CA2163427A1 - Bifunctional selectable fusion genes based on the cytosine deaminase (cd) gene - Google Patents

Abstract

The invention provides selectable fusion genes including a dominant positive selectable gene fused to and in reading frame with a negative selectable gene. The selectable fusion gene encodes a single bifunctional fusion protein which is capable of conferring a dominant positive selectable phenotype and a negative selectable phenotype on a cellular host. A dominant negative selectable phenotype is conferred by the cytosine deaminase (CD) gene for 5-fluorocytosine sensitivity (5-FCs). A dominant positive selectable phenotype is conferred, for example, by the neo gene for G-418 aminoglycoside antibiotic resistance (G-418r), or by the hph gene for hygromycin B resistance (Hmr). The present invention also provides recombinant expression vectors, such as retroviral vectors, which include selectable fusion genes, and cells transduced with the recombinant expression vectors. The bifunctional selectable fusion genes are expressed and regulated as a single genetic entity, permitting co-regulation and co-expression with a high degree of efficiency.

Description

~p 94/28143 PCT/US94/05601 BrFUNCTIONA~ SF~ FCT~RT F FUSION GENES
BA.~Fn ON T~F CYTOSINE DFAM~A.~F (CD) GE~F

Bael~.o.llld The present invention relates generally to genes eAp~ ~Lhg sel~qhle phcho~c,s.
More particularly, the present i~,cnLion relates to genes capable of co l,ApL~3-..g both ~ominq~lt positive sele~ ble and n_gdli~e sP'--t 1~le ph~L~
Genes which eApress a sP~e~le phen~ e are widely used in lL~---b,l,dn~ DNA
terhnnlogy as a means for ide~il~i~ and icol ~i~ host cells into which the gene has been 10 introduced. Typically, the gene eA~r~ the s~ le phenotype is introduced into the host cell as part of a fe.o!-~k;..- ~I tA~r~ssion vector. Positive sPIc~ ~' le genes provide a means to identify and/or isolate cells that have retained i lllL ducod genes in a stable form, and, in this Cd~diil~, have greatly f~ t~q-ted gene transfer and the analysis of gene r. -.- I;o ~ Neg~i~e sele~tqhle genes, on the other hand, provide a means for el;~ .g cells that retain the 15 introduced gene.
A variety of genes are available which confer sPIo~' ble phc~o~ s on animal cells.
The bacterial neo,..~c,n phO~l~hol~r -f t-~a (neo) (Colbere Garapin et al., J. Mol. Biol. 150:1, 1981), hy~:lL~ chl pho~h~J~ rL -~e (~ph) (Santerre et al., Gene 30:147, 1984), and Yi~nthin~guanine pho",ho. il~ - r,..,~ -c C~t) (Mulligan and Berg, Proc. Na~l. Aicad. Sci.
USA 78:2072, 1981) genes are widely used ~o~ 1 positive s~e-t^ble genes. The Herpes sim~lex virus type I l~ idi~e Icinase (HSV-I TK) gene (Wigler et al., CeU 11:223, 1977); the cellular adenine ph~ho, il~sy~ e (APRT) (Wigler et al., Proc. Na~l. Acad. Sci. USA
76:1373, 1979); and h~ pho~-kolil~syl~ fe~ --e (HPRT) genes (Jolly et al., Proc.Natl. Aicad. Sci. USA 80:477, 1983) are c~ used r~i~ positive selectable genes. In general, ~ornin~nt sr-le~ ble genes are more versatile than r~ genes, because ~ie use of recessive genes is limited to mutant cells ~iefirient iDi the sele,li~b'e r~ l;- n~ /he.~s ~iG~ nl genes may be usedi in wild-type cells.
Several genes confer ne~ as well as positive s ~ t^blie phenotypes,; I --~ e th-ie HSV-I TK, HPRT, APRT and ~t genes. Thiesc gcnes encode e~ics which catalyze the conversion of nllrleor;~l~ or purine analogs to e~tu~.~.c ;- t~ Thie .-- ~IP~,side analog WO 94/28143 216 3 4 ~ 7 PCT/US94/05601 ^2-ganciclovir (GCV) is an effi~i~Pnt s~,b~ e for HSV-I TK, but a poor ~sl,~l~ for cellular TK, and therefore may be used for negdlive sPI~tion against the HSV-I TK gene in wild-type cells (St. Clair et al., Antimicrob. Agents Chemother. 31:844, 1987). However, the HSV-I TK gene may only be used effectively for positive SPIP~fion in mutant cells lacking cellular TK activity.
Use of the HPRT and APRT genes for either positive or n~,alive selection is similarly limited to HPRT or APRT cells"~yecliv~ly (Fenwick, "The HGPRT Systemn, pp. 333-373, M.
Gol~ (ed.), Molecl/7nr Cell Genetics, John Wiley and Sons, New York, 1985; Taylor et al., "The APRT System", pp. 311-332, M. ~.,u~ . (ed.), Mole~7~r Cell G~ t~tir~, John Wiley and Sons, New Yorlc, 1985). The gpt gene, on the other hand, may be used for both positive and negative sPIP,ction in wild-type cells. Negative sPIP~tion against the gpt gene in wild-type cells is possible using ~thio~ ne~ which is effiripntly conve,led to a CytotoAic nucleotide analog by the ba.,l~.ial gpt enzyme, but not by the cellular HPRT enzyme (13esnard et al., Mol. CeU. Biol. 7:4139, 1987).
Another negatively selectable gene has recently been l~yo~led by Mullen et al., Proc.
Natl. Acad. Sci. USA 89:33, 1992. The ba~ 1 C~illÇ d~ ce (CD) gene CO~ 5-fluorocytosine (5-FC) to 5-nuo-ou-d.,il (5-FU). 5-FU is further metabolized intrace~ ly to 5-fluoro-uridine-5'-trirhocrh~t-P, and 5-fluoro-2'-deoAy-uridine-5'-,1lonul.hosph~e, which in_ibit RNA and DNA synthesis, causing cell death. Thus, 5-FC can erf~clively ablate cells Cdllyillg and expressing the CD gene. The CD gene is not positively selectable in normal cells.
More recently, at~nfion has turned to selectable genes that may be incol~al~d into gene transfer vectors designed for use in human gene therapy. Gene therapy can be used as a means for aug-- ~nl;i~g normal cellular r---~l;on for e~nple, by introducing a heterologous gene capable of modifying cellular activities or cellular phc~ c, or ~ ely, eA~.~,sing a drug needed to treat a disease. Gene therapy may also be used to treat a hc~ genetic disease which results from a defect in or absence of one or more genes. Collcc~i~ely, such diseases result in si~ifi~ ~t morbidity and mortality. EAamples of such genetic diseases include hemophilias A and B (caused by a d~ of blood c~ n factors VIII and IX, respectively), alpha~ dt!r;- ;ç----~, and ~noSine d~ ce d~fiCi~ncy. In each of these particular cases, the missing gene has been if ~-ontifiP~ and its c~mp~ ç~ - y DNA
(cDNA) molecularly cloned (Wood et al., Nature 312:330, 1984; Anson et al., Nat~re 31S:683, 1984; and Long et al., Biochemistry 23:4828, 1984; n~;ldonq et al., J. Biol. Chem.
259: 12101, 1984). While p. lliative therapy is available for some of these genetic dic~qc~s~
often in the forin of ~.I.";n;.~ ';on of blood products or blood ~ fi~ nc, one way of treating such genetic diseases is to introduce a repl~çm~nt for the defective or missing gene back into ~ 94/28143 ~16 3 ~ 2 ~ PCT/US94/05601 the somatic cells of the patient, a process referred to as "gene therapy" (~n~Pr~ol, Science 226:401, 1984).
The process of gene therapy typically involves the steps of (1) removing somatic (non-germ) cells from the patient, (2) introducing into the cells cx uw a ~elapeulic or replaeemPnt 5 gene via an ~pro~idle vector capable of tA~r~sing the Ihc~ ic or rep~ e-.~ gene, and (3) transplanting or ~ fi~ g these cells back into the patient, where the t~ ;c or repl~^PmPnt gene is t.~r~SCd to provide some ~ I;c benefit. Gene transfer into somatic cells for human gene therapy is presently acLieved ex viw (Kasid et al., Proc. Natl. ~lc~d. Sci.
USA 87:473, 1990; Roscnbe.g et al., N. Engl. J. Med. 323:570, 1990), and this relatively 10 inPffi~iPnt process would be f~ t~tpA by the use of a domin~n~ positive selectable gene for idt;llliîyi.,g and isolating those cells into which the lC~ ---f -~1 gene has been introduced before they are relu"~f~d to the patient. The neo gene, for e~ nple~ has been used to identify genPtir~lly modifiPd cells used in human gene therapy.
In some i~c~ l-es~ however, it is possible that the introduction of genptir~lly m~ifi 15 cells may actually cor,l~,or.,ise the health of the patient. The ability to selectively elimin~e genPtir~lly modified cells in viw would provide an ~d~ition-~l margin of safety for patients undergoing gene therapy, by pe~ g reversal of the p,~,celure. This _ight be 3cco...~ hPd by incc"~,o,ali-~g into the vector a n~,5~ sele~ le (or "suiciden) gene that is capable of functioning in wild-type cells. Inco,~ of a gene capable of co~-i ,g both dc.~
20 positive and neg~ e selc~l~hle phc..cl~,~c,s would ensure co eA~.es~ion and co-reg~ tion of the positive and ncg~ .e sPIe~ le phc.~ cs~ and would ~ --- '7 the size of the vector.
However, positive sP,le~tion for the &~t gene in some i~ e$ requires precise selection conditions which may be difficult to d~ "~;~.r For these reasons, co~ ~r~sio,. of a do...h~
positive sPIP~t~ble ph&nul~ Jc and a ne6dli~e sPIect~le ph&,lol~c is typically achieved by co-25 eA~,~shlg two dirr~re,.l genes which s~ , encode other llo"-;-~ ~ positive and n&6d ive s~ect: ' le r....l-tions, rather than using the &Dt gene.
The existing sl. ~.,6i s for co-eA~,~si-.6 ~o---~ ' positive and negative s~ l; ble ph~noly~vs en~ode~ by dirr~.e.~t genes often present ~mrl~P- rh~lPnges. The most widely used terhni-llle is to co-l,a~r~l two pl~cmir~s jct~ P ~f~l;~'g two ph&nJt~cs ~Vigler et 30 al., Cell 16:777, 1979). However, the effiriPnry of co-transfer is rarely 100%, and the two genes may be subject to independent genetic or epige.lclic ~ io~ A second strategy is to Iink the two gene. on a single pl~cmi~, or to place two in~erPn~lpnt tf ~rr~ il-t;oll units into a viral vector. This method also suffers from the disa~lv~ that the genes may be in~lepenr~P,ntly regulated. In retroviral vectors, su~r~sio~ of one or the other inAPpen~pnt Wo 94/28143 ~ ~,3 ~ PCT/US94/05601--transcription unit may occur (F~ and Temin, Mol. Cell. Biol. 6:792, 1986). In ad~liti~)n, in some cir~ res there may be inc-lffriPnt space to acco... ~atP two functional trancc~irtion units within a viral vector, ~Itho-l~h retroviral vectors with ru.l~lional mllltirle pro.l.ole,~ have been s~lccP-ccfi-lly made (Overell et al., Mol. cen Biol. 8:1803, 1988). A third 5 strategy is to express the two genes as a bicistronic ~NA using a single proluolel. With this method, however, the distal open reading frame is often tra~ pd with variable (and usually reduced) Pffrienr,y (K~llfrnan et al., EMBO J. 6:187, 1987), and it is unclear how t;rr~ive such an expression strategy would be in primary cells.
The present invention provides a method for more PffiriPntly and reliably co~,~.esshlg 10 a ~lomin~nt positive selectable phenotype and a ne~live selectable phenotype encoded by dirre,-illl genes.

