CA2243718A1 - Diabetes therapy - Google Patents

Abstract

The invention provides a gene therapy method for delivering safe and effective, long-term amounts of GLP-1(7-37)-based proteins useful for treating Type I and Type II diabetes. The invention eliminates the need for subcutaneous injections and is able to provide tight glucose control.

Description

DIABETES THERAP~
The present invention relates to gene therapy and provides stable, transformed m~mm~l ian cell lines and vectors useful as vehicles for transferring functional DMA se~uences whose protein products are useful in the treatment of dia~etes mellitus.

The human hormone glucagon is a 29-amino acid hormone produced in pancreatic A-cells. The hormone belongs to a multi-gene family of structurally related peptides that include secretin, gastric inhibitory peptide, vasoactive intestinal peptide and glicentin. These peptides variously regulate carbohydrate metabolism, gastrointestinal mobility and secretory processing. However, the principal recognized actions of pancreatic glucagon are to promote hepatic glycogenolysis and gluconeogenesis, resulting in an elevation of blood sugar levels. In this regard, the actions of glucagon are counter regulatory to those of insulin and may contribute to the hyperglycemia that accompanies diabete~
mellitus (Lund, P.K., et al., Proc. Natl. Acad. Sci. U.S.A., 79:345-349 (1982)).
When glucagon binds to its receptor on insulin producing cells, cAMP production increases which in turn stimulates insulin expression (Korman, L.Y., et al., Dia~etes, 34:717-722 ~1985)). Moreover, high levels of insulin down-regulate glucagon synthesis by a feedback inhibition mechanism (Ganong, W.F., Review of Medical P~ysiology, Lange Publications, Los Altos, California, p. 273 (1979)). Thus, the expression of glucagon is carefully regulated by insulin, and ultimately by serum glucose levels.
Preproglucagon, the precursor form of glucagon, is encoded by a 360 base pair gene and is processed to form -~ proglucagon ~Lund, et al., Proc. Natl. Acad. Sci. U.S.A.
7g:345-349 (1982~). Patzelt, et al. (Nature, 282:260-266 ~1979)) demonstrated that proglucagon is further processed into glucagon and a second peptide. Later experiments demonstrated that proglucagon is cleaved carboxyl to Lys-Arg W O 97~29180 PCTAUS97/01978 or Arg-~rg residues (Lund, P.K., et al., Lopez L.C., et al., Proc. Natl. Acad. Sci. U.S.A., 80:5485-548g (1983), and sell, G.I., et al., Nature 302:71~-718 (1983)). Bell, G.I., et al., also discovered that proglucagon contained three discrete and highly homologous peptide regions which were designated glucagon, glucagon-like peptide 1 (GLP-l), and glucagon-llke peptide 2 (GLP-2). Lopez, et al., demonstrated that GLP-l was a 37 amino acid peptide and that GLP-2 was a 3~ amino acid peptide. Analogous studies on the structure of rat preproglucagon revealed a similar pattern of proteolytic cleavage at Lys-Arg or Arg-Arg residues, resulting in the formation of glucagon, GLP-l, and GLP-2 (Heinrich, G., et a7., Endocrinol., 115:2176-2181 ~1984)). Finally, human, rat, bovine, and hamster se~uences of GLP-l have been found to be identical ~Ghiglione, M., et al., Diabetologia, 27:599-6~0 (1984)).
The conclusion reached by Lopez, et al., regarding the size of GLP-l was confirmed by studying the molecular forms of GLP-l found in the human pancreas (Uttenthal, L.O., et a7. J. Cli~. Endocrinol. Metabol., 61:472-479 (1985)).
Their research showed that GLP-l and GLP-2 are present in the pancreas as 37 and 34 amino acid peptides respectively.
The similarity between GLP-l and glucagon suggested to early investigators that GLP-l might have biological activity. Although some investigators found that GLP-l could induce ra~ brain cells to synthesize cAMP (Hoosein, N.M., et al., Febs Lett. 178:83-86 ~1984)), other investigators failed to identify any physiological role for GLP-l (Lopez, L.C , et a7. supra). The failure to identify any physiological role for GLP-l caused some investigators to question whether GLP-l was in fact a hormone and whether the relatedness between glucagon and GLP-l might be artifactual.
It has now been shown that biologically processed forms of ~LP-l have insulinotropic properties and delay gastrlc emptying. GLP-1(7-34) and GLP-1(7-35) are disclosed in U.S. Patent No: 5,118,666, herein incorporated by WO 97129180 PCTrUS97/01978 reference. GLP-1(7-37) is disclosed in U.S. Patent No:
5,120,712, herein incorporated by reference.
Variants and analogs of GLP-1 are known in the art.
These variants and analogs include, for example, GLP-1(7-36), Gln9-GLP-1(7-37), Thr16-Lys18-GLP-1(7-37), and Lys18-GLP-1(7-37). Derivatives of GLP-1 include, for example, acid addition salts, carboxylate salts, lower alkyl esters, and amides (see, e.g., WO91/11457). Generally, the various disclosed forms of GLP-1 are known to stimulate insulin lC secretion (insulinotropic action) and cAMP formation (see, e.g., Mojsov, S., Int. ~. Peptide Protein Research, 40:333-343 (1992)).
More importantly, numerous investigators have demonstrated a predictable relationship between various in vi~ro laboratory experiments and mammalian, especially human, insulinotropic responses to exogenous administration of GLP-1, GLP-1(7-36) amide, and GLP-1(7-37) acid (see, e.g., Nauck, M.A., et a7., Diabetologia, 36:741-744 (1993); Gutniak, M., et al., New England J. o~ Medicine, ~2~(20):1316-1322 (1992);
Nauck, M.A., et al., ~. Clin. Invest., ~1:301-307 (1993); and Thorens, B., et al., Diabetes, 42:1219-1225 (1993)).
The fundamental defects responsible for causing hyperglycemia in mature onset diabetes include impaired secretlon of endogenous insulin and resistance to the e~fects o~ insulin by muscle and liver tissue ~Galloway, J.S., Dia~etes Care, 13:1209-1239, (1990)). The latter de~ect results in excess glucose production in the liver. Thus, whereas a normal individual releases glucose at the rate of approximately 2 mg/kg/minute, a patient with mature onset diabetes releases glucose at a rate exceeding 2.5 mg~kg/minute, resulting in a net excess o~ at least 70 grams o~ glucose per 24 hours.
Because there exists exceedingly high correlation between hepatic glucose production, fasting blood glucose levels, and overall metabolic control as indicated by glycohemoglobin measurements (Galloway, J.A., supra; and Galloway, J.A., et al., Clin. Therap., 12:460-472 (1990)), it W O 97f29180 PCTrUS97/01978 is readily apparent that control of fasting blood glucose is essential for achieving overall normalization of metabolism sufficient to prevent hyperglycemic complications. Since existing insulin therapies rarely normalize hepatic glucose production without producing significant hyperinsulinemia and hypoglycemia (Galloway, J.A., and Galloway, J.A., et al., supra) alternative approaches for diabetic therapy are needed.
Therapy based on administration of longer acting 10- GLP-l analogs is one such approach. To date however, this approach has failed to deliver long term e~ficacious doses to individuals due in large part because the serum hal~ e o~
GLP-1(7-37) is ~uite short. There~ore, the quest for alternative approaches continues.
Gene therapy o~fers a new avenue ~or treating diseases rooted in hormone de~iciencies because it operates as an i~ vivo protein production and delivery system. This is an especially attractive approach since gene therapy also offers the possibility of physiologically regulating the 2~ production and secretion o~ proteins in response to homeostatic mediators within the body.
Gene therapy can be effected in a number of ways.
~etroviral-mediated gene transfer was suggested ~or treating human diseases involving mal~unctioning bone marrow.
Anderson et al., Science 226: 401 (1984). In addition, PCT
Publication Number W093/09222 (May 13, 1993) and U.S. Patent Number 5,3g9,346 (March 21, 1995) disclose the genetic alteration of primary human cells that are cultured then reintroduced into the body for the treatment o~ a variety of diseases.
Many heritable diseases such as diabetes result from the absence of a functional gene necessary to provide the ~n;~l with an ade~uate supply o~ a vital protein. The goal of gene therapy is to deliver a nucleic acid sequence, 3~ present on an RNA or DNA vector, which is capable of encoding the desired, therapeutic protein. Although a number of methods exist to deliver the DMA or RNA vector containing the W O 97129180 PCTrUS97/01978 desired nucleic acid sequence into the target m~mm~l ian cells, two procedures termed ex vivo and in vivo are generally employed.
Ex vivo gene therapy consists of four primary steps: (1) Primary cells (target cells) are removed from the individual in need of therapy; (2) The gene therapy nucleotide sequence is incorporated into the target cell in vl tro; (3~ The transformed cells expressing the incorporated nucleotide sequence encoding the protein of interest are identified, isolated, and expanded; and (4) The transformed cells are reintroduced into the individual.
~ x vivo therapy generally results in the incorporation of the nucleotide sequence encoding the protein of interest into the chromosomal DNA of the target cell. The critical step of ex vivo gene therapy is the proper introduction of the nucleotide se~uence encoding the desired protein lnto the target cell This transfer of DNA can be accomplished by a number of well documented methods such as:
calcium phosphate precipitation, electroporation, and adenoviral or retroviral vectors (Current Protocols in Molecular 3iolo~v, John Wiley and Sons, 1989; Methods of Cell Bioloov, 43: 161 - 189, 1994; Proc. Natl. Acad. Sci. USA 85:
6460 - 6464, 1985); although, other well known methods are also consistent with this invention.
In vivo gene therapy generally describes the transfer of a desired nucleic acid sequence, located on a transport vector, directly into an individual in need of therapy. This gene therapy approach does not require that the target cells be ~irst removed and manipulated in vitro.
O~ce the vector containing the desired nucleotide sequence is introduced into the individual, the vector moves into the nucleus of the target cell and the nucleotide sequence of interest integrates into the chromosomal DNA of the target cell ~ith varying degrees of efficiency. A number of well documented methods exist for introducing nucleic acids into an individual requiring therapy such as direct injection of WO97/29180 PCTrUS97/01978 D~A and the use o~ recombinant viral vectors (Gene TheraDv, A
Hand~ook for Phvsici~n~, Mary Ann Liebert, Inc., 1994).
The present invention provides a method o~ treating both Type I and Type II diabetics through a di~erent gene therapy approach. A stable m~mm~l ian cell line is trans~ormed by a nucleic acid vector such that it secretes a GLP-1(7-37)-based protein, as de~ined by SEQ ID NO 1, followed by implantation into an individual needing treatment. Once implanted, the G~P-1(7-37)-based protein, in 13 con~unction with high serum glucose levels, causes pancreatic beta cells to produce insulin in non-insulin dependent d~abe~es mellitus (NIDDM~ patients and delays gastric emptying in both NIDDM and insulin dependent diabetes mel 7itus IDDM patients.
Accordingly, one embodiment o~ this invention provides a method o~ treating Type I or Type II diabetes in a mammal in need thereo~ comprising implanting a cell line trans~ormed with a vector comprising a promoter driving expression of a DNA se~uence encoding a protein o~

