Ultra-long acting insulin-fc fusion proteins and methods of use thereof

Abstract

This disclosure provides a recombinantly prepared ultralong-acting insulin-Fc fusion protein for the treatment of diabetes. The insulin-Fc fusion protein comprises an insulin polypeptide linked to a human Fc fragment via a peptide linker. Based on the results obtained, a therapeutically appropriate combination of insulin polypeptide, peptide linker, and Fc fragment, exhibiting adequate in vivo biological activity, demonstrating an extended duration of biological activity, and not exhibiting immunogenicity, is created, requiring a non-obvious combination of these components, including the insulin polypeptide, peptide linker, and Fc fragment, and selective mutations in one or more of these components. In addition to the uses and methods of preparation, exemplary ultralong-acting insulin-Fc fusion proteins, polynucleotides encoding these insulin-Fc fusion proteins, and pharmaceutical formulations of exemplary insulin-Fc fusion proteins are also provided.

Description

[0001] Priority and related applications

[0002] This application relates to and claims priority to U.S. Provisional Patent Application Serial No. 62 / 950,803, filed December 19, 2019, and U.S. Provisional Patent Application Serial No. 62 / 988,441, filed March 12, 2020. The contents of each of the foregoing patent applications are incorporated herein by reference in their entirety. Technical Field

[0003] This technology relates to compositions of insulin-Fc fusion proteins and their use in the treatment of human diabetes. Background Technology

[0004] The following description of the background of this technology is provided only to aid in understanding this technology and does not endorse any prior art that describes or constitutes this technology.

[0005] Diabetes is a chronic condition characterized by insulin deficiency and / or ineffective use of insulin. People with absolute insulin deficiency are classified as having type 1 or insulin-dependent diabetes mellitus (IDDM). Type 1 diabetes is thought to be caused by a genetic predisposition and immune destruction of the insulin-producing beta cells in the pancreas. In contrast, people with diabetes who still produce some insulin but are relatively deficient due to insulin resistance or other dysfunctions are classified as having type 2 or non-insulin-dependent diabetes mellitus (NIDDM). Type 2 diabetes is associated with a genetic predisposition, obesity, and certain medications. Women can also develop temporary insulin resistance during pregnancy, known as gestational diabetes. Some adults are diagnosed with latent adult-onset autoimmune diabetes (LADA), a slowly progressive form of autoimmune diabetes. Similar to type 1 diabetes, in LADA, the insulin-producing beta cells in the pancreas are destroyed, but at a slower rate. A small percentage are diagnosed with juvenile-onset adult-onset diabetes (MODY), which refers to any of several types of inherited diabetes caused by an autosomal dominant gene mutation that disrupts insulin production.

[0006] When the pancreas of patients with type 1 diabetes, LADA, or MODY cannot produce enough insulin, they often exhibit an atypical blood glucose phenotype characterized by hyperglycemia. In these cases, long-term insulin injection therapy is used to treat the patients. Hyperglycemia is also frequently observed in type 2 diabetes and gestational diabetes because patients are unable to properly utilize the insulin produced by the pancreas. In these cases, patients can be treated with oral medications, with or without changes to diet and exercise; however, many subjects eventually develop a condition similar to type 1 diabetes (inflammatory pancreatic disease with a significant reduction in beta cell count) and become dependent on exogenous insulin. If left untreated, diabetes can lead to weight loss, loss of appetite, vomiting, dehydration, motor dysfunction, coma, and even death.

[0007] Approximately 30 million people in the United States, or 9.4% of the population, have diabetes. Type 1 diabetes accounts for about 5% of all diagnosed diabetes cases, affecting about 1.5 million people. Current diabetes treatments include a variety of short-acting insulin products (e.g., (Eli Lilly, Indianapolis, IN) and (Novo Nordisk, Denmark) and long-acting insulin products (e.g., Denmark) (Sanofi, Paris, France) and (Novo Nordisk, (Denmark) Diabetes is administered via multiple daily subcutaneous injections or wearable subcutaneous infusion pumps. The burden of frequent injections leads to poor treatment adherence and underdosing, resulting in poor long-term health outcomes. In fact, more than 7 million American adults are reported to be discharged annually due to diabetes-related cardiovascular events, amputations, and ketoacidosis. Furthermore, more than 14 million American adults are reported to visit emergency rooms annually for conditions such as diabetes-related hypoglycemia and hyperglycemic crises. The estimated prevalence of kidney disease exceeds 36% among American adults aged 20 and older diagnosed with diabetes. Diabetes is the seventh leading cause of death in the United States, with estimated total annual costs exceeding $245 billion. Therefore, cost-effective and less burdensome treatment options are needed for this disease. Summary of the Invention

[0008] On one hand, this disclosure provides a fusion protein comprising an insulin polypeptide and an Fc fragment, wherein the insulin polypeptide and the Fc fragment are linked by a linker, and wherein the fusion protein comprises the following sequence:

[0009] FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGAGGGGAGGGGAGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY SSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:87).

[0010] In some embodiments, this disclosure provides a fusion protein comprising an insulin peptide and an Fc fragment, wherein the insulin peptide and the Fc fragment are linked by a linker, and wherein the Fc fragment is human and comprises the following sequence:

[0011] DKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYX1STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO:77) wherein X1 is S, D, A or R, and wherein the insulin polypeptide is composed of an insulin B chain analog linked to an insulin A chain analog via a C chain, wherein the 16th amino acid (i.e., B16) from the N-terminus of the insulin B chain analog of the insulin polypeptide is alanine (i.e., B16A).

[0012] In some embodiments, the insulin polypeptide comprises the following sequence: FVNQHLCGSX1LVEALALVCGERGFHYGGGGGGSGGGGGGIVEQCCX2STCSLDQLENYC (SEQ ID NO: 9), wherein X1 is not D and X2 is not H. In some embodiments, the insulin polypeptide comprises the following sequence:

[0013] FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYC (SEQ ID NO: 10).

[0014] In some configurations, the connector contains the sequence: GGGGGQGGGGQGGGGGQGGGG (SEQ ID NO:13). In some configurations, the connector contains the sequence: GGGGGAGGGGAGGGGAGGGGG (SEQ ID NO:67). In some configurations, the connector contains the sequence GGGGAGGGG (SEQ ID NO:11).

[0015] In some embodiments, the fusion protein comprises the following sequence:

[0016] FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGAGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYSSTYRVV SEQID NO:89).

[0017] In some embodiments, the fusion protein comprises the following sequence:

[0018] FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY SSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:78).

[0019] In some embodiments, the fusion protein comprises the following sequence:

[0020] FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY DSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:80).

[0021] In some embodiments, the fusion protein comprises the following sequence:

[0022] FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY ASTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:82).

[0023] In some embodiments, the fusion protein comprises the following sequence:

[0024] FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY RSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:84).

[0025] In some embodiments, the fusion protein contains a domain in the following direction from N-terminus to C-terminus: (N-terminus)-insulin polypeptide-linker-Fc fragment-(C-terminus).

[0026] In some embodiments, the fusion protein is a homodimer. In some embodiments, the homodimer percentage of the fusion protein is greater than 90%. In some embodiments, the fusion protein is prepared using HEK293 cells or CHO cells, and the homodimer titer obtained after purification using Protein A beads or Protein A columns is greater than 150 mg / L. In some embodiments, the insulin receptor IC50 of the fusion protein is less than or equal to 5000 nM. In some embodiments, the insulin receptor IC50 of the fusion protein is less than or equal to 2400 nM.

[0027] In some embodiments, the EC50 of the human FcRn receptor of the fusion protein is less than or equal to 1000 ng / mL. In some embodiments, the OD450 ratio of the human Fc(γ)RI receptor is less than or equal to 0.50 at a biotinylated C1q concentration of 3000 ng / mL of the fusion protein. In some embodiments, the OD450 ratio of the human C1q is less than or equal to 0.35 at a biotinylated C1q concentration of 1000 ng / mL. In some embodiments, the fusion protein is formulated into a pharmaceutical composition. In some embodiments, the concentration of the fusion protein in the pharmaceutical composition is about 3 mg / mL or higher. In some embodiments, the pharmaceutical composition is suitable for subcutaneous administration.

[0028] In some embodiments, a physiologically effective amount of the fusion protein or a pharmaceutical composition thereof may be administered to a patient as a method for lowering the patient's blood glucose level. In some embodiments, the patient has been diagnosed with diabetes. In some embodiments, the fusion protein is administered subcutaneously. The fusion protein may be administered to the patient daily, twice weekly, or once weekly. In some embodiments, the fusion protein is administered to the patient once weekly at a dose between 0.025 and 0.500 mg / kg / week.

[0029] In some embodiments, cells can be engineered to express the fusion protein. Cells can be transfected with nucleic acids encoding the fusion protein. In some embodiments, the cells are HEK293 cells or CHO cells.

[0030] In some embodiments, the nucleic acid (cDNA) encoding the fusion protein of SEQ ID NO:87 comprises the following nucleic acid sequence:

[0031] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtgcaggaggcggtggagccggtggaggtggggctggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacagcagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:88)。

[0032] In some embodiments, the nucleic acid (cDNA) encoding the fusion protein of SEQ ID NO:89 comprises the following nucleic acid sequence:

[0033] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtgccggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacagcagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ ID NO:90)。

[0034] In some embodiments, the nucleic acid (cDNA) encoding the fusion protein of SEQ ID NO:78 comprises the following nucleic acid sequence:

[0035] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacagcagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:79)。

[0036] In some embodiments, the nucleic acid (cDNA) encoding the fusion protein of SEQ ID NO:80 comprises the following nucleic acid sequence:

[0037] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacgacagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:81)。

[0038] In some embodiments, the nucleic acid (cDNA) encoding the fusion protein of SEQ ID NO:82 comprises the following nucleic acid sequence:

[0039] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacgccagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:83)。

[0040] In some embodiments, the nucleic acid (cDNA) encoding the fusion protein of SEQ ID NO:84 comprises the following nucleic acid sequence:

[0041] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacagaagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:85)。 Attached Figure Description

[0042] Figure 1 A schematic diagram of an exemplary insulin-Fc fusion protein homodimer is shown.

[0043] Figure 2 The mean fasting blood glucose levels from day 0 to day 3 are shown in N=3 dogs that were administered the homodimer of SEQ ID NO:31 intravenously at 0.2 mg / kg on day 0.

[0044] Figure 3 The image shows a side-by-side sequence comparison of SEQ ID NO: 31, SEQ ID NO: 32, SEQ ID NO: 33, SEQ ID NO: 34, and SEQ ID NO: 35. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0045] Figure 4 The image shows a side-by-side sequence comparison of SEQ ID NO:31, SEQ ID NO:36, SEQ ID NO:37, and SEQ ID NO:38. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0046] Figure 5 The mean fasting blood glucose levels from day 0 to day 7 are shown in N=3 dogs that were administered the homodimer of SEQ ID NO:36 intravenously at 0.2 mg / kg on day 0.

[0047] Figure 6 The mean fasting blood glucose levels from day 0 to day 7 are shown in N=6 dogs that were administered the homodimer of SEQ ID NO:36 intravenously at 0.33 mg / kg on day 0.

[0048] Figure 7 The mean anti-drug antibody titer (μg / mL) of N=3 dogs subcutaneously administered the homodimer of SEQ ID NO:36 on day 0 (0.30 mg / kg), day 28 (0.33 mg / kg), day 35 (0.33 mg / kg), day 42 (0.50 mg / kg), day 49 (1.00 mg / kg), and day 56 (1.00 mg / kg) is shown.

[0049] Figure 8The image shows a side-by-side sequence comparison of SEQ ID NO:39, SEQ ID NO:40, SEQ ID NO:41, and SEQ ID NO:42. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0050] Figure 9 The mean anti-drug antibody titer (μg / mL) of N=1 dogs subcutaneously administered the homodimer of SEQ ID NO:42 on day 0 (0.33 mg / kg), day 7 (0.50 mg / kg), day 14 (0.50 mg / kg), and day 21 (0.50 mg / kg) is shown.

[0051] Figure 10 The mean anti-drug antibody titer (μg / mL) of N=1 dogs subcutaneously administered the homodimer of SEQ ID NO:43 on day 0 (0.33 mg / kg) and day 14 (0.16 mg / kg) is shown.

[0052] Figure 11 The mean fasting blood glucose levels from day 0 to day 7 are shown in N=2 dogs that were subcutaneously administered the homodimer of SEQ ID NO:43 at 0.33 mg / kg on day 0.

[0053] Figure 12 The image shows a side-by-side sequence comparison of SEQ ID NO: 43, SEQ ID NO: 44, SEQ ID NO: 45, SEQ ID NO: 46, SEQ ID NO: 47, and SEQ ID NO: 48. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0054] Figure 13 The following shows a side-by-side sequence comparison of SEQ ID NO: 43, SEQ ID NO: 49, SEQ ID NO: 50, SEQ ID NO: 51, and SEQ ID NO: 52. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0055] Figure 14The image shows a side-by-side sequence comparison of SEQ ID NO:43, SEQ ID NO:48, and SEQ ID NO:53. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0056] Figure 15 The image shows a side-by-side sequence comparison of SEQ ID NO:43, SEQ ID NO:51, SEQ ID NO:52, and SEQ ID NO:54. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0057] Figure 16 The image shows a side-by-side sequence comparison of SEQ ID NO: 26, SEQ ID NO: 28, SEQ ID NO: 43, SEQ ID NO: 55, SEQ ID NO: 56, and SEQ ID NO: 57. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0058] Figure 17 The fasting blood glucose levels of N=1 dogs administered the homodimer of SEQ ID NO:28 subcutaneously at 0.16 mg / kg on day 0 are shown from day 0 to day 7.

[0059] Figure 18 The anti-drug antibody titers (μg / mL) of N=1 dogs subcutaneously administered the homodimer of SEQ ID NO:28 on day 0 (0.16 mg / kg), day 14 (0.16 mg / kg), day 28 (0.16 mg / kg), and day 42 (0.16 mg / kg) are shown.

[0060] Figure 19 The fasting blood glucose levels of N=1 dogs administered subcutaneously with the homodimer of SEQ ID NO:26 at 0.33 mg / kg on day 0 are shown from day 0 to day 7.

[0061] Figure 20 The fasting blood glucose levels of N=1 dogs administered the homodimer of SEQ ID NO:26 subcutaneously from day 0 to day 60 are shown on day 0 (0.33 mg / kg), day 15 (0.16 mg / kg), day 31 (0.16 mg / kg), and day 45 (0.15 mg / kg).

[0062] Figure 21 The anti-drug antibody titers (μg / mL) of N=1 dogs subcutaneously administered the homodimer of SEQ ID NO:26 on day 0 (0.33 mg / kg), day 15 (0.16 mg / kg), day 31 (0.16 mg / kg), and day 45 (0.15 mg / kg) are shown.

[0063] Figure 22 The fasting blood glucose levels of N=1 dogs administered the homodimer of SEQ ID NO:58 subcutaneously at 0.16 mg / kg on day 0 are shown from day 0 to day 7.

[0064] Figure 23 The fasting blood glucose levels of N=1 dogs administered the homodimer of SEQ ID NO:59 subcutaneously at 0.16 mg / kg on day 0 are shown from day 0 to day 7.

[0065] Figure 24 The image shows a side-by-side sequence comparison of SEQ ID NO:61 and SEQ ID NO:62. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0066] Figure 25 The fasting blood glucose levels (%) of N=1 dogs administered the homodimer of SEQ ID NO:61 subcutaneously at 0.16 mg / kg on day 0, and the fasting blood glucose levels of N=1 dogs administered the homodimer of SEQ ID NO:62 subcutaneously at 0.16 mg / kg on day 0, are shown from day 0 to day 7.

[0067] Figure 26 The fasting blood glucose levels of N=1 dogs subcutaneously administered the homodimer of SEQ ID NO:30 from day 0 to day 7 are shown, excluding the timing of dog feeding.

[0068] Figure 27 The image shows a side-by-side sequence comparison of SEQ ID NO:76, SEQ ID NO:91, and SEQ ID NO:78. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0069] Figure 28The diagram shows a side-by-side sequence comparison of SEQ ID NO: 75, SEQ ID NO: 76, SEQ ID NO: 91, SEQ ID NO: 92, SEQ ID NO: 93, SEQ ID NO: 94, and SEQ ID NO: 95. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0070] Figure 29 The image shows a side-by-side sequence comparison of SEQ ID NO: 76, SEQ ID NO: 78, SEQ ID NO: 80, SEQ ID NO: 82, SEQ ID NO: 84, and SEQ ID NO: 86. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0071] Figure 30 The image shows a side-by-side sequence comparison of SEQ ID NO: 87, SEQ ID NO: 96, SEQ ID NO: 78, SEQ ID NO: 97, SEQ ID NO: 89, and SEQ ID NO: 98. An asterisk (*) indicates that the entire sequence is completely homologous at the given sequence position, while a colon (:), a period (.), or a space indicates a conserved, moderately, or very different amino acid mutation at the given sequence position, respectively.

[0072] Figure 31 The complete amino acid sequence is shown, including the leader sequence of the fusion protein (SEQ ID NO: 78) and its corresponding nucleic acid sequence (SEQ ID NO: 79).

[0073] Figure 32 The complete amino acid sequence is shown, including the leader sequence of the fusion protein (SEQ ID NO: 80) and its corresponding nucleic acid sequence (SEQ ID NO: 81).

[0074] Figure 33 The complete amino acid sequence is shown, including the leader sequence of the fusion protein (SEQ ID NO: 82) and its corresponding nucleic acid sequence (SEQ ID NO: 83).

[0075] Figure 34 The complete amino acid sequence is shown, including the leader sequence of the fusion protein (SEQ ID NO: 84) and its corresponding nucleic acid sequence (SEQ ID NO: 85).

[0076] Figure 35The complete amino acid sequence is shown, including the leader sequence of the fusion protein (SEQ ID NO: 87) and its corresponding nucleic acid sequence (SEQ ID NO: 88).

[0077] Figure 36 The complete amino acid sequence is shown, including the leader sequence of the fusion protein (SEQ ID NO: 89) and its corresponding nucleic acid sequence (SEQ ID NO: 90).

[0078] Figure 37 The complete amino acid sequence is shown, including the leader sequence of the fusion protein (SEQ ID NO: 86) and its corresponding nucleic acid sequence (SEQ ID NO: 100).

[0079] Figure 38 The mean fasting blood glucose levels of N=12 Balb / c mice from hour 0 to hour 10 are shown before subcutaneous injection of the homodimers of SEQ ID NO:87 and SEQ ID NO:89 at 300 μg / kg. Detailed Implementation

[0080] Insulin therapy requiring lower frequency of administration (e.g., once-weekly injections) is less burdensome for patients, leading to better adherence, better glucose control, and ultimately better long-term health outcomes. As disclosed herein, the proposed ultra-long-acting insulin therapy for human clinical application includes insulin-Fc fusion proteins that utilize human Fc fragments to prolong their in vivo action. Insulin-Fc fusion proteins suitable for ultra-long-acting diabetes therapy should meet various design objectives. Insulin-Fc fusion proteins suitable for ultra-long-acting diabetes therapy should be able to be prepared in mammalian cells such as human embryonic kidney (HEK, e.g., HEK293) cells and have an acceptable potency of the desired homodimer product (e.g., a homodimer potency greater than 50 mg / L for transiently transfected HEK cells, greater than 75 mg / L for transiently transfected HEK cells, greater than 100 mg / L for transiently transfected HEK cells, greater than 150 mg / L for transiently transfected HEK cells, etc.). It is believed that only human insulin-Fc fusion protein conformations with a homodimer titer greater than 150 mg / L can be used in this invention because experience has shown that homodimer titers below this level are unlikely to lead to commercially viable homodimer titers in stably transfected Chinese hamster ovary (CHO) cells to meet the low production cost requirements of the relatively commercial human insulin market.

[0081] Furthermore, the insulin-Fc fusion protein must bind to the IR with a sensitive affinity (e.g., IC50 less than 5000 nM, IC50 less than 4000 nM, IC50 less than 3000 nM, IC50 less than 2400 nM, IC50 more preferably less than 2000 nM, etc.) as measured in a 4°C IM-9 IR binding assay. Empirically, only molecules exhibiting IR activity with an IC50 value less than 5000 nM are considered likely to exhibit the necessary biological activity. In some preferred embodiments, the insulin-Fc fusion protein exhibits an IR activity with an IC50 value less than 2400 nM, more preferably less than 2000 nM. The insulin-Fc fusion protein conformation must also exhibit sustained biological activity in vivo (e.g., exhibiting glucose-lowering activity for more than about 2 hours, 6 hours, 9 hours, 12 hours, 18 hours, 1 day, 1.5 days, 2 days, 2.5 days, 3 days, 4 days, 5 days, 6 days, 7 days, or longer) to justify a lower dosing frequency. The insulin-Fc fusion protein conformation must also exhibit a prolonged systemic residence time in vivo (e.g., a serum half-life greater than 3 days or longer). The sustained biological activity and prolonged residence time of a given insulin-Fc fusion protein conformation can be predicted by its ability to bind to the FcRn receptor, which is responsible for prolonging the in vivo elimination half-life of both the antibody and the Fc fusion protein. FcRn receptor activity is typically measured by the concentration of insulin-Fc fusion protein that causes it to reach half of its maximum binding (i.e., the EC50 value), as measured in assays using OD 450 nm values ​​(e.g., enzyme-linked immunosorbent assay (ELISA)). Empirically, insulin-Fc fusion protein conformations exhibiting a human FcRn receptor EC50 value less than or equal to 1500 ng / mL (more preferably less than 1000 ng / mL) are most likely to exhibit a sufficiently long half-life to justify weekly dosing.

[0082] Finally, for use in treating chronic diseases such as diabetes, the insulin-Fc fusion protein conformation must not induce the production of anti-drug antibodies, especially those that neutralize the molecular biological activity after repeated administration. The tendency of a given insulin-Fc fusion protein conformation to induce adverse immunogenic responses can be predicted primarily by its ability to bind to the Fc(γ)RI receptor, which plays a crucial role in many immune system effector functions, including the phagocytosis of opsonized molecules, the release of inflammatory mediators, and antibody-dependent cytotoxicity. Fc(γ)RI receptor activity is typically measured by enzyme-linked immunosorbent assay (ELISA) at an absorbance value (OD450) at 450 nm obtained on a microplate reader at a given insulin-Fc fusion protein concentration. Empirically, insulin-Fc fusion protein conformations exhibiting a human Fc(γ)RI receptor OD450 ratio less than or equal to 0.50 at a biotinylated-Fc(γ)RI concentration (where the reference insulin-Fc fusion protein conformation for this ratio is SEQ ID NO:76) may demonstrate sufficiently low immunogenicity to justify repeated weekly dosing. The tendency of a given insulin-Fc fusion protein conformation to induce an adverse immunogenic response can also be predicted by its ability to bind complement component 1q (C1q), which activates the complement cascade, causing phagocytes to clear the bound molecule, attracting additional phagocytes to the inflammatory region, and activating the cell-killing attack membrane complex. C1q activity is typically measured by the OD450 value obtained on a microplate reader at a given concentration of insulin-Fc fusion protein conformation coated on a microplate using an enzyme-linked immunosorbent assay (ELISA). Based on experience, insulin-Fc fusion protein conformations exhibiting a human C1q receptor assay OD450 ratio less than or equal to 0.35 at a biotinylated-C1q concentration of 1000 ng / mL (where the reference insulin-Fc fusion protein conformation for this ratio is SEQ ID NO:76) may exhibit sufficiently low immunogenicity to justify repeated weekly dosing.

[0083] The proposed ultra-long-acting insulin therapy for human clinical application includes insulin-Fc fusion proteins that utilize human Fc fragments to prolong their in vivo action. To understand the behavior of various designed insulin-Fc fusion protein conformations, an insulin-Fc fusion protein conformation suitable for use as an ultra-long-acting insulin in dogs was first considered. Since the human Fc fragment is expected to be immunogenic and thus capable of inducing anti-drug antibodies in dogs, a canine Fc fragment was substituted for the human Fc fragment.

[0084] However, it was unexpectedly found that a simple exchange between a human Fc fragment and any canine Fc fragment in the insulin-Fc fusion protein conformation does not necessarily produce a product with an acceptable homodimer titer (e.g., a homodimer titer greater than 50 mg / L) or a sufficiently high NAOC value (e.g., a NAOC greater than 150 FBGL%·day·kg / mg). For example, in some cases, only a specific isotype of the Fc fragment (e.g., canine IgGB) resulted in an insulin-Fc fusion protein conformation with a sufficiently high homodimer titer to meet the design target (e.g., a homodimer titer greater than 50 mg / L for the canine insulin-Fc fusion protein) and an acceptablely high NAOC value (e.g., a NAOC greater than 150 FBGL%·day·kg / mg). In other cases, specific amino acids of the insulin polypeptide in the insulin-Fc fusion protein conformation are found to be immunogenic in the target species, thus requiring site-directed mutagenesis to find relatively small amounts of the following insulin-Fc fusion protein conformations: non-immunogenic and biologically active in the target species, with an acceptablely high NAOC value (e.g., greater than 150 FBGL%·day·kg / mg) and a NAOCR value greater than 0.50 after the third weekly subcutaneous dosing.

[0085] In other cases, when the Fc fragment was mutated to prevent glycosylation and thereby further reduce the immunogenicity of the insulin-Fc fusion protein conformation, it was unexpectedly found that only specific amino acid mutations in the Fc fragment resulted in the desired homodimer titer (e.g., a homodimer titer greater than 50 mg / L for the canine insulin-Fc fusion protein conformation) and NAOC value (e.g., a NAOC value greater than 150 FBGL%·day·kg / mg). Furthermore, it was found that additional mutations in the insulin components of the insulin-Fc fusion protein conformation were required to produce these Fc-mutated, non-glycosylated insulin Fc-fusion protein conformations with the desired homodimer titer (e.g., a homodimer titer greater than 50 mg / L for the canine insulin-Fc fusion protein conformation) and NAOC value (e.g., a NAOC value greater than 150 FBGL%·day·kg / mg), while also achieving a NAOCR value greater than 0.50 after the third weekly subcutaneous dosing.

[0086] For human ultra-long-acting insulin, to maximize homodimer potency and reduce manufacturing costs, insulin-Fc fusion protein conformations comprising Fc fragments based on different isoforms of human IgG molecules (e.g., IgG1 and IgG2) were produced. Quite unexpectedly, it was found that in the insulin-Fc fusion protein conformations, switching from an IgG2-based Fc fragment to an IgG1-based Fc fragment increased the average homodimer potency by more than 50% without significantly impairing IR or FcRn binding activity. However, given that the human Fc(γ)RI receptor OD450 ratio of the IgG1-derived insulin-Fc fusion protein conformation obtained at a biotinylated-Fc(γ)RI concentration of 3000 ng / mL (where the reference insulin-Fc fusion protein conformation for this ratio is SEQ ID NO:76) is much greater than 0.50, and the human C1q receptor OD450 ratio of the IgG1-derived insulin-Fc fusion protein conformation obtained at a biotinylated C1q concentration of 1000 ng / mL (where the reference insulin-Fc fusion protein conformation for this ratio is SEQ ID NO:76) is much greater than 0.35, it is considered that the obtained IgG1-derived insulin-Fc fusion protein conformation is more likely to interact adversely with the immune system and produce neutralizing antibodies.

[0087] To reduce unwanted immunogenicity, the insulin-Fc fusion protein conformation was mutated to prevent glycosylation of the Fc fragment during synthesis in host cells. Specifically, the conserved asparagine (N)-glycosylation site in the CH2 domain of the IgG1 Fc heavy chain was mutated to different amino acids (e.g., S, D, A, R, and Q) in an attempt to maintain increased homodimer production of the insulin-Fc fusion protein while reducing the interaction between Fc(γ)RI and C1q. Interestingly, the non-glycosylated insulin-Fc fusion protein conformation with S, D, A, and R mutations provided improved homodimer potency compared to the glycosylated parent conformation; however, the homodimer potency of the insulin-Fc fusion protein conformation with the Q mutation at the conserved asparagine (N)-glycosylation site was too low (i.e., below the design target of 150 mg / L for the human insulin-Fc fusion protein conformation) to support the requirement of low preparation cost. Furthermore, for the non-glycosylated insulin-Fc fusion protein conformations with S, D, A, and R mutations, the binding of Fc(γ)RI to C1q was significantly reduced. However, in these insulin-Fc fusion protein conformations, the increased yield and immunogenicity were offset by unexpectedly low FcRn binding affinity and significantly lower IR binding relative to the glycosylated parent conformation, indicating an unacceptable reduction in in vivo residence time and biological activity.

[0088] As unexpectedly observed with the insulin-Fc fusion protein conformation containing the canine Fc fragment, it was found that by altering a single amino acid in the insulin sequence within the human insulin-Fc fusion protein conformation, the increased yield and immunogenicity of the S, D, A, and R-mutated non-glycosylated insulin-Fc fusion protein conformations were maintained or improved, while IR and FcRn binding affinity were enhanced relative to the original glycosylated parent conformation (e.g., lower IR assay IC50 values), rather than decreased. Therefore, the resulting non-glycosylated, mutant insulin-Fc fusion protein conformations were expected to exhibit acceptable in vivo glucose-lowering potency and prolonged residence time. Unexpectedly, altering the same amino acid in the non-glycosylated insulin-Fc fusion protein conformation using a Q mutation at the conserved asparagine (N)-glycosylation site resulted in a significant reduction in the already poor homodimer potency.

[0089] To understand how to further manipulate the properties of S, D, A, and R non-glycosylated mutant insulin-Fc fusion protein conformations, the linker region of the insulin-Fc fusion protein conformation was modified to link the insulin peptide to the Fc fragment. Test results showed that, without the linker, the homodimer potency of the resulting canine insulin-Fc fusion protein conformation was unacceptably low (i.e., less than 50 mg / L). For insulin-Fc fusion protein conformations containing linkers of the same length, certain amino acid sequences were found to be superior to others in terms of homodimer potency, IR binding to the FcRn receptor, and Fc(γ)RI binding to Clq.

[0090] Therefore, this paper provides specific prepareable, high-purity, long-acting, bioactive, non-immunogenic insulin-Fc fusion protein conformations, which respectively contain a mutant insulin peptide, a non-glycosylated Fc fragment, and a linker between the mutant insulin peptide and the non-glycosylated Fc fragment. The insulin-Fc fusion protein conformation meets the following design objectives: acceptablely high homodimer potency (e.g., homodimer potency greater than 150 mg / L), and an IC50 value for IR assay (e.g., less than 5000 nM). The exemplary insulin-Fc fusion protein conformations are expected to exhibit sufficiently low immunogenicity and a sufficiently long half-life to justify repeated weekly dosing for the treatment of diabetes. These conformations are intended to demonstrate IC50 values ​​of less than 2400 nM, more preferably less than 2000 nM; EC50 values ​​of human FcRn receptors (e.g., less than or equal to 1500 ng / L, more preferably less than 1000 ng / mL); OD450 ratios of human Fc(γ)RI receptors (e.g., less than or equal to 0.50 at a biotinylated-Fc(γ)RI receptor concentration of 3000 ng / mL, wherein the reference insulin-Fc fusion protein conformation for this ratio is SEQ ID NO:76); and OD450 ratios of human C1q receptors (e.g., less than or equal to 0.35 at a biotinylated C1q concentration of 1000 ng / mL, wherein the reference insulin-Fc fusion protein conformation for this ratio is SEQ ID NO:76).

[0091] definition

[0092] As used herein, the articles “a” and “a kind” refer to one or more grammatical objects, such as at least one article. When used in conjunction with the term “including” in this text, the words “a” or “a kind” may mean “one / kind”, but they are also consistent with the meanings of “one / kind or more / kinds”, “at least one / kind”, and “one / kind or more than one / kind”.

[0093] As used herein, “about” and “approximately” generally indicate the acceptable degree of error in a quantity measured given the nature or precision of the measurement. Exemplary degrees of error are within 20% of a given range value, typically within 10%, and more often within 5%.

[0094] As used herein, the amount of a molecule, compound, conjugate, or substance that effectively treats a disease (e.g., the disease described herein), the “therapeutic effective amount,” or the “effective amount” means the amount of a molecule, compound, conjugate, or substance that effectively treats a subject after a single or multiple administration to the subject, or that cures, reduces, alleviates, or improves a subject with a disease (e.g., the disease described herein) in a manner that would be expected without such treatment.

[0095] As used herein, the term “analog” refers to a compound or conjugate that has a similar chemical structure to another compound or conjugate but is different in at least one respect (e.g., the compound or conjugate described herein, such as insulin).