SUMMARY OF THE INVENT~ON
The present hl~6~llion provides a sPIP~ble fusion gene Co~ iLg a dominant positive 15 selectable gene fused to and in reading frame with a negalive selectable gene. The selectable fusion gene encodes a single bifi~nrtion~l fusion protein which is capable of c~-rP~ ;n~ a dominant positive selectable phenotype and a negali~e selectable phen.,l~c on a cellular host.
The selectable fusion genes of the present invention c~...l.. ;~e nucleotide sequ~PnrP~s for n~ative selection that are derived from the ba~(r-;a~ ine d~P = ..~ ce (CD) gene.
In a prc;rt;~l~d embodiment, the s~ hle fusion gene co ~ Pes mlrlP4tirl-p~ s~uenr_,s from the bacterial CD gene fused to mlrlP~tide sP~e~nr,P~ from the neo gene, referred to herein as the CD-neo splp~ctahle fusion gene (Se~lPnre Listing No. 1). The CD-neo sel~l~le fusion gene confers both G418 I~ e (G418 ) for do...;.~ positive sPl~ ;on and 5-fluoroc~il,c sensitivity (5-FCS) for negative sPlpxtio~
The present i Ivelllion also provides recombinant t~r~ioll vectors, for e~mplq, .rOVilll:~s, which include the sP~e~table fusion genes, and cells tr^n~ced with tbe -,bin~ll eApressioll vectors.
The selectable fusion genes of the present invention are e,~ A and reg-l~t-P~ as a single genetic entity, pP- .--;l ~ g co-reg~ ion and co l,A~I~ssion with a high degree of effiriPnry.

~ 94/28143 2 16 3 4 2 7 PCTluss4lo56ol BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 shows ~ .. ....c of the e~cpression c~cs~ps c~ ~ in p1~mi~1~
tgCMV/hygro/LTR, tgCMV/neo, tgCMV/hygro-CD, tgCMV/CD-hygro, tgCMV/neo-CD and tgCMV/CD-neo. The ho~ arrows indicate l~ cc- ;ll1 ;0l-~l start sites and direction of S transcription. The open bo~ labeled LTR is the r~ovi~al long le~ ' repeat. The open box labeled CMV is the ~ul~le~alovirus p~ulllùl~r.
Figure 2 shows the results of the ~s-..c d~ assay on e~ctracts p~ ~ from l~a~r~l d pools of NIH/3T3 cells. The e~tracts were assayed by ~ g the conversion of cytosine to uracil.
Figure 3 shows ~ .uc of the proviral structures of r~ovi~al vectors tgLS(+)neo and tgLS(+)CD-neo used in the present invention.
Figure 4 shows the results of the . ~sinc df m.;~ e assay on ~ ;..rr~led (lane 1), tgLS(+)neo-infected (lane 2) and tgLS(+)CD-neo-infected NIH/3T3 (lane 3) cell pools. The results indicate that cells infected with the tgLS(+)CD-neo e~press high levels of ~"~sillc 15 d~ ce activity.
Figure S shows phulu~,-aphs of stained colo~;~s of u~;..r~CI~ NIH/3T3 cells (plates a, b and c) and NIH/3T3 cells infected with the tgLS(+)neo (plates d and e) or tgLS(+)CD-neo (plates f and g) r~ vihu~es. The cells were grown in ~ m alone (plate a) or m~Ainm supplenlPntPd with G418 (plates b, d and f) or G-418+5-FC (plates c, e and g) in a long-term 20 prolifer?tic n assay. The data show that u~ r~d NIH/3T3 cells were s~ ve to G418 and Sii~lant to 5-FC, NIH/3T3 cells infected with tgLS(+)neo are ,esii.l_~ to both G418 and 5-FC, and NIH/3T3 cells infected with tgLS(+)CD-neo are l~DiSI~ . to G~18 and 51"~Silive to 5-FC.

DET~TT Fn DESCRIPTION OF THE INVF~TION
SEQ ID NO:l and SEQ ID NO:2 (~pe~ ;--~ prior to the claims) show specific embodiment~ of the r ^leoti~e se~uenre and co"~ e amino acid sellllPn~e of the CD-neo selectable fusion gene of the present i~ The CD-neo s~e~le fusion gene shown in the Se~llRnre Listing C~ ;-e5 Se~ S from the CD gene (~ rleQ~ Pc ~1281)30 linked to se~u~pnrpc frûm the neo gene (rllrleoti~lp-~ 1282-2073).

