Diabetes mellitus is particularly characterized by hyperglycemia, although other related clinical manifestations include vascular complications and altered metabolism of lipids, proteins and carbohydrates. For example, the altered metabolism of lipids can lead to increased production of ketones, potentially leading to ketonemia and acidosis. Vascular changes which occur include the thickening of the capillary basement membrane, leading to narrowing of the lumen of blood vessels and reduced blood circulation. Such reduced circulation can, in turn, cause retinopathy, neuropathy and gangrene of the extremities. Thus, diabetes mellitus has consequences which extend far beyond hyperglycemia.

Diabetes mellitus can be divided into two clinical syndromes, Type I and Type II diabetes mellitus. Type I, or insulin-dependent diabetes mellitus (IDDM), is a chronic autoimmune disease characterized by the extensive loss of cells in the pancreatic Islets of Langerhans (hereinafter referred to as "islets"), which produce insulin. As these cells are progressively destroyed, the amount of secreted insulin decreases, eventually leading to hyperglycemia (abnormally high level of glucose in the blood) when the amount secreted drops below the level required for euglycemia (normal blood glucose level). Although the exact trigger for this immune response is not known, patients with IDDM have high levels of antibodies against pancreatic cells. However, not all patients with high levels of these antibodies develop IDDM.

One possible explanation for this apparent discrepancy is the capacity of the pancreas to expand cell mass through proliferation of these cells. For example, at the onset of IDDM, the pancreas has been shown to be capable of neoformation of islets [W. Gepts and P.M.

Lecompte, Am. J. Med., 70:105-115, 1981]. cell proliferation and regeneration has also been demonstrated in animal models of diabetes with reduced cell mass, and in transgenic mice with cell damage [S. Bonner-Weir et al., Diabetes, 42:1715-1720, 1993; D. Gu and N.

Sarvetnick, Development, 118:33-46, 1993]. Unfortunately, the lack of significant

proliferative ability of mature p cells is a major impediment to the treatment of diabetes, particularly due to the relatively late diagnosis of the disease.

Without such proliferation of P cells, treatment of IDDM will not be successful, even if the underlying pathology is arrested. For example, immunomodulation alone to date has been a failure for the treatment of IDDM (Marks et al., J. Clin. Endoc. Metab., 72:3-9, 1991).

One reason for this failure is that a certain mass of cells is required to produce sufficient insulin. However, by the time clinical symptoms of diabetes have been detected, the patient generally has lost this required mass of cells. Therefore, immunomodulation alone is not be effective, as it does not increase the number of cells.

Patients with IDDM are usually diagnosed when clinical symptoms are manifested, such as hyperglycemia. However, by the time such symptoms are apparent, the disease has been progressing over an extensive period of time and extensive loss of pancreatic P cells has occurred. Therefore, treatments such as immunoregulators, which seek to inhibit the autoimmune response, can only arrest further progress of IDDM and cannot reverse the extensive pancreatic cell loss. Unless the lost P cells are replaced, the patient will continue to experience clinical symptoms of IDDM, even as the course of the disease itself is halted.

However, simply replacing the P cells will also prove ineffective unless the underlying immunopathology of IDDM is arrested. Thus, a successful treatment for IDDM must both halt the further progression of the disease and induce the replacement of lost pancreatic P cells.

There is therefore an unmet medical need for a treatment for diabetes which both halts the further progress of the disease and which ameliorates or reverses clinical symptoms by inducing the proliferation of pancreatic P cells.

SUMMARY OF THE INVENTION According to the teachings of the present invention, there is provided a composition for the treatment of diabetes, including: (a) an immunoregulator; and (b) a P cell proliferative agent. Preferably, the immunoregulator is selected from the group consisting of Quinoline-3- carboxamide, insulin, vitamin D, nicotinamide, corticosteroid and pharmaceutically acceptable salts and equivalents thereof and a vector for gene transfer. More preferably, the <BR> <BR> immunoregulator i s is Quinoline-3-carboxamide and pharmaceutically acceptable salts and

equivalents thereof. Also more preferably, the immunoregulator is the vector for gene transfer. Most preferably, the vector is an adenoviral vector.

Preferably, the P cell proliferative agent is selected from the group consisting of reg protein, growth hormone, and prolactin. More preferably, the P cell proliferative agent is reg protein. Most preferably, the reg protein is administered as a protein. Alternatively and most preferably, the reg protein is administered as a gene coding for the reg protein.

Preferably, the composition further includes a pharmaceutically acceptable carrier.

According to preferred embodiments of the present invention, there is provided a method for treating diabetes in a subject, comprising the step of administering an effective amount of a composition to the subject, the composition including an immunoregulator and a P cell proliferative agent in a pharmaceutically acceptable carrier. Preferably, the immunoregulator is selected from the group consisting of Quinoline-3-carboxamide, insulin, vitamin D, nicotinamide, corticosteroid and pharmaceutically acceptable salts and equivalents thereof and a vector for gene transfer. Preferably, the P cell proliferative agent is selected from the group consisting of reg protein, growth hormone, and prolactin.

According to other embodiments of the present invention, there is provided a composition for the treatment of diabetes, comprising: (a) quinoline-3-carboxamide and pharmaceutically acceptable salts and bioequivalents thereof; and (b) reg protein; the quinoline-3-carboxamide and the reg protein each being in an amount such that a combination of the quinoline-3-carboxamide and the reg protein has an effect greater than an additive effect of the quinoline-3-carboxamide and the reg protein separately.

According to still other embodiments of the present invention, there is provided a method for treating diabetes in a subject, comprising the step of administering an effective amount of a composition to the subject, the composition including quinoline-3-carboxamide and pharmaceutically acceptable salts and bioequivalents thereof and reg protein in a pharmaceutically acceptable carrier, the quinoline-3-carboxamide and the reg protein each being in an amount such that a combination of the quinoline-3-carboxamide and the reg protein has an effect greater than an additive effect of the quinoline-3-carboxamide and the reg protein separately. Preferably, the subject is glucose intolerant substantially before the composition is administered to the subject. Also preferably, the subject exhibits at least one clinical symptom of diabetes substantially before the composition is administered to the

subject. More preferably, a progression of diabetes in the subject is substantially prevented by administration of the composition.

