CZ20003065A3 - Insulin peptide analogs and their use for treating diabetes mellitus - Google Patents

Abstract

Insulin B chain peptide analogs derived from peptides containing residues (9-23) of wild-type B chain. Analogs are altered in nature sequences in positions (12, 13, 15) and / or (16), and may also be changed in position (19) and / or in other positions. Pharmaceutical compositions comprising these peptide analogs. Peptide analogs are useful in the treatment or inhibition of formation diabetes.

Description

Field of the invention

The present invention relates generally to peptide analogues of insulin, and more particularly to methods of treating diabetes mellitus using peptide analogues derived from residues 9-23 of the human insulin B-chain.

BACKGROUND OF THE INVENTION

Insulin-dependent diabetes mellitus (IDDM) is an organ-specific autoimmune disease affecting nearly a million paceintes in different age groups in the United States. The disease is characterized by widespread destruction of beta cells in insulin-producing pancreatic islets and impaired glucose metabolism regulation that leads to clinical diabetes. The defining feature of IDDM is lymphocyte islet infiltration. Of the infiltrating cells, T cells appear to be one of the major mediators of autoimmune destruction.

Type I diabetes is further characterized by an increased concentration of antibodies against various islet-associated antigens, including insulin, GAD65, GAD67, and ICA512. These antibodies can be detected well in advance of clinical disease and the immune response against such antigens can be used as a predictor of developing diabetes in patients with predisposing genetic (HLA) haplotypes.

Currently, patients are dependent on insulin injections to maintain normoglycaemia. Insulin is a polypeptide hormone consisting of two disulfide-linked chains, A *

a chain consisting of 21 amino acid residues and a B chain consisting of 30 residues. Although insulin administration is of considerable benefit to patients suffering from diabetes, it is difficult to maintain the correct dosage due to the short serum half-life of insulin. Insulin administration may lead to the development of various hypoglycaemic side effects and the generation of neutralizing antibodies.

Because of the problems associated with existing diabetes treatments, there is a need for improved treatment that is more effective and not associated with such disadvantages. The present invention utilizes peptide analogs that antagonize the T-cell response to insulin for the effective treatment of diabetes, and further provides related benefits.

SUMMARY OF THE INVENTION

The present invention provides compounds and methods for treating and preventing diabetes. In some aspects, the present invention provides peptide analogs comprising human insulin B chain residues 9 to 23 (SEQ ID NO: 2), wherein the sequence of these peptide analogs differs from native human B chain residues 9 to 23 of 1-4 amino acid substitutions. Such substitutions may be made at one or more residues selected from the group consisting of residues 12, 13, 15 and 16, with or without further substitutions at other residues. In some preferred embodiments, such substitutions may be present in two or three amino acid residues within residues 9-23 of the B chain of insulin. The substitutions may also be present in residue 19. The substitutions are preferably non-conservative, and analogs in which residues 12, 13, 15, 16 and / or 19 are changed (for example to alanine) are preferred. Also preferred are analogs which further comprise a residue 24 of the insulin B chain. In some other embodiments, the peptide analogs comprise no more than 18 residues, no more than 16 residues, and no more than 15 residues of the human insulin B chain.

In other embodiments, the peptide analogs consist essentially of residues 9 to 23 or 9 to 24 of the human insulin B chain (SEQ ID NO: 2), wherein the sequence of the peptide analogs differs from the sequence of residues 9 to 23 of the natural human B insulin chain in substitutions in 1 to 4 amino acids, and wherein at least one substitution is made at a residue selected from the group consisting of residues 12, 13, 15, and 16.

In another aspect, the invention provides pharmaceutical compositions comprising the peptide analog described above in combination with a physiologically acceptable carrier or diluent.

The present invention further comprises a method for treating and / or inhibiting the development of diabetes, which comprises administering to a patient a therapeutically effective amount of a pharmaceutical composition as described above.

These and other aspects of the invention will be apparent from the following detailed description and the accompanying drawings. In the following, various references are described in detail that describe some procedures or means. These references are all incorporated herein in their entirety.

Description of the figures in the attached drawings

Giant. 1 shows the amino acid sequence of residues 9-23 of the insulin B chain (SEQ ID NO: 2);

Giant. 2 is a graph showing the proliferative response (measured in cpm) of a NOD mouse T cell clone to (9-23) wild type insulin B peptide in the presence of various amounts of representative peptide analogs in which different alanine residues are substituted as indicated.

Giant. 3 is a graph showing the proliferative response (measured in cpm) of a T cell clone of NOD mice to (9-23) wild type insulin B peptide in the presence of various amounts of representative peptide analogs in which amino acids at positions 16 and 19 are substituted with alanine (NBI-6024) , squares). For comparison, a proliferative response in the presence of an unrelated control peptide derived from myelin basic protein (NBI-5096, circles) is also shown. The response is measured as the average CPM + SEM for the three cultures.

Giant. 4-6 are histograms illustrating the proliferative response (measured in cmp) of T cell line from diabetic patients to the natural B chain (9-23) peptide or representative peptide analogs. Peripheral blood mononuclear cells were isolated from. patients with diabetes and were cultured in the presence of (9-23) insulin B chain peptide.

After three cycles of insulin B chain restimulation (9-23), 1 x 10 ⁵ T lymphocytes and 7 x 10 ⁴ irradiated autologous PBMCs were added to each round bottom well of a 96-well plate in complete medium. Cells were cultured for 5 days with NBI 6024 (insulin B chain 9-23 with alanine substitutions at positions 16 and 19), insulin B chain (9-23) or medium only. On day 4, cells were exposed to a ³ H-thymidine pulse and cultured for an additional 18 hours. The cultures were then harvested, read using a liquid scintillation counter, and data expressed as mean counts per minute (cpm) for repeated samples + standard deviation (SEM).

Giant. 7 is a histogram illustrating the proliferative response (measured in cmp) of T cell line from diabetic patients to the natural B chain (9-23) peptide or to representative peptide analogs containing said alanine substitutions. Peripheral blood mononuclear cells were isolated from patients with diabetes and were cultured in the presence of (9-23) insulin B chain peptide. After three cycles of restimulation of insulin B chain (9-23) was added to each well in a round bottom 96 well plate in complete medium added to 1 x ¹⁰ 7 cells and T χ 10 ⁴ irradiated autologous PBMCs. Cells were cultured for 5 days with analog, insulin B chain (9-23) or medium only (BKG) as indicated. On day 4, cells were exposed to a ³ H-thymidine pulse and cultured for an additional 18 hours. The cultures were then harvested, read using liquid scintillation counting, and data expressed as mean counts per minute (cpm) for repeated samples + standard deviation (SEM).