[0096] As used herein, the term "antibody" or "antibody molecule" refers to an immunoglobulin molecule (Ig), specifically the immunologically active portion of an immunoglobulin (Ig) molecule, i.e., a molecule containing an antigen-binding site that specifically binds to an antigen, for example, to induce an immune response. As used herein, the term "antibody domain" refers to the variable or constant region of an immunoglobulin. Antibodies are described in the art as including several classes, such as IgA, IgM, or IgG in the case of mammals (e.g., humans). Immunoglobulin classes can be further subdivided into different isotypes, such as IgGA, IgGB, IgGC, and IgGD for dogs, and IgG1, IgG2, IgG3, and IgG4 for humans. Those skilled in the art will recognize that immunoglobulin isotypes of a given immunoglobulin class will include amino acid sequences, structural and functional properties (e.g., different binding affinities to Fc(γ) receptors). "Specific binding" or "immune reaction" refers to an antibody reacting with one or more antigenic determinants of a desired antigen and having a low affinity for other peptides, such as not reacting with other peptides.

[0097] As used herein, the term “area under the curve” or “AUC” refers to the integral area under the curve representing a subject’s FBGL% (fasting blood glucose level) versus time after administration of a given dose of insulin-Fc fusion protein. As used herein, the term “area above the curve” or “AOC” is used as a measure of the biopotency of insulin-Fc fusion protein, such that AOC equals the difference between the total possible area under the curve representing FBGL% versus time and the AUC value. As used herein, “normalized area above the curve,” “normalized AOC,” or “NAOC” is the AOC value divided by the actual dose of insulin-Fc fusion protein administered. As used herein, the term “normalized AOC ratio” or “NAOCR” is the ratio of the NAOC derived from a specific administration of insulin-Fc fusion protein to the NAOC derived from the first administration of insulin-Fc fusion protein in a series of administrations. Thus, NAOCR provides a measure of the change in bioactivity following repeated administration of insulin-Fc fusion protein.

[0098] As used herein, the terms “bioactivity,” “activity,” “biological activity,” “potency,” “bioactive potency,” or “biopotency” refer to the extent to which the insulin-Fc fusion protein activates the insulin resistance (IR) and / or lowers blood glucose levels in the target subject. As used herein, “in vitro activity” or “IR activity” refers to the affinity of the insulin-Fc fusion protein for binding to the IR and is typically measured by replacing half the concentration of the insulin IR reference standard (i.e., IC50) with the insulin-Fc fusion protein in a competitive binding assay. As used herein, “in vivo activity” refers to the extent and duration of the reduction in fasting blood glucose levels in the target subject following administration of the insulin-Fc fusion protein.

[0099] As used herein, the terms “biosynthesis,” “recombinant synthesis,” or “recombinant preparation” refer to a process in which the insulin-Fc fusion protein is expressed in a host cell by transfecting the host cell with a nucleic acid molecule (e.g., a vector) encoding the insulin-Fc fusion protein (e.g., where the entire insulin-Fc fusion protein is encoded by a single nucleic acid molecule). Exemplary host cells include mammalian cells, such as HEK293 cells or CHO cells. Cells can be cultured using standard methods in the art, and the expressed insulin-Fc fusion protein can be harvested and purified from cell cultures using standard methods in the art.

[0100] As used herein, the term "cell surface receptor" refers to a molecule, such as a protein, that is typically found on the outer surface of a cell membrane and interacts with soluble molecules, such as those circulating in the blood supply. In some embodiments, cell surface receptors may include hormone receptors (e.g., insulin hormone receptors or insulin receptor (IR)) or Fc receptors (e.g., Fc(γ) receptors, such as Fc(γ)RI, or Fc neonatal receptors, such as FcRn) that bind to antibody fragments or Fc regions. As used herein, “in vitro activity” or “Fc(γ) receptor activity” or “Fc(γ) receptor binding” or “FcRn receptor activity” or “FcRn binding” refers to the affinity of the insulin-Fc fusion protein for binding to an Fc receptor (e.g., an Fc(γ) receptor or an FcRn receptor), typically measured by the concentration of insulin-Fc fusion protein that allows it to achieve half of its maximum binding (i.e., the EC50 value), as measured in an assay using an OD450 nm value measured on a microplate reader (e.g., an enzyme-linked immunosorbent assay (ELISA)). Alternatively, the binding affinity of the insulin-Fc fusion protein to an Fc receptor (e.g., an Fc(γ) receptor or an FcRn receptor) is measured by the OD450 nm value obtained on a microplate reader at a given concentration of insulin-Fc fusion protein in an ELISA assay.

[0101] As used herein, the term "C1q" or "complement component 1q" refers to the protein complex involved in the complement system, which is part of the innate immune system. C1q, together with C1r and C1s, forms the C1 complex. C1q plays a role in the specific antigen presentation from dendritic cells to T cells and B cells.

[0102] As used herein, the term “fasting blood glucose level” or “FBGL” refers to the mean blood glucose level of the target subject at the end of the period of fasting and immediately before administration of the insulin-Fc fusion protein. As used herein, the terms “fasting blood glucose level percentage,” “fasting blood glucose level %,” or “FBGL%” refer to the ratio of a given blood glucose level to the fasting blood glucose level multiplied by 100.

[0103] As used herein, the terms “immunogenic” or “immunogenicity” refer to the ability of a given molecule (e.g., the insulin-Fc fusion protein of the present invention) to stimulate the immune system of a target subject such that, upon repeated administration of the molecule, the subject produces antibodies (i.e., antidrug antibodies) capable of specifically binding to that molecule. As used herein, the terms “neutralizing,” “neutralizing antibody,” or “neutralizing antidrug antibody” refer to the ability of an antibody to interfere with the biological activity of a compound in a target subject. As used herein, the terms “immunogenic epitope,” “immunogenic hotspot,” or “hotspot” refer to a mutation or epitope of a given molecule (e.g., the insulin-Fc fusion protein of the present invention) responsible for moderate or strong binding of antidrug antibodies.

[0104] As used herein, the term “insulin reference standard” is any of the following: (i) naturally occurring insulin derived from mammals (e.g., dogs or humans); (ii) an insulin polypeptide that does not contain an Fc fragment; or (iii) standard care insulin (e.g., commercially available insulin).

[0105] As used herein, the term "monomer" refers to a protein or fusion protein comprising a single polypeptide. In some embodiments, a "monomer" is a protein or fusion protein, such as a single polypeptide including an insulin polypeptide and an Fc fragment polypeptide, wherein the insulin and Fc fragment polypeptide are linked by peptide bonds to form a single polypeptide. In some embodiments, a monomer is encoded by a single nucleic acid molecule.

[0106] As used herein, “N-terminus” refers to the starting point of a protein or polypeptide that is initiated by an amino acid containing a free amino group, wherein the free amino group is the α-amino group of the amino acid (e.g., a free amino group covalently linked to a carbon atom adjacent to a second carbon atom, wherein the second carbon atom is part of the carbonyl group of the amino acid). As used herein, “C-terminus” refers to the terminal end of a protein or polypeptide that is terminated by an amino acid containing a carboxylic acid group, wherein the carbon atom of the carboxylic acid group is located adjacent to the α-amino group of the amino acid.

[0107] As used herein, “OD450,” “optical density at 450 nm,” and “absorbance at 450 nm” are used interchangeably and refer to the absorbance of light passing through the sample at 450 nm in assays such as microplate-based assays, enzyme-linked immunosorbent assays, and ELISA assays, as read using an ELISA reader.

[0108] As used herein, the “OD450 ratio” for a specific assay refers to a comparison of the OD450 value obtained by running a first test insulin-Fc fusion protein at a specific time with the OD450 value obtained by running a second test insulin-Fc fusion protein at another time. The OD450 ratio is obtained by dividing the OD450 value of the first test article by the OD450 value of a reference insulin-Fc fusion protein. Similarly, a second OD450 ratio of the second test article can be obtained by dividing the OD450 value of the second test article by the same OD450 value of the reference insulin-Fc fusion protein used to calculate the OD450 ratio of the first test article. As a result, the assay characteristics of the first and second test insulin-Fc fusion proteins can be compared. The reference insulin-Fc fusion protein conformation used to calculate the OD450 ratio is SEQ ID NO:76.

[0109] As used in this article, "pharmacodynamics" or "PD" generally refers to the biological effects of the insulin-Fc fusion protein in subjects. Specifically, in this article, PD refers to the measurement of the decrease in fasting blood glucose levels over time in subjects after administration of the insulin-Fc fusion protein.

[0110] As used herein, “pharmacokinetic” or “PK” generally refers to the characteristic interaction between the insulin-Fc fusion protein and the subject’s body in terms of its absorption, distribution, metabolism, and excretion. Specifically, in this article, PK refers to the concentration of the insulin-Fc fusion protein in the blood or serum of a subject at a given time after administration of the insulin-Fc fusion protein. As used herein, “half-life” refers to the time it takes for the concentration of the insulin-Fc fusion protein in the blood or serum of a subject to reach half of its original value, as calculated from a first-order exponential decay model of drug elimination. Insulin-Fc fusion proteins with larger “half-life” values ​​exhibit a longer duration of action in target subjects.

[0111] As used herein, the terms “sequence identity,” “sequence homology,” “homology,” or “identical” for amino acid or nucleotide sequences describe the presence of identical nucleotide or amino acid residues in the variant and the reference sequence when a specified contiguous segment of the variant’s nucleotide or amino acid sequence is aligned and compared with the nucleotide or amino acid sequence of the reference sequence. Methods for sequence alignment and determining identity between sequences are known in the art, including the use of Clustal Omega, which organizes, aligns, and compares the similarity of sequences, wherein the software highlights each sequence position and compares all sequences at that position, and assigns one of the following scores: "*" (asterisk) indicates a sequence position with a single completely conserved residue; ":" (colon) indicates conservation between groups with highly similar characteristics and a score greater than 0.5 in the Gonnet PAM 250 matrix; "." (period) indicates conservation between groups with weakly similar characteristics and a score less than or equal to 0.5 in the Gonnet PAM 250 matrix; "-" (dash) indicates a sequence gap, which indicates the absence of local homology within a specific set of comparisons within a certain sequence range; and a space "" indicates little or no sequence homology at that specific position in the compared sequences. See, for example, Ausubel et al., eds. (1995) Current Protocols in Molecular Biology, Chapter 19 (Greene Publishing and Wiley-Interscience, New York); and the ALIGN program Dayhoff (1978) in Atlas of Polypeptide Sequence and Structure 5: Suppl. 3 (National Biomedical Research Foundation, Washington, DC). For optimal alignment of two nucleotide sequences, consecutive segments of the variant nucleotide sequence may have additional or missing nucleotides relative to the reference nucleotide sequence. Similarly, for optimal alignment of two amino acid sequences, consecutive segments of the variant amino acid sequence may have additional or missing amino acid residues relative to the reference amino acid sequence. In some embodiments, the consecutive segments used for comparison with the reference nucleotide sequence or reference amino acid sequence will contain at least 6, 10, 15, or 20 consecutive nucleotide or amino acid residues, and may be 30, 40, 50, 100, or more nucleotide or amino acid residues. Increased sequence identity associated with the presence of vacancies in the nucleotide or amino acid sequence of a variant can be corrected by assigning vacancy penalties. Methods for sequence alignment are known in the art.

[0112] In some implementations, the determination of the percentage of identity or "homology" between two sequences is accomplished using mathematical algorithms. For example, the percentage of identity between amino acid sequences is determined using the Smith-Waterman homology search algorithm, which employs an affine 6-vacancy search with a vacancy opening penalty of 12 and a vacancy extension penalty of 2, and a BLOSUM matrix of 62. The Smith-Waterman homology search algorithm is described in Smith and Waterman (1981) Adv. Appl. Math 2:482-489, which is incorporated herein by reference. In some implementations, the percentage of identity between nucleotide sequences is determined using the Smith-Waterman homology search algorithm, which employs a vacancy opening penalty of 25 and a vacancy extension penalty of 5. Such sequence identity determination can be performed using, for example, TimeLogic's DeCypher Hardware Accelerator.

[0113] As used herein, the term "homology" is used to compare two or more proteins by locating common structural features and common spatial distributions, such as β-chains, helices, and folds. Therefore, homologous protein structures are defined through spatial analysis. Measuring structural homology involves computing the geometric topological features of space. One approach used to generate and analyze three-dimensional (3D) protein structures is homology modeling (also known as comparative modeling or knowledge-based modeling), which works by finding similar sequences based on the fact that 3D similarity reflects 2D similarity. Homologous structures do not imply that sequence similarity is a necessary condition.

[0114] As used herein, the terms “subject” and “patient” are intended to include canines and humans suffering from diseases or disorders such as diabetes or other diseases or disorders described herein, or normal subjects.

[0115] As used herein, the terms “potency” or “yield” refer to the amount of fusion protein product (e.g., the insulin-Fc fusion protein described herein) biosynthesized per volume of cell culture (e.g., in mammalian cells, such as HEK293 cells or CHO cells). The amount of product may be determined at any step of the production process (e.g., before or after purification), but yield or potency is always described per volume of original cell culture. As used herein, the terms “product yield” or “total protein yield” refer to the total amount of insulin-Fc fusion protein expressed by cells and purified via at least one affinity chromatography step (e.g., protein A or protein G), and include insulin-Fc fusion protein monomers, homodimers of insulin-Fc fusion protein, and higher-order molecular aggregates of insulin-Fc fusion protein homodimers. As used herein, the terms “homodimer percentage” or “homodimer %” refer to the proportion of fusion protein product (e.g., the insulin-Fc fusion protein described herein) as the desired homodimer. As used herein, the term "homodimer potency" refers to the product of the total protein yield per volume of cell culture reported after the protein A purification step and the homodimer percentage.

[0116] As used herein, the terms "treatment" or "manipulation" for a subject suffering from a disease or disorder refer to administering a treatment regimen to the subject, such as administering a fusion protein like the fusion protein described herein, such that at least one symptom of the disease or disorder is cured, relieved, alleviated, altered, remedied, improved, or aggravated. Treatment includes administering an effective amount to relieve, alleviate, alter, remedy, improve, aggravate, or affect the disease or disorder, or the symptoms of the disease or disorder. The treatment may inhibit the worsening or deterioration of the symptoms of the disease or disorder.

[0117] Insulin-Fc fusion protein components and structure

[0118] This disclosure relates to fusion protein compositions comprising insulin polypeptides linked to species-specific Fc fragments via peptide linkers (i.e., insulin-Fc fusion proteins) and their use in the treatment of diabetes (e.g., in humans and / or dogs). As used herein, the terms "fusion protein" and "insulin-Fc fusion protein" refer to proteins comprising more than one part, for example, from different sources (different proteins, peptides, cells, etc.), covalently linked by peptide bonds. Insulin-Fc fusion proteins are covalently linked as follows: (i) by linking genes encoding each part into a single nucleic acid molecule and (ii) by expressing the protein encoded by the nucleic acid molecule in a host cell (e.g., HEK or CHO): (N-terminus) -- insulin polypeptide -- linker -- Fc fragment -- (C-terminus). A fully recombinant synthetic approach is superior to methods that synthesize insulin polypeptides and Fc fragments separately and then chemically conjugate them. The chemical conjugation step and subsequent purification process increase the complexity of preparation, reduce product yield, and increase cost.

[0119] As used herein, the term "dimer" refers to a protein or fusion protein comprising two covalently linked polypeptides. In some embodiments, two identical polypeptides covalently linked (e.g., via disulfide bonds) form a "homidimer" (in... Figure 1 (Illustrated in the diagram). Disulfide bonds are shown in... Figure 1 In reality, the total number of disulfide bonds can be greater than or less than [the actual number of bonds]. Figure 1 The number shown. In some embodiments, the homodimer is encoded by a single nucleic acid molecule, wherein the homodimer is prepared in a recombinant manner within the cell as described below: first, an insulin-Fc fusion protein monomer is formed, and then further processed within the cell to assemble two identical insulin-Fc fusion protein monomers into a homodimer.

[0120] As used herein, the terms "multimer," "multipolymer," or "multimeric state" refer to a non-covalently associated form of the Fc fusion protein dimer that can be in equilibrium with the Fc fusion protein dimer or can exist as a permanent aggregate form of the Fc fusion protein dimer (e.g., a dimer of the Fc fusion protein homodimer, a trimer of the Fc fusion protein homodimer, a tetramer of the Fc fusion protein homodimer, or a higher-order aggregate containing five or more Fc fusion protein homodimers). It is anticipated that the multimeric form of the Fc fusion protein may possess different physical, stable, or pharmacological activities than the insulin-Fc fusion protein homodimer.

[0121] Insulin peptide

[0122] Insulin polypeptides can be, for example, insulin or insulin analogs produced by β cells in the Langerhans islets of Langerhans within the pancreas. Insulin functions by regulating the uptake of glucose in the blood. Upon stimulation (such as increased protein and glucose levels), insulin is released from β cells and binds to the insulin receptor (IR), initiating a signaling cascade that affects many aspects of metabolism in mammals (e.g., humans). Disruption of this process is directly associated with several diseases, particularly diabetes, insulinoma, insulin resistance, metabolic syndrome, and polycystic ovary syndrome. The insulin analogs disclosed herein may be structurally similar to insulin but contain one or more modifications. In some embodiments, the insulin analogs contain at least one amino acid substitution, deletion, insertion, or chemical modification relative to insulin, which may affect specific characteristics or properties of the insulin-Fc fusion protein conformation. For example, the modifications or alterations described herein, relative to a reference standard, can affect the structure, stability, pH sensitivity, biological activity, or binding affinity to cell surface receptors (e.g., insulin hormone receptors) of the insulin-Fc fusion protein conformation.

[0123] The amino acid sequence of insulin is highly conserved throughout evolution, particularly in vertebrates. For example, natural canine and porcine insulin differs from human insulin by only one amino acid, natural bovine insulin by only three amino acids, and natural feline insulin by only four amino acids. As used herein, the terms "B-chain or B-chain analogue," "C-peptide or "C-chain," and "A-chain or A-chain analogue" refer to peptide segments of the insulin polypeptide, such as... Figure 1 As shown in the diagram. Insulin is a 51-amino acid hormone containing two peptide chains (i.e., the B chain and the A chain) linked by disulfide bonds (e.g., disulfide bonds formed by one or more B-chain cysteine ​​side-chain thiols and one or more A-chain cysteine ​​side-chain thiols). The A chain of insulin is 21 amino acids long, and the B chain of insulin is 30 amino acids long. In the native form of insulin, the A chain contains an intrachain disulfide bond formed by two A-chain cysteine ​​side-chain thiols. For reference purposes, the sequences of the human insulin B chain of SEQ ID NO:1 and the human insulin A chain of SEQ ID NO:2 are shown below:

[0124] FVNQHLCGSHLVEALYLVCGERGFFYTPKT(SEQ ID NO:1)

[0125] GIVEQCCTSICSLYQLENYCN(SEQ ID NO:2).

[0126] As used herein, the term "insulin" or "insulin polypeptide" encompasses mature insulin, proinsulin, proinsulin, and naturally occurring insulin or its analogues. In some embodiments, the insulin polypeptide may be a full-length insulin polypeptide or a fragment thereof. In some embodiments, the insulin polypeptide may comprise one or more fragments derived from mature insulin, proinsulin, proinsulin, or naturally occurring insulin.

[0127] Insulin is typically constructed as an N-terminal B-chain:C-chain:A-chain-C-terminal polypeptide, where the C-chain is cleaved to give it biological activity. For reference purposes, the sequence of the entire human insulin molecule, including the C-chain (i.e., proinsulin), is shown below, with the C-chain underlined:

[0128] FVNQHLCGSHLVEALYLVCGERGFFYTPKT RREAEDLQVGQVELGGGPGAGSLQPLALEGSLQKR GIVEQCCTSICSLYQLENYCN (SEQ ID NO: 3).

[0129] The conversion of a single-chain insulin polypeptide into a biologically active double-chain polypeptide typically occurs within the β cells of Langerhans pancreatic islets, followed by glucose-stimulated insulin secretion by two endonucleases—type I and type II endonucleases. Type I endonucleases PC1 and PC3 disrupt the C-peptide-B chain link, while type II endonucleases PC2 precisely cleave the C-peptide-A chain bond at the correct site. However, this pathway is not available in cellular systems used for the biosynthesis of therapeutic molecules such as insulin (e.g., bacterial, yeast, and mammalian (e.g., HEK and CHO) cell systems), therefore conversion must occur after expression and harvesting of the single-chain polypeptide using chemical or enzymatic methods. All known techniques for cleaving the C-chain after expression and harvest rely on first modifying the C-chain to terminate it at a lysine residue exactly before the N-terminus of the A chain. Then, an enzyme selected from trypsin or the Lys-C family is used, which specifically cleaves the peptide bond at the C-terminus of the lysine residue, cleaving the single-chain insulin polypeptide at the C-terminal lysine of the C-chain and at the 29th C-terminal lysine from the N-terminus of the B chain. In some cases, the resulting bioactive double-chain insulin is used without re-linking the cleaved amino acid at position 30 from the N-terminus of the B chain; in others, an additional enzymatic method is used to add the cleaved amino acid back into the molecule. Such a process is effective for insulin because insulin contains only one lysine in its entire double-chain polypeptide form. However, this process cannot be used for the insulin-Fc fusion protein covered herein because all known Fc fragments contain multiple lysine residues. Therefore, enzymatic cleavage processes digest the Fc fragment into a non-functional portion, thereby eliminating the ability of the Fc fragment to prolong the action of the insulin polypeptide in vivo. Therefore, the insulin-Fc fusion protein of the present invention must contain an insulin polypeptide that does not require C-chain cleavage and is thus bioactive in its single-chain form.

[0130] Many bioactive single-chain insulin polypeptides have been described in the art. In all cases, single-chain insulin polypeptides comprise a C chain of specific length and composition, and A and B chains mutated at specific amino acid sites to achieve electrostatic equilibrium, prevent aggregation, and enhance IR binding and / or downstream signaling to achieve bioactivity comparable to that of native double-chain insulin. In this document, the location of mutations in peptide segments is indicated by the segment name (e.g., B chain, C chain, A chain) and the number of amino acids counted from the N-terminus of the segment. For example, the symbol "B16" refers to the 16th amino acid from the N-terminus of the B chain amino acid sequence. The symbol "A8" refers to the 8th amino acid from the N-terminus of the A chain. Furthermore, if an amino acid is mutated from its native form at a specific position to a new amino acid, that position is appended with a single-letter amino acid code for that new amino acid. For example, B16A refers to an alanine mutation at the 16th amino acid from the N-terminus of the B chain amino acid sequence, and A8H refers to a histidine mutation at the 8th amino acid from the N-terminus of the A chain amino acid sequence.

[0131] US9855318B2 describes a single-chain insulin analog having a C-chain (“first linker”) with the sequence GGSGGGG (SEQ ID NO:72), substitutions in the A-chain, and substitutions and deletions in the B-chain (non-natural amino acids are underlined, and deleted natural amino acids are indicated by an underline Z):

[0132] FVNQHLCGS H LVEAL E LVCGERGF H Y GGGGGGSGGGG GIVEQCCTS T CSL D QLENYC Z (SEQ ID NO:7_NULL).

[0133] The following is a restatement of the sequence shown above, but with the absence of non-existent amino acids indicated by the symbol Z removed from the insulin polypeptide sequence notation. Similarly, as mentioned earlier, non-natural amino acids are underlined. Although there are two different symbols, the paired sequences refer to the exact same insulin polypeptide.

[0134] FVNQHLCGS H LVEAL E LVCGERGF H Y GGGGGGSGGGG GIVEQCCTS T CSL D QLENYC (SEQ ID NO:7)

[0135] In some embodiments, an insulin-Fc fusion protein conformation as described below was unexpectedly found: wherein the alanine at position 16 (i.e., B16) of the N-terminus of the B chain in SEQ ID NO:7 is replaced with glutamate to produce SEQ ID NO:10, resulting in increased homodimer titer, IR binding affinity, and FcRn binding of the insulin-Fc fusion protein, while maintaining reduced immunogenicity, as measured by low Fc(γ)RI and Clq binding affinity. The specific amino acid substitution at B16 was initially prompted by the fact that alanine at that position was known to be less effective at activating insulin-specific T cells (Alleva, DG, Gaur, A., Jin, L., Wegmann, D., Gottlieb, PA, Pahuja, A., Johnson, EB, Motheral, T., Putnam, A., Crowe, PD, Ling, N., Boehme, SA, Conlon, PJ, (2002) Diabetes Vol. 51, No. 7, pp 2126-2134). SEQ ID NO:10 is listed below, with each non-natural amino acid underlined:

[0136] FVNQHLCGS H LVEAL A LVCGERGF H Y GGGGGGSGGGG GIVEQCCTS T CSL D QLENYC (SEQ ID NO:10).

[0137] connector

[0138] In some embodiments, the C-terminus of the insulin polypeptide is directly linked to the N-terminus of the Fc fragment (e.g., no linker or no linker). In other embodiments, successful construction of the recombinantly prepared insulin-Fc fusion protein requires a linker connecting the insulin polypeptide to the Fc fragment. In some embodiments, the insulin-Fc fusion protein conformation described herein includes a peptide linker between the insulin polypeptide and the Fc fragment containing amino acids (e.g., natural or non-natural amino acids). In some embodiments, the peptide linker may be encoded by a nucleic acid molecule, for example, such that a single nucleic acid molecule can encode various peptides within the insulin polypeptide, as well as the peptide linker and the Fc fragment. The choice of peptide linkers (e.g., length, composition, hydrophobicity, and secondary structure) can affect the prepareability (i.e., homodimer titer), chemical and enzymatic stability, biological activity (i.e., NAOC value), biological activity-related parameters (i.e., FcRn EC50 value), and immunogenicity of insulin-Fc fusion protein conformations (Chen, X., Zaro, J., Shen, WC, Adv Drug Deliv Rev. 2013 October 15; 65(10): 1357-1369). Table 1 lists several linkers used to design insulin-Fc fusion protein conformations for the purpose of improving homodimer titer and biological activity.

[0139] Table 1: Peptide linkers between the A chain and Fc fragment in the insulin-Fc fusion protein GGGGAGGGG(SEQ ID NO:11) GGGGSGGGG(SEQ ID NO:12) GGGGGAGGGG(SEQ ID NO:64) GGGGSGGGGSGGGGSGGGG(SEQ ID NO:65) GGGGKGGGGKGGGGKGGGG(SEQ ID NO:66) GGGGGAGGGGAGGGGAGGGGG(SEQ ID NO:67) GGGGGQGGGGQGGGGQGGGGG(SEQ ID NO:13) SGGGGQGGGGQGGGGQGGGGG(SEQ ID NO:68) HGGGGQGGGGQGGGGQGGGGG (SEQ ID NO:69 PGGGGGQGGGGQGGGGQGGGGG(SEQ ID NO:70) GGGGGQGGGGQGGGGQGGGGGQGGGG(SEQ ID NO:99)

[0140] In some embodiments, the peptide linker comprises the sequence: GGGGGQGGGGQGGGGGQGGGG (SEQ ID NO: 13). In other embodiments, the peptide linker comprises the sequence: GGGGSGGGG (SEQ ID NO: 12). In some preferred embodiments, the peptide linker comprises the sequence: GGGGGAGGGGAGGGGAGGGGG (SEQ ID NO: 67) or the sequence: GGGGAGGGG (SEQ ID NO: 11).

[0141] When constructing a recombinant insulin-Fc fusion protein conformation with a peptide linker such as SEQ ID NO:13, care must be taken to avoid undesirable enzymatic cleavage that may occur between the C-terminus of the insulin A chain and the N-terminus of the peptide linker. Cleavage of the linker and Fc fragment from the insulin polypeptide will prevent the insulin-Fc fusion protein conformation from providing a prolonged duration of biological activity. There are known cleavage sites between the asparagine-glycine bond (Vlasak, J., Ionescu, R., (2011) MAbs Vol. 3, No. 3, pp253-263). In many peptide linker embodiments, including the preferred peptide linker of SEQ ID NO:13, the N-terminal amino acid is glycine. Furthermore, the C-terminus of the insulin A chain (i.e., the 21st amino acid from the N-terminus of the A chain (i.e., A21)) is asparagine. Therefore, A21 asparagine was omitted from the insulin peptides of SEQ ID NO:7, SEQ ID NO:9, and SEQ ID NO:10 to eliminate a potentially enzymatically cleavable asparagine-glycine bond that would form between the A chain and the peptide linker in the insulin-Fc fusion protein conformation. Unexpectedly, the insulin-Fc fusion protein conformation constructed from the insulin peptide of SEQ ID NO:8 retained asparagine at the C-terminus of the A chain, demonstrating its ability to be prepared in mammalian cells with: acceptable homodimer potency (i.e., a homodimer potency greater than 50 mg / L for canine insulin-Fc fusion protein), acceptable biological activity in dogs (i.e., a NAOC greater than 150 FBGL%·day·kg / mg), and maintenance of biological activity levels after multiple administrations (i.e., a NAOCR value greater than 0.5 after the third injection in dogs). The results indicate that, contrary to expectations based on previous teachings, there is no risk of enzymatic cleavage or inactivation of insulin-Fc fusion protein conformations containing the Fc fragment sequence disclosed herein, at least for the conformations of insulin-Fc fusion proteins with asparagine-glycine linkages between the insulin peptide and the peptide linker.

[0142] In another embodiment, it was found that for the same insulin polypeptide and Fc fragment composition, mutating glutamine (Q) in SEQ ID NO: 13 to alanine (A) produces a peptide linker of GGGGGAGGGGAGGGGAGGGG (SEQ ID NO: 67), resulting in an insulin-Fc fusion protein conformation with higher homodimer titer, increased binding affinity to IR, and increased binding affinity to FcRn receptor.

[0143] In another embodiment, it was found that for the same insulin peptide and Fc fragment composition, the peptide linker of SEQ ID NO:67 could be shortened without significantly affecting the homodimer titer or binding affinity to the FcRn receptor of the insulin-Fc fusion protein, but with a 60% increase in the IC50 value as determined by IR assay. The shortened peptide linker contains the sequence: GGGGAGGGG (SEQ ID NO:11).

[0144] Fc fragment

[0145] The terms “Fc fragment,” “Fc region,” “Fc domain,” or “Fc polypeptide” are used herein to define the C-terminal region of the immunoglobulin heavy chain. An Fc fragment, region, domain, or polypeptide can be a native sequence Fc region or a variant / mutant Fc region. Although the boundaries of Fc regions of the immunoglobulin heavy chain may vary, they typically contain some or all of the hinge region, CH2 region, and CH3 region of the heavy chain. The hinge region of a canine or human Fc fragment contains an amino acid sequence connecting the CH1 domain and CH2 region of the heavy chain, and this amino acid sequence contains one or more cysteine ​​residues that form one or more interchain disulfide bonds to form a homodimer of the Fc fusion protein from two identical but independent Fc fusion protein monomers. The hinge region may contain all or part of a native amino acid sequence or a non-native amino acid sequence.

[0146] An Fc receptor (FcR) is a receptor that binds to the Fc fragment or Fc region of an antibody. In some embodiments, the FcR is the natural sequence of a canine or human FcR. In some embodiments, the FcR is an FcR (γ receptor) that binds to the Fc fragment or Fc region of an IgG antibody and includes, but is not limited to, receptors of the Fc(γ)RI, Fc(γ)RIIa, Fc(γ)RIIb, and Fc(γ)RIII subclasses, including allelic variants and alternatively spliced ​​forms of these receptors. "FcR" also includes the neonatal receptor FcRn, which is responsible for transferring maternal IgG molecules to the fetus (Guyer et al., 1976 J. Immunol., 117:587; and Kim et al., 1994, J. Immunol., 24:249) and also for prolonging the in vivo elimination half-life of the antibody-Fc fusion protein. Those skilled in the art will understand that mammalian FcRs from one species (e.g., human FcRs) can bind in vitro to insulin-Fc fusion proteins containing Fc fragments from the same species (e.g., human), and sometimes can also bind in vitro to Fc fragments from other mammalian species (e.g., canine). In some embodiments, human FcRs are used in vitro (e.g., in assays) to measure the binding properties of insulin-Fc fusion protein conformations containing human or canine Fc fragments, thereby evaluating their FcR binding properties. In some embodiments, canine FcRs are used in vitro (e.g., in assays) to measure the binding of insulin-Fc fusion protein conformations containing canine Fc fragments.

[0147] In some implementations, the C-terminal lysine (i.e., the lysine representing the last amino acid in the Fc fragment sequence, which is commonly found in the amino acid sequences of Fc fragments of natural canine and human IgG isotypes) is omitted to prevent the accidental generation of undesirable amino acid sequence variants during preparation (e.g., the mixing of Fc fragments containing the C-terminal lysine with Fc fragments that omit the C-terminal lysine, which could occur during the intracellular production of the desired protein (Dick, LW., (2008) Biotechnol Bioeng. Aug 15; 100(6) pp1132-43).