D~r~
As usod herein, the term "se~ hle fusion gene" refers to a rvrl~ti~e sc~ re co"~ h~g a dominant positive sel~hle gene which is fused to and in reading frame with a 35 negative selectable gene and which encodes a single bir.~ ;o~l fusion protein which is capable 2 ~ ~ 3 ~ 2 ~ PCT/US94/05601--of corÇei,ing a dominq~t positive sPIectq-hle phenotype and a ne2;divc selP~ti 'lle phenotype on a cellular host. A "dominant positive selectable gene" refers to a se~uPnre of nucleotides which encodes a protein CollÇcl.i--g a dominq-nt positive sPlecti~1 1e ph~ c on a cellular host, and is ~iccu~c~p~d and exemplified in further detail below. A "ne~alive ce~ le gene" refers to a S sequenre of ml~le~Qti~D$ which encodes a protein coi~rc lh~g a negdi~e. selectable phenotype on a cellular host, and is also IiscllccDd and e~-Dmr!ifi-Dd in further detail below. A "selectq-hle gene" refers ge~ ally to ~lomi~qnt positive sDlD~t~ble genes and ne~;dlivc selDctq-hle genes.
A sDIect~ le gene is "fused to and in reading frame with" another selectable gene if the cJ~ression products of the sDIecti ble genes (i.e., the proteins encoded by the se~P~tq-kle genes) 10 are fuse~ by a peptide bond and at least part of the biological activity of each of the two proteins is retained. With .erc.~,n-~e to the CD-neo sPIDxtYqhle fusion gene disclosed herein, the CD gene (encoding CyLC sinF dP~ ..i~-~ce, which confers a ne~dive sP~P~^'-le phenotype of 5-fluorocytosine sensitivity, or 5-FCS) is fused to and in reading frame with the neo gene (encoding neomycin phocrhol . a-~rP~ce~ which confers the dO.I.ih~ l positive sPIectable 15 ph~lo~c of G~18 r~;~ e, or G~18 ) if the CO and neo proteins are fused by a peptide bond and eA~ cd as a single birl~n~l;Qn~l fusion protein.
The component selectable gene sc~" ~r~ ~ of the present invention are pl~;~ably contiguous; however, it is possible to construct s-DIe~ble fusion genes in which the co~n~
sPIPctq~ble gene se~uPnrP-s are sepq~a'Pd by internal no.~ cl~-Pd ~llrlP~ti~ls s~ ,c -rF~, such as 20 introns. For purposes of the present h~ ion, such nnl~co~ c sclc~,lable gene se~uPnce-c are considered to be fused, provided that ~A~i~ssion of the SPle le fusion gene results in a single bifilnrtionql fusion protein in which the eA~res~io~ products of the co.ll~onclll sel~P~ble gene sequPncPS are fused by a peptide bond.
"Nucleotide se~nPnre" refers to a heteropolymer of deuA~ rl~pQti(lp-c or 25 ribonucleotides, such as a DNA or RNA sc~ e Nuc~PQti~1e sc~l~ cnrçls may be in the forrn of a separate Çla,~ ,nl or as a C...~pon ~ of a larger cor~ ,f~.ably, the mlrleQtir1P, se~ Pnres are in a quantity or con~enl.dion en~li~ irlPntific~irn ~ u~ and l~VVt;
of the se~uence by s~d~d biorhPmi~ ~' ~ho~ for P~ mp'e, using a cloning vector.
Recombinant mlrlPQtide se~ Pnr~Ps are the product of variouLs c~ of cloning, 30 restriction, and ligation steps resulting in a co~i,~.l having a sl~ i ' coding seqnPnre le from homologous sh~ pn~ s found in n~hlral systems. Generally, nvcl~PQti~e se~u~Pnr~PS enCo~iing the structural coding sc~ -re, for e~rnple, the sP~ le fusion genes of the present invention, can be assembled from nucleotide r ~ and short oligon~ PQtirie linkers, or from a series of oli~m~rle~ti~iPs, to provide a synthetic gene which is capable of ~0 94n8143 216 3 ~ 2 7 PCT/US94/05601 being expressed in a lecu~billalll l,ai~sc,il~liondl unit. Such se~uenrP-c are plefelàbly provided in the form of an open reading frame l~llinl~upled by internal ~ cl~Pd se~lPnePs, or introns, which are typically present in eukaryotic genes. Genomic DNA cO.~ g the relevant selectable gene se~uPnn~PS is plef~ ~ably used to obtain ~,r~,i~le m~rlPoti~le se~llPnrPc S encoding selectable genes; ho..~, cDNA r.~ may also be used. Se~uPncP~c of non-translated DNA may be present 5' or 3' from the open reading frame or within the open reading frame, provided such se~uenres do not illle.rere with In~nir~ ion or expression of the coding regions. Some genes, however, may include introns which are nl~e~C~ ~ for proper expression in certain hosts, for example, the HPRT sel~t~ble gene i~ les introns which are 10 necec~y for expression in ~, ~ui~al stèm (ES) cells. As suggested above, the nucleotide se~U~pnrp-c of the present hlvelllioll may also co",l"ise RNA se~uPnrPs, for example, where the nucleotide se~u~Pnr~PrC are pacl~ed as RNA in a ,~ovi,us for infPctir~ a cellular host. The use of re~ovi,àl e,~iession vectors is (licc~lc$ed in greater detail below.
The term "recombinant expression vector" refers to a replicable unit of DNA or RNA
15 in a form which is capable of being tr?nsduce~ into a target cell by ~ ~n~r.~l iOI~ or viral infection, and which codes for tbe e.~ ~sioll of a select~1-1e fusion gene which is l~ansc~ibed into mRNA and tr~ncl~p~ into protein under the control of a genetic element or elpmpntc having a regulatory role in gene e~r~ssion, such as l~allSC~il lion and tr~ncl~ion i..i~ ;OI~ and termination se~mPnrPs. The Lecollll)inàlll expression vectors of the present invention can take 20 the form of DNA constructs replir~t~d in ba~;L,.ial cells and ll~-,r~led into target cells directly, for example, by calcium ph~sph ~ prec;~ild~ioll, el~l~upo,alion or other physical transfer meth9-1.c. The recolllbin~ll eApression vectors which take tne form of RNA co~,L.uc~, may, for e-~mrle, be in the form of inf~tio~c r~ vi,u~D,c~D, pa.,~l by suitable npa L~ g"
cell lines which have pl~riUUSly been ~ r~e~l with a proviral DNA vector and produce a 25 retrovirus cc~ an RNA ~ c. ;I~l of the proviral DNA. A host cell is infected with the r~lruvuu~" and the ~ellUVildl RNA is replir~tP,cl by reverse t-ausc~iplion into a double s ~anded DNA ;.~ ..P~ te which is stably i..l~g..~t~d into cbromos~m~' DNA of the host cell t,o form a provirus. The provirus DNA is then ~Alu~,sP~d in the host cell to produce pol~p~.lides enC~
by the DNA. The ,e~ ou~binanl expression vectors of the present hl~ ion thus include not 30 only RNA constructs present in the i~.r.~i~uC r- hovuus~ but also copies of proviral DNA, which include DNA reverse l - ~,.c~ of a ~ovi,us RNA genome stably i~ ed into chromosomal DNA in a suitable host cell, or cloned copies thereof, or cloned copies of l~ninl~ ~aled inl~.n~li~-~. forms of retroviral DNA. Proviral DNA inCl~d~ps l.,-~c --;pl;-nal elements in independent ol,~.dive ~c~oci~i9n with selected sl~uclu~dl DNA sc~ c~,~ which are 35 transcribed into mRNA and tra~D~P~ into protein when proviral se~ enrPs are eA~,essed in infected host cells. Recol-,biL~l expression vectors used for direct ~ cre~l;on will include DNA se~u~Pn~PS enabling re~lirq-tion of the vector in ba~ ~e~ host cells. Various le~..ll,in~
expression vectors suitable for use in the present i ~ ion are descLib~d below.
"Tr~q-ncduce" means intro~ on of a ~ r~ssion vector co~ h~ a S selectable fusion gene into a cell. Trqnc~ ction meth~c may be p~.y~ical in nature (i.e., rr"l; ~rt), or they may rely on the use of l~.~in~l viral vectors, such as retl'.)VilUSe~S, enr~in~ DNA which can be l.~sclibed to RNA, pa~ed into ;..r~l;Ou- viral particles and used to infect target cells and thereby deliver the desired genetic mqtPri~ql (i.e., infection).
Many diîîel~,-ll types of -- ."",~liqn gene transfer and r~OIll~ e~cpression vectors have 10 been developed (see, e.g., Miller and Caios, Eds., "Gene Transfer Vectors for ~qmm~liqn Cells," Current Cornrn. Mol. Biol., (Cold Spring Harbor Labola~ , New Yorlc, 1987)).
Naked DNA can be physically introduced into .--~ iqn cells by 1., .~r~l;on using any one of a number of terhni-lu~Ps inclllAing~ but not limited to, calcium phosph ~e tr~qncf~rtion (Rermqn et al., Proc. Natl. Acad. Sci. USA 84 81:7176, 1984), DEAE-De~ctran 1l~ .c~ ;o~ (McCutch. n et al., J. Natl. Cancer Inst. 41:351, 1986; T l~thmq~ et al., Nucl. Acids Res. 11:1295, 1983), protoplastfusion (Deans et al., Proc. Natl. Acad. Sci. US~l 84 81:1292, 1984), electroporation (Potter et al., Proc. Natl. Acad. Sci. USA 84 81:7161, 1984), lipof~ioll (Felgner et al., Proc.
Na l. Acad. Sci. USA 84:7413, 1987), Polybrene h~ h~ e bromide 1-~ r~l;n~ (Kawai and Nishizawa, Mol. CeU. Biol. 4: 1172, 1984) and direct gene transfer by laser mic ~ e of cell .lle.llblah~s ~ao et al., Proc. Natl. Acad. Sci. US~ 84:4180, 1987). Various ;~r~~ n techni~lu~Ps have been developed which utilize r~o~ infie~tiol~c virus particles for gene delivery. This represents a piere.l~,d ~loach to the present invention. The viral vectors which have been used in this way include virus vectors derived from simian virus 40 (SV40;
K~rl~on et al., Proc. Natl. ~cad. Sci. US~ 84 82:158, 1985), ade~v~ on et al., EMBO J. 5:2377, 1986), adeno~ oc; ~ed virus (LaFace et al., Virology 162:483, 1988) and retroviruses (Coffin, 1985, pl7-71 in Weiss et al. (eds.), RNA Tumor Viruses, 2nd ed. Vol 2, Cold Spring Harbor T ~s ~, New York). Thus, gene transfer and eA~,~si,ioll methods are IIUlll~UU~ but P~.~Rnti~ y r.- ~ to intrûduce and eApress genetic n~ ri~ n cells.
Several of the above terhniqvp~ have been used t~ tranC~ ce hematopoietic or l~hoid cells, 30 inrl~ calcium pho:~h ~e l.,~r.~l;on (Berman et al., su~ra, 1984), pro~ fusion ~Deans et a~., supra, 1984), eleetlù~)~--';on (Cann et al., Oncogene 3:123, 1988), and ;-,r~ n with recombinant ade.luvilu~ rlcson et al., supra; Reuther et al., Mol. Cell. Biol. 6:123, 1986) adeno-associated virus ~aFace et al., supra) a~d lGh`ùvilu~ vectors (Overell et al., ~ o~,.c 4:1425, 1989). Primary T Iymphocytes have been ~ucc ~srully ~ c-l~ d by el~l~~û~;ol-to 94/28143 216 3 ~ 2 7 PCT/US94/05601 _9 _ (Cann et al., supra, 1988) and by 1~ vhal i.lr~;Lion (Nishihara et al., Cancer Res. 48:4730, 1988; Kasid et al., supra, 1990).

Construction of Selectable Fusion Genes S The sel~Prt~l~le fusion genes of the present hlvelllion c~ ; e a do~ anl positive select ' le gene fused to a negative sPIP~tq-hle gene. A s~lect~tle gene will generally c~ pl;ce7 for eY-q-mple, a gene ~ -r~l;~E a protein capable of cO~r~- h~ an antibiotic fe~ re phP~,ly~e or supplying an ~ hic l~uif~ ll (for ~O~ ' positive sPle~tion)7 or aclivali"g a toAic metabolite (for nc~;divt; selectinrl). A DNA se~uPnre e .r~;~g a bifimrtion^l fusion protein is constructed using r~--.bill~l DNA terhniq~ps to assemble s~p~- ~e DNA r.~ enro~ing a dorninqnt positive sel~live ~ene and a ne~alive ~el~Prt;qhle gene into an ~pfO~iale tAplcsSiOIl vector. The 3' end of the one selectable gene is ligated to the S' end of the other selectable gene, with the reading frames of the sP~u~pnr-p-c in frame to permit tr~clr~inn of the mRNA
sP~uenrPS into a single biologically active b;r~ ;ol-~l fusion protein. The sP-IPctq-hle fusion gene is expressed under control of a single plulllol~
The d()lninq-nt positive select-~le gene is a gene which, upon being trqnCdllrP~ into a host cell, t;A~,esses a domin^ -t phenotype pe- ...;llh~ positive spl~prtion of stable tr~qnc~uct~qntc.
The dominqnt positive sPIP~ ble gene of the present i~ lion is prer~.ably selected from the group co~ ling of the .u.Pllo~l~coside phospho~ r~ase gene (neo or aph) from TnS which codes for L~s;Cl~.re to the antibiotic G418 (Colbere~arapin et al., J. Mol. Biol. 150:1, 1981;
Southern and Berg, J. Mol. Appl. Genet. 1:327, 1982); and the I~ uycin-B
phosrh-)t. a, sre,d~e gene (hph or "hygron) which confers the sel~ti~' le phenotype of hyglo~ Cill ,~;~ e (Hrn ) (Santerre et al., Gene 30: 147, 1984; Sugden et al., Mol. cen Biol. 5:410, 1985; ob1i~h~ l~le from plasmid p~FRol, under ATCC Acc~ss;~-l- No. 39820).
Hyglwll~ B is an ~o~:ly~side antibiotic that inhibits protein synL~s;s by disf-l~l~6 translocation and pfol-wling ",;~ cl ~tion The *p* gene confers Hmr to cells ~ r,ed with the hph gene by phOSphO~ ~lalil~g . nd d~lùAiryh~g the ~libi~lic h~ v~ hl B. Other ~cce~ le dominqnt positive selectable genes include the following: th-e ba ~ ' neo gene ~nr~j~
neomycin phos~ho~ ce (Beclc et al., Gene 19:327, 1982); the ~ h~ guanine phosphorib l~ L- ~ce gene (gpt) from E. coli enrocling f ~ 7;c';--~ce to l~lyCu~k .olir acid (Mnllig~n and Berg, Proc. Natl. Acad. Sci. US,l 78:2072, 1981); the diLy~lr~rulate red~ct~ce (E)HFR) gene from murine cells or E. coli which is ~ cc ~y for biosylllLe~ of purines and can be compelilively i~il; it~d by the drug lll.,I]lolr~Aald (~X) to select for cells co~lilulively tApfessing L,creased levels of DHFR (Si.~onc~ -- and Levinson, Proc. Natl. Acad. Sci. USA