Giant. 8 is a graph showing the percentage of female NOD mice that were diabetic after 9 weeks of treatment with representative peptide analogs. 10 mice were treated subcutaneously from day 24 with 24-chain peptide analogs (9-23) containing alanine substitutions at residue 12 (open triangles), 13 (squares), or 16 (solid triangles). All mice treated with the control peptide, neurotensin (circles), had diabetes.

Giant. 9 is a graph showing the same data as FIG. 8, but showing only the A13 analog treated group and the control peptide (neurotensin) treated group.

Giant. 10 is a graph showing the percentage of female NOD mice that •

were diabetic after 13 weeks of treatment with representative peptide analogs. 10 mice were treated subcutaneously from day 24

400 µg of neurotensin (squares), B chain (9-23) (diamonds) or B chain peptide (9-23) containing an alanine substitution at residue 13 (triangles).

Giant. 11 is a graph showing the effect of representative peptide analogs on the incidence of diabetes in NOD mice. Four week old female NOD mice (n = 9) were treated with subcutaneous injections of 20 mg / kg NBI-6024 (A ^16.1 ®) at weekly intervals for 12 weeks and then every other week until age 35 weeks. The control group (n = 10) consisted of untreated animals. Mice with blood glucose concentrations greater than 200 mg / dL in the two following measurements were considered diabetic mice. Data are expressed as a percentage of non-diabetic mice during a 35-week study. The log-rank test was used to assess whether the results for the two treatment groups were statistically significantly different. The percent of NOD mice that were diabetic following treatment with NBI-6024 (A ^16-1 ®) is given for different times squares, and the percent of mice that were diabetic in the control group, the said wheel.

Giant. 12 is a graph showing the effect of representative peptide analogues on the incidence of diabetes in NOD mice. Four week old female NOD mice (n = 13-15) were treated with subcutaneous injections of 20 mg / kg NBI-6024 (A ¹ ® ³ · ⁹ ) or NBI-6201 (control peptide, neurotensin) at weekly intervals for 12 weeks and then every other week to the age of 35 weeks. Another control group (n = 8) consisted of untreated animals. Mice with blood glucose concentrations greater than 200 mg / dL in the two following measurements were considered diabetic mice. Data are expressed as a percentage of non-diabetic mice during a 35-week study. The log-rank test was used to assess whether the results for the two treatment groups were statistically significantly different. The percent of NOD mice that were diabetic following treatment with NBI-6024 (A ^{16 '3-9)} is given for different times squares percentage of diabetic mice after treatment Neurotensine outlined triangles, and the percent of mice that were diabetic in the control group, the said wheel .

Giant. 13A-13D are graphs illustrating the immunogenicity of representative peptide analogs containing alanine substitution residues 9-23 of B chain at residue 12 (Figure 13A), residue 13 (Figure 13B), residue 15 (Figure 13C) or residue 16 (Figure 13A). 13D). NOD mice were injected 2-4 times subcutaneously with the peptide analogue in soluble form before the proliferative response of the lymph node cells to various concentrations of the peptide analogue or peptide (9-23) of the natural B chain was tested. The proliferative response was determined by the amount of radiolabeled thymidine incorporated into the cells (plotted as the average counts per minute (cpm) of the three culture wells) as determined in a liquid scintillation counter after completion of the culture period.

Giant. 14A-14F are graphs showing the immunogenicity of six different peptide analogs in NOD mice. Peptide analogs with two alanine substitutions (A12,13; A12,15; A12,16; A13,15; A13,16 and A15,16) were injected into NOD mice and after 10 days their lymph node cells were used in a proliferation assay after using different concentrations of the immunizing peptide as a stimulator. The proliferative response was determined by the amount of radiolabeled thymidine incorporated into the cells (plotted as the average counts per minute (cpm) of the three culture wells) as determined in a liquid scintillation counter after completion of the culture period.

• · • ·

Giant. 15A-15D are graphs illustrating the immunogenicity of representative double substituted insulin B chain analogues (9-23). The following peptides were tested for ability to elicit an immune response in NOD mice: A12.19 (Fig. 15A); A13.19 (Fig. 15B); A15.19 (Fig. 15C); A16.19 (Fig. 15D). The proliferative response (counts per minute) of the lymph node in response to the immunization analogue as well as the peptide (9-23) of the natural insulin B chain is shown.

Giant. 16 is a graph showing the ability of a series of triple substituted peptides to elicit a T cell proliferative response in NOD mice. Mice were immunized with individual representative peptide analogs containing the following combinations of substitutions: A12,13,19; A12,15,19; A12,16,19; A13,15,19; or A13,16,19; A15,16,19. The lymph node cells were then used in the proliferation assay and the response to each immunization peptide at different concentrations is shown.

Giant. 17A and 17B are graphs showing the ability of a double-substituted peptide series (A ^xs ' ^x9 ) to elicit an immune response. Five female NOD mice were immunized subcutaneously with 20 mg / kg of soluble NBI-6024 on days 1, 6, and 12. On day 15, mice were sacrificed, inguinal lymph nodes were removed, and node cells were cultured at various concentrations (0-50 µM). ) either NBI-6024 (Fig. 17A) or insulin B chain peptide (9-23) (Fig. 17B). The extent of T-cell proliferation was determined by incorporating ³ H-thymidine.

The reaction is expressed as the average cpm + SEM of the three cultures.

Giant. 18 is a histogram showing a comparison of cytokines produced by immune cells after induction and ^{X 6 'X 9} (NBI-6024) peptide in the presence or absence of adjuvant.

• · · ·

Groups of NOD mice were immunized with NBI-6024 alone or in emulsion with CFA. IL-2 and IL4 cytokine concentrations, as indicated, were measured at 25 µM NBI-6024 and are reported in pg / ml minus basal values.

Before describing the present invention, it is desirable to provide definitions of certain terms for a better understanding.

The term insulin B chain refers to a polypeptide of 30 amino acids constituting one of two disulfide-linked polypeptides in insulin. The sequence of the human insulin chain is shown in SEQ ID NO: 1 and the sequence of residues 9-23 of the human B chain is shown in Figure 1 and in SEQ ID NO: 2.

