EA041034B1 - PARTIAL INSULIN RECEPTOR AGONISTS - Google Patents

Description

Cross-reference to related applications

Background of the Invention (1) The technical field to which the invention pertains.

The present invention relates to insulin dimers and dimers of insulin analogs that act as partial agonists at the insulin receptor.

(2) Description of the prior art.

Insulin is the mainstay of therapy for patients with type 1 diabetes mellitus (T1DM) and many patients with type 2 diabetes mellitus (T2DM), having been prescribed to nearly one-third of patients in the US of all users of antidiabetic drugs over the past decade. The global market for insulin was US$20.4 billion in 2013 and is growing faster than all other antidiabetic drugs combined. However, the problems of current insulin therapies, including the narrow TI to hypoglycemia and weight gain, limit their wider use and the ability for patients to achieve ideal glycemic control.

In addition to secreting prandial insulin in response to meals, the pancreas releases insulin at a basal rate determined primarily by plasma glucose levels to maintain appropriate regulation of fasting glucose levels. This is achieved mainly by controlling hepatic glucose release through the hepatoprotective action of endogenous insulin. Modern insulin analogues include fast-acting and basal insulins, as well as mixtures thereof. Rapid-acting insulin analogs (RAAs) are designed to control postprandial hyperglycemia, while long-acting insulins regulate basal glucose levels. Long-acting insulins use all T1DM (in combination with prandial injections) and most patients with T2DM start insulin therapy with a basal product. Basal insulin consumption is growing rapidly as the worldwide population of diabetics (especially T2DM) is increasing.

Despite ongoing development efforts over the past few decades, the long-acting insulins available are still not optimized compared to physiological basal insulin. This is partly because the focus has been on improving the linearity of the PK of these analogues, but not on eliminating the relative excess insulinization of peripheral tissues that contributes to the increased risk of hypoglycemia. As a result, hypoglycemia remains a major medical risk with a huge burden on patients and causes significant morbidity and mortality.

Brief summary of the invention

The present invention provides compounds containing up to two insulin molecules covalently linked to form a dimer of insulin molecules that can activate the insulin receptor with regular insulin-like activity but with reduced peak activity. These compounds are insulin receptor partial agonists (IPRAs): they behave like other insulin analogues, effectively lowering glucose but with a lower risk of hypoglycemia.

Provides covalent insulin dimers-partial insulin receptor agonists formulated as novel and once-daily converting basal insulins that exhibit an improved therapeutic index (TI) over current standard of care (SOC) basal insulins. In one embodiment, the IPRAs of the present invention can effectively lower glucose levels with a reduced risk of hypoglycemia in a diabetic minipig and have the properties of once-daily (QD) basal insulin. The improved TI may allow practitioners to more aggressively dose the IRPA of the present invention to achieve fasting glucose control targets. Strict control of fasting glucose and HbA1c with IRPA may allow them to serve as 1) a long-acting single-acting insulin with an improved efficacy and safety profile in T2DM and 2) an improved basal basal insulin in T1DM (and sometimes T2DM) for use with additional doses prandial fast-acting insulin analogs (RAA). Thus, the present invention provides the following embodiments.

The present invention provides an insulin receptor partial agonist or insulin dimer comprising a first heterodimer of insulins or insulin analogs and a second heterodimer of insulins or insulin analogs, each heterodimer comprising an A chain polypeptide and a B chain polypeptide, wherein the A chain polypeptide and the B polypeptide -chains are linked together via interchain disulfide bonds; wherein the first and second heterodimers of insulins or insulin analogs are covalently linked together via a linking moiety connecting amino acid side chains at or near the carboxyl ends of the two respective B chain polypeptides; and wherein at least one amino terminus of the A chain polypeptides and B chain polypeptides is covalently linked to the substituent, provided that the linking moiety does not include a disulfide bond. In certain aspects, at least the amino termini of the A chain polypeptide and the B chain polypeptide of the first insulin or insulin analog are covalently linked to a substituent.

- 1 041034

In certain aspects of the insulin receptor partial agonist or insulin dimer, the amino terminus of each A chain polypeptide and each B chain polypeptide is covalently linked to a substituent. In certain aspects, the amino terminus of the A chain polypeptide and the B chain polypeptide of the first insulin or insulin analog and the amino terminus of the A chain polypeptide and the B chain polypeptide of the second insulin or insulin analog are covalently linked to the substituent. In embodiments in which the amino termini of the first and second insulins or insulin analogs are covalently linked to the substituent, the substituent at the amino ends of the A chain and B chain polypeptides of the first insulin or insulin analog may be the same as the substituent at the amino ends of the A polypeptides. β-chain and B-chain of a second insulin or insulin analogue. In embodiments in which the amino ends of the first and second insulins or insulin analogs are covalently linked to the substituent, the substituent at the amino ends of the A chain and B chain polypeptides of the first insulin or insulin analog may be different from the substituent at the amino ends of the A chain and B-chain of a second insulin or insulin analogue.

In another aspect of the insulin receptor partial agonist or insulin dimer, the first and second heterodimers of the insulins or insulin analogs are the same, or the first and second heterodimers of the insulins or insulin analogs are different.