According to still other embodiments of the present invention, there is provided a method for preserving P-cell mass in a subject having an autoimmune disease, the autoimmune disease causing destruction of the P-cell mass, comprising the step of administering an effective amount of a composition to the subject, the composition including quinoline-3-carboxamide and pharmaceutically acceptable salts and bioequivalents thereof and reg protein in a pharmaceutically acceptable carrier, the quinoline-3-carboxamide and the reg protein each being in an amount such that a combination of the quinoline-3-carboxamide and the reg protein has an effect greater than an additive effect of the quinoline-3- carboxamide and the reg protein separately for the preservation of the P-cell mass.

Preferably, the subject is glucose tolerant substantially before the composition is administered to the subject.

According to yet another embodiment of the present invention, there is provided a composition for gene therapy of diabetes, comprising: (a) a reg gene; and (b) a vector for gene transfer. Preferably, the vector is an adenoviral vector.

According to still another embodiment of the present invention, there is provided a method for gene therapy of diabetes in a subject, the method comprising the steps of: (a) placing a reg gene in a vector for gene transfer, to form a reg vector; and (b) administering the reg vector to the subject.

Hereinafter, the term "subject" refers to the human or lower animal to whom the composition of the present invention was administered. Hereinafter, the terms "diabetes", "diabetes mellitus", "Type I diabetes" and "IDDM" all refer to insulin-dependent diabetes mellitus. Hereinafter, "Linomide" is defined as quinoline-3-carboxamide, and pharmaceutically acceptable salts and bioequivalents thereof (Pharmacia Upjohn, Lund, Sweden).

The term "reg gene" includes any polynucleotide sequence, such as reg cDNA, in which at least a portion codes for the reg protein. The term "reg protein" includes those amino acid sequences, and the term "reg gene" includes those polynucleotide sequences, respectively, given as Swiss Prot Accession No. P05451, and EMBL Accession No. M18963, sequences disclosed or identified in PCT Application No. WO 94/12203, European Application No. EP 383453, European Application No. EP 303233 and European Application

No. EP 286114, as well as in Terazono, K., et al., J. Biol. Chem., 263:2111-2114, 1988; Watanabe, T., et al., J. Biol. Chem., 265:7432-7439, 1990; Moriizumi, S., et al., Biochim.

Biophys. Acta, 1217:199-202, 1994, and pharmaceutically active mutants, variants or portions thereof, and combinations of these sequences and pharmaceutically active mutants, variants or portions thereof.

Hereinafter, the term "immunoregulator" is defined as a substance which substantially prevents, reduces or inhibits the autoimmune reaction seen in insulin-dependent diabetes mellitus. A preferred example of such an immunoregulator is Quinoline-3-carboxamide, although other examples include insulin, vitamin D, nicotinamide, corticosteroids and pharmaceutically acceptable salts or equivalents thereof. The term "P cell proliferative agent" is defined as a substance which substantially induces the proliferation of P cells in the pancreas. A preferred example of such a cell proliferative agent is the reg protein, although other examples include growth hormone and prolactin, and pharmaceutically acceptable mutants, variants and portions thereof. Hereinafter, the term "NOD mouse" is defined as a non-obese diabetic mouse. These mice are commercially available (Jackson Laboratories, Bar Harbor, ME, USA). Hereinafter, the term "treat" includes substantially inhibiting, slowing or reversing the progression of a disease, substantially ameliorating clinical symptoms of a disease or substantially preventing the appearance of clinical symptoms of a disease.

BRIEF DESCRIPTION OF THE DRAWINGS The invention is herein described, by way of example only, with reference to the accompanying drawings, wherein: FIGS. 1A and 1B show the ability of the composition of the present invention to substantially reverse the course of diabetes in NOD mice; FIGS. 2A-2F demonstrate P cell proliferation in NOD mice treated with the composition of the present invention; FIGS. 3A and 3B show Northern blot and Western blot analyses, respectively, of pancreatic cell lines infected with Ad reg and Ad luc; FIG. 4 shows localization of luciferase protein to various organs after infection with Ad luc; and FIG. 5 shows the ability of Ad reg to prevent the progression of diabetes in NOD mice.

BRIEF DESCRIPTION OF THE INVENTION The present invention is a treatment for diabetes mellitus, and in particular for insulin- dependent diabetes mellitus (IDDM), which combines an immunoregulator and a cell proliferative agent. As demonstrated by the experimental data below, the combination of the present invention shows unexpectedly efficacy for treatment of diabetes in an animal model of the disease, NOD mice. Without wishing to be bound by a particular hypothesis, the combination may be more effective than either substance alone because the underlying pathology of the disease is both arrested by the immunoregulator and the lost P cell mass is augmented through the effect of the P cell proliferative agent.

Two separate embodiments of the combination of the present invention were tested. In one embodiment, the immunoregulator was quinoline-3-carboxamide, administered orally, and the P cell proliferative agent was reg protein, given by daily injection six days per week.

One disadvantage of reg protein is that proteins are difficult to administer to the subject by a route other than parenteral administration. However, parenteral administration is not a preferred route for administration in human patients, since it is more difficult for the layperson to perform and may result in reduced patient compliance.

In an attempt to overcome the need for daily injections of the reg protein, a second embodiment of the present invention was tested. In this embodiment, the reg protein was delivered locally to the P cells by an adenoviral mediated vector for expression of the gene coding for the reg protein. Quinoline-3-carboxamide was not needed as the immunoregulator, since the adenoviral mediated vector itself appeared to act as the immunoregulator while enabling the P cells and surrounding tissue to produce the reg protein. Thus, adenoviral mediated gene expression appears to enable the combination of the present invention to be delivered locally, substantially without the addition of other exogeneous substances.

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention is a treatment for diabetes mellitus, and in particular for insulin- dependent diabetes mellitus (IDDM), which combines an immunoregulator and a cell proliferative agent. As demonstrated below, such a combination is particularly efficacious not only to substantially slow or halt the progression of IDDM, but even to substantially reduce or reverse the clinical manifestations of the disease.