The term insulin B chain peptide analogues refers to at least 15 amino acid residues derived from residues 9-23 of the human insulin B chain (SEQ ID NO: 2), with at least one different amino acid in sequence between the analog and the native B chain. In the peptide analogue, at least one amino acid sequence difference occurs in residue 12,

13, 15 and / or 16. Further, substituted residue 19 may be substituted and other alterations are possible. Preferably, the peptide analog contains 1 to 4 substitutions at residues 9-23 relative to the natural sequence (9-23) of the insulin B chain, although multiple substitutions (e.g. 5 or 6) are possible. Additional residues from the insulin B chain may be present up to the complete 30 residues of the natural B chain, preferably up to a maximum of 25 residues, preferably up to a total of 16 or 18 residues of the peptide analogue. In a preferred embodiment, the peptide analog also comprises a residue 24 of the insulin B chain. Sequences that are not derived from

the insulin B chain, may or may not be present at the amino and / or carboxyl terminus of the peptide analog. Such sequences can be used, for example, to facilitate the synthesis, purification, or solubilization of a peptide analog.

Unless otherwise indicated, the indicated amino acids are in L-form. The L-amino acid residue in the native peptide sequence can be altered to any of the 20 L-amino acids commonly found in proteins, to any of the respective D-amino acids, rare amino acids such as

4-hydroxyproline or hydroxylysine, or to a non-protein amino acid such as β-alanine or homoserine. Also included within the scope of the present invention are analogs containing amino acids that have been altered by chemical means such as methylation (e.g., -methylvaline); amidating the C-terminal amino acid with an alkylamine such as ethylamine, ethanolamine or ethylenediamine; and / or acylation or methylation of an amino acid side chain group (such as acylation of the epsilon-amino group of a lysine).

The terms residue 12 ", residue 13, residue 15, residue 16, and residue 19 (also referred to as poke 12, position 13, position 15, position 16, and position 19, respectively) refer to amino acids 12, 13, 15, 16, and 19. More specifically, the numbering system for these residues refers to the amino acid positions in the natural human protein, regardless of the length of the peptide analogue or the amino acid position in the analogue. Peptide analogs having alanine substitutions at residues 12, 13, 15 or 16 are referred to as A12, A13, A15, or A16, respectively.

Peptide analogues of insulin B chain

As noted above, the present invention provides peptide analogs comprising at least residues 9-23 of the human insulin B chain and comprising a change in natural L-valine at position 12, L-glutamate at position 13, L-leucine at position 15 and / or L-tyrosine at position 16, to another amino acid. In one embodiment, the peptide analogs comprise further alterations of one to three L-amino acids at positions 12, 13, 15, 16 and / or 19 of the insulin B chain. Preferably, the peptide analogs comprise two or three alterations in which one of the substituted residues is at position 19.

The portion of the peptide analog that is derived from the insulin B chain typically has a length of 15-30 residues, preferably 15-18 residues, more preferably 15-16 residues. Particularly preferred peptide analogs comprise 15 amino acids derived from the insulin B chain.

As noted above, peptide analogs comprising any amino acid changes at the positions listed above are within the scope of the present invention. Preferred peptide analogs include non-conservative substitutions (i.e., substitutions for an amino acid having a different charge, polarity, hydrophobicity, and / or size). Particularly preferred analogs comprise substitution of one or more residues with alanine.

Peptide analogs can be synthesized by standard chemical techniques, including automated synthesis. In general, peptide analogs can be prepared by a solid phase peptide synthesis technique that comprises attaching each protected amino acid residue to a resin support, preferably a 4-methylbenzhydrylamine resin, by activation with dicyclohexylcarbodiimide, to yield the peptide with a C-terminal amide. Alternatively, a chloromethyl resin (Merrifield resin) may be used to obtain peptides with free carboxylic acid at the C-terminus. The side chain functional groups may be protected as follows: benzyl for serine and threonine; cyclohexyl for glutamic and aspartic acids; tosyl for histidine and arginine; 2-chlorobenzyloxycarbonyl for lysine; and

2-bromobenzyloxycarbonyl for tyrosine. After the coupling reaction, the t-butyloxycarbonyl protecting group on the alpha-amino group of the added amino acid can be removed by treatment with trifluoroacetic acid, followed by neutralization with di-isopropylethylamine. The additional protected residue is then attached to the free amino group, resulting in elongation of the peptide chain. After binding of the last residue, the protected peptide-resin is treated with hydrogen fluoride to cleave the peptides from the resin and remove the protecting groups for the side chain functional groups. The crude material can be further purified by gel filtration, HPLC, partition chromatography, or ion exchange chromatography, using well known techniques.

The peptide analogs of the present invention should (a) neither stimulate T cell clones of NOD mice specific for native (9-23) insulin B chain peptides (SEQ ID NO: 2) nor stimulate such clones at a level below the level stimulated with natural peptide; (b) not stimulate human T cells from patients specific for insulin B chain (9-23); (c) not be immunogenic in NOD mice; (d.) they should reduce the incidence of diabetes in NOD mice; and (e) they should inhibit the reaction of T cell clones specific for the peptide (9-23) of the natural insulin B chain (SEQ ID NO: 2). Therefore, potential peptide analogs can be screened for their ability in the treatment of diabetes by assays measuring T cell proliferation, immunogenicity in NOD mice, and the effect on disease incidence in NOD mice. Some representative tests for

uses in evaluating potential peptide analogs are described in more detail below. Those analogs that meet the above criteria are useful therapeutic agents.

Potential peptide analogs may first be tested for the ability to stimulate T cells (9-23) of the native insulin B chain (SEQ ID NO: 2) (from clonal cell lines or isolated from the patient). Such assays can be performed using a direct proliferation assay in which T cell lines reactive with natural B chain (9-23) or T cells from a patient are used as target cells. T cell lines can be obtained, using well known techniques, from lymph nodes obtained from rats injected with the B chain (9-23). Lymph node cells can be isolated and cultured for 5-8 days with B chain (9-23) and IL-2. Viable cells are isolated and a second round of B chain stimulation (9-23) and irradiated splenocytes can be performed as a source of growth factors. After 5 to 6 such passages, the proliferative potential of each cell line is determined. To perform the proliferation assay, B-chain reactive T-cell lines (9-23) can be cultured for 3 days with different concentrations of peptide analogs and irradiated autologous splenocytes. After three days, 0.5-1.0 µCi of [ ³ H] -thymidine is added for 12-16 hours. The cultures are then harvested and the incorporation of radioactivity determined. The mean cpm and standard deviation are calculated from three cultures. Peptide analogs with results that are less than three standard deviations from the average for a comparable B chain concentration (9-23) are considered non-stimulatory. Peptide analogs that do not stimulate proliferation at concentrations less than or equal to 20-50 µM are suitable for further testing.