In yet another aspect of the insulin receptor partial agonist or insulin dimer, the binding moiety covalently links the first heterodimer of insulins or insulin analogs and the second heterodimer of insulins or insulin analogs via an epsilon-amino group of a lysine residue at or near the carboxyl ends of their respective chain polypeptides.

In yet another aspect of an insulin receptor partial agonist or insulin dimer, the substituent has the general formula RC(O)- where R may be R/CH2, R'NH, R'O, and R' may be H, a linear alkyl chain, amino acid, peptide, PEG, saccharides, i.e. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is selected from the group consisting of acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, isobutyl, methoxyacetyl, glycine, aminoethylglucose (AEG), AEG-C6, PEG1, PEG2, N-dimethyl, and alkoxycarbonyl.

In certain aspects of an insulin receptor partial agonist or insulin dimer, each A chain polypeptide independently comprises the amino acid sequence GX ₂ X ₃ EQCCX ₈ SICSLYQLX ₁₇ NX ₁₉ CX ₂₃ (SEQ ID NO: 3) and each B chain polypeptide independently comprises the amino acid sequence X ₂₅ LCGX ₂₉ X ₃₀ LVEALYLVCGERGFX ₂₇ YTX ₃₁ X ₃₂ (SEQ ID NO: 4) or X ₂₂ VNQX ₂₅ X ₂₆ CGX ₂₉ X ₃₀ LVEALYLVCGERGFX ₂₇ YTX ₃₁ X ₃₂ X ₃₃ X ₃₄ X ₃₅ (SEQ ID NO: 5), moreover X ₂ is isoleucine or threonine; X ₃ is valine, glycine or leucine; X8 is threonine or histidine; X ₁₇ is glutamic acid or glutamine; X ₁₉ is tyrosine, 4-methoxyphenylalanine, alanine or 4-aminophenylalanine; X ₂₃ is asparagine or glycine; X ₂₂ is phenylalanine or deaminophenylalanine; X ₂₅ is histidine or threonine; X ₂₆ is leucine or glycine; X ₂₇ is phenylalanine or aspartic acid; X ₂₉ is alanine, glycine or series; X ₃₀ is histidine, aspartic acid, glutamic acid, homocysteine acid or cysteic acid; X ₃₁ is aspartic acid, proline or lysine; X ₃₂ is lysine or proline; X ₃₃ is threonine, alanine or absent; X ₃₄ is arginine or absent; and X ₃₅ is arginine or absent; with the proviso that at least one of X ₃₁ or X ₃₂ is a lysine.

In certain aspects of the insulin receptor partial agonist or insulin dimer, the first and second insulins or insulin analogs are independently native human insulin, insulin lispro, insulin aspart, desB30 insulin, or insulin glargine.

In certain aspects of insulin receptor partial agonists or insulin dimers, the binding moiety may be an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. The linking moiety may be a divalent, straight or branched, saturated or unsaturated, optionally substituted C1- _C20 hydrocarbon chain in which one or more methylene units are optionally and independently substituted with -O-, -S-, -N(R)- , -C(O)-, C(O)O-, OC(O)-, -N(R)C(O)-, -C(O)N(R)-, -S(O)-, -S(O)2-, N(R)SO2-, SO2N(R)-, a heterocyclic group, an aryl group, or a heteroaryl group, wherein each occurrence of R is independently hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic fragment, aryl fragment, heteroaryl fragment or heteroaliphatic fragment.

In yet another aspect of an insulin receptor partial agonist or insulin dimer, the binding moiety is an acyl moiety, -C(O)RC(O)-, where R is an alkyl chain, a poly(ethylene glycol) (PEG) chain containing an amide chain containing triazole(s) chain, cyclooctyne-containing fragment, substituted acyl chain or polyethylene glycol (PEG) chain.

- 2 041034

In another aspect, the linking moiety is an alkyldioyl, -C(O)(CH2) _n C(O)-, where n=0-45, including but not limited to oxalyl (C2) moiety, succinyl (C4) moiety, adipoyl (C6) moiety, suberoyl (C8) moiety, decanedioyl (C10) moiety, dodecanedioyl (C12) moiety, tetradecanedioyl (C14) moiety, or hexadecanedioyl (C16) moiety.

The present invention further provides an insulin receptor partial agonist or insulin dimer having the formula D1-LD ² , wherein D ¹ and D ² are independently an insulin or insulin analog polypeptide, each insulin polypeptide being a heterodimer containing an A chain polypeptide and a B polypeptide chains linked together by interchain disulfide bonds; L is a linking moiety, with one end of the linker moiety attached to an amino acid residue at or near the carboxyl group of D ¹ and the other end of the linker moiety attached to an amino acid residue at or near the carboxyl end of D ² , provided that L is not includes a disulfide bond; and wherein the first and second insulin or insulin analog polypeptides include a substituent attached to the amino terminus of the A chain polypeptide and the B chain polypeptide.

In another aspect of the insulin receptor partial agonist or insulin dimer, D ¹ and D ² are the same, or D ¹ and D ² are different.

In another aspect of the insulin receptor partial agonist or insulin dimer, the binding moiety covalently links D ¹ and D ² via an epsilon-amino group of a lysine residue at or near the carboxyl ends of D ¹ and D ² .

