40%
cr~l;on of the total cell population as cl~t....~ P~l by colu~ lic, c~lorhf ~ l assays for ~gal activity ( Bassel-Duby et al., 1992). Cells were grown to about 80% cc!~-nu. ~e in MeJ;~ 19915% fetal calf serum~ 11 mM glucose (Growth Mc.li~llll) and were re-fed with fresh Growth McL~l one day prior to elec~ ., Cel}s were hal~csled by trypsini7~ion, 25 c.~ e~ lete~ by centrifugation at 1000 rpm for 5 millu~,t~, and ~;,-,s~ cl in Growth Mc~liulll at a dcnsity of 2 x 10~ Ct'lls/IIll. 0.5 ml of cell sus~encion was mixed with 60 ~11 of the following DNAs: either 10 ~g of ~ga} plasmid or 40 nM of oligon~lrleoti~e and 110 ~lg of sQ~ (l salmon spe~n DNA. The cells plus DNA were mixed gently, transferred to a 0.4 mM

WO 97~6321 PCT~US97/00761 cuvette, and ele~iL.o~o.~ted at 600 ~IP, 250 volts using and Electro Cell l~r.ir~ tor 600, BTX
Elcct opu~dtion System. The electlo~ ated cells were removed from the cuvette and diluted into 25 - 30 mls of 37~C Growth Medium cont~inin~ 5 mM butyrate. Following inrub~tion for 12-16 hour at 37~C, 5% CO2 in the growth medium with butyrate, cells were washed once with growth 5 ",.~.1;,..,. and m~int~in~ in growtih ~r.~ .in. In the case of cells transfected with ~gal, cells were ...~ ;..cd 48-72 hours following L~ r~l;on and fixed with 0.59c glutaraldehyde for 10-15 min~lte~ for cylo~l.,..ical detection of ~gal using the 5-bromo-4-chloro-3-indoyl-~D-galaclo~ oside (x-gal) as a substrate (Bassel-Duby et al.~ 1992).

To ~etrrmin~ if con-litio~ o~t;~--;7~d for plasmid DNA would translate to efficient uptake of oligonucleotides, the above cle.,tro~Grdtion protocol was applied to a 62mer that have been radiolabeled with 32p using thc Redi-prime Random Primer labe}ing Kit (A~ hdln Life Scienr~c). Oligonucleotide (40 nM) was elccllo~olated into cells. Cells were analyzed post-tr~n~fectior- at 0, 3, 6, and 24 hours in two ways. First, total r~io~cl.~ily in the media, 15 c~topl~ll~ic cellular fractions, and nuclear cellular fractions was der ~ d by scintillation collntin~ And second nucleic acids were harvested from cellular fractions by phenollchloroforrn/isoamyl extrat tion and fractionated through denaturing pol~ ylal,ude (PAG) gels (Ebbingh~lls et a~., 1996).

2~ There was a marked enh~n~ n~ in nuclear r~dio~ti~ily in the p.~,~ence of ele~ uyolalion as co,~ ,d to control cells that were mi~ed with oli~onllr,leoti~ but not cle~hû~dted~ In the p~ ence of cleclro~ld~ on, about 29, 55, and 66% of total intr~rçllul~r counts se~ ;aled to the nuclear fraction at 0, 3. 6, and 24 hours, r~s~ y. In CQ~ ct only 1 - 24~b of total intr~cl~ r radioaclivily was ~tçc~d in the nuclear fraction through the 24 2~ hour time point. It was also observed that intact, app~il.tly full-length oligonuclcotide could be c~ d from cells which had been cle~hu~ t~d~ as evidenced by fraction~tion on de~ ~tn ;.,g PAG gels and autoradiography. FYttPr~C from cells that had been mixed with the oligon,~ eo~ ;~3e but not ele~;hopul~ted did not yield de-t~rt?l l~ oligon~ eoti~e by this method of analysis W O97/26321 PCT~US97/00761 s~ege~tin~ that radioactivity that was ~et~cted in the non--electroporated cellular fractions was derived from the P~ch~nge of radioiabel~ not from the oligonucleotide.

From these studies it has been concllulPd that the cle~l-oporation protocol described 5 above is a pl~f~llcd method for tr;msfecting both plasmid DNA and oligonucleotides into RIN
cells.

B. Disruption of the neomycin phosphotransferase (NPT) transgene by RDOs.

Multiple RIN cell lines are available that have been en~inPçred to contain an integrated copy of the NPT gene. An RDO for di~lu~ L~on of tr~n~g~nic NPT has been ~1~PC;~P.~I that is c~...pl~ ".l~.~ to ~ leo~ s to 54 to 7~ of NPT countin~ the "A" of the first m~thiorine as 1.
Further, the RDO co~ c a single base change relative to the wild-type NPT (A to C at po~ilio 66). If gene conversion by the RDO is succec~r..l, a T will converted to a G, Tyr22 will be 15 co~ ,.~d to a stop codon, reSiC~n~e to G418 will be lost, and a unique Mae I 1~ o~ site will be in~ ed The RDO also col~ c rc~es previously des~.ibed such as self-~nnP~ling hairpin loops at each end, and 2'-C)-methylation of the ribose sugars. Thc se4u~nce of the RDO
with these fcalu.~;s is (5' to 3' and ~,f~ cd hclealt~.r as AT142):
GCTATTCGGCTAGGACTGGC;( ACAATrrTuu~ c~ ~TCCTAgccgaauagcGCGC~ l~l~l~
20 CGCGC (SEQ ID NO:74), where large caps l~p.ese,lt DNA residues and small, bold letters indica~c RNA re~idlles R~ cell lines with a single integrated copy of NPT will be elecllopol~ted~ as ~scribed in materials and ~ 'th~lC, with varying co~ rdtions of RDO AT142. 4 to 6 hours following 25 tr~ncf~ction ~e~mie DNA from pools of ~ r~ ~ lz~ will be analyzed for detection of a T to G
co~aion at position 66 of the NPT tr~ncg~nP Following isolation of gen~ mie DNA, the first about 200 base pairs of the NPT tr~C~n~ will be ~mplified by the pOlyLu~ âc chain reaction ~PCR) using oligomlrleoti~1~c that flank position 66. Following ~mptifiration, PCR ~ du~,ls WO 97n6321 PCT/US97/00761 will be ~ est~(l with Mae l to ~Ct~ r if any gene conversions have occurred. If the case of s~ ce.~rlll gene inactivation by the RDO, the PCR product will be ~1iges~l into two bands. The wild-type NPI transgene PCR product will be resistant to Mae I digestion. If NPT gene dislu~n is detect~hle by PCRlMae I ~ligestion~ small pools of clones will be analyzed for loss 5 of lc~;c~ ce to G418. Following ele~llopoIdlion, cel}s will be plated into 96 well plates at r~el~s.;l;Ps of 3 to 5 cells/well 3 days following ele~llo~uliltion, cells will be exposed to G418, and each well will be scored for the presence of cell death.

C. Disruption of transgenic GLUT-2 in IUN and 293 cell line~s.
RIN cell lines and 293 cell lines have becn eng.ne.,.cd to express high levels of a t~ sE,~ r GLUT-2 ~ pu~ as detailed herein above. The ~.~sence of this transporter confers se.~sili~,ily to the cytotoxin ~ ozoto~ (STZ~, and thercby provides a means of ncg~i~_ sPI~C!;on (SrhnÇ~l et ~1., 1994). Both RIN and 293 cell lines that express high levels of a GLUT-15 2 tra~ls~r will be transfected with RDOs ~Ci~--d to target and disrupt tr~nS~enic GLIIT-2, and 4 - 6 hours later cells will be e~l-4sccl to cytotoxic levels of STZ. Surviving clones will be al~sly~,cd for the p.~se.,cc of an inactivated GLUT-2 ~ sE,- .r by analysis of g~ no~iC DNA. In the case of the targeted inactivatiorl of tr~nC~nir GLUT-2, leucine at position lO will convertcd tO a stûp codon as a result of a 1' to A conversion, and a unique Avr II restriction site will be 20 created in the tr~n~genir GLUT-2. This unique site can be ~tcrtl d by the ~mrlifir~tion of g~nc~ DNA that flanlcs the site by PCR, followed by ~ligestiQrl of the ~mrlifi~-~ DNA with Avr ~. One such RDO that pot~nti~lly ;~eeo~ licl~f~s the targeted disruption as des.,libed above is the following se4u~,l,ce:
TCACCGGAACCTAGG~ A~ ~laa gugaaagCCTAGguuccgguugaGCGCGTlT
25 TCGCGC (SEQ ID NO:75), where large c~rit~lC ~ sellt DNA residues and small bold letters t RNA reCi~ es CA 0224643l l998-08-l2 Attempts to disrupt tr~ncgPnis GLUT-2 will also be made with non-çhim~ric DNA
oligo~lcleotides that contain phosph~ rothioate m~ifierl b~cLbol f-s to eF~b~nee stability. It has been l~olled that inclusion of phosphorothioate derivatives within the DNA backbone decreases sensitivity to nllelP~ces (Vosberg and Eckstein, 1982; Monia et al., 1996). Oligont-eleotides have 5 been desi~n~d that should selectively target the transgenic GLUT-2 by sy~ g an area of homology that is illt~ll~ted in the endogenous GLUT-2 gene by an intron. If targeting and m~ e~tio~l of the GLllT-2 transgene are cucc~csful, glllt~mine at position 35 will be converted to a stop codon, and a new AJI II site will be introduced into the DNA at this position. DNA
oli~n~rleotides will be ex lminP~ for the ability to target and disrupt the Ir ~ ~g. ~-ic GLUT-2:
oligo name: AT157 (5'to3') GsGTTCCTTCCA~ ~ ;~GATATGACATCGGTGTGATCAATGCACCTTAAGAGGTAAT
AATATCCCATTATCGACA l ~ ~ G(~ ~ l sC (SFQ ID NO:76), oligo name: AT158 (5' to 3') GsAGGAACACCCAAAACATGTCGATAATGGGATATTATTACCI'CTTAAGGTGCATTG
15 ATCACACCGATGTCATATCCGAACTGGAAGGAACsC (SEQ ID NO:77), oligo name: AT159 (5' to 3') GsGATATGACATCGGTGTGATCAATGCACCTTAAGAGGTAATAATATCCCATTATCG
ACATsG (SEQ ID NO:78), and oligo name: AT160 20 CsATGTCGATAATGGGATATTATTACCTCTTAAGGTGCATTGATCACACCGATGTCAT
ATCsC (SEQ ID NO:79).
Each of above the 4 oli~Q~Ilc}eotides have phos~horulhioate morlifir~tion~ in the ba~kbone near the 3' and 5' ends as ir.~ t.~d by "s" in the seq~nre Oligon~cl~tif~es will be introduced into cells both as single-stranded molPc~les and as double-~ ded co..~l-k~cc The following oligor~rletirlP pairs contain compl~ r seq.,~.es and will foml ~urle~Pc: AT157-AT158.
AT157-AT160, AT158-AT159, and AT159-AT160. Cell lines that cxprcss high levels of transgenic GLUT-2 will be eleclro~ol~. ted with oligon~eleotides as des.- ibed above, and e~l,ose1 to levels of STZ that are lethal tO cells t~r~,SSillg non-dismpted tr~ genir Gl,11T-2. Ger~omic CA 0224643l l998-08-l2 W O 97/26321 PCTrUS97/00~61 DNA of surviving cells will be analyzed for the presence of disrupted transgenic GLUT-2 by ~mplifiration of DNA cont~ining thc putative mutation by PCR, followed by digestion with A.fl II.

S* * *
All of the co~ osi~ions andlor mPtho-l~ disclosed and cl~imPd herein can be made and executçd without undue ex~~ ion in light of the present disclosure. While the co...p~silions and methods of this invention have been described in terms of pl~ ,d tmbo~ tc, it will be ~yalc~lt to those of skill in the art that variations may be applied to the 10 co,ll~osilions andlor methods and in the steps or in the sequPnre of steps of the m~thorl described herein without de~lil g from the C'OllCe~l, spirit and scope of the invention. More specific~lly, it will be apparent that certain agents which are both chr~ lly and physiologically related may be ~.II,s~ t~ d for the agents dcsclibf d hcrein while the same or sirnilar results would be acl~,e~cd.
All such sirnilar ~-~h~ vlfs and mo~lifir~tions apparent to those skilled in the art are ~leçn~çd to 15 be within the spirit, scope and con~e~Jt of the invention as defined by the ay~nded claims.

w o 97n6321 PCTAUS97/00761 FERENCES
V. RE
The following references, to the extent that they provide exe-lnrl~y procedural or other details supplell,c.lt~y to those set forth herein, are specific~lly incw~.at~d herein by rtf.,.~nce:

5 Adams et al., "Porin interac~ion with hexokinase and glycerol kinase: Metabolic microc~ t~lion at the outer mi~ochon~lrial lllcl~ e," Bioch. Med. Met. Biol., 45:271-291, 1991.

Aguilar-Bryan et al., "Cloning of the Beta cell high-affinity sulfonylurea receptor: A regulator of insulin secretion," Science, 268:423~26. 1995.

Altman et al., Diabetes 35:625-633t 1986.

f~n.1~ ~oll etal., "Cloning, structure, and e3~ssiol. of the mitochor~lri~l cytoch~vl"c P~S0 sterol 26-h-y~l~u~ylase. a bile acid bio~ tl,~Lic enzyme," J.B.C.. 264:8222-8229, 1989.

Arimura et al., "Regulation of growth honnone secretion." The Pituitary Gland, 2nd F~lition, Chap.9:217-259. H. Imurat Ed., Raven Press, N.Y., 1994.

Arora, et al. J.B.C. 268: 18259-18266, 1993.

Asfari et Q~ Establichm~nt of 2-me~captoethanol-d~ ! n~lF n~ ~lifrC-c ~ rd insulin-secreting cell lines." Endocrinolog~,, 130: 167-178, 1992.

B~h~ e~ al. "Anima~ cell, viral antigens and vaccines," Abs. Pap. ACS, 180:5. 1980.

W O 97~6321 PCT~US97/00761 Baichwal et al., "Vectors for gene ~ fer derived from animal DNA viruses: Tlal~icl1t and stable expression of ~ r~ ,d genes," ~n~ u~h~ ti R. ed. &ene ~ Çtr. New Yorlc: Plenum Press, pp. 11'7-148, 1986.

Baijal, et al., J.E. Arch. Bioch. Biophys. 298:271-278, 1992.

Bailey et al., t'Methylmercury ~c a reversible denaturing agent for agarose gel ele~,lrol~hoIcsis,"
Anal. Biochem., 70:75-85, 1976.

Barr et al., "Systemic delivery of l~col~bi~ l proteins by genP,tic~l~y m~ rlifi~l myoblasts,"
Science,254:1507-lSOg, 1991.

Bassel-Duby, R., Hern~n~l~o7. M.D., Gonzalez, M.A., Krueger, J.K., Williams, R.S., l9g2. A 40 kilodalton protein binds sperifi~ y to an ~ sc~u~-~re elem~n~ esse~ti~l for muscle-specific t~ sc-i~lionofthe humanmyglobin ~ l.Ot. r. Mol. CelL Biol. 12: 5024-5033.

Ba~cl.wit~, R. and Holloman, W.K;. 1990. ~.cQ~ n of the REC2 gene controlling l.,colllbil,ation in Ustilago maydis. Gene 96:285-288.
Beaumont, K., Kennery, M.A~, Yolmg, A.A. and Rink, T.J., "High affinity and amylin binding sites in rat brain," Molecular Phur,. ~c~logy 44:493-497, 1993.

Becker et al., ~Dirrclcll~ial effects of ovc,~,At~.ess~,d gluco~in~ce and hexokinase I in isolated islets: Evidence for functional segl~gaLion of the high and low K " c,r~ es," J. Biol.
Chem., 271:390-394, 1996.

CA 0224643l l998-08-l2 W O 97f26321 PCT~US97/00761 Becker et al., "Use of recombinant adenovirus for metabolic e~gin~ering of m~rnm~ n cells,"
Metkods in Cell Biology, 43:161-189, Roth, M., Ed. New York, AcaA~mic Press, 1994.

Becker et al., "Overexpression of h~xo'-in~ce 1 in i~ol~tPd islets of langerhans via recom~inant adcnovirus," J.B.C., 269:21234-21238, 1994.

Bell, e~ al., "Charactel.2a~ion of the 56-kDa subunit of yeast trehalose-6-phosphate synthase and cloning of its gene reveal its identity with the product of CIFl, a regulator of carbon catabolite inactivation," Eur. J. Biochem., 209:951-959, 1992.
Bell et al., "Nucleotide sc~ n~e of a cDNA clone enroAing human ~ u;~ n-~ Na~ure 282:525-527, 1979.

Rçnj~nn~t et al., ~CO111YaldtiVe bio~.y.~l .es; c, covalent post-tr~nciq~inn~l m~lifi~tions and efficie~cy of ~loseg~ cleaYage of the l~roh.,llllonc conv.,.~.es PCl and PC2:
Glycosylation, sl-lrh~tion and i~ ntifir~tjorl of the intracellular site of prosegrnent cleavage of PC1 and PC2," Biochem. J., 294:735-743, 1993.

Benvenisty and Neshif, "Direction introduction of genes into rats and expression of the genes,"
Proc. Nat'~Acad. Sci. USA, 83:9551-9555, 1986.

Berzal-~erranz, A. et al., Gcnes and De-~el., 6: 129-134, 1992.

Bhogal, R., Smith, D.E., and Bloom, S.R., "Investigation and Cha~ lion of binding sites for islet arnyloid polypeptide in rat ~ lb,~nes, EnclrGci~nology, 130,2:906-913, 1992.

Blasquez et al., "Trehalose-6-pho~plate. a new regulator of yeast glycolysis that inhibits hexokinases," FEBS Le~t. 329:51-54, 1993.

W 0 ~7t26321 PCT~US97/00761 Brewer, "Cytomegalovirus plasmid vectors for pc~ t lines of polarized epith~ l cells," In:
Methods in Cell Biologv, 43:233-245, Roth, M., Ed. New York, AcadeIr~ic Press, l9g2.

S Brewer, "Cytomegalovirus pl~crnid vectors for pe ..~ lines of polarized epith~ cells," In:
~etho~lc in Cell Biolo~OEY. 43:233-245, Roth, M., Ed. New york, Ac~-mic Press, 1994.

Brunstedt et al., "Direct effect of glucose on the ~el"~?insulin mRNA levcl in isolated pancreatic islets," B.B.R.C.. 106: 1383-] 389, 1982.
Burch, et al., "A~pt~ion of glycolytic enzymes. Glucose use and insulin release during fasting and rçfee-ling." Diabe~es 30:923-928.

Burgess et al., ''C~l~s!~ e and regulated secretion of proteins," Ann. ~ev. Cell Biology, 3:243-293, 1987.

Cech et al., "In vitro splicing of the ribosornal E~NA pl-,~,u,sor of Tetrahymena: invol~ "ent of a gll~nosinç n~cleQti~ . in the excision of the intervening sequence," Cell. 27:487-496, 19~1.
~h~llit~ et al., "~ -~rk of expression from a l rtl.VVil~ll vector after tr~ncducr1~n of murine h~ topoietic stem cells is ~csoc~ with methylation in v~vo," Proc. Natl. Acad Sci.
USA~ 91:2567-2571, 1994.

25 Chang et al., "Poreign gene delivery and e~p~S~iOn in hepatocytes using a hP.p,~ ;c B virus vector," Hepatology, 14: 134A, 1991.

W O 97126321 PCTrUS97/00761 Chavez et al., "Expression of exogenous proteins in cells with regulated secretoIy pathways," In:
Methods in Cel~ Biology, 43:263-288, Roth, M., Ed. New York, AC~ mir Press, 1994.

Chen and Okayama, "High-efficiency transfection of m~mm~ n cells by plasmid DNA," Mol.
5Ce~l Biol., 7:2745-2752, 1987.

Chen et al., "Regulated Secretion of Prolactm by the mouse cell line Beta TC-3," Biotechnology, 13:1191-1197, 1995.