His-Xaa1-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-Ser-Tyr-Leu-Xaa2-Gly-Gln-Ala-Ala-Xaa3-Xaa4-Phe-Ile-Ala-Trp-Leu-Val-Lys-Gly-Arg-Xaa5 (gs~ ID NO 1) wherein Xaa1 is Ala, Gly, Val, Thr, and Ile;
Xaa2 is Glu, Gln, Ala, Thr, Ser, and Gly;
Xaa3 is Lys, and Arg;
Xaa4 is Glu, Gln, Ala, Thr, Ser, and Gly; and, Xaa5 is Gly-OH or is absent;
3~ into said m~mm~ I such that it is ;mmllnologically isolated ~rom the mammal's immune system and secretes a protein o~ SEQ
ID NO 1 into said patient.
Pre~erred proteins o~ SEQ ID Mo 1 are those in which Xaal is Ala, Xaa2 is Glu, Xaa3 is ~ys. Xaa4 is Glu, and Xaa5 is Gly-OH. More pre~erred are those in which Xaa1 is Ala, Gly, Val, Thr, and Ile; Xaa2 is Glu, Gln, Ala, Thr, Ser, CA 022437l8 l998-07-2l W O 97/29180 PCTrUS97/01978 and Gly; Xaa3 is Lys, and Arg; Xaa4 i5 Glu, Gln, Ala, Thr, Ser, and Gly; and, Xaa5 is Gly-OH or is absent. Another pre~erred group are those in which Xaa1 is Val and Xaa3 is Lys. A ~urther pre~erred group o~ proteins are those wherein Xaa1 is Ala or Val, Xaa2 Glu, Xaa3 is Lys or Arg, Xaa4 is Glu, and Xaa5 Gly-OH or is absent. Yet another pre~erred group are those in which Xaa1 is Ala, Xaa2 Glu, Xaa3 is Lys, Xaa4 is Glu, and Xaa5 Gly-OH. Still another pre~erred group is when Xaa1 is Val, Xaa2 Glu, Xaa3 is Lys, Xaa4 is Glu, and Xaa5 Gly-OH.
Nucleotide sequences encoding any one o~ the polypeptides o~ SEQ ID NO 1 may be prepared by a variety of means readily apparent to those skilled in the art. Wholly synthetic nucleotide sequences or semi-synthetic sequences deri~ed in part ~rom a natural GLP-1 gene may be used. Owing to the natural degeneracy o~ the genetic code, the skilled artisan will recognize that a sizable yet de~inite number of nucleotide sequences may be constructed which encode the proteins o~ SEQ ID MO 1. A synthetic DNA sequence encoding a G~P-1-based protein o~ SEQ ID NO 1 may be prepared by techni~ues well known in the art in substantial accordance wit~ the teachings o~ Brown, ~ ~1. (1979) Methods in ~zymo70~y, Academic Press, N.Y., Vol. 6~, pgs. 109-151. The DNA seq~ence may be generated using conventional DNA
synthesizing apparatus such as an Applied Biosystems Model 380A or 380B DNA synthesizer (commercially available ~rom Applied Biosystems, Foster City, Cali~ornia). Commercial services are also available for the construction of such nucleotide sequences based on the amino acid sequence.
In one pre~erred embodiment o~ the invention as exempli~ied herein, the coding sequence ~or a protein o~ SEQ
~D N~ 1 is the ~ollowing:
5' - CAT GCT GAA GGG ACC TTT ACC AGT GAT GTA AGT TCT TAT TTG
GAA GGC CAA GCT GCC AAG GAA TTC ATT GCT TGG CTG GTG AAA
35 GGC CGA GGA - 3' ~SEQ ID NO 2).

