CZ20003065A3 - Insulin peptide analogues and their use in the treatment of diabetes mellitus - Google Patents

Abstract

Peptide analogs of the B chain of insulin, which are derived from peptides containing residues (9-23) of the natural sequence of the B chain. The analogs are altered relative to the natural sequence at positions (12, 13, 15) and/or (16), and may also be altered at position (19) and/or at other positions. Pharmaceutical compositions containing these peptide analogs. The peptide analogs are useful in the treatment or inhibition of the development of diabetes.

Description

Field of the invention

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

State of the art

Insulin-dependent diabetes mellitus (IDDM) is an organ-specific autoimmune disease affecting nearly one million patients of all ages in the United States. The disease is characterized by widespread destruction of insulin-producing beta cells in the pancreatic islets and dysregulation of glucose metabolism, leading to clinical diabetes. The defining feature of IDDM is lymphocytic infiltration of the islets. Of the infiltrating cells, T lymphocytes appear to be one of the major mediators of autoimmune destruction.

Type 1 diabetes is further characterized by elevated levels 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 predictive marker of developing diabetes in patients with predisposing genetic (HLA) haplotypes.

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

chain consisting of 21 amino acid residues and a B chain consisting of 30 residues. Although insulin administration provides significant benefits to patients suffering from diabetes, the short serum half-life of insulin makes it difficult to maintain the correct dosage. Insulin administration can lead to the development of various hypoglycemic side effects and the formation of neutralizing antibodies.

Due to the problems associated with existing treatments for diabetes, there is a continuing need for improved treatments that are more effective and that do not suffer from 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 the benefits associated with this effect.

The essence of the invention

The present invention provides compounds and methods for the treatment and prevention of diabetes. In some aspects, the present invention provides peptide analogs comprising residues 9 to 23 of the B chain of human insulin (SEQ ID NO: 2), wherein the sequence of such peptide analogs differs from the sequence of residues 9 to 23 of the native human B chain of insulin by substitutions in 1 to 4 amino acids. 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 additional substitutions at other residues. In some preferred embodiments, such substitutions may be present at two or three amino acid residues within residues 9 to 23 of the B chain of insulin. Substitutions may also be present at residue 19. The substitutions are preferably non-conservative, and analogs in which residues 12, 13, 15, 16 and/or 19 are changed (e.g., to alanine) are preferred. Also preferred are analogs that further comprise residue 24 of the B chain of human insulin. 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 B chain of human insulin.

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

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

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

These and other aspects of the invention will become apparent from the following detailed description and the accompanying drawings. Various references are also included in the text which describe in detail certain methods or means. All of these references are incorporated herein by reference in their entirety.

Description of images in the attached drawings

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

Fig. 2 is a graph showing the proliferative response (measured in cpm) of a NOD mouse T lymphocyte clone to (9-23) native insulin B chain peptide in the presence of varying amounts of representative peptide analogs in which various residues are substituted with alanine as indicated.

Fig. 3 is a graph showing the proliferative response (measured in cpm) of a NOD mouse T cell clone to the (9-23) peptide of the native B chain of insulin 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). The proliferative response in the presence of an unrelated control peptide derived from myelin basic protein (NBI-5096, circles) is also shown for comparison. The response is measured as the mean CPM + SEM of three cultures.

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

After three cycles of restimulation with insulin B chain (9-23), 1 x 10 ⁵ T cells and 7 x 10 ⁴ irradiated autologous PBMCs were added to each well of a round-bottom 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 alone. On day 4, cells were pulsed with ³ H-thymidine and cultured for an additional 18 hours. Cultures were then harvested, counted using a liquid scintillation counter, and data were expressed as mean counts per minute (cpm) for replicates + standard deviation (SEM).

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

Fig. 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 B chain peptide analogs (9-23) containing alanine substitutions at residue 12 (open triangles), 13 (squares), or 16 (filled triangles). All mice treated with the control peptide, neurotensin (circles), developed diabetes.

Fig. 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.