Wo 94/28143 PCT/US94/05601 ~
2~ 6342~
~0:2495, 1983; Simonsen et al., Nucl. ~cids Res. 16:2235, 1988); the S. typhimurium hi.ctiriirln dehydrogenase (hisD) gene (Hartman et al., Proc. Natl. Acad. Sci. USA 85:8047, 1988); the E.
coli L.~phan synthase ,B subunit (trpB) gene (Hartman et al., supra); the ~urulllycin-N-acetyl l.d~ , ce (pac) gene (Vara et al., Nucl. Acids Res. 14:4117, 1986); the ad~P-nncinP- de=-~;nA~e S (ADA) gene (Daddona et al., J. Biol. Chem. 259:12101, 1984); t_e multWrug re~ nre (MDR) gene (Kane et al., Gene 84:439, 1989); the mouse ornithinp decarboAylase (OCD) gene (Gupba and Coffino, J. Biol. Chem. 160:2941, 1985); the E. coli ~ ld~e ha.)sc~ba~ ase catalytic subunit CpyrB) gene (Ruiz and Wahl, Mol. cen. Biol. 6:3050, 1986); and the E. coli asnA gene, enr4ding ~ nA synthPt~ce (Cartier et al., Mol. Cell. Biol. 7:1623, 1987).
The negative sP~ect^'~!e gene is a gene which, upon being tranc~ucPd into a host cell, tA~r~sses a p}ic .lul~c pe- ..;~ l ;ng ne~a1i~,é sclc~lion (i.e., el h~ ;on) of stable I ~. n~ c l~n~;
The pfeîeired negative selectable gene of the present ih~ ion is the ba. leridl CD gene Pn~orling cytosine de~min~ce (Genbank accession number X63656) which confers 5-fluorocytosine sensitivity.
Other e~ suitable for negative sele~ti~n include, but are not limited to, al~aline pho5rh~t~ce useful for COIl~e~ g phl?sph~ g prodrugs such as etoposide-phosl h ~e, doxorubicin-phosrh~te, milul~h~ phosrh~t~, into toAic dephosphorylated metabolites;
arylc~lf~ce useful for ~ll~e.lillg sulfate~-~ g prodrugs into free drugs; ploledses, such as serratia protease, thermolysin, subtilic-in~ CaL1~A~C~ PS and c~thPpsinc (such as cq-thepsinc B and L), that are useful for c~ e~lhlg peptide~ .g prodnugs into free drugs;
D-alanylcalboAy~epti~q-cPs useful for co.,v~lin& prod,..gs that contain D-amino acid sl~hstituPntc; call,ohydla~cleaving e~Ly~es such as ,B-gdlq-~.r~ e and ~ " "h~;~qce useful for cû-lv~ g glycosylated prodrugs into free drugs; ~B-lq- ~ q~e usefill for col~ g drugs d~.iv~liL~d with ,B-lactams into free drugs; and pPnirillin ~ P~ such as peni~illin V amidæe 25 or peni~illin G ;qmi~l~ce~ useful for cû--~elling drugs de.ivàli~d at their amino nillu&~ with phenoAyac~tyl or phe~lac~yl groups, l~e,~ively, into free drugs.
Other enzyme prodrug colll~ alioDs include the b~ ; q~ (for e~mpl ~7 from Pseudomonqc) enzyme ~l~uAy~ ace G2 with the prodrug para-N-bis(2-chloroethyl) alllhlOb~..~yl glutqmi~ acid. Cleavage of the ~ r acid moiety from this cv~uulld30 releases a toxic benzoic acid mustard. Penicillin-V amidase will convert phenoAya~ P
derivatives of doxorubicin and mPIphqlq-n to toxic ~Pt~holitp~s.
Due to the de5~ ,.a~ of the genetic code, there can be ~n~;de~le variation in nucleotide se~uPnres enro~ing the same amino acid ses~uence; ~-~mp!q-y DNA emho~impntc are those coll~ponding to the nucleotide se~nPn~ P-C in Se~uPnre Listing No. 1. Such variants 0 94/28143 21 63~2 7 PCT/US94/05601 will have modified DNA or amino acid sç~lAncPs, having one or more ~ l;ons, fie,letionc~
or additions, the net effect of which is to retain biolcgir l activity, and may be s ~bs~ 'lrd for the specific se~uAnrP~ rlose~ herein. The sequ~nr~ of se1ec~' le fusion genes COlllp~ illg CD and neo are equivalent if they contain all or part of the se~ rPc of CD and neo and are S capable of hybridizing to the nl-rleotif~e se~ e of Sequ~Pnre Listing No. 1 under moderately sllhlgelll conditions (50C, 2 X SSC) and express a biologically active fusion protein.
A "biologically active" fusion protein will share s~lffirip~nt amino acid sequP-nre similarity with the specific em~iimPntc of the present invention disdosed herein to be capable of confe~ g the selectable pht;no~ cs of the component sAIect^~le genes.
In a plerelled embodiment, se~lenr~Ps from the ba.,lelial ~lushlc ~ ce (CD) geneare fused with sequenrPs from the bacterial n&~lll~hl ph~ph')! ~n~r dse (neo) gene. The resulting selectable fusion gene (referred to as the CD-neo selectable fusion gene) encodes a bifi-nrtinn-ql fusion protein that confers G 118 and 5-GC and provides a means by which domin-q-nt positive and ne~;di~e sP~ le ph&.loly~cs may be expressed and regul~fed as a 15 single genetic entity. The CD-neo sP,le~l)le fusion gene may be espe~iqlly advantageous in patient populations likely to receive ganciclovir.

RACOI~;ndn~ EA~iession Vectors The selPrt~l 1e fusion genes of the present h.~,e.llion are utilized to identify, isolate or çlimin~fe host cells into which the sele~f '-le fusion genes are introduced. The selç~f.~le fusion genes are introduced into the host cell by trqn~ ring into the host cell a recombinant expression vector which CQllli~;n~ the sP~I~P~al le fusion gene. Such host cells include cell types from higher eukaryotic origin, such as l"_-ll",~ or insect cells, or cell types from lower prokaryotic origin.
As in~1ir fPd above, such sele. l~' le fusion genes are p,.fel~bly introduced into a particular cell as a ~ ~n~ -1 of a reco~~ nt ~r~ vector which is capable of e.~,~sh,g the s~lP~ le fusion gene within the cell and co--fi ~ g a sel~f~' le ph~ c.
Such recolllbilld~ Apr~;,ioll vectors generally include ~ldll~ic or natural nurl~Qti~le s~P~uçnrPS
COlllpli;~il.g the sPIecf~le fusion gene operably linlced to suitable ' - irfio~' or I -~^fi~n~l control se4u~Pnrçs for eAample, an origin of replir~fion optional Op~d10~ selluPnr~PS to control l.~ls~,iplion, a suitable plOIllul~ and ~ ,~ - ,r_ Iinked to the gene to be eA~,~ssed, and other 5' or 3' fl~ lWng n~n~ .c. ibed se4uPnrP-s, and 5' or 3' ~ cl~fP~ sc~ r-s, such as nP,ce~ ,oso---r binding sites, a polyadenylation site, splice donor and ~rCcy~r sites, and C~ f.. ;n,-~ sc4u---~r~s. Such regulatory sc4u~ --r~s can be derived from WO 94/28143 2 ~ ~ 3 4 % 7 -12- PCT/US94/05601--mqmmqliqn, viral, microbial or insect genes. Nucleotide se~UPnrPs are operably linked when they are fimctionqlly related to each other. ~or eYqmple, a p-u-lloler is operably linked to a selectable fusion gene if it controls the h~,c-i~lion of the sele.,~l~le fusion gene; or a lil)osolllc binding site is operably linked to a sPIectq-hle fusion gene if it is pos;ti~n~d so as to permit 5 trn~l ~tio~ of the selectable fusion gene into a single bir~ n~l fusion protein. Generally, operably linked means corlti~
Specific reco---bi..~l e~cpression vectors for use with ---------~li_n, ba^te~i-q-l, and yeast cellular hosts are described by Pouwels et al. (~oning Vectors: A ~or~ M~
Elsevier, New York, 1985) and are well-lcnown in the art. A detailed ~ ;ol- of 10 recombinant expression vectors for use in animal cells can be found in Rigby, J. Gcn. Virol.
64:255, 1983); Elder et al., Ann. Rev. Genet. 15:295, 1981; and Sub~ i et al., ~nai.
Biochem. 135:1, 1983. Appru~ lt; recombinant expression vectors may also include viral vectors, in particular r~ vhllscs (~ sfd in detail below).
The selectable fusion genes of the present invention are pl~,rerdbly placed under the 15 transcriptional control of a strong e~h~nrP~ and p~u--lùler expression e~c~e~tP. E~amples of such expression c~e~s include the human c~lu~ g~ virus i..,...~ c~ly (HCMV-IE) p~oll~ol~r (Boshart et al., Cell 41:521, 1985), the ,B-actin prol..ul~. (Gunning et al., Proc. Natl.
Acod. Sci. USA 84:5831, 1987), the histone H4 pro...ole. (Guild et al., J. Virol. 62:3795, 1988), the mouse met~lloll.io~ p.ul..~lci (McIvor et al., Mol. Cell. Biol. 7:838, 1987), the rat growth hormone p~o---ul~ (Miller et al., Mol. Cell Biol. 5:431, 1985), the human arlPnosinp de~ ce pro...ote~ (HaL~;Lap lûS et al., Proc. Na~l. Acad. Sci. USA 86:3519, 1989) the HSV
TK promoter (Tabin et al., Mol. Cell. Biol. 2:426, 1982), the a-1 ~Lil,~lJsill çnh~n~e~ (Peng et al., Proc.Natl.Ac~d.Sci.USA 85:8146,1988)andtheil~ uilo~;lobulin e h~ P~ ler (R!2nlrPn.ctçin~ et al., Nucleic Acid Res. 16:10939, 1988), the SV40 early or late p~lllV~.~, the 25 Adel ov-lus 2 major late proulul~l, or other viral plO--wl~.~ derived from polyoma virus, bovine papilloma virus, or other retroviruses or ad~,.lovi u~cs. The plu~ ,. and e .h~ncpv~
elements of immlmqglûbulin (tg) genes confer marlced srecifi~ity to B lymphocytes (13anerji et al., Cell 33:729, 1983; Gillies et al., Cell 33:717, 1983; Mason et al., Cell 41:479, 1985), while the elements controlling t-~u~s~ iûn of the ~B-globin gene r.--- l;on only in e..y L~id cells (van ~çn~lelR et al., Cell 56:969, 1989). Using well-known ~ ';OI- and ligation terhni~lu~s, a~p-(3pli~1e transcliyliûnal control s~-,on~s can be e~cised from various DNA
sources and i~lr~ ed in o~-dive reldiol~ship with the intact. sPIP~ 1e fusion genes to be expressed in accordance wi~ the present ill~6llliOA. Thus, many l,~s~ lional control 21 63~27 ~ 94/28143 PCT/US94/05601 sequPnr~PS may be used succ~P-~fully in rdlluvilal vectors to direct the e~cpression of inserted genes in infected cells.