In yet another aspect of an insulin receptor partial agonist or insulin dimer, the substituent has the general formula RC(O)- where R may be R/CH2, R'NH, R'O, and R' may be H, a linear alkyl chain, amino acid, peptide, PEG, saccharides, i.e. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is selected from the group consisting of acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, isobutyl, methoxyacetyl, glycine, aminoethylglucose (AEG), AEG-C6, PEG1, PEG2, N-dimethyl, and alkoxycarbonyl.

In certain aspects of an insulin receptor partial agonist or insulin dimer, each A chain polypeptide independently comprises the amino acid sequence GX ₂ X ₃ EQCCX ₈ SICSLYQLX ₁₇ NX ₁₉ CX ₂₃ (SEQ ID NO: 3) and each B chain polypeptide independently comprises the amino acid sequence X ₂₅ LCGX ₂₉ X ₃₀ LVEALYLVCGERGFX ₂₇ YTX ₃₁ X ₃₂ (SEQ ID NO: 4) or X ₂₂ VNQX ₂₅ X ₂₆ CGX ₂₉ X ₃₀ LVEALYLVCGERGFX ₂₇ YTX ₃₁ X ₃₂ X ₃₃ X ₃₄ X ₃₅ (SEQ ID NO: 5), moreover X ₂ is isoleucine or threonine; X ₃ is valine, glycine or leucine; X ₈ is threonine or histidine; X17 is glutamic acid or glutamine; X ₁₉ is tyrosine, 4-methoxyphenylalanine, alanine or 4-aminophenylalanine; X ₂₃ is asparagine or glycine; X ₂₂ is phenylalanine or deaminophenylalanine; X ₂₅ is histidine or threonine; X ₂₆ is leucine or glycine; X ₂₇ is phenylalanine or aspartic acid; X29 is alanine, glycine or series; X30 is histidine, aspartic acid, glutamic acid, homocysteine acid or cysteic acid; X ₃₁ is aspartic acid, proline or lysine; X ₃₂ is lysine or proline; X ₃₃ is threonine, alanine or absent; X ₃₄ is arginine or absent; and X ₃₅ is arginine or absent; with the proviso that at least one of X ₃₁ or X ₃₂ is a lysine.

In another aspect, the insulin receptor partial agonist or insulin dimer D ¹ and D ² are independently native human insulin, insulin lispro, insulin aspart, desB30 insulin, or insulin glargine.

In certain aspects of insulin receptor partial agonists or insulin dimers, the binding moiety may be an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. The linking moiety may be a divalent, straight or branched, saturated or unsaturated, optionally substituted C1- _C20 hydrocarbon chain in which one or more methylene units are optionally and independently substituted with -O-, -S-, -N(R)- , -C(O)-, C(O)O-, OC(O)-, -N(R)C(O)-, -C(O)N(R)-, -S(O)-, -S(O)2-, N(R)SO2-, SO2N(R)-, a heterocyclic group, an aryl group, or a heteroaryl group, wherein each occurrence of R is independently hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic fragment, aryl fragment, heteroaryl fragment or heteroaliphatic fragment.

In yet another aspect of an insulin receptor partial agonist or insulin dimer, the binding moiety is an acyl moiety, -C(O)RC(O)-, where R is an alkyl chain, a poly(ethylene glycol) (PEG) chain containing an amide chain containing triazole(s) chain, cyclooctyne-containing moiety, substituted acyl chain, or polyethylene glycol (PEG) chain.

In another aspect of an insulin receptor partial agonist or insulin dimer, the binding moiety is an alkyldioyl, -C(O)(CH2) _n C(O)-, where n=0-45, including but not limited to oxalyl ( C2) fragment, succinyl (C4) fragment, adipoyl (C6) fragment, suberoyl (C8) fragment, decanedioyl (C10) fragment, dodecanedioyl (C12) fragment, tetradecanedioyl (C14) fragment or hexadecanedioyl (C16) fragment.

Also provided are compositions comprising any of the aforementioned partial insulin receptor agonists or an insulin dimer and a pharmaceutically acceptable carrier.

The present invention further provides an insulin receptor partial agonist or insulin dimer comprising a first heterodimer of insulins or insulin analogs and a second heterodimer of insulins or insulin analogs, each heterodimer comprising an Acepi polypeptide and a B chain polypeptide, wherein the A chain polypeptide and the B chain polypeptide the chains are linked together via interchain disulfide bonds; wherein the first and second heterodimers of insulins or insulin analogs are covalently linked together via a linking moiety connecting amino acid side chains at or near the carboxyl ends of the two respective B chain polypeptides; and optionally wherein the amino terminus of at least one of the A chain polypeptides and B chain polypeptides of the first insulin polypeptide or the second insulin polypeptide is covalently linked to the substituent, provided (1) that the linking moiety does not include a disulfide bond, and (2) that when the insulin or insulin analog is not human insulin or insulin analog, and the amino-terminus of the A chain polypeptide and the B chain polypeptide do not include a substituent, then the binding moiety is not an oxalyl (C2) moiety , suberoyl (C8) fragment or dodecanedioyl (C12) fragment.