Potentially suitable peptides that do not stimulate B chain-specific T cells (9-23) and which preferably inhibit the response of such T cells in vitro are further tested for their immunogenicity in NOD mice. Briefly, groups of NOD mice can be immunized with 100-400 µg of potentially useful peptides subcutaneously in mannitol acetate buffer three times within 10-15 days. After the last immunization, the lymph node and / or spleen cells can be used in a proliferation assay in which different concentrations of the immunizing peptides are cultured with the cells for 3-4 days. The last 18 hours of culture can be performed with thymidine-labeled tritium. Cells can then be harvested and read in a scintillation counter and the proliferative response can be expressed as CPM + SEM. Potentially useful peptides that induce proliferation that is at least 2-fold higher than background (no antigen) at 25 µM peptides are considered immunogenic. Alternatively, a potential peptide analog is considered immunogenic if it elicits a proliterative response after immunization with NOD mice in complete Freund's adjuvant. Drainage lymph node or spleen cells, when cultured in the presence of an immunizing analog, should induce a proliferation that is at least 2-fold higher than the baseline (no antigen) proliferation at 25 µM peptides.

Potentially useful peptides that can inhibit B chain-induced proliferation (9-23) are further tested for their ability to reduce the incidence of diabetes in NOD mice. Briefly, the peptides can be administered to NOD mice in soluble form or in an emulsion with, for example, incomplete Freund's adjuvant (IFA). Typically, about 400 µg of peptides per week is sufficient. The incidence of diabetes in treated mice as well as untreated or control mice is then evaluated by weekly monitoring of blood glucose concentrations. A blood glucose value of 200 mg / dl or more in two consecutive measurements indicates the presence of diabetes. Peptide analogs should lead statistically

significantly reducing the percentage of NOD mice affected with diabetes during a follow-up of approximately 25 weeks.

As mentioned above, peptide analogs may also inhibit the B chain (9-23) specific T cell response in vitro. Such inhibition can be measured by competitive assays in which potentially useful peptide analogs are tested for the ability to inhibit native B chain-induced T cell proliferation (9-23) (SEQ ID NO:

2). In such an assay, antigen-presenting cells are first irradiated and then incubated with the competing peptide analog and the natural B chain peptide (9-23). The T cells are then added to the culture. Cultures contain different concentrations of test peptide analogs and can be incubated for a total of 4 days. After incubation, each culture is pulsed, for example, 1 µCi of [ ³ H] -thymidine for an additional 12-18 hours. The cultures can then be harvested on glass fiber filters and read as above. Mean cpm and standard deviation can be determined from data obtained from triplicate cultures. Preferred are those peptide analogs that reduce proliferation by at least 25% at a concentration of 20-50 µM.

Treatment and prevention of diabetes

As noted above, the present invention provides methods for treating and preventing type 1 diabetes in which a therapeutically effective amount of a insulin analogue of the B chain as described herein is administered to a patient. Diabetic patients suitable for such treatment may be identified by criteria for diagnosing clinical diabetes. Such criteria include, for example, low (less than 10 or 1 percentile controls) first phase insulin secretion in an intravenous glucose tolerance test (IVGTT) or persistence of high antibody titer to islet antigens such as insulin, GAD65 and / or ICA512.

Patients without clinical diabetes who may benefit from prophylactic treatment may be identified by any predictive criteria accepted in the art. Diabetes without clinical diabetes can be expected to develop diabetes in the following years (1-5 years) according to the following criteria: (i) family history - first degree relatives are automatically a high risk group unless they have a protective HLA allele; genetic makeup - i.e. the presence or absence of an HLA allele that is associated with a high risk of developing diabetes (e.g. DR3 / 4; DQ8); (iii) the presence or absence of a high titer of autoantibodies in the blood against any of the antigens: insulin, GAD65 and / or ICA512; and (iv) intravenous glucose tolerance test (IVGTT): a low first phase insulin secretion, usually defined as the tenth or first percentile of normal controls, typically prevents the onset of type I diabetes by 1-5 years. Normally, several of the above criteria are taken into account. For example, the risk of developing diabetes in the next 5 years for relatives of first degree diabetes is: 100% for relatives with all 3 autoantibodies listed above; 44% for relatives with two antibodies; 15% for relatives with one antibody; and 0.5% for relatives without antibody. Among the 50 first degree relatives to a patient with type 1 diabetes after developing diabetes, one or more of the above autoantibodies were present in 49/50.

Effective treatment of diabetes can be determined in several different ways. Meeting any of the following criteria, or other criteria accepted in the art, demonstrates effective treatment. The criteria include, without limitation, delaying the onset of clinical hyperglycemia, reducing the frequency of hyperglycemic conditions, and / or prolonging the presence of normal C-peptide levels in patients' blood.

The peptide analogs of the present invention may be administered either alone or in the form of a pharmaceutical composition. Briefly, the pharmaceutical compositions of the present invention may comprise one or more peptide analogs of the present invention, in combination with one or more pharmaceutically or physiologically acceptable carriers, diluents, or excipients. Such compositions may include buffers such as neutral buffered saline, phosphate buffered saline, and the like, carbohydrates such as glucose, mannose, sucrose or dextrans, mannitol, proteins, polypeptides or amino acids such as glycine, antioxidants, chelating agents such as EDTA or glutathione, adjuvants (e.g., aluminum hydroxide) and preservatives. In addition, the pharmaceutical compositions of the present invention may also contain one or more other active ingredients, such as sustained delivery systems or other immunopotentiating agents.

The compositions of the present invention may be in a form suitable for oral, nasal, venous, intracranial, intraperitoneal, subcutaneous or inramuscular administration. Further, the compositions of the present invention may be administered as part of a sustained release implant. In yet other embodiments, the compositions of the present invention may be prepared in the form of lyophilized formulations, using suitable additives suitable to achieve stability of the lyophilized form and / or after rehydration.

The pharmaceutical compositions of the present invention may be administered in a manner suitable for the disease being treated. The amount and frequency of administration will be determined by factors such as the general condition of the patient and the type and severity of the patient's disease. In one embodiment of the present invention, the peptide analog may be administered at a dose ranging from 0.1 to 100 mg / kg, although a suitable dosage may be determined in clinical trials. Patients may be monitored for therapeutic efficacy by delaying the progression of clinical diabetes and reducing insulin use to maintain normoglycaemia.

The following examples are illustrative only and do not limit the scope of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

Example 1: Preparation of peptides

This example describes the synthesis of representative peptide analogs.