[0148] In embodiments of the canine insulin-Fc fusion protein, the Fc fragment comprises the Fc region (e.g., hinge region, CH2 domain, and CH3 domain) of a canine IgGA Fc fragment (SEQ ID NO:14), a canine IgGB Fc fragment (SEQ ID NO:15), a canine IgGC Fc fragment (SEQ ID NO:16), or a canine IgGD Fc fragment (SEQ ID NO:17). Therefore, in the canine insulin-Fc fusion protein conformation, the canine Fc fragment sequence lacking the C-terminal lysine is:

[0149] RCTDTPPCPVPEPLGGPSVLIFPPKPKDILRITRTPEVTCVVLDLGREDPEVQISWFVDGKEVHTAKTQSREQQFNGTYRVVSVLPIEHQDWLTGKEFKCRVNHIDLPSPIERTISKARGRAHKPSVYVLPPSPKELSSSDTVSITCLIKDFYPPDIDVEWQSNGQQEPERKHRMTPPQLDEDGSYFLYSKLSVDKSRWQQGDPFTCAVMHETLQNHYTDLSLSHSPG(SEQ ID NO:14)

[0150] DCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:15)

[0151] CNNCPCPGCGLLGGPSVFIFPPKPKDILVTARTPTVTCVVVDLDPENPEVQISWFVDSKQVQTANTQPREEQSNGTYRVVSVLPIGHQDWLSGKQFKCKVNNKALPSPIEEIISKTPGQAHQPNVYVLPPSRDEMSKNTVTLTCLVKDFFPPEIDVEWQSNGQQEPESKYRMTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQISLSHSPG(SEQ ID NO:16)

[0152] CISPCPVPESLGGPSVFIFPPKPKDILRITRTPEITCVVLDLGREDPEVQISWFVDGKEVHTAKTQPREQQFNSTYRVVSVLPIEHQDWLTGKEFKCRVNHIGLPSPIERTISKARGQAHQPSVYVLPPSPKELSSSDTVTLTCLIKDFFPPEIDVEWQSNGQPEPESKYHTTAPQLDEDGSYFLYSKLSVDKSRWQQGDTFTCAVMHEALQNHYTDLSLSHSPG (SEQ ID NO:17).

[0153] In the human insulin-Fc fusion protein configuration, the human Fc fragment sequence lacking the C-terminal lysine is:

[0154] DKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO:73)

[0155] ECPPCPAPPVAGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVQFNWYVDGVEVHNAKTKPREEQFNSTFRVVSVLTVVHQDWLNGKEYKCKVSNKGLPAPIEKTISKTKGQPREPQVYTLPPSREEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPMLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO:74).

[0156] For dogs, canine IgGA is preferred to minimize any unwanted immunogenicity because the canine IgGA isotype lacks Fc(γ) effector function (similar to the human IgG2 isotype in humans). However, in the insulin-Fc fusion protein embodiment comprising the insulin peptide of SEQ ID NO:4 and the peptide linker of SEQ ID NO:11, it was unexpectedly found that the insulin-Fc fusion protein conformation comprising the canine IgGA fragment (SEQ ID NO:14) was highly aggregated at a low homodimer titer (i.e., the homodimer titer of the canine insulin-Fc fusion protein was less than 50 mg / L). Furthermore, in dogs, the insulin-Fc fusion protein conformation was biologically inactive (i.e., NAOC value less than 150 FBGL%·day·kg / mg), likely due to its high level of aggregation (e.g., low homodimer %). Despite mutations in the insulin peptide of SEQ ID NO:4, the canine IgGAFc fragment (SEQ ID NO:14), and / or the linker, no insulin-Fc fusion protein embodiment containing the canine IgGA Fc fragment exhibited sufficiently low aggregation levels and sufficiently high potency of the desired homodimer. However, replacing the canine IgGA Fc fragment (SEQ ID NO:14) with the canine IgGB Fc fragment (SEQ ID NO:15) in the insulin-Fc fusion protein conformation yielded a compound with significantly less aggregation at a relatively high potency of the desired homodimer. Furthermore, in dogs, the insulin-Fc fusion protein conformation containing the insulin peptide of SEQ ID NO:4 and the canine IgGB Fc fragment (SEQ ID NO:15) was biologically active, exhibiting glucose-lowering activity over multiple days (i.e., NAOC values ​​greater than 150 FBGL%·day·kg / mg).

[0157] In insulin-Fc fusion protein conformations containing the insulin peptide of SEQ ID NO:7 and the peptide linker of SEQ ID NO:13, the canine IgGB Fc fragment was confirmed to be superior to the canine IgGA Fc fragment, both of which are quite different from the insulin peptide of SEQ ID NO:4 and the peptide linker of SEQ ID NO:11. Insulin-Fc fusion protein conformations containing the insulin peptide of SEQ ID NO:7 and the peptide linker of SEQ ID NO:13 were synthesized using Fc fragments from canine IgGA (SEQ ID NO:14), canine IgGB (SEQ ID NO:15), canine IgGC (SEQ ID NO:16), or canine IgGD (SEQ ID NO:17) immunoglobulins. Using conventional purification methods, only the insulin-Fc fusion protein conformations containing canine IgGA and canine IgGB showed any significant protein yield. However, as previously mentioned, the canine IgGA conformation of the insulin-Fc fusion protein aggregates highly at low bioactivity levels, while the canine IgGB conformation of the insulin-Fc fusion protein exhibits low aggregation (i.e., high homodimer percentage), a high homodimer titer (i.e., homodimer titer of the canine insulin-Fc fusion protein conformation greater than 50 mg / L), and significant long-term hypoglycemic bioactivity in dogs (i.e., NAOC value greater than 150 FBGL%·day·kg / mg). Using an alternative purification method, the canine IgGC conformation of the insulin-Fc fusion protein recovers with low aggregation, but it exhibits the lowest bioactivity in dogs (i.e., NAOC value less than 150 FBGL%·day·kg / mg), presumably due to its low affinity for the FcRn receptor. Therefore, for dog-specific products, regardless of the insulin peptide chosen, canine IgGB (SEQ ID NO: 15) is the preferred Fc fragment for all insulin-Fc fusion protein conformations used in dogs.

[0158] Given that the canine IgGB isotype interacts with the canine Fc(γ) receptor with a higher affinity than the canine IgGA isotype, there is a potential risk of undesirable immunogenicity following repeated injections of the insulin-Fc fusion protein conformation containing canine IgGB. Therefore, various mutations in the canine IgGB Fc fragment have been investigated in an effort to maintain a higher homodimer titer for the insulin-Fc fusion protein while reducing its affinity for the canine Fc(γ)RI receptor.

[0159] One method for reducing Fc(γ)RI interaction involves deglycosylation of the Fc fragment or prevention of Fc fragment glycosylation during the synthesis of the insulin-Fc fusion protein in host cells. Each IgG fragment contains a conserved asparagine (N)-glycosylation site in the CH2 domain of each heavy chain in the Fc region. Here, the symbol used to refer to the conserved N-glycosylation site is "cNg". One method for removing attached glycans from the synthesized insulin-Fc fusion protein is to mutate the cNg site to completely prevent glycan attachment during host cell production. Here, the symbol used to describe the cNg mutation is cNg- (the substituted amino acid). For example, if the asparagine at the cNg site is mutated to serine, the mutation is labeled "cNg-S".

[0160] The absolute position of the cNg site relative to the N-terminus of the B chain in the insulin-Fc fusion protein conformation depends on the length of the insulin polypeptide, the length of the linker, and any omitted amino acids in the Fc fragment preceding the cNg site. In this paper, the symbol used to refer to the absolute position of the cNg site in a given insulin-Fc fusion protein sequence (e.g., measured counting from the N-terminus of the B chain) is “NB (number)”. For example, if the cNg site is located at the 155th amino acid position counting from the N-terminus of the B chain, the absolute position of that site is referred to as “cNg-NB155”. As another example, if the cNg site is located at the 155th amino acid position counting from the N-terminus of the B chain, and the asparagine at that site is mutated to a serine residue, the mutation is labeled as “cNg-NB155-S”.

[0161] In the insulin-Fc fusion protein embodiment containing the insulin peptide of SEQ ID NO:4 and a canine IgGB Fc fragment with cNg-Q, cNg-S, cNg-D, and cNg-K mutations, it was unexpectedly found that only the compounds containing cNg-K and cNg-S mutations exhibited the desired homodimer titer greater than 50 mg / L and the lowest Fc(γ)RI binding affinity for the canine insulin-Fc fusion protein. On the other hand, in the insulin-Fc fusion protein embodiment containing the insulin peptide of SEQ ID NO:7 and a canine IgGB Fc fragment with a cNg-S mutation, it was unexpectedly found that the resulting insulin-Fc fusion protein showed significantly reduced bioactivity in dogs compared to the corresponding insulin-Fc fusion protein conformation containing native canine IgGB Fc (i.e., a significantly reduced NAOC value for insulin-Fc fusion protein counterparts containing native glycosylation site amino acids such as cNg-N). When the B16 amino acid was mutated to alanine, as described above for the insulin polypeptide SEQ ID NO:10, the biological activity of the insulin-Fc fusion protein in the cNg-S mutant was unexpectedly restored (i.e., the NAOC value increased significantly). In summary, there is an unexpected and significant interaction between the choice of cNg mutation on the Fc fragment of the insulin-Fc fusion protein conformation and the composition of the insulin polypeptide, therefore, experiments are needed to identify preferred insulin-Fc fusion protein embodiments. In some specific insulin-Fc fusion protein embodiments, a canine IgGB Fc mutant containing the cNg-S mutation is preferred, and the underlined cNg-S sequence is shown below:

[0162] DCPKCPAPEMLGGPSVFIFPPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQF S GTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG (SEQ ID NO: 18).

[0163] Generally, the human Fc IgG2 isoform is superior to other isoforms because it lacks Fc(γ) effector function and therefore has a lower tendency to induce unwanted immunogenicity. As an illustration, in one implementation of Fc(γ)RI binding and Clq binding ELISA, the insulin-Fc fusion protein embodiment containing the human IgG1 fragment of SEQ ID NO:76 showed an OD450 of 2.078 for human Fc(γ)RI binding at a biotinylated-Fc(γ)RI concentration of 3000 ng / mL and an OD450 of 3.006 for human C1q binding at a biotinylated-C1q concentration of 1000 ng / mL. Such high values ​​suggest the potential for immunogenicity of SEQ ID NO:76 in patients. In contrast, in the same implementation of Fc(γ)RI binding and Clq binding ELISA, the insulin-Fc fusion protein conformation (SEQ ID NO:75) containing the insulin peptide of SEQ ID NO:7, the peptide linker of SEQ ID NO:13, and the human IgG2 Fc fragment (SEQ ID NO:74) showed an OD450 of 0.093 for human Fc(γ)RI binding at a concentration of 3,000 ng / mL and an OD450 of 0.928 for human C1q binding at a concentration of 1,000 ng / mL, indicating that SEQ ID NO:75 is significantly less likely to exhibit immunogenicity in patients.

[0164] For the absolute value of the OD450 measurement of the insulin-Fc fusion protein conformation, this value can differ from one ELISA administration to the next, for example, due to minor variations in the assay administration. This makes comparisons of the absolute OD450 values ​​of different insulin-Fc fusion protein conformations across different ELISA administrations unreliable, even when the concentration of the insulin-Fc fusion protein conformation remains constant during the test. In contrast, for the same two insulin-Fc fusion protein conformations across different ELISA administrations, the ratio of the OD450 measurement of one insulin-Fc fusion protein conformation in one ELISA administration to the OD450 measurement of the second insulin-Fc fusion protein conformation in the same ELISA administration (again, keeping the insulin-Fc fusion protein concentration constant) will be relatively stable. Therefore, the target for the insulin-Fc fusion protein designed for OD450 in both human Fc(γ)RI binding assays and C1q binding assays is expressed as the OD450 ratio, where the ratio is the absolute OD450 value of the analyzed insulin-Fc fusion protein to the absolute OD450 value of the reference insulin-Fc fusion protein conformation, with both measurements performed in the same ELISA assay. For all test samples, the biotinylated-Fc(γ)RI concentration in the insulin-Fc fusion protein conformation sample in the Fc(γ)RI binding ELISA OD450 was set to 3000 ng / mL, and the biotinylated-C1q concentration in the insulin-Fc fusion protein conformation sample in the C1q binding ELISA OD450 was set to 1000 ng / mL.

[0165] The insulin-Fc fusion protein conformation of SEQ ID NO: 76 (containing a human IgG1 Fc fragment) was used as the reference insulin-Fc fusion protein for calculating the OD450 ratio of Fc(γ)RI-bound ELISA and Clq-bound ELISA. The Fc(γ)RI-bound ELISA OD450 ratio of the insulin-Fc fusion protein conformation of SEQ ID NO: 75 (containing a human IgG2 Fc fragment) using the measured OD450 values ​​given above is:

[0166]

[0167] The C1q binding ELISA OD450 ratio of the insulin-Fc fusion protein conformation of SEQ ID NO:75, using the measured OD450 values ​​given above, is:

[0168]

[0169] The Fc(γ)RI binding ELISA OD450 ratio and C1q binding ELISA OD450 ratio of the insulin-Fc fusion protein conformation of SEQ ID NO: 75 (hIgG2), relative to the SEQ ID NO: 76 (hIgG1) conformation, represent an ideal benchmark for creating design targets to evaluate Fc(γ)RI binding and C1q binding of different insulin-Fc fusion protein conformations to reduce undesirable immunogenicity. Therefore, the design target established for human Fc(γ)RI binding (where the biotinylated Fc(γ)RI concentration of the insulin-Fc fusion protein being tested is 3000 ng / mL) is an OD450 ratio <0.50, and the design target established for human C1q binding (where the biotinylated C1q concentration of the insulin-Fc fusion protein being tested is 1000 ng / mL) is an OD450 ratio <0.35.

[0170] The average homodimer potency obtained from two separate synthesis of this insulin-Fc fusion protein conformation (SEQ ID NO: 75) containing the human IgG2 Fc fragment was 117 mg / L.

[0171] By comparison, the insulin-Fc fusion protein embodiment (SEQ ID NO:76), comprising the same insulin polypeptide as SEQ ID NO:7, the same peptide linker as SEQ ID NO:13, and a human IgG1 fragment (SEQ ID NO:73), exhibited an average homodimeric titer of 180 mg / L from two separately synthesized proteins. This is more than 50% higher than the titer obtained by the insulin-Fc fusion protein conformation analog containing the human IgG2 fragment. Given that the human IgG1 isotype interacts with the human Fc(γ)RI receptor with a higher affinity than the human IgG2 isotype, there may be an undesirable risk of immunogenicity after repeated injections of the insulin-Fc fusion protein conformation containing human IgG1.

[0172] Therefore, various mutations in the human IgG1 Fc fragment were investigated in an effort to maintain a larger homodimer titer of the insulin-Fc fusion protein while reducing affinity for the human Fc(γ)RI receptor and C1q.

[0173] As mentioned above, one method for reducing Fc(γ)RI interaction involves deglycosylation of the Fc fragment or prevention of Fc fragment glycosylation during the synthesis of the insulin-Fc fusion protein in host cells. One method for removing the attached glycan from the synthesized insulin-Fc fusion protein is to mutate the cNg site to completely prevent glycan attachment during host cell production.

[0174] The general conformation of the mutant human IgG1 Fc fragment is as follows, with the cNg site underlined:

[0175] DKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY X1 STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:77), wherein instance X1 is S, D, A, R or Q.

[0176] The human IgG1 Fc fragment containing the insulin polypeptide of SEQ ID NO:7, the peptide linker of SEQ ID NO:13, and the human IgG1 Fc fragment of SEQ ID NO:77 is shown below, where X1 is S (i.e., having the cNg-NB155-S mutation):

[0177] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY S STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 91).

[0178] Unexpectedly, the non-glycosylated insulin-Fc fusion protein embodiment of SEQ ID NO:91 produced improved homodimer potency compared to the glycosylated parent material (SEQ ID NO:76). Furthermore, for the insulin-Fc fusion protein embodiment of SEQ ID NO:91 compared to the reference insulin-Fc fusion protein of SEQ ID NO:76, the OD450 ratio measured by human Fc(γ)RI at a biotinylated-Fc(γ)RI concentration of 3000 ng / mL was less than 0.50, and for the insulin-Fc fusion protein conformation of SEQ ID NO:91 compared to the reference insulin-Fc fusion protein of SEQ ID NO:76, the OD450 ratio measured by human C1q binding at a biotinylated C1q concentration of 1000 ng / mL was less than 0.35. However, the non-glycosylated insulin-Fc fusion protein of SEQ ID NO:91 exhibited decreased IR binding (increased IC50 value by IR assay) and decreased FcRn binding affinity (increased EC50 value) compared to its parent glycosylated insulin-Fc fusion protein of SEQ ID NO:76, suggesting that its residence time in vivo and biological activity were likely unacceptably reduced.

[0179] Insulin-Fc fusion protein

[0180] This document provides an insulin-Fc fusion protein conformation comprising an insulin polypeptide, an Fc fragment, and a linker between the insulin polypeptide and the Fc fragment. In some embodiments, the insulin polypeptide comprises a domain in the following direction from N-terminus to C-terminus: (N-terminus)--B chain--C chain--A chain--(C-terminus). In some embodiments, the insulin polypeptide is located on the N-terminal side of the Fc fragment. In some embodiments, the fusion protein comprises a domain in the following direction from N-terminus to C-terminus: (N-terminus)--insulin polypeptide--linker--Fc fragment--(C-terminus) (e.g., (N-terminus))--B chain--C chain--A chain--linker--Fc fragment--(C-terminus)), such as... Figure 1 As shown.

[0181] canine insulin-Fc fusion protein

[0182] In some embodiments, the preferred linker of SEQ ID NO:13 is used to combine the preferred non-glycosylated, bioactive insulin polypeptide of SEQ ID NO:5 with the preferred canine IgGB Fc fragment of SEQ ID NO:15 to produce a family of high homodimer potency-yield, non-aggregating, bioactive, non-immunogenic canine insulin-Fc fusion protein conformations of SEQ ID NO:20, exhibiting a homodimer potency greater than 50 mg / L, a NAOC value greater than 150 FBGL%·day·kg / mg in dogs, and a NAOCR value greater than 0.5 after the third injection in a series of repeated injections in dogs. SEQ ID NO:20 is shown below, where non-natural amino acids are underlined:

[0183] FVNQHLCGS X1 LVEAL E LVCGERGF H Y GGGGGGSGGGG GIVEQCC X2 S T CSL D QLENYC X3 GGGGGQGGGGQGGGGGQGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:20), where X1 is not D, X2 is not H, and X3 is nonexistent or N.

[0184] In the preferred canine insulin-Fc fusion protein embodiment comprising SEQ ID NO:20, X1 is H, X2 is T, and X3 is absent or N. This selection yields a canine insulin-Fc fusion protein conformation of SEQ ID NO:21 with high homodimer potency-yield, non-aggregation, biological activity, and non-immunogenicity, exhibiting a homodimer potency greater than 50 mg / L, a NAOC value greater than 150 FBGL%·day·kg / mg in dogs, and a NAOCR value greater than 0.5 after the third injection in a series of repeated injections in dogs. SEQ ID NO:21 is shown below, where non-natural amino acids are underlined:

[0185] FVNQHLCGSHLVEAL E LVCGERGF H Y GGGGGGSGGGG GIVEQCCTS T CSL D QLENYC X3 GGGGGQGGGGQGGGGQGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVNNKA LPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:21), where X3 does not exist or is N.

[0186] In some preferred embodiments, X3 is absent in SEQ ID NO:21 to produce the canine insulin-Fc fusion protein of SEQ ID NO:26, which exhibits a homodimer titer greater than 50 mg / L, a NAOC value greater than 150 FBGL%·day·kg / mg in dogs, and a NAOCR value greater than 0.5 after the third injection in a series of repeated injections in dogs. SEQ ID NO:26 is shown below, where non-natural amino acids are underlined:

[0187] FVNQHLCGSHLVEAL E LVCGERGF H Y GGGGGGSGGGG GIVEQCCTS T CSL DQLENYCGGGGGQGGGGQGGGGQGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVN NKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:26).

[0188] In some preferred embodiments, X3 in SEQ ID NO:21 is N to produce the canine insulin-Fc fusion protein of SEQ ID NO:28, which exhibits a homodimeric titer greater than 50 mg / L, a NAOC value greater than 150 FBGL%·day·kg / mg in dogs, and a NAOCR value greater than 0.5 after the third injection in a series of repeated injections in dogs. SEQ ID NO:28 is shown below, where non-natural amino acids are underlined:

[0189] FVNQHLCGSHLVEAL E LVCGERGF H Y GGGGGGSGGGG GIVEQCCTS T CSL D QLENYCNGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVN NKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:28).

[0190] In some preferred embodiments, the preferred linker of SEQ ID NO:13 is used to combine the preferred non-glycosylated, cNg-S mutated canine IgGB Fc fragment of SEQ ID NO:18 with the preferred B16A mutated insulin polypeptide sequence of SEQ ID NO:9 to produce a canine insulin-Fc fusion protein of SEQ ID NO:22 with high homodimer potency-yield, non-aggregating, biologically active, and non-immunogenic properties, exhibiting a homodimer potency greater than 50 mg / L, a NAOC value greater than 150 FBGL%·day·kg / mg in dogs, and a NAOCR value greater than 0.5 after the third injection in a series of repeated injections in dogs. SEQ ID NO:22 is shown below, where non-natural amino acids are underlined:

[0191] FVNQHLCGS X1 LVEAL A LVCGERGF H Y GGGGGGSGGGG GIVEQCC X2 S T CSL D QLENYCGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQF S GTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG (SEQ ID NO:22), where X1 is not D and X2 is not H.

[0192] In a preferred embodiment, X1 in SEQ ID NO:22 is H and X2 is T to produce the canine insulin-Fc fusion protein of SEQ ID NO:30, which exhibits a homodimeric titer greater than 50 mg / L, a NAOC value greater than 150 FBGL%·day·kg / mg in dogs, and a NAOCR value greater than 0.5 after the third injection in a series of repeated injections in dogs. SEQ ID NO:30 is shown below, where non-natural amino acids are underlined:

[0193] FVNQHLCGS H LVEALA LVCGERGF H Y GGGGGGSGGGG GIVEQCCTS T CSL D QLENYCGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQF S GTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG (SEQ ID NO: 30).

[0194] An unexpected discovery of the highly prepareable and effective ultra-long-acting canine insulin-Fc fusion protein embodiment of SEQ ID NO:30 led the inventors to attempt to produce a similar prepareable and effective ultra-long-acting insulin-Fc fusion protein conformation for human patients.

[0195] Human insulin-Fc fusion protein

[0196] Further experiments were conducted to determine whether other non-glycosylated insulin-Fc fusion protein conformations exhibited the same behavior. The general conformation of the human IgG1 Fc fragment is shown below, with the cNg site underlined:

[0197] DKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY X1 STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:77), where X1 is S, D, A, R or Q.

[0198] The insulin-Fc fusion protein embodiment comprises the insulin polypeptide of SEQ ID NO:7 (without B16A mutation), the linker of SEQ ID NO:13, and a human IgG1 Fc fragment conformation with cNg mutation selection to prevent glycosylation (SEQ ID NO:77, where X1 is S, D, A, R, or Q), as shown below:

[0199] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY S STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:91)

[0200] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY D STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:92)

[0201] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY A STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:93)

[0202] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY R STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:94)

[0203] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYQSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:95)

[0204] In embodiments containing the insulin polypeptide of SEQ ID NO:7, the preferred linker of SEQ ID NO:13 and the human IgG1 Fc fragment of SEQ ID NO:77 are used, wherein X1 is S (cNg-NB155-S–SEQ ID NO:91), D (cNg-NB155-D-SEQ ID NO:92), A (cNg-NB155-A–SEQ ID NO:93), R (cNg-NB155-R–SEQ ID NO:94) or Q (cNg-NB155-Q–SEQ ID NO:95). Unexpectedly, each of these non-glycosylated insulin-Fc fusion proteins was found to provide improved homodimer potency relative to the glycosylated parent material (SEQ ID NO:76). Furthermore, for each of the deglycosylated insulin-Fc fusion proteins of SEQ ID NO:91, SEQ ID NO:92, SEQ ID NO:93, SEQ ID NO:94, and SEQ ID NO:95, the OD450 ratio of human Fc(γ)RI at a biotinylated-Fc(γ)RI concentration of 3000 ng / mL (wherein the reference insulin-Fc fusion protein used in the OD450 ratio calculation is the insulin-Fc fusion protein of SEQ ID NO:76) was less than 0.50, and the OD450 ratio of human C1q binding at a biotinylated-C1q concentration of 1000 ng / mL (wherein the reference insulin-Fc fusion protein used in the OD450 ratio calculation is the insulin-Fc fusion protein of SEQ ID NO:76) was also less than 0.35. However, compared to the glycosylated parent compound (SEQ ID NO:76), the deglycosylated insulin-Fc fusion proteins of SEQ ID NO:91, SEQ ID NO:92, SEQ ID NO:93, SEQ ID NO:94, and SEQ ID NO:95 all exhibited lower IR binding affinity (higher IR-assay IC50 value) and lower FcRn binding affinity (higher EC50 value), suggesting that these compounds are likely to exhibit unacceptable reductions in in vivo biological activity and residence time. Unexpectedly, in an embodiment containing the insulin peptide of SEQ ID NO:7, using the preferred linker of SEQ ID NO:13 and the human IgG1 Fc fragment of SEQ ID NO:77, where X1 is Q (SEQ ID NO:95), the resulting non-glycosylated insulin-Fc fusion protein homodimer titer was 136 mg / L, thus this insulin-Fc fusion protein conformation did not meet the design target of 150 mg / L homodimer titer.

[0205] Figure 28The diagram shows a side-by-side sequence comparison of SEQ ID NO: 75 (containing natural human IgG2) and SEQ ID NO: 76 (containing natural human IgG1) with SEQ ID NO: 91, SEQ ID NO: 92, SEQ ID NO: 93, SEQ ID NO: 94, and SEQ ID NO: 95, which contain variants of SEQ ID NO: 77. An asterisk (*) indicates complete homology of the entire sequence at a given position, while colons (:), periods (.), or spaces indicate conserved, moderately, or very different amino acid mutations at a given position, respectively.

[0206] Applying the learning from the unexpected results regarding the previously discussed canine insulin-Fc fusion protein conformations, the 16th amino acid (B16) on the B chain of the insulin polypeptide in the insulin-Fc fusion protein embodiments of SEQ ID NO:91, SEQ ID NO:92, SEQ ID NO:93, and SEQ ID NO:94 is mutated to alanine, resulting in SEQ ID NO:78 (with cNg-NB155-S), SEQ ID NO:80 (with cNg-NB155-D), SEQ ID NO:82 (with cNg-NB155-A), and SEQ ID NO:84 (with cNg-NB155-R). When the B16 amino acid was mutated to alanine as described above to obtain the insulin polypeptide of SEQ ID NO: 10, in the insulin-Fc fusion protein conformations of SEQ ID NO: 78, SEQ ID NO: 80, SEQ ID NO: 82 and SEQ ID NO: 84, the acceptable insulin receptor binding was unexpectedly restored (i.e., the IR binding and FcRn receptor binding affinity were significantly increased), without impairing the increase in homodimer titer or the decrease in the binding affinity of Fc(γ)RI and Clq.

[0207] In a preferred embodiment, the human insulin-Fc fusion protein conformation of SEQ ID NO:78 comprises the insulin polypeptide of SEQ ID NO:10, the linker of SEQ ID NO:13, and the cNg-NB155-S mutant of the human IgG1 Fc fragment of SEQ ID NO:77. The insulin-Fc fusion protein of SEQ ID NO:78 exhibits a homodimer titer greater than 150 mg / L. The insulin-Fc fusion protein of SEQ ID NO:78 is shown below, wherein the B16A and cNg-NB155-S mutations are underlined:

[0208] FVNQHLCGSHLVEAL ALVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY S STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 78).

[0209] Similar to the unexpected discovery of the canine insulin-Fc fusion protein conformation, the biological activity of the non-glycosylated insulin-Fc fusion protein embodiment containing the B16A mutation and a human IgG1 Fc fragment with the cNg-NB155-S mutation in SEQ ID NO:78 was unexpectedly restored compared to the biological activity of the non-glycosylated insulin-Fc fusion protein embodiment lacking the B16A mutation in SEQ ID NO:91 (i.e., significantly enhanced IR binding and FcRn receptor binding affinity). The embodiment of the insulin-Fc fusion protein of SEQ ID NO:78 shows an IC50 value of less than 2400 nM, more preferably less than 2000 nM, an EC50 value of less than 1500 ng / L, more preferably less than 1000 ng / mL, an OD450 ratio of less than 0.50 at a biotinylated-Fc(γ)RI concentration of 3,000 ng / mL (wherein the reference insulin-Fc fusion protein used in the OD450 ratio calculation is the insulin-Fc fusion protein of SEQ ID NO:76) at a biotinylated-C1q concentration of 1,000 ng / mL (wherein the reference insulin-Fc fusion protein used in the OD450 ratio calculation is the insulin-Fc fusion protein of SEQ ID NO:76) at a biotinylated-C1q concentration of 1,000 ng / mL (wherein the reference insulin-Fc fusion protein used in the OD450 ratio calculation is the insulin-Fc fusion protein of SEQ ID NO:76) at a biotinylated-C1q concentration of 1,000 ng / mL.

[0210] In a preferred embodiment, the insulin-Fc fusion protein of SEQ ID NO:80 comprises the insulin polypeptide of SEQ ID NO:10, the linker of SEQ ID NO:13, and the cNg-NB155-D mutant of the Fc fragment of SEQ ID NO:77 (X1 is D). The insulin-Fc fusion protein of SEQ ID NO:80 exhibits a homodimer titer greater than 150 mg / L, an IC50 value less than 2400 nM by IR assay, an EC50 value less than 1000 ng / mL by human FcRn assay, an OD450 ratio less than 0.50 by human Fc(γ)RI at a biotinylated-Fc(γ)RI concentration of 3,000 ng / mL, and an OD450 ratio less than 0.35 by human C1q binding assay at a biotinylated-C1q concentration of 1,000 ng / mL. SEQ ID NO:80 is shown below, where the B16A and cNg-NB155-D mutations are underlined:

[0211] FVNQHLCGSHLVEAL A LVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY D STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 80).

[0212] In a preferred embodiment, the insulin-Fc fusion protein of SEQ ID NO:82 comprises the insulin polypeptide of SEQ ID NO:10, the linker of SEQ ID NO:13, and the cNg-NB155-A mutant of the Fc fragment of SEQ ID NO:77 (X1 is A). The insulin-Fc fusion protein of SEQ ID NO:82 exhibits a homodimer titer greater than 150 mg / L, an IC50 value less than 2400 nM by IR assay, an EC50 value less than 1000 ng / mL by human FcRn assay, an OD450 ratio less than 0.50 by human Fc(γ)RI at a biotinylated-Fc(γ)RI concentration of 3,000 ng / mL, and an OD450 ratio less than 0.35 by human C1q binding assay at a biotinylated-C1q concentration of 1,000 ng / mL. SEQ ID NO:82 is shown below, where the B16A and cNg-NB155-A mutations are underlined:

[0213] FVNQHLCGSHLVEAL A LVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY A STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 82).

[0214] In a preferred embodiment, the insulin-Fc fusion protein of SEQ ID NO:84 comprises the insulin polypeptide of SEQ ID NO:10, the linker of SEQ ID NO:13, and the cNg-NB155-R mutant of the Fc fragment of SEQ ID NO:77 (X1 is R). The insulin-Fc fusion protein of SEQ ID NO:84 exhibits a homodimer titer greater than 150 mg / L, an IC50 value less than 2400 nM by IR assay, an EC50 value less than 1000 ng / mL by human FcRn assay, an OD450 ratio less than 0.50 by human Fc(γ)RI at a biotinylated-Fc(γ)RI concentration of 3,000 ng / mL, and an OD450 ratio less than 0.35 by human C1q binding assay at a biotinylated-C1q concentration of 1,000 ng / mL. SEQ ID NO:84 is shown below, where the B16A and cNg-NB155-R mutations are underlined:

[0215] FVNQHLCGSHLVEAL A LVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY R STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 84).

[0216] Similar to SEQ ID NO:78, when the 16th amino acid (B16) on the B chain of the insulin polypeptide is mutated to alanine (resulting in the insulin polypeptide of SEQ ID NO:10), the biological activity of the non-glycosylated insulin-Fc fusion protein conformations (SEQ ID NO:92, SEQ ID NO:93, SEQ ID NO:94, and SEQ ID NO:95) containing human IgG1 FC fragments with cNg-NB155-D, cNg-NB155-A, cNg-NB155-R, and cNg-NB155-Q mutations is restored (i.e., significantly increased IR binding and FcRn receptor binding affinity), without impairing the increase in homodimer titer or the decrease in Fc(γ)RI and C1q binding affinity of the non-glycosylated insulin-Fc fusion protein conformations of SEQ ID NO:92, SEQ ID NO:93, and SEQ ID NO:94 (as previously stated, SEQ ID NO:95) The homodimer titer of the insulin-Fc fusion protein conformation at NO:95 did not meet the design target of 150 mg / L.