In another aspect of the insulin receptor partial agonist or insulin dimer, the first and second heterodimers of the insulins or insulin analogs are the same, or the first and second heterodimers of the insulins or insulin analogs are different.

In another aspect of the insulin receptor partial agonist or insulin dimer, the binding moiety covalently links the first heterodimer of insulins or insulin analogs and the second heterodimer of insulins or insulin analogs via an epsilon-amino group of a lysine residue at or near the carboxyl ends of their respective B chain polypeptides.

In yet another aspect of an insulin receptor partial agonist or insulin dimer, the substituent has the general formula RC(O)- where R may be R/CH2, R'NH, R'O, and R' may be H, a linear alkyl chain, amino acid, peptide, PEG, saccharides, i.e. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is selected from the group consisting of acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, isobutyl, methoxyacetyl, glycine, aminoethylglucose (AEG), AEG-C6, PEG1, PEG2, N-dimethyl, and alkoxycarbonyl.

In another aspect of the insulin receptor partial agonist or insulin dimer, the first and second insulins or insulin analogs are independently native human insulin, insulin lispro, insulin aspart, desB30 insulin, or insulin glargine.

In certain aspects of an insulin receptor partial agonist or insulin dimer, each A chain polypeptide independently comprises the amino acid sequence GX ₂ X ₃ EQCCX ₈ SICSLYQLX ₁₇ NX ₁₉ CX ₂₃ (SEQ ID NO: 3) and each B chain polypeptide independently comprises the amino acid sequence X ₂₅ LCGX ₂₉ X ₃₀ LVEALYLVCGERGFX ₂₇ YTX ₃₁ X ₃₂ (SEQ ID NO: 4) or X ₂₂ VNQX ₂₅ X ₂₆ CGX ₂₉ X ₃₀ LVEALYLVCGERGFX ₂₇ YTX ₃₁ X ₃₂ X ₃₃ X ₃₄ X ₃₅ (SEQ ID NO: 5), moreover X ₂ is isoleucine or threonine; X ₃ is valine, glycine or leucine; X ₈ is threonine or histidine; X ₁₇ is glutamic acid or glutamine; X ₁₉ is tyrosine, 4-methoxyphenylalanine, alanine or 4-aminophenylalanine; X ₂₃ is asparagine or glycine; X ₂₂ is phenylalanine or deaminophenylalanine; X ₂₅ is histidine or threonine; X ₂₆ is leucine or glycine; X ₂₇ is phenylalanine or aspartic acid; X29 is alanine, glycine or series; X30 is histidine, aspartic acid, glutamic acid, homocysteine acid or cysteic acid; X ₃₁ is aspartic acid, proline or lysine; X ₃₂ is lysine or proline; X ₃₃ is threonine, alanine or absent; X ₃₄ is arginine or absent; and X ₃₅ is arginine or absent; with the proviso that at least one of X ₃₁ or X ₃₂ is a lysine.

In certain aspects of insulin receptor partial agonists or insulin dimers, the binding moiety may be an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. The linking moiety may be a divalent, straight or branched, saturated or unsaturated, optionally substituted C1- _C20 hydrocarbon chain in which one or more methylene units are optionally and independently substituted with -O-, -S-, -N(R)- , -C(O)-, C(O)O-, OS(O)-, -N(R)C(O)-, -C(O)N(R)-, -S(O)-, -S(O)2-, -N(R)SO2-, SO2N(R)-, a heterocyclic group, an aryl group, or a heteroaryl group, where each occurrence of R is independently hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl

- 4 041034 fragment, aliphatic fragment, aryl fragment, heteroaryl fragment or heteroaliphatic fragment.

In yet another aspect of an insulin receptor partial agonist or insulin dimer, the binding moiety is an acyl moiety, -C(O)RC(O)-, where R is an alkyl chain, a poly(ethylene glycol) (PEG) chain containing an amide chain containing triazole(s) chain, cyclooctyne-containing fragment, substituted acyl chain or polyethylene glycol (PEG) chain.

In another aspect of the insulin receptor partial agonist or insulin dimer, the binding moiety is a C ₂ -C ₂₀ acyl moiety.

In certain aspects, the linking moiety is an alkyldioyl, -C(O)(CH2) _n C(O)-, where n=0-45, including, but not limited to, oxalyl (C2) moiety, succinyl (C4) moiety, adipoyl (C6) moiety, suberoyl (C8) moiety, decanedioyl (C10) moiety, dodecanedioyl (C12) moiety, tetradecanedioyl (C14) moiety, or hexadecanedioyl (C16) moiety.

Also provided are compositions containing any of the aforementioned partial insulin receptor agonists or insulin dimers and a pharmaceutically acceptable carrier.

The present invention further provides an insulin receptor partial agonist or insulin dimer having the formula D1-LD ² , wherein D ¹ and D ² are independently an insulin or insulin analog polypeptide, each insulin polypeptide being a heterodimer containing an A chain polypeptide and a B polypeptide chains linked together by interchain disulfide bonds; L is a linking moiety, with one end of the linker moiety attached to an amino acid residue at or near the carboxyl group of D ¹ and the other end of the linker moiety attached to an amino acid residue at or near the carboxyl end of D ² , provided that L is not includes a disulfide bond; and, optionally, at least one of D ¹ or D ² includes a substituent attached to the amino terminus of the A chain polypeptide or the D ¹ or D ² B chain polypeptide; provided (1) that the linking moiety does not include a disulfide bond, and (2) that when the amino termini of the A chain polypeptide and the B chain polypeptide do not include a substituent, the linking moiety is not an oxalyl (C2) moiety , suberoyl (C8) fragment or dodecanedioyl (C12) fragment.