Peptides were synthesized by solid phase synthesis on a peptide synthesizer (Beckman model 990). Peptides with an amidated carboxyl terminus were prepared with p-methylbenzhydrylamine resin (MBHA resin); for peptides with a free carboxyl terminus, a Merrifield resin with a suitably protected amino acid was used. Both resins were obtained from Bach Fine Chemicals (Torrance, CA). Derivatized amino acids (Bachem Fine Chemicals) used in the synthesis were L-configurations, unless otherwise indicated, and had an N-alpha-amino function protected exclusively by a t-butyloxycarbonyl group. The side chain functional groups were protected as follows: benzyl for serine and threonine; cyclohexyl for glutamic and aspartic acids; tosyl for histidine and arginine; 2-chlorobenzyloxycarbonyl for lysine and 2-bromobenzyloxycarbonyl for tyrosine. Coupling of the carboxy-terminal amino acid to the MBHA resin was accomplished with dicyclohexylcarbodiimide and additional amino acids were coupled with dicyclohexylcarbodiimide according to Ling et al. (Proc. Natl. Acad. Sci. USA 81: 4302, 1984). After incorporation of the last amino acid, the t-butyloxycarbonyl protecting group was removed and the peptide-resin conjugate was reacted with a mixture of 14 mL of hydrofluoric acid (HF), 1.4 mL of anisole and 0.28 mL of methyl ethyl sulfide per gram of resin conjugate at -20 ° C for 0 , 5 hours and at 0 ° C for 0.5 hours. HF was removed in vacuo at 0 ° C and the resulting peptide-resin mixture was washed twice with diethyl ether and twice with chloroform and diethyl ether alternately. The peptide was extracted five times with 2 M acetic acid and the extract was lyophilized.

The lyophilized material was first purified on a Sephadex G-25 fine column (Pharmacia-LKB, Piscataway, NJ) developed in 30% acetic acid to remove truncated fragments and inorganic salts (Ling et al., 1984). The peptides were further purified by CM-32 carboxymethylcellulose cation exchange chromatography (Ling et al., 1984). Final purification was accomplished by partition chromatography on Sephadex G-25 fine (Ling et al., 1984). Alternatively, the crude peptide can be purified by preparative HPLC on a Biotage KP-100 gradient HPLC system. Synthetic material was characterized by amino acid analysis, mass spectrometry and reverse phase HPLC.

Example 2: Long-term T-cell lines

This example illustrates the preparation of long-term insulin-specific T-cell lines from NOD mice.

• β 9 · · *

Insulin-specific NOD T cell lines were prepared by culturing lymphocytes isolated from islet infiltrating populations by in vitro stimulation with porcine insulin at a dose of 25 µg / ml and irradiated islet NOD cells as antigen-presenting cells and cytokines. To obtain infiltrating lymphocytes, the following procedures were performed (see Wegmann et al., Eur. J. Immunol. 24: 1853, 1994): pancreas from NOD mice were digested with collagenase and individual islets were isolated manually. Infiltrating lymphocytes were then obtained by gently digesting the islets with trypsin. Insulin-specific T-cell lines or clones were propagated by serial stimulation in the presence of NOD cells of spleen, porcine insulin and lymphokines. Clones were obtained by limiting dilution of the B chain-specific T cells (9-23) in the presence of antigen-presenting cells and porcine insulin at a dose of 25 µg / ml. Wells with a growing cell population after limiting dilution were expanded in an appropriate medium and after one growth cycle were tested for reactivity to the insulin B chain peptide (9-23) by assessing the proliferative response.

Example 3: Effect of peptide analogs on the proliferation of insulin-specific NOD T cell clones

This example illustrates the effect of representative peptide analogs on T cell proliferation.

Insulin B chain (9-23) specific (NOD) T cell clones (SEQ ID NO: 2) were isolated from infiltrating lymphocytes as described in Example 2. Peptide analogs with one alanine substitution were prepared as described in Example 1. The effect of each analog on T cell proliferation was then assessed using a flat bottom 96-well assay (see Daniel et al., Eur.

4 · · 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 «· · · · · ··· *

J. Immunol. 25: 1056 (1995). Briefly, 25,000 T lymphocytes with 10,000,000 irradiated spleen NOD cells were cultured in triplicate in the presence of 50 µg / ml 9-23 insulin B chain peptide or in the presence of any alanine substituted peptide listed below. The plates were incubated for a total of 72 hours in a 7% carbon dioxide atmosphere with a pulse of 1 µCi / well of tritiated thymidine for the last 6-8 hours of culture. Cells were harvested on a glass fiber filter and associated radioactivity was counted in a liquid scintillation counter. Results are expressed as mean counts per minute for three wells.

Data obtained from five separate T cell clones show either lack of proliferation or significant reduction in proliferation (relative to native 9-23 insulin B chain peptide; SEQ ID NO: 2) in the presence of the following alanine-substituted analogs: A12, A13, A15 , A16, A17, and A18. These data are shown in Tables 1 and 2.

«« ·

Table 1: Response (cpm) of insulin-specific NOD T cell clones

T cell clone

M P odifikovaná Natural _oz -i _ce rest Substitution PD6-4.3 PD12-2.40 9 WITH AND 12861 42234 10 H AND 12507 1409 11 L AND 14148 2594 12 IN AND 8292 671 13 E AND 142 519 14 AND none 15 Dec L AND 161 1422 16 Y AND 98 539 17 L AND 553 19321 18 IN AND 234 44785 19 Dec C AND 7678 34212 20 May G AND 2440 38685 21 E AND 91 39087 22nd R AND 6555 51722 23 G AND 14304 75441 No antigen Native 9-23 163 682 10463 32221

• ·

Table 2: Response (cpm) of insulin specific NOD T cell clones

Modified position Natural residue Substitution T cell clone PD12-4.4 PD 12-4.29 PD12-4.34 9 with AND 1000 18422 259 10 H AND 823 15484 356. 11 L AND 474 18416 190 12 IN AND 1129 15041 194 13 E AND 373 891 179 14 AND none «5 15 Dec L AND 675 809 191 16 Y AND 779 63 £ 202 17 L AND 332 1460 4360 18 IN AND 225 1193 721 19 Dec C AND 4295 6054 689 20 May G AND 1323 13736 466 21 E AND 7900 4904 773 22nd R AND 1313 12635 1555 23 G AND 3228 18422 791 3ez antigens 350 789 231 Native .9-23 i. ------------ 10000 14820 3614

Table 3 shows the response of four different T cell clones obtained from NOD mice to the double alanine-substituted peptide analog A16, A19 (NBI-6024; 16Y> A / 19C> A). NOD T cell clones were incubated in the presence of 50 μΜ of either native B chain peptide (9-23) or NBI-6024. The data in Table 3 represents the mean of three samples + standard deviation from the mean. In Table 3, S.I. (stimulation index) = proliferation (cpm) in the presence of peptide / proliferation (cpm) in the medium alone. These data show a significant response when cells are cultured with natural B chain peptide (9-23) but little or no proliferation against medium alone (baseline) in the presence of NBI-6024.