10Chen et al., Proc. Nat'l Acad. Sci. USA 93: 14795 (1996).

Chowlira, B.H. et al., Biochemistry, 32, 1088-1095.

Chowr~ra, B.H. et al., "J~ vitro and in vivo co.,.~ on of h~.. 1.~ 7~ hairpin, and hep~titic 15delta virus self-p~cesc;.~ ribo~y"le c~ r s." J.B.C., 269:25856-25864, 1994.

Clark et al., "Mo~ tiQr~ of g}ucose-in~llce~l insulin secretion from a rat clonal beta-cell line,"
Endocrinology, 127:2779-2788, 1990.

20Clark et al., "Islet cell culture in defined selum-free ~ , " Endocrinology, 126: 1895- 1903, 1990.

Coffin. "Retroviridae and thelr replir~-iQn," In: Virology, Pields et ~l. (eds.), New York: Raven Press, pp. 1437-1500, 1990.
Cole-Strauss, A., Yoon, K., Wiang, Y. Byrne, B.C., Rice. M.C. Gryn, J., Holloman, W.K., and Kmiec, E.B. 1996. Co,.~ C~iOI of the mutation l~s~onsi~lc for sickle ccll anemia by and RNA-DNA oli~Qn~ leotide. Science 273: 1386-1389.

Cone et al., Proc. Nat'l Acad. Sci. I~.S A. 81 :6349-6353, 1984 Cordell et al., "Isolation and charaL~ ation of a cloned rat insulin gene." Cell, 18:533-543, 1979.

Couch et al., "Tmmllni7~tion with types 4 and 7 adenovirus by selective infection of the intestin~
tract," Am. Re~. Resp. Dis., 88:~94403, 1963.
~0 Coupar et al., "A general method for the con~l~u~tiol~ of recombinant vaccinia virus e~ ssing mllltirle foreign genes." Gene, 68:1-10. 1988.

Curry, "~nsulin content and insulinogenesis by the perfused rat pancreas: Effects of long term glucose stimlll~tiQn," Endoorinology, 118;170-175, 1986.
~lttitt~ "Peptide amidation: Signature of bioactivity," The Anatomical Record, 236:87-93, 1993.

D'Santos, C.S., Gatti, A., Poyner. D.R., and Hanley, M.R., ''Stim~ tion of Adenylate Cyclase by Amylin in CHO-Kl Cells," Mo~ecular Pharmacology 41 :894-899, 1992.
Dai et al., "Ce~ r and humoral i .. ~ .c ~ C~S to adenoviral vectors corlt~inin~ factor IX
gene: TG1e. ;~ n of factor IX and vector ~ntigerlc allows for long-term e~ ,ssion,"
P.N~.S. USA, 92:1401-14()5. 1995.

Danos e~ al., Proc. Nar 'l Aca~. Sci. U.S.A. 85:6460-6464, 1988.

Day ct a~., "Distribution and le~ on of the p~O~O~ .c convertases PCl and PC2 in the rat uitdl~" MolecularEndocrinology. 6:485497,1992.

W O 97/26321 PCTrUS97/00761 Dhawan et al., "Systemic delivery of human gro~,vth hG~ one by injection of gen~ y en~;l.ecred myoblasts,~' Science, 254: 1509- 1512, 1991.

Dickerson et al., "Tr~ncfected human Neu,o~c~Lide Y cDNA expression in mouse pituitary cells," J.B.C., 262:1364~13653t 1987.

Drucker et al., "Cell-specific post-translational procescing of preproglucagon expressed from a metallo~hio~in-glucagon fwion gene, ' J.B.C., 261:9637-9643, 1986.
DulJcr~y el al., "Direct tr~ncfecrion of viral and Fl~Cmi~l DNA into the liver or spleen of mice,"
Proc. Nat. Acad. Sci. USA, 81:7529-7533, 1984.

Ebbingh~c, S.W., Vigneswaran, N., Miller, C.R., Chee-Awai, R.A., Mayfield, C.A., Curiel, D.T., and Miller, D.M. 1996. F.ffi~ .n~ delivery of triplex forming oligo~ leo~ides to tumor cells by adenovirus-polylysine complexes. Gene 7herapy 3: 287-297.

Efrat, et a~., "Ribozyme-m~ t~ ~tt~n-l~tion of pancreatic Beta-cell glncolcin~ce expression in g~ l~ic mice results in impaired glucose-in~luce.l insulin secretion," Proc. Natl. Acad.
Sci. USA, 91:2051-2055, 1994.

Efrat et al., "Murine inc~llinoma cell line with normal glurose-regulated insulin secretion,"
Diabetes, 42:901-907, 1993.

~5 Efrat et al., "Beta-cell lines derived from transgenic mice e~L~ressing a hy~rid insulin gene-Ol~G~ f-," P.N.A.S.~ 85:9037-9041, 1988.

W O 97/26321 PCTrUSg7/00761 Efrat et al., "Conditional transfroln~tion of a ~ cl~,atic beta cell line derived from transgenic mice e~ ssi~lg a ~acyclinc-regulated oncogene," Proc. Natl. Acad. Sci., USA, 92:3576-3580, 199~.

5 Eipper et al., "Al~ GSp~ ing and endoproteolytic processing generate tissue-specific forms of ~:)iLuil~uy peptidylglycine alpha-~miA~'ing monooxygenase (PAM)," J.B.C., 267:4008-4015, 1992b.

Eipper et al., "The biosyntbesis of n~(~lides: Peptide alpha-~mi~l~tion," Annu. Rev.
Neuroscience, 15:57-85, 1992a.

Eizirik, D.L., Korbutt, G.S., H~ , C., (1992), "Prolonged e~osu~, of human ~ ic islets to high glucose con~Pntrations in vitro impairs the beta-cell r~ ;on ", J. Clin.
Invest., 90(4): 1263- 1268.
Frstein, et al., "Expression of yeast h~Y-.~insce in ~&~Cl~a~C beta-cells of ll~nc~ nuce reduces blood glucose, ~nh~nres insulin secretion, and dec,._dses ~ betes~" Proc. Natl.
Acad. Sci. USA. 89: 12038-12Q42, 1992.
~0 Ercolani et al., "Isolation and ~D~ ,'tt- se~ n~e of a filnrtion~l human glyceraldehyde-3-~h-ss~.h~ delly-l-u~ ,ase gene," J. B~ol. Chem., 263;15335-15341, 1988.

Eshkol, "M~mm~ n cell-derived recomhin~nt human growth ho~ e: Pl.~ cology, metabolism and clinical results," Hormonc Research, 37:1-3, 1992.
Fanciulli et al., "Glycolysis and growth rate in normal and in hPxo~in~ce-tr~n~fect~l NIH-3T3 cells," Oncology Res., 6(9): 405409, 1994.

reehh ~ et al., "Transfection of m~Tnm~ n cells with plasmid DNA by scrape loading and sonir~qtion loading," Proc. Nat'l Acad. Sci. USA, 84:8463-8467, 1987.

Felgner,etal.,Arch.Bioch.Biophys. 182:282-94, 1977.
s Ferber, et al., "GLUT-2 Gene Transfer into Tnclllinnm~ Cells Confers Both Low and High Affinity Glucose-stimlll~t~d insulin Release", Tke Journal of Biological Chemistry, 269:12523-12529, 1994.

Ferkol et al., FASEB J., 7: 1081-1091, 1993.

ri~Ol~ et al., "Scle~ u expression of the insulin I gene in rat in~lllinom~-derived cell lines,"
Mol. En~ocrin., 4:990-999, 1990.

Fiek et al., "Evidence for identity between hrxrlinq~e-binding protein and the mitochon~
porin in the outer ~ e of rat liver mi~chnn~1Ti~ " Biochem. Biophys. Acta, 688:429-440, 1982.

Fincgood, D.T., Scaglia, L., Bonner-Weir, S., Elliot, P.,"A simple m~th~ thir~l model,"
Diabetes, 44(3):249-256, 199S.

Forstcr & Symons, "Self-cleavage of plus and minus RNAs of a virusoid and a SLIu~;lu~al model fortheactivesites," Cell, 49:211-220, 1987.
~5 Fort et al., "Various rat adult tissues express only one major species from the glyceraldehyde-3-~hnspk~e-dehydrogenase m~ eni~ family," NucleicAcids~2P,~ ich, 13:1431-1442, lg85.

W O 97/26321 PCTrUS97/00761 Fraley et al., "Ern~aymel~t of a b~ct ~ l pl~cmi~l in phospholipid vesicles: Poten~ial for gene tlallarer~" Proc. Nat'l Acad. Sci. USA, 76:3348-3352, 1979.

Fricker, "C~l,o~y~ Lidase E." Ann. Rev. Physiology, 50:309-321, 1988.

~;P.i.~Vm, "~ ,SS toward human gene therapy," Science, 244:1275-1281, 1989.

Fritschy et al., Diabetes, 40:37, 1991.
~0 Frougel et al., "Familial hyperglycemia due to mutations in glucokinase," N. EngL J. Med., 328:105-112, 1993.

Gazdar et al. ''Contint~ c~ clonal, insulin- and s-~m~tost~tin-sec~ g cell lines est~lichel1 from a tr~n.cpl~nt ~' le rat islet cell tumor," Proc. Natl. Acad. Sci. USA, 77:3519-3523, 1980.
edlllin B., E. Larson, et al., (1993) 'Yhe selective amylin antagonist, AC187, e~h~-~res the insulin ~ onse during i~ a ~ellous glucose tol~,ai ce tests in ~n~sth~ti7.~.~ rats." Diabetes 42(Suppl.1):229A.
~0 &elb et al., "Targeting of hex~l-in~ce 1 to liver the hepatoma mitochQl t3~ Proc. Natl. Acad.
Sci USA, 89:202-206. January, 1992.

Gerlach et al., "Cons~ clion of a plant disease re~;ct~ce gene from the satellite RNA of tobacco rinspot virus," Nature (~ondon). 328:802-805, 1987.
German et al., "The insulin and islet amyloid polypeptide genes contain sirnilar cell-spe~ ific promoter el.~ that bind iA~ntir~l ~cell nuclear comrl~Ps," Mol. Cell. Biol., 12:1777-1788, 1992.

Ghosh and Bachha-val, "Targeting of liposomes to hr~a~ es," In: Wu G. and C. Wu ed. Liver ~lice~es, targeted .~ si.c and therapy using specific receptors and lig~nr~. New York:
Marcel Dekker, pp. 87-104, 1991.

Ghosh-Choudhury et al., "Protein lX, a minor cG~ Jollell~ of the human adenovirus capsid, is es~nti~l for the p~ ing of full-length genomPS~ ~ EMBO J., 6: 1733-1739, 1987.

Giddings et al., "Effects of glucose on proinsulin m~Ssçn~er ~NA in rats in vivo," Diabetes, 31:624-629, 1982.

Goding, In: Monoclonal Antibodies~ .les and Practice, 2d ed., ~c;~1~mic Press, Orlando, FL, pp. 60-61, 65-66, 71-74, 1986.

Gomez-Foix et al., "Adenovirus-mediated lr~l~f~r of the muscle glycogen ~hns~hhl~lase gene into hepatocytes confers altered regulation of glycogen," J. Biol. Chem., 267:25129-25134, 1992.

Gopal, "Gene transfer method for transient gene e~ ession, stable sr~n~c~ion~ and co~ sfec~ion of ~ or cell cultures," Mol. Cell Biol., 5: 1188-1190, 1985.

Gosscn et al., "Tight control of gene c~ ssion in ..- - ,,.,~ cells by tetracycline-~s~nsive l.lulllGt~"~," Proc. Natl. Acad. Sci USA, 89:5547-5551, 1992.

~5 Graham & Prevec, "Adenovirus-based e~plcssion vectors and l~coll~binant ~rc;l.~,s,"
Biotechnology, 20:363-3gO,lg92.

W O97126321 PCT~USg7/00761 Graham and Prevec, "~nirulz~tion of adenovirus vector," ~n: E.J. Murray (ed.), Methods in Molecular Biology: Gene Transfer and Expression Protocol, Clifton, NJ: Ilulllalla Press, 7:109-128, 1991.

5 &raham and van der Eb, "A new technique for the assay of infectivity of human adenovirus 5 DNA", Virology, 52:456~67, }973.

Graham et al., "Ch~ ;ctics of a human cell line transformed by DNA from human adenovirus type 5", J. Gen. Virol., 36:59-72, 1977.
Grampp et al., "Use of regulated secretion in protein production from animal cells: An overview," Advances in Bioc~lemical Engincering, 46:35-62, 1992.

Gross et al., "Par~ial diversion of a mutant proinsulin (B 10 aspartic acid) from the regulated to the co~ e S~lctOI~ patll~ ay in l.~sfGcled AtT-20 cells," P.N.A.S., 86:41074111,1989.

Grunhaus and ~orwitz, "Adenovirus as cloning vector," Seminar in Virology, 3:237-252, 1992.
~0 Hakes et al., "Tcol~tion of two compl~mPnt~ry deoxyribonllrki~ acid clones from a rat in~ulin~m~ cell line based on ~imil~riti~s to Kex2 and furin sequences and the specific loc~li7A~ion of each transcnpt to e..Aor.; .c and ~ ;..e tissues in rat,"Endocnnology, 129:3053-3063, 1991.

~5 Halban, "Structural ~lom~in~ and moleclll~r lifestyles of insulin and its ~IC.,Ul~Ol:~ in the ~c~ca~.c beta cell," Dtabetologia. 34:767-778, 1991.

Halban et al.. "Abnormal glucose metabolism ~CQ~ es failure of glucose to stimnl~te insulin release from a pancreatic cell line (RINmSf~," Biochem. J. . 212:439-443, 1983.

Halban et al.. "High~ Çu~ -.re liquid chlo...~tography (HPLC): a rapid, fle~cible and sensitive S method for se~ islet prnincl-lin and insulin," Diabetologia, 29:893-896, 1986.

Halban et al., "IntrAre~ r degradation of insulin stores by rat ~ cl~atic islets in vitro: An re p~ wa~/ for h~.." o~ cic of ~ alic insulin content," J.B.C, 255:6003-6(~06, 1980.
Harland & We.nl...~, "Translation of m~mm~ n mRNA inject~d into ~Pnopl-c oocytes is ,pec;fl~Ally inhi1~jt~.-1 by ~ e R~A," J. Cell Biol., 101:1094-1099, 1985.

}~loff, T. and (3~rl~rh, W.L. "Simple RNA cnzy..les with new and highly specific endo~ibol.uclcAiG activities." Nature, 334:585-591, 1988.

Heartlein e~ al., "Long-tcrm production and delivery of human growth h-nnnne in vivo,"
P.N.A.S., 91:10967- 10971, 1994.

ITei~ h et al., "Pre-glucagon messenger ribo~ rleic acid: Nucleotide and encoded a~uno acid sey--~n~es of the rat par, ~.e~l-c comple.,lemal y deo~.y.ib~ rl~ic a~id., Endocrinology, 115:2176-2181, lg84.

J~.... ...oll~1 and MULYC~, "Use of aderlo~c~oc;~ted virus as a i~ n DNA cloning vector:
Tr~nc(hlc-ion of neomycin le~;ct~n~e into ~.a~ n tissue culture cells," Proc. Nat'l Acad. Sci USA, 81 :6466-6470, 1984.

WO 97/26321 PcTrusg7/00761 .. rr~ et al., "F.ffiri~nt gene ~ in live mice using a unique retroviral pqc~qein~ line,"
DNA Cell Biol., 9:713-723,1990.

Herz and Gerard, "Ad~,n~vi~ m~ ~d transfer of low density li~u~l~ot~ln receptor gene acutely S acceler. tes eholPst~ rol cle;~ce in nûrmal mice," Proc. Nat'l Aca~l. Sci. USA
90:2812-2816,1993.

H~cner et al., "Mûlecular physiology of the islet amyloid polypeptide (LAAP)/ amylin gene in man, rat, and transgemc mice," J.Cell. Biochem., 55S:39-53,1994.
Horwich et al. "Sylllhcsis of h.,p~df l~ovims particles that contain replication-defective duck h~.pa-itic B virus gen~ s in cultured HuH7 cells," J. Virol., 64:642-650, 1990.

Hoso~a, et al., "Upregulated h~Yol-inqce activit,v in icol~tP(l islets frûm f~ etir 90%
pancreat~o~ l rats." DiaJetes 44:132~-1333, l99S.

Hughcs et al., "Tr~cfec-tion of AtT~20ins Cells with GLUT-2 bUt Not GLUI-l Confers Glucose-stim~ P~ Insulin Secretion", The Journal of Biological C~zemistry, 268: 1520S-15212, 1993.
Hughes et al., ~r"2 jn~ of Ghlrose stim~ tPd Insulin Secretiûn and Biosynthesis in Non-islet Cells"t Proc. Na~l. Acad Sci. USA, 89:688-692, 1992.

Hughes et al., "E~ ,SSiOI of normal and novel ~'l co'-in~ mRNA's in ant~,ior ~ and islet cells," J. Biol. Chem.. 266:4521-4S30, lg91.

W O 97126321 PCTrUS97100761 Hughes et al., "Transfection of AtT-20ins cells with GLUT-2 but not GLUT- 1 confers glucose-stim~ ted insulin secretion: relationship to glucose metabolism," J.B.C, 268:15205-15212, 19g3.

~ Inagaki et al., "Reco~ on of IKATP: An inward rectifier subunit plus the sulfonylurea ,~ce~,)tor," Science, 270: 1166-1170, 1995.

Ic~ con et al., "Mode of action of piouitary growth hormone on target cells," Ann. Rev. Physiol., 47:483~99, 1985.
TChibqc~li et al., "Hyperchole~ ol~ ia in Low Density Li~o~lvt.,hl R~csptQ~ Knoc~out Mice and its Reversal by Adcnovirus-m~ Gene Delivery," J. Clin. Invest, 92:883-893, 1993.

Jenkins et al., "Getting the glycosylation right: ~mrlir~tion~ for the biotechnology industry"
Nature Biotechnol. 14: 975-976, 1996 Johnson el al., "Und~ xt,l~ssion of Beta cell high Km ~I-)cose L.~ls~)olLers in nonin.c~llin de~.ldent ~ et~s~ ~ Science, 250:546-549, 1990.

~0 Jones and Shenk, "~col~tion of deletion and s~lbstinl~ion ml~t~nt~ of adenovirus type 5," Cell, 13:181-188, 1978.

Jonsson et al., "Insulin-~o l~t~-factor 1 is required for pan. leas devel~"lent in rnice," Nature, 371:606-609, 1994.
Joyce, "RNA evolution and the origins of life," Nature, 338:217-244, 1989.

W O 97/26321 PCT~US97100761 K~n~;q et al., "lncreased eAI~Iei,sion of DNA coi~ oduced with nuclear protein in adult rat liver," Science, 243:37S-378,1989.

K~lsson e~ al., EMBO J., 5:2377-2385,1986.
s Karlsson et al., "A m~t~tion~l analysis of the insulin gene ~ scli~nion control region:
e~ ssion in ~-cells is de~e.,de,lt on two related sequences within the enh~nrer," Proc.
Natl. Acad. Sci. USA, 84:8819-8823,1987.

0 K~rlcson et al., "Individual protein-binding dom uns of the insulin gene el-h~ el positively ~li..rate ,B-cell-specific L.~ clil.l;on," Mol. Cell. Biol., 9:823-827,1989.

Kato et al., "Expression of hepatitis ~ virus surface antigen in adult rat liver," J. Biol. Chem., 266:3361-3364,1991.

~s~nfm~n and Sharp, "Construction of a m~ r dihydrofolate reduct~se cDNA gene: Analysis of signals utilized for efficient ci~essi~n," Mol. Cell. Biology, 2:1304-1319,1982.

Keller et al., "Insulin prophylaxis in individuals at high risk for the development of type I
~i~het~S mPllihlC," Lancet., 341:927-928,1993.

KPnn~y et al., "The miniQ~t~llite in the ~ etes suscei~til.ility locus ID D M2 regulates insulin transcription," Nature Genetics. 9:293-2g8, 1995.