[0217] However, the insulin-Fc fusion protein of SEQ ID NO:86, a cNg-NB155-Q mutant containing the insulin peptide of SEQ ID NO:10, the linker of SEQ ID NO:13, and the Fc fragment of SEQ ID NO:77 (X1 is Q), did not exhibit the reduced Fc(γ)RI and C1q binding affinity of the non-glycosylated insulin-Fc fusion protein conformation of SEQ ID NO:95. However, the homodimer potency of the insulin-Fc fusion protein conformation of SEQ ID NO:86 further decreased from that of SEQ ID NO:95 to 111 mg / L, which does not meet the design target of an insulin-Fc fusion protein homodimer potency greater than 150 mg / L. SEQ ID NO:86 is shown below, where the B16A and cNg-NB155-Q mutations are underlined:

[0218] FVNQHLCGSHLVEAL A LVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY QSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 86).

[0219] These results indicate an unexpected and significant interaction between the selection of cNg mutations on the human IgG1 Fc fragment and the composition of the insulin polypeptide, thus requiring further experiments to identify the preferred conformation related to the IR and FcRn binding affinity of the resulting insulin-Fc fusion protein.

[0220] In some configurations, the insulin-Fc fusion protein configuration described herein does not include a leader amino acid sequence at the N-terminus. In other configurations, the insulin-Fc fusion protein configuration described herein includes, for example, a leader sequence at the N-terminus. In some embodiments, an exemplary leader sequence includes the amino acid sequence: MEWSWVFLFFLSVTTGVHS (SEQ ID NO:24). In some embodiments, the insulin-Fc fusion protein configuration described herein is encoded by a nucleic acid molecule containing the leader sequence, for example, for expression (e.g., recombinant expression) in cells (e.g., eukaryotic cells, such as mammalian cells). In some embodiments, the leader sequence is cleaved during expression, for example, in a cell culture. Exemplary nucleic acid sequences encoding the leader sequence include the following nucleic acid sequences:

[0221] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactcc (SEQ ID NO:23). In some embodiments, the exemplary nucleic acid of SEQ ID NO:23 encodes the exemplary leader sequence of SEQ ID NO:24.

[0222] In a preferred embodiment of the insulin-Fc fusion protein conformation including SEQ ID NO:78, the nucleic acid sequence (lead sequence underlined) is:

[0223] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacagcagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:79)。

[0224] In a preferred embodiment including the insulin-Fc fusion protein conformation of SEQ ID NO:80, the nucleic acid sequence (lead sequence underlined) is:

[0225] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacgacagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:81)。

[0226] In the embodiment including the insulin-Fc fusion protein conformation of SEQ ID NO:82, the nucleic acid sequence (lead sequence underlined) is:

[0227] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacgccagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:83)。

[0228] In a preferred embodiment of the insulin-Fc fusion protein conformation including SEQ ID NO:84, the nucleic acid sequence (lead sequence underlined) is:

[0229] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacagaagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:85)。

[0230] In a preferred embodiment including the insulin-Fc fusion protein conformation of SEQ ID NO:86, the nucleic acid sequence (lead sequence underlined) is:

[0231] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtaccaaagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:100)。

[0232] In some applications, covalent modification of insulin-Fc fusion proteins may be necessary to further improve their properties. For example, insulin-Fc fusion proteins can be conjugated to polyethylene glycol (PEG) molecules of varying lengths to further extend their cycling half-life, as described in Dozier, J.; Distefano, M. Site-Specific PEGylation of Therapeutic Proteins. Int. J. Mol. Sci. 2015, 16(10), 25831-25864). One such conjugation method relies on an enzyme called transglutaminase (TGase), which forms a covalent bond via an acyl transfer reaction between a primary amine (e.g., at the end of the PEG molecule) and the carboxylamine group of glutamine (e.g., Q on the target protein, such as Q on the target insulin-Fc fusion protein) (Dozier, J.; Distefano, M., page 25853). The Q at position NB153 of the insulin-Fc fusion proteins of SEQ ID NO:86 and SEQ ID NO:95 is a preferred TGase conjugation site, but because each insulin-Fc fusion protein contains a GGGGGQGGGGGGGGGGGGGGG (SEQ ID NO:13) linker, this linker will also conjugate at the Q residue. Therefore, to minimize unwanted TGase conjugation at the linker Q site, alanine (A) in the linker for glutamine (Q) mutation was tested. Thus, in a preferred embodiment, the insulin-Fc fusion protein of SEQ ID NO:87 comprises the insulin polypeptide of SEQ ID NO:10, a linker of GGGGGAGGGGAGGGGAGGGGGG (SEQ ID NO:67), and a cNg-NB155-S mutant of the Fc fragment of SEQ ID NO:77 (X1 is S). The insulin-Fc fusion protein of SEQ ID NO:87 exhibits a homodimer titer greater than 150 mg / L, an IC50 value less than 2400 nM by IR assay, an EC50 value less than 1000 ng / mL by human FcRn assay, an OD450 ratio less than 0.50 by human Fc(γ)RI at a biotinylated-Fc(γ)RI concentration of 3,000 ng / mL, and an OD450 ratio less than 0.35 by human C1q binding assay at a biotinylated-C1q concentration of 1,000 ng / mL. SEQ ID NO:87 is shown below, with the B16A and cNg-NB155-S mutations and the adapter sequence underlined:

[0233] FVNQHLCGSHLVEAL A LVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYC GGGGGAGGGGA GGGGAGGGGG DKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY S STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 87).

[0234] In the embodiment including the insulin-Fc fusion protein of SEQ ID NO:87, the nucleic acid sequence (lead sequence underlined) is:

[0235] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtgcaggaggcggtggagccggtggaggtggggctggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacagcagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ IDNO:88)。

[0236] It is known that the linker length in Fc fusion protein conjugates affects the properties of the fusion protein, ranging from yield to biological activity. Therefore, conformations of insulin-Fc fusion proteins with shorter linkers were tested to determine any increase or loss in yield or function due to linker length. In another preferred embodiment, the insulin-Fc fusion protein of SEQ ID NO:89 comprises the insulin polypeptide of SEQ ID NO:10, the linker GGGGAGGGG (SEQ ID NO:11), and a cNg-NB155-S mutant of the Fc fragment of SEQ ID NO:77 (X1 is S). The insulin-Fc fusion protein of SEQ ID NO:89 exhibited a homodimer titer greater than 150 mg / L, an IC50 value less than 2400 nM by IR assay, an EC50 value less than 1000 ng / mL by human FcRn assay, an OD450 ratio less than 0.50 by human Fc(γ)RI at a biotinylated-Fc(γ)RI concentration of 3,000 ng / mL, and an OD450 ratio less than 0.35 by human C1q binding assay at a biotinylated-C1q concentration of 1,000 ng / mL.

[0237] The following shows SEQ ID NO:89, where the B16A and cNg-NB155-S mutations and the adapter sequence are underlined:

[0238] FVNQHLCGSHLVEAL A LVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYC GGGGAGGGG DKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY S STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 89).

[0239] In the embodiment including the insulin-Fc fusion protein of SEQ ID NO:89, the nucleic acid sequence (lead sequence underlined) is:

[0240] atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagcacctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtgccggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcataatgccaagacaaagccgcgggaggagcagtacagcagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ ID NO:90)。

[0241] In a generally preferred configuration, a family of high-homogeneity, non-glycosylated, cNg-mutated human IgG1 Fc fragments of SEQ ID NO: 77 (X1 is S, D, A, or R) are combined with a preferred B16A-mutated insulin polypeptide of SEQ ID NO: 10 and various linkers (SEQ ID NO: 11, SEQ ID NO: 13, SEQ ID NO: 67) to produce a family of high homodimer potency-yield, bioactive, non-immunogenic human insulin-Fc fusion protein configurations that are expected to exhibit sufficient in vivo potency and duration of action for long-term once-weekly dosing in diabetic patients.

[0242] insulin-Fc fusion protein production

[0243] In some implementations, such as those described in more detail in the Examples section, the fusion protein can be expressed in cells.

[0244] Expression and purification

[0245] In some embodiments, the insulin-Fc fusion protein can be recombinantly expressed, for example in eukaryotic cells, such as mammalian or non-mammalian cells. Exemplary mammalian cells for expression include HEK cells (e.g., HEK293 cells) or CHO cells. CHO cells can be subdivided into different strains or subclasses (e.g., CHO DG44, CHO-M, and CHO-K1), and some of these cell lines can be genetically engineered to be optimized for use with specific types of nucleic acid molecules (e.g., DNA-containing vectors) or specific cell growth medium compositions described in the Examples section. In some embodiments, cells are transfected with a nucleic acid molecule encoding the insulin-Fc fusion protein (e.g., a vector) (e.g., where the entire insulin-Fc fusion protein is encoded by a single nucleic acid molecule). In some embodiments, HEK293 cells are transfected with a vector encoding the insulin-Fc fusion protein, but only for a period of time (e.g., 3, 4, 5, 7, 10, 12, 14 days or longer) before the host cells cease expressing a significant level of the insulin-Fc fusion protein (i.e., transient transfection). Transient transfection of HEK293 cells with a nucleic acid sequence encoding the insulin-Fc fusion protein typically allows for more rapid production of recombinant proteins, which facilitates the preparation and screening of multiple insulin-Fc fusion protein candidates. In some embodiments, CHO cells are transfected with a vector that permanently incorporates host cell DNA and results in consistent and permanent expression of the insulin-Fc fusion protein (i.e., stable transfection), provided the cells are properly cultured. Stable transfection of CHO cells and CHO cell lines with nucleic acids encoding the insulin-Fc fusion protein typically requires a longer development time, but they generally yield higher total protein yields and are more suitable for preparing low-cost products (e.g., products for the relatively commercialized human insulin market). Cells and cell lines can be cultured using standard methods in the art.

[0246] In some preferred embodiments, the insulin-Fc fusion protein is expressed using HEK cells containing any one of the following cDNA sequences: SEQ ID NO:79 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:78), SEQ ID NO:81 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:80), SEQ ID NO:83 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:82), SEQ ID NO:85 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:84), SEQ ID NO:88 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:87), and SEQ ID NO:90 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:89). In some preferred embodiments, the insulin-Fc fusion protein is expressed using CHO cells containing any one of the following cDNA sequences: SEQ ID NO:79 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:78), SEQ ID NO:81 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:80), SEQ ID NO:83 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:82), SEQ ID NO:85 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:84), SEQ ID NO:88 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:87), and SEQ ID NO:90 (insulin-Fc fusion protein embodiment corresponding to SEQ ID NO:89).

[0247] In some embodiments, the insulin-Fc fusion protein is purified or isolated from cells (e.g., by cell lysis). In other embodiments, the insulin-Fc fusion protein is secreted by cells and purified or isolated from the cell culture medium in which the cells are grown. Purification of the insulin-Fc fusion protein may include using column chromatography (e.g., affinity chromatography) or other separation methods based on differences in the size, charge, and / or affinity of certain molecules. In some embodiments, purification of the insulin-Fc fusion protein involves screening or enriching proteins containing the Fc fragment, for example, by using protein A beads or a protein A column to bind the protein containing the Fc fragment to protein A covalently bound to the protein A beads at a neutral solution pH with high affinity. The bound insulin-Fc fusion protein can then be eluted from the protein A beads by a change in solution variables (e.g., a decrease in solution pH). Other separation methods such as ion exchange chromatography and / or gel filtration chromatography may also be used alternatively or additionally. In some embodiments, purification of the insulin-Fc fusion protein further includes filtering or centrifuging the protein preparation. In some embodiments, further purification of the insulin-Fc fusion protein includes percolation, ultrafiltration, and filtration through porous membranes of various sizes, as well as final formulation using excipients.

[0248] A variety of methods can be used to characterize purified insulin-Fc fusion proteins, such as purity, total protein yield, structure, and / or activity. These methods include absorbance at 280 nm (e.g., for determining total protein yield), size exclusion or capillary electrophoresis (e.g., for determining molecular weight, aggregation percentage, and / or purity), mass spectrometry (MS) and / or liquid chromatography-MS (LC-MS) (e.g., for determining purity and / or glycosylation), and / or ELISA (e.g., for determining the degree of binding to anti-insulin antibodies, such as affinity). Exemplary characterization methods are also described in the Examples section.

[0249] In some embodiments, the total protein yield of the insulin-Fc fusion protein produced in transiently transfected HEK cells and purified from protein A is greater than 5 mg / L, 10 mg / L, or 20 mg / L. In some preferred embodiments, the total protein yield of the human insulin-Fc fusion protein produced in transiently transfected HEK cells and purified from protein A is greater than 100 mg / L (e.g., greater than 150 mg / L). In some embodiments, the homodimer percentage of the insulin-Fc fusion protein produced in transiently transfected HEK cells and purified from protein A is greater than 70% (e.g., greater than 80%, greater than 85%, greater than 90%, greater than 95%, greater than 96%, greater than 97%, greater than 98%, greater than 99%). In some preferred embodiments, the homodimer titer of the human insulin-Fc fusion protein produced in transiently transfected HEK cells and purified from protein A is greater than 100 mg / L (e.g., greater than 150 mg / L), and this homodimer titer is calculated as the product of the total insulin-Fc fusion protein yield and the homodimer percentage. It is considered that only human insulin-Fc fusion protein conformations with a homodimer titer greater than 150 mg / L are suitable for use in this invention because experience suggests that homodimer titers below this level are unlikely to result in commercially viable titers in CHO cells to meet the low production cost requirements of the relatively commercial human insulin market.

[0250] In some embodiments, the total protein yield of insulin-Fc fusion protein produced and purified from protein A in stably transfected CHO cells (e.g., CHO cell lines or CHO cell clones) is greater than 100 mg insulin-Fc fusion protein / L (e.g., mg / L of culture medium). In some preferred embodiments, the total protein yield of insulin-Fc fusion protein produced and purified from protein A in stably transfected CHO cells (e.g., CHO cell lines or CHO cell clones) is greater than 150 mg insulin-Fc fusion protein / L of culture medium (e.g., greater than 200 mg / L, greater than 300 mg / L, greater than 400 mg / L, greater than 500 mg / L, greater than 600 mg / L or more). In some embodiments, the homodimer percentage of insulin-Fc fusion protein produced in stably transfected CHO cells (e.g., CHO cell lines or CHO cell clones) and purified from protein A is greater than 70% (e.g., greater than 80%, greater than 85%, greater than 90%, greater than 95%, greater than 96%, greater than 97%, greater than 98%, greater than 99%). In some embodiments, the homodimer titer of insulin-Fc fusion protein produced in stably transfected CHO cells (e.g., CHO cell lines or CHO cell clones) and purified from protein A is greater than 150 mg / L (e.g., greater than 200 mg / L, greater than 300 mg / L, greater than 400 mg / L, greater than 500 mg / L, greater than 600 mg / L or more), and this homodimer titer is calculated as the product of the total insulin-Fc fusion protein yield and the homodimer percentage.

[0251] Functional characteristics of insulin-Fc fusion protein

[0252] This document describes a method for interacting with insulin receptors to lower blood glucose in a target subject (e.g., a dog or a human), wherein the method includes administering an insulin-Fc fusion protein, such as the fusion protein described herein, to the subject. In some embodiments, the subject has been diagnosed with diabetes (e.g., canine diabetes in dogs, or type 1 or type 2 diabetes in humans).

[0253] In some embodiments, the insulin-Fc fusion protein described herein binds to the IR with significant affinity, as measured by the IC50 (e.g., IC50 < 5000 nM, IC50 < 4000 nM, IC50 < 3000 nM, IC50 < 2400 nM, IC50 < 2000 nM) in the 4°C IM-9 IR binding assay described in Example 12. Based on experience, only compounds exhibiting IR activity with IC50 values ​​less than 5000 nM, preferably less than 2400 nM, and more preferably less than 2000 nM are considered likely to exhibit biological activity in the target subject. Generally, higher affinity IR binding (i.e., lower IC50 values) is preferred. However, it is well known that the clearance of insulin and insulin analogs (e.g., the insulin peptides described herein) is primarily regulated by binding to the IR followed by its internalization and degradation within the cell. Therefore, insulin-Fc fusion protein conformations with excessively high IR binding affinity (i.e., excessively low IC50, such as less than 500 nM IC50) may be cleared from circulation too quickly, resulting in a shorter duration of glucose-reducing biological activity in target subjects than the desired duration.

[0254] In some embodiments, the insulin-Fc fusion protein conformation described herein binds to the FcRn receptor with an affinity higher than that of the insulin-Fc fusion protein reference standard measured according to Example 19. In some embodiments, the EC50 value of the FcRn receptor affinity of the insulin-Fc fusion protein conformation as described herein and measured by a human FcRn receptor assay is less than or equal to 1500 ng / mL, more preferably less than or equal to 1000 ng / mL.

[0255] In some embodiments, the insulin-Fc fusion protein described herein reduces glucose levels (e.g., blood glucose levels) after administration to a subject. In some embodiments, the glucose-reducing activity of the insulin-Fc fusion protein is higher than the glucose-reducing activity of an insulin reference standard. In some embodiments, the duration of activity of the insulin-Fc fusion protein can be measured by a decrease in fasting blood glucose relative to pre-administration fasting blood glucose levels, such as a statistically significant decrease. In some embodiments, the duration of activity of the insulin-Fc fusion protein (e.g., the duration for which a subject's fasting blood glucose level is statistically significantly reduced relative to pre-administration levels) exceeds about 2 hours. In some embodiments, the duration of activity of the insulin-Fc fusion protein (e.g., the duration for which a subject's fasting blood glucose level is statistically significantly reduced relative to pre-administration levels) exceeds about 6 hours, 9 hours, 12 hours, 18 hours, 1 day, 1.5 days, 2 days, 2.5 days, 3 days, 4 days, 5 days, 6 days, 7 days, or longer. In some embodiments, the insulin-Fc fusion protein is long-acting (e.g., having a long half-life, e.g., in serum).

[0256] In some embodiments, the serum half-life of the insulin-Fc fusion protein described herein in the target subject is longer than the serum half-life of an insulin reference standard or control formulation. In some embodiments, the serum half-life of the insulin-Fc fusion protein in the target subject (e.g., in the blood of the subject after administration) exceeds about 2 hours. In some embodiments, the serum half-life of the insulin-Fc fusion protein in the target subject is about 0.5 days, 1 day, 2 days, or 2.5 days. In some preferred embodiments, the serum half-life of the insulin-Fc fusion protein in the target subject is about 3 days or longer.

[0257] In some embodiments, the combination of potency and duration of biological activity of the insulin-Fc fusion protein described herein can be quantified by calculating the area above a curve of fasting blood glucose percentage (FBGL%) normalized to a given dose in mg / kg (NAOC), in FBGL%·day·kg / mg. In some embodiments, the NAOC of the insulin-Fc fusion protein described herein is greater than 150 FBGL%·day·kg / mg (e.g., greater than 200 FBGL%·day·kg / mg, greater than 250 FBGL%·day·kg / mg, or greater). Again, empirically, when the NAOC value is greater than 150 FBGL%·day·kg / mg, the dose requirement for the target subject will be low enough to achieve an acceptable treatment cost. In some embodiments, after repeated dosing in the target subject, the NAOC of the insulin-Fc fusion protein must remain (i.e., the ratio of the NAOC after the third dose of the insulin-Fc fusion protein to the NAOC after the first dose) greater than 0.5 (e.g., greater than 0.6, greater than 0.7, greater than 0.8, greater than 0.9, or greater).

[0258] In some embodiments, the insulin-Fc fusion protein conformation described herein binds to the Fc(γ) receptor with an affinity lower than that of the insulin-Fc fusion protein reference standard as measured according to Example 15. In some embodiments, the ratio of the Fc(γ) receptor affinity of the insulin-Fc fusion protein to the Fc(γ) receptor affinity of the insulin-Fc fusion protein reference standard (i.e., the insulin-Fc fusion protein of SEQ ID NO: 76) is less than 0.50 (e.g., less than 0.40, less than 0.30, less than 0.20). In some embodiments, the Fc(γ) receptor affinity of the insulin-Fc fusion protein is less than or equal to 0.50, as measured by determining the OD450 ratio (relative to the reference insulin-Fc fusion protein of SEQ ID NO: 76) at a human Fc(γ)RI receptor at 3000 ng / mL biotinylated Fc(γ)RI.

[0259] In some embodiments, the insulin-Fc fusion protein conformation described herein binds to C1q with an affinity lower than that of the insulin-Fc fusion protein reference standard as measured according to Example 16. In some embodiments, the ratio of the C1q binding affinity of the insulin-Fc fusion protein to the C1q binding affinity of the insulin-Fc fusion protein reference standard (i.e., the insulin-Fc fusion protein of SEQ ID NO: 76) is less than 0.50 (e.g., less than 0.40, less than 0.30, less than 0.20). In some embodiments, the C1q binding affinity of the insulin-Fc fusion protein is less than or equal to 0.35, as measured by the OD450 ratio (relative to the reference insulin-Fc fusion protein of SEQ ID NO: 76) of human C1q binding at a concentration of 1000 ng / mL biotinylated C1q.

[0260] Characteristics of treatment methods and subject selection

[0261] This article describes a method for treating diabetes (e.g., type 1 or type 2 diabetes in humans) that includes administering an insulin-Fc fusion protein (e.g., the insulin-Fc fusion protein conformation described herein) to a target subject.

[0262] In some embodiments, the reference standard used in any of the methods described herein includes a reference treatment or reference therapy. In some embodiments, the reference includes a standard care agent for the treatment of diabetes (e.g., a standard care agent for canine diabetes, or a standard care agent for human type 1 diabetes, or a standard care agent for human type 2 diabetes). In some embodiments, the reference standard is commercially available insulin or an insulin analogue. In some embodiments, the reference standard includes long-lasting insulin, intermediate-lasting insulin, short-lasting insulin, rapid-acting insulin, short-acting, intermediate-acting, and long-acting insulin. In some embodiments, the reference standard for canine insulin includes... Insulin NPH, Insulin Glargine Or recombinant human insulin. In some implementations, the reference standard for human insulin includes... (Novo Nordisk, Denmark) (Novo Nordisk, Denmark) (Eli Lilly,Indianapolis,IN)、 (Eli Lilly,Indianapolis,IN)、 and Or general recombinant human insulin.

[0263] In some implementations, the reference standards used in any of the methods described herein include the results of diabetes treatments (e.g., canine diabetes treatments or human diabetes treatments), such as the results described herein.

[0264] In some embodiments, the reference standard is the level of a biomarker (e.g., blood glucose or HbA1c) in the target subject prior to the initiation of treatment (e.g., the insulin-Fc fusion protein treatment described herein); wherein the target subject has diabetes. In some embodiments, the blood glucose level in the target subject prior to the initiation of treatment is greater than 200 mg / dL (e.g., greater than 250 mg / dL, 300 mg / dL, 350 mg / dL, 400 mg / dL, or higher). In some embodiments, the fructosamine level in dogs prior to the initiation of treatment is greater than 250 μmol / L, 350 μmol / L (e.g., greater than 400 μmol / L, 450 μmol / L, 500 μmol / L, 550 μmol / L, 600 μmol / L, 650 μmol / L, 700 μmol / L, 750 μmol / L, or higher). In some embodiments, prior to the initiation of treatment, the target human subject has an HbA1c level greater than 7 mmol / L (e.g., greater than 7.5 mmol / L, 8 mmol / L, 9 mmol / L, 10 mmol / L, 11 mmol / L, 12 mmol / L, or higher). In some embodiments, the reference criterion is a measure of the presence, progression, or severity of disease. In some embodiments, the reference criterion is a measure of the presence or severity of disease symptoms prior to the initiation of treatment (e.g., insulin-Fc fusion protein therapy as described herein); wherein the target subject has diabetes.

[0265] Pharmaceutical Compositions and Routes of Administration

[0266] This article provides pharmaceutical compositions comprising the insulin-Fc fusion protein conformation as described herein, which can be used to lower blood glucose in target subjects. The amount and concentration of the insulin-Fc fusion protein in the pharmaceutical composition, and the amount of the pharmaceutical composition administered to the target subject, can be selected based on clinically relevant factors, such as the subject's medical characteristics (e.g., age, weight, sex, other medical conditions), the solubility of the compound in the pharmaceutical composition, the potency and activity of the compound, and the route of administration. For more information on the route of administration and dosage regimen, please refer to Chapter 25.3 in Volume 5 of Comprehensive Medicinal Chemistry (Corwin Hansch; Chairman of Editorial Board), Pergamon Press 1990.

[0267] The formulations disclosed herein include those suitable for parenteral administration. As used herein, the phrases “parenteral administration” and “via parenteral administration” refer to routes of administration other than enteral and local administration, typically via intravenous or subcutaneous injection.

[0268] Examples of suitable aqueous and non-aqueous carriers that can be used in the pharmaceutical compositions of this disclosure include water, ethanol, polyols (such as glycerol, propylene glycol, polyethylene glycol, etc.) and suitable mixtures thereof, vegetable oils such as olive oil, and injectable organic esters such as ethyl oleate. For example, suitable flowability can be maintained by using a coating material such as lecithin to maintain the desired particle size in the case of a dispersion and by using a surfactant such as a Tween-like surfactant. In some embodiments, the pharmaceutical composition (e.g., as described herein) comprises a Tween-like surfactant, such as polysorbate-20, Tween-20, or Tween-80. In some embodiments, the pharmaceutical composition (e.g., as described herein) comprises a Tween-like surfactant, such as Tween-80, at a concentration of about 0.001% to about 2%, or about 0.005% to about 0.1%, or about 0.01% and about 0.5%.

[0269] In some embodiments, the concentration of the insulin-Fc fusion protein in the aqueous carrier is about 3 mg / mL. In some embodiments, the concentration of the insulin-Fc fusion protein in the aqueous carrier is about 6 mg / mL. In some embodiments, the concentration of the insulin-Fc fusion protein in the aqueous carrier is about 8 mg / mL, 9 mg / mL, 10 mg / mL, 12 mg / mL, 15 mg / mL, or higher.

[0270] In some embodiments, the insulin-Fc fusion protein is administered via bolus, infusion, or intravenous bolus. In some embodiments, the fusion protein is administered via syringe injection, pump, pen, needle, or indwelling catheter. In some embodiments, the insulin-Fc fusion protein is administered via subcutaneous bolus injection. The delivery method can also be provided using rechargeable or biodegradable devices. In recent years, various sustained-release polymer devices have been developed and tested in vivo for the controlled delivery of drugs, including protein-based biopharmaceuticals. A variety of biocompatible polymers, including both biodegradable and non-degradable polymers (including hydrogels), can be used to form implants for the sustained release of compounds at specific target sites.

[0271] dose

[0272] The actual dose level of the insulin-Fc fusion protein with the conformation described herein can be varied to obtain the amount of active ingredient required to effectively achieve a desired therapeutic response in a specific target subject (e.g., a dog or a human). The selected dose level will depend on a variety of factors, including the activity of the specific fusion protein used or its esters, salts, or amides, the route of administration, the timing of administration, the excretion rate of the specific compound used, the duration of treatment, other drugs, compounds, and / or materials used in combination with the specific fusion protein used, the age, sex, weight, condition, general health status, and medical history of the subject being treated, and similar factors known in the medical field.

[0273] Generally, the appropriate dose of insulin-Fc fusion protein will be the lowest dose that effectively produces a therapeutic effect. This effective dose typically depends on the factors mentioned above. Generally, the intravenous and subcutaneous dose range of insulin-Fc fusion protein for target subjects is approximately 0.001 to approximately 1 mg per kilogram of body weight per day (e.g., mg / kg), such as approximately 0.001 to 1 mg / kg / day, approximately 0.01 to 0.1 mg / kg / day, approximately 0.1 to 1 mg / kg / day, or approximately 0.01 to 1 mg / kg / day. In other embodiments, the fusion protein is administered at a dose of 0.025 to 4 mg per kilogram of body weight per week, such as 0.025 to 1.0 mg / kg / week.

[0274] This disclosure covers the formulation of insulin-Fc fusion proteins in any of the above-described pharmaceutical compositions and formulations. Furthermore, this disclosure covers administration via any of the above-described routes of administration. Those skilled in the art can select appropriate formulations and routes of administration based on the condition being treated and the overall health, age, and body type of the subject receiving treatment.

[0275] Example

[0276] The present technology is further illustrated by the following examples and should not be construed as limiting it in any way.

[0277] General methods, determinations and materials

[0278] Example 1: Method and synthesis of insulin-Fc fusion protein in HEK293 cells

[0279] The insulin-Fc fusion protein was synthesized as described below. The target gene sequence was constructed using specialized software (LakePharma, Belmont, CA) and cloned into a high-expression mammalian vector. HEK293 cells were seeded in shake flasks 24 hours prior to transfection and cultured in a serum-free, chemically defined medium. The DNA expression construct encoding the target insulin-Fc fusion protein was transiently transfected into the HEK293 cell suspension using standard operating procedures for transient transfection (LakePharma, Belmont, CA). After 20 hours, cell counts were performed to determine viability and live cell count, and the results were analyzed using… (Pall FortéBio LLC, Fremont, CA) Measure titer. Obtain additional readings throughout the transient transfection production run. Harvest cultures on or after day 5.

[0280] Example 2: Method and Synthesis of Insulin-Fc Fusion Protein in CHO Cells

[0281] The CHO cell line was originally derived from CHO-K1 (LakePharma, Belmont, CA), with the endogenous glutamine synthase (GS) gene knocked out using recombinant technology employing methods known in the art. Stable expression DNA vectors were designed and optimized for CHO expression and GS selection, and integrated into a high-expression mammalian vector (LakePharma, Belmont, CA). The sequence of each completed construct was confirmed before initiating amplification experiments. CHO cells adapted to suspension were cultured in a chemically defined medium (CD OptiCHO; Invitrogen, Carlsbad, CA) at 37°C under humidified 5% CO2 conditions. No serum or other animal-derived products were used in the culture of CHO cells.

[0282] use The system (MaxCyte, Inc., Gaithersburg, MD) was used to transfect approximately 80 million suspension-adapted CHO cells cultured in CD OptiCHO medium during the exponential growth phase with 80 μg of DNA via electroporation to generate stable CHO cell lines (DNA constructs containing the full-length sequence of the insulin-Fc fusion protein) for each insulin-Fc fusion protein. After 24 hours, the transfected cells were counted and selected for stable integration of the insulin-Fc fusion gene. The transfected cells were then cultured at a density of 0.5 × 10⁻⁶ cells / year. 6Cells were seeded at a density of [number] cells / mL into shake flasks in CD OptiCHO selective medium containing 0-100 μM methionine sulfonamide (MSX) and incubated at 37°C and 5% CO2. During selection, cells were centrifuged and resuspended in fresh selective medium every 2 to 3 days until the CHO stable pool recovered its growth rate and viability. Cell culture growth and titer were monitored.

[0283] Culture the cells to 2.5 × 10⁻⁶. 6 Cells / mL. Viability was above 95% at cell banking. Cells were then centrifuged, and the cell pellet was resuspended in CD OptiCHO medium containing 7.5% dimethyl sulfoxide (DMSO) until a cell count of 15 × 10⁶ cells / mL was achieved. 6 Cells / mL / vial. Store vials frozen in liquid nitrogen.

[0284] Small-scale production was performed using CHO cells as described below. Cells were expanded for production at 37°C in CDOptiCHO growth medium containing 100 μM MSX, fed every 2 to 4 days as needed, with glucose and additional amino acids added to the CDOptiCHO growth medium as needed, and maintained for approximately 14 to 21 days. The conditioned medium supernatant harvested from the stable tank production run was clarified by centrifugation. The protein was passed through a Protein A (MabSelect, GE Healthcare, Little Chalfont, UK) column pre-equilibrated with binding buffer. The wash buffer was then passed through the column until the OD280 value (NanoDrop, Thermo Scientific) was measured to be at or near the background level. The insulin-Fc fusion protein was eluted using a low-pH buffer, the elution fractions were collected, and the OD280 value of each fraction was recorded. The fractions containing the target insulin-Fc fusion protein were combined and optionally further filtered using a 0.2 μM membrane filter.

[0285] Optionally, the cell line is further subcloned into single clones, and optionally, limiting dilution (a method known to those skilled in the art) is used to further select clones expressing high-titer insulin-Fc fusion protein. After obtaining a cell line expressing a high-titer monoclonal insulin-Fc fusion protein, insulin-Fc fusion protein production is performed as described above in MSX-free growth medium or optionally in MSX-containing growth medium to obtain a cell culture supernatant containing recombinant, CHO-prepared insulin-Fc fusion protein. Optionally, the MSX concentration is increased over time to impart additional selectivity to clones capable of producing higher product titers.