Table 3: Proliferative response of murine T-lymphocyte clones specific for insulin B chain (9-23) at 50 .mu.M of the B chain (9-23) or an analogue and ^{X 6 'X 9} (NBI-6024)

Clone of Tlymphocytes Try C. AND Only medium '-'mj nsulin B chain (9-23) NBI-6024 (A ¹⁶ · ¹⁹⁾ Average cpm + sem S.I. Diameter · cpm + sem S.I. PD12-2.35 1 688 ± 227 120.886 ± 7.171 175.7 841 ± 88 1.22 2 493 ± 20 100.521 ± 1.581 203.89 452 ± 179 0.91 PD 12-2.40 1 170 ± 8 16.730 ± 3.835 98.4 272 ± 34. 1.16 2 1.834 ± 638 176.359 ± 36.306 96.16 1,863 ± 451 1.01 PD12-4.1 1 215 ± 17 28.593 ± 4664 132.99 566 ± 30 2.63 PD12-4.9 1 9.111 ± 1.889 45.541 ± 5.222 4.99 12,313 ± 1,372 1.35 2 7.202 ± 2.773 65.624 ± 4.979 9.1 6.171 ± 725 0.85

Example 4: Antagonizing the T cell proliferation assay

This example illustrates the inhibition of the response of murine T cell clones specific for peptide (9-23) of the insulin B chain by representative peptide analogs.

B chain peptide analogs (9-23) containing alanine substitutions at residue 12, 13, 15 or 16 or double substitutions at positions 16 and 19 (A ^xe '^x9; NBI-6024) were prepared as described in Example 1. Antagonist action T cells were detected by assessing the ability of peptide analogs to inhibit natural B chain-induced T cell proliferation (9-23) (SEQ ID NO: 2). In this assay, antigen-presenting cells were first irradiated and then incubated with the competitive peptide analog and the peptide (9-23) of the natural B-chain. T lymphocytes were then added to the culture. Different concentrations of potential peptide analogs were included in cultures that were incubated for a total of 4 days. After this incubation period, each culture was exposed to a pulse of 1 µCi [ ³ H] -thymidine for an additional 12-18 hours. Cells were then harvested on glass fiber filters and counted as described above. Mean cpm and standard deviation were calculated from data obtained in three experiments. The results presented in Figure 2 show that peptide analogs containing alanine substitutions at residue 12, 13 or 16 are capable of attenuating the response of insulin B chain-specific pathogenic T cells (9-23).

The ability of the double substituted peptide to inhibit insulin-dependent T cell proliferation is shown in Table 4 and Figure 3. In Table 4, the control peptide NBI-5096 is an unrelated peptide of myelin basic protein. Percent inhibition was calculated as: (1-experimental cpm / cpm for insulin peptide) x 100%.

• ·

Table 4: Inhibition of insulin B chain peptide (9-23) response in two mouse NOD T cell clones by peptide analogue

Clone Conditions CPM ± SEM % Inhibition PD12-2.35 only medium 213 ± 17 B (9-23) 5μΜ 7.840 ± 528 B (9-23) 5µΜ + 10µΜ NBI-6024 4.441 ± 626 43.0 B (9-23) 5µΜ + 50µΜ NBI-6024 1.389 ± 218 82.0 B (9-23) 5µΜ + 10µΜ NBI-5096 10.089 ± 1.113 ON B (9-23) 5µΜ + 50µΜ NBI-5096 10.125 ± 887 ON PD12-2.40 only medium 305 ± 13 B (9-23) 5μΜ 9.149 ± 1.062 B (9-23) 5µΜ + 10µΜ NBI-6024 6.379 ± 1.485 30.0 B (9-23) 5µΜ + 50µΜ NBI-6024 4.305 ± 941 52.9 B (9-23) 5µΜ + 10µΜ NBI-5096 12.336 ± 1.556 ON B (9-23) 5µΜ + 50µΜ NBI-5096 17.988 ± 584 ON

N / A - not applicable because no inhibition was observed

The ability of NBI-6024 to block B chain peptide (9-23) induced stimulation of T cell clones derived from NOD mice suggests that alterations at positions 16 and 19 of the native insulin B chain peptide (9-23) do not alter analog recognition by pathogenic T cells. In addition, these results suggest that the analog also binds to MHC with sufficient affinity for recognition by insulin B chain-specific T cells (9-23).

Example 5: Effect of peptide analogs on the proliferation of T-cell lines and clones from diabetic patients

This example demonstrates the lack of stimulation of T-cell lines and clones obtained from diabetic patients by representative peptide analogs.

Peptide analogues of B chain (9-23) containing alanine substitutions at zybtcích 13, 515, 16 or 17 or the doubly substituted alanine analog and ^{3-6 '3-9} (NBI-6024) were prepared as described under Example 1. Lines T- lymphocytes from diabetic paceint were prepared by isolating lymphocytes from paceint blood so that the blood was processed by density separation. The isolated lymphocytes were then cultured in the presence of peptide (9-23) insulin B chain (10 µM) and recombinant human IL-2 in the presence of 5-10% autologous serum and irradiated peripheral blood autologous lymphocytes in the culture medium. After four days or five days, the cells were harvested and the cycle repeated 2 to 3 times.

Proliferation of the T cell line, in response to peptide (9-23) of the natural B chain (SEQ ID NO: 2) or peptide analogs, was measured by culturing 25000 to 100000 T cells in the presence of 50,000-200,000 irradiated autologous PBLs and various peptide concentrations (9-23) insulin B chain or peptide analog in three cultures. After 4-5 days of culture, including the last 18 hours with radiolabeled thymidine, cells were harvested and associated radioactivity was measured in a liquid scintillation counter. Results are reported as the average number of pulses per minute for each peptide analogue tested.

The results shown in Figures 4-7 show that T cell clones and lines that proliferate in response to peptide (9-23) of the natural insulin B chain (SEQ ID NO: 2) are not stimulated by peptide analogs. The results from these and other patients are summarized in Table 5.