W O g7129180 PCT~US97/01978 This sequence encodes the following protein o~ SEQ ID NO 1:
H2N- 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 -OH (SEQ ID NO 3).
In another preferred embodiment of the invention as - exempli~ied herein, the coding se~uence ~or a protein of SEQ
ID NO 1 is the following:
5 ' - CAT GTT GAA GGG ACC TTT ACC AGT GAT GTA AGT TCT TAT ~TG
GAA GGC CAA GCT GCC AAG GAA TTC ATT GCT TGG CTG GTG AAA
GGC CGA GGA - 3~ (SEQ ID No 4).
This sequence encodes the following protein of SEQ ID NO 1:
H2N- His Val 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 -OH (SEQ ID NO 5).
15_ Coding regions for SEQ ID NO 1 may be fused to a leader sequence which signals the cell to secrete a precursor peptide of SEQ ID NO 1 that is subsequently processea by the cell during the secretion process to a protein of SEQ ID NO
l Many such leader sequences are know in the art. One well known and preferred leader sequence is the hybrid tissue plasminogen activator/protein C prepropeptide described in serg et a 7 . Biochem. Biophys. Res. Commun. 179: 1289-1296 ~1991). Typically nucleotide se~uences encoding precursor peptides of SEQ ID NO 1 are flanked by linker DNA to facilitate enzymatic ligation into expression vectors as is later exemplified herein.
Once a suitable coding sequence of SEQ ID N~ 1 iS
constructed and optionally fused and flanked by an apprcpriate leader sequence and linker DNA, the construct is ligated into an expression vector which is then introduced into an ap~L ~L iate cell line. Construction of suitable vectors containing the desired coding and control sequences may be constructed by standard ligation techniques. Isolated WO 97129180 PCTrUS97/01978 plasmids or nucleotide ~ragments are cleaved, tailored, and religated in the manner necessary to achieve the desired plasmids.
To ef~ect the expression of a polypeptides o~ SEQ
ID NO 1, one ligates a nucleotide se~uence encoding the polypeptide into an appropriate recombinant nucleotide expression vector through the use o~ appropriate enzymes.
The nucleotide sequence encoding a polypeptide o~ SEQ ID NO 1 is designed to possess restriction endonuclease cleavage sites a~ either end of the DNA to ~acilitate isolation from and integration into these ampli~ication and expression plasmids. The coding sequence may be readily modi~ied by the use of synthetic linkers to facilitate the incorporation of the coding sequence into the desired cloning vectors by techniques well known in the art. The particular en~onucleases employed will be dictated by the restriction endonuclease cleavage pattern of the parent expression vector to be employed. The choice of restriction sites are chosen so as to properly orient the coding sequence such that it is properly associated with the promoter and ribosome binding site of the expression vector, both o~ which are ~unctional in the host cell in which a compound of SEQ ID NO 1 is to be ex~ressed.
In addition to the desired nuceotide sequence which will code for the therapeutic protein, the expression vector m~y contain several other ~unctional elements. One such element is the promoter and upstream regulatory sequences which control the level of expression of the protein o~
interest. Some expression vectors contain promoters and regulatory sequences which normally regulate transcription o~
cellular genes. One such promoter is the mouse metallothionein-I promoter which has been shown to function both in-vitro and in-vivo (Palmiter et al., Nature 300: 611 -615, 1982). In addition, promoters and regulatory sequences from viruses are ~requently used in expression vectors ~Dijkema et al., ~MBO J. 4, 471, 1985; Gorman et al., Proc.
Natl. Acad. ~ci. 79: 6777, 1982; Boshart, et al. Cell 41:

W O97/2918~ PCT~US97/01978 521, lg85). Although, Verma et al. have shown that the retrovirus promoter and enhancer se~uences do not ~unction for long periods of time in-vivo. In addition, the expression vector also carries an origin of replication as well as marker se~uences which are capable o~ providing phenotypic selection in transformed cells.
Because the proteins useful in the present invention do not re~uire post-translational processing mechanisms other than enzymatic removal of the propeptide leader sequence, many stable human cell lines are consistent with the practice of the invention. One preferred cell line i~ the human embryonal kidney cell line 293, available from the permanent collection of the American Type Culture ~ollection.
A number of well known methods exist for introducing the genetic material into target cells such as chemical ~calcium phosphate precipitation), physical (electorporation and microin~ection), and viral methods ~adenovirus, retrovirus, and adeno-associated virus) (Methods ~or Gene Trans~er, Gene Thera~v, Mary Ann Liebert, Inc., lg~4). All such methods are consistent with the practice of the prese~t invention. The techniques o~ transforming mammalian cells with the aforementioned vector types are well known in the art and may be found in such general references as Maniatis, et al. (1989) Molecul~r Cloninc: A Laboratory M~nll~l, Cold Spring Harbor Press, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York or Current Protocols in Molecular Bioloov, Vol. 1, (1988), Wiley Interscience, and supplements.
3~ Stable transformed cell lines that express proteins Q:E SE~ ID NO 1 must then be implanted into the individual in need of such treatment. Because such transformed cell lines generally will be histologically incompatible with the indivi~uals receiving them, the cells must to be protected from the recipient's immune system. Once way o~ protecting the implanted cells is by masking them with F(abl)2 fragments speci~ic ~or HLA class I antigens. Immunological masking W O97/29180 PCTrUS97/01978 methods are well known in the art. For example, see Faust et al. ~cience 252:1700-1702 (1991). Other means for protecting the implanted cells from the recipient's immune system are consistent with this invention. Such methods include but are not limited to encapsulation in semi-permeable membranes (Lanza et al. Diabetes 41: 1503 - 1510) and through the use o~ immunosuppressants (Rynasiewicz et al., Diabetes 31: 92 -108, 1982).
By way o~ illustration, the ~ollowing examples are provided to help describe how to make and practice the various embodiments of the invention. These examples are in no way meant to limit the scope o~ the invention.

~i~Amr~le Construction of Intermediate Plasmid ~LP53-tLB+~TP-1 A. Pre~aration of B~lTI-Muna sean-AvrII Diaested ~P53-tLB
The plasmid pLP53-tLB was isolated ~rom E. coli K12 AG1 lon deposit under terms of the Budapest Treaty and made part of the permanent stock culture collection of the Northern Regional Research Laboratories (NRRL), agricultural Research Service, U.S. Dept. of Agriculture, Peoria, IL 61604 under accession number NRRL B-18714) using the Plasmid Puri~ication Midi Kit (Qiagen, Inc., 9600 DeSoto Avenue, Chatsworth, CA 91311).
Sixty ~l (approximately 20 ~g) o~ pLP53-tLB DNA was digested with 2 ~l (20 units) of ~glII in a 70 ~1 reaction volume containing 50mM Tris-HCl (pH 8.0), 10mM MgCl2, and l~M NaCl. The sample was incubated at 37~C for one hour.
17.5 ~l of 5x stop mix (25% glycerol, 2% SDS, 0.05%
bromophenol blue, 0.05% xylene cyanol in water) was added and then the reaction was heated at 70~C for 15-20 minutes to inactivate the restriction enzyme. The mixture was spin dialyzed using G-50 Sephadex Quick Spin columns (Boehringer Mannheim Corporation, P.O. Box 50414, 9115 Hague Road, CA 022437l8 l998-07-2l WO97/29180 PCT~US97/01978 Indianapolis, IN 46250-0414) to remove the reaction components.
The 5' protruding ends created by cleavage with ~glII were removed using Mung Bean Nuclease. The BalII
digested pLP53-tLB DNA was incubated with 0.3 ~l (approximately 3.3 units) of Mung Bean Nuclease in a 100 ~l reaction volume containing 10mM Tris-HCl (pH 7.9 at 25~C), lOmM MgC12, 50mM NaCl, lmM DTT, and lmM ZnS04. The reaction was allowed to proceed for 3D minutes at 30~C. One ~l of 1%
~DS was added to inactivate the nuclease. Due to an - incomplete BalII digest, the digested and undigested DNA was separated by gel electrophoresis. Ten ~l o~ gel loading dye (0.25% bromophenol blue, 0.25~ xylene cyanol, 30% glycerol in water) was added to the reaction. The reaction was loaded into the preparative well of a 1.5% NuSieve GTG agarose (EMC
Bioproducts, 191 Thomaston Street, Rockland, ME 04841)/TAE
buf~er gel and then electrophoresed. The gel was stained with ethidium bromide and the DNA wa$ visualized by ultraviolet light. The digested DMA band was excised with a scalpel and placed into two micro-tubes. The DNA was purified from the low melting point agarose usiny the Wizard PCR Preps DNA Purification System (Promega, 2800 woods Hollow Road, Madison, WI 53711-5399).
One hundred ~1 of BalII-Mung Bean digested pLP53-tLB DNA was further digested with 4 ~l (approximately 16 ~ units) of AvrII in a reaction volume of 120 ~1 containing lOmM Tris-HCl (pH 7.9 at 25~C~, 10mM MgC12, 50mM NaCl, lmM
DTT . The sample was incubated at 37~C for 30 minutes. To prevent recircularization, the B~lII-Mung Bean-AvrII digested pLP53-tLB DNA was dephosphorylated (removal of 5~ phosphate groups) by the addition of 2 ~l (2 units) of calf intestinal alkaline phosphatase to the reaction. The sample was incubated at 37~C for an additional 30 minutes. Twenty-four ~l of 5x stop mix was added. The sample was heated at 70~C
for 15-20 minutes to inactivate the enzymes and then spin dialyzed using G-50 Sephadex Quick Spin columns (Boehringer Mannheim Corporation, P.O. Box 50414, 9115 ~ague Road, CA 022437l8 l998-07-2l W O 97/29180 PCT~US97/01978 Indianapolis/ IN 46250-0414) in order to remove the ~II
produced small DNA fragments. The DNA was precipitated by addition o~ O.1 volume oE 3M sodium acetate (pH 5.2) and 2.5 volumes of absolute ethanol. This mixture was mixed 5 thoroughly and then chilled to -20~C. The precipitate was coll ected by centrifugation for 30 minutes. The supernatant was discarded and the pellet was washed with 700 ~l of cold 70% ethanol. The sample was centrifuged for 15 minutes. The supernatant was discarded, the DNA pellet was dried and then 10 resuspended in 25 ,Ul of water.