Although the present invention is not limited to this single combination, the immunoregulator tested was Quinoline-3-carboxamide, while the cell proliferative agent was reg protein. Such a combination was neither taught nor suggested in the prior art, although each of these substances had been examined separately in prior art references.

Quinoline-3-carboxamide, the immunoregulator, was shown to prevent diabetes from developing in NOD mice if administered sufficiently early, before clinical symptoms have developed [D.J. Gross et al., Diabetologia, 37:1195-1201, 1994]. However, once clinical symptoms have appeared in these mice, Quinoline-3-carboxamide was shown to only partially ameliorate the disease. Unfortunately, as noted above, human patients often are diagnosed only at such a relatively late stage of the disease, so that Quinoline-3-carboxamide alone would not be effective as a treatment for patients with IDDM. In order to overcome the depletion of P cell mass seen in NOD mice with clinical symptoms of diabetes, prior art studies indicated that transplantation of islet cells followed by immunoregulation with Quinoline-3-carboxamide were required [S. Slavin et al., Cell Transplantation, 5:627-630, 1996]. However, such transplantations would be difficult to perform in human patients for a number of reasons. First, sufficient donor material must be located, which is generally a problem for human transplantation. Second, extensive drug treatment is required to prevent rejection of the transplanted tissue. Finally, transplantation itself is a complicated, difficult medical procedure. Thus, other treatments for patients with IDDM are required.

Reg protein, the P cell proliferative agent used in the composition of the present invention, is a C-type lectin originally cloned from a cDNA library prepared from regenerating islets, hence its name [K. Terazono et al., J. Biol. Chem., 263:2111-2114, 1988].

Reg expression has been demonstrated at the protein and mRNA levels in replicating islets, as well as in the Syrian golden hamster and rat islet regeneration models [K. Terazono et al., Diabetologia, 33:250-252, 1990; R. Rafaeloff et al., Diabetologia, 38:906-913, 1995; and M.E. Zenilman et al., Surgery, 119:576-584, 1996]. However, other studies have failed to demonstrate reg expression in pancreatic islets or cell lines [C. Miyaura et al., Mol.

Endocrinol., 5:226-234, 1991; and F.E. Smith et al., Diabetologia, 37:994-999, 1994]. Thus, the exact role of reg is unclear from the prior art teachings.

In one study, reg protein was found to ameliorate diabetes in rats resulting from removal of the pancreas [T. Watanabe et al., PNAS, 91:3589-3592,1994]. However, against the teachings of this prior art reference, reg protein alone was not found to be effective in

inhibiting or reversing the progression of diabetes, as shown in Examples 1 and 2 below.

Thus, the prior art only taught the efficacy of reg protein alone, while the results shown below not only teach against the prior art, but also demonstrate the efficacy of the combination of Quinoline-3-carboxamide and reg protein. Such a combination was neither taught nor suggested in the prior art.

The principles of a treatment for diabetes according to the present invention may be better understood with reference to the Examples, drawings and the accompanying description. Examples of the methods described or mentioned herein may be found in standard reference books such as Current Protocols in Human Genetics, N.G. Dracopoli et al., eds., John Wiley & Sons, Inc., New York, New York, 1997 or Current Protocols in Molecular Biology, Ausubel, F. M. et al., eds., John Wiley & Sons, Inc., New York, New York, 1997, incorporated herein by reference merely as examples.

Example 1 Occurrence of Diabetes in NOD Mice Female NOD (non-obese diabetic) mice are characterized by displaying IDDM with a course which is similar to that found in humans, although the disease is more pronounced in female than male NOD mice. Hereinafter, unless otherwise stated, the term "NOD mouse" refers to a female NOD mouse. NOD mice have a progressive destruction of P cells which is caused by a chronic autoimmune disease. Thus, NOD mice begin life with euglycemia, or normal blood glucose levels. By about 15 to 16 weeks of age, however, NOD mice start becoming hyperglycemic, indicating the destruction of the majority of their pancreatic P cells and the corresponding inability of the pancreas to produce sufficient insulin. Thus, both the cause and the progression of the disease are similar to human IDDM patients.

In particular, the course of this disease in both NOD mice and in humans includes the progressive destruction of cells which is caused by a chronic autoimmune disease.

Therefore, both NOD mice and humans with IDDM begin life with euglycemia, or normal blood glucose levels. However, eventually the majority of cells in the pancreas are destroyed, so that clinical symptoms of diabetes begin to appear. The disease is not diagnosed in human patients until at least one clinical symptom of diabetes, and often the entire constellation of symptoms, has started to appear. Thus, glucose intolerant mice and human

patients diagnosed with IDDM are actually at a highly similar or even identical clinical point in the progression of the disease of diabetes.

Because of these similarities, NOD mice are widely considered to be a highly acceptable model for studying IDDM. Indeed, NOD mice are the most widely used animal model by those of ordinary skill in the art and in fact by all workers in this field. In particular, glucose intolerant NOD mice are actually extremely similar to human patients diagnosed with IDDM for the following reasons. First, both NOD mice and humans with IDDM suffer from the progressive destruction of cells. Second, clinical symptoms, and in particular glucose intolerance, is only shown after the majority of these cells have been destroyed in both mice and human patients. Third, glucose intolerant mice are at the same clinical stage of the disease as human patients diagnosed with IDDM, since glucose intolerance is the main clinical symptom which leads to the diagnosis of IDDM in human patients. Thus, the cause of the disease, the course of the disease and the clinical manifestations of the disease are all similar or identical in NOD mice and in human patients.

For the experiments performed in both this Example and Example 2, female NOD mice were used. These mice were either untreated (control), treated with Quinoline-3- carboxamide alone, reg alone or a combination of reg and Quinoline-3-carboxamide. The effect of these various treatments on the progression of diabetes was measured. As shown in Figures 1A and 1B, only the combination of reg and Quinoline-3-carboxamide was able to substantially halt and even reverse the course of diabetes in glucose intolerant mice. The experimental method was as follows.