[0286] Example 3: Method and Synthesis of Insulin-Fc Fusion Protein in CHO Cells

[0287] The CHO cell line was originally derived from CHO-K1 (LakePharma, Belmont, CA), with the endogenous glutamine synthase (GS) gene knocked out using recombinant technology employing methods known in the art. Stable expression DNA vectors were designed and optimized for CHO expression and GS selection, and integrated into a high-expression mammalian vector (LakePharma, Belmont, CA). The sequence of each completed construct was confirmed before initiating amplification experiments. CHO cells adapted to suspension were cultured in a chemically defined medium (CD OptiCHO; Invitrogen, Carlsbad, CA) at 37°C under humidified 5% CO2 conditions. No serum or other animal-derived products were used in the culture of CHO cells.

[0288] use The system (MaxCyte, Inc., Gaithersburg, MD) was used to transfect approximately 80 million suspension-adapted CHO cells cultured in CD OptiCHO medium during the exponential growth phase with 80 μg of DNA via electroporation to generate stable CHO cell lines (DNA constructs containing the full-length sequence of the insulin-Fc fusion protein) for each insulin-Fc fusion protein. After 24 hours, the transfected cells were counted and selected for stable integration of the insulin-Fc fusion gene. The transfected cells were then cultured at a density of 0.5 × 10⁻⁶ cells / year. 6 Cells were seeded at a density of [number] cells / mL into shake flasks in CD OptiCHO selective medium containing 0-100 μM methionine sulfonamide (MSX) and incubated at 37°C and 5% CO2. During selection, cells were centrifuged and resuspended in fresh selective medium every 2 to 3 days until the CHO stable pool recovered its growth rate and viability. Cell culture growth and titer were monitored.

[0289] Culture the cells to 2.5 × 10⁻⁶. 6 Cells / mL. Viability remained above 95% at cell banking. Cells were then centrifuged, and the cell pellet was resuspended in CDOptiCHO medium containing 7.5% dimethyl sulfoxide (DMSO) until a cell count of 15 × 10⁶ cells / mL was achieved. 6 Cells / mL / vial. Store vials frozen in liquid nitrogen.

[0290] Small-scale production was performed using CHO cells as described below. Cells were expanded for production at 37°C in CDOptiCHO growth medium containing 100 μM MSX, fed every 2 to 4 days as needed, with glucose and additional amino acids added to the CDOptiCHO growth medium as needed, and maintained for approximately 14 to 21 days. The conditioned medium supernatant harvested from the stable tank production run was clarified by centrifugation. The protein was passed through a pre-equilibrated Protein A (MabSelect, GE Healthcare, Little Chalfont, UK) column with binding buffer. The wash buffer was then passed through the column until the OD280 value (NanoDrop, Thermo Scientific) was measured to be at or near the background level. The insulin-Fc fusion protein was eluted using a low-pH buffer, the elution fractions were collected, and the OD280 value of each fraction was recorded. The fractions containing the target insulin-Fc fusion protein were combined and optionally further filtered using a 0.2 μM membrane filter.

[0291] Optionally, the cell line is further subcloned into single clones, and optionally, limiting dilution (a method known to those skilled in the art) is used to further select clones expressing high-titer insulin-Fc fusion protein. After obtaining a cell line expressing a high-titer monoclonal insulin-Fc fusion protein, insulin-Fc fusion protein production is performed as described above in MSX-free growth medium or optionally in MSX-containing growth medium to obtain a cell culture supernatant containing recombinant, CHO-prepared insulin-Fc fusion protein. Optionally, the MSX concentration is increased over time to impart additional selectivity to clones capable of producing higher product titers.

[0292] Example 4: Purification of insulin-Fc fusion protein.

[0293] Purification of the insulin-Fc fusion protein was performed as follows. Conditioned medium supernatant containing secreted insulin-Fc fusion protein was harvested from transiently transfected HEK, stably transfected HEK, or stably transfected CHO production runs and clarified by centrifugation. The supernatant containing the desired insulin-Fc fusion protein was passed through a Protein A column and washed with various wash buffers, including 0.15–0.50 M sodium chloride, followed by elution with a low pH solution. The elution fractions containing the desired protein were then combined, and the buffer was replaced with 200 mM HEPES, 100 mM NaCl, 50 mM NaOAc, pH 7.0 buffer. A final filtration step was performed using a 0.2 μm membrane filter. The final protein concentration was calculated from the optical density of the solution at 280 nm. Further optional purification was performed, if necessary, by ion exchange chromatography (e.g., using anion exchange beads or cation exchange beads), gel filtration chromatography, or other methods.

[0294] Example 5: Purification of insulin-Fc fusion protein.

[0295] Purification of the insulin-Fc fusion protein was performed as follows. Conditioned medium supernatant containing secreted insulin-Fc fusion protein was harvested from transiently transfected HEK, stably transfected HEK, or stably transfected CHO production runs and clarified by centrifugation. The supernatant containing the desired insulin-Fc fusion protein was passed through a Protein A column and washed with various wash buffers, including 0.15–0.50 M sodium chloride, followed by elution with a low pH solution. The elution fractions containing the desired protein were then combined, and the buffer was replaced with 200 mM HEPES, 100 mM NaCl, 50 mM NaOAc, pH 7.0 buffer. A final filtration step was performed using a 0.2 μm membrane filter. The final protein concentration was calculated from the optical density of the solution at 280 nm. Further optional purification was performed, if necessary, by ion exchange chromatography (e.g., using anion exchange beads or cation exchange beads), gel filtration chromatography, or other methods.

[0296] Example 6: Structure confirmed by non-reduction and reduction CE-SDS.

[0297] exist Purified insulin-Fc fusion protein solutions dissolved in 200 mM HEPES, 100 mM NaCl, 50 mM NaOAc, pH 7.0 buffer were analyzed by capillary electrophoresis with sodium dodecyl sulfate (CE-SDS) in a GXII (Perkin Elmer, Waltham, MA) apparatus, and electrophoretic patterns were plotted. Under non-reducing conditions, samples were run according to protein standards of known molecular weight (MW), and the elution peaks represented the "apparent" MW of the insulin-Fc fusion protein homodimer.

[0298] Under reducing conditions (e.g., by using β-mercaptoethanol to break the disulfide bonds of the insulin-Fc fusion homodimer), comparing the apparent molecular weight (MW) of the resulting insulin-Fc fusion protein monomer with half the molecular weight of the insulin-Fc fusion protein homodimer may be accurate in determining the structural purity of the insulin-Fc fusion protein.

[0299] Example 7: Structure confirmed by non-reduction and reduction CE-SDS.

[0300] exist Purified insulin-Fc fusion protein was analyzed by capillary electrophoresis with sodium dodecyl sulfate (CE-SDS) in a GXII (Perkin Elmer, Waltham, MA) buffer solution containing 200 mM HEPES, 100 mM NaCl, 50 mM NaOAc, pH 7.0, and an electrophoresis pattern was plotted. Under non-reducing conditions, samples were run according to protein standards of known molecular weight (MW), and the elution peaks represented the "apparent" molecular weight (MW) of the insulin-Fc fusion protein homodimer.

[0301] Under reducing conditions (e.g., by using β-mercaptoethanol to break the disulfide bonds of the insulin-Fc fusion homodimer), comparing the apparent molecular weight (MW) of the resulting insulin-Fc fusion protein monomer with half the molecular weight of the insulin-Fc fusion protein homodimer may be accurate in determining the structural purity of the insulin-Fc fusion protein.

[0302] Example 8: Sequence identification by LC-MS using polysaccharide removal from polysaccharide-containing compounds.

[0303] To accurately estimate the mass of the insulin-Fc fusion protein by mass spectrometry (MS) in cases of native glycosylation, the sample was first treated to remove native glycans that might interfere with MS analysis. First, 100 μL of 2.5 mg / mL insulin-Fc fusion protein buffer, dissolved in 200 mM HEPES, 100 mM NaCl, 50 mM NaOAc, pH 7.0 buffer, was replaced with 0.1 M Tris, pH 8.0 buffer containing 5 mM EDTA, using a Zeba desalting column (Pierce, ThermoFisher Scientific, Waltham, MA). Then, 1.67 μL of PNGase F enzyme (Prozyme N-glycanase) was added to this solution to remove N-linked glycans (e.g., glycans linked to the asparagine side chain at the cNg-N site) present in the fusion protein, and the mixture was incubated overnight at 37°C. The sample was then analyzed by LC-MS (NovaBioassays, Woburn, MA) to obtain the molecular weight corresponding to the desired homodimer without the glycan. This mass was then further corrected because the enzymatic process used to cleave the glycan from cNg-asparagine also deaminates the asparagine side chain to form aspartic acid, and in doing so, the enzymatically treated homodimer as a whole gains 2 Da, corresponding to a mass of 1 Da per chain present in the homodimer. Therefore, the actual molecular weight is the measured mass minus 2 Da to correct for the enzymatic modification of the insulin-Fc fusion protein structure in the analyzed sample.

[0304] Accurate estimates of the mass of the insulin-Fc fusion protein can be obtained directly using LC-MS (NovaBioassays, Woburn, MA) without molecule pretreatment with PNGase enzyme, provided that the amino acid composition of the insulin-Fc fusion protein prevents native glycosylation at the cNg site.

[0305] Example 9: Sequence identification by LC-MS using polysaccharide removal from polysaccharide-containing compounds.

[0306] To accurately estimate the mass of the insulin-Fc fusion protein by mass spectrometry (MS) in cases of native glycosylation, the sample was first treated to remove native glycans that might interfere with MS analysis. First, 100 μL of 2.5 mg / mL insulin-Fc fusion protein buffer, dissolved in 200 mM HEPES, 100 mM NaCl, 50 mM NaOAc, pH 7.0 buffer, was replaced with 0.1 M Tris, pH 8.0 buffer containing 5 mM EDTA, using a Zeba desalting column (Pierce, ThermoFisher Scientific, Waltham, MA). Then, 1.67 μL of PNGase F enzyme (Prozyme N-glycanase) was added to this solution to remove N-linked glycans (e.g., glycans linked to the asparagine side chain at the cNg-N site) present in the fusion protein, and the mixture was incubated overnight at 37°C. The sample was then analyzed by LC-MS (NovaBioassays, Woburn, MA) to obtain the molecular weight corresponding to the desired homodimer without the glycan. This mass was then further corrected because the enzymatic process used to cleave the glycan from cNg-asparagine also deaminates the asparagine side chain to form aspartic acid, and in doing so, the enzymatically treated homodimer as a whole gains 2 Da, corresponding to a mass of 1 Da per chain present in the homodimer. Therefore, the actual molecular weight is the measured mass minus 2 Da to correct for the enzymatic modification of the insulin-Fc fusion protein structure in the analyzed sample.

[0307] When the amino acid composition of the insulin-Fc fusion protein prevents native glycosylation at the cNg site, an accurate estimate of the mass of the insulin-Fc fusion protein can be obtained directly using LC-MS (NovaBioassays, Woburn, MA) without molecule pretreatment with PNGase enzyme.

[0308] Example 10: Obtaining homodimer by size exclusion chromatography.

[0309] Size exclusion chromatography (SEC-HPLC) of insulin-Fc fusion protein was performed at 280 nm using a Waters 2795HT HPLC (Waters Corporation, Milford, MA) connected to a 2998 photodiode array. 100 μL or less of the sample containing the target insulin-Fc fusion protein was injected into a MAbPac SEC-1, 5 μm, 4 × 300 mm column (ThermoFisher Scientific, Waltham, MA) at a flow rate of 0.2 mL / min. The mobile phase consisted of 50 mM sodium phosphate, 300 mM NaCl, and 0.05% w / v sodium azide at pH 6.2. The MAbPac SEC-1 column operates based on molecular size separation principles. Therefore, larger soluble insulin-Fc aggregates (e.g., multimers of the insulin-Fc fusion protein homodimer) eluted at earlier cutoff times, while unaggregated homodimers eluted at later cutoff times. The purity of the insulin-Fc fusion protein solution was determined as a percentage of non-aggregated homodimers when the homodimer mixture was separated from aggregated polydimers by analytical SEC-HPLC.

[0310] Example 11: Obtaining homodimer by size exclusion chromatography.

[0311] Size exclusion chromatography (SEC-HPLC) of the insulin-Fc fusion protein was performed at 280 nm using a Waters 2795HT HPLC (Waters Corporation, Milford, MA) connected to a 2998 photodiode array. 100 μL or less of the sample containing the target insulin-Fc fusion protein was injected into a MAbPac SEC-1, 5 μm, 4 × 300 mm column (ThermoFisher Scientific, Waltham, MA) at a flow rate of 0.2 mL / min. The mobile phase consisted of 50 mM sodium phosphate, 300 mM NaCl, and 0.05% w / v sodium azide at pH 6.2. The MAbPac SEC-1 column operates based on molecular size separation principles. Therefore, larger soluble insulin-Fc aggregates (e.g., multimers of the insulin-Fc fusion protein homodimer) eluted at earlier cutoff times, while unaggregated homodimers eluted at later cutoff times. The purity of the insulin-Fc fusion protein solution was determined as a percentage of non-aggregated homodimers by separating the homodimer mixture from aggregated polydimers using analytical SEC-HPLC.

[0312] Example 12: In vitro IM-9 insulin receptor (IR) binding of an exemplary insulin-Fc fusion protein at 4°C.

[0313] Human IM-9 cells (ATTC#CCL-159) expressing human IR were cultured in complete RPMI 5% FBS medium and maintained at 70-80% confluence. The IM-9 cell culture was centrifuged at 250×g (~1000rpm) for 10 minutes to pellet the cells. The cells were washed once with HBSS or PBS buffer and resuspended in cold FACS staining medium (HBSS / 2mM EDTA / 0.1% sodium azide + 4% horse serum) to a concentration of 8×10⁻⁶. 6 Cells / mL, and keep on ice or at 4°C until the test solution is prepared. Dilute insulin-Fc protein to 2× concentration in FACS buffer at a 1:3 ratio in 1.2 mL tubes (each dilution volume is approximately 60 μL), and keep the solution chilled on ice until ready for pipetting.

[0314] Biotinylated RHI was diluted to a concentration of 1.25 μg / mL in FACS staining medium. 40 μL of serially diluted test compound and 8 μL of 1.25 μg / mL biotin-RHI were added to each well of a V-bottom microtiter plate, mixed by slow vortexing, and placed on ice. Then, 40 μL of IM-9 cell suspension (8 × 10⁸ cells / well) was pipetted using a multichannel pipette. 6 Cells / mL were added to each well, gently mixed again, and incubated on ice for 30 min to allow competitive IR binding on IM-9 cells. Cells were then washed twice with 275 μL of ice-cold FACS wash buffer (HBSS / 2 mM EDTA / 0.1% sodium azide + 0.5% horse serum) by centrifugation at 3000 rpm for 3 min on a V-bottom plate and aspiration of the supernatant. Cells were then resuspended on ice in 40 μL of FACS staining medium containing 1:100 diluted Streptavidin-PE (Life Technologies) for 20 min. Cells were then washed once with 275 μL of ice-cold FACS buffer and finally fixed with 3% paraformaldehyde for 10 min at room temperature. Cells were then washed once with 275 μL of ice-cold FACS buffer and resuspended in 250 μL of FACS buffer for analysis.

[0315] Cell-containing V-shaped substrates were then analyzed using a Guava 8-HT flow cytometer (Millipore). For each concentration of the test compound, biotinylated RHI binding to IR was quantified by median fluorescence intensity (MFI) of cells in the FACS FL-2 channel. Control wells were labeled with biotinylated-RHI only and used to calculate the percentage inhibition (%) at each test compound concentration. The percentage inhibition of biotinylated RHI binding on IM-9 cells by the test compound was plotted against the logarithmic concentration of the test compound, and the resulting IC50 value of the test compound was calculated using GraphPad Prism (GraphPad Software, La Jolla, CA). Therefore, a lower IC50 value for the test compound indicates a higher level of biotinylated-RHI inhibition at lower concentrations, suggesting a stronger binding of the insulin-Fc fusion protein to IR. Control compounds such as unlabeled recombinant human insulin (RHI) are also used as internal standards to generate RHI IC50, which yields the ratio of IC50 of a given compound to RHI IC50 (IC50(compound) / IC50(RHI)). Lower IC50 ratios indicate a more similar binding to RHI (stronger binding to IR), while higher IC50 ratios indicate a weaker binding to IR relative to RHI.

[0316] Example 13: In vitro IM-9 insulin receptor (IR) binding of an exemplary insulin-Fc fusion protein at 4°C.

[0317] Human IM-9 cells (ATTC#CCL-159) expressing human IR were cultured in complete RPMI 5% FBS medium and maintained at 70-80% confluence. The IM-9 cell culture was centrifuged at 250×g (~1000rpm) for 10 minutes to pellet the cells. The cells were washed once with HBSS or PBS buffer and resuspended in cold FACS staining medium (HBSS / 2mM EDTA / 0.1% sodium azide + 4% horse serum) to a concentration of 8×10⁻⁶. 6 Cells / mL, and keep on ice or at 4°C until the test solution is prepared. Dilute insulin-Fc protein to 2× concentration in FACS buffer at a 1:3 ratio in 1.2 mL tubes (each dilution volume is approximately 60 μL), and keep the solution chilled on ice until ready for pipetting.

[0318] Biotinylated RHI was diluted to a concentration of 1.25 μg / mL in FACS staining medium. 40 μL of serially diluted test compound and 8 μL of 1.25 μg / mL biotin-RHI were added to each well of a V-bottom microtiter plate, mixed by slow vortexing, and placed on ice. Then, 40 μL of IM-9 cell suspension (8 × 10⁸ cells / well) was pipetted using a multichannel pipette. 6Cells / mL were added to each well, gently mixed again, and incubated on ice for 30 min to allow competitive IR binding on IM-9 cells. Cells were then washed twice with 275 μL of ice-cold FACS wash buffer (HBSS / 2 mM EDTA / 0.1% sodium azide + 0.5% horse serum) by centrifugation at 3000 rpm for 3 min on a V-bottom plate and aspiration of the supernatant. Cells were then resuspended on ice in 40 μL of FACS staining medium containing 1:100 diluted Streptavidin-PE (Life Technologies) for 20 min. Cells were then washed once with 275 μL of ice-cold FACS buffer and finally fixed with 3% paraformaldehyde for 10 min at room temperature. Cells were then washed once with 275 μL of ice-cold FACS buffer and resuspended in 250 μL of FACS buffer for analysis.

[0319] Cell-containing V-shaped substrates were then analyzed using a Guava 8-HT flow cytometer (Millipore). For each concentration of the test compound, biotinylated RHI binding to IR was quantified by median fluorescence intensity (MFI) of cells in the FACS FL-2 channel. Control wells were labeled with biotinylated-RHI only and used to calculate the percentage inhibition (%) at each test compound concentration. The percentage inhibition of biotinylated RHI binding on IM-9 cells by the test compound was plotted against the logarithmic concentration of the test compound, and the resulting IC50 value of the test compound was calculated using GraphPad Prism (GraphPad Software, La Jolla, CA). Therefore, a lower IC50 value for the test compound indicates a higher level of biotinylated-RHI inhibition at lower concentrations, suggesting stronger binding of the insulin-Fc fusion protein to IR. Control compounds such as unlabeled recombinant human insulin (RHI) are also used as internal standards to generate RHI IC50, which yields the ratio of IC50 of a given compound to RHI IC50 (IC50(compound) / IC50(RHI)). Lower IC50 ratios indicate a more similar binding to RHI (stronger binding to IR), while higher IC50 ratios indicate a weaker binding to IR relative to RHI.

[0320] Example 14: Determination of in vitro human Fc(γ)RI binding affinity.

[0321] The binding of insulin-Fc fusion protein to Fc(γ)RI was performed using an ELISA assay at pH 7.4 using human Fc(γ)RI (i.e., rhFc(γ)RI). Insulin-Fc fusion protein was diluted to 10 μg / mL in sodium bicarbonate buffer at pH 9.6 and plated onto Maxisorp (Nunc) microtiter plates overnight at 4°C. The microplates were then washed five times with PBST buffer (PBS / 0.05% Tween-20) and blocked with Superblock blocking reagent (ThermoFisher). Serial dilutions of biotinylated rhFc(γ)RI (recombinant human Fc(γ)RI; R&D Systems) from 3000 ng / mL to 4.1 ng / mL were prepared in PBST / 10% Superblock buffer and loaded at 100 μL / well onto the microplates coated with insulin-Fc fusion protein. The microtiter plate was incubated at room temperature for 1 hour, then the microplate strips were washed 5 times with PBST. 100 μL of streptavidin-HRP diluted 1:10000 in PBST / 10% Superblock buffer was then loaded into each well. After incubation for 45 minutes, the microplate strips were washed 5 more times with PBST. TMB was added to visualize the bound Fc(γ)RI protein, and the ELISA was terminated with an ELISA termination reagent (Boston Bioproducts). The plate was read at 450 nm in an ELISA reader, and the OD450 values ​​(proportional to the binding of rhFc(γ)RI to insulin-Fc protein) were plotted against the logarithmic concentration of rhFc(γ)RI added to each well to generate binding curves using GraphPad Prism software. For groups of compounds with somewhat similar curves, the OD450 at one of the higher concentrations, such as 3000 ng / mL of rhFc(γ)RI, could be used to identify differences between coated insulin-Fc fusion protein compounds. To compare the differences between various insulin-Fc fusion proteins running at different times, the OD450 ratio of human Fc(γ)RI was calculated as the OD450 value of the test insulin-Fc fusion protein compound obtained at a biotinylated-Fc(γ)RI concentration of 3000 ng / mL divided by the OD450 value of the reference insulin-Fc fusion protein of SEQ ID NO:76 obtained at a biotinylated-Fc(γ)RI concentration of 3000 ng / mL.

[0322] Example 15: Determination of in vitro human Fc(γ)RI binding affinity.

[0323] The binding of insulin-Fc fusion protein to Fc(γ)RI was performed using an ELISA assay at pH 7.4 using human Fc(γ)RI (i.e., rhFc(γ)RI). Insulin-Fc fusion protein was diluted to 10 μg / mL in sodium bicarbonate buffer at pH 9.6 and plated onto Maxisorp (Nunc) microtiter plates overnight at 4°C. The microplates were then washed five times with PBST buffer (PBS / 0.05% Tween-20) and blocked with Superblock blocking reagent (ThermoFisher). Serial dilutions of biotinylated rhFc(γ)RI (recombinant human Fc(γ)RI; R&D Systems) from 3000 ng / mL to 4.1 ng / mL were prepared in PBST / 10% Superblock buffer and loaded at 100 μL / well onto the microplates coated with insulin-Fc fusion protein. The microtiter plate was incubated at room temperature for 1 hour, then the microplate strips were washed 5 times with PBST. 100 μL of streptavidin-HRP diluted 1:10000 in PBST / 10% Superblock buffer was then loaded into each well. After incubation for 45 minutes, the microplate strips were washed 5 more times with PBST. TMB was added to visualize the bound Fc(γ)RI protein, and the ELISA was terminated with an ELISA termination reagent (Boston Bioproducts). The plate was read at 450 nm in an ELISA reader, and the OD450 values ​​(proportional to the binding of rhFc(γ)RI to insulin-Fc protein) were plotted against the logarithmic concentration of rhFc(γ)RI added to each well to generate binding curves using GraphPad Prism software. For groups of compounds with somewhat similar curves, the OD450 at one of the higher concentrations, such as 3000 ng / mL of rhFc(γ)RI, could be used to identify differences between coated insulin-Fc fusion protein compounds. To compare the differences between various insulin-Fc fusion proteins running at different times, the OD450 ratio of human Fc(γ)RI was calculated as the OD450 value of the test insulin-Fc fusion protein compound obtained at a biotinylated-Fc(γ)RI concentration of 3000 ng / mL divided by the OD450 value of the reference insulin-Fc fusion protein of SEQ ID NO:76 obtained at a biotinylated-Fc(γ)RI concentration of 3000 ng / mL.

[0324] Example 16: In vitro C1q binding affinity determination

[0325] The binding of insulin-Fc fusion protein to complement component C1q was performed using an ELISA assay at pH 7.4 as described below. Insulin-Fc fusion protein was diluted to 10 μg / mL in sodium bicarbonate buffer (pH 9.6) and plated onto Maxisorp (Nunc) microtiter plates overnight at 4°C. The microplates were then washed five times with PBST buffer (PBS / 0.05% Tween-20) and blocked with Superblock blocking reagent (ThermoFisher). Serial dilutions of biotinylated complement component C1q (human complement component C1q; Sigma-Aldrich) from 1000 ng / mL to 1.4 ng / mL were prepared in PBST / 10% Superblock buffer and loaded at 100 μL / well onto the microplates coated with insulin-Fc fusion protein. The microtiter plate was incubated at room temperature for 1 hour, then the microplate strips were washed 5 times with PBST. 100 μL / well of streptavidin-HRP diluted 1:12000 in PBST / 10% Superblock buffer was then loaded. After incubation for 45 minutes, the microplate strips were washed 5 more times with PBST. TMB was added to visualize the bound complement C1q protein, and the ELISA was terminated with an ELISA termination reagent (Boston Bioproducts). The absorbance (OD450) of the plate was read at 450 nm using an ELISA reader, and the OD450 values ​​(proportional to the binding of complement component C1q to insulin-Fc protein) were plotted against the logarithmic concentration of complement component C1q added to each well to generate binding curves using GraphPad Prism software. For groups of compounds with somewhat similar curves, the differences between coated insulin-Fc fusion protein compounds could be identified using the OD450 at one of the higher concentrations, such as 1000 ng / mL of complement component C1q. To further compare the differences between various insulin-Fc fusion proteins running at different times, the human C1q assay OD450 ratio was calculated as the OD450 of the test insulin-Fc fusion protein compound obtained at a biotinylated-C1q concentration of 1000 ng / mL divided by the OD450 of the reference insulin-Fc fusion protein of SEQ ID NO:76 obtained at a biotinylated-C1q concentration of 1000 ng / mL.

[0326] Example 17: In vitro C1q binding affinity determination

[0327] The binding of insulin-Fc fusion protein to complement component C1q was performed using an ELISA assay at pH 7.4 as described below. The insulin-Fc compound was diluted to 10 μg / mL in sodium bicarbonate buffer (pH 9.6) and coated onto Maxisorp (Nunc) microtiter plates overnight at 4°C. The microplates were then washed five times with PBST buffer (PBS / 0.05% Tween-20) and blocked with Superblock blocking reagent (ThermoFisher). Serial dilutions of biotinylated complement component C1q (human complement component C1q; Sigma-Aldrich) ranging from 1000 ng / mL to 1.4 ng / mL were prepared in PBST / 10% Superblock buffer and loaded at 100 μL / well onto the microplates coated with the insulin-Fc fusion protein. The microtiter plate was incubated at room temperature for 1 hour, then the microplate strips were washed 5 times with PBST. 100 μL / well of streptavidin-HRP diluted 1:12000 in PBST / 10% Superblock buffer was then loaded. After incubation for 45 minutes, the microplate strips were washed 5 more times with PBST. TMB was added to visualize the bound complement C1q protein, and the ELISA was terminated with an ELISA termination reagent (Boston Bioproducts). The absorbance (OD450) of the plate was read at 450 nm using an ELISA reader, and the OD450 values ​​(proportional to the binding of complement component C1q to insulin-Fc protein) were plotted against the logarithmic concentration of complement component C1q added to each well to generate binding curves using GraphPad Prism software. For groups of compounds with somewhat similar curves, the differences between coated insulin-Fc fusion protein compounds could be identified using the OD450 at one of the higher concentrations, such as 1000 ng / mL of complement component C1q. To further compare the differences between various insulin-Fc fusion proteins running at different times, the human C1q assay OD450 ratio was calculated as the OD450 of the test insulin-Fc fusion protein compound obtained at a biotinylated-C1q concentration of 1000 ng / mL divided by the OD450 of the reference insulin-Fc fusion protein of SEQ ID NO:76 obtained at a biotinylated-C1q concentration of 1000 ng / mL.

[0328] Example 18: In vitro measurement of the affinity of insulin-Fc fusion protein for canine FcRn receptor.

[0329] The in vitro binding affinity of an insulin-Fc fusion protein containing a canine IgG-derived Fc fragment to the canine FcRn receptor was measured using an ELISA technique performed at a solution pH 5.5. A slightly acidic pH is a preferred binding environment for Fc-fragment-containing molecules to bind to the FcRn receptor. In vivo, cells express FcRn on their surface and inside endosomes. When Fc-fragment-containing molecules enter cells via natural processes such as pinocytosis or endocytosis, the pH in the endosome becomes lower, where the FcRn receptor binds to the Fc-fragment-containing molecules, which would otherwise be degraded in the endosome-lysosome compartment, allowing these molecules to recycle back to the cell surface at a more neutral pH (e.g., pH 7.0–7.4). Neutral pH is unfavorable for binding to the FcRn receptor and allows the release of Fc-fragment-containing molecules back into circulation. This is a major mechanism by which Fc-fragment-containing molecules exhibit a prolonged circulating pharmacokinetic half-life in vivo.

[0330] The insulin-Fc fusion protein containing the canine Fc fragment was diluted to 10 μg / ml in pH 9.6 sodium bicarbonate buffer and coated in duplicate onto Maxisorb ELISA strips at room temperature for 1 to 2 hours. The strips were then washed four times with PBST (PBS / 0.1% Tween-20) buffer and blocked with Superblock blocking reagent (ThermoFisher). The strips for FcRn binding were then washed twice more with pH 5.5 MES / NaCl / Tween (50 mM MES / 150 mM NaCl / 0.1% Tween-20) buffer, and then FcRn reagent (biotinylated canine FcRn; Immunitrack) was added. Prepare serial dilutions (1:3X) of biotinylated FcRn reagent at concentrations ranging from 1000 ng / mL to 0.45 ng / mL in pH 5.5 MES / NaCl / Tween / 10% Superblock buffer, and load 100 μL / well onto strips coated with the insulin-Fc fusion protein compound using a multichannel pipette. Incubate the assay plate at room temperature for 1 hour. Wash the FcRn binding strips four times with pH 5.5 MES / NaCl / Tween buffer, then load 100 μL / well of streptavidin-HRP diluted 1:10000 in pH 5.5 MES / NaCl / 10% Superblock buffer. After incubation for 45 minutes, wash the strips four more times with pH 5.5 MES / NaCl / Tween buffer. Finally, add TMB to visualize the bound biotinylated canine FcRn reagent, and stop the colorimetric reaction with an ELISA stop reagent. The plate was read at a wavelength of 450 nm in an ELISA reader. The OD values ​​(proportional to the binding of canine-FcRn to the insulin-Fc fusion protein test compound) were plotted against the logarithmic concentration of FcRn added to each well to generate binding curves using GraphPadPrism software. The EC50 value for each binding curve was calculated to compare different compounds.

[0331] Example 19: In vitro measurement of the affinity of the insulin-Fc fusion protein for the human FcRn receptor.

[0332] The in vitro binding affinity of an insulin-Fc fusion protein containing a human IgG-derived Fc fragment to the human FcRn receptor was measured using an ELISA technique performed at a solution pH 5.5. A slightly acidic pH is a preferred binding environment for Fc-fragment-containing molecules to bind to the FcRn receptor. In vivo, cells express FcRn on their surface and inside endosomes. When Fc-fragment-containing molecules enter cells via natural processes (e.g., pinocytosis or endocytosis), the pH in the endosome becomes lower, where the FcRn receptor binds to the Fc-fragment-containing molecules, which would otherwise be degraded in the endosome-lysosome compartment, allowing these molecules to recycle back to the cell surface at a more neutral pH (e.g., pH 7.0–7.4). Neutral pH is unfavorable for binding to the FcRn receptor and allows the release of Fc-fragment-containing molecules back into circulation. This is a major mechanism by which Fc-fragment-containing molecules exhibit a prolonged circulating pharmacokinetic half-life in vivo.

[0333] The insulin-Fc fusion protein containing the Fc fragment was diluted to 10 μg / mL in pH 9.6 sodium bicarbonate buffer and coated in duplicate onto Maxisorb ELISA strips at room temperature for 1 to 2 hours. The strips were then washed five times with PBST (PBS / 0.1% Tween-20) buffer and blocked with Superblock blocking reagent (ThermoFisher). The strips used for FcRn binding were then washed three times with pH 5.5 MES / NaCl / Tween (50 mM MES / 150 mM NaCl / 0.1% Tween-20) buffer, and then FcRn reagent (biotinylated human FcRn; Immunitrack) was added. Prepare serial dilutions (1:3X) of biotinylated FcRn reagent at concentrations ranging from 6000 ng / mL to 8.23 ​​ng / mL in pH 5.5 MES / NaCl / Tween / 5% Superblock buffer, and load 100 μL / well onto strips coated with the insulin-Fc fusion protein compound using a multichannel pipette. Incubate the assay plate at room temperature for 1.5 h. Wash the FcRn binding strips four times with pH 5.5 MES / NaCl / Tween buffer, then load 100 μL / well of streptavidin-HRP diluted 1:10000 in pH 5.5 MES / NaCl / 5% Superblock buffer. After incubation for 45 min, wash the strips five more times with pH 5.5 MES / NaCl / Tween buffer. Finally, add TMB to visualize the bound biotinylated FcRn reagent, and stop the colorimetric development with an ELISA stop reagent. The plate was read at a wavelength of 450 nm in an ELISA reader. The OD450 values ​​(proportional to the binding of human-FcRn to the insulin-Fc fusion protein test compound) were plotted against the logarithmic concentration of FcRn added to each well to generate binding curves using GraphPad Prism software. The EC50 value for each binding curve was calculated to compare different compounds.