Retroviruses Retroviruses can be used for highly efficiPnt trar~llction of the sPIectq-llle fusion genes of the present h~ve~llion into eukaryotic cells and are pr.,f~,L~I for the delivery of a selectable fusion gene into primary cells. Moreover, I~`UVila~ l talces place in a controlled fashion and results in the stable i ~ n of one or a few copies of the new genetic i~rc~ n per cell.
R~Lroviluses are a class of viruses whose genome is in the form of RNA. The genomic RNA of a lc:~rOVilui~ contains ~Tans-acting gene s~u~ ~e~ coding for viral proteins, inclu~
structural proteins (enr~ed by the gag region) that associate with the RNA in the core of the virus particle; reverse l-~c-i~t~e (en~o~ed by thepol region) that makes the DNAc4mrlemPnt, and an envelope ~ Iyco~ lein (en~ocled by the env region) that resides in the lS lipopioLein envelope of the particles and binds the virus to the surface of host cells on infection.
~PE!lic~tion of the retrovirus is re~liqtPd by cis-acting elc.l~cn~, such as the prol~ ~ for trqm~rirtion of the proviral DNA and other nllrleotide s~u- n~Ps ,~ c~ ~ for viral r~plicqtion The cis-acting elem~ntC are present in or ''~jR~ ~I to two itlenti-~q~ untran~ P~d long tP-rminql repeats (LTRs) of about 600 base pairs present at the S' and 3' ends of the r~ ovilal genome. Re;llovi~ replicate by copying their RNA genome by reverse 1- .~-.c. ;l~l;on into a double-stranded DNA ;--1r--- P 1;~'~, using a VilU_ e ~odeA, RNA~irected DNA polyl.~ se, or reverse ll -CC~ e The DNA ;--~ Pd;-~e is; ~ 'Pd into cl~o---oso---~' DNA of an avin or l"~"""~liqrl host cell. The ;.~IL.gr.~ed I~UV~ DNA is called a pluvhuS. The pr~vilus serves as te~ 'e for the s~lllhcsis of RLNA chains for the 1~ ;n-` of hlr~iliùui~ virus particles. Forward !~ c ~ r of the prUVillli~ and assembly into inrL~ J~s virus particles occurs in the pr~cnce of an ~pr~liale helper virus having endogenous trans-acting genes r~uir~ for viral re~lir~fiûn RelLrùvilu ,~ are used as vectors by r-P~I~ ng one or more of the ~dGgu~ui~ trons-acting genes of a proviral form of the r~UVilllS with a r~ l)in~ull th~ fi~ gene or, in the case of the present invention, a sclP~ble fusion gene, and then trqn~rlllcing the ~ hlall~
provirus into a cell. The trans-acting genes include the gag, pol and env genes which encode, respectively, proteins of the viral core, the enzyme reverse l~r ~c- ;pl~ce and cQnctihlpntc of the envelope protein, all of which are nr,C~ for production of intact virions. Rccû."~m~l retroviruses defi~iPnt in the trans-acting gag, pol or env genes cannot synthP~i7e e~P ~fi~l WO 94/28143 21~ 3 4 ~ 7 PCT/US94/05601---1~
proteins for replication and are acco~li~l~ replication-defective. Such replication~efective recombinant retroviruses are ~l) ~;d~i using pa"~L ~;~ cell lines. These parL ~i.~g cell lines contain hllt;g~ ed ~hovilal g~no~es which provide all trans-acting gene se~l~.e~re~ l-PreD~-. y for production of intact virions. Proviral DNA se~ 5 which are ~ c.J"~r~1 into such 5 p~ ng cells lines are l,~c-ibed into RNA and enr~s~ Pd into i.~rl~i.JU~ virions c~ ;..i..g the sP~ect~hle fusion gene (andtor ~ gene), but, lac~ing the ~ans-acting gene products gag, pol and env cannot synthesize the nP~P~c~- ~ gag, pol and env proteins for enrarsi~ting the RNA into particles for ~ r~~ g other cells. The resulting infectious retrovirus vectors can therefore infect other cells and ;~1r~ r~ a sP~Ir~ e fusion gene into the cellular DNA of a host cell, but cannot replicate. Mann et al. (Cell 33: 153, 1983), for example, describe the development of various pa ~L~j"g cell lines (e.g., ~2) which can be used to produce helper virus-free stoclcs of r~,l,bina,ll r~_h~ovi~us. Fn~rsi~ i()n in a cell line h~lw.i"g Irans-acting elements enr~ding an C;C~I.u~iC viral envelope (e.g., ~2) provides ecotropic ~imited host range) plOgG~ virus. A~ dively~ assc,.-bl~ in a cell line c ~
~ -pho!~upic par~ g genes (e.g., PA317, ATCC CRL 9078; Miller and nu~ , Mol.
cen Biol. 6:2895, 1986) provides au,~ho~pic (broad host range) progeny virus.
Nullle~ous p-uVif~s co~l.u-,ls have been used ~lcc-7-r~ y to e~cpress foreign genes (see, e.g., Coffin, in Weiss et al. (eds.), RNA Tumor Viruses, 2nd Ed., Vol. 2, (Cold Spring Harbor Làbû~alO~)r, New York, 1985, pp. 17-71). Most proviral elPm~nt~ are derived from 20 murine retroviruses. Re1luvh..~es a~-q~rt~' le for use in acco~ance with the present h~ iùn can, how~ver, be derived from any avian or ,- "".~ ., cell source-. Suitable .~ovi u,es must be capable of i..~ cells which are to be the reci~ie"~ of the new genetic n-~riql to be trq~ red using the r~ovilal vector. F~qmpl~ 5 of suitable nhùviluSeS include avian retroviruses, such as avian e~ ublq~ocic virus (AEV), avian lell~nsi~ virus (ALV), avian 25 myPIobl~lJ~ virus (AMV), avian siil~-llâ virus (ASV), I~ujinau,i s~.lla virus (FuSV), spleen necrosis virus (SNV), and Rous sarcoll-a virus (RSV); bovine l~ n-iq virus (BLV);
feline .._t~ovi,l.ses, such as feline le~ rn;q virus (FeLV) or feline sarcoma virus (FeSV);
murine r~uvhubes~ such as murine lelllrPmiq vws (MuLV); mouse ~ ~mor virus (MM'rV), and murine sarcoma virus (MSV); and primate .~hovh~es~ such as human T-cell 30 Iy~ hol~(Jpic viruses 1 . nd 2 (HTLV-1, and -2), and simian sa~ a vi~us (SSV). Many other suit. ble retrovimses . re known to those s~illed in the art. A ~ n-~ of ~ Ovil~s is p~ovided by Teich, in Weiss et al. (eds.), RNA TlLmor viruses 2d ed., Vol. 2 (Cold Spring H rbor Labo.àlo.y, New York, 1985, pp. 1-160). P~f~r~ l~Jvi.~..,cs for use in co-~nr~ n with the present invention are the murine r~uv,.. se~ known as Moloney murine le-~Pm;q 0 94/28143 21 6 3 ~ 2 7 PCT/US94/05601 virus (MoMLV), Moloney murine sarcoma virus (MoMSV), Harvey murine sal~a virus (HaMSV) and Kirsten murine sarcoma virus (KiSV). The se~uenrPes l~uif-,d to construct a retroviral vector from the MoMSV genome can be obtained in conjunclion with a pBR322 plasmid se~uP~ n~e such as pMV (ATCC 37190), while a cell line producer of KiSV in K-BALB
5 cells hæ been deposited as ATCC CCL 163.3. A deposit of pRSVneo, derive~ from pBR322 inrlv~ the RSV LTR and an intact ne~ drug r~ e marlcer is available from ATCC
under AccP~cs;on No. 37198. Plasmid pPB101 cc~ the SNV genome is available as ATCC 45012. The viral genolllPs of the above r~oviluSeS are used to construct replic~iQn-defective retrovirus vectors which are capable of h~ h~g their viral gc,lûl--es into the 10 chromosomal DNA of an infected host cell but which, once i~ are incapable of replic~tion to provide hlre~iliuus virus, unless the cell in which it is introduced cont~inc other proviral elRm~ntC Pn~ing fiJnc`tion~l active trans-acting viral p-ùleil s.
The selectable fusion genes of the present ihlve.ltion which are ~l~nc~lu~ed by retroviruses are expressed by placing the sPIe~ !e fusion gene under the tl~--C~ on~l control 15 of the enh~nrPI and prol,~ter ihleol~laled into the l~:ll'OVilal LTR, or by placing them under the control of heterologous l,~"c. ~;l.l;on~l control se~--e~r~c inserted b~ the LTRs. Use of both heterologous transcriptional control se~u~n~P~s and the LTR ll~s~ iollal control se~uPnrPS enables CO~ SSiOI~ of a therapeutic gene and a sP-l~ble fusion gene in the vector, thus allowing selection of cells ~Apr~ g specific vector seqllenrPc enr~i~ the desired 20 thsr~reuti~ gene product. Ol,tai~g high-level eA~-essio~ may require placing the th~ eu~ic gene and/or sele~t~hle fusion gene within the l~llUV-l~ under the I. nc. ~ l;on~l control of a strong heterologous e~ fP~ and pro~olèr eA~ ion c~Cse~tP~. Many dirLie~t heterologous enh~nl~Pr.c and pru~lûle.~ have been used to e~press genes in 1~VVi1dI vectors. Such ~
or plollw~ can be derived from viral or cellular sources, il-~l",~ g ..~ n ~cw~s, and 25 are pre~.ably cQr~s~ V-e in nature. Such heterologous ~ ,c~ -t;o..~' control se~1enc~Ps are tiicc~cced above with r~fef~l.ce to l~OI-Ibi~ l eAL~L~c3sion vectors. To be eAl,r~s~ in the tr~nc-l~lre~ cell, DNA se~uPn~es introduced by any of the above gene transfer mPtho lc are usually eA~.~ssed under the control of an RNA polylll~dse II plu~ er.
Particularly pref~,.l~ recombinant eA~ vectorc include pLXSN, pLNCX and 30 pLNL6, and derivatives thereof, which are ~P ~- ;bed by Miller and Ptlcm ~n, Ri~teC~tniq~ s 7:980, 1989. These vectors are capable of tAplessi~ heterologous DNA under the !- ~ns~ ~ ;pl;on~l control of the .~Ov~lal LTR or the CMV pro~ol~,., and the neo gene under the control of the SV40 early region pru-l~Olel or the l~lùvhal LTR. For use in the present invention, the neo gene is replaced with the bi~ul-~lional s~ç~l~le fusion genes (1ic~losed WO 94/28143 2,~.G3 4~ PCT/US94/05601--herein, such as the CD-neo scle~l~ble fusion gene. Construction of useful re~lir~tion~f~Live retroviruses is a matter of routine skill. The resulting ~O~ in~l ~huvi-~P~ are capable of integration into the ~.Loluosolnal DNA of an infected host cell, but once ;~ eA arè
inr~ra~le of rPplir~ion to provide ;~ ;u~s viru_, unless the cell in which it is introduced S contains another proviral insert Pnro.lii~g fimrtionqlly active trans-acting viral proteins.