In another aspect of the insulin receptor partial agonist or insulin dimer, D ¹ and D ² are the same, or D ¹ and D ² are different.

In another aspect of the insulin receptor partial agonist or insulin dimer, the binding moiety covalently links D ¹ and D ² via an epsilon-amino group of a lysine residue at or near the carboxyl ends of D ¹ and D ² .

In yet another aspect of an insulin receptor partial agonist or insulin dimer, the substituent has the general formula RC(O)- where R may be R/CH2, R'NH, R'O, and R' may be H, a linear alkyl chain, amino acid, peptide, PEG, saccharides, i.e. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is selected from the group consisting of acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, isobutyl, methoxyacetyl, glycine, aminoethylglucose (AEG), AEG-C6, PEG1, PEG2, N-dimethyl, and alkoxycarbonyl.

In another aspect, the insulin receptor partial agonist or insulin dimer D ¹ and D ² are independently native human insulin, insulin lispro, insulin aspart, desB30 insulin, or insulin glargine.

In certain aspects of an insulin receptor partial agonist or insulin dimer, each A chain polypeptide independently comprises the amino acid sequence GX ₂ X ₃ EQCCX ₈ SICSLYQLX ₁₇ NX ₁₉ CX ₂₃ (SEQ ID NO: 3) and each B chain polypeptide independently comprises the amino acid sequence X ₂₅ LCGX ₂₉ X ₃₀ LVEALYLVCGERGFX ₂₇ YTX ₃₁ X ₃₂ (SEQ ID NO: 4) or X ₂₂ VNQX ₂₅ X ₂₆ CGX ₂₉ X ₃₀ LVEALYLVCGERGFX ₂₇ YTX ₃₁ X ₃₂ X ₃₃ X ₃₄ X ₃₅ (SEQ ID NO: 5), moreover X ₂ is isoleucine or threonine; X ₃ is valine, glycine or leucine; X ₈ is threonine or histidine; X ₁₇ is glutamic acid or glutamine; X ₁₉ is tyrosine, 4-methoxyphenylalanine, alanine or 4-aminophenylalanine; X ₂₃ is asparagine or glycine; X ₂₂ is phenylalanine or deaminophenylalanine; X ₂₅ is histidine or threonine; X ₂₆ is leucine or glycine; X ₂₇ is phenylalanine or aspartic acid; X29 is alanine, glycine or serine; X30 is histidine, aspartic acid, glutamic acid, homocysteine acid or cysteic acid; X31 is aspartic acid, proline or lysine; X32 is lysine or proline; X33 is threonine, alanine or absent; X34 is arginine or absent; and X ₃₅ is arginine or absent; with the proviso that at least one of X ₃₁ or X ₃₂ is a lysine.

In certain aspects, insulin receptor partial agonists or insulin dimers

- 5 041034 linking moiety may be an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl and heterocyclic. The linking moiety may be a divalent, straight or branched, saturated or unsaturated, optionally substituted C1-C2o hydrocarbon chain in which one or more methylene units are optionally and independently substituted with -O-, -S-, -N(R)-, -C(O)-, C(O)O-, OC(O)-, -N(R)C(O)-, -C(O)N(R)-, -S(O)-, - S(O) ₂ -, N(R)SO ₂ -, SO2N(R)-, a heterocyclic group, an aryl group, or a heteroaryl group, wherein each occurrence of R is independently hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic fragment, aryl fragment, heteroaryl fragment or heteroaliphatic fragment.

In yet another aspect of an insulin receptor partial agonist or insulin dimer, the binding moiety is an acyl moiety, -C(O)RC(O)-, where R is an alkyl chain, a poly(ethylene glycol) (PEG) chain containing an amide chain containing triazole(s) chain, cyclooctyne-containing fragment, substituted acyl chain or polyethylene glycol (PEG) chain.

In another aspect of the insulin receptor partial agonist or insulin dimer, the binding moiety is a C ₂ -C _2o acyl moiety.

In certain aspects, the linking moiety is an alkyldioyl, -C(O)(CH ₂ ) _n C(O)-, where n=0-45, including, but not limited to, oxalyl (C2) moiety, succinyl (C4) moiety, an adipoyl (C6) moiety, a suberoyl (C8) moiety, a decanedioyl (C10) moiety, a dodecanedioyl (C12) moiety, a tetradecanedioyl (C14) moiety, or a hexadecanedioyl (C16) moiety.