[0334] Example 20: In vitro measurement of the affinity of the insulin-Fc fusion protein for the human FcRn receptor.

[0335] The in vitro binding affinity of an insulin-Fc fusion protein containing a human IgG-derived Fc fragment to the human FcRn receptor was measured using an ELISA technique performed at a solution pH 5.5. A slightly acidic pH is a preferred binding environment for Fc-fragment-containing molecules to bind to the FcRn receptor. In vivo, cells express FcRn on their surface and inside endosomes. When Fc-fragment-containing molecules enter cells via natural processes such as pinocytosis or endocytosis, the pH in the endosome becomes lower, where the FcRn receptor binds to the Fc-fragment-containing molecules, which would otherwise be degraded in the endosome-lysosome compartment, allowing these molecules to recycle back to the cell surface at a more neutral pH (e.g., pH 7.0–7.4). Neutral pH is unfavorable for binding to the FcRn receptor and allows the release of Fc-fragment-containing molecules back into circulation. This is a major mechanism by which Fc-fragment-containing molecules exhibit a prolonged circulating pharmacokinetic half-life in vivo.

[0336] The insulin-Fc fusion protein containing the Fc fragment was diluted to 10 μg / mL in pH 9.6 sodium bicarbonate buffer and coated in duplicate onto Maxisorb ELISA strips at room temperature for 1 to 2 hours. The strips were then washed five times with PBST (PBS / 0.1% Tween-20) buffer and blocked with Superblock blocking reagent (ThermoFisher). The strips used for FcRn binding were then washed three times with pH 5.5 MES / NaCl / Tween (50 mM MES / 150 mM NaCl / 0.1% Tween-20) buffer, and then FcRn reagent (biotinylated human FcRn; Immunitrack) was added. Prepare serial dilutions (1:3X) of biotinylated FcRn reagent at concentrations ranging from 6000 ng / mL to 8.23 ​​ng / mL in pH 5.5 MES / NaCl / Tween / 5% Superblock buffer, and load 100 μL / well onto strips coated with the insulin-Fc fusion protein compound using a multichannel pipette. Incubate the assay plate at room temperature for 1.5 hours. Wash the FcRn binding strips four times with pH 5.5 MES / NaCl / Tween buffer, then load 100 μL / well of streptavidin-HRP diluted 1:10000 in pH 5.5 MES / NaCl / 5% Superblock buffer. After incubation for 45 minutes, wash the strips five more times with pH 5.5 MES / NaCl / Tween buffer. Finally, add TMB to visualize the bound biotinylated FcRn reagent, and stop the colorimetric development with an ELISA stop reagent. The plate was read at a wavelength of 450 nm in an ELISA reader. The OD450 values ​​(proportional to the binding of human-FcRn to the insulin-Fc fusion protein test compound) were plotted against the logarithmic concentration of FcRn added to each well to generate binding curves using GraphPad Prism software. The EC50 value for each binding curve was calculated to compare different compounds.

[0337] Example 21: General procedures for determining in vivo drug efficacy (PD) after a single dose of canine insulin Fc fusion protein in dogs Procedure.

[0338] The effect of insulin-Fc fusion proteins on fasting blood glucose levels in dogs was determined as follows. N=1, 2, 3, or more healthy, antibody-naïve dogs weighing approximately 10 to 15 kg were used, with each insulin-Fc fusion protein administered to one dog. Animals were also observed twice daily for signs of allergic reactions, lethargy, distress, pain, etc., and optionally, for some compounds, subcutaneous injection was continued for an additional three weeks or longer to observe whether the glucose-lowering capacity of the compounds decreased over time, a key marker of potential induction of neutralizing anti-drug antibodies. On day 0, animals are given a single injection of a drug composition, which comprises insulin Fc fusion protein homodimer at a concentration of 1 to 10 mg / mL in a solution of 10 to 50 mM sodium hydrogen phosphate, 50 to 150 mM sodium chloride, 0.005 to 0.05% v / v Tween-80, and optionally 0.02 to 1.00 mg / mL of an antibacterial agent (e.g., phenol, m-cresol, or methylparaben), at a pH between 7.0 and 8.0, at a dose of 0.08 to 0.80 mg insulin-Fc fusion protein / kg (or approximately equivalent to 1.2 to 12.3 nmol / kg or approximately equivalent to 0.4 to 4.0 U / kg insulin equivalent, in moles). Blood was collected from a suitable vein on day 0, immediately before injection, and at 15, 30, 45, 60, 120, 240, 360, and 480 minutes after injection, and on days 1, 2, 3, 4, 5, 6, and 7.

[0339] At each time point, collect at least 1 mL of whole blood. Use a glucose meter ( AvivaPlus immediately determines glucose level readings, requiring approximately one drop of blood. Mean fasting blood glucose levels (FBGL%) from day 0 to day 7 are plotted to evaluate the biological activity of the given insulin-Fc fusion protein.

[0340] Example 22: General procedure for determining in vivo drug efficacy (PD) after repeated administration of canine insulin Fc fusion protein to dogs. Procedure.

[0341] The effect of insulin-Fc fusion protein on fasting blood glucose levels was assessed after repeated injections as described below. Healthy, antibody-naïve dogs weighing approximately 10-20 kg were administered a dose of insulin-Fc fusion protein to each animal. Animals were observed twice daily for signs of allergic reactions, lethargy, distress, pain, and other adverse side effects. Optionally, for some compounds, treatment was continued with up to two to five additional subcutaneous injections to observe whether the compound's hypoglycemic effect decreased over time, indicating the possible presence of neutralizing anti-drug antibodies. On day 0, animals are given a single subcutaneous injection of a drug composition comprising insulin Fc fusion protein in a solution of 10 to 50 mM sodium hydrogen phosphate, 50 to 150 mM sodium chloride, 0.005 to 0.05% v / v Tween-80, and optionally 0.02 to 1.00 mg / mL of an antibacterial agent (e.g., phenol, m-cresol, or methylparaben), at a pH between 7.0 and 8.0, at a dose of 0.08 to 0.80 mg insulin-Fc fusion protein / kg (or approximately equivalent to 1.2 to 12.3 nmol / kg or approximately equivalent to 0.4 to 4.0 U / kg insulin equivalents, in molars). Blood is collected from a suitable vein immediately before injection and at 15, 30, 45, 60, 120, 240, 360, and 480 minutes after injection, and on days 1, 2, 3, 4, 5, 6, and 7.

[0342] Subsequent subcutaneous injections are administered no more than once a week, and in some cases, at different time intervals depending on the efficacy of the given insulin-Fc fusion protein formulation. Subsequent injections of each insulin-Fc fusion protein are adjusted to higher or lower doses based on proven efficacy. For example, if the first injection on day 0 is found to be ineffective in lowering blood glucose, the dose level of subsequent insulin-Fc fusion protein injections is increased. Similarly, if the first injection on day 0 is found to lower glucose too strongly, the dose level of subsequent insulin-Fc fusion protein injections is decreased. It has also been found that intermediate or final doses can be adjusted in a similar manner as needed. For each dose, blood is collected from a suitable vein immediately before injection and at 15, 30, 45, 60, 120, 240, 360, and 480 minutes after injection, and on days 1, 2, 3, 4, 5, 6, and 7 (optionally 14 days). At least 1 mL of whole blood is collected at each time point. A glucose meter is used. The Aviva Plus immediately determines glucose level readings, requiring approximately one drop of blood. Plotting the mean fasting blood glucose percentage (FBGL%) over time throughout the study allows for the determination of the fusion protein's biological activity.

[0343] To determine the bioactivity of each dose, an area above the curve (AOC) analysis was performed as described below. After constructing the FBGL% data relative to time, the data was then entered into data analysis software (GraphPad Prism, GraphPad Software, San Diego, CA). The software was first used to perform an area below the curve (AUC) to integrate the area under the curve for each dose's FBGL% relative to time. To convert the AUC data into the desired AOC data, the following equation was used: AOC = TPA – AUC; where TPA is the total possible area obtained by multiplying each dose lifetime (e.g., 7 days, 14 days, etc.) by 100% (where 100% represents y = 100% of the FBGL% relative to time curve). For example, given a 7-day dose lifetime and a calculated AUC of 500 FBGL%·days, the AOC was calculated as follows: AOC = (100 FBGL% x 7 days) – (500 FBGL%·days) = 200 FBGL%·days. Each injection dose in a series of injection doses is analyzed to obtain the AOC value for injection 1, injection 2, injection 3, etc.

[0344] Since the dosage of insulin-Fc fusion protein can vary as previously mentioned, it is generally more convenient to normalize all calculated AOC values ​​for a given insulin-Fc fusion protein to a specific dose of that insulin-Fc fusion protein. This facilitates comparisons of the glycemic efficacy of insulin-Fc fusion protein over multiple injections, even if dose levels vary within a given study. The normalized AOC (NAOC) for a given dose is calculated as follows: NAOC = AOC / D, in units of FBGL%·days·kg / mg; where D is the actual dose injected into the animal, in mg / kg. The NAOC value is calculated for each injection in a series of injections for a given animal and averaged across groups of animals receiving the same insulin-Fc fusion protein formulation.

[0345] For a given animal, the NAOC ratio (NAOCR) of each injection in a series of injections is calculated by obtaining the NAOC value for each injection (e.g., injections 1, 2, 3, ... N) and dividing each NAOC of a given injection by the NAOC from injection 1, as follows: NAOCR = (NAOC(Nth injection) / NAOC(first injection)). By evaluating the NAOCR of a given insulin-Fc fusion protein formulation injected Nth in a series of injections, it is possible to determine whether the in vivo glucose-lowering activity of the given insulin-Fc fusion protein is substantially maintained after a series of N administrations (e.g., NAOCR after Nth administration is greater than 0.5) or whether the in vivo glucose-lowering activity of the given insulin-Fc fusion protein has lost most of its potency after a series of N administrations (e.g., NAOCR after Nth administration is less than 0.5), indicating the possible formation of neutralizing antidrug antibodies in vivo. In some preferred embodiments, the ratio of the NAOC after the third subcutaneous injection to the NAOC after the first subcutaneous injection is greater than 0.5 (i.e., the NAOCR of the third subcutaneous injection is greater than 0.5).

[0346] Example 23: General procedures for determining the in vivo pharmacokinetics (PK) of canine insulin Fc fusion protein in canine serum Procedure.

[0347] An assay for measuring the concentration of insulin-Fc fusion protein containing the canine isotype Fc fragment in canine serum was constructed as described below. The assay comprises a sandwich ELISA format, in which the therapeutic compound in the serum sample is captured by an anti-insulin / proinsulin monoclonal antibody (mAb) coated onto an ELISA plate, and then detected by an HRP-conjugated anti-canine IgG Fc-specific antibody followed by color development using a TMB substrate system. Maxisorp ELISA plates (Nunc) were coated overnight at 4°C with the anti-insulin mAb clone D6C4 (Biorad) in coating buffer (pH 9.6 sodium carbonate-biocarbonate buffer) at a concentration of 5 μg / ml. The plates were then washed five times with PBST (PBS + 0.05% Tween 20) and blocked with SuperBlock blocking solution (ThermoFisher) for at least one hour at room temperature (or overnight at 4°C). Serum samples were diluted to a 1:20 ratio in PBST / SB / 20% HS sample dilution buffer (PBS + 0.1% Tween 20 + 10% SuperBlock + 20% horse serum). To prepare a standard curve, the target insulin-Fc fusion protein was serially diluted 1:2.5 in sample dilution buffer (PBST / SB / 20% HS) + 5% combined beagle serum (BioIVT) to concentrations ranging from 200 ng / ml to 0.82 ng / ml. Standards and diluted serum samples were added in duplicate (100 μl / well) to the blocking plate and incubated at room temperature for 1 hour. After incubation, the samples and standards were washed five times with PBST. The HRP-conjugated goat anti-dog IgG Fc (Sigma) detection antibody was diluted to approximately 1:15,000 in PBST / SB / 20% HS buffer, and 100 μl was added to all wells and incubated at room temperature in the dark for 45 minutes. The plate was washed five times with PBST and once with deionized water, and then developed at room temperature for 8 to 10 minutes by adding 100 μl / well TMB (Invitrogen). Development was then stopped by adding 100 μl / well ELISA stop solution (Boston Bioproducts), and absorbance was read at 450 nm using a SpectraMax plate reader (Molecular Devices) within 30 minutes. The concentration of the insulin-Fc fusion protein compound in the samples was calculated by interpolation on the 4-PL curve using SoftMaxPro software.

[0348] Example 24: Determination Procedure for Measuring Antidrug Antibodies in Canine Serum

[0349] To measure ADA against the test compound, Maxisorp ELISA plates (Nunc) were coated overnight with the target insulin-Fc fusion protein diluted 10 μg / mL in coating buffer (pH 9.6 carbonate-biocarbonate buffer) at 4°C. To measure ADA against the insulin moiety of the insulin-Fc fusion protein containing the Fc fragment of canine IgG, the plates were coated with 30 μg / mL purified insulin in coating buffer. The plates were then washed five times with PBST (PBS + 0.05% Tween 20) and blocked with SuperBlock blocking solution (Thermo Fisher, Waltham MA) for at least 1 hour (or overnight). To calculate ADA in canine IgG units, strips of canine IgG (Jackson Immunoresearch Laboratories, West Grove PA) serially diluted 1:2 in Carb-Biocarb coating buffer (pH 9.6) were coated overnight at 4°C and used to create a 7-point pseudo-standard curve. Similarly, wash and seal the standard strips with SuperBlock sealing solution for at least 1 hour (or overnight).

[0350] Dilute the test serum sample to a concentration greater than or equal to 1:100 (usually 1:200 for testing) in PBST / SB / 20% HS sample dilution buffer (PBS + 0.1% Tween 20 + 10% SuperBlock + 20% horse serum) and add 100 μL / well in duplicate to strips coated with insulin-Fc fusion protein (or RHI). Also add two copies of the standard strip coated with canine IgG to each plate, and fill each well with PBST / SB (PBS + 0.1% Tween 20 + 10% SuperBlock) buffer. Incubate the plates at room temperature for 1 hour, then wash 5 times with PBST. To detect ADA, HRP-conjugated goat anti-cat IgG F(ab')2 (anti-cat IgG F(ab')2 reagent cross-reacts with canine antibody; Jackson Immunoresearch Laboratories, West Grove PA) was diluted to 1:10000 in PBST / SB and 100 μL / well was added to both sample and standard wells, and incubated at room temperature in the dark for 45 minutes. The plates were washed five times with PBST and once with deionized water, and then developed for 15–20 minutes by adding 100 μL / well of TMB substrate (Invitrogen, Thermo Fisher Scientific, Waltham MA) at room temperature in the dark. Development was then stopped by adding 100 μL / well of ELISA stop solution (Boston Bioproducts), and absorbance was read at 450 nm using a SpectraMax plate reader within 30 minutes. Anti-antibody concentrations were determined by inserting OD values ​​into a 4-PL pseudo-standard curve using SoftMax Pro software (Molecular Devices, San Jose CA).

[0351] To demonstrate the specificity of the detected ADA, an "inhibition" assay was performed. In the drug-induced ADA inhibition assay, serum samples were diluted 1:100 in PBST / SB / 20% HS buffer and mixed with an equal volume of 300 μg / mL of the relevant therapeutic compound (the final sample was diluted 1:200 and the final inhibitory compound was 150 μg / mL) and incubated at room temperature for 30 to 40 minutes to allow the anti-drug antibody to bind to the free inhibitor (i.e., the therapeutic compound). After pre-incubation, samples were added in duplicate at 100 μL / well to strips coated with insulin-Fc fusion protein (or RHI). Samples diluted 1:200 in PBST / SB / 20% HS buffer without the inhibitory compound were also tested in sample plates and standard duplicate strips coated with canine IgG. The remaining steps of the assay procedure were performed as described above. The ADA measured in the drug-inhibited wells was matched with the concentration of uninhibited ADA to determine the specificity of ADA. If a significant inhibition of the ADA signal was observed in the drug-inhibited wells, this indicated that ADA was specific to the therapeutic compound.

[0352] Example 25: General procedures for determining in vivo drug efficacy (PD) after a single dose of human insulin Fc fusion protein in mice. Procedure.

[0353] The effect of insulin-Fc fusion protein on fasting blood glucose levels was determined as follows. Data were collected from N=3 blb / c mice or diabetic wt NOD mice (Jackson Laboratories) per group. Animals were fasted for one hour prior to the experiment, and then at time = 0 hours, mice received a single subcutaneous administration of a drug composition containing insulin-Fc fusion protein homodimer at a concentration of 300 μg / kg insulin-Fc fusion protein in a solution of 10-50 mM sodium hydrogen phosphate, 50-150 mM sodium chloride, and 0.005-0.05% v / v Tween-80; and an optional antibacterial agent (e.g., phenol, m-cresol, or methylparaben) at a concentration of 0.02 to 1.00 mg / mL, with the final solution pH adjusted to between 7.0 and 8.0 using sodium hydroxide and / or hydrochloric acid.

[0354] At each time point, blood samples were collected and immediately used with a blood glucose meter. 2. A pet blood glucose meter is used to determine blood glucose level readings, which requires approximately one drop of blood. The mean fasting blood glucose level % (FBGL%) from 0 to 9 hours is plotted to evaluate the biological activity of a given insulin-Fc fusion protein.

[0355] Example 26: General procedures for determining in vivo drug efficacy (PD) after a single dose of human insulin Fc fusion protein in mice. Procedure.

[0356] The effect of insulin-Fc fusion protein on fasting blood glucose levels was determined as follows. Data were collected from N=3 blb / c mice or diabetic wt NOD mice (Jackson Laboratories) per group. Animals were fasted for one hour prior to the experiment, and then at time = 0 hours, mice received a single subcutaneous administration of a drug composition containing insulin-Fc fusion protein homodimer at a concentration of 300 μg / kg insulin-Fc fusion protein in a solution of 10-50 mM sodium hydrogen phosphate, 50-150 mM sodium chloride, and 0.005-0.05% v / v Tween-80; and an optional antibacterial agent (e.g., phenol, m-cresol, or methylparaben) at a concentration of 0.02 to 1.00 mg / mL, with the final solution pH adjusted to between 7.0 and 8.0 using sodium hydroxide and / or hydrochloric acid.

[0357] At each time point, blood samples were collected and immediately used with a blood glucose meter. 2. A pet blood glucose meter is used to determine blood glucose level readings, which requires approximately one drop of blood. The mean fasting blood glucose level % (FBGL%) from 0 to 9 hours is plotted to evaluate the biological activity of a given insulin-Fc fusion protein.

[0358] Example 27: Assay Procedure for Identification of Immunoeptopes

[0359] A library of insulin-Fc fusion protein homodimer compounds with known amino acid sequences was coated onto Maxisorp ELISA microplates (Nunc), and the coated plates were blocked in a manner similar to that described in the antidrug antibody ELISA assay of Example 25, except that each compound in the library was coated onto an individual strip in a separate ELISA microplate well. The compounds in the library comprised a series of insulin-Fc fusion proteins with different insulin polypeptide amino acid compositions, including various B, C, and A chain amino acid mutations, different linker compositions, and different Fc fragment compositions, including some human-derived compounds. Individually, some strip wells were coated directly with serially diluted 1:2 canine IgG (Jackson Immunoresearch Laboratories, West Grove PA) to calculate antidrug antibody (ADA) units in canine IgG units, as described in Example 24.

[0360] First, serum samples obtained from individual dogs receiving repeated administrations of the insulin-Fc fusion protein were screened for the anti-drug antibody ELISA assay in Example 24. Serum samples showing intermediate or high positivity (e.g., intermediate or high titers of antibodies) in the assay of Example 24 were serially diluted (1:200 to 1:8000) in PBST / SB / 20% HS sample dilution buffer (PBS + 0.1% Tween 20 + 10% SuperBlock + 20% horse serum) and added to plates coated with a library of the insulin-Fc fusion protein compound at room temperature and incubated for 1 hour. After incubation, the plates were washed 5 times with PBST. To detect canine antibodies capable of cross-reacting with the coated compound library, HRP-conjugated goat anti-cat IgG F(ab')2 (Jackson Immunoresearch Laboratories, West Grove PA) that cross-reacts with canine IgG was diluted to 1:10000 in PBST / SB and added at 100 μL / well to both sample and standard wells and incubated at room temperature in the dark for 45 minutes. The plate was washed five times with PBST, then once with deionized water. Color development was then performed by adding 100 μL / well of TMB substrate (Invitrogen, ThermoFisher Scientific, Waltham MA) at room temperature in the dark for 15–20 minutes. Color development was then stopped by adding 100 μL / well of ELISA stop solution (Boston Bioproducts, Ashland MA), and absorbance was read at 450 nm using a SpectraMax plate reader within 30 minutes. The concentration of cross-reactive antibodies against the anti-compound in the serum samples was determined by interpolating the OD values ​​from the 4-PL pseudo-standard curve relative to a directly coated canine IgG antibody control using SoftMax Pro software (MolecularDevices, San Jose CA).

[0361] By correlating the resulting antibody concentration from the assay with the known amino acid composition of the coated insulin-Fc fusion protein library, it is possible to determine in the assay whether a specific amino acid mutation or epitope contributes to, or does not contribute to, the total antibody signal, indicating whether it binds poorly, or strongly to various insulin-Fc fusion protein homodimers. In this paper, mutations or epitopes responsible for moderate or strong binding are referred to as immunogenic “hot spots.”

[0362] Example 28: For obtaining acute and repeated-dose bioactive compounds with high homodimer titer and acceptable levels. The design process of a sex-linked canine insulin-Fc fusion protein.

[0363] The process for achieving the design objectives described in the detailed description of the invention includes the following steps. First, the insulin polypeptide of SEQ ID NO:4 or SEQ ID NO:5 is combined with a species-specific Fc fragment and linker of a specific IgG isotype such that the resulting insulin-Fc fusion protein is most likely to produce a long-acting, minimally immunogenic bioactive product (e.g., selecting a species-specific IgG isotype with minimal Fc(γ) receptor I binding). The DNA sequence encoding the desired fusion protein is prepared, cloned into a vector (LakePharma, San Carlos, CA), and then transiently transfected into HEK cells using the vector according to the procedure described in Example 1. The insulin-Fc fusion protein is then purified according to Example 4, and the total protein yield and homodimer percentage are measured according to Example 10. Only candidates with a homodimer titer greater than 50 mg / L are considered acceptable, as titers below this level are unlikely to produce commercially viable titers that meet the stringent low-cost requirements for veterinary products. The selected insulin-Fc fusion protein is then screened for bioactivity indicators by in vitro insulin receptor binding studies as described in Example 12. Based on experience, only compounds exhibiting an IC50 value of less than 5000 nM for IR activity are considered likely to exhibit biological activity in the target species. While the in vitro IR IC50 value is a useful qualitative screening tool, it utilizes human IM-9 cells expressing the human insulin receptor, and therefore may not capture some of the subtle differences in affinity between canine and human IR. Furthermore, factors other than insulin receptor binding can affect the in vivo biological activity of compounds (e.g., affinity for canine FcRn to prolong the elimination half-life in in vivo pharmacokinetics). Therefore, selected insulin-Fc fusion proteins acceptable from the perspective of preparation and IR activity IC50 value are further screened for biological activity in dogs to screen for any substances with lower than the desired potency and / or duration of biological activity (e.g., NAOC less than 150 FBGL%·day·kg / mg). Again, based on experience, when the NAOC value is greater than 150 FBGL%·day·kg / mg, the dose requirement for the target species will be low enough to achieve an acceptable therapeutic cost. Finally, an additional evaluation criterion is added, which is rarely mentioned in the art. As discussed in more detail in the following examples, many insulin-Fc fusion protein implementations exhibit acceptable NAOC levels in the target species after the first dose, but unexpectedly fail to maintain this level of bioactivity after repeated doses. Furthermore, in most cases, the decrease in bioactivity after repeated doses in the target species is associated with the production of neutralizing antidrug antibodies. This tendency to produce antidrug antibodies and the inability to maintain activity precludes the use of this insulin-Fc fusion protein for the treatment of chronic diseases such as canine diabetes.Therefore, it is considered that only insulin-Fc fusion proteins exhibiting acceptable levels of repeated-dose biological activity (e.g., a NAOCR value greater than 0.5 for the third dose relative to the first dose) and the lowest levels of antidrug antibodies can be accepted for use in this invention.

[0364] Example 29: For obtaining acute and repeated-dose bioactive compounds with high homodimer titer and acceptable levels. The design process of the human insulin-Fc fusion protein.

[0365] The process for achieving the design objectives described in the detailed description of the invention includes the following steps. First, the insulin polypeptide of SEQ ID NO:7 or SEQ ID NO:10 is combined with a human Fc fragment of a specific IgG isotype (IgG2 or IgG1) and a linker such that the resulting insulin-Fc fusion protein is most likely to produce a long-acting bioactive product. The DNA sequence encoding the desired fusion protein is prepared, cloned into a vector (LakePharma, San Carlos, CA), and then transiently transfected into HEK cells using the vector according to the procedure described in Example 1. The insulin-Fc fusion protein is then purified according to Example 4, and the total protein yield and homodimer percentage are measured according to Example 10, and the homodimer potency is calculated. Only candidates with a homodimer potency greater than 150 mg / L are considered acceptable, because homodimer potency below this level cannot be converted into a high homodimer potency CHO stably transfected cell line, and therefore cannot produce a commercially viable potency that meets the low production cost requirements of the relatively commercially available human insulin market. Then, as described in Example 12, selected insulin-Fc fusion protein conformations were screened for bioactivity indicators using in vitro IR binding studies. Empirically, it is believed that only compounds exhibiting IR activity with IC50 values ​​less than 2400 nM, more preferably less than 2000 nM, are likely to exhibit bioactivity in vivo. The in vitro IR IC50 value is a useful qualitative screening tool because this assay utilizes human IM-9 cells expressing human IR, and it is expected that higher binding (lower IR IC50 value) provides greater in vivo potency for a given dose compared to compounds with reduced binding (higher IRIc50 value). Furthermore, factors other than IR binding can affect the in vivo bioactivity of compounds. For example, the affinity of Fc fusion proteins for human FcRn receptors is related to the elimination half-life of compounds in vivo pharmacokinetics. As measured by an in vitro human FcRn binding assay, prolonged in vivo half-lives of insulin-Fc fusion proteins, measured over several days, are closely associated with EC50 values ​​less than 1500 ng / mL, more preferably less than 1000 ng / mL (Example 19).

[0366] By testing the binding affinity to the human Fc(γ)RI receptor and human C1q, selected insulin-Fc fusion protein conformations acceptable in terms of preparation, IC50 value of IR activity, and EC50 value of human FcRn activity were further screened to target their immunogenic potential in vivo. Molecules that bind more firmly to these immune system components are more likely to undergo increased antigen-cell presentation (APC) and exhibit stronger immunogenic characteristics as measured by anti-drug antibodies. Anti-drug antibodies are undesirable as they may impair the molecule's in vivo pharmacokinetic half-life, neutralize drug activity, or may perform both functions simultaneously. The potential impact of anti-drug antibodies on the in vivo performance of insulin-Fc can be quite problematic. Therefore, work was carried out to screen candidate insulin-Fc fusion proteins for Fc(γ)RI (Example 14) and C1q (Example 16) to mitigate the risks of unwanted antigen presentation and potential anti-drug antibody-induced immunogenicity (Guilliams, Martin & Bruhns, Pierre & Saeys, Yvan & Hammad, Hamida & Lambrecht, Bart. (2014). The function of Fc gammareceptors in dendritic cells and macrophages. Nature reviews. Immunology. 14.10.1038 / nri3582). The design target for human Fc(γ)RI binding (where the biotinylated-Fc(γ)RI concentration of the tested insulin-Fc fusion protein was 3000 ng / mL) was OD450 ratio <0.50 (relative to the insulin-Fc fusion protein conformation of SEQ ID NO:76), and the design target for C1q binding (where the biotinylated C1q concentration of the tested insulin-Fc fusion protein was 1000 ng / mL) was OD450 ratio <0.35 (relative to the insulin-Fc fusion protein conformation of SEQ ID NO:76).

[0367] As described elsewhere, it was unexpected that the insulin-Fc fusion protein conformation containing human IgG1 Fc yielded significantly higher yields than the conformation containing human IgG2 Fc, and was therefore preferred in terms of prepareability. However, the insulin-Fc fusion protein conformation containing human IgG1 Fc, when containing native N-linked glycosylation at the cNg site, typically exhibits very high binding to both Fc(γ)RI and C1q, and strong binding to one or both of these moieties is associated with a high potential for enhanced antigen presentation and immunogenicity in vivo (Kouser, L., Madhukaran, SP, Shastri, A., Saraon, A., Ferluga, J., Al-Mozaini, M., & Kishore, U. (2015). Emerging and Novel Functions of Complement Protein C1q. Frontiers in Immunology, 6, 317. doi:10.3389 / fimmu.2015.00317). Therefore, by first screening compounds with mutations at the native glycosylation site (cNg) to prevent native glycosylation during biosynthesis, several variants of the hIgG1 isotype, Fc-containing insulin-Fc fusion protein conformation were tested. These molecules were then further mutated to improve their properties by manipulating the insulin polypeptide composition. Finally, several adapter variants with different compositions and lengths were investigated to identify whether various designed properties could be further optimized. After multiple rounds of optimization and screening, only insulin-Fc fusion protein conformations exhibiting acceptable levels of homodimeric titer, IR binding, FcRn binding, Fc(γ)RI binding, and Clq binding were deemed acceptable for use in this invention.

[0368] Results - Insulin-FC fusion protein containing canine Fc fragment

[0369] Example 30: Canine insulin-Fc fusion protein containing canine Fc IgGA isotype.

[0370] An attempt was made to use the peptide linker of SEQ ID NO:11 to generate an insulin-Fc fusion protein containing the insulin polypeptide sequence of SEQ ID NO:4 and the Fc fragment of the canine IgGA isotype (SEQ ID NO:14).

[0371] The complete amino acid sequence of the obtained insulin-Fc fusion protein is as follows:

[0372] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGRCTDTPPCPVPEPLGGPSVLIFPPKPKDILRITRTPEVTCVVLDLGREDPEVQISWFVDGKEVHTAKTQSREQQFNGTYRVVS VLPIEHQDWLTGKEFKCRVNHIDLPSPIERTISKARGRAHKPSVYVLPPSPKELSSSDTVSITCLIKDFYPPDIDVEWQSNGQQEPERKHRMTPPQLDEDGSYFLYSKLSVDKSRWQQGDPFTCAVMHETLQNHYTDLSLSHSPG(SEQ ID NO:31).

[0373] The insulin-Fc fusion protein of SEQ ID NO: 31 was synthesized in HEK cells according to Example 1 and purified according to Example 4. The protein yield after the protein A purification step was 22 mg / L. The structure of the insulin-Fc fusion protein was confirmed by non-reducing and reducing CE-SDS according to Example 6, and the sequence was further identified by LC-MS and glycan removal according to Example 8. The homodimer percentage was measured by size exclusion chromatography according to Example 10 and determined to be 24%, indicating a high degree of homodimer aggregates. Therefore, the homodimer potency obtained was only 5 mg / L. In summary, the preparation of the insulin-Fc fusion protein of SEQ ID NO: 31 in HEK cells yielded high levels of aggregates and a low homodimer potency (5 mg / L), which does not meet the design target of a homodimer potency greater than 50 mg / L.

[0374] However, the bioactivity of the insulin-Fc fusion protein of SEQ ID NO: 31 was evaluated. First, the insulin receptor binding of the insulin-Fc fusion protein of SEQ ID NO: 31 was measured according to Example 12, yielding an IC50 value of 2,733 nM, indicating that the compound may have bioactivity in vivo (i.e., IC50 less than 5000 nM).

[0375] Next, according to Example 21, the in vivo efficacy (PD) of the insulin-Fc fusion protein of SEQ ID NO: 31 was measured after a single intravenous administration of the compound to N=3 dogs. Figure 2The fasting blood glucose level percentage of SEQ NO: 31 is shown as a function of time. The NAOC of SEQ ID NO: 31 was calculated as 105 FBGL%·day·kg / mg according to the procedure of Example 22. Using the method of Example 23, the in vivo half-life of SEQ ID NO: 31 was calculated to be less than 1 day. The relatively low NAOC may be a result of the large amount of aggregates in the sample (i.e., low homodimer %), but the pharmacokinetic elimination half-life of the soluble homodimers remaining in circulation is still less than one day, which is considered unlikely to support weekly dosing.

