WO2003066832A2

WO2003066832A2 - Generation of new insulin cells from progenitor cells present in adult pancreatic islets

Info

Publication number: WO2003066832A2
Application number: PCT/US2003/003859
Authority: WO
Inventors: Teitelman Gladys
Original assignee: The Research Foundation Of The State University Of New York
Priority date: 2002-02-07
Filing date: 2003-02-07
Publication date: 2003-08-14
Also published as: WO2003066832A3; AU2003217357A8; AU2003217357A1

Abstract

The present invention is directed to an in vitro method for converting noninsulin producing cells to insulin-producing cells. A continuous source of insulin-producing cells is obtained. Insulin from these insulin-producing cells may be collected and administered to a mammal, or the insulin-producing cells may themselves be transplanted into a mammal. Also provided are insulin-producing cells made by the methods of the invention, methods for culturing insulin-producing cells, and methods for treating diabetes with the insulin-producing cells.

Description

GENERATION OF NEW INSULIN CELLS FROM PROGENITOR CELLS PRESENT IN ADULT PANCREATIC ISLETS

GOVERNMENT RIGHTS This invention was made with government support under NIH Grant #DK- 053870. The Government has certain rights in the invention.

BACKGROUND

1. Technical Field

The present invention relates to a unique source of purified immature cells obtained from the islets of Langerhans which may be used to provide a continuous source of insulin-producing cells. These cells may be utilized to produce insulin or may be transplanted into an individual suffering from diabetes.

2. Background of Related Art

The pancreas is composed of an exocrine and endocrine component. The endocrine tissue is scattered throughout the gland in clusters of cells called the islets of Langerhans. Each pancreatic islet is composed of four islet cell types, which produce glucagon (o; cells), insulin (β cells), somatostatin (δ cells), and pancreatic polypeptide (PP cells), respectively. Insulin (IN) containing β cells form the core of the mature islets, whereas the periphery contains lower numbers of the other endocrine cell types.

Lack of insulin occurs when there is injury or death of the β cells which leads to disregulation of circulating blood glucose levels and diabetes. One of the main areas of research in the diabetes field is the availability of new β cells suitable for transplantation. The possible sources of cells for transplantation currently being investigated are insulin-producing cell lines, embryonic stem cells, and "neural" type stem cells present in adult islets.

One of the main disadvantages in using insulin-producing cell lines is that these cells are neoplastic cells which require the successful downregulation of tumor genes utilized to form the cell lines prior to their transplantation.

While embryonic stem cells can be exposed to growth factors to differentiate into insulin-producing cells, only a small percentage of the stem cells will transform into β cells and those that do generate only a minimal amount of insulin. However, for reasons still unknown, the insulin-producing cells eventually lose their β cell traits and return to their original undifferentiated phenotype.

"Neural" type stem cells are found in adult islets. These cells express nestin, a marker of neuronal stem cells, and are reported to initiate expression of insulin and glucagon producing cells after exposure to a cocktail of growth factors.

Other sources of β cells are human islets, obtained from the pancreas post- mortem, and pig islets. However, the number of human islets obtained is low and both human and pig islets have to be encapsulated prior to transplantation to prevent their destruction by the recipient's immune system. The technology for encapsulation is still in a primitive stage. In addition, it is not possible to expand the original β cell population because these cells become senescent following a prolonged period of proliferation and lose their responsiveness to glucose.

Previous in vivo studies in mice have noted that when the population of β cells present in the islets of Langerhans are eliminated by a β cell toxin, new insulin containing cells differentiate in the islets. Fernandes, et al. "Differentiation of New Insulin-Producing Cells is Induced by Injury in Adult Pancreatic Islets", Endocrinology, Vol. 138, No. 4, pp. 1750-1762 (1997); Guz, et al. "Regeneration of Pancreatic β Cells from Intra-Islet Precursor Cells in an Experimental Model of Diabetes", Endocrinology, Vol. 142, No. 11, pp. 4956-4968 (2001); Guz, et al. "Detrimental Effect of Protracted Hyperglycaemia on Beta-Cell Neogenesis in a Mouse Murine Model of Diabetes", Diabetologia, 45:1689-1696 (2002). In the above-referenced studies in mice, the biochemical properties of the progenitor cells were examined in order to determine whether these cells expressed embryonal properties. The different endocrine cell types appear sequentially during development. Glucagon producing (GLU+) cells and insulin producing (IN+) cells were first seen at postnatal day 9.5 and 10 respectively, somatostatin (SOM) at postnatal day 15 and pancreatic polypeptide (PP) at postnatal day one (Alpert et al., "Hybrid Insulin Genes Reveal a Developmental Lineage for Pancreatic Endocrine Cells and Imply a Relationship With Neurons", Cell, 53: 295-308 (1988); Gittes et al., "Onset of Cell-Specific Gene Expression in the Developing Mouse Pancreas", Proc. Natl. Acad. Set USA, 89: 1128-1132 (1992); Guz et al., "Expression of Murine STF- 1 , a Putative Insulin Gene Transcription Factor, in β Cells of Pancreas, Duodenal Epithelium and Pancreatic Exocrine and Endocrine Progenitors During Ontogeny", Development, 121: 11-18 (1995); Herrera et al. "Embryogenesis of the Murine Endocrine Pancreas; Early Expression of the Pancreatic Polypeptide Gene", Development, 113: 1257-1265 (1991)) whereas acinar cell differentiation starts at e- 15 (embryonic day 15) (Gittes et al. supra; Pictet et al., "Development of the

Embryonic Pancreas", n Handbook of Physiology, Section 7 (Steiner et al. eds.) pp. 25-69 (American Physiological Society, 1972)).

One of the first molecular markers expressed by islet stem cells during development is a homeodomain protein, pancreas duodenum homeobox gene-1 (Pdx- 1). .Pdx-1 plays a key transcriptional role during the development of both the endocrine and exocrine pancreas as demonstrated by targeted mutagenesis experiments in mice. Pdx-1, previously termed STF-1, IDX-1, or IPF-1, is transiently expressed by precursor cells present in the pancreatic duct and by each of the four endocrine cell types when they first differentiate. Pancreatic Pdx-1 expression becomes progressively restricted during development such that it is almost exclusively localized in β cells in adults, where it is believed to play a role in the regulation of IN gene transcription. In addition to β cells, a small subset of the δ and PP cells express Pdx-1.