~~5 Kim & Cech, "Three ~iimPn~jonal model of the active site of the self-splicing rRNA ~ or or Te~ahyrnena," Proc. Natl. Acod. Sci. VSA, 84:8788-8792,1987.

W O 97/26321 PCT~US97/00761 Klein et al., "High-velocity rmcroprojectiles for delivering nucleic acids into living cells,"
Nature, 327:70-73, 1987.

Krniec, E.B., Cole, A., and Holloman, W.K. 1994. The REC2 gene encodes the homologous S pairing protein of Ustilago maydis. Mol. Cell. Biol. 14: 7163-7172.

Knaack et al., "Clonal Insulinoma Cell Line That Stably ~ Correct Glucose Rci~nsi~ness." Dia~etes, 43:1413-1417, Dec~...bcr, 1994.

Koh}er ~c Milstein, Nature, 256:495497, 1975.

Kohlcr & ~ilctçin, Eur. J. Immunol., 6:511-519, 1976.

Kreymann and Williams, "Glucagon-like peptide- 1 7-36: a physiological incretion in man,"
Lancet, 2: 1300- 1303, 1987.

Kruse et al., "An ~n~locnn~-specific elem~t is an integral co~ enl of an exoc,ine specific pallc~e~ic e~h~nrer," Gcnes and Dev., 7:774-786, 1993.

Kyte and Doolittle, J.MoLBiol., 157:105-132, 1982.

Lacy et al., Science, 254: 1782-1784. 1991.

Larsson and Litwin, "The growth of polio virus in human diploid fibroblast grown with cellnlose ~ OCail.el~ in s~lspen~ion culture," De~. Biol. Standard., 66:385-390, 1987.

Le Gal La Salle et al., "An &d~nO~ uS vector for gene transfer into neurons and glia in the brain,"
Science, 259:988-990, 1993.

WO 97126321 rCT~US97100761 Leffert et al., "Rat amylin: Cloning and tissue-specific expression in pancreatic islets," Proc.
Natl. Acad. Sci. USA, 86: 3127-3130, 1989.

S Leonard e~ af., "Ch~ n of Som~os~tin Transactivating Factor-l, a Novel Homeobox Factor That Stim~ t~s Som~tost~tin Expression in Pallcl~atic Islet Cells," Mol Endocrinol, 7: 1275-1283, 1993.

Levrero e- al., "Defectivc and nondefective adenovirus vectors for e~ sshlg foreign genes in 0 vitro and in vivo," Gene, 101: 195-202, 1991.

Li et al., "Effect of disruption of actin fil~m~r ts by Clostridium botlllimlm C2 toxin on insulin secretion in H~-T15 cells and pallcle~lic islets," M.B. C, 5: 1199- 1213, 1994.
~5 Liang et al., "Effects of ~lt~rn~ RNA spliring on glucokinase isoform activities in the p~lclcatic islet, liver, and ant~"iur piLuiL~ ," J.B.C., 266:6999-7007, 1991.

Liang et aL, "Enhanced and Swi~l hs~le Expression Systems for Gene-Transfer," ~. Cell.
Biochem. S~ le... nt 21A:379~ lg95.
Liang, e~ al. "~luçose regulates glu~ in~sÇ activity in cultured islets from rat pal'lC;l~.ZiS." J. Biol.
Chcm. 265:16863-16866, 1990.

Licber, A. and Strauss, M. "Sel~lio~ of effiri~nt cleavage sites in target RNAs by using a ribo~yllle eA~,~s~ion libraty." Mol. Cell. Bwl., 15: 540 551, 1995.

Lim, "rn~ o~ of biological ma~erials," US Patent 4,352,883, October 5, 1982.

CA 0224643l l998-08-l2 W O 97/26321 PCTnUS97/00761 Lin et al., "Pit- 1 -depen-len~ yiluil~y cell proliferation involves activation of the growth hl)rmQn~
rPleacing factor receptor gene," Nature, 360:765-768, 1992 Linden et al., "Pore protein and the h~xcl-in~ce-binding protein from the outer membrane of rat livermitocho~lri~ are ~lenti~ql," FEBSLett., 141 189-192, 1982 McGarry, D J, "Disold~"~,d metabolism in ~ etes: Have we und~ h ~ ed the fat co~ o~ t?", Journal of Cellular Biochemistr~ 55, supp:29-38, 1994 ~0 Macejak and Sarnow, "Tntern~l injti~-jon of trar~slation m~ t~d ~y the 5' leader of a cellular rrR~A," Nature, 353:90~94, 1991 Madsen, et al., Proc. Natl. Ac~d. Sci. U.S.A., 85:6652-6656, ~988 ~5 Mann et al., "Construction of a retrovirus p~ck~ging mutant and its use to produce helper-free defective retrovirus," Cell, 33:153-159, 1983 Marchetti et al., "Pulsatile insulin secretion from irol~tçd human ~a~ c.Galic islets," Diabetes, 43:827-830, 1994 Marie et a~., "The pyruvate kinase gene is a model for studies of glucose-d~ ~ nt regulation of gene eAp~ ,ion in the entloçrine ~al~ClcatiC Beta-cell type," J.B.C., 268:23881-23890, ~5 Markowitz et al., "A safe packaging line for gene l~ srG~ Separating ~riral genes on two dirÇ~ t pl~cmi/1$~ J Vi~ol., 62:1120-1124, 1988 w o 97n6321 PCT~US97/00761 Matchinsky, "CJIUeC' jnqCe as as sensor and ".~ hQIic signal generator in ~ tic Beta-Cells and hepatocytes," Diabetes, 39:647-652, 1990.

Mayo, "Mc l~c~ r cloning and e7~yl~ssion of a ~ilui~y-specific receptor for growth hol o~-c S rr1e~Cin~ honno~," Mol. Endocnnol., 1734-1744, 1992.

Meglasson and ~S~h;~-CL y, "ran~ lic islet glucose metabolism and rc~ ation of insulin secretion," Diabetes Metab. Rev., 2:163-214, 1986.

10 Michel & Westhof, '~I~od~lin~@ of the three~ mPn~iQn~l arehi~ect~lre of group I catalytic introns based on colllp~ua~ e se~l~e-nce analysis," J. Mol. Biol., 216:585-610, 1990.

Milgram el al., "Expression of individual forms of peptidylglycine alpha-arnidating monooxygenase in AtT-20 cells: Endoplu ~ cllytic procescing and routing to Se~ ,lOl~/
lS granules," J.C.B., 117:717-728, 1992.

Miller et al., "IDX- 1: a new homeodom~in l~ on factor expressed in rat ~lcn~a~c islets and duodenllm that transactivates the ~om~ost~tin gene," EMBO J., 13:114S-1 156, 1994.

20 Miya_aki et al., "Establi~hm~t of a pancreatic Beta-cell line that retains glucose-inducible insulin se~ o~: speci~ l~,f~ e to e~ ssion of ~ ose !-~ ,isofollLs,' Endocrinolo~y, 127:126-132, 1990.

Mizrahi, "Production of human inle.f~,.olls: An overview," Process Biochem., (August):9-12, 2~ 1983.

Mojsov et al., "Both amidated and non~mi~ted forms of glucagon-like peptide 1 are srth~i7pd in the rat il~tfj~ and the ~ ,as," J.B.C., 265:8001-8008, 1990.

W O 97~6321 PCT~US97/00761 Mojsov et al., "Preglucagon gene ~A~ ,ssion in the pancreas and intestine diversifies at the level of post-translational proces~in~ r~ J.B.C., 261: 11880-11889, 1986.
Monia, B.P, Johnston, J.F., Sasmor, H., and c~u~ ..s, L.L. 19g6. Nuclease recict~nce and S :mti~e~.ce activity of modified oligonucleotides targeted to Ha-ras. J. Biol. Chem. 271:
14533-14540.

Moore e- al., "E~pn_c.5in8 a human proinsulin cDNA in a mouse ACTH-secr~ting cell.
Intr~-~ell~ r storage, proteolytic ~loces~ , and secretion on s~imul~tit)n~ Ce~l, 35:531-538, 1983.

Moore et al., "Se~;~el ~- ~ protein targeting in a ~iLuil~ ~ cell line: Dirr~.~.. ial ~ of foreign secretory proteins to distinct sc~ t~ alh~ ~s," J.C.B., 101:1773-1781, 1985.

Muir et al., "Antigen specific lmmllnoll ~ ~y. Oral tol~ ri ~-re and s~lbc~ ous i.~ tion in the L~c~ nt of insulin-depecd~nt diabetes," Diabe~es. Melab. Rev., 9:279-287, 1993.

Muir e~ al., "Insulin immnni7~tion of nonobese ~ hetic mice in~ cec a protective inculitic ch. r~reri7~d by ~imini~h~d intraislet hllc~on-gam~na transcription," J. Clin. Invest., 95:628-634, 19g~.

Mulligan, "The Basic Science of Gene Therapy," Science, 260:926-932, 1993.

~Illli~n et al., "Selection for animal cells that express the Escherichia coli gene coding yp~lthine-guan~nephosphoribosylll~lsre~se~" Proc. Natl.Acad. Sci. USA, 78:2072-2076, 1981.

W O 97/26321 PCT~US97/00761 M~lllig~n et al., "Expression of a b2~rten~l gene in m~mm~ n cells," Science, 209 1422-1427, 1980.

r~ 1ni, M., S. Su~i, et~l. (1990) "Effects of amylin on insulin secretion from RINm5F
S cells." Diabetes 39(Suppl 1):266A

N~ S and Steiner DP Endocrinology 130: (2) p748-54, 1992 Nce.~ -Arbez M et al., J Biol Chem, 268 (22) pl6098-100, 1993 Ne. gar~, and McGarly, "Metabolic coupling factors in ~arlC~d~iC Beta-cell signal ~n~d~lcti~ n "
Ann. Rev. Biochcm., 64:689-719, 1995 N~ gLd et al., "Gl~co'~ ce and glucose ~ spo~ler e~ s~ion in liver and islets: impli~ )n~
forcontrol of glucose hol. ro~ ," Biochcm. Soc. Trans., 18:851-853, 1990 Nicolas and Rubinstein, "R~t~ u~ al vectors," In: Vectors: A survey of m~lec.~lAr cloning vectors ~nd their uses, Rodriguez and Denhardt (eds ), Stol~eh~ UIL~ orth, pp. 494-513, lg88.
Nicolau and Sene, "Liposome-m~ DNA ~ r~,r in eukaryotic cells," Biochem. Biophys.
Acta, 721: 185-190, 1982 Nicolau et al., "Lipos~ s as carriers for in vivo gene ~nsÇ~. and e~.es~ion~" Methods En~yrnol., 149:157-176, lg87.

Nilsson and Mosbach, "Il~ lobilized animal cells," Dev. Biol. Standard., 66:183-193, 1987 Nielsen Acta Endocrinol: 266 pl -39, 1985 Wo 97/26321 PCTAUS97J00761 O'Shea and Sun, Diabetcs 35:943-946,1g86.

Ogawa et al., "Loss of glucose-in~ueed insulin secretion and GLUT-2 expression in tr~n.spl~n Beta-cells," Di~betes, 44:75-79,1995.

Ohagi et al., "~der-tifi~ ation and analysis of the gene çncorling PC2, a plùho~ une col,l,el~ase e~cssed in ne.uu~ orrinç tissues," P.N.A.S., 89:4977-4g81,1992.

10 Ohlsson et al., "IPF}. a hom~QdQln~in-col~t~ining transactivator of the insulin gcne," EMBO J., 12:4251-4259,1993.

Ohsawa, H., A. K~n~tQulca, et aL, (1989) "Islet amyloid polypeptide inhibits glucose-stimulated insulin secretion from icol~çd rat p~.c~c islets." Biochcm.Biophys Rcs. Camm.
160:961-967.

Orci et al., "Evidence that down-reg~ tior~ of Bcta-cell gl~cQse tl~ Vl~:~a in non-insulin-~iPpçn~l~nt ~ ktes may be the cause of ~ tic l,y~rgl~rcemia," P.N.A.S., 87:9953-9957, lg90.

Orskov et al., "Biological effects and metabolic rates of gluc~el-nlikç peptide-l 7-36 amide and eh~c~onli~ pe~d~- 1 7-37 in healthy ~ ~bJe -~c are in~ictin~ c~able~ Diabetes, 42:658-661,1993.

25 Orskûv et al., "~o~ Y~ sequçnre of gluca~on-like peptide- 1 from human and pig small ;-~r~-s~ e," J.B.C, 264:12826-12829,1989.

W O 97126321 PCTrUS97/00761 Ouafik et al., "Developm~nt~l regulation of peptidylglycine alpha-~miA~in~ monooxygenase (PAM) in rat heart atrium and ventricle," J.B.C., 264:5839-5845, 1989.

Ouafik et al.t "The m~ ;r...~.-tional peptidylglycine alpha-~mirl~ting monooxygenase gene:
Exon/intron org~ni7~tion of catalytic, ~rUceCcil-g, and routing dotT~sinc," Molecular Endocrinology, 6:1571-1584, 1992.

Owerbach and Aagaard, "Analysis of a 1963-bp polymo~phic region fl~nlring the hurnan insulin gene," GENE, 32:475-479, 1984.
Palmer et al., "Production of human Pactor IX in ~nim~lc by gen~ti~ / m~tlifi~l skin fib~obla~ Potential therapy for hemophilia B," Blood, 73:438445, 1989.

Palmer et al., r'~Jenlo,tic~lly mor1tfi~ rl slcin fibrobl~ts persist long after t~ n~lion but gradually in&~;LiYale introduced genes," Proc. Natl. Acad Sci. USA, 88:1330-1334, 1991.

p~ ?itiC, P. et al., " f'~ .on of a viroid ~ccoci. ~d with avacado sunblotch disease."
Virology, 99: 145-151, 1979.

~0 Pas~ind et al., "Depçn~en~e of moloney murine l~uk~rni~ virus production on cell growth,"
Virology, 67:242-248, 1975.

Pavla~is and Hamer, "~ tjoll of a metallolhion~in-growth honnone hybrid gene in bovine papilloma virus," P.N.A.S. 80:397 401, 1983.
2~
Peers et ~1., "Insulin Expression in ~ ,dliC Islet Cells Relies on Coopcl~ti~c Intcla~tions be~ ,en the Helix Loop Helix Factor E47 and the Hol~lcobox F~tor STF- 1," Mol Endocrin, 8:1798-1806, 1994.

W O 97/26321 PCT~JS97/00761 Pelletier and Sonel be.g, "Internal ini~ia~ion of translation of eukaryotic mRNA directed by a se~uellce derived from poliovirus RNA," Nature, 334:320-325. 1988.

S Perales et al., Proc. Natl. Acud. Sci. USA, 91 :408~4090, 1994.

Percy, A.J., Trainor, D.A., et al., (1996) "Development of sensitive immt~noass~ys to detect amylin and amylin-like peptides in unextracted plasma." Clin. Chem. 42:576-585.

10 F'e.~ , R. et al., "F~t~nr~eA target-site speçificity for a h~,nç,l~ead ribozyme." Gene, 113:157-163, 1992.

Perrotta and Been, Biochem. 31: 16, 1990.

15 ~ctl ;r~ i, "Should c~ntin~lQus cell lines be used as ~ ,dtes for biological products?," Dev.
Biol. Standard., 66:3-12, 1985.

Philippe et al., "Multipo~nti~l phenotypic expression of genes enco~ling peptide honno~es in rat inclllinoma lines," J. Clin. ~nvest., 79:351-358, 1987.
Phillips eJ al., In: Largc Scale M~ n Cell Culture (Feder, l. and Tolbert, W. R., eds.), A~P demir Press, Or}ando, FL, U.S.A., 1985.

Pieber et aL, "Direct plasma r~dioimmllno~c~y for rat amylin-(1-37): conrçrtrations with ac~luir~d and g¢netic obesity," Am. J. Physiol., 267:E156-E164, 1994.

Pittner, l~.A., Albrandt, K., et al., (1994) "M~lPc~ r physiology of amylin." J.Cell Biochem.
55:19-28.

W O 97126321 PCTnUS97/00761 Polakis and Wilson, "An Intact Hyd~hobic N-Terminal Sequence Is Critical for Binding of ~at Brain Hexokinase to Mitochondria," Arch. Biochem. Biophys., 236(1):328-337, January, 1985.
s Potter et al., "F.nh~nf çr-dependent e~lcssion of human k imml~noglo~ulin genes introduced into mouse pre-B lymphocytes by eleetlu~lation," Proc. Nat'lAcad. Sci. USA, 81:7161-7165, lg84.
~0 Powell et al., "Efficient targeting to storage granules of hurnan proincl~linc with altered pn,~ptide dom~in " J. CB., 106: 1843- 1851, 1988.

Prody, G.A. et al., "Autolytic proces~in~ of dimeric plant virus satellite RNA." Science, 231, 1577-1580, 1986.
Quaade et al., "Analysis of the protein products enco~ed by variant gll-~al-in~ce Ll~nscll~ls via ~,A~l~,ssion in ~cten~ " FEBSLett, 280:47-52, 1991.

Racha et al., Biotechnology Techni4ues, 9:169-174, 1995.
Radvallyi et al., "Pancreatic beta bells cultured from individual pr~nf oplscti~ foci in a mllltict~e tnmnrigpn~cic pathway: a pot~nti~lly gcneral t~r~ e for isolating p~ ologically r~l)~selltative cell lines," Molec. and Cen Biol.,13(7): 42234232, 1993.

25 Ragot e~ al., "~ffiit i~rlt ad~,.lo~ d h~f~l of a human mini~ly~llui)hin gcne to skeletal muscle of md~ mice," Nature, 361:647-650, 1993.

WO 97~6321 P ~ ~US97/00761 nhn~ Hurek & Shub, "Sel~splicing introns in tRNA genes of widely divergent b~t~ri~ "
Na~re, 351:173-176, 1992.

Renan, "Cancer genes: current status, future pros~ts, and applicants in radiotherapy/oncology,"
S ~adiother. Oncol., 19: 197-218, ] 990.

Rhodes and Alarcon, "What beta-cell defect could lead to hy~ ,oin.c.liinemi~ in NIDDM?,"
Diabe~es, 43:511-517, 1994.
~0 Rhodes and Halban, "The intr~ellul~r h~n-llinf~ of insulin-related pep~ides in isolated ~ancrea~ic islets," Biochem. J., 251:23-30, 1988.

Rich et al., "Development and analysis of rtcolllbinant ad~,no~ ses for gene therapy of cystic fibrosis," Hurn. Gene Ther., 4:461-476, 1993.
1~
Ridgeway, "M~mm~ n expression vectors," ln: R~ e7 RL, Denhardt DT, ed. Vectors: A
survey of molecular cloning vectors and their uses. Stc!l~rh~ Bul~ O,lh,pp~ 467492, lg88.
~0 Rippe et al., ''DNA-m~ tec~ gene transfer into adult rat hep~ocytes in primary culture," Mol.
CellBiol., 10:689-695, 1990.

I2il~ nk~ c, J., Chait, B.T., et al. (1996) "Hcterogcneity of naturally-occ~ g human amylin due to glyco~l~ion." Dia~etes 45 (Sup.2):234A.
2~
R~se~f~l~l et al., "A~ s-rnediated transfer of a recombin~t al-~~ in gene to the lung epith~ m in vivo," Science, 252:431 434, 1991.

W O 97126321 PCT~US97~761 Rosenfeld et aL, "ln vivo 1-~ sf_r of the human cystic fibrosis tr~n~ .h~ ~e cond~ t~ e regulator gene to the ~rway epithelium," Cell, 68: 143-155, 1992.

Rouille el al., ~Dirr~r~ al procec~ing of proglucagon by the subtilisin-like prohr.rrn~-nP
5convertases PC2 and PC3 to ~,.,n~,.ate either glucagon or glucagon-like pepti~e." J.B.C., 270:26488-26496, 1995.