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

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

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

Fig. 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 35 weeks of age. The control group (n=10) consisted of untreated animals. Mice with blood glucose concentrations greater than 200 mg/dl on two subsequent measurements were considered diabetic mice. Data are expressed as a percentage of non-diabetic mice over the course of the 35-week study. The log-rank test was used to assess whether the results for the two treatment groups were statistically significantly different. The percentage of NOD mice that were diabetic after treatment with NBI-6024 (A ¹⁶ - ¹ ® ) is shown for different times by squares, and the percentage of mice that were diabetic in the control group is shown by circles.

Fig. 12 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=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 until 35 weeks of age. Another control group (n=8) consisted of untreated animals. Mice with blood glucose concentrations greater than 200 mg/dl on two subsequent measurements were considered diabetic mice. Data are expressed as a percentage of non-diabetic mice over the course of the 35-week study. The log-rank test was used to assess whether the results for the two treatment groups were statistically significantly different. The percentage of NOD mice that were diabetic after treatment with NBI-6024 (A ¹⁶ ' ³ - ⁹ ) is shown for different times by squares, the percentage of diabetic mice after treatment with neurotensin is shown by triangles, and the percentage of mice that were diabetic in the control group is shown by circles.

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

Fig. 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 using various concentrations of the immunizing peptide as a stimulator. The proliferative response was determined by the amount of radiolabeled thymidine incorporated into the cells (indicated in the graph as the average count per minute (cpm) of three culture wells), which was determined in a liquid scintillation counter after the completion of the culture period.

• · • ·

Fig. 15A-15D are graphs illustrating the immunogenicity of representative double-substituted insulin B chain peptide analogs (9-23). The following peptides were tested for their 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 (in beats per minute) of the descending lymph node cells in response to the immunizing analog as well as to the native insulin B chain peptide (9-23) is shown.

Fig. 16 is a graph showing the ability of a series of triply substituted peptides to elicit a T-lymphocyte proliferative response in NOD mice. Mice were immunized with individual representative peptide analogs containing the following substitution combinations: A12,13,19; A12,15,19; A12,16,19; A13,15,19; or A13,16,19; A15,16,19. Lymph node cells were then used in a proliferation assay and the response to each immunizing peptide at various concentrations is shown.

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

The response is expressed as the mean cpm + SEM of three cultures.

Fig. 18 is a histogram showing a comparison of cytokines produced by immune cells following induction with A ^x6 ' ^x9 (NBI-6024) peptide in the presence or absence of an adjuvant.

• · · ·

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

Before describing the present invention, it is useful to provide definitions of some terms for a better understanding thereof.

The term insulin B chain refers to a polypeptide of 30 amino acids forming 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 peptide analogs of the insulin B chain 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 the sequence between the analog and the native B chain. In a peptide analog, at least one difference in the amino acid sequence occurs at residue 12,

13, 15 and/or 16. Furthermore, 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 a larger number of substitutions (e.g. 5 or 6) is also 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, more preferably up to a total of 16 or 18 residues of the peptide analog. In a preferred embodiment, the peptide analog also contains residue 24 of the insulin B chain. Sequences not derived from

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

Unless otherwise stated, the amino acids listed are in the L-form. An L-amino acid residue in a natural peptide sequence may be altered to any of the 20 L-amino acids commonly found in proteins, to any of the corresponding D-amino acids, to 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 modified by chemical means such as methylation (e.g., -methylvaline); amidation of the C-terminal amino acid with an alkylamine such as ethylamine, ethanolamine, or ethylenediamine; and/or acylation or methylation of a side chain group of the amino acid (e.g., acylation of the epsilon-amino group of lysine).

The terms residue 12”, residue 13, residue 15, residue 16 and residue 19 (also referred to as position 12, position 13, position 15, position 16 and position 19, respectively) refer to amino acids 12, 13, 15, 16 and 19 of the insulin B chain, as shown in Fig. 1. More specifically, the numbering system for these residues refers to the amino acid positions in the native human protein, regardless of the length of the peptide analog or the amino acid position in the analog. Peptide analogs having alanine substitutions at residues 12, 13, 15 or 16 are referred to as A12, A13, A15 or A16, respectively.