The present invention further provides a dimer of insulin analogs comprising: a first heterodimer of insulin analogs and a second heterodimer of insulins or insulin analogs, each heterodimer comprising an A chain polypeptide and a B chain polypeptide, wherein the A chain polypeptide and the B chain polypeptide are linked together via interchain disulfide bonds; wherein the first and second heterodimers of insulins or insulin analogs are covalently linked together via a linking moiety connecting amino acid side chains at or near the carboxyl ends of the two respective B chain polypeptides; wherein the insulin analog is selected from insulin lispro, insulin aspart, and insulin glargine; and optionally, the amino terminus of at least one of the A chain polypeptides and B chain polypeptides of the first insulin polypeptide or the second insulin polypeptide is covalently linked to the substituent, provided that the linking moiety does not include a disulfide bond.

In another aspect of the insulin receptor partial agonist, the first and second heterodimers of the insulins or insulin analogs are the same, or the first and second heterodimers of the insulins or insulin analogs are different.

In another aspect of the insulin receptor partial agonist, the binding moiety covalently links the first heterodimer of insulins or insulin analogs and the second heterodimer of insulins or insulin analogs via an epsilon-amino group of a lysine residue at or near the carboxyl ends of their respective B chain polypeptides.

In yet another aspect of an insulin receptor partial agonist, the substituent has the general formula RC(O)-, where R may be R/CH2, R'NH, R'O, and R' may be H, a linear alkyl chain, an amino acid, a peptide , PEG, saccharides, i.e. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is selected from the group consisting of acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, isobutyl, methoxyacetyl, glycine, aminoethylglucose (AEG), AEG-C6, PEG1, PEG2, N-dimethyl, and alkoxycarbonyl.

In another aspect of the partial insulin receptor agonist, at least one of the first and second insulin or insulin analog is further conjugated to a polyethylene glycol, a sugar moiety, or a heterocycle.

In certain aspects of insulin receptor partial agonists, the binding moiety may be an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. The linking moiety may be a divalent, straight or branched, saturated or unsaturated, optionally substituted C1- _C20 hydrocarbon chain in which one or more methylene units are optionally and independently substituted with -O-, -S-, -N(R)- , -C(O)-, C(O)O-, OS(O)-, -N(R)C(O)-, -C(O)N(R)-, -S(O)-, -S(O)2-, -N(R)SO2-, SO2N(R)-, a heterocyclic group, an aryl group, or a heteroaryl group, wherein each occurrence of R is independently hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic moiety, an aryl moiety, a heteroaryl moiety, or a heteroaliphatic moiety.

In yet another aspect of the insulin receptor partial agonist, the binding moiety is an acyl moiety, -C(O)RC(O)-, where R is an alkyl chain, a poly(ethylene glycol) (PEG) chain containing an amide chain, a triazole(s) containing chain, a cyclooctyne moiety, a substituted acyl chain, or a polyethylene glycol (PEG) chain.

In another aspect of the insulin receptor partial agonist, the binding moiety is a C ₂ -C ₂₀ acyl moiety.

In certain aspects, the linking moiety is an alkyldioyl, -C(O)(CH2) _n C(O)-, where n=0-45, including, but not limited to, oxalyl (C2) moiety, succinyl (C4) moiety, adipoyl (C6) moiety, suberoyl (C8) moiety, decanedioyl (C10) moiety, dodecanedioyl (C12) moiety, tetradecanedioyl (C14) moiety, or hexadecanedioyl (C16) moiety.

The present invention further provides an insulin analog dimer having the formula D1-LD ² , wherein D ¹ and D ² are independently an insulin or insulin analog polypeptide, wherein each insulin polypeptide is a heterodimer having an A chain polypeptide and a B chain polypeptide linked together via interchain disulfide bonds; L is a linking moiety, with one end of the linker moiety attached to an amino acid residue at or near the carboxyl group of D ¹ and the other end of the linker moiety attached to an amino acid residue at or near the carboxyl end of D ² , provided that L is not includes a disulfide bond; wherein the insulin analog is selected from insulin lispro, insulin aspart, and insulin glargine; and optionally, at least one of D ¹ or D ² includes a substituent attached to the amino terminus of the Acepi polypeptide or the D ¹ or D ² B chain polypeptide; with the proviso that the linking moiety does not include a disulfide bond.

In another aspect of the insulin receptor partial agonist, D ¹ and D ² are the same, or D ¹ and D ² are different.

In another aspect of the insulin receptor partial agonist, the binding moiety covalently links D ¹ and D ² via the epsilon-amino group of a lysine residue at or near the carboxyl ends of D ¹ and D ² .

In yet another aspect of an insulin receptor partial agonist, the substituent has the general formula RC(O)-, where R may be R/CH2, R'NH, R'O, and R' may be H, a linear alkyl chain, an amino acid, a peptide , PEG, saccharides, i.e. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is selected from the group consisting of acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, isobutyl, methoxyacetyl, glycine, aminoethylglucose (AEG), AEG-C6, PEG1, PEG2, N-dimethyl, and alkoxycarbonyl.

In certain aspects of insulin receptor partial agonists, the binding moiety may be an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. The linking moiety may be a divalent, straight or branched, saturated or unsaturated, optionally substituted C1- _C20 hydrocarbon chain in which one or more methylene units are optionally and independently substituted with -O-, -S-, -N(R)- , -C(O)-, C(O)O-, OC(O)-, -N(R)C(O)-, -C(O)N(R)-, -S(O)-, -S(O) ₂ -, -N(R)SO2-, SO2N(R)-, a heterocyclic group, an aryl group or a heteroaryl group, wherein each occurrence of R is independently hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic moiety, an aryl moiety, a heteroaryl moiety, or a heteroaliphatic moiety.