Rouille ef al, "Proglucagon is processed to glucagon by proho~c-ne conv.r.lase PC2 in alphaTCl -6 cells," P.N.A.S, 91 :3242-3246, 1994.
E~oux et al., "A versatile and pot~nti~lly general apl~loach to the targetmg of s~ ifir cell types by retrovil~ses: Appli~tio~ to the infection of human cells by means of maJor hi~toco...~ ility c~ le~ class I and class II qnti~,-nC by mouse CCOI~o~ic murine le~L~ "iq. virus-derived viruses," Proc. Natl. Acad. Sci. USA, 86:9079-9083, 1989.
15Rubin, B.P., ~ lSOn, D.O., and Holl~m~n, W.K. 1994. S~ of REC2, a recombinational repair gene of Ustilago maydis, and its fimetion in homologous ~c.l..l.;..ation ~I-.~n p)qcmi~ and chromnsom~l sequences. Mol. Cell. Biol.. 14: 6287-6296.

.S7 ~P1~jn e~ al., "Prevention of type I (liq-l-etes in NOD mice '~y adjuvant immlmotherapy~"
20I~iabe~es, 39:583-589. 1990.

S~mh~ni.c et al., "Use of regulated s.~C,~,IiOl~ in protein produc~ion from animal cells: An ev~qln~ ion with the AtT-20 n~del cell line," Biotechnology and Bioengineenng, 35:771-780, 1990.
Sqmhq~1;C etaL, "A model of se~ t~l~ protein tr~ffi~ in~ in r .~CO~~.h ~ 1 AtT-20 cells,"
Biotechnologyand Bioen~ e~ri..g, 36:28~295, 1991.

SarLke et al., "An islet amyloid peptide is derived from an 89-amino acid precursor by proteolytic plocecs;..g," ~.B.C., 263:17243-17246, 1988.

Santerre et ~1., "Insulin synthesis in a clonal cell line of simian virus 40-transformed h~mcter pancreatic beta-cells," P.N.A.S., 78:43394343, 1981.

Sarver, et al, "Ribozymes as potential anti-HIV-1 tl~la~e.ltic agents," Science, 247:1222-1225, 1990.

Sato et al., Proc. Nat'l Acad. Sci. U.S.A. 48: 1184-1190, 1962.

Sauer, "Manipulation of transgenes by site-specifir recombina~ion: Use of Cre l~co,llbil~ase,"
Methods in Enzymology, 225:890 900, 1993.

Sc~nlQn et al., "RiboLyllle-m~ Je~l cleavages of c-fos mRNA reduce gene ~ ssion of DNA
~yl~thesis e~ "les and m~t~llothior~in," Proc. Nat'l Acad. Sci. USA, 88:10591-10595, ~991.

Scharfmann et al., "Long-term in vivo eApression of retrovirus~ tP~I gene transfer in mouse fibroblast-mrl~ntc,"P.N.A.S. USA,88:4626-4630, 1991.

Scharp et al., "F~ote~l;QI~ of f 1~ .s~lated huïnan islets imrl~nted without immll~o~u~ ssion in pati~l.ts with Type I or Type n diabetes and in nor~ etiC control subiects," Dia~etes, 43: 1167-1170, 1994.
S~hmi~lt et al., "Synthesis and ta,~,ting of insulin-like growth factor-l to the hormnr e storage granules in an euloc~ e cell line," J.B.C, 269:27115-27124, 19!~4.

Schnedl et al., "STZ ll~spoll and cytotoxicity: Specific enh~ e~ t in GLUT-2-e~ ;ssi~g cells," l)iabetes, 43: 1326-1333, 1994.

SchnPt7ler et al., "Ac~ptatio~ to ~ h~iological levels of insulin gene e~ ,s~ion in S transgenic mice: Evidence for the im~.~ ce of l o~lL.~rcriptional regu}ation," J. Clin.
Invest., 92:272-280, 1994.

Schwab et aL, "Complete amino acid sequence of rat brain hexokinase. ~led--~ed from the cloned cDNA, and ~ posed st~ucnlre of a m~mm~ n hexokinase," Proc. NatL Acad. Sci. lJSA.
86:2563-2567, 1989.

See~ " "The human growth hv~ gene family: Nucleotide seqllence~ show recent di~ "ce and predict a new polypeptide ~~ e.," DNA, 1 ;239-249, 1982.

~5 Selden et ~1., "Tmrl~nt : ;ol- of gen~tir~lly en~ d fibroblasts into mice: Tmplir~tion~ for gene therapy," Science, 236:714-718. 1987.

Sevarino e~ al., "Cell-specific proces~ing of pl~,roso.. ~ost~tin in cultured n~ ocrine cells," J.B.C., 262:49874993, 1987.
Sevarino et al., "Tllylutlu~in rele~cing holl.lone (TRH) ~.~.,~so, IJ~ocesc~ J.B.C., 264:21529-2153S, 1989.

~;hP.h ~leh et al., "Effect of adjuvant therapy on development of r~ etes in mouse and rnan,"
~cel., 343:706-707, 1994.

Sioud e~ al., "~.,fo,---ed r;boL~ c de~L~ tumour ~ecl~sis factor mRNA in human cells," . J
Mol. Biol., 223:831-835, 1992.

W O 97~6321 PCT~US97/00761 Si7nnr-nko and ~qlbqn~ "Differential rates of con~r~.sion of rat prl~inclllinc I and II: Evidence for slow cleavage at the B-chain/C-chain junction of proinsulin II." Biochemistry J.. 278:621-625, 1991.
s S~ cL~ .~c and Steiner, "k3~-ntific~ti~n of a human insulinoma cDNA en~o~ a novel mqmmqliqn protein sLIu~;Lul~lly related to the yeast dib. sic proce~cing protease Kex2,"
J.B.C., 265:2997-3000, 1990.

Smith and Wilson, Arch. Bioch. Biophys. 287:359-366, 1991.

Steiner et al., "The new enzymology of pi.,~,ul~ procescin~ endoprote. ses," J.B.C., 267:23435-23438, 1992.

Stoffers et al., "~ ti~e mRNA SFii~in~ generates mllltiple forms of peptidyl-glycine alpha-~mi~qtin~ monooxygenase in rat atrium," P.N.A.S., 86:735-739, 1989.

Stoffers et a~., "Ch~ tion of novel rnRN~'s encoding e,lzrl~es mvolved in peptide alpha-,qmi~ ion," J.B.C., 266:1701-1707. 1991.
Stossel, T., "The Molo~l~lq-~ B. sis of Blood Di~--es," Ch-qrter 14, pp. 499-533, W.B. Sa~n~-s Co., Ph~ lfl~hiq PA, 1987.

Stratford-Pe~nrq~ld~t and Pe.lie~dr-t "Gene ll~if~l into qnimqlc: the p~ se of adenovirus," p.
51-61, In: H~man Gene Tran~er, Eds, O. Cohen-Haguenauer and M. Boiron, F litirncohn Libbey ~ ,te~, France~ 1991.

W O 97126321 PCT~US97/00761 Sl-alÇord-Pernc~ let et al., "Ev~ tic-n of the transfer and expression in rnice of an enzyme-encoding gene using a human adenovirus vector," Hum. Gene Ther., 1:241-256, 1990.

~ S~ ..A~-i e~ al., "Expression of the mouse dihydrofolate re~l~lct~e co...pl.~~..Fn~
deox~ .eleic acid in simian virus 40 vcctors," Mol. Cell. Bio~., 1 :854-864, 1981.

Sullivan et al., Science 252:718-721, 1991.
~0 Swarovsky et al., "Sodium butyrate in~ ces neulûc~ ocrine cytodirrtl~,ntiation in the in~lllinorr cell line RINm5F," Pancreas, 9:460~68, 1994.

Symons, R.H., "Avacado sunblotch viroid: primaty seqllen~e and ~rûl)oscd seco.~ . y structure."
N.A.R., 9:6527-6537, 1981.
Symons, R.~I. "Small catalytic RNAs." Annu. Rev. Biochem., 61:641-671, 1992.

Takeda et al., "Organiza~ion of the human GLUT-2 (Pal~.,nvatic Beta-cell and hepatocyte) g}ucose transporter gene," Diabetes, 42:773-777, 1993.
Tal~euchi et al., "Expression of human pancreatic peptide in heterologous cell 1ines," J.B.C., 266:17409-17415, 1991.

Temin, ''Rel~ùvilus vectors for gene transfer: F.ff~ient intcgldiou into and eA~.~ssion of 2~ exogenous DNA in vertcbra~e cell ~,c ~ n: Gene Tran~er, ~ h~rl~p~ti (ed.), New York: P~enumPress,pp. 149-188, 1986.

Tho...l.son, J.D. et al., "Ribu~yl-~es in gene therapy." Nature Medicine, 1:277-278, 1995.

W O 97126321 PCT~US97/00761 Thorens, "Expression cloning of the ~an. .~,atic betacell receptor for the gluco-incretion hormone glucagon-like peptide 1," P.N.A.S., 89:8641-8646, 1992.

~ Thorens et al., "Cloning and functional expression in bacteria of a novel glucose l-~lspolt~,l present in liver, i~t~.ali~r.~ kidney and beta-pancreatic islet cells." Cell, 55:281-290, 1988.

Thorens et al., "The loss of GLUT-2 expression by glucose-u,ll~,syonsive beta cells of d~/db mice is reversible and is inrlncetl by the ~ hetic environr~ t," J. Clin. Invest., gO:77-85, }O 1992b.

Thorcns et al., "l~odl~Ge~3 e~ ssion of the liverlbeta-cell glucose transporter isofonn in glucose-in~e~ beta-cells of tli~bctie rats," Proc. Natl. Acad. Sci. USA. ~7:6492-6496, 1990.
Thorne et al., "E~yl~,ssiOIl of mouse proopiom~ ocolliu in an incl-linorn~ cell line," J.B.C., 264:3545-3552, 1989.

Ting and Lee, "Human gene enco{ling the 78,000-Dalton glucose-regulated protein and its pseudo gene: SLIU~IU~C, co~ dtion~ and reg~ tion," DNA, 7:275-286, 198B.

Tokuyama el aL, "Evolution of beta~ell ~ru~ lion in the male Zucker ~i~hetic fatty rat,"
Dia~e~es, 44(12):1447-1457, l9gS.

Top et al., "~,.. -;~ -: ;ol- with live types 7 and 4 adenovirus vaccines. II. Antibody ll,syol~e and yl~lc~;~ive effect against acute respiratory disease due to adcno~u~ type 7," J. ~nfec~.
Dis., 124:155-160, 1971.

Ts~ d~ T., Bauchwitz, R, and Hollorn~n, W.K 1989. Isolation of the REC1 gene controlling lcc~lllbination in Ustilago maydis. Gene 85:335-341 Tur-Kasp~ et al., "Use of ele~;l ulx~ldtion to introduce biologically acti- e foreign genes into S prirnary rat hepatocytes," Mol. Cell Biol.. 6.716-718, 1986 Unger, "Diabetic hy~,lycernia: Link to i~ cd glucose lldnS~)Oll in pancreatic Beta cells,"
Science, 251:120û-1205, 1991.

Unger and Orci, "Glucagon and the Alpha cell," N.E. J.M., 304: 1518- 1524, 1981.
Usdin et al., UGastric inhibitory peptide receptor, a ~ of the secretin-vasoactive intesrin~
peptide l~,cel)tor family, is widely distributed in ~ Jh.,,~l organs and the brain,"
Endocrinology, 133:2861-2870, 1993.
van Wezel, "Growth of cell-strains and ~ cells on rnicrocarriers in homogeneous culture,"
Nature, 216:6~65, 1967 Var nus et al., "Retroviruses as m~ ~nc Insertion and excision of a nontransrolllling provirus alter the eAy~s~ of a resident tran~Çolllling provirus," Cell, 25:23-36, 1981.

Vollenweider et al., "E~)ces~ing of prQinclllin by furin, PC2 and PC3 in (co)tral-gfect~d COS
(monkey Kidney) cells," Diabetes, 44:107~-1080, 1995.

25 Vosberg, H.T. and FrlrC~ein F 1982. Effect of deoxynu~leoei-le phosphorothioates incol~ol~ed in DNA on cleal~dgc by ~ n e~ cs. J. Biol. Chem. 257:6S95-6599.

WO 97n6321 PCT~US97/00761 Voss-McGowan, M.E., Xu, B., and F.pstçin, P.N., "Insulin syn~h~sic~ secretory co...~ e~-re, and glucose ~-tili7~-ion are sel.~;t;i~ d by tr~ncge!-~ir yeast hp-~cl-in~ce7~ J. Biol. Chcm.
269:15~14-15818, 1~94.
~ Waeber e~ al., "Characteri7~tion of the murine high ~m glucose transportcr GLUT-2 gene and its sc~ ional regulation by ~I~JCOSe in a dirf~e -ti~d insulin-secreting cell line," J.
Biological Chemistry, 43:26912-26919, 1994.

Wagner et al., Science, 260: 1510- 1513, 1990.
Wagoner, P.K., Chen, C., et al.. (1993) "Amylin ~d~ es Beta-cell glucose sensing via effects on stim-ll--c-secretion collpling ~ Proc.NatLAcn~.~i, USA, 90:9145-9149.

Wang et al., "Design and performance of a packed bed bio~ or for the production of lceolllbillallt protein using semm-free .. ~ .. ," P~ocee~l;.. g of the J ~ ce Society for Animal Cell Technology, 1994.

Wang et al., "High density perfusion culture of hybrid--m~ cells for production of tn~rloclon~
antibodies in the CelliGen packed bed reactor, ~n: Anunal Cell ~echnology: Basic &
Applied Aspects, S. K~ o~. a et al., (eds), vol. 5, pp463-469, Kluwer Publishers, lelelh ..l~nAc, 1993.

Wang e~ al., "Mo-lified CelliGen-packed bed bioreactor for hybridorna cell cultures,"
Cytotechnology, 9:4149, 1992.
Watad~ et al.. "The human gl~CoL;~ce gene ,B-cell-type plolllo~l," Diabetes, 15:1478-1488, W O 97126321 PCT~US97/00761 Welsh et aL, "~t~ qtion of growth ho~ u~e synthesis by glucose in islets of Langerhans iSol~(l from transgenic n~ice," J.B.C., 261: 12915- 12917, 1986.

White and Wilson, Arch. Bioch. Biophys. 259:402~11, 1987.
s White and Wilson, Arch. Bioch. Biophys. 277:26-34, 1990.

Whitesell et al., "Transport and metabolism of glucose in an insulin-secreting cell line, BetaTC-1," Biochemistry, 30: 11560- 11566, 1991.
Willnow and Herz, "~omologous Recombina~ion for Gene ~e~lace,l,en~ in Mouse Cell Lines,"
Methods in Cell Biology, 43 pt A:305-334, 1994.

Wilson, "Brain hexokinase: A ~u~oscd relation ~l~ cn soluble-particulate distribution and .ctivity in vivo," J. Biol. Chem., 243:364~3647, 1968.

Wilson, "Hexokinqces," In: ~?e-views of Physiology, Biochemistry cmd Pha~nacology, Pette, D.
(Ed.), 126:65-174, 1994.
~0 Wilson, "Ligand inrhlred conf;l...qtio~c of rat brain hexokinase: Effects of glucose-6-phosphqtç
and il~or~ ~ic phospk~ ," Arch. Bioche~ Biophys., 159:543-549, 1973.

Wilson, "Regulation of ~ qn hexokinase activity," ~n: Regulation of Carboh~rate Me~abolism, Beitner, R. (Ed.) Vol. 1, CRC, Boca Raton, 4S-81, 1985.

Wong etal., "A~al~ce of ~-la~!;-.~qcG activity in an~mal cells upon ~ osul~ Ai~l~d gene transfer," Gene, 10:87-94. 1980.

WO 97126321 PCT~US97/00761 Wu and Wu, "Evidence for targeted gcne delivery to HepG2 hepatoma cells in vitro,"
Biochemisrry, 27:887-892. 1988.

WuandWu,Adv.DrugDelive~yReY., 12:159-167, 1993.

Wu and Wu, "Receptor-..~r~ t~d in vilro gene ~ br~lions by a soluble DNA carrier system," J.
Biol. Chem., 262:44294432, 1987.

Xie and Wilson, "Rat brain he~c' ir~ce the hy~l,ul~hobic N-~elllii.,us of the mitochon~ lly bound el~yllle is inserted in Ihe lipid bilayer," Arch. Biochem. Biophys., 267:803-810, 1988.

Yang et al., "/n vivo and in vitro gene transfer to .1~ -.n~ n somatic cells by particle bolnbal~:llll.,nt," Proc. Nat'l Acad. Sci. USA, 87:9568-9572, 1990.
1~
Yang et al., "Cellular immunity to viral ~ntige~ limits El-deleted adenoviruses for gene therapy," P.N~.S. USA, 91:4407 4411, 1994.

Yoon, K. Cole-Strauss, A., and Kmiec, E.B. 1996. Targeted gene correction of episomal DNA in ~I.z.. ~.n.~ n cells mrr~ d by a ch;.. ,~ ~ ,c RNA-DNA oligorn~cleotide Proc. Natl. Acad.
Sci. USA 93: 2071-1076.

Young, A.A., P. Carlo, et al., (1992). ~-8 37hCGRP, an amylin receptor antagonist, er h~nres the insulin l.,;.ponse and ~I.Iul~S the glucose l~,pol~ce to infused ~g~inc in ~ sl~- ti~e~
rats." Mol. Cell. Endocrin. 84: Rl-R5.

Yuan, Y. and Altman, S. "Seleclion of guide se~lu~lces that direct ~-f~lrier~t cleavage of rnRNA
by human rih~nnrle~e P." Science, 263:1269-1273, 1994.

W O 97~6321 PCTrUS97/00761 Yuan, Y. et al., "Targeted cleavage of mRNA by human RNase P." P.N.A.S., 89:8006-8010, lg92.

~ Yun and Eipper, ''AdAition of an endopl~smi~ reticulum retention/retrieval signal does not block maturation of e~-- A~ic~lly active peptidylglycine alpha~~ mnnoo~,g~ ase,"
J.B.G, 270: 15412-15416, 1995.

Zelenin et al., "High-velocity ,.,~ch~ni~ ~l DNA transfer of the Ch1O1A~ hC ~ jCOI aCetY~ Sie~Se gene into rodent liver. kidney and ~ r gland cells in organ eYpl~~ts and in vivo,"
FEBS Lett., 280:94-96~ 1991.