B. Pre~ration of GLP-l Linker The following single stranded DNA segments were conventionally synthesized by methods well known in the art 15 on an automated DNA synthesizer (model 394 Applied Biosystems 850 Lincoln Center Drive, Foster City, CA 94404-1128) using ~-cyanoethyl phosphoramidite chemistry.

(~T,P--1 . 1 20 5' - GACATGCTGA AGGGACCTTT ACCAGTGATG TAAGTTCTTA TTTGGAAGGC
CAAGCTGCCA AGGAATTCAT TGCTTGGCTG GTGAAAGGCC GAGGATAGGG
ATCCC - 3' (SEQ ID NO 6) GLP-1.2 25 5' - CTAGGGGATC CCTATCCTCG GCCTTTCACC AGCCAAGCAA TGAATTCCTT
GGCAGCTTGG CCTTCCAAAT AAGAACTTAC ATCACTGGTA AAGGTCCCTT
CAGCATGTC - 3' (SEQ ID NO 7) GLP-l.l and GLP-1.2 are complementary DNA
30 molecules. The synthetic DNA moIecules were dissolved in water and stored at less than 0~C.
To anneal the DNA strands, approximately 92.7 pmoles each of GLP-l.l and GLP-1.2 were mixed in 50mM Tris-HCl (pH 7.4) and lOmM MgC12 in a total volume of 80 111 and 35 boiled for~5 minutes. The mixture was slowly brought to room temperature and then transferred to 4~C overnight. This process allowed the two complementary strands to anneal and W O 97/29180 PCTrUS97/01978 form the double stranded DNA linker known as GLP-1. The linker was stored at -20~C. In order to be able to ligate into the dephosphorylated ~g1II-Mung Bean-AvrII digested pLP53-tLB DNA segment, the GLP-1 linker must have 5' phosphate groups. The phosphate groups were added by the use of the enzyme T4 polynucleotide kinase. The kinase reaction contained 80 ~l of the GLP-1 linker, 0.33~M ATP, 70mM Tris-HCl (pH 7.6), 10mM MgCl2, 100mM KCl, lmM ~-mercaptoethanol and 37.2 ,ul t372 units) of T4 polynucleotide kinase. The reaction was incubated at 37~C ~or 30 minutes. Sixteen ~l of 500mM EDTA was added to stop the reaction. The reaction was extracted once with a mixture of phenol:chloroform:isoamyl alcohol (25:24:1) followed by an extraction with chloroform:isoamyl alcohol (24:1). One hundred ~l of the a~ueous layer was spin dialyzed using G-50 Sephadex Quick Spin columns (Boehringer Mannheim) in order to remove the reaction components. The DNA was precipitated by addition of 0.1 volume o~ 3M sodium acetate (pH 5.2), 0.1 volume of 100mM
MgCl2 and 2.5 volumes of absolute ethanol. This mixture was 2G mixed thoroughly and then chilled at -20~C. The precipitate was collected by centrifugation for 30 minutes. The supernatant was discarded and the pellet was washed with 700 ~l of cold 70% ethanol. The sample was centrifuged for 15 minutes. The supernatant was discarded, the DNA pellet was dried and then resuspended in 10 ~l of water.

C. Final Construction of ~TP53-tLB+GLP-1 The DNA prepared in Example lA was ligated with linker GLP-1. Two ~l of DNA prepared in Example lA and 4 ~l o~ GLP-1 linker were ligated in a reaction that contained 2 ,Ul (2 units) of T4 DNA ligase, 50mM Tris-HCl (pH 7.6), 10mM
MgC12, lmM ATP, lrnM DTT, and 50% (w/v) polyethylene glycol-8000 in a total volume of 10 ~l. The mixture was incubated at 16~C for 16 hours. The ligation was used to transform E.
coli K12 I~VaF' cells as generally described below.

CA 022437l8 l998-07-2l W O 97/29180 PCTrUS97/01978 D. Tr~nsformation Procedure Frozen competent E. coli K12 INVaF' cells were = obtained from Invitrogen (3985 B Sorrento Valley Boulevard, San Diego, CA 92121). Two ~l of 0.5M ~-mercaptoethanol were added to 50 ~l of thawed competent cells. About 1-2 ~l of the ligation reaction was mixed with the cells. The cell-DNA
mixture was incubated on ice for 30 minutes, heat-shocked at 42~C for exactly 30 seconds and then chilled on ice for 2 minutes. The cell-DNA mixture was diluted into 450~1 of SOC
media (2% tryptone, 0.05% yeast extract, lOmM NaCl, 2.5mM
KCl, lOmM MgCl2, lOmM MgSO4, 20mM glucose in distilled water) and lncubated at 37~C for one hour in a rotary shaker set at about 225 rpm. Aliquots o~ up to 200 ~l were plated on TY-agar plates ~1% tryptone, 0.5% yeast extract, 1% NaCl, and 1.5% agar, pH 7.4) containing lOO~g/ml ampicillin and then incubated at 37~C until colonies appear.