In yet another aspect of the insulin receptor partial agonist, the binding moiety is an acyl moiety, -C(O)RC(O)-, where R is an alkyl chain, a poly(ethylene glycol) (PEG) chain, an amide containing chain, containing triazole(s ) chain, cyclooctyne moiety, substituted acyl chain, or polyethylene glycol (PEG) chain.

In another aspect of the insulin receptor partial agonist, the binding moiety is a C ₂ -C ₂₀ acyl moiety.

In certain aspects, the linking moiety is an alkyldioyl, -C(O)(CH ₂ ) _n C(O)-, where n=0-45, including, but not limited to, oxalyl (C2) moiety, succinyl (C4) moiety, an adipoyl (C6) moiety, a suberoyl (C8) moiety, a decanedioyl (C10) moiety, a dodecanedioyl (C12) moiety, a tetradecanedioyl (C14) moiety, or a hexadecanedioyl (C16) moiety.

The present invention provides an insulin receptor partial agonist comprising a first heterodimer of insulins or insulin analogs and a second heterodimer of insulins or insulin analogs, each heterodimer comprising an A chain polypeptide and a B chain polypeptide, wherein the A chain polypeptide and the B chain polypeptide are linked together via interchain disulfide bonds; wherein the first and second heterodimers of insulins or insulin analogs are covalently linked together via a linking moiety connecting amino acid side chains at or near the carboxyl ends of the two respective B chain polypeptides; optionally, the amino terminus of at least one of the A chain polypeptides and the B chain polypeptides

- 7 041034 of the first insulin polypeptide or the second insulin polypeptide is covalently linked to a substituent; and wherein the partial insulin receptor agonist has a maximum response to the human insulin receptor (IR) that is approximately 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, or 70% of the maximum response of native human insulin. in relation to IR determined using a functional analysis of phosphorylation; or a first heterodimer of insulins or insulin analogs and a second heterodimer of insulins or insulin analogs, each heterodimer comprising an A chain polypeptide and a B chain polypeptide, the A chain polypeptide and the B chain polypeptide being linked together via interchain disulfide bonds; wherein the first and second heterodimers of insulins or insulin analogs are covalently linked together via a linking moiety connecting amino acid side chains at or near the carboxyl ends of the two respective B chain polypeptides; optionally, the amino terminus of at least one of the A chain polypeptides and the B chain polypeptides of the first insulin polypeptide or the second insulin polypeptide is covalently linked to the substituent; and wherein the partial insulin receptor agonist has a maximum response to the human insulin receptor (IR) that is in the range of between 20 and 70, 40 and 70, 50 and 70, 40 and 60, or 20 and 40% of the maximum response of native human insulin in relation to IR determined using a functional analysis of phosphorylation; or a first heterodimer of insulins or insulin analogs and a second heterodimer of insulins or insulin analogs, each heterodimer comprising an A chain polypeptide and a B chain polypeptide, the A chain polypeptide and the B chain polypeptide being linked together via interchain disulfide bonds; wherein the first and second heterodimers of insulins or insulin analogs are covalently linked together via a linking moiety connecting amino acid side chains at or near the carboxyl ends of the two respective B chain polypeptides; optionally, the amino-terminus of at least one of the A chain polypeptides and the B chain polypeptides of the first insulin polypeptide or the second insulin polypeptide is covalently linked to the substituent; and wherein the partial insulin receptor agonist has a maximum response to the human insulin receptor (IR) that is at least 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, or 70% of the maximum response of the native human insulin in relation to IR determined using a functional analysis of phosphorylation; or a first heterodimer of insulins or insulin analogs and a second heterodimer of insulins or insulin analogs, each heterodimer comprising an A chain polypeptide and a B chain polypeptide, the A chain polypeptide and the B chain polypeptide being linked together via interchain disulfide bonds; wherein the first and second heterodimers of insulins or insulin analogs are covalently linked together via a linking moiety connecting amino acid side chains at or near the carboxyl ends of the two respective B chain polypeptides; optionally, the amino terminus of at least one of the A chain polypeptides and the B chain polypeptides of the first insulin polypeptide or the second insulin polypeptide is covalently linked to the substituent; and wherein the partial insulin receptor agonist has a maximum human insulin receptor (IR) response that is less than 70% of native human insulin's maximum IR response as determined by functional phosphorylation assay; or a first heterodimer of insulins or insulin analogs and a second heterodimer of insulins or insulin analogs, each heterodimer comprising an A chain polypeptide and a B chain polypeptide, the A chain polypeptide and the B chain polypeptide being linked together via interchain disulfide bonds; wherein the first and second heterodimers of insulins or insulin analogs are covalently linked together via a linking moiety connecting amino acid side chains at or near the carboxyl ends of the two respective B chain polypeptides; optionally, the amino terminus of at least one of the A chain polypeptides and the B chain polypeptides of the first insulin polypeptide or the second insulin polypeptide is covalently linked to the substituent; and wherein the partial insulin receptor agonist has a maximum response to the human insulin receptor (IR) that is approximately 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, or 70% of the maximum response of native human insulin. in relation to IR determined using a functional analysis of phosphorylation.