[0376] Example 31: Mutation in the Fc fragment region of an insulin-Fc fusion protein containing a canine IgGA isotype.

[0377] To increase the homodimer content (%), enhance bioactivity, and prolong the half-life of the insulin-Fc fusion protein of SEQ ID NO:31, a mutation was inserted into the CH3 region of the Fc fragment to attempt to prevent intermolecular association (e.g., intermolecular Fc fragment-Fc fragment interactions) and promote stronger binding to the FcRn receptor (e.g., higher affinity for FcRn) to increase recirculation and systemic circulation time. The insulin-Fc fusion protein was synthesized in HEK cells according to Example 1, purified according to Example 4, and tested according to Examples 6, 8, and 10. The results are shown in Table 2 below. Sequence alignment and amino acid sequence differences between SEQ ID NO:32, SEQ ID NO:33, SEQ ID NO:34, and SEQ ID NO:35 and SEQ ID NO:31 are shown in Table 2. Figure 3 middle.

[0378] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGRCTDTPPCPVPEPLGGPSVLIFPPKPKDILRITRTPEVTCVVLDLGREDPEVQISWFVDGKEVHTAKTQSREQQFNGTYRVVS VLPIEHQDWLTGKEFKCRVNHIDLPSPIERTISKARGRAHKPSVYVLPPSPKELSSSDTVSITCLIKDFYPPDIDVEWQSNGQQEPERKHRMTPPQLDEDGSYFLYSKLSVDKSRWQQGDPFTCAVLHEALHSHYTQKSLSLSPG(SEQ ID NO:32)

[0379] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGRCTDTPPCPVPEPLGGPSVLIFPPKPKDILRITRTPEVTCVVLDLGREDPEVQISWFVDGKEVHTAKTQSREQQFNGTYRVVSVLPIEHQDWLTGKEFKCRVNHIDLPSPIERTISKARGRAHKPSVYVLPPSPKELSSSDTVSITCLIKDFYPPDIDVEWQSNGQQEPERKHRMTPPQLDEDGSYFLYSKLSVDKSRWQQGDPFTCAVLHETLQSHYTDLSLSHSPG(SEQ ID NO:33)

[0380] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGRCTDTPPCPVPEPLGGPSVLIFPPKPKDILRITRTPEVTCVVLDLGREDPEVQISWFVDGKEVHTAKTQSREQQFNGTYRVVSVLPIEHQDWLTGKEFKCRVNHIDLPSPIERTISKARGRAHKPSVYVLPPSPKELSSSDTVSITCLIKDFYPPDIDVEWQSNGQQEPERKHRMTPPQLDEDGSYFLYSKLSVDKSRWQQGDPFTCAVMHETLQSHYTDLSLSHSPG(SEQ ID NO:34)

[0381] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGRCTDTPPCPVPEPLGGPSVLIFPPKPKDILRITRTPEVTCVVLDLGREDPEVQISWFVDGKEVHTAKTQSREQQFNGTYRVVSVLPIEHQDWLTGKEFKCRVNHIDLPSPIERTISKARGRAHKPSVYVLPPSPKELSSSDTVSITCLIKDFYPPDIDVEWQSNGQQEPERKHRMTPPQLDEDGSYFLYSKLSVDKSRWQQGDPFTCAVLHETLQNHYTDLSLSHSPG(SEQ ID NO:35)

[0382] Table 2 lists the insulin-Fc fusion proteins based on canine IgGA variants, along with their corresponding protein yields, homodimer percentages, and homodimer titers. The results indicate that various mutations in the IgGA Fc fragment did not increase homodimer percentages or titers; instead, they produced highly aggregated proteins with extremely low homodimer titers, below 5 mg / L. Therefore, the in vivo biological activity and pharmacokinetics of the compounds could not be assessed.

[0383]

[0384] Example 32: Canine insulin-Fc fusion protein using other canine Fc fragment isotypes.

[0385] As mentioned above, canine IgGA is considered the preferred isotype of the Fc fragment for producing non-immunogenic insulin-Fc fusion proteins in dogs because it lacks Fc(γ)I effector function in dogs (similar to the human IgG2 isotype in humans). However, insulin-Fc fusion proteins prepared from canine IgGA Fc fragments exhibit high aggregation, unacceptably low homodimer titers, and unacceptably low levels of biological activity and duration of action. Therefore, Fc fragments from other canine IgG isotypes (canine IgGB of SEQ ID NO:15, canine IgGC of SEQ ID NO:16, and canine IgGD of SEQ ID NO:17) have been evaluated as alternatives to the canine IgGA Fc fragment of the insulin-Fc fusion protein of SEQ ID NO:31. Three insulin-Fc fusion proteins comprising Fc fragments based on canine IgGB, IgGC, and IgGD isotypes were synthesized using the same insulin peptide of SEQ ID NO:4 and the peptide linker of SEQ ID NO:11 used in the preparation of the insulin-Fc fusion protein of SEQ ID NO:31. The protein was prepared in HEK293 cells according to Example 1. The insulin-Fc fusion protein was then purified using a protein A column according to Example 4. The structure of the insulin-Fc fusion protein was confirmed by non-reducing and reducing CE-SDS according to Example 6, and the sequence was further identified by LC-MS and glycan removal according to Example 8. The homodimer percentage was measured by size exclusion chromatography according to Example 10. Their sequences are shown below, and their sequence alignment with SEQ ID NO:31 is shown in [reference needed]. Figure 4 middle:

[0386] FVNQHLCGSDLVEALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:36)

[0387] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGCNNCPCPGCGLLGGPSVFIFPPKPKDILVTARTPTVTCVVVDLDPENPEVQISWFVDSKQVQTANTQPREEQSNGTYRVVSV LPIGHQDWLSGKQFKCKVNNKALPSPIEEIISKTPGQAHQPNVYVLPPSRDEMSKNTVTLTCLVKDFFPPEIDVEWQSNGQQEPESKYRMTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQISLSHSPG(SEQID NO:37)

[0388] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGCISPPVPESLGGPSVFIFPKPKDILRITRTPEITVCVLDLGREDPEVQISWFVDGKEVHTAKTQPREQQFNSTYRVVSVL PIEHQDWLTGKEFKCRVNHIGLPSPIERTISKARGQAHQPSVYVLPPSPKELSSSDTVTLTCLIKDFFPPEIDVEWQSNGQPEPESKYHTTAPQLDEDGSYFLYSKLSVDKSRWQQGDTFTCAVMHEALQNHYTDLSLSHSPG(SEQ ID NO:38)

[0389] The resulting protein yields, homodimer percentages, and homodimer titers are given in Table 3. Surprisingly, only the insulin-Fc fusion protein of SEQ ID NO:36, containing an Fc fragment based on a canine IgGB isotype, showed a homodimer titer greater than the design criterion of 50 mg / L. The insulin-Fc fusion protein of SEQ ID NO:37, containing an Fc fragment based on a canine IgGC isotype, produced no compound at all, and the insulin-Fc fusion protein of SEQ ID NO:38, containing an Fc fragment based on a canine IgGD isotype, showed significant protein yield but exhibited high aggregation, resulting in an unacceptably low homodimer titer.

[0390] The in vitro insulin receptor binding of the insulin-Fc fusion proteins of SEQ ID NO:36 and SEQ ID NO:38 was tested according to the procedure of Example 12. The insulin-Fc fusion protein of SEQ ID NO:38 showed an IC50 greater than 5000 nM, indicating that this compound is highly unlikely to exhibit biological activity in vivo. However, the insulin-Fc fusion protein of SEQ ID NO:36 showed an IC50 of 28 nM, indicating that this sequence may have biological activity in vivo.

[0391]

[0392] *DNM = Not measured

[0393] Example 33: Insulin-Fc Fusion Containing Insulin Peptide of SEQ ID NO: 4 and Canine IgGB Isotype Fc Fragment The in vivo efficacy of the protein.

[0394] Based on the promising homodimer titer and insulin receptor activity results in Example 32, the in vivo bioactivity of the insulin-Fc fusion protein of SEQ ID NO:36 was tested according to Example 21 after intravenous injection in each of N=3 healthy, antibody-naïve beagle dogs weighing approximately 10 kg. In a separate experiment, the compound was administered subcutaneously to N=3 healthy, antibody-naïve beagle dogs. Figure 5 The relationship between FBGL% and time for a single intravenous administration of the insulin-Fc fusion protein of SEQ ID NO: 36 is shown. Figure 6 The relationship between FBGL% and time for a single subcutaneous administration of the insulin-Fc fusion protein of SEQ ID NO: 36 is shown, both of which demonstrate significant biological activity of the insulin-Fc fusion protein of SEQ ID NO: 36 in dogs.

[0395] The NAOC was calculated according to the procedure in Example 22 to determine the relative bioactivity and duration of action of the insulin-Fc fusion protein. The NAOC of the intravenously administered insulin-Fc fusion protein of SEQ ID NO:36 was 399 FBGL%·day·kg / mg, which was 3.8 times that of the intravenously administered insulin-Fc fusion protein of SEQ ID NO:31, indicating a significant increase in bioactivity compared to the insulin-Fc fusion protein containing the canine IgGA Fc fragment. The NAOC of the subcutaneously administered insulin-Fc fusion protein of SEQ ID NO:36 was 366 FBGL%·day·kg / mg, indicating that the bioactivity level obtained via subcutaneous administration was similar to that obtained via intravenous administration.

[0396] Example 34: Repeated subcutaneous administration of an insulin peptide containing SEQ ID NO: 4 and a canine IgGB isotype Fc fragment In vivo immunogenicity screening following insulin-Fc fusion protein.

[0397] The bioactivity of the insulin-Fc fusion protein of SEQ ID NO:36 under repeated subcutaneous administration was tested in dogs according to the method described in Example 22. Subcutaneous administration was performed on N=3 animals on days 0, 35, and 42, and the FBGL% was measured in a 7-day window following each administration, as described in Example 22. The NAOC and NAOCR for each repeated subcutaneous injection were calculated according to the procedure of Example 22. As shown in Table 4, repeated subcutaneous administration in dogs unexpectedly revealed a significant decrease in bioactivity at the third administration, such as a significant reduction in the measured NAOCR (i.e., the NAOC of the third injection was only 0.40, or 40%, of the NAOC of the first injection).

[0398]

[0399] Without being bound by any particular interpretation, it is assumed that the significant decrease in the biological activity of the insulin-Fc fusion protein of SEQ ID NO: 36 after a third repeated subcutaneous administration in dogs is due to the neutralization of its biological activity by the production of antidrug antibodies. Antidrug antibodies may target the insulin polypeptide, linker, or Fc fragment portion of the insulin-Fc fusion protein. Immunogenic reactions manifest as interactions between antigen-presenting cells, T helper cells, B cells, and their associated cytokines, which may result in the production of endogenous antibodies against the drug (e.g., antidrug antibodies). Binding antibodies are all isotypes capable of binding the insulin-Fc fusion protein, and these can be detected in an immunoassay as described in Example 24. Neutralizing antibodies that inhibit the functional activity of the insulin-Fc fusion protein typically target epitopes essential for biological activity. To assess whether this is the case, serum collected before each dose administration and at the end of the experiments described in Examples 11 and 12 was tested to quantify the levels of antidrug antibodies, according to Example 24. Figure 7 As shown, the level of antidrug antibodies did indeed increase with repeated subcutaneous administration of the compound, suggesting that the generation of neutralizing antidrug antibodies may be the reason for the decrease in NAOCR after the third injection of the insulin Fc fusion protein of SEQ ID NO: 36.

[0400] Example 35: Non-glycosylated insulin-Fc fusion protein comprising the insulin polypeptide of SEQ ID NO: 4 and canine IgGB Homotype Fc fragments can be used to reduce the potential risk of immunogenicity.

[0401] As shown in Examples 32 and 33, the insulin-Fc fusion protein of SEQ ID NO: 36 exhibited acceptable homodimer content %, homodimer titer, and bioactivity in dogs; however, its use for chronic diseases such as diabetes was compromised by decreased bioactivity following repeated subcutaneous administration (Example 34) and the generation of antidrug antibodies (Example 34). Without being bound by any particular theory, one possible reason for the generation of antidrug antibodies and decreased bioactivity is the increased interaction between the canine IgGB Fc fragment and various receptors of the canine immune system (e.g., Fc(γ) receptors, such as Fc(γ)RI). However, the canine IgGB isotype is the only one of the four canine IgG isotypes that produces an insulin-Fc fusion protein (Example 28) that meets the design goals of prepareability and single-dose bioactivity when used for the Fc fragment. As described in the detailed description of the invention, a method for reducing Fc(γ) interaction involves mutating the cNg site of the Fc fragment to prevent glycosylation during synthesis in host cells. Therefore, a cNg site mutation was performed on the Fc fragment region of SEQ ID NO:36 to reduce the binding affinity of the Fc fragment to the Fc(γ) receptor in vivo, as measured by the binding in the in vitro human Fc(γ)RI assay described in Example 14. Verification of glycan deficiency was performed using the LC-MS method of Example 8, but the PNGase F treatment step was omitted. The cNg site in the insulin-Fc fusion protein of SEQ ID NO:36 is cNg-NB139. Mutations in SEQ ID NO:36 include SEQ ID NO:39 containing the cNg-NB139-Q mutation, SEQ ID NO:40 containing the cNg-NB139-S mutation, SEQ ID NO:41 containing the cNg-NB139-D mutation, and SEQ ID NO:42 containing the cNg-NB139-K mutation. The full amino acid sequence of the cNg-mutated insulin-Fc fusion protein (NB139 position is underlined) is listed below, and the resulting sequence alignment is shown in [reference needed]. Figure 8 Clustal Omega:

[0402] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQF QGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:39)

[0403] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGDCPKCPAPEMLGGPSVFIFPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQF S GTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:40)

[0404] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGDCPKCPAPEMLGGPSVFIFPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQF D GTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:41)

[0405] FVNQHLCGSDLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCHSICSLYQLENYCNGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQF K GTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG (SEQ ID NO: 42).

[0406] Insulin-Fc fusion protein was prepared in HEK293 cells according to Example 1 and purified using a protein A column according to Example 4. The structure of the insulin-Fc fusion protein was confirmed by non-reducing and reducing CE-SDS according to Example 6, and the sequence was further identified by LC-MS and glycan removal according to Example 8. Homodimer percentage was measured by size exclusion chromatography according to Example 10. As shown in Table 5, the homodimer titers of insulin-Fc fusion proteins of SEQ ID NO:40, SEQ ID NO:41, and SEQ ID NO:42 met the design target, while the insulin-Fc fusion protein of SEQ ID NO:39 containing the cNg-NB139-Q mutation unexpectedly did not meet the homodimer titer design target.

[0407]

[0408]

[0409] To determine which of the remaining three compounds was most likely to exhibit reduced immunogenicity, Fc(γ) receptor binding was measured according to the procedure of Example 15. Low Fc(γ) receptor binding was most likely associated with minimal immunogenicity. Table 6 compares the Fc(γ) receptor I binding of these insulin-Fc fusion proteins with that of the insulin-Fc fusion protein of SEQ ID NO:36, unexpectedly demonstrating that the insulin-Fc fusion protein containing the cNg-D mutation of SEQ ID NO:41 exhibited approximately twice the Fc(γ) receptor binding activity of the insulin-Fc fusion proteins containing the cNg-S mutation of SEQ ID NO:40 and the cNg-K mutation of SEQ ID NO:42. Therefore, it is considered that only insulin-Fc fusion proteins containing the latter two compounds containing the cNg-S and cNg-K mutations are suitable for repeated-dose bioactivity testing in dogs.

[0410]

[0411] Example 36: Evaluation of SEQ using non-glycosylated cNg-K and cNg-S canine IgGB isotype Fc fragments Pancreas ID NO: 4 In vivo bioactivity and immunogenicity of insulin peptides.

[0412] To determine whether the insulin-Fc fusion protein containing the cNg-S mutation of SEQ ID NO: 40 improved repeated-dose bioactivity in dogs, the compound was administered subcutaneously to N=1 dogs on days 0, 7, 14, and 28 according to the protocol of Example 22. When the dogs' FBGL% decreased too low, they were fed to raise their blood glucose to a safe level. The NAOC of the first injection was 191 FBGL%·day·kg / mg, indicating that the insulin-Fc fusion protein of SEQ ID NO: 40 has satisfactory bioactivity in vivo. The NAOC and NAOCR of each subsequent dose were also measured according to the general protocol of Example 22, calculated from the time of administration to the time prior to the next administration. The NAOC and NAOCR shown in Table 7 illustrate that the insulin-Fc fusion protein of SEQ ID NO: 40 exhibited a significantly reduced NAOCR at the 3rd and 4th doses of the four-dose regimen. Therefore, in dogs, the insulin-Fc fusion protein containing the cNg-S mutation of SEQ ID NO:40 could not demonstrate biological activity with repeated administration, even though its Fc(γ)RI binding was four times lower than that of the insulin-Fc fusion protein of SEQ ID NO:36.

[0413]

[0414] To determine whether the insulin-Fc fusion protein containing the cNg-K mutation of SEQ ID NO: 42 improved repeated-dose bioactivity in dogs, the compound was administered subcutaneously to N=1 dogs on days 0, 7, 14, and 28 according to the protocol of Example 22. When the dogs' FBGL% decreased too low, they were fed to raise their blood glucose to a safe level. The NAOC of the first injection was 449% FBGL·day·kg / mg, indicating that the insulin-Fc fusion protein of SEQ ID NO: 42 has satisfactory bioactivity in vivo. The pharmacokinetic properties of the compound were also measured using ELISA according to the method of Example 23, and its elimination half-life was determined to be approximately 0.9 days by fitting a two-compartment model to the data. NAOC and NAOCR were also measured for each subsequent dose according to the general protocol of Example 22, calculated from the time of administration to the time prior to the next administration. The NAOC and NAOCR shown in Table 8 indicate that the insulin-Fc fusion protein of SEQ ID NO: 42 maintained a NAOCR greater than 0.6 over four doses. Therefore, unexpectedly, the insulin-Fc fusion protein of SEQ ID NO: 42 containing the cNg-K mutation is the only non-glycosylated mutant of the insulin-Fc fusion protein of SEQ ID NO: 36, resulting in significantly improved bioactivity with repeated dosing in dogs.

[0415]

[0416]

[0417] According to Example 24, the levels of anti-drug and anti-insulin antibodies were also measured throughout the treatment course (28 days) and for an additional two weeks. Figure 9 This indicates that repeated subcutaneous administration of the insulin-Fc fusion protein of SEQ ID NO:42 to dogs still produces anti-drug antibodies, but the anti-drug antibody titers are much lower than those produced by the insulin-Fc fusion protein of SEQ ID NO:36 (Example 32).

[0418] Example 37: Screening canine serum containing anti-drug antibodies and identifying the B10D and A8H positions of the insulin peptide. Potential immunogenic epitopes at the site.

[0419] Mutating the cNg site of the canine IgGB Fc fragment to Lys (i.e., cNg-K) did indeed improve the repeated-dose bioactivity of the insulin fusion protein containing the insulin peptide of SEQ ID NO:4 and the peptide linker of SEQ ID NO:11 (Example 36), but the resulting insulin-Fc fusion protein of SEQ ID NO:42 still produced anti-drug antibodies (Example 36). Therefore, it is speculated that the insulin peptide of SEQ ID NO:4 may unexpectedly contain a specific epitope (i.e., an immunogenic "hotspot") targeted by the canine immune system. Therefore, the binding specificity of the antibodies present in the serum samples described in Example 24 was evaluated according to the general procedure of Example 27. Analysis of antibody-containing serum samples from repeatedly administered insulin-Fc fusion protein of SEQ ID NO:36 (Example 32) against a coated insulin-Fc fusion protein library revealed two unexpected major "hot spots" within the insulin polypeptide sequence of SEQ ID NO:5: an aspartic acid mutation at position 10 from the N-terminus of the B chain (i.e., B10), and, individually, a histidine mutation at position 8 from the N-terminus of the A chain (i.e., A8). The results indicate that insulin-Fc fusion proteins containing an amino acid composition of insulin polypeptides with these two specific amino acid mutations may be immunogenic in dogs, and therefore may generate anti-drug antibodies that neutralize biological activity after repeated injections. Therefore, insulin polypeptides that do not contain B10 aspartic acid and A8 histidine are determined to be preferred for insulin-Fc fusion proteins requiring long-term, repeated administration in dogs (e.g., for the treatment of canine diabetes).

[0420] Example 38: An insulin-Fc fusion protein comprising the insulin polypeptide and non-glycosylated form of SEQ ID NO:4 The canine IgGB isotype Fc fragment, in which the B10D and A8H mutations in the insulin peptide were restored to the natural composition to reduce potential There is an immunogenicity risk.

[0421] To assess whether replacing the "hotspot" mutation would improve the immunogenicity and repeated-dose bioactivity of the insulin-Fc fusion protein comprising the insulin peptide of SEQ ID NO:4 and the canine IgGB isotype fragment, an exemplary insulin-Fc fusion protein (SEQ ID NO:43) was synthesized in which the B10 and A8 amino acids of the insulin peptide were restored to their native histidine and threonine compositions as given below (SEQ ID NO:63), wherein the non-native amino acids are underlined.

[0422] FVNQHLCGSHLVEAL A LVCGERGFFYT DP T GGGPRR GIVEQCCTSICSLYQLENYCN (SEQ ID NO: 63).

[0423] Furthermore, considering the additional potential benefits of the non-glycosylated cNg mutant, the insulin-Fc fusion protein of SEQ ID NO: 43 contains the cNg-Q mutation. The full amino acid sequence of the insulin-Fc fusion protein of SEQ ID NO: 43 is shown below:

[0424] FVNQHLCGSHLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCTSICSLYQLENYCNGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFQGTYRVVSV LPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:43).

[0425] The insulin-Fc fusion protein of SEQ ID NO:43 was prepared in HEK293 cells according to Example 1 and purified using a protein A column according to Example 4. The yield of the obtained protein was only 21 mg / L. The structure was confirmed by non-reducing and reducing CE-SDS according to Example 6, and the sequence was further identified by LC-MS and glycan removal according to Example 8. The homodimer percentage measured by size exclusion chromatography according to Example 10 was 98.0%, indicating that the protein was relatively free of aggregates.

[0426] Despite the relatively low homodimer titer of 21 mg / L, the in vivo bioactivity and immunogenicity of the insulin-Fc fusion protein of SEQ ID NO:43 were evaluated in dogs according to the procedures of Examples 22, 23 and 24, respectively. Figure 10 This indicates that restoring the B10D and A8H mutations to their native amino acids (i.e., B10H and A8T) in the insulin-Fc fusion protein of SEQ ID NO:43 does indeed significantly reduce the immunogenicity of the parent compound (SEQ ID NO:36).

[0427] However, as Figure 11 As shown, the insulin-Fc fusion protein containing natural B10 and A8 amino acids in SEQ ID NO: 43 has no biological activity (i.e., NAOC is essentially zero).

[0428] Example 39: An attempt to incorporate additional B-chain and A-chain mutations into the insulin peptide of SEQ ID NO:63 to improve... Biological activity of the insulin-Fc fusion protein containing the canine IgGB Fc fragment.

[0429] The fact that the insulin-Fc fusion protein of SEQ ID NO:43 does not produce antidrug antibodies (Example 38) compared to the insulin-Fc fusion protein of SEQ ID NO:36 (Example 33) provides strong evidence for the theory that the B10D and A8H mutations in the insulin polypeptide of SEQ ID NO:4 may be the immunogenic epitopes responsible for the production of antidrug antibodies. However, the insulin-Fc fusion protein of SEQ ID NO:43 lacks in vivo potency compared to SEQ ID NO:36, suggesting that these two amino acid mutations are also the reason for achieving acceptable levels of biological activity. As measured by insulin receptor binding assay according to the method of Example 12, the lack of in vivo potency of the insulin-Fc fusion protein of SEQ ID NO:43 is associated with its high IC50 (shown in Table 9 below). Therefore, further efforts are needed to improve the bioactivity of the insulin-Fc fusion protein (i.e., to reduce the IC50 value of insulin receptor binding assay to less than 5000 nM, or more preferably less than 4000 nM, or even more preferably less than 3000 nM), while maintaining low immunogenicity by retaining native B10 and A8 amino acids in the insulin peptide.

[0430] As is well known, portions of the B and A chains of insulin are essential for strong binding to the receptor (Hubbard SR, "Structural biology: Insulin meets its receptor", Nature. 2013; 493(7431):171-172). Therefore, portions of the B or A chain are modified while retaining B10 and A8 identical to those in native insulin, while the C chain and peptide linker remain unchanged. Some of these insulin-Fc fusion proteins were prepared in HEK293 cells according to Example 1 and purified using a protein A column according to Example 4. Their structures were confirmed by non-reducing and reducing CE-SDS according to Example 6, and their sequences were further identified by LC-MS and glycan removal according to Example 8. Their homodimer content (%) was measured by size exclusion chromatography according to Example 10, and their insulin receptor binding affinity was measured according to Example 12. Their sequences are shown below, and the resulting sequence alignment with SEQ ID NO:43 is shown in [reference needed]. Figure 12 (Clustal Omega)

[0431] FVNQHLCGSHLVQALYLVCGERGFFYTDPTGGGPRRGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:44)

[0432] FVNQHLCGSELVEALALVCGERGFFYTDPTGGGPRRGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:45)

[0433] FVNQHLCGSHLVEAALALVCGEAGFFYTDPTGGGPRRGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:46)

[0434] FVNQHLCGSHLVEALALVCGERGFYYTDPTTGGGPRRGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSV LPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:47)

[0435] FVNQHLCGSHLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSV LPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:48)

[0436]

[0437]

[0438] Compared to SEQ ID NO:43, the proposed mutations improved IR binding (i.e., reduced IC50 values) in only three cases (SEQ ID NO:44, SEQ ID NO:45, and SEQ ID NO:47). However, none of the mutations resulted in the compound meeting the preparation design target of a homodimer potency greater than 50 mg / L, and in some cases, the mutations led to a significant reduction in prepareability (e.g., homodimer potency less than 20 mg / L).

[0439] Example 40: An attempt to incorporate a C-chain mutation into the insulin peptide of SEQ ID NO:63 to enhance the inclusion of canine IgGB Biological activity of insulin-Fc fusion protein related to Fc fragment.

[0440] The results obtained in Example 39 indicate that all attempts to mutate the A and B chains of the insulin polypeptide of SEQ ID NO:63 resulted in unacceptably low HEK homodimer titers of the associated insulin-Fc fusion (i.e., homodimer titers less than or equal to 25 mg / L). Therefore, further experiments are required. In this example, the C chain composition of the insulin polypeptide of SEQ ID NO:63 was mutated by making it longer or by increasing its flexibility. Natural insulin (e.g., human insulin) has been shown to undergo significant conformational changes involving the folding and movement of the B and A chains due to their binding to the insulin receptor (e.g., as described in Menting, et al., Nature, 2013; 493(7431):pp241–245). Unlike the insulin polypeptide of the present invention, natural insulin is free to undergo such conformational changes at the insulin receptor because it is a naturally occurring double-chain polypeptide linked only by two disulfide bonds without the C chain restricting the mobility of the A and B chains. Unbound by any particular theory, it is assumed that the C-chain of the insulin polypeptide contained in SEQ ID NO:63 is too inflexible (e.g., lacking an amino acid composition and sequence that allows for easy movement between the B and A chains) and / or too short (e.g., insufficient amino acids between the C-terminus of the B chain and the N-terminus of the A chain) to prevent the insulin polypeptide from undergoing the necessary changes in molecular shape required for strong binding to the insulin receptor. Therefore, several insulin-Fc fusion proteins were synthesized based on the insulin-Fc fusion protein of SEQ ID NO:43, exhibiting the changes in the C-chain of the insulin polypeptide as shown below, with the resulting sequence alignment to SEQ ID NO:43 shown in [image / description]. Figure 13 (ClustalOmega)

[0441] FVNQHLCGSHLVQALYLVCGERGFFYTDPTQRGGGGGQRGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSV LPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQID NO:49)

[0442] FVNQHLCGSHLVVEALALVCGERGFFYTDPTGGGGGGSGGGGGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVV SVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPDIVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:50)

[0443] FVNQHLCGSHLVVEALALVCGERGFYTDPGGGGGGGGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPKPKDTLLIARTPEVTCVVVDLDPEPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSVLPIG HQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPDIVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQIDNO:51)

[0444] FVNQHLCGSHLVEALALVCGERGFFYTPGGGGGGGGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:52)

[0445]

[0446]

[0447] Insulin-Fc fusion proteins were prepared in HEK293 cells according to Example 1 and purified using a protein A column according to Example 4. Their structure was confirmed by non-reducing and reducing CE-SDS according to Example 6, and the sequence was further identified by LC-MS and glycan removal according to Example 8. Their homodimer content (%) was measured by size exclusion chromatography according to Example 10, and their insulin receptor binding affinity was measured according to Example 12. The results are shown in Table 10. In only one case, the insulin-Fc fusion protein containing the longest C chain (GGGGGGSGGGG–SEQ ID NO:71) (SEQ ID NO:50) showed a significantly increased insulin receptor binding affinity (IC50 < 3000 nM) compared to the insulin-Fc fusion protein of SEQ ID NO:43. However, none of these C-chain mutated insulin-Fc fusion proteins exhibited the homodimer potency greater than 50 mg / L, the target for the preparation design. In fact, in one case (SEQ ID NO:49), the C-chain mutation unexpectedly resulted in a significantly reduced homodimer potency.

[0448] Example 41: An attempt to incorporate peptide linker mutations into an insulin peptide containing SEQ ID NO:63 and canine IgGB Fc Fragments of insulin-Fc fusion protein are used to enhance biological activity.

[0449] Unbound by any particular theory, another possible reason for the poor insulin receptor binding of the insulin-Fc fusion protein of SEQ ID NO:43 involves steric hindrance between the insulin peptide and the insulin receptor, caused by the close proximity of the much larger Fc fragment molecule to the insulin peptide via a peptide linker. It is thought that shorter peptide linkers or more tightly folded peptide linkers might exacerbate this problem, while longer peptide linkers or self-folding peptide linkers (e.g., linkers with higher molecular stiffness) could alleviate it by creating more space between the insulin peptide and the Fc fragment. The increased space between the insulin peptide and the Fc fragment would also increase the distance between the insulin receptor and the Fc fragment, thereby reducing interference during insulin receptor binding. It is hypothesized that the peptide linker of SEQ ID NO:11 (i.e., GGGGAGGGG) used to construct the insulin-Fc fusion protein of SEQ ID NO:43 might be too short and / or too flexible because the amino acids constituting the linker do not contain side chains (i.e., it contains only glycine and alanine amino acids). Therefore, to test this hypothesis, two additional insulin-Fc fusion protein variants of the insulin-Fc fusion protein of SEQ ID NO: 43 were synthesized. The insulin-Fc fusion protein of SEQ ID NO: 48 contains the same peptide linker as the insulin-Fc fusion protein used to construct SEQ ID NO: 43, but has an insulin polypeptide in which asparagine is absent at position 21 (i.e., A21) starting from the N-terminus of the A chain. This specific mutation was incorporated to observe whether the connection between the A chain and the peptide linker affects protein yield and / or the bioactivity of the molecule. Another insulin-Fc fusion protein of SEQ ID NO: 53 contains this des-A21N A chain mutation and a peptide linker that is more than twice the length of the insulin-Fc fusion protein used to construct SEQ ID NO: 43. In this longer peptide linker, alanine is undesirable and is replaced by glutamine, which contains a polar amide side chain. Glutamine substitution is expected to increase the hydrophilicity of the peptide linker and may prevent the linker from folding itself. The sequence is shown below, and the sequence alignment with SEQ ID NO: 43 is shown in the figure. Figure 14 (Clustal Omega)

[0450] FVNQHLCGSHLVVEALALVCGERGFFYTDPTGGGPRRGIVEQCCTSICSLYQLENYCGGGGGQGGGGQGGGGQGGGGDGCPKCPAPEMLGGPSVFIFPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGT YRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPDIVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHPG(SEQ ID NO:53)

[0451] FVNQHLCGSHLVEALALVCGERGFFYTDPTGGGPRRGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:48)

[0452]

[0453] Two insulin-Fc fusion proteins were prepared in HEK293 cells according to Example 1 and purified using a protein A column according to Example 4. Their structures were confirmed by non-reducing and reducing CE-SDS according to Example 6, and their sequences were further identified by LC-MS and glycan removal according to Example 8. Their homodimer content (%) was measured by size exclusion chromatography according to Example 10, and their insulin receptor binding affinity was measured according to Example 12. The results are given in Table 11. As measured by a significant decrease in IC50 value, the incorporation of longer peptide linkers of different compositions (GGGGGQGGGGGGGGGGGGGGG (SEQ ID NO:13) for SEQ ID NO:53 vs. GGGGAGGGGG (SEQ ID NO:11) for SEQ ID NO:43) did indeed improve insulin receptor binding, suggesting that longer linkers may be a strategy to increase insulin receptor binding to other insulin-Fc fusion proteins. However, the incorporation of longer linkers still did not increase the homodimer potency to a level higher than the design target of greater than 50 mg / L.