U.S. Patent 4,892,538 U.S. Patent S,399,346 U.S. Patent S,011,472 WO 89~01967 WO 89~01967 25 WO 90~02580 WO 90~15637 PCTNS97tO0761 WO 91/104?0 CA 0224643l l998-08-l2 W O 97/26321 PCT~US97/00761 ~U~NC~ LISTING

(1) GENERAL INFORMATION~
(i) APPLICANT: Newgard, Christopher B.
Halban, Philippe Normington, Rarl D.
Clar~, Samuel A.
Thigpen, Anice E.
Quaade, Christian Kruse, Fred (ii) TITLE OF INVENTION~ Recombinant Expression of Proteins From Secretory Cell Lines ~iii) NUMBER OF SEQUENCES: 69 ~iv) CORRESP~ ADDRESS:
(A) ~nnR~-cs~: Arnold, White & Durkee (B) STREET: P. O. Box 4433 (C~ CITY: Houston ID) STATE: TX
~E) COu.~.~Y: USA
(F) ZIP: 772l0-4433 (v) COMPUTER READABLE FORM:
(A) MEDIUM TYPE: Floppy disk (B) CO_I~U~K: IBM PC compatible (C) OPERATING SYSTEM: PC-DOS/MS-DOS
(D) SOFTWARE: PatentIn Release #1.0, Verslon ~:1.30 tVi ) ~UKk~N'l APPLICATION DATA:
(A~ APPLICATION NUMBER: US Unknown (B) FILING DATE: Concurrently Herewith (C) CLASSIFICATION: Un~n-_,,"
(viii) ALlORN~Y/AGENT INFORMATION:
(A) NAME: Hi~hl~n~r, Steven L.
(B) REGISTRATION NUMBER: 47,642 (C) R~:~n~N~DOCKET NUMBER: UTSD:426\HYL
tix) TFrr~COM~n~rTCATION INFORMATION:
(A) TELEPHONE: (512) 418-3000 (B) TELEFAX: (512) 474-7S77 t2) INPORMATION FOR SEQ ID NO:l:

W O 97/26321 PCT~US97100761 (i) S~Q~ CHARACTERISTICS:
(A) LENGTH: 515 base pairs (B) TYPE: nucleic acid (C) STR~Nn~DN~S: single ~D) TOPOLOGY: linear (xi) ~Uu~ DESCRIPTION: SEQ ID NO:l:
GAATTCCGGG G~lC~ G CCAlGGCCCl GTGGATGCGC ~r~GCCCC TGCTGGCGCT 60 ~C~IaGC~l~ TGGGGACCTG ACCCAGCCGC AGC~lll~LG AACCAACACC l~lGCGG~ 120 ACAC~-l~rG GAAG~lGl ACCTAGTGTG CGGGGAACGA GG~lr~ ACACACCCAA 180 GACCCGCCGG GAGGCAGAGG ACCTGCAGGT GGGGCAGGTG GAGCTGGGCG GGGGCC~-~GG 240 TGCAGGCAGC CTGCAGCCCT TGGCC~lGGA GGGGlCCClG CAGAAGCGTG GCATTGTGGA 300 ACAA~G~l ACCAGCATCT G~l'CC~-r~A CCAGCTGGAG AACTACTGCA ACTAGACGCA 360 GCCCGCAGGC AGCCCCC~AC CCGCCGC~lC CTGCACCGAG AGAGATGGAA TAAAGCCCTT 420 GAACCAGCAA AAAAAAAAAA AiUU~UU4AAA AiUUUUhhhAA AUUUU WAAAC CCCCCCCCCC 480 CCCC'--~~CAG CAATGGCAAC AAC~ll~CGG AATTC 5l5 (2) INFORMATION FOR SEQ ID NO:2:
N,~'~ CHARACTERISTICS:
(A~ LENGTH: ll0 amino acids (B) TYPE: amino acid (C) STR~"~'.I INI-:.C S
(D) TOPOLOGY: linear ~xi) SEQu~N~- DESCRIPTION: SEQ ID NO:2:
Met Ala Leu Trp ~et Arg Leu Leu Pro Leu Leu Ala Leu Leu Ala Leu l 5 l0 15 Trp Gly Pro Asp Pro Ala Ala Ala Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr Arg Arg Glu Ala Glu Asp Leu Gln Val Gly Gln Val Glu Leu Gly Gly Gly Pro Gly Ala Gly Ser Leu Gln Pro Leu 2~9 W O97126321 PCTnUS97/00761 Ala Leu Glu Gly Ser Leu Gln Lys Arg Gly Ile Val Glu Gln Cys Cys Thr Ser Ile Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Cys Asn l00 105 ll0 (2) INFORMATION FOR SEQ ID NO:3 ~i) S~UU~N~ CHARACTERISTICS:
(A~ LENGTH: 359 base pairs (B) TYPE nucleic acid (C) STR~N~ S: single (D) TOPOLOGY linear ~xi) ~Uu~C~ DESCRIPTION SEQ ID NO:3 CCGGGGATCC 11~GCCATG GCCCl~GGA TGC&CCTCCT GCCC~lG~lG GCG~lG~GG 60 CC~.~-laGGG ACCTGACCCA GCCGr~qCCT TTGTGAACCA ACAC~ GC GGCTCACACC 120 .G~GGAAGC TCTCTACCTA ~l~GCGGGG AACGAGGCTT CTTCTACACA CCCAAGACCC 180 GCCGGGAGGC AGAGGACCTG CAG~rGGGGC AGGTGGAGCT GGGCGGGGGC Cu~ GCAG 240 GCAGCCTGCA GC-~l~GCC CTGGAGGGGT CC~-~GCAGAA GCGTGGCATT GTGGAACAAT 300 GCAGTACTAG CAl.lGC~CC CTCTACCAGC TGGAGAACTA CAGCAACTAG ATCTAGCCC 359 ~2) INFORMATION FOR SEQ ID NO:4 (i) ~Q~N~ CHARACTERISTICS:
(A) LENGTH: ll0 amino acids (B) TYPE: amino ac$d (C) STRAN~ -CS:
(D) TOPOLOGY: linear (xi~ ~Uu~ DESCRIPTION SEQ ID NO:4 Met Ala Leu Trp Met Arg Leu Leu Pro Leu Leu Ala Leu Leu Ala Leu l 5 l0 15 Trp Gly Pro Asp Pro Ala Ala Ala Phe Val Asn Gln His Leu Cys Gly Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly Phe W O 97~6321 PCTrUSg7/00761 Phe Tyr Thr Pro Lys Thr Arg Arg Glu Al~ Glu Asp Leu Gln Val Gly Gln Val Glu Leu Gly Gly Gly Pro Gly Ala Gly Ser Leu Gln Pro Leu Ala Leu Glu Gly Se~ Leu Gln Lys Arg Gly Ile Val Glu Gln Cys Ser Thr Ser Ile Cys Ser Leu Tyr Gln Leu Glu Asn Tyr Ser Asn l00 105 ll0 ~2) INFORMATION FOR SEQ ID NO:5:
~UU~N~ CHARACTERISTICS:
~A) LENGTH: 867 base pairs (B) TYPE: nucleic acid (C~ STRA~ N~:Cs: single ~D~ TOPOLOGY: linear ~xi~ ~UU~N~ DESCRIPTION: SEQ ID NO:5:
GGATCCATGA ACAGTGAGGA GCAGTACTAC GCGGCC~C~C AGCTCTACAA GGACCC~T~C 60 GCATTCCAGA GGGGCCC~1 GCCAGAGTTC AGCGCTAACC CCC--1GC~1G CCTGTACATG 120 GGCCGCCAGC CCCCACCTCC GCCGCrACCC CAGTTTACAA GCTCGCTGGG ATCACTGGAG 180 CAGGGAAGTC CTCCGGACAT ~1CCCCATAC GAAG~GCCCC CG~CGCClC CGACGACCCG 240 G~l~GCGC1C ACCTCCACCA CCACCTTCCA GCTCAGCTCG GGCTCGCCCA TCCACCTCCC 300 GGAC~-111CC CGAATGGA-A-c CGAGCCTGGG GGC~1~GAAG AGCCCAACCG CGTCCAGCTC 360 ~ CCC~, GGATGAAATC CACCAAAGCT CACGC~GGA AAGGc~TG GGCAGGAGGT 420 ~-~G~GGAGC TGGAGAAGGA ATTCTTATTT AACAAATACA 1~-CCCGGCC CCGCCGGG~G 540 GAGCTGGCAG TGATGTTGAA CTTr-ACCg~G AGACACATCA AAAl~la~-l CCAAAACCGT 600 CGCATGAAGT GGAAAAAAGA GGAAGATAAG AAACGTAGTA GCGGGACCCC GA~l ~ GGGC 660 G~aGGGGCG AA~AC-CCGGA GCAAGATTGT GcG~AccT CGGGCGAGGA GC~ CA 720 GTGCC~CCGC TGCCACCTCC C~GAGGTGCC ~GCCCCCAG GC~rCCCAGC TGCAGTCCGG 780 WO 97126321 PCTnUs97/00761 GA~GGC~rAC TGC~-1GGGG CGTTAC,CGTG TCGCCACAGC CCTCCAGCAT CGCGCCACTG 840 CGACCGCAGG AACCCCGGTG A~GATCT 867 (2) INFORMATION FOR SEQ ID NO:6:
(i~ SEQUENCE CHARACTERISTICS:
(A) LENGTH: 284 amino acids (B) TYPE: amino acid (C) STRANnRnNF~S:
(D) TOPOLOGY: linear (xi~ ~E~U~N~ DESCRIPTION: SEQ ID NO:6:
Met Asn Ser Glu Glu Gln Tyr Tyr Ala Ala Thr Gln Leu ~yr Lys Asp l 5 l0 15 Pro Cys Ala Phe Gln Arg Gly Pro Val Pro Glu Phe Ser Ala Asn Pro Pro Ala Cys Leu Tyr Met Gly Arg Gln Pro Pro Pro Pro Pro Pro Pro Gln Phe Thr Ser Ser Leu Gly Ser Leu Glu Gln Gly Ser Pro Pro Asp Ile Ser Pro Tyr Glu Val Pro Pro Leu Ala Ser Asp Asp Pro Ala Gly Ala His Leu His His His Leu Pro Ala Gln Leu Gly Leu Ala His Pro Pro Pro Gly Pro Phe Pro Asn Gly Thr Glu Pro Gly Gly Leu Glu Glu l00 105 ll0 Pro Asn Arg Val Gln Leu Pro Phe Pro Trp Met Lys Ser Thr Lys Ala llS 120 125 His Ala Trp Lys Gly Gln Trp Ala Gly Gly Ala Tyr Thr Ala Glu Pro Glu Glu Asn Lys Arg Thr Arg Thr Ala Tyr Thr Arg Ala Gln Leu ~eu 145 lS0 155 160 Glu Leu Glu Lys Glu Phe Leu Phe Asn Lys Tyr Ile Ser Arg Pro Arg Arg Val Glu Leu Ala Val Met Leu Asn Leu Thr Glu Arg ~is Ile Lys lB0 185 l90 W O 97/26321 PCT~US97100761 Ile Trp Phe Gln Asn Ar~ Arg Met :Lys Trp Lys Lys Glu Glu Asp Lys Lys Ar~ Ser Ser Gly Thr Pro Ser Gly Gly Gly Gly Gly Glu Glu Pro Glu Gln Asp Cys Ala Val Thr Ser Gly Glu Glu Leu Leu Ala Val Pro Pro Leu Pro Pro Pro Gly Gly Ala Val Pro Pro Gly Val Pro Ala Ala Val Arg Glu Gly Leu Leu Pro Ser Gly Leu Ser Val Ser Pro Gln Pro Ser Ser Ile Ala Prc Leu Arg Pro Gln Glu Prc Arg (2) INFORMATION FOR SEQ ID NO:7:
Q~NL~: CHARACTERISTICS:
(A) LENGTH: 677 base pairs (B1 TYPE: nucleic acid ( C ~ STRA~nNFSS: single (D) TOPOLOGY: }inear ~xi) S~Qu~N~ DESCRIPTION: SEQ ID NO:7:
AAGCTTCAGG CTGCCAGCAC ACTATCTGTT A~ 1GCCA CTGCCCACTG AAAGGGATCT 60 TGAGACATGA GGTGCATCTC CAGGCTGCCA GL~ CC TCA1CLL~-lC GGTGGCACTC 120 TGCAACACAG CCACATGTGC CACACAACGT CTGGr~CT l~ CG cTcr~rAar 240 AACL~-~GlC CAGLCL-.CCC ACCAACCAAT GTGGGATCCA ATACATATGG GAAGAGGAAT 300 GTGGCAGAGG ATCCAAATAG GGAATCCCTG GA~ Ac lCClv~AAAG TCAATGTACT 360 CCCGTATCTC TTATTACTTC CTGTGTAAAT GCTCTGATGA ~ AAT AATGTAACAG 420 l~C~CAAC C L GCL ~ G 1~C 1 1~ a ~ 1 TGTAAATTCT TATTCTAAGA C~ L~ L ~1AA 4 8 0 ACTGAGTGTT GATAAAGGTC AGGGTGAATA CL----~--AAT CACAACATGT l~-.~GL.~l 540 ACATCGATAT CGTAGGAACA CTTAAAATTT ~lLll~ AC CTTGTAACTC TATGACTCAA 600 W O 97/26321 PCTnUS97/00761 (2) INFORMATION FOR SEQ ID NO:8:
(i) ~yu~Nc~ CHARACTER:[STICS:
~A) LENGTH: 93 am:ino acids (B) TYPE: amino acid (C) STRA~ S:
( D ) TOPOLOGY: linlaar (xi) ~:Qu~ DESCRIPTION: SEQ ID NO:8:
Met Arg Cys Ile Se:r Ar!~ Leu Pro Ala Val Leu Leu Ile Leu Ser Val l 5 l0 l5 Ala Leu Gly His Leu Ar~ Ala Thr Pro Val Gly Ser Gly Thr Asn Pro Gln Val Asp Lys Ar~ Lys Cys Asn Thr Ala Thr Cys Ala Thr Gln Arg Leu Ala Asn Phe Leu Val Arg Ser Ser Asn Asn Leu Gly Pro Val Leu Pro Pro Thr Asn Val Gly Ser Asn Thr Tyr Gly Lys Arg Asn Val Ala Glu Asp Pro Asn Arg Glu Ser Leu Asp Phe Leu Leu Leu (2) lN~OK~TION FOR SEQ ID NO:9:
~QD~N~E CHARACTERISTICS:
(A) LENGTH: 2086 b~e pairs (B) TYPE: nucleic acid (C) STRP~ Y:~S: single (D) TOPO~OGY: line~r ~xi) ~Qu~C~ DESCRIPTION: SEQ ID NO:9:
GGATCCAAGG CCCAACTCCC CGAACCACTC AGG~lC~l GGACAGCTCA CCTAGCTGCA 60 ATGGCTACAG GTAAGCGCCC CTAAAATCCC l~lGGCACAA l~G~ GA GGGGAGAGGC 120 A~-C~CCTGT AGATGGGACG GGGGCACTAA CCCTCA¢GGT ~ GG~ GAATGTGAGT l80 CA 0224643l l998-08-l2 ATCGCCATCT AAGCCCAGTA ~ GCCAAT CTCAGAAAGC TCCTGGCTCC CTGGAGGATG 240 rA~ArAr~AAA AACAAACAGC TCCTGGAGCA GGGAGAGTGT TGGC~ G ~l- CCGGCl 300 CC~ G CC~ 16~ ~-lCCCC~GG ~lCCC6GACG TCC~lG~-lCC lGGC~ GG 360 C~-l~ GC ClGCC- GGC TTCAAGAGGG CA~lGC~i-lC CCAACCATTC CCTTATCCAG 420 G~ l-aAC AACGCTATGC TCCGCGCCCA lC~l~-lGCAC CAG~l~GCCl TTGACACCTA 480 CCAGGAGTTT GTAAGCTCTT GGGGAATGGG TGCGCATCAG GGGlGGCAGG AAGGGGTGAC 540 l..CCCCCGC TGGAAATAAG AGGAGGAGAC TAAGGAGCTC AGG~ll~llC CCGACCGCGA 600 AAATGCAGGC AGATGAGCAC ACGCTGAGCT AG61iCCCAG AAAAGTAAAA TGGGAGCAGG 660 TCTCAGCTCA GAC~Ll~lG GGCG~lC~-l CTCCTAGGAA GAAGCCTATA TCCCAAAGGA 720 ACAGAAGTAT TCA~ ~C AGAACCCCCA GACC~CC~-lC ~ CAG AGTCTATTCC 780 GACACCCTCC AACAGGGAGG AAAr~AA~ GA~AATCCGTG AGT6GATGCC ll~CCCAG 840 GCGGGGATGG GGGAGACCTG TAGTCAGAGC CCCCGGGCAG CACAGCCAAT GCCCGlC~l 900 GCCC~-lGCAG AACCTAGAGC lG~l.'CGCA'r ~lCCC~GClG CTCATCCAGT C~l~G~aGA 960 GCCC-~lGCAG ~cclcAGGA Gl~l~ CGC CAACAGCCTG GTGTACGGCG CCTCTGACAG 1020 CAACGTCTAT GAC~-~CC~AA A6GACCTAGA GGAAGGCATC CAAACGCTGA TG6GGGTGAG 1080 ~GGCGC~A GGG~lCCCCA ATCCTGGAGC CCCACTGACT TTGAGAGACT GTGTTAGAGA 1140 AACACTGGCT GCCC~ ll TAGCAGTCAG GCCCTGACCC AAGAGAACTC ACCTTATTCT 1200 TCA ~ ;CCC TCGTGAATCC TCCAGGCCTT TCTCTACACT GAAG6GGAGG GAG&AAAATG 1260 AATGAATGAG AAAG~r~r-GG AACAGTACCC AAGCG~ 6G C~ C~ ~A 1320 GCAAGTTCGA CACAAACTCA CACAACGATG ACGCACTACT CAAGAACTAC GGG~lG~-l 1440 A~GC~lCAG 6AAGGACATG GACAA6GTCG AGACATTCCT GCGCATCGTG CAGTGCCGCT 1500 ~GAGGG CAG~-~GC TTCTAGCTGC CCGGC~GCA lCC~ ~GAC CCC,;CCCAG 1560 ~GC-~ ~.~CT GGCC~l~GAA G~CCACTC CAGTGCCCAC CAGC~ C CTAATAAAAT 1620 TAAGTTGCAT CAl~..~,., GACTAGGTGT CCl..,ATAA TATTATGGGG TGGAGGGGGG 1680 TGGTATGGAG CAAGGGGCCC AAGTTGGGAA GACAACCTGT AGGGCClGCG GGGTCTATTC 1740 GGGAACCAAG CTGGAGTGCA GT~C~ T ~ll~G~.AC TGCAATCTCC GC~'GCl~GG 1800 TTCAAGCGAT l~lC~ GC~ CAGCCTCCCG AGllb~lGGG ATTCCAGGCA TGCATGACCA 1860 GGCTCAGCTA A'~ l~lAGA GACGGG~l-l. CACCATATTG GCCAGGCTGG l920 TCTCCAACTC CTAATCTCAG GTGATC'rACC CAC~ilGGCC TCCCAAATTG CTGGGATTAC 1980 AGGCGTGAAC CACla~ -'CC 11CCC~G'~ ~C TTCTGATTTT AAAATAACTA TACCAGCAGG 2040 AGGACGTCCA GACACAGCAT AGGCTA_CTG CCATGGCCCA ACCGGT 2086 (2) INFORMATION FOR SEQ ID NO l0 ~UU~N-~ CHARACTERISTICS
(A) LENGTH 2l7 amino acids (B) TYPE amlno acid ~ C ) STR~ h ~ :-c s (D) TOPOLOGY linear (xi) S~UU~N~ DESCRIPTION SEQ ID NO l0 Met Ala Thr Gly Ser Arg Thr Ser Leu Leu Leu Ala Phe Gly Leu Leu 1 5 l0 15 Cys Leu Pro Trp Leu Gln Glu Gly Ser Ala Phe Pro Thr Ile Pro Leu Ser Arg Leu Phe Asp Asn Ala Me~ Leu Arg Ala Hls Arg Leu His Gln g0 45 Leu Ala Phe Asp Thr Tyr Gln Glu Phe Glu Glu Ala Tyr Ile Pro Lys W O 97/26321 PCT~US97/00761 Glu Gln Lys Tyr Ser Phe Leu Gln Asn Pro Gln Thr Ser Leu Cys Phe Ser Glu Ser Ile Pro Thr Pro Ser Asn Arg Glu Glu Thr Gln Gln Lys Ser Asn Leu Glu Leu Leu Arg Ile Ser Leu Leu Leu Ile Gln Ser Trp l00 105 ll0 Leu Glu Pro Val Gln Phe Leu Ar~ Ser Val Phe Ala Asn Ser Leu Val Tyr Gly Ala Ser Asp Ser Asn Val Tyr Asp Leu Leu Lys Asp Leu Glu Glu Gly Ile Gln Thr Leu Met Gly Arg Leu Glu Asp Gly Ser Pro Arg lg5 150 155 160 Thr Gly Gln Ile Phe Lys Gln Thr Tyr Ser Lys Phe Asp Thr Asn Ser His Asn Asp Asp Ala Leu Leu Lys Asn Tyr Gly Leu Leu Tyr Cys Phe 180 185 l90 Arg Lys Asp Met Asp Lys Val Glu Thr Phe Leu Arg Ile Val Gln Cys Arg Ser Val Glu Gly Ser Cys Gly Phe 12~ INFORMATION FOR SEQ ID NO ll (i) S~yu~NCE CHARACTERISTICS:
~A) LENGTH 235 base pairs (B) TYPE: nucleic acid ( C ) sTRA~n~nN~s: single (D) TOPOLOGY linear ~xi) S~YU~N~E DESCRIPTION: SEQ ID NO ll:
CCGGATCCAG GTCGACGCCG GCCAAGACAG r~ TTGACCTATT GGG~ C 60 GCGAGTGTGA GAGGGAAGCG CCGCGGCCTG TATTTCTAGA CCl~CC~ C GC~G~,CG 120 lGGCGC~-G TGACCCCGGG CCCCl~G~C CTGCAAGTCG AAAll~CG~l GlG~lC~. 180 GCTACGGCCT ~GG~G~Ac lGC~ C TGCCCAACTG GCTGGCAAGA TCTCG 235 W O 97~i6321 PCTAUS97/00761 ~2) INFORMATION FOR SEQ ID NO:12:
(i) ~u~' CHARACTERISTICS:
(A) LENGTH: 441 ~ase pairs (Ei~ TYPE: nucleic acid (C) STRA~n~n~S: single (D~ TOPOLOGY: linear (xi) ~ QU~ DESCRIPTICN: SEQ ID NO:12:
GGTACCGGGC CCCCC~'CGA GGTCCTGAGC TAAGAATCCA GCTATCAATA GAAACTATGA 60 AACAGTTCCA GGGACAAAGA TACCAGGTCC CCAACAACTG CAA~-~ G GGAAATGAGG 120 TGGAAAATGC TCAGCCAAGG AAAAAGAGGG CCTTACCCTC TCTGGGACAA TGA~l~lGC-l 180 GTGAACTGCT TCATCACGGC AlClGCiCCCC TTGTTAATAA TCTAATTACC CTAGGTCTAA 240 GTAGAGTTGT TGACGTCCAA TGAGCG~- L l 1 CTGCAGACTT AGCACTAGGC AA~ .lGG 300 AAATTACAGC TTCAGCCCCT CTCGCCATCT GCCTACCTAC CC~~ AGA &CCCTTAATG 360 GGCCAAACGG CAAAGTCCAG GGGGCAGAGA GGAGGl~ TGGACTATAA AGCTAGTGGA 420 (2) INFOR~ATION FOR SE2 ID NO:13:
~QU~N~ CHARACTERISTICS:
(A) LENGTH: 1252 ba~e pairs (B) TYPE: nucleic acid (C) STRAr~ s: single (D) TOPOLOGY: lin~!ar [xi~ ~Uu~ DESCRIPTIC)N: SEQ ID NO:13:
GGGTATCTTT GCCCCG~TGG A~ ~ CC TAGCCAGGGG CTCCCAGTCC CCAGGC~-~iG 60 GGTAGAAGTT GGACTCTATA GTCACCTAAG GGCCTATGTT GCA~lC~l~iG TCTCAGGCAC 120 GCGGC~-l~CA GGA~-~CTT AAAAGAAGAG AAACTGCACA CAATGGCTAG GTCACCGGCG 180 TTAAAGCTAA GCAAACCCAG CGCTAC'TCCT GGGCAGCAAC TGCAAAGCGT ~ lCAGGT 240 CCCTCACCTG TAGAATCAGA GCGGT~GTCG C~lc~GCATG TCTGAGTTCT TACATGTCGT 300 AATGTACAAA CACGATTTCC CCTCAATCAC CGCCCG~AAC AGTACCTCCA A~CCCAGA 360 CCCGGATGCC CCAAGAGCCA GAGTAC;GGTG GGAAAATCGG GACAGGCCCC CAAATTCCAC 420 lCGGGGGC~:l TGAGCTCTTA CAlG~ CA CGGGGGCAGG TAGTTTGGGT TTAGCAATGT 480 GAACTCTGAC AATTTGGGAT GT~GAGCTGG TGGGCCATCG TGGGACGCCA AGCATCATCC 540 TTAGAGTTTG GAlC~ AG GGCAGGCAGG CACAGGGACC CAGTGCGAGA TCAGTGAAGC 600 CGCCCAGTTT CGGL-llCCGC lClllllCCA CGCCCACTTG CGlG~ 'ClC CAACAGTGTG 660 GATGGGAGGG GlGGGGGACG AGCCCTAATC TCCGAGGAAG GG~ il~GCC CCGl-lC~,L~i-. 720 TCTCCAGTTT GTGGCGTCCT GGATCTGTCC lClGt,1CCCC TCCAGATCGT GTCCCACACC 780 CACCC~illCA GGCATGGCAC lGlGCCGCCA CGCGTGACCG lGCG~lCCll ACGl ~GGGA 840 CGTGCAGGGT GCTGCCTCCT ~l'CGGlGCG GGAGGGAGCG GCCCii~-lL~C TC~ Ll~:lG 900 GC-lGGGAAGC CCCAGCCATT GCG~;lGCAGA GGAGACTTGC AGCCAATGGG GACTGAGGAA 960 Gl-3GGC~GGC lGGCG~l-l-/l CACC~lC~-CG GGGACCGGAG CTCCGAGGTC TGGAGAGCGC 1020 AGG~ r-~CGC CCGCCCCGCC CGGGGACTGA GGGGGAGGAG CGAAGGGAGG AGGAGGTGGA 1080 Gl~_-CCGATC TGCCGCTGGA GGACCACTGC TCACCAGGCT ACTGAGGAGC CA~-l~GCCCC 1140 ACAC'-l~ TTCCGCATCC CCCACCGTCA GCATGATCGC CGCGCAACqA ClGGC~ATT 1200 (2) INFORMATION FOR SEQ ID NO:14:
( i ) SE~QU~ ; CHARACTERISTICS:
(A) LEWGTH: 410 base pairs IB) TYPE: nucleic acid (C) STRl~Nl~:llN~ S single (D) TOPOLOGY: linear (xi~ Qu~ C-~: DESCRIPTION: SEQ ID NO:14:
GAAl-Cl~;ll GGG~:lCGCGG TTGACCACAA A~:l~l-lCGCG ~lCCAG TACl~ GGA 60 TCGGAAACCC ~,lCG~jC~;lCC GAACGGTACT CCGCCACCGA GGGACCTGAG CGAGTCCGCA 120 TCGACCGGAT CGGAAAACCT CTCGACTGTT GGGGTGAGTA ~_lCCt~ lCA AA~GCGGG-'D. 180 TGACTTCTGC GCTAAGATTG TCA~jl l'lCCA AAAA~ G~ GGATTTGATA TTCACCTGGC 240 CC~'G~C~AT GC~ AGG GIGGCCG~il CCAl~-l~lC AGAAAAGACA Al~ l 300 CA 0224643l l998-08-l2 W O 97/26321 PCT~US97/00761 TGTCAAGCTT GAG~-~lGGC AGGCTTGAGA TCTGGCCATA CACTTGAGTG ACAATGACAT 360 CCA~.llGCC T1~ CAC ACAGGTGTCC ACTCCCAGGT CCAACTGCAG 410 ~2) INFORMATION FOR SEO ID NO:lS:
~i) S~Q~N~ CHARACTERISTICS:
~A) LENGTH 175 base pairs ~B) TYPE: nucleic acid ~C) STRA~ S single (D) TOPOLOGY linear (xi~ S~UU~N~ DESCRIPTION SEQ ID NO:15 GAlCCul~CA TCAGGCCATC TGGCCC~-l-lG TTAATAATCG ACTGACCCTA GGTCTAAGAT 60 CC~lCATCA GGCCATCTGG C'CC~ l'-'A ATAATCGACT GACCCTAGGT CTAAGATCCC 120 TTCATCAGGC CAl~GGCCC ~l~ AATA ATCGACTGAC CCTAGGTCTA AGATC 175 (2) INFORXATION FOR SEQ ID NO:16:
~QU~N~ CHARACTERISTICS:
(A) LENGTH 31 bae;e pairs ~B) TYPE: nucleic acid ~C) STRA~ S single ~D) TOPOLOGY linear ~xi) S~Qu~ DESCRIPTIC)N: SEQ ID NO:16 'C'C'LlCC. AGCACCGCCC GGAACAGTAC C 31 ~2) INFORMATION FOR SEQ ID No :17:
(i) ~Qu~N~E CHARACTERISTICS:
(A) LENGTH: 30 base pairs (B) TYPE: nucleic acid (C) ST}U~r ~ S: single (D) TOPOLOGY linear (xi) ~QU~N~ DESCRIPTION: SEQ ID NO:17:
.~GCGC~-~C GAGCATGCTG ACGGT~GGGG 30 ~2) lN~RIIATION FOR SEQ ID NO:18:

W O 97/26321 PCT~US97/00761 ~UU~NC~' CXARACTERISTICS:
(A~ LENGTH: 31 base pairs (B~ TYPE: nucleic acid (C~ STR~N~ NI-:C S: single (DJ TOPOLOGY: linear (xi) ~:Qu~c~: DESCRIPTION: SEQ ID NO:18:
GTTGGACTCG AGAGTCACCT AAGGGCCTAT G 3l (2) INFORMATION FOR SEQ ~D NO:19:
(i) S~yU~NC~: CHARACTERISTICS:
(A) LENGTH: 26 base pairs (B) TYPE: nucleic acid (C) STR~Nn~nNF~S: single (D) TOPOLOGY: line~r txi) ~U~N~ DESCRIPTION: SEQ ID NO:l9~

(2) INFORNATION FOR SEQ ID NO:20:
:QU~N~ CHARACTERISTICS:
tA) LENGTH: 24 base pairs (B) TYPE: nucleic acid tc~ STR~nEnNFCS single (D) TOPOLOGY: linear ~xi) ~uu~ DESCRIPTION: SEQ ID NO:20:
A~.CGCC.~"1' GCATGTCTGA GTTC 24 t2) lN~-Ofi~ATION FOR SEQ ID NO:2l:
Q~kN~ CHARACTERISTICS:
(A) LENGTH: 24 base pairs (B) TYPE: nucleic acid ( C ) STRAr -- ~r~ -ss: single (D) TOPOLOGY: linear (xi) ~Q~ NC~ DESCRIPTION: 5EQ ID NO:2l:

W O 97126321 rCTAUS97/00761 (2) INFORMATION FOR SEO ID NO:22:
~i) SE4u~C~ CHARACTERISTICS:
~A) LENGTH: 2l base pairs (B) TYPE: nucleic acid (C) STRANn~n~S: single (D) TOPOLOGY: linear (Xi ) ~QU~N~: DESCRIPTION: SEQ ID NO:22:
TCCCCAGGCG TGGGGTAGAA C; 2l ~2) INFORMATION FOR SEQ ID NO:23:
CHARACTERI.STICS:
(A) LENGTH: 24 bac;e pairs (B) TYPE: nucleic acid (C) STRAN~ S: single (D) TOPOLOGY: lin~!ar (xi) ~Qu~NC;~: DESCRIPTION: SEQ ID NO:23:
CAACCG~lGG GACATTTGAG TTCC 24 (2) INFORMATION FOR SEQ ID No: 24:
u~ CHARAC'TER:iSTICS:
(A) LENGTH: 24 ba~;e pairs ~B) TYPE: nucleic acid (C) ST~A~ S: single ~D) TOPOLOGY linear (xi) SEQu~ DESCRIPTION: SEQ ID NO:24:

~2) lNrO~TION FOR SEQ ID NO:25:
li) ~Uu~N~ CHARACTERISTICS:
(A) LENGTH: 24 base pairs (B) TYPE: nucleic acid ~C) STR~ ;SS: single (D) TOPOLOGY: linear (xi) ~:Uu~ DESCRIPTION: SEQ ID NO:25:

W O 97126321 PCTnUS97100761 (Z) INFORMATION FOR SE~ ID NO:26:
UU~N-~ CHARACTERISTICS:
(A) LENGTH: 24 base pairs (B) TYPE: nucleic acid ( C ) STR~Nl ]~ r~l ~ C s: 5 ingle (D) TOPOLOGY: linear ~xi) ~yu~ DESCRIPTION: SEQ ID NO:26:

(2) INFORMATION FOR SEQ ID NO:27:
:yu~-E CHARACTERISTICS:
(A) LENGTH: 25 base pairs (B) TYPE: nucleic acid ~C) STR~N~ N~:~S: single (D~ TOPOLOGY: linear (xi) ~Q~NC~: DESCRIPTION: SEQ ID NO:27:

(2) INFORMATION FOR SEQ ID NO:28:
(i) ~Q~N~ CHARACTERISTICS:
(A) LENGTH: 22 base pairs (B) TYPE: nucleic acid (C) STR~ N~:~S: single (D) TOPOLOGY: linear ~xi) ~UU~N~ DESCR:IPTION: SEQ ID NO:28:

(2~ lN~O~ ~TION FOR SEQ ID NO:29:
CHARACTERISTICS:
~A) LENGTH: 5~ base pairs ~B) TYPE: nucleic acid ~C) STR~r~ N~:~S: single ~D) TOPOLOGY: linear Ixi) ~Uu~N~ DESCRIPTION: SEQ ID NO:29:

W O 97126321 PCTnUS97/00761 GA1CCu-lCA TCAGGCCATC TGGCCCCTTG TTAATAATCG ACTGACCCTA GGTCTAA 57 ~2) INFORMATION FOR SEQ ID NO:30:
Q~ CHARACTERISTICS:
~A) LENGTH: 57 base pairs ~B) TYPE: nucleic acid ~C) STRPNnFnN~-CS: single ~D) TOPOLOGY: llnear ~xi) ~Qu~NC~ DESCRIPTI~N: SEQ ID NO:30:

~2) INFORMATION FOR SEQ ID NO:3l:
:QD~u~ CHARACTERISTICS:
~A) LENGTH: 27 base pairs (B) TYPE: nucleic acid (C) STRAr~ SS: single (D) TOPOLOGY: linear ~xi) ~:uu~u~: DESCRIPTION: SEQ ID NO:31:

(2) INFOR~ATION FOR SE~ ID NO:32:
(i) ~u~C~' CHAR~CTER.ISTICS:
(A) LENGTH: 28 base pairs (B) TYPE: nucleic acid (C) ST~N~ C~S: single (D) TOPOLOGY: lin,ear (xi) ~u~C~: DESCRIPTION: SEQ ID NO:32:
GGGCAA~CTA GGTACTGGAC CTTCI~ATC 28 (2) INFORMATION FOR SEQ ID NO:33:
(i) S~UU~NC~ CHARACTERISTICS:
(A) LENGTH: 25 b2,se pairs (B) TYPE: nucleic acid ( C ) STl~ N l ~ N l~: c s single (D) TOPOLOGY: linear W O g7/26321 PCTrUS97/00761 ~xi) ~:Uu~:N~ DESCRIPTION: SEQ ID NO:33:
GGGTCTAGAG GAC~1~11CC CACCG 25 ~2) INFORMATION FOR SEQ ID NO:34:
U~N~ CHARACTERISTICS:
(A) LENGTH: 28 base pairs (B) TYPE: nucleic acid (C) STR~N~nNF~S: single (D) TOPOLOGY: linear (xi) SEQu~ DESCRIPTION: SEQ ID NO:34:

(2) INFORMATION FOR SEQ ID NO:35:
(i) ~UU~N~ CHARACTERISTICS:
(A) LENGTH: 35 base pairs (B) TYPE: nucleic acid (C) STR~N~ N~SS: sin~le (D) TOPOLOGY: linear ~xi) ~:yu~N.~ DESCRIPTION: SEQ ID NO:35:

(2) INFORMATION FOR SEQ ID NO:36:
U~h CHARACTERISTICS:
(A) LE~GTH: 22 base pairs (B) TYPE: nucleic acid (C) STR~N~ N~S: single (D) TOPOLOGY: linear ~xi) S~u~E DESCRIPTION: SEQ ID NO:36:
CCAC~1~1~. ACAC~lC~1C TC 22 ~2) lN~O~IATION FOR SEQ ID NO:37:
(i) ~Qu~N~E CHARACTERISTICS:
(A) LENGTH: 22 base pairs (B) TYPE: nucleic acid W O 97126321 PCT~US97100761 ~C~ STRANDEDNESS: single ~D) TOPOLOGY: linear (xi) ~QU~L~ DESCRIPTIt)N: SEQ ID NO:37:
GTAATCCAGG ~~lC~1GACT GC 22 (2) INFORMATION FOR SEQ ID NO:38:
U~NC~ CHARACTER:[STICS:
(A~ LENGTH: 24 base pairs (B~ TYPE: nucleic acid ~C~ STRA~ S: single (D~ TOPOLOGY: lin~ar ~xi~ SEQUENCE DESCRIPTION: SEQ ID NO:38:
CCGGGGATCC ~lulGC~ATG GCCC 24 ~2) lN~O.~_'.TION FOR SEQ ID ]NO:39:
u~ C~ARACTERrSTICS:
(A) LENGTH: 69 base pairs (B~ TYPE: nucleic acid (C~ ST~A~ S: single ~D~ TOPOLOGY: linear (xi~ ~Uu~ DESCRtPTION: SEQ ID NO:39:
GGGCTAGATC TA~l~C~ A~lL~lCCAG CTGGTAGAGG GAGCAGATGC TAGTACTGCA 60 ,~.lC~AC 69 ~2) INFORMATION FOR SEQ ID NO:40:
yukN~: CHARACTERISTICS:
(A) LENGTH: 27 base pairs ~B) TYPE: nucleic acid ~C) STR~ cs: single (D) TOPOLOGY: linear (xi) ~ u~C~: DESCRIPTION: SEQ ID NO:40:

~2) lNPO~.TION FOR SEQ ID NO:4l:

W O 97126321 PCTrUS97100761 (i) ~Q~ CHARACTERISTICS:
(A) LENGTH: 24 base pairs (B) TYPE: nucleic acid (c) STRAN~r;~N~SS: single ~D) TOPOLOGY: linear (xi) ~:U~N~ DESCRIPTION: SEQ ID NO:41:

(2) INFOR~ATION FOR SEQ ID NO:42:
Quh~ CHARACTERISTICS:
(A) LENGTH: 24 base pairs (B) TYPE: nucleic acid (C) STRANn~nN~CS sing}e (D) TOPOLOGY: linear ~xi) ~QU~N~: DESCRIPTION: SEQ ID NO:42:
AGATCTTCAC CGGG~l-CL GCGG 24 t2~ INFORNATION FOR SEQ ID NO:43:
r;4~r~0~ CHARACTERISTICS:
(A) LENGTH: 24 base pairs tB) TYPE: nucleic acid (C) STRA~ lrSS: single (D) TOPOLOGY: linear ~xi) ~yur~: DESCRIPTION: SEQ ID NO:43:

~2) l~rO~ ~TION FOR SEQ ID NO:44:
~i) Sbyur~ CHARACTERISTICS:
tA) LENGTH: 23 base pairs ~B) TYPE: nucleic acid (C) STRA~ tCS: single ~D) TOPOLOGY: linear ~xi) ~ ~N~: DESCRIPTION: SEQ ID NO:44:

W O 97~6321 PCT~US97100761 12~ INFO~MATION FOR SEQ ID NO:4~:
(i) S~:uu~u~ CHARACTERISTICS:
(A) LENGTH: 22 ~ase pairs (B) TYPE: nucleic acid (C) STRANDEDNESS: single (D) TOPOLOGY: linear (xi) ~UU4N~E DESCRIPTION: SEQ ID NO:45:

12) INFORMATION FOR SEQ ID NO:46:
;uu~ CHARACTER'rSTICS:
tA) LENGTH: 23 base pairs (B) TYPE: nucleic acid ~C) STR~ N~ S: single (D) TOPOLOGY: linear ~xi) ~:Q~N~ DESCRIPTION: SEQ ID NO:46:

~2) INFORMATION FOR SEQ ID NO:47:
(i) SEQUENCE CHARACTERISTICS:
(A) LENGTH: 23 base pairs (B) TYPE: nucleic acid (C) STR~ MF~S: single (D) TOPOLOGY: linear (xi) ~yu~N~: DESCRIPTION: SEQ ID NO:47:

12) INFORMATION FOR SEQ ID NO:48:
UU~NC~ CHARACTER.ISTICS:
(A) LENGTH: 22 base pairs ~B) m E: nucleic acid ~C) STRAr~ b:cs: single ID~ TOPOLOGY: lin.ear ~xi) S~u~C~ DESCRIPTI:ON: SEQ ID NO:48:

CA 0224643l l998-08-l2 GGAGATCTGT TAG-l--ll L~1 TC 22 ~2) INFORMATION FOR SEQ ID NO:49:
(i) ~u~ CHARACTERISTICS:
(A) LENGTH: 22 base pairs ~B) TYPE: nucleic acid ~C) ST}~N~ NI~:CS single (D) TOPOLOGY: linear (xi) S~Qu~N~: DESCRIPTION: SEQ ID NO:49:
CCGGATCCAT GAGCTTCAAT AC' 22 (2) INFORMATION FOR SEQ ID NO:50:
(i) S~UU~N~: CHARACTERISTICS:
(A) LENGTH: 23 base pairs tB) TYPE: nucleic acid ~C) ST~A~ :Cs: single (D) TOPOLOGY: linear ~xi) S~YU~N~ DESCRIPTION: SEQ ID NO:50:

~2) INFORMATION FOR SEQ ID NO:51:
(i) ~U~kN~: CHARACTERISTICS:
(A) LENGTH: 37 amino acids (B) TYPE: amino acid (C) ST~ANI~ Nl.:C~S
tD) TOPOLOGY: linear txi) ~UU~N~ DESCRIPTION: SEQ ID NO:51:
Lys Cys Asn Thr Ala Thr Cys Ala Thr Gln Arg Leu Ala Asn Phe Leu l 5 l0 15 Val Arg Ser Ser Asn Asn Leu. Gly Pro Val Leu Pro Pro Thr Asn Val Gly Ser Asn Thr Tyr (2) lN~O~IATION FOR SEQ ID NO:52:

W O 97126321 PCT~US97/00761 ( i ) ~k4U~NG~ CHARACTERISTICS:
(A) LENGTH: 491 ba.se pairs (B) TYPE: nucleic acid ( C ) STRANV~VNI~ Ss: single (D) TOPOLOGY: line!ar (xi~ uu~ DESCRIPTION: SEQ ID NO:52:
~,.~lGATA TTGCTGACAT TGAAAC'ATTA AAAGAAAATT TGAGAAGCAA TGGGCATCCT 60 GAAGCTGCAA GTA~r~ LCA ~lu~lG~ C ~ ATTG AACCATCTGA AAGCTACACC 120 CATTGAAAGT CATCAGGTGG AAAAGC'GGAA ATGCAACACT GCCACATGTG CAACGCAGCG l80 C~lGG~AAAT TTTTTAGTTC ATTCCAGCAA CAAO~Ll~l GCCATTCTCT CATCTACC~A 240 CC.~G~ATCC AATACATATG GCAAG~LGGAA TGCAGTAGAG GTTTTAAAGA GAGAGCCACT 300 GAATTACTTG CCC~-lL~AGA GGACAATGTA ACTCTATAGT TA~ A ~LL~rAGTG 360 A-l.C~l~lA TAATTTAACA ~C~ -l CATCTCCAGT GTGAATATAT G~ ~L~ 420 TCTGATGTTT ~lL~lAGGA CATATACCTT CTCAAAAGAT l~rlllATAT GTAGTACTAA 480 CTAAGG~CCC A 49l (2) INFORMATION FOR SEQ ID NO:53:
EQU~N~: CBARACTERISTICS:
(A) LENGTH: 89 a~LnO acids (B) TYPE: amino acid ~C) STR.~ :lINl-:.'~C
(D) TOPOLOGY: lin~ar ~xi) SEyu~N~ ~F-s~RTpTIoN: SEQ ID NO:53:
Met Gly Ile Leu Lys Leu Gln Val Phe Leu Ile Val Leu Ser Val Ala l 5 l0 15 Leu Asn His Leu Lys Ala Thr Pro Ile Glu Ser His Gln Val Glu Lys Arg Lys Cys Asn Thr Ala Thr Cys Ala Thr Gln Arg Leu Ala Asn Phe Leu Val His Ser Ser As~n Asn Phe Gly Ala Ile Leu Ser Ser Thr Asn W O 97126321 PCT~US97/00761 Val Gly Ser Asn Thr Tyr Gly Lys Arg Asn Ala Val Glu Val Leu Lys Arg Glu Pro Leu Asn Tyr Leu Pro Leu (2) INFORNATION FOR SEQ ID NO:54 ~yu~ CHARACTERISTICS:
(A~ LENGTH 37 amino acids (B) TYPE amino acid (C) STR~N~nN~CS
(D) TOPOLOGY inear (xi) ~Qu~ DESCRIPTION: SEQ ID NO:54 Lys Cys Asn Thr Ala Thr Cys Ala Thr Gl~ Arg Leu Ala Asn Phe Leu l 5 l0 15 Val His Ser Ser Asn Asn Phe Gly Ala Ile Leu Ser Ser Thr Asn Val Gly Ser Asn Thr Tyr (2) INFORNATION FOR SEQ ID NO:55 ~i) S~YU~N~ CHARACTERISTICS:
(A) LENGTH 895 base pairs (B) TYPE nucleic acid (C) STR~NL~ S single (D) TOPOLOGY linear ~xi~ S~UU~N~ DESCRIPTION: SEQ ID NO:55 CCAC~-~.. l ACAC'~--C-~ TCAGCTCAGT CCCACAARGC AGAATAAAAA AATGAAGACC 60 GTTTACATCG l~G~.~ATT G~ AATG CTGGTACAAG GCAGCTGGCA GCA.GCCC'~l 120 CAGGACACGG AGGAGAACGC CAGATCATTC CCAGC-~C'C AGACAGAACC ACTTGAAGAC l80 CCTGATCAGA TAAACGAAGA C~ACGCr~T TCACAGGGCA CATTCACCAG TGACTACAGC 240 AAATACCTAG A~-CCC~CCG TGCTCAAGAT ~ AGT GGTTGATGAA ~CCA~ G 300 ACCAGTGATG TGA~ A CTTGGAGG&C CAGGCAGCAA AGGAATTCAT lG~.~a~ 420 ~.

W O 97~6321 PCTrUS97/00761 GTGAAAGGCC GAGGAGAACG AGA~-llCCGG GAAGAAGTCG CCATAGCTGA GGAACTTGGG 480 CGCAGACATG CTGATGGATC ~l'1'~r~l~AT GAGATGAACA CGATTCTCGA TAAC~ll~CC 540 AGrAr~AGA-cT TCATCAACTG GCTGATTCAA ACCAAGATCA CTGACAAGAA ATAGGAATAT 600 TTCACCATTC ACAACCATCT TCACAACATC l~l~C~AGT CACTTGGGAT GTACATTTGA 660 ~.Ar.~ATATCc GAAGCTATAC l~C~ CAT GCGGACGAAT ACAl-ll~G~l TTAGCGl-L~l 720 GTAACCCAAA G~l~rAAAT GGAATAAAGT TTTTCCAGGG TGTTGATAAA GTAACAACTT 780 TACAGTATGA AAATG~ A TTCTCAAATT ~~ l TTTGAAGTTA CCGCCCTGAG 840 ATTACTTTTC l~rG~'ATAA ATTGTAAATT ATCGCAGTCA CGACACCTGG ATTAC 895 l2) lN~ ATIoN FOR SEQ ID NC~:56:
[i~ ~UU~N~ CHAPACTERI';TICS:
lA) LENGTH: 180 ami.no acids (B) TYPE: amino aci.d (C) STR~N~ S:
tD) TOPOLC~GY: inee~r ~xi) ~Uu~N~t DFcG~TPTION: SEQ ID NO:56:
Met Lys Thr Val Tyr Ile Val Ala Gly Leu Phe Val Met Leu Val Gln l 5 10 15 Gly Ser Trp Gln His Ala Pro Gln Asp Thr Glu Glu Asn Ala Arg Ser Phe Pro Ala Ser G1n Thr Glu Pro Leu Glu Asp Pro Asp Gln Ile Asn Glu Asp Lys Arg His Ser Gln Gly Thr Phe Thr Ser Asp Tyr Ser Lys Tyr Leu Asp Ser Arg Arg Ala Gln Asp Phe Val Gln Trp Leu Met Asn Thr Lys Arg Asn Arg Asn A~n Ile Ala Lys Arg His Asp Glu Phe Glu Arg His Ala Glu Gly Thr Phe Thr Ser Asp Val Ser Ser Tyr Leu Glu 100 105 ll0 G}y Gln Ala Ala Lys Glu Phe Ile Ala Trp Leu Val Lys Gly Arg Gly w o 97n6321 PCT~US97100761 Glu Arg Asp Phe Pro Glu Glu Val Ala Ile Ala Glu Glu Leu Gly Arg Arg His Ala Asp Gly Ser Phe Ser Asp Glu Met Asn Thr Ile Leu Asp Asn Leu Ala Thr Arg Asp Phe Ile Asn Trp Leu Ile Gln Thr Lys Ile Thr Asp Lys Lys (2~ lN~u~.TION FO~ SEQ I~ NO:57:
U~N~ CHARACTERISTICS:
(A) LENGTH: 955 base pairs ~B) TYPE: nucleic acid ~C) STRA~ )N~S: single ~D) TOPOLOGY: lin-ar (xi) ~u~N-~ DESCRIPTION: SEQ ID NO:57:
CAGCACACTA CCAGAAGACA GrA~AATGA AAAGCA m A ~L ~L~laGCl GGGTTATTTG 60 TAA1~ .1 ACAAr~rAGC TGG~AArGTT CC~ AAGA CACAGAGGAG AAATCCAGAT 120 CATTCTCAGC TTCCCAGGCA GACCCACTCA GTGATCCTGA TCAGATGAAC GAGGACAAGC l80 GCCATTCACA GGGCACATTC ACCAGTGACT ACAGCAAGTA TCTGGACTCC AGGC~l~.CCC 240 AAGATTTTGT GCA~~ G ATGAATACCA AGAGGAACAG GAATAACATT GCCAAACGTC 300 AAGGCCAAGC TGCCAAGGAA TTCATTGCTT ~G~L~l~AA Ar~;CCG~G~ Ar~GC~-~r-A~T 420 TCCCAGAAGA GalCGCCATT GT~GAAGAAC ~ GCCG~'G ACATGCTGAT G~~ ~l 480 CTGATGAGAT GA~rArr~TT CTTGATAATC ll~CCGC~4G GGACTTTATA AA~ l~A 540 ATCACCTGCT ~GcrAcGTGG GA~ AA ATGTTAAGTC CTGTAAATTT AAGAGGTGTA 660 TTCTGAGGCC ACAl~ GCATGCCAAT AAATAAA m ~ Ll-lAGTG ~ AGCC 720 AAAAATTACA AATGr~APT~ AG m TATCA AAATATTGcT A~AATATCAG C m AAAATA 780 ,.

W O 97126321 rCTrUS97/00761 TGAAAGTGCT AGA~ l A~ ~l TAll~l~GAT GAAGTACCCC AACCTGTTTA 840 CATTTAGCGA TAAAATTATT TTmCTAl'GAT ATAATTTGTA AATGTAAATT ATTCCGATCT 90O

t2~ INFORMATION FOR SEQ ID NO:58:
Ii) ~UUkN~ CHARACTERI';TICS:
~A) LENGTH: 180 amu.no acids ~B) TYPE: amino aci.d ~C) STRA~n~nNF~S:
~D) TOPOLOGY: linear ~Xi) ~:UU~ F-~CRTPT$0N: SEQ ID NO:58:
Met Lys Ser Ile TYI Phe Val Ala Gly Leu Phe Val Met Leu Val Gln Gly Ser Trp Gln Arg Ser Leu Gln Asp Thr Glu Glu Lys Ser Arg Ser Phe Ser Ala Ser Gln Ala Asp Pro Leu Ser Asp Pro Asp Gln Met Asn Glu Asp Lys Arg His Ser Gln Gly Thr Phe Thr Ser Asp Tyr Ser Lys Tyr Leu Asp Ser Arg Arg Ala Gln Asp Phe Val Gln Trp Leu Met Asn Thr Lys Arg Asn Arg Asn Asn Ile Ala Lys Ar~ His Asp Glu Phe Glu Arg His Ala Glu Gly Thr Phe Thr Ser Asp Val Ser Ser Tyr Leu Glu Gly Gln Ala Ala Lys Glu Phe Ile Ala Trp Leu Val Lys Gly Arg Gly Arg Arg Asp Phe Prv Glu Glu Val Ala Ile Val Glu Glu Leu Gly Arg Arg His Ala Asp Gly Ser Phe Ser Asp Glu Met Asn Thr Ile Leu Asp Asn Leu Ala Ala Arg Asp Phe Ile Asn Trp Leu Ile Gln Thr Lys Ile 2&4 W O 97126321 PCTrUS97/00761 Thr Asp Arg Lys (2) INFORMATION FOR SEQ ID NO:59:
yU~N~: CHARACTERISTICS:
~A) LENGTH: 531 base pairs ~B~ TYPE: nucleic acid ~C) STRPN~ S: single ~D) TOPOLOGY: linear ~xi) ~uu~ DESCRIPTION: SEQ ID NO:Sg:
AGCTTAAGTG ACCAGCTACA ATCATAGACC ATCAGCAAGC AGGTATGTAC 1~1C~1~G~1 60 r~A~c~ w ~1- CCCCCAGCCA AAACTCTAGG GACTTTAG6A AGGATGTGGG l.C~l~u~1- 120 ACATGGACCT ~ AGCC TCAACCCTGC CTA~-.c~A GGTCATTGTT CCAACATGGC l80 C~-~ATG C'GC.-~'T~C CC~u~C,aGC C~ ,C~C ~ ~AGC CCAAGCCTGC 240 CCAGG~lL-. GTCAAACAGC AC~ G~aG TCCTCACCTG GTGGAGGCTC TGTACCTGGT 300 G~-~G4GAA CG1~ ~. TCTACACACC CAA~.CCC~1 CGTGAAGTGG AGGACCCGCA 360 AGTGCCACAA CTGGAGCTGG GTGGAGGCCC GGAGGCCGGG GATCTTCAGA CC1~G~ACT 420 GGAG~.~CC C w CAGAAGC GTGGCATTGT GGATCAGTGC TGCACCAGCA lC~ CCCl 480 CTACCAACTG GAGAACTACT GCAACTGAGT CCACCACTCC CCGCCCAGAC G 53l (2) INFORMATION FOR SEO rD NO:60::
~i) S~Ou~N~ CHARACTERISTICS:
~A) LENGTH: 955 b~se pairs ~B) TYPE: nucleic acid ~C) STR~ S: single (D) TOPOLOGY: linear (xi) x~uu~N~ nF~rRTPTION: SEQ ID NO:60:
CAGCACACTA CCAGAAGACA GCAGAAATGA AAAGCATTTA C L 1 1~ L~ GGATTATTTG 60 TAAi~ ACAA w CAGC TGGrAArGTT Cc~L~cAAGA CACAGAGGAG AAATCCAGAT 120 CATTCTCAGC TTCCCAGGCA r~ACCr~CTCA GTGATCCTGA TCAGATGAAC GAGGACAAGG 180 CA 0224643l l998-08-l2 W O 97/26321 PCT~US97/00761 CCCATTCACA GGGCACATTC ACTAGTGACT ACAGCAAGTA TCTGGACTCC AGGCGlGCCC 240 AAGATTTTGT GCA~lG~llG ATGAAT~CCA AGAGGAACAG GAATAACATT GCCAAACGTC 300 TCCC~r-~AGA G~lCGCCATT GTTGAAGAAC ll~ CCG~AG ACATGCTGAT G~rl~ l 480 CTGATGAGAT GAACACCATT CTTGATAATC ll~CCGCCAG GGACTTTATA AA~G~A 540 ATCACCTGCT AGCCACGTGG GAl~l-ll~AA ATGTTAAGTC CTGTAAATTT AAGAGGTGTA 660 TTCTGAGGCC ACAll~ r GCATGCCAAT AAATAAATTT L~ ~ lU-lAGTG l l~ r~LAGcc 72 0 AAAAATTACA AATGG~ATA~ AGT m ~TCA AAATATTGCT AAAATATCAG CTTTAAAATA ~80 TGAAAGTGCT AGAll~l~Lr All-l l~l-l~'L TA~ l~GAT GAAGTACCCC AAO~ ~l-. lA 840 (2) INFORNATION FOR SEQ ID NO:61:
(i) ~4U~N~ CHARACTERISTICS
~A) LENGTH 180 amino acids (B) TYPE: amino acid (C) STRA~ CS
(D) TOPOLOGY linear (xi) ~U~N~ DESCRIPTION: SEQ ID NO:61 Met Lys Ser Ile Tyr Phe Val Ala Gly Leu Phe Val Met Leu Val Gln l 5 l0 15 Gly Ser Trp Gln Arg Ser Leu G}n Asp Thr Glu Glu Lys Ser Arg Ser Phe Ser Ala Ser G1n Ala Asp Pro Leu Ser Asp Pro Asp Gln Met Asn Glu Asp Lys Ala His Ser Gln Gly Thr Phe Thr Ser Asp Tyr Ser Lys Tyr Leu Asp Ser Arg Arg~ Ala Gln Asp Phe Val Gln Trp Leu Met Asn CA 0224643l l998-08-l2 W O 97/26321 PCTÇUS97/00761 Thr Lys Arg Asn Arg Asn Asn Ile Ala Lys Arg His Asp Glu Phe Gl~

Arg His Ala Glu Gly Thr Phe Thr Ser Asp Val Ser Ser Tyr Leu Glu Gly Gln Ala Ala Lys Glu Phe Ile Ala Trp Leu Val Lys Gly Arg Gly Arg Arg Asp Phe Pro Glu Glu Val Ala Ile Val Glu Glu Leu Gly Arg Arg His Ala Asp Gly Ser Phe Ser Asp Glu Met Asn Thr Ile Le~ Asp Asn Leu Ala Ala Arg Asp Phe Ile Asn Trp Leu Ile Gln Thr Lys Ile Thr Asp Arg Lys (2) INFORMATION FOR SEQ ID NO:62:
u N~: CHARACTERISTICS:
~A) LENGTH: 18 base pairs (B) TYPE: nucleic acid (C) STR~N~ N~ S: single (D) TOPOLOGY: linear (xi) ~Q~N~ DESCRIPTION: SEQ ID NO:62:
lGATA r~ AC 18 (2~ lN~vK~ATION FOR SEQ ID NO:63::
~yu N~ CHARACTERISTICS:
~A) LENGTH: 18 base pairs (B) TYPE: n~cleic acid (C) STR~N~ )Nr:~S: single ~D) TOPOLOGY: linear (xi) ~Uu N~ DESCRIPTION: SEQ ID NO:63:

w o 97n6321 PCTAUS97/00761 (2) INFORMATION FOR SEQ ID NO:64:
~i) S~UU~N~: CHARACTERISTICS:
(A) LENGTH: 22 base pairs (B) TYPE: nucleic acid (C) STR~N~ ,N ~ S S: single (D) TOPOLOGY: linear ~xi) ~:QukNL~: DESCRIPTION: SEQ ID NO:64:
r~A~CTA CCAGAAGACA GC' 22 ~2) INFORMATION FOR SEQ ID NO:65:
(i) ~Q~kN~: CHARACTERIC,TICS:
(A) LENGTH: 22 base pairs ~B) TYPE: nucleic a.cid ) ST~ANI Jl~ l ~N~ s: single ~D) TOPOLOGY: linea.r (xi) S~uu~ P~CrRTPTION: SEQ ID NO:65:

(2) INFORNATION FOR SEQ _D NCt:66:
(i) S~yu~ CHARACTERI';TICS:
~A) LENGTH: 21 base pairs ~B) TYPE: nucleic acid ~C) STR~N~:I )N~-C S: single (D) TOPOLOGY: linear ~xi) ~uu~ D~Cr~TpTIoN: SEQ ID NO:66:
r~ r~CTA CCAGAAGACG C 21 ~2) INFORNATION FOR SEQ ID NO:67:
:UU~N~ CHARACTERI';TICS:
tA) LENGTH: 45 base ~tairs ~B) TYPE: nucleic e~cid ~C) STR~ ~ N~:~S: singie ~D) TOPOLOGY: line~r ~Xi ) S~UU~N~ CrRTpTIoN: SEU ID NO:67:
CTGTAGTCAC TAGTGAATGT GCC~ L~AA ~GGCC~ GGTCG 45 2~8 W O 97~6321 PCTAUS97/00761 ~2) INFORNATION FOR SEQ _D NO:68:
Uu~N~ CHARACTERISTIC5:
~A) LENGTH: 18 base pairs ~B) TYPE: nucleic acid ~C) STRA~ N~sS: single ~D) TOPOLOGY: linear ~xi) ~u~N~E DESCRIPTION: SEQ ID NO:68:
ACTAGTGAAT ~1GCC~ lB

~2) INFORMATION FOR SEQ ID NO:69:
:Uu~N~ CHARACTERISTICS:
~A) LENGTH: 18 base pairs (B) m E: nucleic acid (C) STR~ N~:~S: single (D) TOPOLOGY: linear (xi) ~:Q~N~: D~rRTPTION: SEQ ID NO:69:
ACTABTGACT ACAGCAAG l8

Claims (126)

What is claimed is:

1. A method of engineering a mammalian cell:

a) providing a starting cell that is a secretory cell;
b) introducing into said starting cell an amylin-encoding gene operatively linked to a first promoter; and c) selecting a cell from step b that exhibits increased amylin production as compared to said starting cell, wherein said amylin is biologically active.