E. DNA Tsolation Following transformation, ampicillin resistant cells were picked and inoculated into 3 ml of TY broth (1%
tryptone, 0.5~ yeast extract, 1% NaCl, pH 7.4) cont~;n;ng lOO~g/ml ampicillin. These cultures were grown for about 16 hours at 37~C with aeration. Plasmid DNA was isolated from cultures using Wizard Minipreps obtained from Promega (2800 Woods Hollow Road, Madison, WI 53711-5399). Recombinant plasmids were identified by digestion with restriction endonucleases followed by gel electrophoresis analysis.
To obtain larger amounts of pLP53-tLB+GLP-1 plasmid DNA, large scale isolation was performed using the Plasmid Purification Midi Kit (Qiagen, Inc.).
Exam~le 2 Construction of DGT-h+tLB+GLP-1 A. Preparation of BclI Diaested DGT-h The plasmid pGT-h was isolated from ~. coli K12 GM48 (on deposit under terms of the Budepest Treaty and made part W O 97t29180 PCT~US97/01978 - of the permanent stock culture collection of the NRRL under accession number B-18592 using the Plasmid Puri~ication Midi Kit (Qiagen, Inc.).
Ten ~g (37.5 ~l) of pGT-h DNA was digested to completion with 2 ~1 (20 units) of ~clI in a 45 ~l reaction volume containing 50mM Tris-HCl (pH 8.0), lOmM MgCl2, 50mM
NaCl. The sample was incubated at 50~C for 1 hour. Eleven ~l of 5x stop mix was added to the reaction mixture. The mixture was heated at 70~C for 15-20 minutes to inactivate the restriction enzyme and then spin dialyzed using G-50 Sephadex Quick Spin columns (Boehringer Mannheim) in order to remove the reaction components. The DNA was precipitated by adding 0.1 volume of 3M sodium acetate (pH 5.2) and 2.5 volumes of absolute ethanol. This mixture was mixed thoroughly and then chilled at -20~C. The precipitate was collected by centrifugation for 30 minutes. The supernatant was discarded and the pellet was washed with 700 ~l of cold 70% ethanol. The sample was centrifuged for 15 minutes. The supernatant was discarded, the DNA pellet was dried and then resuspended in 20 ~1 o~ water.
Calf intestinal alkaline phosphatase was used to remove the 5' phosphate groups from the DNA segment in order to prevent recircularization of the pGT-h. Ten ~l of the BclI digested pGT-h DNA was treated with 1 ~1 ~1 unit) of calf intestinal alkaline phosphatase in a 15 ~l reaction cont~;ning 50mM Tris-HCl (pH 8.5 at 20~C) and O.lmM EDTA.
The reaction was allowed to proceed for 45 minutes at 37~C.
The phosphatase was inactivated by the addition of 1~1 of 500mM EDTA and then heating at 65~C for 10 minutes. The reaction volume was increased to 100~1 with water and then extracted twice with phenol:chloroform:isoamyl alcohol (25:24:13 followed by extraction with chloroform: isoamyl alcohol (24:1). The DNA was recovered from the a~[ueous layer by the addition of 0.1 volumes o~ 3M sodium acetate (pH 5.2) and 2.5 volumes of absolute ethanol. The mixture was mixed thoroughly and then chilled at -20~C. The precipitate was collected by centrifugation for 30 minutes. The supernatant W O 97/29180 PCTrUS97101978 was discarded and the pellet was washed with 700 ~l of cold 70% ethanol. The sample was centrifuged for 15 minutes. The supernatant was discarded, the DNA pellet was dried and then resuspended in 25 ~l of water.
B. Pre~aration of ~LP53-tLB+GLP-l BamHI Fra~ment Thirty-five ~1 (10.6 ~g) of pLP53-tLB+GLP-l DNA, prepared in Example 1, was digested with 0.5 ~l (25 units) of BamHI in a 40 ~1 reaction volume containing 50mM Tris-~Cl (pH
8.0), 10mM MgCl2, and 100mM NaCl. The reaction was allowed to proceed at 37~C for one hour. Five ~l of gel loading dye was added to the reaction. The reaction was loaded into the preparative well of a 4% NuSieve GTG agarose/TAE bu~er gel.
The DNA was electrophoresed for about one hour at 70 constant volts. The gel was stained with ethidium bromide and then the DNA was visualized by ultraviolet light. The desired 213 base pair DNA band was excised using a scalpel. The DNA was purified from the low melting point agarose using Wizard PCR
preps (Promega).

C. Final Construction o~ ~GT-h+tLB+GLP-l The DNA prepared in Example 2A was ligated with DNA
prepared in Example 2B. One ~l of DNA from Example 2A and 10.5 ~1 of DNA ~rom Example 2B were ligated in a reaction that contained 2 ~1 (2 units) of T4 DNA ligase, 50mM Tris-HCl 25 (pH 7.6), 10mM MgCl2, lmM ATP, lmM DTT, and 50% (w/v) polyethylene ~lycol-8000 in a total volume of 20 ~l. The mixture was incubated at 16~C for 16 hours. The ligation reaction was used to trans~orm E. coli K12 INVo~F~ as described in Example lD. Plasmid DNA was isolated from ampicillin resistant cultures as described in Example lE.
To obtain larger amounts o~ pGT-h+tLB+GLP-l plasmid DNA ~or the purpose o~ transfection of m~mm~l ian cells, large scale isolation was performed using the alkaline lysis method.

W O 97/29180 PCT~US97/01978 Ex~m~le 3 Construction o~ ~GT-h+tTB+V~l8G~P-l The plasmid pGT-h+tLB+Val8GLP-1 was constructed substantially in accordance with Examples 1 and 2. To accommodate the change in the amino acid se~uence (Ala to val at Xaa1 in SEQ ID NO 1) the following coding se~uences were substituted for those described in Example ~B-V~18GLP-1.1 5~ - GACATGTTGA AGGGACCTTT ACCAGTGATG TAAGTTCTTA TTTGGAAGGC
CAAGCTGCCA AGGAATTCAT TGCTTGGCTG GTGAAAGGCC GAGGATAGGG
ATCCC - 3' (~EQ ID NO 8) V~18~T,P-1.2 5' - CTAGGGGATC CCTATCCTCG GCCTTTCACC AGCCAAGCAA TGAATTCCTT
GGCAGCTTGG CCTTCCAAAT AAGAACTTAC ATCACTGGTA AAGGTCCCTT
CAACATGTC - 3~ (SEQ ID NO 9) ~ le 4 Construction of Tntermediate Plasmid pM100-neo A. Pre~aration of ~coRI Diaested ~M100 The plasmid pM100 (pOK12) was isolated from E. coli K12 RRl~M15 usiny Magic Minipreps (Promega). See Vieira and Messing, Gene 100:189-94, 1991). Forty ~l of pM100 DMA was digested to completion with 3 ,Ul of EcQRI in a reaction volume of 50 ~l containing 50mM Tris-HCl (pH 8.0), lOmM MgCl2, and lOOmM NaCl. The sample was incubated at 37~C for 2 hours. The digested DNA was precipitated with ethanol. The final DNA pellet was resuspended in 30 ~1 of water.
B. Pre~aration of ~BK-neo ~coR:r Fra~rment The plasmid pBK-neo 1 (described in U S Patent No:
5,550,036, herein incorporated by reference, and available from the American Type Culture Collection under terms of the W O97/29180 PCT~US97/01978 Budapest Treaty via Accession Mo Atcc 37224) was isolated from E. coli K12 HB-1 using the large scale alkaline lysis ~ method. Three ~l (26.3 ~g) of pBK-neo DNA was digested to completion with 7 ~l (15 units) of ~_RI in a reaction volume of 200 ~l containing 50mM Tris-HCl (pH 8.0), lOmM MgCl2, and lOOmM NaCl. The sample was incubated at 37~C. Fifty ~l o~
the reaction was loaded into the preparative well o~ an agarose/TAE bu~er gel and then electrophoresed. The desired 4.2 Kb DNA band was isolated ~rom the agarose by electrophoresing onto DEAE-cellulose membrane. Following elution from the DEAE-cellulose membrane, the DMA was precipitated with ethanol. The final DNA pellet was resuspended in 30 ~l of water.