At 14 weeks of age, the female NOD mice were phenotyped according to glucose tolerance. Glucose tolerance was measured with the intraperitoneal glucose tolerance test (IPGTT). Briefly, blood was drawn from the paraorbital plexus at 0 minutes and 60 minutes after the intraperitoneal injection of glucose (1 g/kg body weight). Normal tolerance was defined as plasma glucose at 0 minutes of less than 144 mg%, or at 60 minutes of less than 160 mg%. Blood glucose levels were determined with a Glucometer Elite apparatus.

Based upon this phenotypic analysis, animals were allocated to the different experimental groups. In particular, animals with more elevated blood glucose levels were assigned to the impaired glucose tolerance group. The mice were fed ad libitum and were supplied with acidified water (pH 2.3).

The glucose tolerant and intolerant mice were further subdivided into control, Quinoline-3 -carboxamide, reg protein and Quinoline-3 -carboxamide/reg protein combination

groups. Mice in the control group received an interperitoneal injection of vehicle daily, six times per week. Mice in the Quinoline-3-carboxamide group received 0.5 mg/ml of Quinoline-3-carboxamide in normal, non-acidified drinking water on a daily basis. Mice in the reg protein group received an interperitoneal injection of 1 mg/kg body weight of reg protein in vehicle daily, six times per week. Mice in the Quinoline-3-carboxamide/reg combination group received both Quinoline-3-carboxamide and reg protein as described above, according to the same amount and rate of dosage. Thus, mice receiving Quinoline-3- carboxamide, alone or in combination with reg protein, received 0.5 mg/ml of Quinoline-3- carboxamide in normal, non-acidified drinking water on a daily basis; while mice receiving reg protein, alone or in combination with Quinoline-3-carboxamide, received an injection of 1 mg/kg body weight of reg protein daily, six times per week.

The reg protein was prepared as follows. The human reg cDNA encompassing the coding sequence was introduced into a Pichia pastoris expression vector for production of recombinant human reg protein [K.A. Barr et al., Pharm. Eng., 12:48-51, 1992]. Yeast supernatant, containing reg protein, was concentrated by precipitation with 60% saturated solution of ammonium sulfate. The precipitate was dissolved and then further concentrated and desalted using an Intersep apparatus (Intersep Filtration Systems, Berkshire, UK) with a 10 kDa cut-off membrane. The resultant reg protein migrated on SDS-PAGE to the same position as the recombinant reg protein in unprocessed yeast medium and after ion-exchange chromatography, showing a double-band at 16.5 kDa. Furthermore, the resultant protein was also detected by a human reg anti-serum on a western blot. Prior to injection, the lyophilized protein was reconstituted at a concentration of 1 mg/ml in 50 mM acetic acid.

The level of urine glucose in the NOD mice was determined on a bi-weekly basis using Labstix (Bayer Diagnostics, Hampshire, England). Weight and fluid intake were also determined on a bi-weekly basis. The onset of diabetes was defined after the appearance of glucosuria on two consecutive determinations. After 10 weeks of treatment, an additional IPGTT was performed and animals were sacrificed the following day. Results are shown in Figures 1A and 1B, as well as in Table 1.

Over the 10 week course of treatment, control animals in both the glucose tolerant (Figure 1A) and glucose intolerant (Figure 1B) developed diabetes at a rate of 60% and 86%, respectively. Thus, high rates of diabetes occur even in NOD mice which were initially glucose tolerant if no intervention is made.

Quinoline-3-carboxamide alone was able to prevent the development of diabetes in glucose tolerant animals, as shown in Figure 1 A. This result is expected, since Quinoline-3- carboxamide inhibits the autoimmune response which causes diabetes in NOD mice. Since these animals were still euglycemic at the onset of treatment, Quinoline-3-carboxamide was effective in preventing progression of the disease. However, Quinoline-3 -carboxamide alone was not able to inhibit or reverse progression of diabetes in glucose intolerant mice, since these mice had presumably already suffered significant P cell loss at the onset of treatment, so inhibition of further progression would not prevent diabetes in these animals.

The reg protein alone conferred only partial protection for glucose tolerant mice, as shown in Figure 1A. Only 20% of glucose tolerant mice developed diabetes when treated with reg protein, while 60% of glucose tolerant mice in the control group developed diabetes.

However, as shown in Figure 1B, the reg protein alone did not inhibit or reverse the course of diabetes in glucose intolerant mice, possibly because the autoimmune disease process was not affected by the reg protein. Thus, although the reg protein promotes P cell proliferation as shown in Example 2, these cells continued to be destroyed by the autoimmune reaction and so the development of diabetes was not affected in these mice.

The combination of Quinoline-3-carboxamide and reg protein completely prevented the development of diabetes in glucose tolerant mice, as shown in Figure 1A. However, it should be noted that Quinoline-3-carboxamide alone was also able to prevent the development of diabetes in these mice. Thus, the inhibition of the immune process was clearly the deciding factor in the protection of the glucose tolerant mice.

Glucose intolerant mice, by contrast, initially rapidly developed diabetes even when treated with the combination of Quinoline-3-carboxamide and reg protein, as shown in Figure 1B. After the first 2 weeks of treatment, the incidence of diabetes was 30-50% in all glucose intolerant mice. However, the incidence of diabetes actually decreased after 4 weeks of treatment with the combination of reg protein and Quinoline-3-carboxamide. By the end of the treatment period, only 13% of mice given the combination of reg protein and Quinoline-3- carboxamide had diabetes, as compared to 80% of mice in the control group (p<0.0001).

Thus, the combination of Quinoline-3-carboxamide and reg protein actually reversed the course of the disease in glucose intolerant NOD mice, rather than merely inhibiting its further progression.

This result is confirmed by the measurement of blood glucose levels in NOD mice, before and after treatment. Blood glucose levels were measured as described above in both glucose tolerant and intolerant mice in all four groups. Results are shown in Table 1. The numbers in parentheses indicate the number of mice tested in each group.