Uses of Bifunction-q-l Selectable Fusion Genes The sclP~ le fusion genes of the present i ~ t;on are particularly prt;ît;,red for use in gene therapy as a means for idc.lli~ing, jc"l~i~ or e~ ;ng cells, such as somatic cells, 10 into which the selectable fusion genes are introduced. In gene therapy, somatic cells are removed from a patient"~ c~luçed with a ~ecou,bin.ull eApression vector co.~ .;ng a ther~rel~tir gene and the s~ k~t: ' le fusion gene of the present invention, and then reintroduced back into ~e patient. Somatic cells which can be used as vehicles for gene therapy include hematopoietic ~one Ill~ru~. derived) cells, L ~fi~ yl~r" hepatocytes, endothelial cells and fibroblasts (Friedman, Science 244:1275, 1989). Allt;~ liv~ly, gene therapy can be arcomrli~hed through the use of inject~ble vectors which l- ~ ce somatic cells in vivo. T~e feasibility of gene transfer in humans has been demon~ '~ (Kasid et al., Proc. Natl. ~cad.
Sci. USA 87:473, 1990; Rosenbelg et al., N. Engl. J. Med. 323:57Q, 1990).
The selectable fusion genes of ~e present h~ lioll are particularly useful for 20 e~ g gPnetic~lly mo lified cells in vivo. In vivo el~ fion of cells tA~ tSillg a nc 5~live selectable phenotype is particularly useful in gene therapy as a means for ablating a cell graft, thereby providing a means for ~ ing the gene therapy procedure. For e~-q-mple, it has been shown that a-l".i";-l- dion of the anti-herpes virus drug g~q~riclovir to ll~s~_nic animals t;A~ressing the HSV-I TK gene from an ;--~ ogloblllin plu~lel results in the sele~live 25 ablation of cells ~A~le~si~.g the HSV-I TK gene (EIeyman et al, Proc. Natl. ~lcad. Sci. US,~
86:2698, 1989). Using the same tl~sg~,nic mice, GCV has also been shown to induce full ~eglcssion of Abelson le~ miq virus-induced l~ p~ q~ (~fooll~,~ et al., H~Lman Genc ~herapy 1: 125, 1990). In a third study, in which a murine salcu~a (K3T3) was infected with a retrovirus t;Ayressin~ HSV-I TK a~d tr?~spl ~ed into s~ neic mice, the tumors induced by 30 the sarcoma cells were compl~tely e~?~ d ~ollowing l~ea~nl wi~ GCV (M~olten and Wells, J. Natl. Cancer Inst. 82:297, 1990).
The selectable fusion genes of the present iu~ io~ also are ~cn~ r.l :~' in tumor ablation therapy as it has been pr~ti~ed by Oldfield et al., Human Gene lherapy 4:39, 1993.
p~ ~jn~ cells (about 106 109) pr~lu-,iLg the tgLS(+)CD-neo l~hUVi~;~l vectors are 21 63~27 ~ 94/28143 PCT/US94/05601 inoculated intra-tumorally. After a period of several days, during which the newly produced lf~tLuViLus~s infect the v~lj7r~ ~l rapidly ~-owL.g tumor cells, the patient is given about 5~200 mg of S-FC per kg body weight (orally or ihllLa~ ou; ly) daily (when the tgLS(+)CD-neo retroviral vector has been used) to selectively ablate the infected tumor cells.The bir,.fi~ l;f~nql selectable fusion genes of the present ul~e~ ion can also be used to f~ilitqte gene m :~-lifif- -~ion by homologow 1~4"'~i" ' ;f r Reid et al., Proc. Na~l. Acad. Sci.
USA 87:4299, 1990 has recently dP~ il,ed a two-step procelu.e for gene mfndifif-q~ion by l omologoll~ recor-h~ ;f~n in ES cells (nin-outn hon~olrgow o:~Lubi~lion) wing the HPRT
gene. Briefly, this procedure iuvol~s two steps: an f'inf' step in which the HPRT gene is embedded in target gene seqU~pnrps~ transfP~ted into HPRT host cells and homologous recombinants having iucol~,oldled the HPRT gene into the target locus are iriPntifiPA by their growth in HAT ~m~PAillm and genf~mir analysis using PCR. In a second "out" step, a construct c4n~ g the desired re~l ce--~f;ul se4u~ ~PS embedf~f~PA. in the target gene se~uRnf-Ps (but without the HPRT gene) is t.-~ Cr~pd into the cells and homologous r~lllbiu.3ûLs having the re~!~rPm~nt sequ~Rnres (but not the HPRT gene) are isolated by nc,galiv~ selection against HPRT~ cells. ~lthough this pr~cPAIul~e allows the i..1.~1ucl;o~ of subtle ..---l; ~;o.-.~ into a target gene without ih~lL~nlucil,g se~t-~' le gene sP~qu~.rP~ into the target gene, it requires positive selection of L,~n,ro,,..~ in a HPRT cell line, since the HPRT gene is rec~;,i~ for positive selection. Also, due to the ;~l rl;, en~ iOIl of the HPRT gene in ES cells, it is n~Pc~
20 to use a large 9-kbp HPRT mini-gene which comI lir~s the co~LLu~lion and ~,a2;a~;on of homologous l~coulbhlalion vectors. The sPlert~')le fusion genes of the present i- venlioLL
provide an improved means whereby "in-out" homologous r~ io~- may be pPlr~nulPd.Rerq~l~e the sclP~L,~le fusion genes of the present hl~ t;on are ~ l for positive sPlf~
any wild-type cell may be used (i.e., one is not limited to use of cells ~efici~pnt in the s~PlF~ hle 25 phenotype). Moreover, the size of the vector col~ the sPlccl;~ble fusion gene is reduced ~ignifirqrltly relative to the large HPRT mini-gene.
By way of i~ ;o.~, the CD-neo sP~ ble fusion gene is used as follows: In the first "in" step, the CD-neo s_le~L~le fusion gene is em~-P~Pd in target gene se~lPnres l..,h~re~l~ into a host cell, and hom-)logous rxu~ having inc~o~at~ the CD-neo 30 sP-IF~l ble fusion gene into the target locus are identifiP~ by their growth in m-Pdillm c~ g G 418 followed by genome analysis using PCR. The CD-neo cells are then used in the second "out" step, in which a co~llu~ the desired re~l~r~ 1 se~1., .rf~ emhPdd in the target gene SC~I~-P~f s ~DUt without the CD-neo s~l';lble fusion gene) is "~ rr~i into 22627/~000346l 147700 2163~27 the cells. Homologous recombinants are isolated by selective elimin~tion of CD-neo cells using 5-FC followed by genome analysis using PCR.