C. Final Construction o~ ~M100-neo The DNA prepared in Example 4A was ligated with DNA
prepared in Example 4B. Two ~l of DNA from Example 4A and 2 ~l of DNA from Example 4B were ligated in a reaction that contained 1 ~l (1 unit) of T4 DNA ligase, 50mM Tris-HCl ~pH
7.6), lOmM MgCl2, lmM ATP, lmM DTT, and 50% (w/v) polyethylene glycol-8000 in a total volume of 16 ~l. The mixture was incubated at 16~C for 16 hours. Ten ~l o~ the ligation reaction was used to transform E. coll K12 RRl~M15 as described in Example 5D. Aliquots of up to 200 ~l were plated on TY-agar plates (1% tryptone, 0.5% yeast extract, 1%
NaCl, and 1 5% agar, pH 7.4) containing kanamycin and then incubated at 37~C until colonies appear.

D. DNA Isol~t~on Following transformation, kanamycin resistant cells were picked and inoculated into 5 ml of TY broth (1%
tryptone, 0.5% yeast extract, 1% NaCl, pH 7.4) containing kanamycin. These cultures were grown for about 8 hours at 37~C with aeration. Plasmid DNA was isolated from cultures using Magic Minipreps obtained from (Promega). Recombinant plasmids were identified by digestion with restriction endonucleases ~ollowed by gel electrophoresis analysis To obtain larger amounts of pM100-neo plasmid DNA, large scale isolation was performed using the Plasmid Purification Kit (Qiagen, Inc.).
~xam~le 5 ~onstruction of ~MT-h+tLB+Val8GLP-l A. Pre~aration of ~stllO7I-BclI Diaested ~GT-h The plasmid pGT-h was isolated as described in substantial accordance with Example 2A. Three ~g (~.6 ~1) of pGT-h DNA was digested to completion with 1 ~1 (8 units) of ~1107I in a 10 ~1 reaction volume containing lOmM Tris-HCl, lOmM MgCl2 , lOOmM KCl (pH 8.5 at 37~C). The sample was incubated at 37~C ~or 1 hour. The BstllO7I digested pGT-h DNA was purified from the reaction components using the Wizard DNA Clean-Up System (Promega). The stllO7I digested pGT-h DNA was concentrated in a 6.7 ~l volume using a microcon 50 (Amicon, Inc. 72 Cherry Hill Drive, Beverly, MA
01915).
The 6.7 ~1 of ~g~1107I digested pGT-h DNA was further digested with 1 ~l (10 units) of ~lI in a reaction volume of 10 ~1 containing 50mM Tris-HCl (pH 8.0), lOmM
MgC12, 50mM NaCl. The sample was incubated at 50~C for 1 hour. To prevent any possible recircularization, the ~stllO7I-BclI digested pGT-h was dephosphorylated (removal of 5' phosphate groups) by the addition of 1 ~l (1 unit) of calf intestinal phosphatase to the sample. The sample was incubated at 37~C for 30 minutes. Gel loading dye (0.25%
bromophenol blue, 0.25% xylene cyanol, 30% glycerol in water) was added to the reaction. The reaction was loaded into the preparative well of a 1% SeaPlague GTG agarose/TAE bu~fer gel and then electrophoresed. The gel was stained with ethidium bromide and then the DNA was visualized by ultraviolet light.
The desired 6.2 Kb DNA band was excised with a scalpel and placed into a micro-tube. The DNA was puri~ied from the low melting point agarose using the Wizard PCR Preps DNA
Puri~ication System (Promega).
-W O 97/29180 PCT~US97/01978 B. Pre~aration of ~M10~-neo EcoRV-BalII Fra~ment Five ~g (3.8 ~l) o~ pM100-neo DNA constructed in Example 4 was digested to completion with 1 ~l (10 units) of ~coRV in a 10 ~1 reaction volume containing 50mM Tris-HCl (pH
8.0), lOmM MgCl2, 50mM NaCl. The reaction was incubated at 37~C for 1 hour. The ~coRV digested pM100-neo was further digested with 1 ~l (lOunits) of ~glII in a 20 ~l reaction volume containing 50mM Tris-HCl (pH 8.0), lOmM MgCl2, 50mM
NaCl. The sample was incubated at 37~C ~or 1 hour. Gel loading dye (0.25% bromophenol blue, 0.25% xylene cyanol, 30%
glycerol in water) was added to the reaction. The reaction was loaded into the preparative well of a 1~ SeaPlaque GTG
agarose/TAE buffer gel and then electrophoresed. The gel was stained with ethidium bromide and then the DNA was visualized by ultraviolet light. The desired 1.8 Kb DNA band was excised with a scalpel and placed into a micro-tube. The DNA
was puri~ied ~rom the low melting point agarose using the Wizard PCR Preps DNA Purification System (Promega).
C. F;n~l Construction of ~MT-h+tLB+Val8GLP-1 The DNA prepared in Example 5A was ligated with the DNA prepared in Example 5B and the DNA prepared in Example 3 (pLP53-tLB+Val8GLP~ HI fragment). Four ~l of DNA from Example 5A, 3 ~l of DNA from Example 5B and 12 ~l of DNA
prepared in Example 3 (pLP53-tLB+Val8GLP-1 E~mHI fragment) were ligated in a reaction that contained 1 ~l (1 unit) of T4 DNA ligase, 50rnM Tris-HCl (pH 7. 6), lOmM MgCl2, lmM ATP, lmM
DTT, and 50% ~w/v) polyethylene glycol-8000 in a total volume of 25 ~l. The mixture was incubated at 16~C for 16 hours.
Frozen competent ~. coli K12 DH5a cells were transformed using about 3-4 ~l of the ligation reaction in substantial accordance with Example lD, and the plasmid DNA
was isolated in substantial accordance with Example lE.