It was recently found that precursors expressing the pancreas duodenal homeobox gene Pdx-1 (Wright et al., "XIHbox 8: A Novel Xenopus Homeoprotein Restricted to a Narrow Band of Endoderm", Development, 104: 787-794 (1988)) give rise to endocrine, exocrine and ductal cells expressing Neurogenin 3 (Ngn3) (Gu et al., "Direct Evidence for the Pancreatic Lineage: Ngn3+ Cells are Islet Progenitors and are Distinct From Duct Progenitors", Development, 129: 2447-2458 (2002)). Ngn3 is a member of the basic helix-loop-helix family of proneural genes expressed during early development which gives rise to endocrine precursors. These progenitors turned off Ngn3 expression prior to the initiation of hormone synthesis (Jenssen et al., "Independent Development of Pancreatic cuand β cells From Neurogenin-3 Expressing Precursors", Diabetes, 49: 163-176 (2000); Schwittzgebel et al., "Expression of Neurogenin 3 Reveals an Islet Cell Precursor Population in the

Pancreas", Development, 127: 3533-3542 (2000)). In addition to Pdx-1 andNgn3, endocrine precursors express glucose transporter 2 (GLUT-2), a membrane protein involved in the active transport of glucose into β cells. GLUT-2 expression is downregulated in non-/3 cells during maturation but its expression remains in β cells throughout life. However, The cell lineage relationship of pancreatic islet cells remains controversial. One approach assumes that expression of common molecular markers by embryonic cells indicates a cell lineage relationship. This model (Fig. 1) suggests that islet cells differentiate from precursors that first express Pdx-1, then initiate synthesis of GLUT-2 and Ngn3 and turn off Ngn3 expression prior to the initiation of hormone synthesis.

The increase in β cell number during embryonic development and early postnatal life occurs by continued differentiation of new islet precursor cells from the pancreatic duct and by proliferation of differentiated insulin producing cells. In adults, new insulin producing cells arise mostly from division of existing β cells.

Neogenesis, the differentiation of insulin cells from stem-like cells present in the pancreas, has been reported to occur in several animal models in addition to the in vivo mouse studies described above, including after β cell destruction with toxins in neonatal rats (Dutrillaux, et al., "Ultrastructural Study of Pancreatic β Cell Regeneration in Newborn Rats After Destruction by Streptozotocin", Vir chows Arch (Cell Pathol), 39:173-185 (1982)), after cellophane wrapping of the pancreas in adults (Rosenberg, et al., "Induction of Endocrine Cell Differentiation: A New Approach to Management of Diabetes", J Lab Clin Med, 114:75-83 (1989)), after partial pancreatectomy (Bonner-Weir, et al., "A Second Pathway for Regeneration of Adult Exocrine and Endocrine Pancreas", Diabetes 42: 1715-1720 (1993)), and in some transgenic mouse models of autoimmune diabetes (Gu, et al., "Epithelial Cell Proliferation and Islet Neogenesis in IFN-g Transgenic Mice", Development, 118:33- 46 (1993); and Gu, et al., "Transitional Cells in the Regenerating Pancreas", Development, 120:1873-1881 (1994)). However, all these studies were limited to in vivo approaches and attempts to demonstrate how the new β cells were generated were hampered by the inability to unambiguously identify the precursor cells.

J The present invention provides methods for identifying and isolating these cells in vitro and developing a continuous and/or renewable source of insulin- producing cells.

SUMMARY

The present invention provides a method for converting non-insulin producing cells of the islets of Langerhans into insulin-producing cells, said method comprising isolating islets from pancreatic tissue; contacting the islets with a toxic substance which selectively kills the β cells; and sustaining the remaining cells in culture under , conditions which convert the non-insulin producing cells to insulin-producing cells. Suitable toxic substances include streptozotocin, alloxan, interleukin-1 beta, tumor necrosis alpha, interferon-γ, and combinations thereof. The methods of the present invention can provide a continuous source of actively proliferating insulin-producing cells. The proliferation of the remaining cells of the islets may be sustained by embedding them in a matrix, including collagen and basement membrane derivatives extracted from tumor cells, or by adding a growth factor. Suitable growth factors for use in the methods of the present invention include hepatocyte growth factor/scatter factor, activin, exendin 4, betacellulin, nicotinamide, culture media supplement B27, culture media supplement ITS, epidermal growth factor, basic fibroblast growth factor, and glucose.

The proliferation of the non-insulin producing cells of the islets may also be amplified by transfecting non-insulin producing precursor cells with a construct combining an oncogene and a promoter of a gene expressed by the precursor cells but not by the insulin-producing cells. Preferably, the precursor cells are cells which are transfected with a construct combining an oncogene and a promoter of the glucagon gene.

The present invention is also directed to methods for in vitro culturing insulin- producing cells which results in expanding their population and maintaining their proliferation. Preferably, the insulin-producing cells are maintained under conditions of normoglycemia at a temperature equivalent to the body temperature of the mammal from which the islets were obtained.

The present invention is also directed to methods for producing insulin from these insulin-producing cells, which can be collected for administration to an individual suffering from diabetes. The insulin-producing cells may also be transplanted into a hyperglycemic individual suffering from diabetes.

The present invention also encompasses pharmaceutical compositions containing the insulin-producing cells and a pharmaceutically acceptable carrier.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a graphical depiction of one suggested model of the differentiation of islet cells.

FIGs. 2A-D are photomicrographs of control islets, islets treated with lOmM streptozotocin (SZ) at 1 day in vitro (DIV), and cultures fixed 24 hours after SZ treatment (2 DIV).

FIGs. 3A-C are photomicrographs of: a control islet at 3 DIV stained for insulin (IN) and Pdx-1; an islet treated with 10 mM SZ and maintained with media containing 11 mM glucose; and NBCs produced in accordance with the present invention that have differentiated in vitro. FIG. 4 is a series of four photomicrographs of control NBCs andNBCs subjected to hyperglycemia demonstrating the detrimental effects of hyperglycemia on the differentiation of the NBCs.

FIGs. 5A-B are photomicrographs of control islets and islets treated with SZ and maintained for 2 days in vitro in the presence of BrdU demonstrating the proliferation of the NBC's.

FIG. 6 is a photomicrograph of SZ-treated islets that were maintained in culture for 2 days and then fixed with guinea pig antisera to rat C-peptide which demonstrates that the cells produce insulin rather than taking up the hormone from the culture media.