Insulin B chain peptide analogs

As mentioned above, the present invention provides peptide analogs comprising at least residues 9-23 of the human insulin B chain and comprising a change of the native 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 further comprise 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 is typically 15-30 residues in length, preferably 15-18 residues, more preferably 15-16 residues. Particularly preferred peptide analogs contain 15 amino acids derived from the insulin B chain.

As noted above, peptide analogs containing any amino acid changes at the positions listed above are within the scope of the present invention. Preferred peptide analogs contain non-conservative substitutions (i.e., substitutions for amino acids having a different charge, polarity, hydrophobicity, and/or size). Particularly preferred analogs contain substitutions 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 solid phase peptide synthesis techniques, which involve coupling each protected amino acid residue to a resin support, preferably a 4-methyl-benzhydrylamine resin, via activation with dicyclohexylcarbodiimide, to yield a peptide with a C-terminal amide. Alternatively, a chloromethyl resin (Merrifield resin) can be used to yield peptides with a free carboxylic acid at the C-terminus. Side chain functional groups can 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 reaction with trifluoroacetic acid followed by neutralization with di-isopropylethylamine. Another protected residue is then coupled to the free amino group, resulting in peptide chain extension. After the last residue is coupled, the protected peptide-resin is treated with hydrogen fluoride to cleave the peptides from the resin and to 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) not stimulate NOD mouse T cell clones specific for the native (9-23) insulin B chain peptide (SEQ ID NO: 2), nor should they stimulate such clones at a level that is lower than the level stimulated by the native peptide; (b) should not stimulate human T cells from patients specific for the insulin B chain (9-23); (c) should not be immunogenic in NOD mice; (d.) should reduce the incidence of diabetes in NOD mice; and (e) should inhibit the response of T cell clones specific for the native insulin B chain peptide (9-23). Therefore, potential peptide analogs can be screened for their ability to treat diabetes using assays measuring T cell proliferation, immunogenicity in NOD mice, and effect on disease incidence in NOD mice. Some representative assays for

The methods used 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 can first be tested for their ability to stimulate T cells specific for the peptide (9-23) of the natural insulin B chain (SEQ ID NO: 2) (from clonal cell lines or isolated from a patient). Such tests can be performed using a direct proliferation assay in which T cell lines reactive with the natural B chain (9-23) or T cells from the patient are used as target cells. T cell lines can be obtained, using well-known techniques, from lymph nodes obtained from rats injected with B chain (9-23). Lymph node cells can be isolated and cultured for 5 to 8 days with B chain (9-23) and IL-2. Viable cells are isolated and a second round of stimulation can be performed with B chain (9-23) and irradiated splenocytes as a source of growth factors. After 5 to 6 such passages, the proliferative potential of each cell line is determined. To perform a proliferation assay, B chain-reactive T lymphocyte lines (9-23) can be cultured for 3 days with various concentrations of peptide analogs and irradiated autologous splenocytes. After three days, 0.5-1.0 pCi of [ ³ H]-thymidine is added for 12-16 hours. The cultures are then harvested and radioactivity incorporation is 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 mean for a comparable concentration of B chain (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 lymphocytes (9-23) and that preferentially inhibit the response of such T lymphocytes 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 over 10-15 days. After the last immunization, lymph node and/or spleen cells can be used in a proliferation assay in which various concentrations of the immunizing peptide are cultured with the cells for 3-4 days. The last 18 hours of culture can be performed with tritium-labeled thymidine. The cells can then be harvested and counted in a scintillation counter and the proliferation response can be expressed as CPM + SEM. Potentially useful peptides that induce proliferation that is at least 2-fold greater than background (antigen-free) at 25 μM peptides are considered immunogenic. Alternatively, a potential peptide analog is considered immunogenic if it elicits a proliferative response after immunization of NOD mice in complete Freund's adjuvant. Draining lymph node or spleen cells, when cultured in the presence of the immunizing analog, should induce proliferation that is at least 2-fold greater than background proliferation (antigen-free) 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, peptides can be administered to NOD mice in soluble form or in an emulsion with, for example, incomplete Freund's adjuvant (IFA). Usually, a weekly administration of approximately 400 µg of peptides is sufficient. The incidence of diabetes in treated mice, as well as in untreated or control mice, is then assessed by weekly monitoring of blood glucose concentrations. A blood glucose value of 200 mg/dl or more on two consecutive measurements indicates the presence of diabetes. Peptide analogs should lead to a statistically significant