EXAh~PLES
FY~ m Ple 1 Construction and Characterization of Plasmid Vectors Containing CD-neo Selectable Fusion Gene A. ConstructiQn of the Bifimctional CD-neo Selectable Fusion Gene.
Plasmid tgCMV/hygro/LTR (Figure 1) consists of the following elements: the BalI-SstII
fragment co~t~ining the HCMV IE94 promoter (Boshart et al., Cell 41:521, 1985); an oligonucleotide containing a sequence conforming to a consensus translation initiation sequrnre for m~mm~ n cells (GCCGCCACC ~ (SEQ ID NO:5) (Kozak et al., Nucl. Acids Res.
15:8125, 1987); nucleotides 234-1256 from the hph gene (Kaster et al., Nucl. Acia's Res.
15 11:6895, 1983), encoding hygromycin phospholld"s~,ase; sequences from nucleotide 7764 and through the 3' LTR of MoMLV (shinnirlr et al., Nature 293:543, 1981), containing a polyadenylation sequ~nre: the NruI-AlwNI fragment from pML2d (Lusky and Botchan, Nature 293:79, 1981), cont~ining the bacterial replication origin; the AlwNI-AatII fragment from pGEM1 (Promega Corp.), containing the ~ e gene.
Plasmids tgCMV/neo, tgCMV/CD, tgCMV/CD-hygro, tgCMV/neo-CD, and tgCMV/CD-neo are all similar in structure to tgCMVlhygro/LTR and contain the consensus ~r~n~l~tion initiation sequenrer however, each contains different sequtonrp~ in place of the hph sequçnr~. Plasmid tgCMV/neo contains an oligonucleotide encoding three amino acids (GGA
TCG GCC) (SEQ ID NO:7) and nucleotide 154-945 from the bacterial neo gene encoding 25 neomycin phospho~,a~ .dse (Beck et al., Gene 19:327, 1982), in place of the h~h sequenres.
Plasmid tgCMV/CD contains nucleotides 1645-2925 from the bacterial CD gene encoding cytosine d~amin~ce (Genbank accession number X63656), in place of the hph sequences. The CD sequenc~s were amplified by PCR from plasmid pCD2 (Mullen et al., Proc. Na~l. Acod.
Sci. ~SA 89:33, 1992). Plasmid tgCMV/hygro-CD contains nucleotides 234-1205 from the 30 hph gene fused to nucleotides 1645-2925 from the CD gene in place of the )~ph sequences.
Plasmid tgCMV/CD-hygro contains nucleotides 1645-2922 from the CD gene fused to nucleotides 23~1256 from the hph gene in place of the hph sequ~onces. Plasmid tgCMV/neo-CD contains ~n oligonucleotide encoding an additional three amino acids (GGA TCG GCC) (SEQ ID NO:7) and nucleotides ''''6'~7/"000346/147;'00 _ ~163~27 154-942 from the bacterial neo gene fused to nucleotides 1645-2925 from the CD gene in place of the hph sequences. Plasmid tgCMV/CD-neo contains nucleotides 1645-2922 from the CD
gene fused to nucleotides 154-94~ from the neo gene in place of the hph sequences.
Plasmid tgCMV/hygro/LTR was constructed using standard techniques (Ausubel et al., Glr,~-ent Protocols in Molecular Biology (Wiley, New York), 1987) as follows: Plasmid HyTK-CMV-IL2 was constructed ~irst by ligating the large HindIII-StuI fragment from tgLS(+)HyTK
(Lupton et al., Mol. Cell. Biol. 11:3374, 1991) with the HindIII-StuI fragment sp~nning the HCMV IE94 promote~ from tgI_S(-)CMV/HyTK (Lupton et al., supra, 1991), and a t`ragment containing human IL-2 cDNA sequences. The fragment co~t~ining human IL-2 cDNA
sequenr~s was amplified from a plasmid containing the human IL-2 cDNA by PCR using oligonucleotides 5'-CCCGCTAGCCGCCACCATGTACAG&ATGCAACTCC-3' (SEQ ID
NO:8) and 5'-CCCGTCGACTTAATTATCAAGTCAGTGTT-3'(SEQ 1D NO:9). Following arnplification, the PCR product was first treated with T4 DNA polymerase to render the ends blunt, tben digested with N'neI, before ligation to the fragments from tgLS(+)HyTK and tgLS(-)CMV/HyTK. To generate plasmid tgCMV/hygro/LTR, the SalI-PvuI fragment sp~nning the SV40 polyadenylation signal of tgCMV/hygro (Lupton et al., supra, 1991) was replaced witb, the SalI-PvuI fragment com~ining the Moloney lel-k~mi~ virus LTR (which contains the retroviral polyadenylation signal) from HyTK-CMV-lL2.
Plasmid tgCMV/neo was constructed using standard te~hni~ues (Ausubel et al., supra, 1987) as follows: A PvuI-NheI fragment sp~nning the HCMV IE94 piuu~ er from tgCMV/hygro was ligated to a NheI-Hindm Ga~ Sp~nning the neo gene from tgLS(+)neo (the Hindm site was treated with T4 DNA polymerase to render the end blunt) and ligated to SalI-PvuI fragment co.l~hlil.g the Moloney lellkRnni~ virus LTR (which contains the ~ vhal polyadenylation signal) from HyTK-CMV-IL2.
Plasmid tgCMV/CD was constructed using standard techniques (Ausubel et al., supra, 1987) as follows: A PvuI-Nhel fragment sp~nning the HCMV E94 promoter from tgCMV/hygro was ligated to â synthetic DNA fragment (pl~,~ed by ~nnP~Iing oligonucleotides ~'-CTAGCCGCCACCATGTCGAATAACGC~ ACAAACAATTATTAACGCCCG-3' (SEQ
ID NO:1ûj and 5'-GTAACCGGGCGTTAATAATTGTTTGTAAAGCGTTATTCGACATG&TGGCGG-3') (SEQ ID NO: 11), the BstE2-AluI fragment con~ining the r~nl~inri~r of the CD coding region from pCD2 (Mullen et al., Proc. Na~l. ACad. Sci. USA 8~.33, 1992), and the SalI-PvuI
fragment containing the Moloney le -k~mi~ virus LTR (which contains the retroviral polyadenylation signal) from _ 6~7/~000346l 147700 ~163427 ,,o HyTK-CMV-IL2. The SalI sile in the latter fragment was treated with T4 DNA polymerase to render the end blunt before ligation.
Plasmid tgCMV/CD-hygro was constructed using standard techniques (Ausubel et al., supra, 1987) as follows: The large ClaI-SalI fragment from tgCMV/CD was ligated to a Cla~-NcoI fragment amplified from tgCMV/hygro by PCR using oligonucleotides 5'-CCCATCGATTACAAACGTAAAAAGCCTGAACTCACCGCGAC-3' (SEQ ID NO:12) and ~'-GCCATGTAGTGTATT&ACCGAl-rCC-3' (SEQ ID NO:13) (the PCR product was digested with ClaI and NcoI before ligation), and an NcoI-SalI fragment containing tne rem~in~er of the hph coding region from tgCMV/hygro/LTR.
Plasmid tgCMV/hygro-CD was constructed using standard techniques (Ausubel et al., supra, 1987) as ~ollows: The large SpeI-BstE2 fragment from tgCMV/CD was ligated to a SpeI-Scal fragment containing the hph coding region from tgCMV/hygro/LTR, and a synthetic DNA fragment (prepared by ~nnP~ling oligonucleotides 5'-ACTCTCGAATAACGCmACAAACAATTATTAACGCCCG-3' (SEQ ID NO:14) and 1~ 5'-GTAACCGGGCGTTAATAATTGmGTAAAGCGTTATTCGAGAGT-3') (SEQ ID
NO~
Plasmid tgCMV/CD-neo was constructed using standard tprhnir~ s (Ausubel et al., supra, 1987) as follows: The large ClaI-Asp718 fragment from tgCMV/CD was ligated to a synthetic DNA fragment (p,e~ared by annealing oligonucleotides 20 ~'-CGATTACAAACGTATTGAACAAGATGGATTGCACGCAGGTTCTCC-3' (SEQ ID
NO:16) and 57-GGCCGGAGAACCTGCGTGCAATCCATCTTGTTCAATACGmGTAAT-37) (SEQ ID
NO:17), and an Eagl-Asp718 fragment containing the rern~in-l~r of the nco gene coding region from tgCMV/neo.
2~ Plasmid tgCMV/neo-CD was constructed using standard techniques (Ausubel et al., supra, 1987) as follows: The large SphI-SalI fragment from tgCMV/neo was Iigated to a ClaI-NcoI fii~gmPnt ~mrlified from tgCMV/neo by PCR using oligonucleotides 5'-CGAACTGTTCGCCAG&CTC-3' (SEQ ID NO:18) and ~' -CCCGGTAACCGGGCGTTAATAATTGmGTAAAGcGTTATTcGAGAA
30 GAACTCGTCAAGAAG&C-3' (SEQ ID NO: 19) (t'ne PCR product was digested with SphI and BstE2 before ligation), and a BstE2-SalI fragment containing the rem~in~r of the CD gene coding region from tgCMV/CD.

''''6 ~/''000346/14/700 B. Dominant Positive Selection of Cells containin~ CD Fusion Genes~
To demonstrate that the CD fusion gene encodes both neo and hph activities, the frequencies with which the various plasmids conferred drug resistance in NIH/3T3 cells were determined.

_ ~ 94/28143 216 3 4 2 7 PCT/US94/05601 First, NIH/3T3 cells were grown in Dulbecco Modified Eagle MeAillm (DMEM;
available from Gibco Labord~-ies) snrplc~ PA with 10% bovine calf serum (Hyclone), 2 mM L-~ e, 50 U/ml pçnirillin~ and S0 ~g/l Shey~ ~-,.,- at 37C in a hllmiriifiPd ~mnsphpre snrpl~mpntpd with 10% C02. For tr~n~r~;l;QI~, e-l Ollr~ lly ~,IOWillg cells were ha.v~,led by l~y~ , washed free of serum, and r~ ~cpen~lPd in DMEM at a conr~-.l. ~ ;on of 107 cells/l. Plasmid DNA (5~Lg) was added to 800 ,ul of cell su~",e~ion (8 ~c 106 cells), and the mi~cture was subjected to clech-)pG-alion using the Biorad Gene Pulser and C~a-~ ce FYt~PnrlPr (200-300 V, 960 ~F, 0.4 cm electrode gap, at ambient lé ll~,.a~ule).
Following ele~ po~dion~ the cells were ~u~,.ed to 10 cm dishes and grown in non-selective mPAillm After 24 hours, the cells were l,~ ~l, seeded at 6 x 10 cells/10 cm dish, and allowed to attach o~e..,igl-~. The non-selective mPAi--m was replaced with selective medium (co..~ g 500 U/ml of Hm or 800 ~g/ml of G-418), and s~lecti~n was c~ntinl-~A for 10-14 days. The plates were then fixed with n~- II.al-ol, stained with methylene blue and colonies were counted. The number of coloni~s rë~-led in Table 1 is the average number of 15 coloniP-s per 10 cm dish.
U..l~ r~æd cells were not h~g.~ Hmr) or G-418 resistant (G-418r).
The results indicate that the hygro-CD and CD-hygro fusion genes encode Hm, but the activity of the CD-hygro fusion gene is lower than that of the hygro-CD fusion gene. The CD-neo fusion gene confers G-418r, but the neo-CD fusion gene doe~s not.
Table 1 Do~-lh~a--l Positive Selection Tl~ r~ d No. Hm Colonies No. G-418 Coloni~s Pl~cmi-l Trial 1 Trial 2 Trial 1Trial 2 None 0 0 tgCMV/hygro/LTR 89 34 nt nt tgCMV/hygro-CD 96 34 nt nt tgCMV/CD-hygro 7b 13b nt nt tgCMV/neo nt nt 28 73 tgCMV/neo-CD nt nt 0 0 tgCMV/CD-neo nt nt 29 64 nt = not tested 35 b = small, slowly gro~,ng co~ni WO 94/28143 ~ ~ 3 4 ~ -22- PCT/US94/05601 1 C. Cytosine Deq~min-qse Assay on T~ r~ed Cell Pools.
To det~rminr whether the fusion genes had retained cytosine d~A-Aminqce (CD) activity, the Hm and G-418r NIH/3T3 cQloni-A~ as ~ d in Table 1, were pooled and e~pqn~P~ into cell lines. FYtr~tc were pr~a, d and assayed for CD activity by .--rA~-~. h-~ the co~ ion of S ~losine to uracil e~çntiqlly as ~ )usly described (Mullen et al., Proc. Natl. Acad. Sci. US~
89:33, 1992), except that [14C]~yLosine was used in place of [3H~y~siile. A 10 cm dish was seeded with 1 x 106 cells, and the cells were ;~ ed for two days. The cells were then washed in Tris buffer (100 mM Tris, pH 7.8, 1 mM EDTA, 1 mM dithiothreitol) and scraped from the dish in 1 ml of Tris buffer. The cells were then cAntrifi~ged for 10 sec at 24,000 rpm 10 in an Eppendorf microfiuge, r~us~e"ded in 100 ~l of Tris buffer and subjected to five cycles of rapid freezing and thawing. Following c~ ;r..gd~ for 5 min at 6,000 rpm in an Eppendorf microfuge, the suprrn~t~nt was l~ 9~.~r~ . xl to a clean tube.
The cQnr~ alion of protein in the e~tract was d~ ...inP~ using a Biorad protein assay kit. A 25 ~l aliquot of cell extract (or an equivalent amount of protein in a volume of 25 ~l) l~ was then mixed with 1 ~l of [14C]-cytosine (0.6 mCi/ml, 53.4 mCi/mmol; Sigma ChPmir~l Co.), and the reaction allowed to proceed at 37C for 1-4 h. One half of the reaction was then applied to a thin-layer cl-lu-"àtl)gram and ~,l.ro...d~graphed in a mi~ture of 869~ l-butanol and 14% water. Following devP~opm~nt the thin-layer cl~ro-l-d~gram was exposed to Kodak X-OMAT AR X-ray film for 8-14 h. The result is shown in Figure 2.
The results indicate that the CD-neo, CD-hygro and hygro-CD fusion genes encodedCD activity, but the activities of the CD-hygro and hygro-CD fusion genes were lower than that of the CD-neo fusion gene.