DESCRIPTION OF PREFERRED EMBODIMENTS

In accordance with the present invention, in vitro methods are provided for converting non-insulin producing cells of the islets of Langerhans into insulin- producing cells. The method utilizes, as starting material, islets removed from an animal. By increasing the number of precursor cells, which are non-insulin producing cells, and taking advantage of their ability to convert into insulin-producing cells and proliferate, the methods of the present invention have the potential to provide an unlimited supply of insulin-producing cells. As used herein, the terms "insulin-producing cells", "new Beta cells", and

"NBCs" are used interchangeably to refer to those cells found in the islets of Langerhans which are not β cells and do not produce insulin in an untreated pancreas, but after treatment in accordance with the methods of the present invention do, in fact, produce insulin. First, islets of Langerhans are isolated from pancreatic tissue by methods known to those skilled in the art. Islet of Langerhans cells can be prepared from the pancreas of a single mammal or several mammals of the same species. For example, the pancreas may be digested using collagenase and the islets collected with a fine tip pipette under a microscope. For pancreases from larger animals (other mammals and humans), the islet of Langerhans cells can be further isolated using conventional gradients. Preferably, islet purification is based on density differences between islets and exocrine tissue and is performed on a discontinuous Ficoll gradient. After density centrifugation, the islets are collected, washed several times and put in suitable tissue culture, such as different formulations of MEM or Hams F12 based media (available from Sigma-Aldrich, St. Louis, MO; Gibco-Invitrogen Corporation, Carlsbad, CA).

Suitable human sources of pancreatic tissue include cadavers and patients suffering from diabetes. Where a patient is the source of pancreatic tissue, in a preferred embodiment the patient is also the recipient of the insulin-producing cells or insulin made by such cells .

Once collected, the islets are then contacted with a toxic substance which selectively kills β cells present therein. Such substances include, but are not limited to, streptozotocin (SZ), Alloxan, and cocktails of cytokines (e.g., interleukin-1 beta, tumor necrosis alpha, interferon-γ, and combinations thereof). Streptozotocin, which consists of 1 -methyl- 1 nitrosourea attached to the C-2 of D-glucose, is used for rodent islets and cytokines are used for human islets. It is believed that the basis of the selective effect of streptozotocin on β cells is its glucose moiety, which reacts with glucose sensing mechanisms within β cells to deplete nicotinamide adenine dinucleotide (NAD), producing long-lasting impairment and death of β cells. Photomicroscopy and radio-immunoassay (RIA) of insulin release into culture media by dying cells may be conducted after the islets are treated with the toxic substance to confirm the death of pre-existing β cells.

The remaining living cells of the islets of Langerhans are then sustained under conditions which convert non-insulin producing cells therein to proliferating insulin- producing cells, after which time the insulin-producing cells are isolated. These conditions include, but are not limited to, normoglycemia; embedding the remaining cells in a matrix such as soluble basement membrane derivatives extracted from tumor cells; the prior removal of pre-existing β cells; and combinations thereof. Suitable methods for culturing the cells can be determined by those skilled in the art using known methodologies. Some of the more relevant parameters are: substrate and/or matrix for cell growth; cell density and cell contact; the medium; temperature; and gas phase.

Specific substrates and growth factors can be used to induce or increase the proliferation of these precursor cells so that they develop into NBCs. Preferably a matrix, into which the cells are embedded, is used. Suitable matrices include, but are not limited to, collagen and basement membrane derivatives extracted from tumor cell lines including, but not limited to, human tumor bladder-9 cells. Commercially available matrices such as Matrigel™ (Becton Dickinson and Company, Franklin Lakes, NJ) may also be used.

Basement membranes are thin extracellular matrices underlying cells in vivo. Their major components are laminin and collagen, and they may also contain growth factors. Basement membranes may be extracted from tumor cell lines. Such basement membranes are soluble and can be induced to polymerize. In a polymerized state, they provide a three dimensional matrix that maintains the cell-cell and cell- matrix interaction similar to that of whole tissue in vivo.

In addition, it is possible to expand the population of non-insulin producing precursor cells in the islets, thus amplifying their proliferation, by the transfection of precursor cells with constructs combining an oncogene with a promoter of a gene expressed by the precursor cells that becomes inactive once the precursor cells are converted to insulin-producing cells and initiate insulin synthesis. Preferably, the population of cells in the islets may be expanded by the transfection of cells with constructs combining an oncogene with the promoter of the glucagon gene. The oncogene will induce the proliferation of the a cells, but once these transformed glucagon cells are induced to convert into insulin-producing cells, the glucagon promoter will be inactive, the expression of the oncogene will be turned off (because the cells no longer express glucagon), and the cells will cease proliferating.

Once embedded in a matrix, the cells are then placed in a culture medium and allowed to gel. Suitable culture media are known to those skilled in the art and include, but are not limited to, fetal calf serum in combination with commercially available salt solutions (such as RPMI 1640) (Sigma- Aldrich, St. Louis, MO; Gibco- Invitrogen Corporation, Carlsbad, CA) (RPMI is a commercially available salt solution named for Roswell Park Memorial Institute). Antibiotics, including penicillin and streptomycin, may also be added to the culture medium.

Culture media may contain classical media with or without HEPES. Such media include, but are not limited to Dulbecco, Modified Dulbecco plus nutrient mixtures, RPMI 1640, Media 199, Minimum Essential Media with different salt components and Balanced Salt Solutions (Dulbecco's, Hanks, or Earle's Balanced Salt Solutions), which are all commercially available. The media may contain 10% fetal calf serum, which may be heat inactivated, or a combination of fetal calf serum and horse serum. The media may also be a serum free formulation. (All of the above media components may be obtained from Sigma-Aldrich, St. Louis, MO; Gibco- Invitrogen Corporation, Carlsbad, CA; HyClone Laboratories, Logan, UT.) Optionally, growth factors may also be included in the culture. Suitable growth factors include, but are not limited to, hepatocyte growth factor/scatter factor (HGF/SF), activin (a member of the TGF/3 supergene family), exendin 4 (an analog of GLP-1), betacellulin, nicotinamide, culture media supplement B27 (Gibco BRL, Gaithersburg, MD), culture media supplement ITS (insulin/transferrin/selenium) (Sigma), epidermal growth factor, basic fibroblast growth factor, and glucose.

Preferably, culture conditions are maintained such that the insulin-producing cells themselves proliferate. In a preferred embodiment, glucose is added to the cultures to maintain the insulin-producing cells in normoglycemia at a temperature equal to the body temperature of the mammal from which the islets were obtained. Without wishing to be bound by any theory, it is also believed that the prior removal of pre-existing β cells enhances the ability of the insulin-producing cells to proliferate. Cultures are generally allowed to incubate, preferably in a CO2 incubator

having a CO2 concentration of about 5% (and about 95% O2) for a period of time

ranging from about 24 hours to about 240 hours, more preferably from about 40 hours to about 96 hours, and most preferably from about 48 hours to about 72 hours. Where the islets are obtained from a human, the temperature of the culture is preferably from about 34°C. to about 40°C, with a temperature of about 37°C. being most preferred. Where the source of the islets is a mouse, glucose is preferably added to the culture at a range of from about lOmM to about 12mM, with about 1 lmM glucose being most preferred. Where the source of the islets is human, glucose is preferably added to the culture at a range of from about 4.6mM to about 6.6mM, with about 5.6mM glucose being most preferred.