a significant reduction in the percentage of NOD mice affected by diabetes during a follow-up period of approximately 25 weeks.

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

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

Diabetes treatment and prevention

As mentioned above, the present invention provides methods for the treatment and prevention of type I diabetes, wherein the patient is administered a therapeutically effective amount of an insulin B chain peptide analog as described herein. Diabetic patients suitable for such treatment can be identified according to criteria for the diagnosis of clinical diabetes. Such criteria include, for example, low (less than 10th or 1st percentile of controls) first phase insulin secretion in an intravenous glucose tolerance test (IVGTT) or persistence of high titer antibodies to islet antigens such as insulin, GAD65 and/or ICA512.

Patients without clinical diabetes who may benefit from prophylactic treatment can be identified by any of the following accepted predictive criteria. In patients without clinical diabetes, the development of diabetes in the following years (1-5 years) can be predicted by the following criteria: (i) family history - first-degree relatives are automatically at high risk if they do not have a protective HLA allele; genetic make-up - 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): low first-phase insulin secretion, usually defined as the tenth or first percentile of normal controls, usually precedes the development of type 1 diabetes by 1-5 years. Several of the above criteria are usually considered. For example, the risk of developing diabetes in the next 5 years for relatives of individuals with 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 with no antibodies. Among 50 first-degree relatives of a patient with type 1 diabetes after the onset of diabetes, 49/50 had one or more of the above autoantibodies.

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, but are not limited to, delaying the onset of clinical hyperglycemia, reducing the frequency of hyperglycemic episodes, and/or prolonging the presence of normal C-peptide levels in the patient's blood.

The peptide analogs of the present invention can be administered either alone or in the form of a pharmaceutical composition. Briefly, the pharmaceutical compositions of the present invention can contain one or more peptide analogs of the present invention, in combination with one or more pharmaceutically or physiologically acceptable carriers, diluents or additives. Such compositions can contain 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 can also contain one or more other active ingredients, such as sustained release systems or other immunopotentiating agents.

The compositions of the present invention may be in a form suitable for oral, nasal, intravenous, intracranial, intraperitoneal, subcutaneous or intramuscular administration. Furthermore, the compositions of the present invention may be administered as part of a sustained release implant. In still other embodiments, the compositions of the present invention may be prepared in the form of lyophilized compositions, using suitable additives suitable for achieving stability of the lyophilized form and/or upon rehydration.

The pharmaceutical compositions of the present invention may be administered in a manner appropriate to the condition being treated. The amount and frequency of administration will be determined by such factors as the general condition of the patient and the type and severity of the patient's condition. 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 appropriate dosages may be determined in clinical trials. Patients may be monitored for therapeutic efficacy by delaying progression of clinical diabetes and reducing insulin use to maintain normoglycemia.

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

Examples of embodiments 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 amidated carboxyl termini were prepared with p-methylbenzhydrylamine resin (MBHA resin); for peptides with free carboxyl termini, Merrifield resin with appropriately protected amino acid was used. Both resins were obtained from Bachem Fine Chemicals (Torrance, CA). Derivatized amino acids (Bachem Fine Chemicals) used in the synthesis were of the L-configuration unless otherwise noted and had the N-alpha-amino function protected exclusively by a t-butyloxycarbonyl group. 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 carried out using dicyclohexylcarbodiimide and subsequent amino acids were coupled using 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 hydrofluoric acid (HF), 1.4 ml anisole and 0.28 ml methylethylsulfide per gram of resin conjugate at -20°C for 0.5 h and at 0°C for 0.5 h. The 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 performed 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. The synthetic material was characterized by amino acid analysis, mass spectrometry, and reverse-phase HPLC.