F~ 'e 2 Construction and Cha~a~ dion of Re~rovil~l Vectors Containin~ neo or CD-neo Selectable Fusion Genes A. Con~truction of R~ vhdl Vectors.
The retroviral plasmids tgLS(+)neo and tgLS(+)CD-neo consist of the following elements: the 5' LTR and se~lPnr~,s through the PstI site at n~rl~otiAe 984 of MoMSV (Van Beveren et al., Cell 27:97, 1981); se~u~'nr~s from the PstI site at r--rleoti~le 563 to mlrl~oti-l9 la40 of MoMLV (Shinni~ et al., Nature 293:S43, 1981); a r.~ from tgCMV/ne~ or tgCMV/CD-neo, co.~;ni~g the neo or CD-neo coding regions, l~)C~ively~ se~UPnrr-~ from nllrle4ti-1e 7764 and through t_e 3' LTR of MoMLV (~hinnirL- et al., su~ra, 1981); the NruI-~ ~627r~ooo3~6/14~7oo 2 1 6 3 4 2 7 _~)3_ AlwNI fragment from pML2d (Lusky and Botchan, supra, 1981), containing the bacterial replication origin: the AlwNI-AatII fragment from pGEMI (Promega Corp.), containing the ~-lart~m~ce gene.
Plasmid tgLS(+)neo was cons~ructed using standard techniques (Ausubel et al., supra, 1987) as follows: Plasmid tgLS(+)hygro was constructed first, by ligating an EcoRI-Clal fragment from tgLS(+)HyTK tO an EcoR~-Asp718 fragment from tgCMV/hygro, and a synthetic DNA fragment (prepared by ~nne~ling oligonucleotides 5'-GTACAAGCTTGGATCCCTCGAGAT-3' (SEQ ID NO:20) and 5'-CGATCTCGAGGGATCCAAGCTT-3' (SEQ ID NO:21)). Plasmid tgLS(+)neo was then 10 constructed by replacing the NheI-HindIII ~ragment sp~nning the hygro gene with a NheI-HindIII fragment arnplified from pSV2neo (Southern and Berg, ~. Mol. Appl. Gen. 1:327, 1982) by PCR using oligonucleotides 5 ' -CCCGCTAGCCGCCGCCAC CATGGGATCGGCCATTGAACAAGATGGATTGCAC-3 ' (SEQ ID NO:22) and 5'-CCCAAGCTTCCCGCTCAGAAGAACTCGTC-3' (SE(2 ID NO:23) 15 (the PCR product was digested with Nhel and HindIII before ligation).
Plasmid tgLS(+)CD-neo was constructed using standard techniques (Ausubel et al.,s~ra, 1987) as follows: The NheI-SalI fragment spanning the HCMV E94 pLo,llole. and human IL-2 cDNA from HyTK-CMV-IL2 was replaced with the NheI-SalI fragment from tgCMV/CD-neo.
Figure 3 shows the proviral structures of the retroviral vectors tgLS(+)neo and tgLS(+)CD-neo. In the figure "LTR" signifies the long terminal repeat segrn~ntc of the l~LIovilal vector, "neo" signifies the bacterial neomycin phosphoL-al~,r~.~,e gene, and "CD-neo" represents the CD/neo~ ch1 phosphotlall,~l~,e fusion gene. The neo and CD-neo genes are operably linked to the LTR transcriptional control region. The arrows show the direction 25 of Ll~,-ni~ion from the transcriptional control regions. "A " represents the polyadenylation sequenre B. C,eneration of Stable Cell Lines Infected With Retroviral Vectors.
To derive stable NIH/3T3 cell lines infected with tgLS(+)neo and tgLS(+)CD-neo, the 30 retroviral plasmid DNAs were ~ ,r~e~ed into ~2 ecotropic p~cl~ging cells. The transfected ~2 cells were then ~ 71`elled to a lO cm tissue culture dish contail~ing 10 ml of complete growth medium supplemented with 10 mM sodium butyrate (Sigma Chemical Co.) and allowed to attach overnight. After 15 h. the medium was removed and replaced with fresh m~Aillm 1q6~ 000346l l47700 -23/1- 2 1 6 3 ~ 2 7 After a further 24 hours~ the medium containing transiently produced ecotropic virus particles was harvested. centrifuged at 2000 rpm for 10 minutes and used to infect NIH/3T3 cells.

= = ~

-2~ 2~3~7 ExponPnti~lly dividing NIH/3T3 cells were h~lL~Ied by tL~ ";Qn and seeded at a density of 2.5 x 10 cells/35 mm well in two ~well tissue culture trays. On the following day, the mPAillm was replaced with serial t~ ti~n~ of virus~4~ g, cell-free ,LI~e~
ml/well) in mRAillm suppl~mpn~Rd with 4 ~g/ml Pol~l~e h~ -e bromide (Sig_a5 ChRmi~l Co.). T.~r~;OI. was allowed to proceed overnight. Then the su~,e~ was replaced with complRte growth mRAillm After a further 8-24 hours of growth, the infected NIH/3T3 cells were selected for drug ,~ e to G-418 (Gibco) at a final col-~e~ alion of 800 ~g/ml (Hmr cells). After a total of 12-14 days of growth, one tray of cultured G418r resis~l cells was fixed with 100% ~ nol and stained with methylene blue. The coloniPs 10 were counted and the number of colonies in each well was used to P~st~hli~h the titers of the retrovirus present in the tr~n~i~ntly infected ~upe~ (Table 2).

Table 2 Titers of F~otropic R~ vi~u,cs Produced Tr~n~iPntly in ~2 p~L~;~ Cells on N~H/3T3 Cells G4l8r Virus CFU/ml tgLS(+)neo s ~c 105 tgLS(+)CD-neo 1 x 105 From the other tray of G418 cells, the colonies of G418 -cells were pooled and eYr~nAPd into bulk cultures for analysis. Extracts were pr~,d from the bulk cultures and assayed for CD activity by ,~ g the coL-~ion of eylv~iL e to uracil generally aspreviously described (Mullen et al., 1992), except that [ C]~ ,si--e was used in place of ~3H~-cyto;,inc. A 10 cm dish was seeded with 1 ~c 10 cells, and the cells were ;--~ aled for 2 days. The cells were then washed in Tris buffer (100 mM Tris, pH 7.8, 1 mM EDTA, 1 mM
dithh~llLL~ ~I) and scraped from the dish in 1 ml of Tris buffer.
The cells were then c~ ;rL.~od for 10 seconds at 14,000 rpm in an E~endol~
microfuge, r~ ed in 100 ~1 of Tris buffer and s-lbjected to five cycles of rapid L~i, g and t_awing. Following cP~ ;r~a~in-~ for 5 min at 6,000 rpm in an Eppendorf microfuge, the su~ AI;3~l was l~ ,--- r~, . dl to a clean tube. The co ~ l ;n~- of protein in the extract was d~t~ ned using a Biorad protein æsay l~it. A 25 ~1 aliquot of cell e~ctract (or an equivalent amount of protein in a volume of 25 ~1) was then mixed with 1 ml of [ C]~i..c (0.6 mCi/ml, 53.4 mCi/mmol; Sigma Ch~mi~Al Co.), and the reaction was allowed to proceed at 35 37 for 1~ hours. One half of the reaction mi~cture was then applied to a thin-layer ) s4ns143 PCT/US94/05601 c]~o~logram, and Clln~ "O~h~d in a mixture of 86% 1-butanol and 14% water.
Following developlllc.ll, the thin-layer chromatogram wæ exposed to Kodalc X-OMAT AR X-ray film for 8-14 hours. The results shown in Figure 4 indicate that cells infected with the tgLS(+)CD-neo ~hovi~d vector express high levels of cytosine ~ e activity.

C. NP~tive Sel~liol- of Cells (~o..1i.;1~i"~ thP (~n-neo Selectable Fllci~ n Gene. To ,al~ the utility of the neo and CD-neo s~ e fusion genes for ne~dive s~l~lion, the color--s resulting from each tra~f~.~;on were pooled and e-~n-1~d into cell lines for fur~er analysis. The NIH/3T3 cells, or NIH/3T3 cells infected with the tgLS(+)neo or tgLS(+)CD-10 neo retroviruses were assayed for 5-FC using a long-term prolir ~;o.. assay.
First, 1 x 104 cells were seeded into 10 cm tissue culture dishes in c~mrl~P~e growth mA~ m and allowed to attach for 4 hours. The mAdi~m was then suppl~mAntp~d with various co~ e~ ;on.C of G 118 and/or 5-FC (Sigma), after which the cells were j~ ed for a further 1~14 days. The mA~ m was replaced every 2~ days. The cells were then fLlced in situ with 15 100% ~ )l and stained with methylene blue.
Photographs of ~f~enL~Iivc; stained plates are shown in Figure 5. Plate a had NIH/3T3 cells grown in drug-free mAdi~m- Plate b had NIH/3T3 cells grown in m~Aillm CQnl~ E 800 ~g/ml G~18. Pl?te c had NIH/3T3 cells grown in mA~linnt ~41~ g 100 ~Lglm~ 5-FC. Plate d had NIH/3T3 cells infected with tgLS(+)neo and grown in meAillm c~ g 800 ~g/ml G 418. Plate e had NIH/3T3 cells infected with tgLS(+)neo and grown in mA~illm c~n'~ c 800 ~g/ml G~18 and 100 ~g/l 5-FC. Plate f had NIH/3T3 cells infected with tgLS(+)CD-neo and grown in m~i~ c~ g 800 ~g/ G418. Plate g had NIH/3T3 cells infected with tgLS(+)CD-neo and grown in medi~m CQIIt~ 800 ~Lg/ml G-418 and 100 ~g/_~ 5-FC.
These results indicate that 1) lln;--~ NIH/3T3 cells are se~ili~ to G-418 and l~si;~ to 5-FC, 2) NIH/3T3 cells infected with tgLS(+)neo are r~i;,~ to both G-418 and 5-FC, and 3) NIH/3T3 cells infected with tgLS(+)CD-neo are resistant to G-418 but se~ilive to 5-FC.

Claims (15)

-26-We claim:

1. A selectable fusion gene comprising a dominant positive selectable gene fusedto and in reading frame with a negative selectable gene, wherein the selectable fusion gene encodes a single bifunctional fusion protein which when expressed confers a dominant positive selectable phenotype and a negative selectable phenotype on a cellular host;
wherein the negative selectable gene is cytosine deaminase (CD).

2. A selectable fusion gene according to claim 1, wherein the dominant positive selectable gene is selected from the group consisting of hph and neo genes.

3. A selectable fusion gene according to claim 2, wherein the dominant positive selectable gene is neo.

4. A selectable fusion gene according to claim 3 encoding the sequence of amino acids 2-690 of SEQ ID NO:2.

5. A selectable fusion gene according to claim 3 encoding the sequence of nucleotides 4-2073 of SEQ ID NO: 1.

6. A recombinant expression vector comprising a selectable fusion gene accordingto claim 2.

7. A recombinant expression vector comprising a selectable fusion gene accordingto claim 3.

8. A recombinant expression vector comprising a selectable fusion gene accordingto claim 4.

9. A recombinant expression vector according to claim 6, wherein the vector is aretrovirus.

10. A recombinant expression vector according to claim 7, wherein the vector is a retrovirus.

11. A recombinant expression vector according to claim 8, wherein the vector is a retrovirus.

12. A cell transduced with a recombinant expression vector according to claim 6.

13. A cell transduced with a recombinant expression vector according to claim 9.

14. A method for conferring a dominant positive and negative selectable phenotype on a cell, comprising the step of transducing the cell with a recombinant expression vector according to claim 6.

15. A method for conferring a dominant positive and negative selectable phenotype on a cell, comprising the step of transducing the cell with a recombinant expression vector according to claim 9.