Photomicroscopy, measurement of insulin release by glucose, and determination of insulin mRNA by RT-PCR, northern blots, western blots and/or ELISAs, may then be conducted to confirm the presence of insulin containing NBCs.

In those cases where the NBCs prepared by the methods of the present invention are used to produce insulin for injection by patients suffering from diabetes, the production of insulin may also be confirmed and measured by analysis of insulin protein expression (preproinsulin, proinsulin, insulin, or c-peptide) utilizing methods known to those skilled in the art including, but not limited to, western blots, radioimmune assays, ELISAs, and the like. Downstream effects of insulin modulation can also be examined.

Insulin may be collected by methods known to those skilled in the art including, but not limited to, the assays described above such as radioimmune assays, chromatography, including cation exchange and anion exchange chromatography, and HPLC.

NBCs prepared by the methods of the present invention may be collected from the culture by means known to those skilled in the art, including, but not limited to, removing the NBCs from the matrix using an appropriate probe such as a rubber policeman and centrifugation. Once collected, the NBCs may be transplanted into an individual suffering from a pancreatic disorder characterized by a deficiency of insulin, especially diabetes.

In a preferred embodiment, matrices containing the NBCs are lifted from the culture, washed by centrifugation, collected, and then transplanted into a mammal. If isolated NBC's are needed, the cells can be released from the matrix using enzymes such as Dispase (available from BD Biosciences (San Jose, CA); Gibco-Invitrogen Corporation). Other cells, including but not limited to Sertoli cells, may also be included with the NBCs for transplantation.

Suitable transplantation methods are conventional techniques capable of introducing the cells into the mammal such as parenteral administration, subcutaneous administration, biocompatible scaffolds, sponge or matrix delivery, or intraperitoneal administration.

Preferred locations for transplantation of the NBCs have an available blood supply to provide nourishment for the transplanted cells and dense tissue to keep the transplanted cells within close proximity of each other. Suitable transplantation sites include, but are not limited to, the renal subcapsular space, subcutaneous facie, salivary glands, spleen, liver and the hepatic portal vein. Cell stimulatory factors may be included in transplanting the NBCs to enhance the viability of the NBCs.

In a preferred embodiment, the present invention provides a method of treating diabetes mellitus by transplanting NBCs into a mammal, most preferably a human. As used herein, the term "allograft" is understood to mean the transfer of tissues or cells between two different mammals of the same species. The term "xenograft" as used herein describes the transfer of tissues or cells between two mammals of different species. In another preferred embodiment, the islets utilized to generate the NBCs are obtained from a mammal suffering from diabetes, the NBCs are produced by the methods of the present invention, and the resulting NBCs are then re-implanted in the donor mammal as an autograft.

The matrices containing the NBCs or the isolated NBC's can be transplanted into a mammal using a variety of conventional techniques including, but not limited to: injection into a tissue mass; microencapsulation inside biocompatible membranes, hydrogels, or reticulated thermoplastics which are then injected, transplanted or introduced subcutaneously into a tissue mass; introduction into a pouch, such as an intestinal pouch, an omental pouch, or a gastric pouch; introduction into a biocompatible polymeric scaffold, sponge or matrix consisting of, e.g., polylactic acid, or reticulated thermoplastics such as acylnitrile vinyl chloride copolymer (PAN- PVC) and the like. The cells are then transplanted into the recipient. Optionally, other cells such as Sertoli cells may also be introduced with the NBCs.

Where the NBCs are allografts or xenografts, encapsulation of the NBCs is one method that may be utilized to prevent rejection by the recipient's immune system. Methods by which the NBCs may be encapsulated are known to those skilled in the art and include those procedures described by Goosen, et al. in Biotech. Bioeng., 27: 146-150 (1985), and Weber, et al. U.S. Pat. No. 5,227,298, the contents of each of which are incorporated by reference herein.

In one embodiment, NBCs, which may be in a matrix or isolated, are encapsulated in an insulin permeable encapsulant. Preferably, such encapsulant is hypoallergenic, is easily and stably situated in a target tissue, and provides added protection to the implanted structure such as to avoid destruction of the NBCs.

Other methods for encapsulation are known to those skilled in the art and include the following: (1) suspending a pharmaceutically effective amount of NBCs in combination with a gelling agent in an aqueous medium which is physiologically compatible with the cells and extruding the NBC/gelling agent mixture to form a droplet containing the NBCs;

(2) subjecting the product of step (1) to an effective amount of network forming cations to form discrete capsules capable of encapsulating the NBCs; (3) forming a semipermeable membrane around the capsules to obtain a single-walled bead encapsulating the NBCs; and

(4) contacting the single-walled bead with a second gelling agent to form a second semi-permeable membrane encapsulating the product of step (3). Preferred gelling agents are water soluble, natural or synthetic polysaccharide gums such as alkali metal alginate.

Procedures for transplanting cells in pouches, such as omental or gastric pouches, are known to those skilled in the art and include the methods described by Amiri, et al. (1990) Arch. Surg. 125:1472-1474 and Bayat, et al. (1995) Surg. Res. Commun. 17 : 87-91 , the contents of each of which are incorporated by reference herein.

Procedures for incorporating NBCs in polymeric scaffolds are also known to those skilled in the art. Preferably, the polymeric template is biodegradable and comprises a polyvinyl alcohol, e.g., poly-L-lactic acid. Polyvinyl alcohol-based templates are preferred because of their porosity, which permits rapid tissue ingrowth and prevascularization before cell transplantation.

Preferably, the NBCs are transplanted into the peritoneal space, the renal subcapsular space or subcutaneous facie.

In those cases where the NBCs are allografts or xenografts, it may also be necessary to provide the transplant recipient with immunosuppressive agents to prevent transplant rejection. Suitable immunosuppressive agents include, but are not limited to, cyclosporine, tacrolimus, atrazine, despergualin and monoclonal antibodies to, e.g., T cells. In a preferred embodiment the immunosuppressive agent is cyclosporine administered at a dosage of from about 0.5 mg to about 200 mgkg body weight, more preferably from about 5 mg to about 40 mg/kg body weight. The immunosuppressive agent should be administered for a time sufficient to permit the transplanted NBCs to be functional and capable of producing therapeutically effective amounts of insulin. In a preferred embodiment, the immunosuppressive agent is administered for a time period ranging from about 40 to about 100 days following transplantation, more preferably from about 50 to about 60 days.

In some cases, a recipient mammal may also achieve systemic tolerance by injecting antibodies to mask surface antigens on transplanted cells, which renders the transplanted cells invisible to the host's immune system. In addition, a sample of pancreatic tissue from a diabetic patient may be removed, the islets treated to generate a large supply of NBCs, and these cells may then be re-implanted into the patient to regulate blood glucose levels. This approach will circumvent the problem of allograft rejection.