Table 1. Blood Glucose Measurements Group: Glucose Tolerant Glucose Intolerant Age (weeks) 14 24 14 24 Control 6.5+0.2 (15) 15.4+2.7 (15) 8.2+0.5 (14) 21.5+3.1(9) reg alone 6.7+0.2 (15) 9.2+1.9 (13) 8.2+0.4 (14) 19.1+3.4 (13) Quinoline-3- 6.4+0.2 (16) 7.3+0.5 (16) 7.8+0.5 (15) 20.9+2.9 (13) carboxamide reg+ 6.5+0.2 (17) 7.8+0.9 (12) 8.4+0.4 (16) 8.9+0.6 (9) Quinoline-3- carboxamide As can be seen from Table 1, NOD mice in the control group which were initially glucose tolerant had high blood glucose levels and had become glucose intolerant by the end of the treatment period. However, initially glucose tolerant mice in all three treatment groups, Quinoline-3-carboxamide alone, reg protein alone and Quinoline-3-carboxamide + reg protein, maintained blood glucose levels similar to those seen before the start of treatment.

Thus, all three treatments effectively prevented the development of high blood glucose levels in initially glucose tolerant mice.

Mice which were initially glucose intolerant started with higher blood glucose levels before treatment and developed significantly higher blood glucose levels by the end of treatment in all groups except for the group treated with the combination of reg protein and Quinoline-3-carboxamide (p<0.0005). Thus, only the combination of reg protein and Quinoline-3-carboxamide was able to prevent the development of higher blood glucose levels in initially glucose intolerant mice.

IPGTT determinations were also performed as described above, both before and after treatment, in reg-treated glucose intolerant mice, as shown in Table 2. Mice began treatment

at age 14 weeks and ended it 10 weeks later, at age 24 weeks. Blood glucose was measured at the start of the IPGTT determination (0 minutes) and at the end (60 minutes).

Table 2. IPGTT in Glucose Intolerant Mice Blood Glucose (mmol/L) Age (weeks) 0 minutes 60 minutes 14 8.2+2 15.3+6.2 24 7.9+2.7 10.2+2.4 Reg protein treatment prevented the deterioration of glucose tolerance in 6 of 10 mice which did not develop frank diabetes by the end of the treatment period. By contrast, all mice succumbed to diabetes in the control group. Quinoline-3-carboxamide treatment alone also did not prevent the deterioration of glucose tolerance (results not shown). Thus, only reg protein treatment enabled initially glucose intolerant mice to maintain relatively stable responses to the IPGTT.

Therefore, the effects of Quinoline-3-carboxamide and reg protein, alone or in combination, clearly depended upon the extent to which the mice had developed the pathology of diabetes. Quinoline-3-carboxamide and reg protein alone clearly each had a substantial beneficial effect on initially glucose tolerant mice, as shown in Table 1, as well as a very small effect on initially glucose intolerant mice. Therefore, in the amounts administered, clearly each ingredient had some effect on blood glucose levels, even in glucose intolerant mice. However, the effect of the combination on glucose intolerant mice was clearly far greater than mere additivity of the effect of each ingredient separately, as well as being greater than the effect on initially glucose tolerant mice. As noted previously, glucose intolerant mice more closely model the extent of the pathology of IDDM in human patients when the initial clinical diagnosis of diabetes is made. Thus, the combination of Quinoline-3 - carboxamide and reg protein shows unexpected synergistic activity in the amelioration of the disease of diabetes for this model of human IDDM, in that the combination clearly shows a far greater effect than an additive effect of each ingredient separately.

Example 2 Histological Examination of NOD Mice Histological examination of tissue samples from NOD mice demonstrated the ability of the composition of the present invention, a combination of an immunoregulator and a cell proliferation agent, to increase the relative concentration of P cells in the pancreas, as shown in Figures 2A-2F. Such stimulation was not seen in NOD mice treated with either Quinoline-3-carboxamide or reg protein alone. The experimental method was as follows.

The mice from Example 1 were sacrificed at the end of the treatment period and tissue samples were taken from the pancreas. The samples were fixed in 10% formalin in 0.9% saline and embedded in wax. Two sets of 5 serial 5 pm sections were cut for immunolabelling at a cutting interval of 150 pm. Sections were immunolabelled for insulin (guinea pig anti-insulin antisera dilution 1:1000, ICN Thames U.K.) and glucagon (rabbit anti-pancreatic glucagon antisera dilution 1:2000) and detected with peroxidase conjugated anti-guinea pig (Dako, High Wycombe, U.K.) or peroxidase conjugated anti-rabbit antisera (dilution 1:50, Dako). Specific histological samples are shown in Figures 2A-2F, and overall proportions of P cells for glucose tolerant mice are given in Table 3.

Glucose intolerant mice in the control (Figure 2A) or Quinoline-3-carboxamide-treated (Figure 2B) group had small islets consisting largely of glucagon cells with no discernible P cells. Small numbers of cells were observed in one animal from the reg protein-treated group and in one animal from the group treated with the combination of Quinoline-3- carboxamide and the reg protein (data not shown). Thus, the composition of the present invention did not have as strong an effect on the visible mass of P cells as it did on the clinical manifestations of diabetes in glucose intolerant animals.

In the glucose tolerant group, 4 of 14 control mice did have visible P cells (Figure 2C).

Quinoline-3-carboxamide-treated glucose tolerant mice did not show significantly better results: only 3 of 11 mice had P cells (Figure 2E). Even relatively few reg protein-treated glucose tolerant mice had P cells, with only 3 of 9 animals showing P cells (Figure 2D).

However, the combination of reg protein and Quinoline-3-carboxamide clearly increased the proportion of mice with visible P cells to 50% (Figure 2F). Furthermore, as shown in Figure 2F, islet architecture was relatively maintained with a peri-insulitis pattern.

The ranges of the ratio of P cell mass to total pancreatic cell mass ( cell/total pancreas), as well as of the ratio of P cell mass to islet cell mass ( cell/islet), are as follows for all four groups of glucose tolerant mice. First, the P cell/total pancreas ratio for control mice ranged from 0.0001 to 0.007, while the P cell/islet ratio for control mice ranged from 0 to 0.7. Second, the P cell/total pancreas ratio for reg alone mice ranged from 0.0009 to 0.005, while the P cell/islet ratio for reg alone mice ranged from 0 to 0.6. Third, the P cell/total pancreas ratio for Quinoline-3-carboxamide alone mice ranged from 0.0004 to 0.006, while the P cell/islet ratio for Quinoline-3-carboxamide alone mice ranged from 0 to 0.6. Finally, the P celUtotal pancreas ratio for reg protein + Quinoline-3-carboxamide mice ranged from 0.0002 to 0.007, while the cell/islet ratio for reg protein + Quinoline-3-carboxamide mice ranged from 0 to 0.8.