Example 2: Long-term T-lymphocyte lineages

This example illustrates the preparation of long-term insulin-specific T-lymphocyte 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 25 μg/ml porcine insulin and irradiated NOD islet 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 was digested with collagenase and individual islets were isolated manually. Infiltrating lymphocytes were then obtained by mild trypsin digestion of the islets. Insulin-specific T cell lines or clones were propagated by serial stimulation in the presence of NOD spleen cells, porcine insulin and lymphokines. Clones were obtained by limiting dilution of B chain-specific T-lymphocyte lines (9-23) in the presence of antigen-presenting cells and porcine insulin at a dose of 25 µg/ml. Wells with a growing population of cells after limiting dilution were expanded in appropriate medium and after one growth cycle were tested for reactivity to the insulin B chain peptide (9-23) by proliferative response assay.

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

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

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

4« · · · ···· ··· · · · · * · ·· · ··» 0 · · » * · < · ·· ♦ ······ « 0 · · « «··· • « · · · · · · · · · · · · · *

J. Immunol. 25: 1056, 1995). Briefly, 25,000 T cells with 1,000,000 irradiated NOD spleen cells were cultured in triplicate in the presence of 50 pg/ml of insulin B chain peptide 9-23 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. The cells were harvested on a glass fiber filter and the associated radioactivity was counted in a liquid scintillation counter. The results are expressed as the average number of counts per minute for three wells.

Data obtained from five separate T-lymphocyte clones show either a lack of proliferation or a significant reduction in proliferation (relative to the 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 presented in Tables 1 and 2.

« «·

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

T-lymphocyte clone

M P odified Natural _oz -i _ce remainder Substitution PD6-4.3 PD12-2.40 9 With A 12861 42234 10 H A 12507 1409 11 L A 14148 2594 12 In A 8292 671 13 E A 142 519 14 A none 15 L A 161 1422 16 Y A 98 539 17 L A 553 19321 18 In A 234 44785 19 c A 7678 34212 20 G A 2440 38685 21 E A 91 39087 22 R A 6555 51722 23 G A 14304 75441 Antigen-free Native 9-23 163 682 10463 32221

• ·

Table 2: Response (cpm) of insulin-specific mouse NOD T-lymphocyte clones

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

Table 3 shows the response of four different T-cell clones derived 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 the native B chain peptide (9-23) or NBI-6024. The data in Table 3 represent the mean of three samples + standard deviation of the mean. In Table 3 S.I. (stimulatory index) = proliferation (cpm) in the presence of peptide/proliferation (cpm) in medium alone. These data show a significant response when cells are cultured with the native 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 mouse T-lymphocyte clones specific for insulin B chain (9-23) to 50 μM B chain (9-23) or analog A ^x6 ' ^x9 (NBI-6024)

T-lymphocyte clone Attempt No. I Only medium '-'mj insulin B chain (9-23) NBI-6024 (A ^{16 19} ) Average cpm+sem S.I. Average · 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-lymphocyte proliferation assay

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

Peptide analogs of the B chain (9-23) containing alanine substitutions at residues 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. T cell antagonistic activity was detected by assessing the ability of the peptide analogs to inhibit T cell proliferation induced by native B chain (9-23) (SEQ ID NO: 2). In this assay, antigen-presenting cells were first irradiated and then incubated with a competitive peptide analog and native B chain peptide (9-23). T cells were then added to the culture. Various concentrations of the potential peptide analogs were included in the cultures, which were incubated for a total of 4 days. After this incubation period, each culture was pulsed with 1 µCi [ ³ H]-thymidine for an additional 12-18 hours. Cells were then harvested onto glass fiber filters and counted as described above. The mean cpm and standard deviation were calculated from the data obtained in three experiments. The results, shown in Fig. 2, show that peptide analogs containing alanine substitutions at residue 12, 13 or 16 are able to attenuate the response of pathogenic T lymphocytes specific for the insulin B chain (9-23).