Another aspect of the present invention provides a pharmaceutical composition comprising NBCs and a pharmaceutically acceptable carrier. In one embodiment additional cells, such as Sertoli cells, are included in the pharmaceutically acceptable carrier. A further embodiment of the present invention comprises using porcine, bovine or human NBCs in combination with porcine, bovine or human Sertoli cells. As used herein, a "pharmaceutically acceptable carrier" includes any and all biological and non-biological biocompatible membrane materials, and can include any conventional solvents, dispersion media, coatings, antibacterial and antifungal agents, isotonic agents and the like known to those skilled in the art.

The method of the present invention provides a continuous source of endocrine hormones, especially insulin, which can be recovered from the culture medium or which can be directly released into an animal by implantation of the NBCs into an animal. Applying this system to human islets removed from cadavers could provide an unlimited supply of β cells that could be especially beneficial in treating hyperglycemic patients suffering from diabetes.

EXAMPLES

The following examples are merely illustrative of certain aspects of the invention and should not be construed as limiting the invention in any manner.

EXAMPLE 1 Islets isolated from mouse pancreatic tissue were collected in 2 ml of Hanks

Balanced Salt Solution (HSS, from Sigma-Aldrich Company, Gibco-Invitrogen Corporation), exposed to a β cell toxin, streptozotocin 10 mM (USB, Cleveland, Ohio; also available from Sigma Aldrich), for 15 minutes, washed with 8 ml HSS, transferred to cell culture coverslips (13mm diameter from Nalge Nunc International, Rochester, NY), embedded in 20 microliters of a 1 : 1 dilution of Matrigel™ (Becton Dickinson and Company, Franklin Lakes, NJ) and Hanks Balanced Salt Solution, and placed in 2 mL of culture media. The culture media contained 10% heat-inactivated fetal calf serum, and 90% of a salt solution (RPMI 1640) (Sigma-Aldrich Company, also available from Gibco-Invitrogen Corporation) in combination with glucose in a concentration of 1 ImM, a combination of penicillin and streptomycin at 10,000 U/ml (also available from Sigma-Aldrich Company, and Gibco-Invitrogen Corporation) and 15 mM of HEPES buffer. The glucose concentration of 1 ImM is within the physiological range for murine islets. The cultures were then placed in a CO2

incubator at 5% CO2 a temperature of 37°C for 48 hours, and constant humidity achieved through the process of water evaporation from a stainless steel pan placed on the bottom of the incubator.

EXAMPLE 2 The cultures prepared in Example 1 were processed for microscopic analysis to confirm that the toxin (streptozotocin (SZ)) eliminated the pre-existing β cells.

These cultures revealed the presence of cells that contained both glucagon and insulin, and cells containing only insulin, suggesting that the glucagon cells converted to NBCs. Matrices containing NBCs were obtained, fixed with 4% paraformaldehyde (Electron Microscopy Sciences; Fort Washington, PA), washed with a solution of O.lM Tris buffer (TS) containing 0.25 Triton ((T), Fisher Biotech, Sigma-Adrich) and 0.1% goat serum (GS, Gibco-Invitrogen Corporation, also available from Sigma- Aldrich and HyClone (Logan, Utah) ("TSTGS")), incubated with antibodies to glucagon and insulin respectively, washed with a solution of TSTGS, incubated with fluorescent-tag secondary antibodies and then subjected to confocal photomicrography.

Source of antibodies and purified peptides

For primary antiserum, guinea pig antibodies to bovine insulin and rat C- peptide were purchased from Linco Research (Eureka, Mo., USA). Rabbit antiserum to human glucagon was purchased from Calbiochem (San Diego, Calif., USA). Rabbit antisera to human PP and to somatostatin were supplied by Peninsula Labs (Belmont, Calif., USA). Mab antibody to human glucagon and insulin were purchased from Sigma (St. Louis, Mo., USA). Rabbit antisera to Nestin was generously provided by Dr. M. Vallejo (UA, Madrid, Spain). Rabbit antisera to GLUT-2 was purchased from Chemicon (Temecula, Calif., USA). Affinity purified antiserum to the N-terminal domain of Pdx-1 was a generous gift from CN.E. Wright (Nanderbilt University, Nashville, Term., USA).

Antibodies were used at the following dilutions: guinea pig anti-bovine insulin - 1:400; rabbit antisera to GLUT-2 - 1:1000; rabbit anti-human glucagon - 1 : 12,000; rabbit anti-human somatostatin - 1 :8000 for control sections and 1 :20,000 for sections of SZ-treated pancreases; rabbit anti-human pancreatic polypeptide - 1 : 100,000; rabbit anti-Nestin - 1 :5000; rat anti-somatostatin - 1 :2000; Mab to glucagon - 1 :6000 and Mab to insulin - 1 :2000. Rabbit anti-mouse Pdx-1 was used at a 1 : 8000 dilution. This antibody is highly specific for the nuclear form of Pdx-1. For secondary antibodies, Alexa Fluor 488 anti-mouse, anti-rat and anti-rabbit

IgG, Alexa Fluor 594 anti-guinea pig, anti-rabbit and anti-mouse IgG were purchased from Molecular Probes (Eugene, Ore., USA).

All secondary IgGs were used at 1:200 dilution for 1 hour. Cells labeled with fluorescent probes were covered with 2 to 3 drops of Prolong Antifade solution (Molecular Probes) after completion of the staining procedure and were dried at room temperature before examination. Immunolabelling of cells using peroxidase techniques

In brief, the cells were incubated sequentially in an empirically derived optimal dilution of control serum or primary antibody raised in species "X" containing 1% goat serum in Tris-saline solution (TS; 0.9% NaCl in 0.1 mol/1 Tris, pH 7.4) containing 0.5% Triton and 1% Goat Serum (TSTGS) for 18 hours; followed by a 1:200 dilution of anti-(species x) biotinylated IgG solution in TSTGS for 2 hours. After these incubations, the discs containing the NBC's were covered with an Antifade solution (Vectashield, Vector Laboratories, Burlingame, CA). For double labelling, cells were incubated with a cocktail of rabbit antibody to glucagon and mouse antibody to insulin and processed as provided above. In this case, the secondary antibodies (anti-rabbit and antimouse) used were coupled to fluorophores of different wavelengths. Con focal microscopy

An LSM 510 laser scanning confocal microscope (Zeiss) fitted with an Axiovert 100 M microscope (Zeiss) was used with a 63X 1.4NA pan Apochromat objective (Zeiss) to examine the stained cells. Excitation on LSM 510 unit was with a 15 mW argon ion laser running at 75% power emitting at 488 nm, a 1 ,0- helium/neon laser emitting at 543 nm, and a 5.0 mW helium/neon laser emitting at 633 nm. Emissions were collected using a 505 to 530-nm band pass filter to collect Alexa green emissions and a 560 to 615-nm band pass filter to collect Alexa red emissions. Typically, 1 um vertical steps were used with a vertical optical resolution of less than 1.0 um. Results