2. The method of Claim 1, further comprising introducing into said selected cell an insulin-encoding gene operatively linked to a second promoter.

3. The method of Claim 1, wherein said starting cell produces amylin.

4. The method of Claim 1, wherein said starting cell does not produce amylin.

5. The method of Claim 1, wherein said amylin is proteolytically processed.

6. The method of Claim 1, wherein said amylin is amidated.

7. The method of Claim 1, wherein said amylin is glycosylated.

8. The method of Claim 8, wherein said glycosylated amylin is O-glycosylated.

9 The method of Claim 8, wherein said glycosylated amylin is N-glycosylated.

10. The method of Claim 8, wherein said glycosylated amylin comprises an oligosaccharide linked to threonine-9 of an amylin of SEQ ID NO:51 or SEQ
ID NO:53.

11. The method of Claim 8, wherein said glycosylated amylin comprises an oligosaccharide linked to threonine-6 of an amylin of SEQ ID NO:51 or SEQ
ID NO:53.

12. The method of Claim 11, wherein said glycosylated amylin comprises an oligosaccharide structure having between about 3 and about 10 saccharide units.

13. The method of Claim 12, wherein said glycosylated amylin comprises an oligosaccharide structure having between about 3 and about 10 saccharide units.

14. The method of Claim 8, wherein saccharides are selected from the group consisting of ribose, arabinose, xylose, Iyxose, allose, altrose, glucose, mannose, fructose, gulose, idose, galactose, talose, ribulose, sorbose, tagatose, gluconic acid, glucuronic acid, glucaric acididuronic acid, rhamnose, fucose, N-acetyl glucosamlne, N-acetyl galactosamine, N-acetyl neuraminic acid, sialic acid, amino glycal and a substituted amino glycal.

15. The method of Claim 8, wherein said glycosylated amylin comprises [NeuAc,HexNAc2]Gal(.beta.1-3)GalNac.

16. The method of Claim 1, wherein said first promoter is selected from the group consisting of CMV IE, SV40 IE, RSV LTR, RIP, modified RIP, POMC
and GH.

17. The method of Claim 2, wherein said second promoter is selected from the group consisting of CMV IE, SV40 IE, RSV LTR, RIP, modified RIP, POMC and GH.

18. The method of Claim 1, wherein the starting cell is a human cell.

19. The method of Claim 1, wherein the starting cell is a non-human cell.

20. The method of Claim 1, wherein the said starting cell is a neuroendocrine cell.

21. The method of Claim 1, wherein said starting cell is a beta cell.

22. The method of Claim 1, wherein said starting cell is a pituitary cell.

23. The method of Claim 1, wherein said starting cell is secretagogue responsive.

24. The method of Claim l, wherein said starting cell is glucose responsive.

25. The method of Claim 1, wherein said starting cell is non-glucose responsive.

26. The method of Claim 1, wherein the starting cell is derived from a .beta.TC,RIN, HIT, BHC, CM, TRM, TRM6 AtT20, PC12 or HAP5 cell.

27. The method of Claim 1, wherein said amylin encoding gene is a human amylin-encoding gene.

28. The method of Claim 1, wherein said amylin-encoding gene is linked to a selectable marker.

29. The method of Claim 28, wherein said selectable marker is selected from a group consisting of hygromycin resistance, neomycin resistance, puromycin resistance, zeocin, gpt, DHFR and histadinol.

30. The method of Claim 2, wherein said insulin gene is linked to a selectable marker gene.

31. The method of Claim 30, wherein said selectable marker is selected from a group consisting of hygromycin resistance, neomycin resistance, puromycin resistance, zeocin, gpt, DHFR and histadinol.

32. The method of Claim 1, wherein said amylin is an analog of human amylin.

33. The method of Claim 1, wherein said amylin is a non-amyloidogenic analog.

34. The method of Claim 1, wherein said amylin-encoding gene is a rat amylin-encoding gene.

35. The method of Claim 1, wherein said amylin is a rat amylin analog.

36. A method of providing amylin to a mammal comprising a) providing a starting cell that is a secretory cell;
b) introducing into said starting cell an amylin-encoding gene operatively linked to a first promoter;
c) selecting a cell of step b that exhibits increased amylin production as compared to said starting cell, wherein said amylin is biologically active; and d) administering said selected cell to a mammal.

37. The method of Claim 36, wherein said mammal exhibits at least one pathologic condition selected from the group consisting of angiogenesis, gastricemptying, anorexia, obsesity, hypertension, hypercalcernia, Pagets disease and osteoporosis.

38. The method of Claim 36, further comprising introducing into said selected cell an insulin-encoding gene operatively linked to a second promoter.

39. The method of Claim 36, further comprising (i) encapsulating said selected cell in a biocompatible coating or (ii) placing said cells into a selectively permeable membrane in a protective housing.

40. The method of Claim 36, wherein said first promoter is selected from the group consisting of CMV IE, SV40 IE, RSV LTR, RIP, modified RIP, POMC and GH.

41. The method of Claim 39, wherein said second promoter is selected from the group consisting of CMV IE, SV40 IE, RSV LTR, RIP, modified RIP, POMC and GH.

42. The method of Claim 37, wherein said cell is administered intraperitoneally, subcutaneously or via the CNS.

43. The method of Claim 37, wherein the cell is contained within a selectively semi-permeable device, said device being connected to the vasculature of the mammal.

44. A method of producing mammalian amylin comprising:
a) providing a starting cell that is a secretory cell;
b) introducing into said starting cell an amylin-encoding gene operatively linked to a first c) promoter;
d) selecting a cell of step b that exhibits increased production of amylin as compared to said starting cell, wherein said amylin is biologically active; and e) culturing said selected cell.

45. The method of Claim 44, further comprising the step of purifying said amylin.

46. The method of Claim 44, wherein said starting cell produces amylin.

47. The method of Claim 44, wherein said starting cell does not produce amylin.

48. The method of Claim 44, wherein said amylin is proteolytically processed.

49. The method of Claim 44, wherein said amylin is amidated.

50. The method of Claim 44, wherein said amylin is glycosylated.

51. The method of Claim 50, wherein said glycosylated amylin is O-glycosylated.

52. The method of Claim 50, wherein said glycosylated amylin is N-glycosylated.

53. The method of Claim 44, wherein said promoter is selected from the group consisting of CMV IE, SV40 IE, RSV LTR, RIP, modified RIP, POMC
and GH.

54. The method of Claim 44, wherein said starting cell is a human cell.

55. The method of Claim 44, wherein said starting cell is a non-human cell.

56. The method of Claim 44, wherein said amylin-encoding gene is linked to a selectable marker.

57. The method of Claim 56, wherein said selectable marker is selected from the group consisting of hygromycin resistance, neomycin resistance, puromycin resistance, zeocin, gpt, DHFR and histadinol.

58. The method of Claim 44, wherein said amylin-encoding gene is a human amylin-encoding gene.

59. The method of Claim 44, wherein said amylin is an analog of human amylin.

60. The method of Claim 44, wherein said amylin is a non-amyloidogenic analog.

61. The method of Claim 44, wherein said amylin-encoding gene is rat amylin encoding gene.

62. The method of Claim 44, wherein said amylin is an analog of rat amylin.

63. Amylin produced according to a process comprising the steps of:
a) providing a starting cell that is a secretory cell;
b) introducing into said starting cell an amylin-encoding gene operatively linked to a first promoter;
c) selecting a cell of step b that exhibits increased production of amylin as compared to d) said starting cell wherein said amylin is biologically active; and e) culturing said selected cell.

64. The amylin of Claim 63, wherein said amylin is secreted.

65. The process of Claim 63, further comprising the step of purifying said amylin.

66. The process of Claim 63, further comprising the step of introducing into said cell an insulin-encoding gene operatively linked to a second promoter.

67. The method of Claim 63, wherein said amylin is proteolytically processed.

68. The method of Claim 63, wherein said amylin is amidated.

69. The method of Claim 63, wherein said amylin is glycosylated.

70. The method of Claim 69, wherein said glycosylated amylin is O-glycosylated.

71. The method of Claim 69, wherein said glycosylated amylin is N-glycosylated.

72. The method of Claim 63, wherein said first promoter is selected from the group consisting of CMV IE, SV40 IE, RSV LTR, RIP, modified RIP, POMC and GH.

73. The method of Claim 66, wherein said second promoter is selected from the group consisting of CMV IE, SV40 IE, RSV LTR, RIP, modified RIP, POMC and GH.

74. The method of Claim 66, wherein the amylin-to-insulin content of said selected cell is between about 0.002 to about 10Ø

75. A method of regular blood glucose levels in a mammal comprising:
a) providing a starting cell that is a secretory cell;

b) introducing into said starting cell an amylin-encoding gene operatively linked to a first promoter;
c) selecting a cell of step b that exhibit increased amylin secretion as compared to said starting cell; and d) administering said selected cell to said mammal, whereby said secreted amylin is biologically active and regulates blood glucose levels of said mammal.

76. The method of Claim 75, further comprising the step of providing said cell with an insulin-encoding gene operatively linked to a second promoter.

77. A method for modulating the circulating levels of insulin in a mammal comprising the steps of:
a) providing a starting cell;
b) introducing into said cell an amylin-encoding gene operatively linked to a promoter;
c) selecting a cell of step b that exhibits increased amylin secretion as compared to said starting cell; and d) administering said selected cell to said mammal, whereby said secreted amylin modulates glucose-stimulated insulin secretion in said mammal.

78. A method for decreasing glycogen synthesis in a mammal comprising the steps of:
a) providing a starting cell that is a secretory cell;
b) introducing into said cell with an amylin-encoding gene operatively linked to a first promoter;

c) selecting a cell of step b that exhibits increased amylin secretion as compared to said starting cell wherein said amylin is biologically active; and d) administering said selected cell to said mammal, whereby said secreted amylin reduces glycogen synthesis in said mammal.

79. The method of Claim 78, wherein said secretory cell further is transfected with an insulin-encoding gene operatively linked to a promoter.

80. A method of screening for an amylin receptor comprising:
a) obtaining amylin from a recombinant amylin expressing secretory cell;
b) admixing said amylin with a composition comprising a putative amylin receptor; and c) detecting an amylin receptor bound to said amylin.

81. The method of Claim 80, wherein said composition comprising a putative amylin receptor is a composition comprising a population of recombinant cells transfected with portions of a DNA library.

82. The method of Claim 80, further comprising obtaining a DNA segment from said DNA library that expresses an amylin receptor.

83. An amylin receptor gene, prepared by the process of:
a) obtaining amylin from a recombinant amylin expressing secretory cell;
b) admixing said amylin with a composition comprising a putative amylin receptor said composition comprising a population of recombinant cells transfected with portions of a DNA library; and c) obtaining a DNA segment from said DNA library that expresses an amylin receptor.

84. An amylin receptor-like gene, wherein at least a portion of said gene hybridizes to at least a portion of the amylin receptor gene of Claim 83 under low stringency hybridization conditions.

85. A purified amylin composition comprising an amidated and glycosylated amylin polypeptide.

86. A method for the production of a polypeptide comprising the steps of:
a) providing a secretory host cell that is a secretory cell;
b) blocking the production of an endogenous, secreted polypeptide;
c) contacting with said host cell an exogenous polynucleotide comprising a gene encoding an exogenous polypeptide, wherein said gene is under the control of a promoter active in eukaryotic cells; and d) culturing said secretory host cell under conditions such that said exogenous polynucleotide expresses said exogenous polypeptide.

87. The method of Claim 86, wherein said promoter is selected from the group consisting of CMV, SV40 IE, RSV LTR, GAPHD and RIP1.

88. The method of Claim 86, wherein said exogenous polynucleotide further comprises an adenovirus tripartite 5' leader sequence and intron.

89. The method of Claim 88, wherein said intron comprises the 5' donor site of the adenovirus major late transcript and the 3' splice site of an immunoglobulin gene.

90. The method of Claim 86, wherein said exogenous polynucleotide further comprises a polyadenylation signal.

91. The method of Claim 86, wherein said secretory host cell is a neuroendocrine cell.

92. The method of Claim 91, wherein said secretory host cell is an insulinoma cell.

93. The method of Claim 92, wherein said insulinoma cell is a rat insulinoma cell.

94. The method of Claim 92, wherein said insulinoma cell is a human insulinoma cell.

95. The method of Claim 86, wherein said exogenous polypeptide is secreted.

96. The method of Claim 95, wherein said exogenous polypeptide is fusion protein.

97. The method of Claim 95, wherein said exogenous polypeptide is amidated.

98. The method of Claim 97, wherein said amidated polypeptide is selected from the group consisting of calcitonin, calcitonin gene related peptide (CGRP), .beta.-calcitonin gene related peptide, hypercalcemia of malignancy factor (1-34) (PTH-rP), parathyroid hormone-related protein (107-128) (PTH-rP), parathyroid hormone-related protein (107-111) (PTH-rP), cholecystokinin (26-32) (CCK), galanin message associated peptide, preprogalanin (64-105), gastrin I, gastrin releasing peptide, glucagon-like peptide (GLP-1), pancreastatin, pancreatic peptide, peptide YY, PHM, secretin, vasoactive intestinal peptide (VIP), oxytocin, vasopressin (AVP), vasotocin, enkephalins, enkephalinamide, metorphinamide (adrenorphin), alpha melanocyte stimulating hormone (alpha-MSH), atrial natriurc-tic factor (5-27) (ANF), amylin, amyloid P component (SAP-1), corticotropin releasing hormone (CRH), growth hormone releasing factor (GHRH), luteinizing hormone-releasing hormone (LHRH), neuropeptide Y, substance K (neurokinin A), substance P and thyrotropin releasing hormone (TRH).

99. The method of Claim 95, wherein said exogenous polypeptide is a hormone.

100. The method of Claim 99, wherein said hormone is selected from the group consisting of growth hormone, prolactin, placental lactogen, luteinizing hormone, follicle-stimulating hormone, chorionic gonadotropin, thyroid-stimulating hormone, leptin, adrenocorticotropin (ACTH), angiotensin I, angiotensin II, .beta.-endorphin, .beta.-melanocyte stimulating hormone (.beta.-MSH), cholecystokinin, endothelin I, galanin, gastric inhibitory peptide (GIP), glucagon, insulin, lipotropins, neurophysins and somatostatin.

101. The method of Claim 95, wherein said exogenous polypeptide is a growth factor.

102. The method of Claim 101, wherein said growth factor is selected from the group consisting of epidermal growth factor, platelet-derived growth factor, fibroblast growth factor, hepatocyte growth factor and insulin-like growth factor 1.

103. The method of Claim 86, wherein said endogenous, secreted polypeptide and said exogenous polypeptide are the same.

104. The method of Claim 103, wherein said endogenous, secreted polypeptide and said exogenous polypeptide are insulin.

105. The method of Claim 86, wherein said exogenous polypeptide enhances the production and/or secretion of at least one polypeptide produced by said cell.

106. The method of Claim 105, wherein said exogenous polypeptides is selected from the group consisting of a protein processing enzyme, a receptor and a transcription factor.

107. The method of Claim 105, wherein said exogenous polypeptide is selected from the graup consisting of hexokinase, glucokinase, GLUT-2, GLP-1, IPFI, PC2, PC3, PAM, glucagon-like peptide I receptor, glucose-dependent insulinotropic polypeptide receptor, BIR, SUR, GHRFR and GHRHR.

108. The method of Claim 86, wherein said step (c) further comprises contacting said secretory host cell with a polynucleotide comprising a gene for a selectable marker and step (d) further comprises culturing under drug selection.

109. The method of Claim 108, wherein said selectable marker gene is flanked by LoxP sites.

110. The method of Claim 109, further comprising e) contacting the secretory host cell with a polynucleotide encoding the Cre protein, wherein said polynucleotide is under the control of a promoter active in eukaryotic cells; and f) replicate culturing said cell with and without drug selection.

111. The method of Claim 109, wherein said selectable marker is hygromycin resistance and said drug is hygromycin.

112. The method of Claim 109, wherein said selectable marker is neomycin and said drug is G418.

113. The method of Claim 109, wherein said selectable marker is GLUT-2 and said drug is streptozotocin.

114. The method of Claim 109, wherein the genes for said exogenous polypeptide and said selectable marker are part of the same polynucleotide.

115. The method of Claim 114, wherein the genes for said exogenous polypeptide and said selectable marker are separated on the same polynucleotide by an internal ribosome entry site.

116. The method of Claim 86, wherein said secretory host cell is glucose-responsive.

117. The method of Claim 86, wherein said secretory host cell is not glucose-responsive.

118. A secretory host cell comprising an exogenous polynucleotide comprising a gene encoding a first exogenous polypeptide, wherein said secretory host cell is blocked in the expression of at least one endogenous, secreted polypeptide.

119. The secretory host cell of Claim 118, wherein said exogenous polynucleotide is inserted into the coding or regulatory region of said endogenous, secreted polypeptide.

120. The secretory host cell of Claim 118, wherein said exogenous polynucleotide further comprises a promoter active in eukaryotic cells.

121. A method of preventing type I diabetes comprising the steps of:

a) identifying a subject at risk of type I diabetes; and b) providing to said subject a polynucleotide comprising a gene for a disulfide mutant of human insulin, wherein said mutant gene is under the control of a promoter active in eukaryotic cells.

122. The method of Claim 121, wherein said promoter is selected from the group consisting of CMV, SV34 IE, RSV LTR, GAPHD and RIPI.

123. The method of Claim 121, wherein said providing comprises introducing said polynucleotide to a cell of said subject in vivo.

124. The method of Claim 121, wherein said providing comprises contacting with a secretory host cell ex vivo and administering said secretory host cell tosaid subject.

125. A method for treating a subject afflicted with diabetes comprising the steps of:

a) identifying a subject afflicted with diabetes; and b) providing to said subject a secretory host cell, wherein (i) the production of an endogenous. secreted polypeptide has been blocked and (ii) wherein the secretory host cell comprises an exogenous polynucleotide comprising a gene encoding insulin, wherein said gene is under the control of a promoter active in eukaryotic cells.

126. A method for providing a polypeptide to an animal comprising the step of providing to an animal a secretory host cell, wherein (i) the production of an endogenous, secreted polypeptide in said secretory host cell has been blocked and (ii) wherein the secretory host cell comprises an exogenous polynucloetide comprising a gene encoding said polypeptide wherein said gene is under the control of a promoter active in eukaryotic cells.