From these results, the ratio of P cell mass to total pancreatic cell mass is not significantly different between the four groups. However, the proportion of P cells within the islets is significantly higher for the group treated with reg protein and Quinoline-3- carboxamide as compared to the control (p<0.05). Thus, clearly the combination of reg protein and Quinoline-3-carboxamide preserved P cell mass in glucose tolerant mice.

Furthermore, the composition of the present invention was also able to increase the number of animals with visible cells.

From these results, as well as those shown in Example 1, clearly the combination of an immunoregulator, Quinoline-3-carboxamide, and a cell proliferation agent, reg protein, was able to significantly reverse the progression of diabetes in glucose intolerant mice, and to visibly preserve cell mass in glucose tolerant mice, which had relatively less severe disease.

Furthermore, even mice which had lost such a significant proportion of P cells as to show overt, frank diabetes were actually able to benefit from this combination, with an accompanying amelioration or complete reversal of the clinical symptoms. Such an effect is of paramount importance for the treatment of human patients with IDDM, since the appearance of clinical symptoms in humans accompanies the destruction of a majority of the P cells in the pancreas. For a treatment for IDDM to be effective, the course of the disease must therefore not only be stopped, but actually reversed, in order to achieve a reduction in the level and frequency of clinical symptoms such as hyperglycemia.

The composition of the present invention, a combination of an immunoregulator and a P cell proliferative agent, fulfills this stringent requirement. Such a requirement is particularly difficult since mature P cells do not normally proliferate, as noted above. For a reversal of the disease state, the process of the autoimmune reaction must be stopped, and the normal inhibition against proliferation of mature P cells must be lifted. Furthermore, the P cells which had been involved in the pathological process of diabetes mellitus must also be able to recover sufficiently to both proliferate and to produce enough insulin to reduce or reverse clinical symptoms of the disease. Thus, the ability of the composition of the present invention to achieve both effects, and then to actually reverse the course of diabetes, is unexpected, and was neither taught nor suggested by the prior art.

Example 3 Adenoviral Vector Mediated Gene Transfer for Reg Protein The efficacy of the expression of reg gene after transfer by an adenoviral vector was examined in female NOD (non-obese diabetic) mice. The effects of transfection with an adenoviral vector containing a luciferase reporter gene, or of injections of saline, were examined as a control. Mice receiving the saline injections clearly showed progression of the disease so that only twenty percent of the mice were not diabetic by twenty weeks of age.

Mice receiving injections of the adenoviral vector containing the luciferase reporter gene showed a lower incidence of diabetes by twenty weeks, such that sixty percent of the mice were not diabetic. Mice receiving injections of the adenoviral vector containing the reg gene showed by far the lowest incidence of diabetes by twenty weeks, such that ninety percent of the mice were not diabetic. Thus, clearly the combination of the expression of the reg gene and of the effect of transfection by the adenoviral vector showed the greatest efficacy for preventing the development of diabetes in NOD mice.

Such efficacy is particularly unexpected because the adenoviral vector has been shown to cause histological damage to pancreatic tissue, which was manifested as long as the gene carried by the adenoviral vector was being expressed (McClane, S. J. et al., Human Gene Therapy, 8:739-746, 1997). The observed tissue damage included edema, inflammation, cell destruction and vacuolization. Since the pathology of IDDM results in a significant reduction in pancreatic function through the destruction of the P cells, the further destructive effects of

the adenoviral vector would have been expected to result in even greater damage to the pancreas, thereby increasing the incidence of diabetic mice. Thus, the prior art clearly taught against the use of adenoviral vector mediated gene transfer for the treatment of diabetes.

By contrast, the combination of the adenoviral vector and the expression of the reg gene clearly demonstrated high efficacy for preventing the onset of clinical diabetes in the experimental model of NOD mice. Thus, the adenoviral vector containing the reg gene showed unexpectedly efficacy for the treatment of IDDM. The experimental method was as follows.

First, the reg adenoviral expression vector was formed by introducing the human reg cDNA encompassing the coding sequence into the pJM17 plasmid and co-transfected with the plasmid pACCMV.pLpa into 293 cells as previously described (Becker, T.C. et al., Methods in CellBiol., 43:161-189, 1994). Recombinant virus, hereinafter referred to as "Ad reg", was recovered from the cell lysate and replated on 293 cells for cloning. Viral DNA derived from recombinant Ad reg were screened by Southern blot according to methods well known in the art (see for example Current Protocols in Human Genetics, N.G. Dracopoli et al., eds., John Wiley & Sons, Inc., New York, New York, 1997), and were found to contain the expected reg coding sequence. The resultant virus allows tissue non-specific expression using a CMV promoter and SV40 polyadenylation signal.

The luciferase reporter adenoviral expression vector, hereinafter referred to as "Ad luc", was constructed in a similar manner to Ad reg, with an adenovirus 5 plasmid, the CMV promotor and the SV40 polyadenylation signal, except that the coding sequence coded for luciferase rather than for reg protein.

Adenovirus stocks of both the reg adenoviral expression vector (Ad reg) and of the luciferase reporter adenoviral expression vector (Ad luc) were then prepared as follows.

Duplicates of subconfluent 293 (60 mm) plates were infected with serial dilutions of the presumably polyclonal recombinant virus stock, ranging from a 10-5 to a 10-9 dilution of the infectious lysate, in 2% fetal bovine serum in DMEM (Dulbecco's Modified Eagle's Medium) medium (FBS-DMEM). After a 1 hour incubation, the infection medium was removed and replaced with 5 ml complete DMEM for an overnight incubation. Next, the medium was replaced with an agar overlay, enabling isolation of recombinant adenovirus clones from the individual plaques formed. Several well spaced plaques were isolated 6-8 days after infection, by removing the agar-plug above the plaque with a pipette tip. The agar plug was then placed in 2 ml of 2% FBS-DMEM and subjected to freeze-thaw cycles to release the virions from the

agar. A 2 ml aliquot was used to infect subconfluent 293 cells, which underwent lysis within a week when the clone isolation was successful. The medium and remaining cells were collected from the lytic plates, the remaining intact cells were lysed by freeze-thaw cycles, and the debris was pelleted by microcentrifugation at 3,000 rpm for 5 minutes. The supernatant (designated as P1) was stored at -20 OC. To increase the certainty of isolating a single viral clone, the procedure was repeated twice more.