Figure 2A is a photomicrograph of control islets maintained for 2 days in vitro (2 DIV) stained for insulin (red) and glucagon (green). Figure 2B is a photomicrograph of an islet treated with 10 mM SZ at 1 day in vitro (1 DIV) and stained 6 hours later for glucagon (green) and insulin (red). The islets in Figure 2B only contained glucagon producing (GLU+) cells. A diffuse red staining, representing debris from dead β cells, was visible in some but not all islets. Figures 2C and 2D are photomicrographs of cultures fixed 24 hours after SZ treatment (2 DIV), both of which illustrate the presence of insulin containing new beta cells (NBC's) (red) and few Glu+ cells (green). Figure 3 A is a photomicrograph of a control islet at 3 days in vitro (3 DIN) stained for IN (red) and Pdx-1 (green) (Bar:40um). Figure 3B, a photomicrograph of an islet treated with 10 mM SZ and maintained with media containing 11 mM glucose, reveals the presence of many NBC's (red) which lack Pdx-1 (green) (Bar: 20um). The large number of NBC's present in Fig 3B were most likely generated by precursors from several islets; it was common to see groups of 3-4 islets in close proximity. As can be seen by the photomicrograph set forth in Fig. 3C, NBC's express IN (red) and GLUT-2 (green) (Bar: 15um). The periphery of cells is yellow because the transporter is localized to the cell membrane where both colors overlap. As can be seen in Figure 3C, the NBC's that differentiated in vitro had a round shape and were 50 % smaller than β cells of controls (Area β cell control: 98um±0.4, N=50; area NBC: 48um±l .3; N=58). The NBC's that differentiated in vitro expressed GLUT-2 (Fig. 3C) but did not contain Pdx-l(Fig. 3B), even after 4 days in culture. The above experiment demonstrates that an SZ concentration of lOmM effectively destroyed the pre-existing β cell population a few hours after treatment and the treated islets resembled those shown in Fig 2. At 2 DIV islets contained NBC's and a variable number of GLU+ cells. Similar results were obtained when islets were exposed to SZ in vitro or when SZ was injected into mice and the islets transferred to culture. These observations indicate that β cell neogenesis occurred independently of factors present in vivo.

EXAMPLE 3 To determine the effect of high glucose levels on the appearance of new β cells (NBC's), islets were isolated, transferred to culture and fed with culture media containing 5.6 mM glucose. The following day (1 day in vitro = 1 DIN) the cultures were exposed to 10 mM SZ, re-fed with media containing either 5.6 mM, 11 mM, 17 mM or 28 mM glucose, and examined at 2, 3 and 5 days in vitro (2 DIN, 3 DIV and 5 DIV). In all cases, islet shrinkage and dead cells appeared 5-7 hours after SZ treatment. As indicated in Table 1 below and as set forth in Figure 4, the time of appearance and number of ΝBC's present in cultures was inversely correlated with the glucose levels. Table 1- Effect of hyperglycemia onΝBC differentiation in vitro mM glucose 2DIV 3DIV

5.6 + + 11 + +

17 - +

28 - -/+

(+ = ΝBCs present; - = no ΝBCs present; -/+ = few ΝBCs present.)

Figure 4 are photomicrographs of the cells following the procedures provided above. As set forth in Figure 4, severe hyperglycemia reduced the number of ΝBCs that differentiate. Cultures illustrated by the photomicrographs in the upper row of Figure 4 were maintained in media containing 28 mM glucose. The left photomicrograph is the control, the right photomicrograph is the SZ-treated islet. The arrow in the photomicrograph of the SZ-treated islet indicates the few ΝBCs that differentiated in those cultures. The photomicrographs in the lower row of Figure 4 illustrate cultures maintained in media containing 5.6 mM glucose. Again, the left photomicrograph is the control; the right photomicrograph is the SZ-treated islet. Insulin-producing cells showed up as red and glucagon producing cells showed up as green. All cultures were examined at 3 DIV. A large number of NBC's appeared in cultures maintained in 5.6 or 11 mM glucose at 2 DIV while cultures maintained in media containing 17 mM glucose did not contain a significant number of NBC's until one day later, at 3 DIV. Few NBC's first appeared at 3 DIV in cultures maintained at 28 mM glucose (Fig 4 and Table 1) and the number of cells did not increase with time (at 5 DIV) or when the cultures were refed with media containing 5.6 mM glucose. These observations indicate that hyperglycemia delayed the appearance of NBC's. In addition, the number of NBC's decreased significantly at extremely high glucose levels. It should be noted that a glucose concentration of 11 mM was not deleterious to β cell regeneration. This is in agreement with reports by others indicating that this concentration was optimal for rodent islets.

In previous in vivo studies examining SZ-treated mice with circulating blood glucose levels in the range of 500 mg/dl (which is equivalent to 28 mM) severely hyperglycemic mice had poor β cell neogenesis. (Guz et al., "Regeneration of Pancreatic β Cells From Intra-Islet Precursor Cells in an Experimental Model of

Diabetes", Endo, 142: 4956-4968 (2001)). Similarly, cultures maintained in vitro in media containing 28 mM glucose generated few NBCs. However, while β cell neogenesis in vivo improved following the restoration of normoglycemia by insulin therapy, the replacement in vitro of media containing 28 mM glucose by one containing physiological levels of glucose had no effect on the number of NBCs that differentiate. Therefore, it is presumed factors present in vivo but not in culture participate in reverting the deleterious effect of hyperglycemia. EXAMPLE 4 The ability of NBCs to proliferate in culture was examined. Islets were isolated, treated with SZ and maintained in vitro for 2 days in the presence of lOOuM 5-bromo-2'deoxyuridine (BrdU) (Sigma, St. Louis, MO). Cells were first stained for insulin and then rinsed and incubated with the corresponding IgG linked to an Alexa fluorophore. Cells were rinsed overnight with 0.1M PBS, pH 7.2, fixed with 4% paraformaldehyde for 30 minutes, rinsed and incubated overnight at 37°C with a cocktail containing 50 U/ml of DNase I (Roche Diagnostics, Indianapolis, IN) and 1:40 dilution of monoclonal antibody for BrdU (DAKO Corp., Carpinteria, CA; 1:100 diluted with a 1% solution of goat serum in PBS). Cells were rinsed and incubated with goat antimouse IgG labeled with Alexa fluorophore for 2 hours. Cells were mounted with coverslips using Prolong.