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

• ·

Table 4: Inhibition of the response to insulin B chain peptide (9-23) in two mouse NOD T-lymphocyte clones by peptide analog

Clone Terms and 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 N/A B(9-23) 5μΜ + 50μΜ NBI-5096 10.125 ±887 N/A 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 N/A B(9-23) 5μΜ + 50μΜ NBI-5096 17.988 ±584 N/A

N/A - not applicable as 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 recognition of the analog by pathogenic T cells. Furthermore, these results suggest that the analog also binds to MHC with sufficient affinity for recognition by insulin B chain (9-23)-specific T cells.

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

This example shows the lack of stimulation of T-lymphocyte lines and clones obtained from diabetic patients ··· · « · · · » * · ··

Peptide analogs of the B chain (9-23) containing alanine substitutions at residues 13, 515, 16 or 17 or a double alanine-substituted analog at ^{residues 3-6} , ^3-9 (NBI-6024) were prepared as described in Example 1. T-lymphocyte lines from diabetic patients were prepared by isolating lymphocytes from the blood of the patients by density separation. The isolated lymphocytes were then cultured in the presence of insulin B chain peptide (9-23) (10 µM) and recombinant human IL-2 in the presence of 5-10% autologous serum and irradiated autologous peripheral blood lymphocytes in culture medium. After four or five days, the cells were harvested and the cycle was repeated 2 to 3 times.

Proliferation of a T cell line in response to the native B chain peptide (9-23) (SEQ ID NO: 2) or peptide analogs was measured by culturing 25,000 to 100,000 T cells in the presence of 50,000-200,000 irradiated autologous PBL and varying concentrations of insulin B chain peptide (9-23) or peptide analog in triplicate cultures. After 4-5 days of culture, including the final 18 hours with radiolabeled thymidine, the cells were harvested and the associated radioactivity was counted in a liquid scintillation counter. The results are reported as the average number of counts per minute for each peptide analog tested.

The results, shown in Figures 4-7, demonstrate that T-lymphocyte clones and lines that proliferate in response to the peptide (9-23) of the native 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 patient PBL to natural insulin peptide or to the analog NBI-6024

Patient number Patient ID ¹ Stimulation Index* Insulin B (9-23) [50 μΜ] NBI-6024 [50 μΜ] l 100 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 (without 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 the APL NBI-6024 binds with similar affinity to DQ8 antigens. Therefore, the lack of stimulation of diabetic T cells by NBI-6024 is not due to any incompatibility of the peptide with the presenting MHC molecules, but rather to altered recognition by T cells specific for the B chain (9-23).

Example 6: Reducing 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 spontaneously develop diabetes at approximately 3 months of age (Makino et al., Current Topics in Clinical and Experimental Aspects of Diabetes mellitus, Sakamoto et al., ed., 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 peptide analogues of the B chain (9-23) containing alanine substitutions at residues 12, 13 or 16 were administered to NOD mice at weekly intervals. At each dose, 400 µg of each peptide was administered to each of ten animals. After 9 cycles, the percentage of mice in each group that developed diabetes was determined by measuring blood glucose concentrations at weekly intervals using a glucometer. Blood glucose concentrations greater than 200 mg/dl on two subsequent measurements were considered indicative of clinical diabetes.

As shown in Fig. 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 are also shown in Fig. 9.

In another experiment, 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 μg of each peptide was administered to each of ten animals. After 13 cycles, the percentage of mice in each group that developed diabetes was determined using the method described above. As shown in Fig. 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 peptide _A i6,i9 (NBI-6024) to control the development of diabetes in NOD mice, the peptide was administered to young animals. Female mice (n=9, age • · approximately 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 until week 35. From the age of 9-10 weeks, mice were monitored weekly for hyperglycemia by measuring blood glucose concentrations. A group of 10 untreated mice served as controls.

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

0.004) .

These observations were then confirmed and extended in a 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 modified peptide NBI-6024 resulted in a reduction in the incidence of diabetes compared to the neurotensin-treated group or the untreated group.