[0454] Example 42: Attempting to delete a portion of the insulin peptide B chain of SEQ ID NO:63 to enhance the canine IgGB-containing Fc Homodimer titer of the relevant insulin-Fc fusion protein fragment.

[0455] The results of Example 41 demonstrate that the peptide linker can be modified to enhance the insulin receptor binding affinity of the insulin-Fc fusion protein of SEQ ID NO: 43, which contains native B10 and A8 amino acids. However, the peptide linker mutation failed to increase the homodimer titer sufficiently to meet the design goals. Since homodimer titer is a function of several properties, including intracellular synthesis and intracellular processing, it is presumed that insulin-Fc molecules may self-associate (i.e., aggregate) intramolecularly between the two monomers of the homodimer or intermolecularly between two or more individual homodimers during and after synthesis. Such aggregation would result in unacceptably low homodimer titers obtained from cell culture supernatants during the production processes described in Examples 1, 4, and 10. This potential interaction between insulin-Fc fusion protein molecules may be partly attributed to insulin's known tendency to self-associate and form aggregates. One method known in the art for reducing insulin's tendency to self-associate involves mutating amino acids near the C-terminus of the B chain. For example, lispro insulin (B28K; B29P mutation) and aspart insulin (B28D mutation) are well-known commercially available double-chain insulins with non-natural B-chain mutations that prevent association and aggregation, resulting in insulin in solution primarily in monomeric form. Another method to prevent aggregation involves amino acid deletions. For example, a double-chain insulin called depentapeptide insulin (DPPI); see Brange J., Dodson G.G., Edwards J., Holden PH, Whittingham JL 1997b. The “fibrinolytic insulin model derived from the X-ray crystal structure of monomeric insulin (depentapeptide insulin)” (Proteins 27 507–516) is identical to natural double-chain human insulin except for the removal of the five C-terminal amino acids of the B chain (YTPKT). Compared to natural double-chain human insulin, DPPI has a lower binding affinity to the insulin receptor, but it is completely monomeric in solution, meaning there is no significant association or aggregation between DPPI molecules. Therefore, to reduce the possibility of intramolecular and intermolecular self-association and improve the homodimer of the insulin-Fc fusion protein, several variant titers of the insulin-Fc fusion protein of SEQ ID NO: 43 were constructed using partial B-chain amino acid truncation and B-chain amino acid mutations as described above for DPPI, lispro insulin, and aspart insulin. The sequences are shown below, and the sequence alignment with SEQ ID NO: 43 is shown in [image / description]. Figure 15 Clustal Omega:

[0456] FVNQHLCGSHLVEAALALVCGERGFFYTDPGGGGGGGGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:51)

[0457] FVNQHLCGSHLVEALALVCGERGFFYTPGGGGGGGGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:52)

[0458] FVNQHLCGSHLVEALALVCGERGFFYTQGGGGGGGGIVEQCCTSICSLYQLENYCGGGGAGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFSGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFPPDILVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ IDNO:54)

[0459]

[0460]

[0461] Insulin-Fc fusion protein was prepared in HEK293 cells according to Example 1 and purified using a protein A column according to Example 4. Its structure was confirmed by non-reducing and reducing CE-SDS according to Example 6, and the sequence was further identified by LC-MS and glycan removal according to Example 8. Its homodimer content (%) was measured by size exclusion chromatography according to Example 10, and its insulin receptor binding affinity was measured according to Example 12. The results are given in Table 12. The homodimer potency of the obtained compounds was significantly increased only in one case (SEQ ID NO: 51), but unexpectedly, the insulin receptor affinity was increased in all mutant compounds (SEQ ID NO: 51, SEQ ID NO: 54, and SEQ ID NO: 52).

[0462] Example 43: Attempting to mutate the B, C, and A strands, truncate the B strand, and modify the adapter with SEQ Pancreatic islet ID NO:43 The combination of xin-Fc fusion proteins is used to further enhance homodimer potency and biological activity.

[0463] As demonstrated in Examples 39, 40, 41, and 42, no single strategy successfully combined an insulin polypeptide comprising non-immunogenic native B10 and A8 amino acids with a canine IgGB Fc fragment to form an insulin-Fc fusion protein with acceptable insulin receptor activity and homodimeric titer. Therefore, the concept of combining a longer C-chain, a longer peptide linker, and truncation of the C-terminal amino acid of the B-chain was adopted. Furthermore, to potentially further reduce the tendency for self-association and aggregation, additional point mutations were introduced at the hydrophobic amino acid residue sites of native insulin, using less hydrophobic amino acids, including those with side groups that carry a negative or positive charge at physiological pH. Exemplary mutations included tyrosine to alanine, tyrosine to glutamic acid, isoleucine to threonine, and phenylalanine to histidine. Additionally, to simplify analysis, in all cases, the cNg site of the canine IgGB Fc fragment was restored to its native asparagine. The sequences of these insulin-Fc fusion protein variants are shown below, with the resulting sequence alignment to SEQ ID NO:43 shown in [image / description missing]. Figure 16 Clustal Omega:

[0464] FVNQHLCGSHLVEALELVCGERGFFYTPKTGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:55)

[0465] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCNHGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:56)

[0466] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCNGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:28)

[0467] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:26)

[0468] FVNQHLCGSHLVEALELVCGERGFFYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:57)

[0469]

[0470]

[0471] Insulin-Fc fusion proteins were prepared in HEK293 cells according to Example 1 and purified using a protein A column according to Example 4. Their structure was confirmed by non-reducing and reducing CE-SDS according to Example 6, and the sequence was further identified by LC-MS and glycan removal according to Example 8. Their homodimer content (%) was measured by size exclusion chromatography according to Example 10, and their insulin receptor binding affinity was measured according to Example 12. The results are given in Table 13. The results indicate that a combination of reducing the hydrophobicity of certain B and A chain amino acids, using a longer and more flexible C-peptide sequence, truncating several C-terminal B chain amino acids, and using a longer peptide linker results in several useful insulin-Fc fusion proteins that meet the design criteria for minimum homodimer potency and insulin receptor binding activity. Compared to SEQ ID NO:43 or SEQ ID NO:55, SEQ ID NO:56, SEQ ID NO:28, SEQ ID NO:26, and SEQ ID NO:57 show preferred insulin receptor IC50 values ​​(less than 3000 nM) and preferred HEK homodimer potency values ​​(greater than 100 mg / L). Surprisingly, changing just a few amino acids resulted in a multiple-fold increase in insulin receptor affinity, and in the case of the insulin-Fc fusion protein of SEQ ID NO:26, the homodimer titer was significantly increased compared to the original insulin-Fc fusion protein of SEQ ID NO:43.

[0472] Example 44: Composed of the insulin polypeptide of SEQ ID NO:7, the peptide linker of SEQ ID NO:13, and SEQ ID NO:15 In vivo bioactivity, repeated-dose bioactivity, and immunogenicity of insulin-Fc fusion protein constructed from canine IgGB Fc fragment. sex.

[0473] Based on the positive homodimer titer and insulin receptor binding activity results from Example 43, two of the most promising insulin-Fc fusion proteins (SEQ ID NO:26 and SEQ ID NO:28) were tested in dogs to evaluate their bioactivity and immunogenicity under repeated dosing. Each compound comprises the longer, more hydrophilic peptide linker of SEQ ID NO:13 and the more readily available, less aggregated canine IgGB Fc fragment of SEQ ID NO:15. Crucially, both insulin-Fc fusion proteins comprise an insulin polypeptide (i.e., generally SEQ ID NO:6) with presumably low immunogenicity of the native B10 and A8 amino acids. In the case of the insulin-Fc fusion protein of SEQ ID NO:28, asparagine is present at position A21 (i.e., the insulin polypeptide comprises SEQ ID NO:8). In the case of the insulin-Fc fusion protein of SEQ ID NO:26, asparagine is absent at position A21 (i.e., the insulin polypeptide comprises SEQ ID NO:7).

[0474] Following the procedure of Example 21, the in vivo bioactivity of the insulin-Fc fusion protein of SEQ ID NO:28 was tested in N=1 dogs. Figure 17 The results of a single subcutaneous administration demonstrate that the insulin-Fc fusion protein of SEQ ID NO:28 is indeed biologically active in vivo, with a NAOC calculated according to the procedure of Example 22 of 1076 FBGL%·day·kg / mg. The pharmacokinetic properties of the insulin-Fc fusion protein of SEQ ID NO:28 were measured using ELISA according to the method of Example 23, and its elimination half-life was determined to be 3.5 days by fitting a two-compartment model to the data.

[0475] Then, following the procedure of Example 15, the bioactivity of repeated dosing was assessed by subcutaneously administering the insulin-Fc fusion protein of SEQ ID NO: 28 to N=1 dogs on days 14, 28, and 42 after the initial injection. When the dogs' FBGL% decreased too low, they were fed to raise their blood glucose to a safe level. The NAOC and NAOCR of each subsequent dosing were measured according to the general procedure of Example 22, calculated from the time of administration until the next administration. The NAOC and NAOCR shown in Table 14 demonstrate that the insulin-Fc fusion protein of SEQ ID NO: 28 maintained a NAOCR greater than 0.8 over four dosings, thus meeting the bioactivity design target for repeated dosing.

[0476]

[0477] The immunogenicity of the insulin-Fc fusion protein of SEQ ID NO: 28 was tested according to the procedure in Example 24. Figure 18 This demonstrates that the insulin-Fc fusion protein of SEQ ID NO: 28 did not exhibit significant immunogenicity in vivo, consistent with maintaining in vivo bioactivity throughout the repeated dosing study.

[0478] The bioactivity of the insulin-Fc fusion protein of SEQ ID NO:26, which lacks asparagine at A21 of the insulin polypeptide chain, was also evaluated in dogs after repeated administration. The compound was administered subcutaneously to N=1 dogs on days 0, 14, 28, and 42, according to the protocol of Example 22. When the FBGL% of the dogs decreased too low, they were fed to raise their blood glucose to a safe level. The NAOC of the first injection was an impressive 2278 FBGL%·day·kg / mg, indicating that the insulin-Fc fusion protein of SEQ ID NO:26 has satisfactory bioactivity in vivo, almost twice as potent as the insulin-Fc fusion protein of SEQ ID NO:28. The pharmacokinetic properties of the insulin-Fc fusion protein were measured using ELISA according to the method of Example 23, and its elimination half-life was determined to be approximately 4.1 ± 0.7 days by fitting a two-compartment model to the data. Figure 19 and Figure 20 Single-dose and multi-dose, multi-week blood glucose controls in animals receiving the homodimer of SEQ ID NO:26 are shown. NAOC and NAOCR were also measured for each subsequent dose according to the general procedure of Example 22, calculated from the time of administration to the time prior to the next administration. The NAOC and NAOCR shown in Table 15 demonstrate that the insulin-Fc fusion protein of SEQ ID NO:26 maintains a NAOCR greater than or equal to 1.0 over four administrations, thus meeting the repeated-dose bioactivity design target described in Example 29.

[0479] The immunogenicity of the insulin-Fc fusion protein of SEQ ID NO: 26 was tested according to the procedure in Example 24. Figure 21 This demonstrates that the insulin-Fc fusion protein of SEQ ID NO: 26 did not exhibit significant immunogenicity in vivo, consistent with maintaining in vivo bioactivity throughout the repeated dosing study.

[0480]

[0481] As discussed in the detailed description of this invention, there are known cleavage sites between the asparagine-glycine bonds (Vlasak, J., Ionescu, R., (2011) MAbs Vol.3, No.3pp 253-263). In the insulin-Fc fusion protein of SEQ ID NO:26, which contains a peptide linker of SEQ ID NO:13, asparagine is omitted at the 21st amino acid (i.e., A21) of the insulin polypeptide of SEQ ID NO:7, eliminating the possibility of enzymatic cleavage of the asparagine-glycine bond between the C-terminus of the A-chain and the N-terminus of the peptide linker. However, the insulin-Fc fusion protein of SEQ ID NO:28 contains the peptide linker of SEQ ID NO:13 and the insulin polypeptide of SEQ ID NO:8, which retains the asparagine at A21. Therefore, the insulin-Fc fusion protein of SEQ ID NO:28 is expected to be enzymatically digested in vivo during synthesis or after subcutaneous administration. However, quite unexpectedly, the insulin-Fc fusion protein of SEQ ID NO: 28 could be prepared in HEK cells with an acceptable homodimer titer and exhibited acceptable biological activity in vivo, with no indication that enzymatic digestion impaired its biological activity.

[0482] Example 45: Canine IgGB isotype Fc fragment for preferred insulin peptides containing SEQ ID NO:7 and SEQ ID NO:7 Optimal manufacturability and in vivo efficacy of insulin-Fc fusion protein with preferred peptide linker ID NO:13.

[0483] As described in Examples 43 and 44, a novel combination of insulin peptide and peptide linker was found to produce a non-immunogenic, high-yield, high-purity, and highly bioactive insulin-Fc fusion protein. However, the question remains regarding whether the canine IgGB Fc fragment remains the preferred isotype in terms of homodimer titer and bioactivity, as was the case with the insulin-Fc fusion proteins in Examples 32 and 33. Therefore, additional insulin-Fc fusion proteins were designed in which the insulin peptide (SEQ ID NO:7) and peptide linker (SEQ ID NO:13) of the insulin-Fc fusion protein of SEQ ID NO:26 remain unchanged, and the canine IgGB Fc fragment of SEQ ID NO:15 is replaced by the canine IgGA Fc fragment of SEQ ID NO:14, the canine IgGC Fc fragment of SEQ ID NO:16, or the canine IgGD Fc fragment of SEQ ID NO:17. The sequences of these resulting insulin-Fc fusion protein variants are shown below:

[0484] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQFNGTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:26)

[0485] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGRCTDTPPCPVPEPLGGPSVLIFPPKPKDILRITRTPEVTCVVLDLGREDPEVQISWFVDGKEVHTAKTQSREQQFNGTYRVVSVLPIEHQDWLTGKEFKCRVNHIDLPSPIERTISKARGRAHKPSVYVLPPSPKELSSSDTVSITCLIKDFYPPDIDVEWQSNGQQEPERKHRMTPPQLDEDGSYFLYSKLSVDKSRWQQGDPFTCAVMHETLQNHYTDLSLSHSPG(SEQ ID NO:58)

[0486] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGCNNCPCPGCGLLGGPSVFIFPPKPKDILVTARTPTVTCVVVDLDPENPEVQISWFVDSKQVQTANTQPREEQSNGTYRVVSVLPIGHQDWLSGKQFKCKVNNKALPSPIEEIISKTPGQAHQPNVYVLPPSRDEMSKNTVTLTCLVKDFFPPEIDVEWQSNGQQEPESKYRMTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQISLSHSPG(SEQ ID NO:59)

[0487] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGCISPCPVPESLGGPSVFIFPPKPKDILRITRTPEITCVVLDLGREDPEVQISWFVDGKEVHTAKTQPREQQFNSTYRVVSVLPIEHQDWLTGKEFKCRVNHIGLPSPIERTISKARGQAHQPSVYVLPPSPKELSSSDTVTLTCLIKDFFPPEIDVEWQSNGQPEPESKYHTTAPQLDEDGSYFLYSKLSVDKSRWQQGDTFTCAVMHEALQNHYTDLSLSHSPG(SEQ ID NO:60)。

[0488] Insulin-Fc fusion protein was prepared in HEK293 cells according to Example 1 and purified using Protein A or Protein G columns according to Example 4. Its structure was confirmed by non-reducing and reducing CE-SDS according to Example 6, and the sequence was further identified by LC-MS and glycan removal according to Example 8. Its homodimer content (%) was measured by size exclusion chromatography according to Example 10, and its insulin receptor binding affinity was measured according to Example 12. Additionally, the affinity of the insulin-Fc fusion protein for the canine FcRn receptor was measured according to Example 15. As shown in Table 16, the insulin-Fc fusion protein containing the canine IgGB Fc fragment of SEQ ID NO: 26 exhibited the highest homodimer titer among these sequences. The insulin-Fc fusion protein containing the canine IgGA Fc fragment of SEQ ID NO: 58 exhibited a poorer homodimer titer when purified using a Protein A column; however, when purified using a Protein G column, the homodimer titer was significantly improved, exceeding the design target of greater than 50 mg / L. The same applies to the insulin-Fc fusion protein containing the canine IgGC Fc fragment of SEQ ID NO: 59. The insulin-Fc fusion protein containing the canine IgGD Fc fragment of SEQ ID NO: 60 does not produce any compounds when purified with a protein A or protein G column. Therefore, as demonstrated by the insulin-Fc fusion protein containing a different insulin polypeptide (SEQ ID NO: 4) and a peptide linker (SEQ ID NO: 11) of SEQ ID NO: 36, canine IgGB is the preferred Fc fragment in terms of homodimer titer (see Example 32).

[0489]

[0490]

[0491] DNM = Not measured; # =Purified via protein A; = Purified via protein G.

[0492] The in vivo bioactivity of the insulin-Fc fusion protein containing the canine IgGA Fc fragment of SEQ ID NO: 58 purified via protein G was tested according to the procedure in Example 21. Figure 22 The results shown indicate that the insulin-Fc fusion protein of SEQ ID NO:58 has only certain biological activity in vivo, with a NAOC calculated according to Example 22 of only 174FBGL%·day·kg / mg.

[0493] The in vivo bioactivity of the insulin-Fc fusion protein containing the canine IgGC Fc fragment of SEQ ID NO: 59, purified via protein G, was tested according to the procedure of Example 21. Figure 23 The results shown indicate that the insulin-Fc fusion protein of SEQ ID NO:59 has only certain biological activity in vivo, with a NAOC calculated according to Example 22 of only 39FBGL%·day·kg / mg.

[0494] Therefore, as demonstrated by the insulin-Fc fusion protein containing different insulin peptides (SEQ ID NO:4) and peptide linkers (SEQ ID NO:11) of SEQ ID NO:36, canine IgGB is the preferred Fc fragment in terms of biological activity (see Examples 32 and 33 above and Table 16).

[0495] Example 46: No Glycosylated insulin-Fc fusion protein contains the insulin polypeptide of SEQ ID NO: 8, SEQ ID NO: 8, and SEQ ID NO: 8. The peptide linker at NO:13 and the canine IgGB Fc fragment are used to reduce the potential risk of immunogenicity.

[0496] While the insulin-Fc fusion protein of SEQ ID NO:26 satisfies all design objectives (Example 28), there may be or may not be an immunogenicity risk regarding prolonged treatment duration (e.g., 6 months, 1 year, 2 years, or longer), which could potentially affect the use of this insulin-Fc fusion protein for diabetes treatment. As detailed in the invention and in Examples 34 and 35, one possible reason for decreased bioactivity after repeated dosing is an undesirable interaction between the canine IgGB Fc fragment and the canine immune system, leading to the production of neutralizing antidrug antibodies. However, the results shown in Example 45 unexpectedly demonstrate that the canine IgGB isotype is the only option among the four canine IgG isotypes to produce the desired preparability and bioactivity. Therefore, further Fc mutations were explored to obtain a non-glycosylated insulin-Fc fusion protein with low Fc(γ)RI receptor binding, which would reduce the risk of long-term, chronic immunogenicity.

[0497] As described in the detailed description of the invention, a method for reducing Fc(γ)RI interaction involves mutating the cNg site of an Fc fragment to prevent glycosylation during synthesis in host cells. Therefore, a cNg site mutation is performed on the Fc fragment region of SEQ ID NO:26 to reduce the binding affinity of the Fc fragment to the Fc(γ) receptor in vivo, as measured by binding in an in vitro human Fc(γ)RI assay as described in Example 15. The cNg site in the insulin-Fc fusion protein of SEQ ID NO:26 is located at cNg-NB151. Mutations in SEQ ID NO:26 include SEQ ID NO:62 containing the cNg-NB151-S mutation and SEQ ID NO:61 containing the same cNg-NB151-S mutation and the NB119-A mutation. The inclusion of NB119-A was an attempt to further reduce interaction with Fc(γ)RI, as described in Lo, M. et al., “Effector attenuating substitutions that maintain antibody stability and reduce toxicity in mice”, J. Biol. Chem. (2017), pp. 1-20. The full amino acid sequences of the resulting insulin-F are only those described for the mouse antibody. The full amino acid sequence of the resulting insulin-Fc fusion protein (NB119 and NB151 sites are underlined for clarity) is listed below along with their sequence alignment (Clustal Omega). Figure 24 middle:

[0498] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPPKPKDTLLIARTPEVTCVVV A LDPEDPEVQISWFVDGKQMQTAKTQPREEQF S GTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG(SEQ ID NO:61)

[0499] FVNQHLCGSHLVEALELVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQF S GTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG (SEQ ID NO: 62).

[0500] Insulin-Fc fusion protein was prepared in HEK293 cells according to Example 1 and purified using a protein A column according to Example 4. Its structure was confirmed by non-reducing and reducing CE-SDS according to Example 6, and the sequence was further identified by LC-MS and glycan removal according to Example 8. Its homodimer content (%) was measured by size exclusion chromatography according to Example 10, and its insulin receptor binding affinity was measured according to Example 12. As shown in Table 17, the cNg-NB151-S mutation in the incorporated Fc fragment reduced the homodimer content (%), indicating an unacceptably high level of polymerization (i.e., the homodimer content decreased to slightly above 70%).

[0501]

[0502] Following the procedure of Example 21, the in vivo bioactivity of the insulin-Fc fusion proteins of SEQ ID NO:61 and SEQ ID NO:62 was tested in N=1 dogs. Figure 25 The results of a single subcutaneous administration shown in the figure indicate that the in vivo bioactivity of both compounds was significantly lower than that of the insulin-Fc fusion protein of SEQ ID NO:26 (NAOC = 574 FBGL%·day·kg / mg for SEQ ID NO:62; NAOC = 921 FBGL%·day·kg / mg for SEQ ID NO:61). The results suggest that incorporating the cNg-NB151-S mutation into the Fc fragment to produce a non-glycosylated variant of the insulin-Fc fusion protein of SEQ ID NO:26 unexpectedly reduced the in vivo bioactivity of the resulting compounds.

[0503] To reduce aggregation and improve the bioactivity of the insulin-Fc fusion protein containing the cNg-NB151-S mutation of SEQ ID NO:62, various insulin-peptide B-chain variants were investigated, and the mutation in this region was considered to be the cause of aggregation. The insulin-Fc fusion protein was prepared in HEK293 cells according to Example 1 and purified using a protein A column according to Example 4. Its structure was confirmed by non-reducing and reducing CE-SDS according to Example 6, and the sequence was further identified by LC-MS and glycan removal according to Example 8. Its homodimer content (%) was measured by size exclusion chromatography according to Example 10. Unexpectedly, among the tested B-chain variants, an insulin-Fc fusion protein (SEQ ID NO:30) containing a tyrosine-to-alanine substitution at the 16th amino acid (i.e., B16) from the N-terminus of the B chain exhibited high homodimer potency (105 mg / L) and low aggregation (99% homodimer), yielding a homodimer potency of 104 mg / L. The insulin receptor binding measured according to Example 12 was acceptable, with an IC50 of 2040 nM. The FcRn receptor binding affinity EC50 measured according to Example 16 was 1194 ng / mL. The pharmacokinetic properties of the insulin-Fc fusion protein of SEQ ID NO:30 were measured using ELISA according to the method of Example 23, and its elimination half-life was determined to be approximately 4.1 ± 0.7 days by fitting a two-compartment model to the data. The sequence of SEQ ID NO:30 is shown below (B16A and cNg-NB151-S mutations are underlined for clarity).

[0504] FVNQHLCGSHLVEAL A LVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGQGGGGGDCPKCPAPEMLGGPSVFIFPPPKPKDTLLIARTPEVTCVVVDLDPEDPEVQISWFVDGKQMQTAKTQPREEQF S GTYRVVSVLPIGHQDWLKGKQFTCKVNNKALPSPIERTISKARGQAHQPSVYVLPPSREELSKNTVSLTCLIKDFFPPDIDVEWQSNGQQEPESKYRTTPPQLDEDGSYFLYSKLSVDKSRWQRGDTFICAVMHEALHNHYTQESLSHSPG (SEQ ID NO: 30).

[0505] The repeated-dose bioactivity of the insulin-Fc fusion protein of SEQ ID NO: 30 was then evaluated in dogs. Following the protocol of Example 22, the compound was administered subcutaneously to N=1 dogs on days 0, 7, 14, and 28. When the dogs' FBGL% decreased too low, they were fed to raise their blood glucose to a safe level. Unexpectedly, the NAOC of the insulin-Fc fusion protein containing the B16A mutation of SEQ ID NO: 30 was significantly higher (1185 FBGL%·day·kg / mg) compared to the insulin-Fc fusion protein of SEQ ID NO: 62. The in vivo bioactivity after the first dose is illustrated in [Figure / Image]. Figure 26 The pharmacokinetic properties of the compound were also measured using ELISA according to the method of Example 23, and its elimination half-life was determined to be 3.5 days by fitting a two-compartment model to the data. The NAOC and NAOCR of each subsequent dose were also measured according to the general procedure of Example 22, calculated from the time of administration to the time prior to the next administration. The NAOC and NAOCR shown in Table 18 demonstrate that the insulin-Fc fusion protein of SEQ ID NO: 30 maintains a NAOCR greater than or equal to 0.6 over four administrations, thus meeting the design target for repeated-dose bioactivity. In summary, the results indicate the necessity of mutating the insulin B chain sequence to obtain a suitable, non-glycosylated cNg-S variant of SEQ ID NO: 26. Therefore, for a non-glycosylated insulin-Fc fusion protein containing a cNg-mutated canine IgGB Fc fragment, the insulin peptide of SEQ ID NO: 10 is preferred.

[0506]

[0507] Finally, following the procedure of Example 15, the potential for interaction between the selected compounds and the immune system was tested by measuring their Fc(γ) receptor binding activity. Table 19 compares the Fc(γ) receptor binding of these insulin-Fc fusions with the Fc(γ) receptor binding of the insulin-Fc fusion protein of SEQ ID NO:36. It can be seen that the non-glycosylated insulin-Fc fusion protein (achieved through a cNg-S mutation) exhibits the lowest Fc(γ) receptor binding rate with SEQ ID NO:36.

[0508]

[0509]

[0510] Example 47: Preparation of a superior strain containing a canine IgGB-derived Fc fragment using a stably transfected CHO cell line. An exemplary CHO-based production run of the insulin-Fc fusion protein is selected.

[0511] A single CHO cell line stably transfected with the vector encoding SEQ ID NO:26 or SEQ ID NO:30 was constructed as described in Example 2. A 14-day production run (0.5-2.0 L culture medium scale) was conducted in batch shake flasks seeded at 500,000 cells / mL in an incubator-shakeer set to 37°C and 5% CO2, and performed as described in Example 2 above, except that CD OptiCHO was replaced with Dynamis as the medium (ThermoFisher) and high-efficiency nutrient C (ThermoFisher) was used as the nutrient. Nutrients were added at 3% v / v...

Claims

1. A fusion protein comprising an insulin polypeptide and an Fc fragment, wherein the insulin polypeptide and the Fc fragment are linked by a linker consisting of a domain (N-terminus)-insulin polypeptide-linker-Fc fragment-(C-terminus) in a direction from N-terminus to C-terminus. in, The Fc fragment is human in origin and is shown in the following sequence: DKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYX1STYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEK TISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:77), where X1 is S or R, The insulin polypeptide described herein is as follows: FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYC (SEQ ID NO: 10); and The connectors described herein are in the following sequence: GGGGGAGGGGAGGGGAGGGGG (SEQ ID NO: 67); GGGGGQGGGGQGGGGQGGGGGQGGGG (SEQ ID NO:99); or GGGGGQGGGGQGGGGQGGGGG (SEQ ID NO: 13).

2. The fusion protein of claim 1, comprising an insulin polypeptide and an Fc fragment, wherein the insulin polypeptide and the Fc fragment are linked by a linker, and wherein the fusion protein has the following sequence: FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGIVEQCCTSTCSLDQLENYCGGGGGAGGGGAGGGGAGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQY SSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG(SEQ ID NO:87).

3. The fusion protein according to claim 1, wherein the fusion protein has the following sequence: FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYSSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 78).

4. The fusion protein according to claim 1, wherein the fusion protein has the following sequence: FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGGQGGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYRSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 84).

5. The fusion protein according to claim 1, wherein the fusion protein has the following sequence: FVNQHLCGSHLVEALALVCGERGFHYGGGGGGSGGGGGGIVEQCCTSTCSLDQLENYCGGGGGQGGGGQGGGGGQGGGGGGQGGGGDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYSSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG (SEQ ID NO: 96).

6. The fusion protein of claim 1, wherein the fusion protein is a homodimer, and wherein the homodimer percentage of the fusion protein is greater than 90%.

7. The fusion protein according to claim 1, wherein the fusion protein is prepared using either HEK293 cells or CHO cells, and the homodimer titer obtained after purification using protein A beads or protein A columns is greater than 150 mg / L.

8. The fusion protein according to claim 1, wherein the insulin receptor IC50 of the fusion protein is less than or equal to 5000 nM.

9. The fusion protein according to claim 1, wherein the human FcRn receptor EC50 of the fusion protein is less than or equal to 1000 ng / mL.

10. The fusion protein according to claim 1, wherein the OD450 ratio of the fusion protein to the human Fc(γ)RI receptor at a biotinylated-Fc(γ)RI concentration of 3000 ng / mL is less than or equal to 0.

50.

11. The fusion protein according to claim 1, wherein the OD450 ratio of human C1q at a biotinylated-C1q concentration of 1000 ng / mL is less than or equal to 0.

35.

12. A pharmaceutical composition comprising a fusion protein according to any one of claims 1 to 11.

13. The pharmaceutical composition of claim 12, wherein the fusion protein is present in the pharmaceutical composition at a concentration of 3 mg / mL or higher.

14. The pharmaceutical composition according to claim 12, wherein the composition is suitable for subcutaneous administration.

15. Use of the fusion protein according to any one of claims 1 to 11 or the pharmaceutical composition according to any one of claims 12 to 14 in the preparation of a medicament for lowering blood glucose levels in patients.

16. The use according to claim 15, wherein the patient is diagnosed with diabetes.

17. The use according to claim 15, wherein the fusion protein is administered subcutaneously.

18. A cell engineered to express the fusion protein of any one of claims 1 to 11.

19. The cell of claim 18, wherein the cell is transfected with a nucleic acid encoding the fusion protein.

20. The cell of claim 18, wherein the cell is a HEK293 cell or a CHO cell.

21. A cDNA encoding the fusion protein of any one of claims 1 to 11.

22. A cDNA encoding a fusion protein of SEQ ID NO:87, comprising the following nucleic acid sequence: atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagc acctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatc gtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtgcaggaggcggtggagccggtggaggtggggctgg aggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccc tcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcata atgccaagacaaagccgcgggaggagcagtacagcagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ ID NO:88).

23. A cDNA encoding a fusion protein of SEQ ID NO:78, which comprises the following nucleic acid sequence: atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagc acctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcata atgccaagacaaagccgcgggaggagcagtacagcagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ ID NO:79).

24. A cDNA encoding a fusion protein of SEQ ID NO:84, comprising the following nucleic acid sequence: atggaatggagctgggtctttctcttcttcctgtcagtaacgactggtgtccactccttcgtgaaccagc acctgtgcggctcccacctggtggaagctctggcactcgtgtgcggcgagcggggcttccactacgggggtggcggaggaggttctggtggcggcggaggcatcgtggaacagtgctgcacctccacctgctccctggaccagctggaaaactactgcggtggcggaggtggtcaaggaggcggtggacagggtggaggtgggcagggaggaggcgggggagacaaaactcacacatgcccaccgtgcccagcacctgaactcctggggggaccgtcagtcttcctcttccccccaaaacccaaggacaccctcatgatctcccggacccctgaggtcacatgcgtggtggtggacgtgagccacgaagaccctgaggtcaagttcaactggtacgtggacggcgtggaggtgcata atgccaagacaaagccgcgggaggagcagtacagaagcacgtaccgtgtggtcagcgtcctcaccgtcctgcaccaggactggctgaatggcaaggagtacaagtgcaaggtctccaacaaagccctcccagcccccatcgagaaaaccatctccaaagccaaagggcagccccgagaaccacaggtgtacaccctgcccccatcccgggatgagctgaccaagaaccaggtcagcctgacctgcctggtcaaaggcttctatcccagcgacatcgccgtggagtgggagagcaatgggcagccggagaacaactacaagaccacgcctcccgtgctggactccgacggctccttcttcctctacagcaagctcaccgtggacaagagcaggtggcagcaggggaacgtcttctcatgctccgtgatgcatgaggctctgcacaaccactacacgcagaagagcctctccctgtctccgggttag(SEQ ID NO:85)。

Images