Photomicrographs were taken following the procedures provided in Example 2, with red indicating insulin producing cells and green indicating BrdU. The results are set forth in the photomicrographs of Fig. 5.

As shown by the photomicrograph of Fig. 5 A, β cells of untreated islets did not divide. In contrast, the newly differentiated IN+ cells proliferated in vitro and were labeled with BrdU (Fig 5B).

EXAMPLE 5

A recent report indicated that insulin staining of embryonic stem cells is due to insulin taken up from the media rather than from the initiation of insulin gene expression by those cells (Rajagopal et al., "Insulin Staining of ES Cell Progeny From Insulin Uptake", Science, 299:363 (2003)). To determine whether NBC's initiate insulin synthesis, SZ-treated islets were stained and maintained in culture for 2 days with guinea pig antisera to (rat) C-peptide (Linco Re). (C-peptide is known to be a very sensitive marker for beta cell function.) Photomicrographs of the cells were prepared following the procedures set forth in Example 2.

As illustrated by Figure 6, cultures of NBC's processed for C-peptide staining revealed the presence of stained cells, which indicates that proinsulin is produced by the NBC's, and that the prohormone is processed to generate insulin and C-peptide.

EXAMPLE 6 This example demonstrates the ability of NBCs to secrete insulin in response to glucose stimulation.

The NBCs are isolated and obtained as provided in Examples 1 and 2. Insulin secretion is measured by radioimmune assay (RIA).

The cultured NBCs are washed in basal medium RPMI (5 times for 20 minutes per wash) before the experiment. NBCs are cultured in the medium provided in Example 1, with the difference that the medium contained 5.6mM glucose instead of 11 mM glucose, for 3-1/2 hours. The NBCs are then transferred to 1 ml of the basal medium containing an additional 17mM glucose and cultured for another 3-1/2 hours. RIA is conducted to confirm and determine the amount of insulin secretion.

EXAMPLE 7 To test the effectiveness of these NBCs in the regulation of blood glucose levels, immunodeficient mice (obtained from commercial sources) are injected with streptozotocin (60 mg/Kg) to produce diabetes. Diabetic mice are anesthetized with pentobarbital ( 50 mg/Kg). The fur is shaved, disinfected with 80% ethanol and a small incision made in the abdominal wall. The kidneys are exposed and 10-50 cell clusters of the NBCs are deposited under the kidney capsule by means of a braking pipette. The kidney is then returned to the abdominal cavity and the abdominal incision closed with wound clips. Mice are allowed to awaken and are given food and water. Animals are allowed to survive for one week to a month and blood glucose levels are tested with commercial kits. Blood glucose levels are monitored, which levels are expected to be normal compared to a diabetic mouse. At different times after surgery the animals are killed and the kidney containing the transplant examined using histological techniques.

EXAMPLE 8 Rhesus monkeys are transplanted with allograft NBCs in their testes to examine the survival of these transplants. The recipients are made diabetic by means of a near total pancreatectomy, followed by an intravenous injection of 35 mg streptozotocin kg body weight two weeks later. This procedure will induce severe diabetes mellitus. Malabsorption is prevented by the oral administration of VACUOUS®, one tablet given twice daily before each meal.

Islets are isolated from female Rhesus monkeys. First, the pancreases is removed, pooled and chopped finely into smaller fragments. The fragments are then digested by collagenase in a water bath at 37° C. Islets are separated from exocrine tissues and other cellular debris on at least two Ficoll gradients, prepared in tandem. The islets are washed three times by centrifugation in ice-cold Hanks' buffer and transferred in groups of 150 to biologic grade Petri dishes. Each dish contains 6 mL of culture medium CMRL-1066 supplemented with 5% fetal calf serum, glucose at a concentration of 250 mg/dL, penicillin (100 U/mL), and streptomycin (100 U/mL). Incubation of islets is carried out at 35° C. in 5% C0₂ and air for 4 to 6 days. The islets are transferred to fresh medium at 48 hour intervals. Viability and counting of the islets is facilitated by uptake of the dye dithizone. Islets are exposed to streptozotocin and/or a cocktail of cytokines to eliminate pre-existing β cells and allow the differentiation of proliferating NBCs. The NBCs population is maintained in culture conditions containing growth factors to sustain their proliferation. Each monkey receives a number of NBCs equivalent, with

regards to insulin production and secretion, to an average of about 10^ islets/kg body weight by injection into both testes. In some of the test animals the testes are elevated into the abdominal cavity, whereas in other recipients the grafted organs are anchored into the inguinal canal. Cyclosporine is administered, in varying doses to the grafted animals over a 30 day period, whereas the animals with the organs anchored in the inguinal canal are given 7 injections of Cyclosporine (20 mg/kg) on days -4 to +3.

Oral sustacal tolerance tests are done on day 30, and then at intervals in the normoglycemic animals, as follows.

The monkeys are housed individually in cages and given standard monkey chow and fruit twice daily. In addition, a pancreatic enzyme is mixed with the food since the monkeys are pancreatectomized to make them diabetic before transplantation.

Animals are fasted for 12 hours. They are then anesthetized and prepared for the test meal. Sustacal is used as the test agent. Sustacal consists of a physiologic mixture of carbohydrates, proteins and fat which closely mimics a standard meal and which is a powerful stimulus for the release of insulin.

Sustacal is injected directly into the stomach of a sleeping animal through a nasogastric tube. Blood samples are then obtained at times 0, 15, 30, 60, 90, 120 and

180 minutes. The samples are centrifuged and the serum is stored at -20° C. until measurements for insulin or C-peptide can be carried out. Following transplantation assays are performed to monitor glucose responses, which are expected to be restored to normal levels. C-peptide responses to oral sustacal tolerance tests following transplantation are also expected to be restored to normal. The above Examples indicate that, following elimination of the pre-existing β cells, precursor cells present in the islets initiated insulin expression. Although the glucagon cells could be the major source of NBCs, other precursor cells present in the islets, such as newly discovered dormant Ngn3+ cells present in adult islets (Gu et al. "Direct Evidence for the Pancreatic Lineage: Ngn3+ Cells are Islet Progenitors and are Distinct From Duct Progenitors", Development, 129: 2447-2458 (2002)) could convert into NBCs.

Claims

WHAT IS CLAIMED IS:

1. An in vitro method for converting non-insulin producing cells of the islets of Langerhans into insulin-producing cells, said method comprising: a) isolating islets of Langerhans from pancreatic tissue; b) contacting the islets of Langerhans with a toxic substance which selectively kills β cells present therein; and c) sustaining the remaining living cells of the islets of Langerhans under conditions which convert non-insulin producing cells therein to insulin-producing cells.

2. The method of Claim 1 wherein the toxic substance is selected from the group consisting of streptozotocin, alloxan, interleukin-1 beta, tumor necrosis alpha, interferon-γ, and combinations thereof.