W O 97/29180 PCTnUS97/01978 E~mnl e 6 Construction of St~hle Cell lines Plasmid DNA, either pGT-h+tLB+GLP-l, pGT-h+tLB+val8GLP-l, or p~T-h+tLB+val8GLP-l was transfected into human embryonic kidney 293 cells using the stable CaPQ4 method contained in the Mammalian Transfection Kit available from Stratagene (11011 Morth Torrey Pines Road, LaJolla, CA
92037). Selection was achieved by the addition o~ 300~g/ml of hygromycin B (Eli Lilly and Company In~;~n~polis, IN
46285) to the culture medium. Monoclonal cell lines were expanded and screened ~or the ability to secrete the corresponding protein of SEQ ID NO 1 into t~e culture medium.
The presence of biologically active GLP-1(7-37)-based protein in the culture medium was determined by measuring the amount of luci~erase enzyme present in a biological system that expressed luci~erase enzyme following stimulation with GLP-l.
~ ~le 7 Im~lantation The transformed 293 cell were cultured then surgically transplanted under the kidney capsule o~ 8 week old zucker Diabetic Fatty (ZDF/GmiTM-~a/~a) male rats. Under iso~urane anesthesia, a dorsal incision was made just posterior to the diaphram, and using a rib spreader, the kidney was exposed. Approximately 20 million trans~ormed 293 cells, in 200 ~l of Hank's buffer, were in~ected just under the kidney capsule using a 23 gauge blunt needle. The incision was sutured and protected from chewing with wound clips .

Claims (18)

We Claim:

1. A method of treating Type I or Type II
diabetes in a mammal in need thereof comprising implanting a cell line transformed with a vector comprising a driving expression of a DNA sequence encoding a protein of the formula:
His-Xaa1-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-Ser-Tyr-Leu-Xaa2-Gly-Gln-Ala-Ala-Xaa3-Xaa4-Phe-Ile-Ala-Trp-Leu-Val-Lys-Gly-Arg-Xaa5 (SEQ ID NO 1) wherein Xaa1 is Ala, Gly, Val, Thr, and Ile;
Xaa2 is Glu, Gln, Ala, Thr, Ser, and Gly;
Xaa3 is Lys, and Arg;
Xaa4 is Glu, Gln, Ala, Thr, Ser, and Gly; and, Xaa5 is Gly-OH or is absent;
into said mammal such that it is immunologically isolated from the mammal's immune system and secretes a protein of SEQ
ID NO. 1 into said patient.

2. A method of treating Type I or Type II diabetes in a mammal in need thereof comprising implanting a cell line transformed with a vector comprising a promotor driving expression of a DNA sequence encoding a protein of SEQ ID NO
1 into said mammal wherein said mammal is under immunosuppression therapy.

3. The method of Claims 1 or 2 wherein Xaa1 is Ala or Val, Xaa2 Glu, Xaa3 is Lys or Arg, Xaa4 is Glu, and Xaa5 Gly-OH or is absent.

4. The method of Claims 1 or 2 wherein Xaa1 is Ala, Xaa2 Glu, Xaa3 is Lys, Xaa4 is Glu, and Xaa5 Gly-OH.

5. The method of Claims 1 or 2 wherein Xaa1 is Val, Xaa2 Glu, Xaa3 is Lys, Xaa4 is Glu, and Xaa5 Gly-OH.

6. The method of Claims 1 or 2 wherein the promotor is a viral promotor.

7. The method of Claims 1 or 2 wherein the promotor is a metallothionein promotor.

8. The method of Claims 1 or 2 wherein the DNA
coding sequence is 5' - CAT GCT GAA GGG ACC TTT ACC AGT GAT GTA AGT TCT TAT TTG
GAA GGC CAA GCT GCC AAG GAA TTC ATT GCT TGG CTG GTG AAA
GGC CGA GGA - 3'. (SEQ ID NO 2)

9. The method of Claims 1 or 2 wherein the DNA
coding sequence is 5' - CAT GTT GAA GGG ACC TTT ACC AGT GAT GTA AGT TCT TAT TTG
GAA GGC CAA GCT GCC AAG GAA TTC ATT GCT TGG CTG GTG AAA
GGC CGA GGA - 3'. (SEQ ID NO 4)

10. The method of Claims 1 or 2 wherein the cell line is the human embryonal kidney cell line 293 transformed with a vector selected from the group consisting of pGT-h+tLB+GLP-1, pGT-h+tLB+Val8GLP-1, or pMT-h+tLB+Val8GLP-1.

11. The method of Claims 1 or 2 wherein the cell line is the human embryonal kidney cell line 293 transformed with the vector pGT-h+tLB+GLP-1.

12. The method of Claims 1 or 2 wherein the cell line is the human embryonal kidney cell line 293 transformed with the vector pGT-h+tLB+Val8GLP-1.

13. A method of treating Type I or Type II
diabetes in a mammal in need thereof comprising injecting an expression vector of any one of Claims 1 to 12 directly into the mammal such that the expression vector is incorporated into a cell of the mammal and secretes a protein of SEQ ID
NO. 1.

14. A stable transformed cell line of any one of Claims 1 to 12.

15. A vector of any one of Claims 1 to 12.

16. A method of treating Type I or Type II
diabetes in a mammal in need thereof comprising implanting a cell line transformed with a vector comprising a promotor driving expression of a DNA sequence encoding a protein of the formula:
His-Xaa1-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Val-Ser-Ser-Tyr-Leu-Xaa2-Gly-Gln-Ala-Ala-Xaa3-Xaa4-Phe-Ile-Ala-Trp-Leu-Val-Lys-Gly-Arg-Xaa5 (SEQ ID NO 1) wherein Xaa1 is Ala, Gly, Val, Thr, and Ile;
Xaa2 is Glu, Gln, Ala, Thr, Ser, and Gly;
Xaa3 is Lys, and Arg;
Xaa4 is Glu, Gln, Ala, Thr, Ser, and Gly; and, Xaa5 is Gly-OH or is absent;
into said mammal such that it is immunologically isolated from the mammal's immune system substantially as hereinbefore described with reference to any one of the Examples.

17. A vector encoding a protein of SEQ ID NO 1 substantially as hereinbefore described with reference to any one of the Examples.

18. A stable mammalian cell line transformed with a vector capable of secreting a protein of SEQ ID NO 1 substantially as hereinbefore described with reference to any one of the Examples.