These results show that the altered peptide ligand NBI-6024, derived from the peptide (9-23) of the insulin B chain, is able to prevent the spontaneous onset of diabetes in animals at risk. It is likely that T cells that recognize other pancreatic antigens are present in these animals, but they are also likely regulated by insulin APLs. The administration was probably done at the same time that autoreactive lymphocytes begin to infiltrate the pancreas and initiate the destructive process. These results give hope that early intervention using these APLs may be effective in delaying or preventing the onset of type 1 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 of the peptide analogs subcutaneously in mannitol-acetate buffer three times over 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 various concentrations of the immunizing peptide for 3-4 days. Tritiated thymidine was added to the culture for the final 18 hours. Cells were harvested and counted in a scintillation counter and the response is expressed as cpm + sem. These results, shown in Figs. 13-16, demonstrate that these representative peptide analogs can bind to murine MHC molecules and can be recognized by appropriate T-lymphocytes.

The ability of the double-substituted peptide NBI-6024 (A ^16.19 ) 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, the mice were sacrificed, spleen and inguinal lymph node cells were removed and collected, and a single cell suspension was prepared. The cells were cultured in the presence of various concentrations (0-25 μM) of NBI-6024. The ability of lymphoid cells to proliferate after 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 the mean cpm + s.e.m. from three cultures. Lymph node cells isolated from mice immunized with the analog in CFA had a strong proliferative response to stimulation with the immunizing analog that was dose-dependent (Table 6). These results indicate that the alterations made in the sequence (9-23) of the insulin B chain 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-lymphocyte 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 and lymph node cells isolated from mice given the soluble peptide exhibited a strong proliferative response to APL after stimulation with NBI-6024 in vitro (Table 7 and Figs. 17A and 17B). Even more significant was the finding that NBI-6024-reactive lymphocytes from mice immunized with the soluble peptide also reacted to the insulin B chain (9-23). This cross-reactivity was not observed for the peptide emulsified in CFA. This ability of the soluble peptide to induce cross-reactivity may be useful in the control of diabetes because it may help mobilize protective NBI-6024-specific T cells to pathogenic target tissue.

• ·

Table 7: Proliferative response of T-lymphocytes from NOD mice immunized with soluble NBI-6024 to NBI-6024 or to native insulin B chain (9-23)

ΝΒΪ-6024 induced cell line Peptide conc. [μΜ] CPM ± SEM NBI-6024 CPM ± SEM Insulin B(9-23) 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, the supernatants of immune lymphoid cell cultures were removed 48 hours after the start of the culture and the concentrations of various cytokines were measured using standard ELISA technology. Surprisingly, it was found that T-lymphocytes from mice immunized with soluble NBI-6024 produced the Th2 cytokines, interleukin-4 (Fig. 18) and interleukin-5 (Table 8), and not the Th1 cytokine interleukin-2. In Table 8, the values are given in pg/ml as the mean of three experiments + SEM. As a control, NBI-6024 in emulsion with CFA was used, which induced • ·

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

Table 8: Cytokine response of soluble NBI-6024-induced T-lymphocytes 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 ability to induce Th-2-like lymphocytes by subcutaneous administration of soluble NBI-6024 is a desirable property because 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 cytokines have potent anti-inflammatory effects that suppress the emergence of autoreactive Th1 lymphocytes that produce pro-inflammatory cytokines that cause disease.

It will be apparent from the foregoing description that while specific embodiments of the invention have been described, various modifications thereof are possible without departing 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 of treating diabetes mellitus using peptide insulin analogues <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 His Leu Val Glu Ala Leu Tyr 1 5 10 15

Leu Val Cys Gly Glu Arg Gly Phe Phe Tyr Thr Pro Lys Thr 20 25 30 <210> 2 <211> 15 <212> PRT <213> Homo sapiens <220>

<223> Residues 9-23 of the B chain of human insulin <400> 2

Ser His Leu Val Glu Ala Leu Tyr Leu Val Cys Gly Glu Arg Gly

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.