The resultant clones were used to infect 293 plates for DNA isolation, which was subjected to restriction enzyme digestion and Southern blot analysis, to verify successful subcloning of the gene of interest in the recombinant adenovirus, either the reg gene or the luciferase reporter gene. The P1 stock of the chosen Ad clone was then used to infect 293 cells for further amplication of the viral stock. The supernatant of lysed 293 plates infected with P 1 (after freeze-thaw cyles to release virions from cells which remained intact, and pelleting the debris by microcentrifugation at 3000 rpm for 5 minutes) was designated as P2.

The viral titer was then determined by infecting subconfluent 293 cells with duplicates of serial dilutions ofthe viral stock, ranging from 10-3 to 10-1° particles. Volumes of 1 ml FBS-DMEM were used to infect 60 mm plates for one hour, after which most of the infection medium was removed and gently replaced with DMEM containing 10% FCS. After an overnight incubation, before the appearance of plaques, the medium was replaced with a mixture of 2X DMEM (with 4% FCS and 2% glutamine, warmed to 37 "C) and 1.3% Noble agar (cooled to 45 "C); and allowed to solidify. Plaques were visible to the naked eye by day four, and were counted on the tenth day after infection. The final viral stocks for in vivo use were purified on CsCl gradients and desalted as previously described (Yee, J.-K., "Vectors for gene therapy", in Current Protocols in Human Genetics, N.G. Dracopoli et al., eds., John Wiley & Sons, Inc., New York, New York, 1997, pp. 12.4.6-12.4.15).

Once the viral stocks had been prepared, gene transfer and expression was analyzed as follows. First, the ability of both Ad reg and Ad luc to infect pancreatic cell lines in vitro was examined. The two pancreatic cell lines tested were PTC-tet and AR42J. The cells were incubated with Ad reg at a concentration of either 10 pfu/cell or 20 pfu/cell, or with Ad luc as a control, essentially as previously described (Becker, T.C. et al., Methods in Cell Biol., 43:161-189, 1994). Both Northern blot analysis and Western blot analysis were performed as previously described (see for example Current Protocols in Molecular Biology, F.M. Ausubel

et al., eds., John Wiley & Sons, Inc., New York, New York, 1997). The results are shown in Figures 3A (Northern blot analysis) and 3B (Western blot analysis).

The Northern blot shows that no endogenous expression of the reg gene is seen in the P cell line PTC-tet (lane 1), although some endogenous expression is seen in the amphicrine pancreatic cell line AR42J (lane 4). For both cell lines, expression of the reg gene is related to the amount of Ad reg given in a dose dependent manner ("L" indicates a dose of 10 pfu/cell, while "H" indicates a dose of 20 pfu/cell).

The Western blot shows the reg protein migrating as two bands at about 14 kDa and 16 kDa. Reg protein is clearly seen in the AR42J cell line (lanes 6 and 7) while no specific protein band is seen when Ad luc was given (lane 8). Low levels of reg protein are also seen in the PTCtet line (lanes 3 and 4). Thus, the Western blot confirms that Ad reg can actually cause pancreatic cell lines to produce reg protein.

In addition, expression of reg was found at both the mRNA and peptide levels in infected RIN (rat insulinoma cell line) and AtT-20 cells (pituitary corticotroph cell line, as an example of a peptide-secreting non-P cell line) (data not shown).

Next, the expression of Ad luc in NOD (non-obese diabetic) mice was examined.

Nine male NOD mice were injected intravenously with 10 pfu/kg of Ad luc. Three mice were then sacrificed four days, seven days or fourteen days after infection. The pancreas, liver, spleen, kidney and heart were removed from the mice. The tissues of each organ were homogenized on ice by tissue Dounce homogenization in PLB passive lysis buffer (Promega, Madison, Wisconsin, USA). Luciferase activity was then determined by luminometry with the Luciferase Assay Kit (Promega, Madison, Wisconsin, USA). The assay was performed according to the instructions of the manufacturer. The results are shown in Figure 4.

Expression of the transgene in various organs of NOD mice was clearly greatest in the liver, at the level of about two orders of magnitude greater. Expression in the pancreas is not seen until day seven and continues to be seen at day fourteen. Thus, a single injection of Ad luc viral particles resulted in expression of the luciferase gene at least between seven days and fourteen days after injection.

Finally, the ability of Ad reg to substantially prevent or to delay the onset of clinical diabetes in female NOD mice was examined. The animals were divided into four groups of ten animals each: control (no injection), injection of saline, injection of Ad luc (both as controls for injection), and injection of Ad reg. Female NOD mice received daily intravenous

(i.v.) injections of Ad reg (1010 particles per injection per animal), or of Ad luc (1010 particles per injection per animal) or saline as a control, from 5 to 12 weeks of age. The volume of saline injection was the same for all injected animals. As a further control, one group of animals did not receive any injections. The mice were fed and supplied with water ad libitum.

At five weeks of age, the animals were glucose tolerant, and hence not diabetic. The presence of diabetes, and hence of glucose intolerance, was determined by two consecutive glucose positive glucose levels in the urine. The level of urine glucose in the NOD mice was determined on a bi-weekly basis using Labstix (Bayer Diagnostics, Hampshire, England).

Weight and fluid intake were also determined on a bi-weekly basis.

Figure 5 shows the percentage of mice remaining non-diabetic (y-axis) against the age of the mice in weeks (x-axis). The period of injections is indicated by the box labelled "virus infusions". As shown control (non-injection) mice began to develop diabetes by week 14, and all of the mice were diabetic by week 18. Onset of diabetes was seen for mice receiving saline injections by week 15, and only twenty percent of the mice were non-diabetic by week twenty. Thus, without treatment, almost all of the non-diabetic mice eventually developed diabetes.