• ·

Table 5: Proliferative response of a PBL patient to natural insulin peptide or NBI-6024 analogue

Patient number Patient ID ¹ Stimulation Index * Insulin B (9-22) [50 μΜ] NBI-6024 [50 μΜ] l 100 ALIGN! 9.9 0.9 2 200 5.3 1.2 O 3 400 7.8 1.0 4 500 2.1 0.9 5 600 5.8 1.6 · 6 700 3.2 1.5 7 900 2.6 0.9 8 1100 3.7 0.8

* Stimulation index = cpm with antigen / cpm with medium alone (no antigen)

The results clearly show that cells from diabetic patients that are reactive to the insulin B chain peptide (9-23) do not respond to the altered peptide ligand NBI-6024, which has substitutions at positions 16 and 19. We also found that APL NBI- 6024 binds with similar affinity to DQ8 antigens. Therefore, the absence of T-cell stimulation in diabetic patients NBI-6024 is not due to any inactivity of the peptide with the presenting MHC molecules, but rather to altered recognition by B-chain specific T cells (9-23).

Example 6: Reduction of the incidence of diabetes in NOD mice

This example illustrates the ability of representative peptide analogs to prevent the development of diabetes in NOD mice.

• · «· * * * * * * * *

NOD mice develop spontaneously diabetes at about 3 months of age (Makino et al., Current Topics in Clinical and Experimental Aspects of Diabetes Mellitus, Sakamoto et al., Eds., Pp. 25-32 (Elsevier, Amsterdam, 1985) ). The onset of the disease is preceded by cellular infiltration of T lymphocytes into the pancreas, which occurs as early as 1 month of age. Soluble B chain peptide analogs (9-23) containing alanine substitutions at residues 12, 13, or 16 were administered to NOD mice at weekly intervals. At each dose, 400 of each peptide was administered to each of ten animals. After 9 cycles, the percentage of mice that developed diabetes was determined in each group by measuring the blood glucose concentration at weekly intervals with a glucometer. Blood glucose levels above 200 mg / dL in the following two measurements were considered as an indicator of clinical diabetes.

As shown in Figure 8, treatment with each of the alanine-substituted analogs resulted in a significant reduction in the incidence of diabetes. Data for the A13 substituted analog is also shown in Figure 9.

In another experiment, the B chain (9-23), the A13 substituted analog, or neurotensin (as a control) were administered subcutaneously to NOD mice at weekly intervals. At each dose, 400 of each peptide was administered to each of ten animals. After 13 cycles, the percentage of mice that developed diabetes was determined in each group using the method described above. As shown in Figure 10, the B chain peptide (9-23) reduced the incidence of diabetes. This reduction was most pronounced for the A13 substituted analog.

To determine the ability of the double substituted peptides _A 16, 19 (NBI-6024) to control the development of diabetes in NOD mice, the peptide was administered to animals of low age. Female mice (n = 9, age • about 4 weeks) were treated with subcutaneous injections of 20 mg / kg (40 µg / mouse) of NBI-6024 for 12 weeks and then every other week to week 35. From age 9-10 weeks mice were monitored weekly for hyperglycemia by measuring blood glucose concentration. A group of 10 untreated mice was used as a control.

As can be seen, NBI-6024 treatment significantly reduced the incidence of diabetes by approximately 60-70% compared to the untreated group (p <

0.004).

These observations were then confirmed and expanded in the second experiment. In this experiment, animals (n = 13-15) were treated with either NBI-6024, or an unrelated peptide, neurotensin, NBI-6201, as described above. Another group (n = 8) was untreated. As shown in Fig. 12, treatment with 20 mg / kg of the altered peptide NBI-6024 resulted in a reduction in the incidence of diabetes compared to the neurotensin treated or untreated group.

These results indicate that the altered peptide ligand NBI-6024, derived from the insulin B chain peptide (9-23), is capable of preventing spontaneous diabetic development in high-risk animals. It is likely that T lymphocytes that recognize other pancreatic antigens are present in these animals, but are also likely to be regulated by insulin APL. Administration was probably performed at the same time as autoreactive lymphocytes begin to infiltrate the pancreas and initiate a destructive process. These results give hope that early intervention using these APLs may be effective in delaying or preventing the onset of type I diabetes in humans.

Example 7: Immunogenicity of Representative Peptide Analogs

This example illustrates the immunogenicity of representative peptide analogs in NOD mice.

Groups of 3-4 mice were immunized with 100-400 µg peptide analogs subcutaneously in mannitol acetate buffer three times within 10-15 days. After the last immunization, lymph node cells and / or spleen cells were used in a proliferation assay in which they were cultured with different concentrations of the immunization peptide for 3-4 days. Triitium-labeled thymidine was added to the culture for the last 18 hours. Cells were harvested and read on a scintillation counter and the response is expressed as cpm + sem. These results, shown in Figures 13-16, show that these representative peptide analogs can bind to murine MHC molecules and can be recognized by the respective T cells.

The ability of the double substituted peptide NBI-6024 ( ^1616.19 ) to induce a cellular immune response in NOD mice was then determined. Two female NOD mice were immunized with 10 mg / kg NBI-6024 either as an aqueous suspension or - as a control - as an emulsion with complete Freund's adjuvant (CFA). On day 8, three days after the last injection, mice were sacrificed, spleen cells and inguinal lymph nodes were removed and collected, and a single cell suspension was prepared. Cells were cultured in the presence of various concentrations (0-25 µM) of NBI-6024. Lymphoid cell proliferation capacity upon stimulation with NBI-6024 was measured in vitro by [ ³ H] -thymidine incorporation.

The results are shown in Table 6, in which the response is expressed as mean cpm +. here from three cultivations. Lymph node cells isolated from mice immunized with the analog in CFA had a strong proliferative response to dose-dependent stimulation with the immunizing analog (Table 6). These results indicate that the alterations made to the insulin B chain sequence (9-23) at positions 16 and 19 did not affect the ability of the peptide to bind to the MHC haplotype molecule associated with the disease in NOD mice and - more significantly 32

that T cell recognition was not impaired.

Table 6: Proliferative response of lymph node cells to NBI-6024 in NOD mice immunized with NBI-6024 in CFA

NBI-6024 Proliferative response (μΜ) (CPM ± SEM) 0 2.445 ± 137 1 140.061 ± 7.289 5 187.711 ± 2.548 25 218.149 ± 4.462

Furthermore, both spleen cells and lymph nodes isolated from mice receiving a soluble peptide showed a strong proliferative response to APL upon stimulation with NBI-6024 in vitro (Table 7 and Figures 17A and 17B). More importantly, it was found that NBI-6204-reactive lymphocytes from mice immunized with a soluble peptide also responded to the insulin B chain (9-23). This cross-reactivity was not observed for the peptide emulsified in CFA. This ability of soluble peptides to induce cross-reactivity may be useful in the control of diabetes, as it may assist the mobilization of protective NBI-6024 specific T cells into pathogenic target tissue.