3. The method of Claim 1 further comprising sustaining the proliferation of the remaining cells of the islets capable of producing insulin.

4. The method of Claim 3 wherein the proliferation of the remaining cells of the islets is sustained by embedding them in a matrix.

5. The method of Claim 4 wherein the matrix is selected from the group consisting of collagen and basement membrane derivatives extracted from tumor cells.

6. The method of Claim 3 wherein the proliferation of the remaining cells of the islets is sustained by adding a growth factor.

7. The method of Claim 6 wherein the growth factor is selected from the group consisting of hepatocyte growth factor/scatter factor, activin, exendin 4, betacellulin, nicotinamide, culture media supplement B27, culture media supplement ITS, epidermal growth factor, basic fibroblast growth factor, and glucose.

8. The method of Claim 3 wherein the proliferation of the remaining cells of the islets is amplified by transfecting non-insulin producing precursor cells with a construct combining an oncogene and a promoter of a gene expressed by the precursor cells but not by the insulin-producing cells.

9. The method of Claim 8 wherein the proliferation of the remaining cells of the islets is amplified by transfecting a cells with a construct combining an oncogene and a promoter of the glucagon gene.

10. A method for in vitro culturing insulin-producing cells which comprises: a) isolating islets of Langerhans from pancreatic tissue; b) contacting the islets of Langerhans with a toxic substance which selectively kills β cells present therein; c) sustaining the living cells of the islets of Langerhans under conditions which convert non-insulin producing cells therein to insulin-producing cells; and d) maintaining the insulin-producing cells in a state of normoglycemia at a temperature equivalent to the body temperature of the mammal from which the islets of Langerhans were obtained.

11. The method of Claim 10 wherein the toxic substance is selected from the group consisting of streptozotocin, alloxan, interleukin-1 beta, tumor necrosis alpha, interferon-γ, and combinations thereof.

12. The method of Claim 10 further comprising sustaining the proliferation of the remaining cells of the islets capable of producing insulin.

13. The method of Claim 12 wherein the proliferation of the remaining cells of the islets is sustained by embedding them in a matrix.

14. The method of Claim 13 wherein the matrix is selected from the group consisting of collagen and basement membrane derivatives extracted from tumor cells.

15. The method of Claim 12 wherein the proliferation of the remaining cells of the islets is sustained by adding a growth factor.

16. The method of Claim 15 wherein the growth factor is selected from the group consisting of hepatocyte growth factor/scatter factor, activin, exendin 4, betacellulin, nicotinamide, culture media supplement B27, culture media supplement ITS, epidermal growth factor, basic fibroblast growth factor, and glucose.

17. The method of Claim 12 wherein the proliferation of the remaining cells of the islets is amplified by transfecting non-insulin producing precursor cells with a construct combining an oncogene and a promoter of a gene expressed by the precursor cells but not by the insulin-producing cells.

18. The method of Claim 17 wherein the proliferation of the remaining cells of the islets is amplified by transfecting cells with a construct combining an oncogene and a promoter of the glucagon gene.

19. The method of Claim 10 wherein the insulin-producing cells are cultured for a period of time ranging from about 40 hours to about 96 hours.

20. The method of Claim 10 wherein the islets are obtained from a mouse and about lOmM to about 12mM of glucose is added to maintain the insulin- producing cells in a state of normoglycemia.

21. The method of Claim 10 wherein the islets are obtained from a human and about 4.6mM to about 6.6mM of glucose is added to maintain the insulin- producing cells in a state of normoglycemia.

22. An isolated insulin-producing cell made by: a) isolating islets of Langerhans from pancreatic tissue; b) contacting the islets of Langerhans with a toxic substance which selectively kills β cells present therein; c) sustaining the living cells of the islets of Langerhans under conditions which convert non-insulin producing cells therein to insulin-producing cells; and d) isolating the insulin-producing cells.

23. A pharmaceutical composition for treating diabetes comprising the insulin-producing cells of Claim 22 in a pharmaceutically acceptable carrier.

24. A method for producing insulin comprising: a) isolating islets of Langerhans from pancreatic tissue; b) contacting the islets of Langerhans with a toxic substance which selectively kills β cells present therein; c) sustaining the remaining living cells of the islets of Langerhans under conditions which convert non-insulin producing cells therein to insulin-producing cells; and d) collecting insulin from the insulin-producing cells.

25. The method of Claim 24 wherein the toxic substance is selected from the group consisting of streptozotocin, alloxan, interleukin-1 beta, tumor necrosis alpha, interferon-γ, and combinations thereof.

26. The method of Claim 24 further comprising sustaining the proliferation of the remaining cells of the islets capable of producing insulin.

27. The method of Claim 26 wherein the proliferation of the remaining cells of the islets is sustained by embedding them in a matrix.

28. The method of Claim 27 wherein the matrix is selected from the group consisting of collagen and basement membrane derivatives extracted from tumor cells.

29. The method of Claim 26 wherein the proliferation of the remaining cells of the islets is sustained by adding a growth factor.

30. The method of Claim 29 wherein the growth factor is selected from the group consisting of hepatocyte growth factor/scatter factor, activin, exendin 4, betacellulm, nicotinamide, culture media supplement B27, culture media supplement ITS, epidermal growth factor, basic fibroblast growth factor, and glucose.

31. The method of Claim 26 wherein the proliferation of the remaining cells of the islets is amplified by transfecting non-insulin producing precursor cells with a construct combining an oncogene and a promoter of a gene expressed by the precursor cells but not by the insulin-producing cells.

32. The method of Claim 31 wherein the proliferation of the remaining cells of the islets is amplified by transfecting a cells with a construct combining an oncogene and the promoter of the glucagon gene.

33. A method for treating diabetes in a mammal comprising: a) isolating islets of Langerhans from pancreatic tissue; b) contacting the islets of Langerhans with a toxic substance which selectively kills β cells present therein; c) sustaining the remaining living cells of the islets of Langerhans under conditions which convert non-insulin producing cells therein to insulin-producing cells; d) isolating the insulin-producing cells; and, e) implanting the insulin-producing cells into a subject suffering from diabetes.

- 34. The method of Claim 33 further comprising maintaining the insulin- producing cells in a state of normoglycemia at a temperature equivalent to the body temperature of the mammal from which the islets of Langerhans were obtained to expand the population of insulin-producing cells.

35. The method of Claim 33 wherein Sertoli cells are implanted with the insulin-producing cells.

36. The method of Claim 33 wherein the insulin-producing cells comprise an allograft.

37. The method of Claim 33 wherein the insulin-producing cells comprise a xenograft.

38. The method of Claim 33 wherein the insulin-producing cells comprise an autograft.

9. The method of Claim 33 wherein the mammal is a human.