Injections of Ad luc showed some protective effect, with no mice developing diabetes before 17 weeks of age, and only forty percent of the mice developing diabetes by 20 weeks of age. Ad reg showed an even greater protective effect, with no development of diabetes seen until 20 weeks of age. Even then, only 10 percent of the mice developed diabetes. Thus, although Ad luc shows some protective effect, the greatest effect was clearly seen with Ad reg.

Furthermore, Ad reg was able to both prevent the onset of diabetes in nearly all mice tested, and to substantially delay the onset of diabetes in those mice which did develop diabetes, in a system in which all or nearly all of the mice develop diabetes if not treated. By contrast, the previous Examples show the effects of reg and Quinoline-3-carboxamide in a system in which the mice are selected according to the presence of diabetes, at an age at which diabetes is normally displayed in the NOD mouse model. In that system, reg and Quinoline- 3-carboxamide separately showed protective effects only for NOD mice which appeared to show resistance to the onset of clinical diabetes, while failing to prevent the further deterioration of mice which already had diabetes. Thus, Ad reg is able to prevent the onset of diabetes in substantially all mice tested, regardless of any inherent susceptibility to the onset of clinical diabetes symptoms.

In addition, Ad luc did show some protective effect for some mice. Without wishing to be bound by a particular hypothesis, the effect may be mediated through the elicitation of a TNFa (Tumor Necrosis Factor) response which has been previously observed (Lieber, A., et al., J. Virol., 71:8798-8807, 1997). Other cytokines may also be involved. Although the possibility of an effect by the luciferase protein itself cannot be eliminated, such an effect seems unlikely.

Regardless of the mechanism of mediation of the effect, since Ad luc shows a protective effect for only a fraction of the mice, the effect may be limited to mice already somewhat resistant to the development of clinical symptoms of diabetes. The far more extensive effect of Ad reg clearly shows that the combination of the adenoviral vector, possibly as an immunomodulator, and of reg protein is able to protect nearly all mice. Thus, presumably treatment with Ad reg would be efficacious for both substantially halting and reversing the progress of IDDM in human patients.

Example 4 Suitable Formulations for Administration of the Composition The composition of the present invention, a combination of an immunoregulator and a P cell proliferative agent, can be administered to a subject in a number of ways, which are well known in the art. For example, administration may be done topically (including ophtalmically, vaginally, rectally, intranasally and by inhalation), orally, or parenterally, for example by intravenous drip or intraperitoneal, subcutaneous, or intramuscular injection. If the cell proliferative agent is a protein such as reg, administration is preferably done parenterally.

If the immunoregulator is the adenoviral vector, and the P cell proliferative agent is the reg protein, then the composition is preferably administered either parenterally, for example by injection, or topically, for example intranasally. The designation of the adenoviral vector as an "immunoregulator" is not intended to limit this embodiment to any particular mechanism or mechanisms, but is done only for the sake of clarity.

In addition, the combination of the reg gene and a suitable vector, such as the adenovirus vector, for example as described in Example 3 above, can be described as a reg

vector. Such a reg vector is a composition for gene therapy for treating IDDM according to the present invention.

Formulations for topical administration may include but are not limited to lotions, ointments, gels, creams, suppositories, drops, liquids, sprays and powders. Conventional pharmaceutical carriers, aqueous, powder or oily bases, thickeners and the like may be necessary or desirable.

Compositions for oral administration include powders or granules, suspensions or solutions in water or non-aqueous media, sachets, capsules or tablets. Thickeners, diluents, flavorings, dispersing aids, emulsifiers or binders may be desirable.

Formulations for parenteral administration may include but are not limited to sterile aqueous solutions which may also contain buffers, diluents and other suitable additives.

Dosing is dependent on the severity of the symptoms and on the responsiveness of the subject to the composition of the present invention. Persons of ordinary skill in the art can easily determine optimum dosages, dosing methodologies and repetition rates.

Example 5 Method of Treatment of Diabetes As noted above, the composition of the present invention, a combination of an immunoregulator and a cell proliferative agent, has been shown to be an effective treatment for diabetes. The following example is an illustration only of a method of treating diabetes with the composition of the present invention, and is not intended to be limiting.

The method includes the step of administering the composition of the present invention, in a pharmaceutically acceptable carrier as described in Example 4 above, to a subject to be treated. The composition is administered according to an effective dosing methodology, preferably until a predefined endpoint is reached, such as the absence of further progression of diabetes in the subject, the amelioration of clinical symptoms, the reversal of the progression of diabetes or the prevention of the appearance of clinical symptoms.

According to another embodiment of the present invention for gene therapy, the method is as follows. First, the reg gene is placed in a suitable vector, such as the adenovirus vector, for example as described in Example 3 above, to form a reg vector. Many different such vectors are known in the art, as are many different methods for placing the gene in such a

vector. Next, the reg vector is administered to the subject by a suitable route of administration, for example parenterally or topically. Particularly preferred routes of administration include injection (intravenously or intramuscularly) and intranasally. Thus, such a method for gene therapy also provides a treatment for IDDM, regardless of the mechanism of action of the reg vector.

Example 6 Method of Manufacture of a Medicament The following is an example of a method of manufacturing a medicament containing the composition of the present invention. First, both the immunoregulator and the P cell proliferative agent are synthesized in accordance with good pharmaceutical manufacturing practice. Examples of such methods are well known in the art. Next, the immunoregulator and the P cell proliferative agent are placed in a suitable pharmaceutical carrier, as described in Example 4 above, again in accordance with good pharmaceutical manufacturing practice.

As another example for the method of manufacturing a medicament for gene therapy, first, the reg gene is placed in a suitable vector, such as the adenovirus vector, for example as described in Example 3 above, to form a reg vector. Many different such vectors are known in the art, as are many different methods for placing the gene in such a vector. Next, preferably, the reg vector is placed in a suitable pharmaceutical carrier prior to administration to the subject.

It will be appreciated that the above descriptions are intended only to serve as examples, and that many other embodiments are possible within the spirit and the scope of the present invention.