• ·

Table 7: T-cell proliferative response from NOD mice immunized with soluble NBI-6024 to NBI-6024 or natural insulin B chain (9-23)

ΝΒΪ-6024 induced cell line Conc. peptide [μΜ] CPM ± SEM NBI-6024 Insulin B (9-23) CPM ± SEM Mouse _; # 1 0 19.130 ± 2191 19.130 ± 2191 1 48.319 ± 191 » 16.870 ± 4469 5 160.673 ± 2269 21.292 ± 3216 25 268.005 ± 11198 33.317 ± 3619 I Mouse # 2 0 21.588 ± 2326 21.588 ± 2326 1 54.519 ± 5666 17.262 ± 602 5 126.123 ± 13851 19.648 ± 2169 25 202.707 ± 8125 30.612 ± 3557 Mouse # 3 0 24.006 ± 2803 24.006 ± 2803 1 64.239 ± 9493 25.825 ± 3841 5 140.836 ± 11778 58.567 ± 2737 25 240.278 ± 15015 113.366 ± 515

To determine the type of T lymphocytes produced after administration of soluble NBI-6024, supernatants of immune lymphoid cell cultures were removed 48 hours after the start of the culture, and various cytokines were measured using standard ELISA technology. Surprisingly, T-lymphocytes from mice immunized with soluble NBI-6024 were found to produce Th2 cytokines, interleukin-4 (Fig. 18) and interleukin-5 (Table 8), and not Th1 cytokine interleukin-2. In Table 8, values are given in pg / ml as the average of three experiments + SEM. As a control, NBI-6024 in CFA emulsion was used, which induced •

expected Th1-profile of cytokines (IL-2) in immune T cells after in vitro stimulation.

Table 8: Cytokine response of soluble NBI-6024-induced T cells to NBI-6024

NBI-6024 [mM] Cytokine (pg / ml) IL-2 IL-4 IL-5 0 <15 pg / ml 134 ± 0 1.814 ± 332 1 <15 pg / ml 684 ± 92 9.999 ± 503 5 <15 pg / ml 1.653 ± 51 23.496 ± 684 25 <15 pg / ml 2.102 ± 85 28.062 ± 141

The possibility of inducing Th-2-like lymphocytes by subcutaneous administration of soluble NBI-6024 is a desirable property since such cells are associated with recovery in diabetes and other organ-specific autoimmune diseases (Sarventick, J. Exp. Med. 184: 1597-1600, 1996; Shaw et al., 1997; Balasa et al., J. Exp. Med. 186: 385-391, 1997). These Th2-derived lymphocyte-derived cytokines have potent anti-inflammatory effects that suppress the development of autoreactive Th1 lymphocytes producing pro-inflammatory cytokines that cause disease.

From the above description, it is clear that although specific embodiments of the invention have been described, various modifications thereof exist that do not depart from the scope and spirit of the present invention. Accordingly, the invention is limited only by the appended claims.

• ·

Sequence Listing <110> Neurocrine Biosciences, Inc.

<120> Method for the treatment of diabetes mellitus using peptide analogues of insulin <130> 690068.448PC <140> PCT <141> 23.2.1999 <160> 2 <170> Patentln ver. 2.0 <210> 1 <211> 30 <212> PRT <213> Homo sapiens <220>

<223> human insulin B chain <400> 1

Phe Val Asn Gin His Leu Cys Gly Ser

Leu Val Cys Gly Arg Gly Phe Phe Thr Pro Lys Thr 20 25 30 <210> 2 <211> 15 <212> PRT <213> Homo sapiens

<223> Human insulin B chain residues 9-23 <400> 2

Ser His Leu Val Clu Glu Ala Leu Tyr

Claims (27)

A peptide analog comprising residues 9-23 of human insulin B chain that differs from the sequence of residues 9-23 of natural human B chain insulin by substitution at 1-4 amino acid positions, wherein at least one substitution is present in a residue selected from the group consisting of residues 12, 13, 15 and 16. The peptide analogue of claim 1, having a sequence that differs from the natural human B chain of insulin at two amino acid residues. The peptide analogue of claim 1 having a sequence that differs from the natural human B chain of insulin at the three amino acid residues. The peptide analog of claim 1, wherein the amino acid substitution is present at residue 19. The peptide analogue of claim 1, wherein at least one amino acid substitution is non-conservative. The peptide analog of claim 1, wherein residue 12 is substituted. The peptide analogue of claim 6, wherein residue 12 is an alanine residue. The peptide analogue of claim 1, wherein residue 13 is substituted. The peptide analog of claim 8, wherein residue 13 is an alanine residue. The peptide analog of claim 1, wherein residue 15 is substituted. The peptide analogue of claim 10, wherein residue 15 is an alanine residue. The peptide analog of claim 1, wherein residue 16 is substituted. The peptide analogue of claim 12, wherein residue 16 is an alanine residue. The peptide analogue of any one of claims 6-13, wherein residue 19 is substituted. The peptide analog of claim 14, wherein residue 19 is an alanine residue. The peptide analogue of claim 1, further comprising a residue 24 of human insulin B chain. The peptide analog of claim 1, which contains no more than 18 residues of human insulin B chain. The peptide analog of claim 1, which contains no more than 16 residues of human insulin B chain. The peptide analogue of claim 1, which does not contain more than 15 human insulin B chain residues. 20. A peptide analogue consisting essentially of residues 9 to 9 of a peptide analogue. · • · 23 of a human insulin B chain that differs from the sequence of residues 9-23 of the native human insulin B chain by substitutions at 1-4 amino acid positions wherein at least one substitution is present in a residue selected from the group consisting of residues 12, 13, 15, and 16 . 21. A peptide analogue consisting essentially of residues 9 through 24 of a human insulin B chain that differs from the sequence of residues 9-23 of the native human insulin B chain by substitutions at 1-4 amino acid positions wherein at least one substitution is present in a residue selected from the group consisting of residues 12, 13, 15, and 16 . 22. A pharmaceutical composition comprising the peptide analogue of any one of claims 1-19 in combination with a pharmaceutically acceptable carrier or diluent. 23. A method for inhibiting the development of diabetes comprising administering to a patient a therapeutically effective amount of a pharmaceutical composition of claim 22. 24. A method for treating diabetes comprising administering to a patient a therapeutically effective amount of the pharmaceutical composition of claim 22. 25. A peptide analog comprising residues 9-23 of human insulin B chain that differs from the sequence of residues 9-23 of natural human B chain insulin by substitutions at residues 16 and 19. 26. A pharmaceutical composition comprising the peptide analog of claim 25 in combination with a pharmaceutically acceptable carrier or diluent. · * 00 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 27. A method for inhibiting diabetes, comprising administering to a patient a therapeutically effective amount of a pharmaceutical composition of claim 26.