In the above embodiments, the functional phosphorylation assay may be an insulin receptor (IR) AKT phosphorylation assay.

In another aspect of an insulin receptor partial agonist, the binding moiety does not include a disulfide bond, and when the amino termini of the A chain polypeptide and the B chain polypeptide do not include a substituent, the binding moiety is not an oxalyl (C2) moiety, suberoyl ( C8) fragment or dodecanedioyl (C12) fragment.

In another aspect of the insulin receptor partial agonist, the first and second heterodimers of the insulins or insulin analogs are the same, or the first and second heterodimers of the insulins or insulin analogs are different.

In another aspect of an insulin receptor partial agonist, the binding moiety is covalently

- 8 041034 links the first heterodimer of insulins or insulin analogs and the second heterodimer of insulins or insulin analogs via the epsilon-amino group of a lysine residue at or near the carboxyl ends of their respective B chain polypeptides.

In yet another aspect of an insulin receptor partial agonist, the substituent has the general formula RC(O)-, where R may be R/CH2, R'NH, R'O, and R' may be H, a linear alkyl chain, an amino acid, a peptide , PEG, saccharides, i.e. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is selected from the group consisting of acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, isobutyl, methoxyacetyl, glycine, aminoethylglucose (AEG), AEG-C6, PEG1, PEG2, N-dimethyl, and alkoxycarbonyl.

In another aspect of the partial insulin receptor agonist, the first and second insulins or insulin analogs are independently native human insulin, insulin lispro, insulin aspart, desB30 insulin, or insulin glargine.

In certain aspects of an insulin receptor partial agonist, each A chain polypeptide independently comprises the amino acid sequence GX ₂ X ₃ EQCCX ₈ SICSLYQLX ₁₇ NX ₁₉ CX ₂₃ (SEQ ID NO: 3), and each B chain polypeptide independently comprises the amino acid sequence X ₂₅ LCGX ₂₉ X ₃₀ LVEALYLVCGERGFX ₂₇ YTX ₃₁ X ₃₂ (SEQ ID NO: 4) or

X ₂₂ VNQX ₂₅ X ₂₆ CGX ₂₉ X ₃₀ LVEALYLVCGERGFX ₂₇ YTX ₃₁ X ₃₂ X ₃₃ X ₃₄ X ₃₅ (SEQ ID NO: 5), wherein X ₂ is isoleucine or threonine; X ₃ is valine, glycine or leucine; X ₈ is threonine or histidine; X ₁₇ is glutamic acid or glutamine; X ₁₉ is tyrosine, 4-methoxyphenylalanine, alanine or 4-aminophenylalanine; X23 is asparagine or glycine; X22 is phenylalanine or deaminophenylalanine; X25 is histidine or threonine; X ₂₆ is leucine or glycine; X ₂₇ is phenylalanine or aspartic acid; X29 is alanine, glycine or series; X30 is histidine, aspartic acid, glutamic acid, homocysteine acid or cysteic acid; X ₃₁ is aspartic acid, proline or lysine; X ₃₂ is lysine or proline; X ₃₃ is threonine, alanine or absent; X ₃₄ is arginine or absent; and X ₃₅ is arginine or absent; with the proviso that at least one of X ₃₁ or X ₃₂ is a lysine.

In certain aspects of insulin receptor partial agonists, the binding moiety may be an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. The linking moiety may be divalent, straight or branched, saturated or unsaturated, optionally substituted with a _C1-20 hydrocarbon chain, wherein one or more methylene units are optionally and independently substituted with -O-, -S-, -N(R)-, -C( O)-, C(O)O-, OS(O)-, -N(R)C(O)-, -C(O)N(R)-, -S(O)-, -S(O )2-, -N(R)SO2-, SO2N(R)-, a heterocyclic group, an aryl group, or a heteroaryl group, wherein each occurrence of R is independently hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic moiety, an aryl fragment, heteroaryl fragment or heteroaliphatic fragment.

In yet another aspect of the insulin receptor partial agonist, the binding moiety is an acyl moiety, -C(O)RC(O)-, where R is an alkyl chain, a poly(ethylene glycol) (PEG) chain, an amide containing chain, containing triazole(s ) chain, cyclooctyne moiety, substituted acyl chain, or polyethylene glycol (PEG) chain.

In another aspect of the insulin receptor partial agonist, the binding moiety is a C2-C20 acyl moiety.

In certain aspects, the linking moiety is an alkyldioyl, -C(O)(CH ₂ ) _n C(O)-, where n=0-45, including, but not limited to, oxalyl (C2) moiety, succinyl (C4) moiety, an adipoyl (C6) moiety, a suberoyl (C8) moiety, a decanedioyl (C10) moiety, a dodecanedioyl (C12) moiety, a tetradecanedioyl (C14) moiety, or a hexadecanedioyl (C16) moiety.

The present invention further provides an insulin dimer comprising a first B29 or B28 Lys of a first insulin heterodimer molecule comprising a first A chain polypeptide and a first B chain polypeptide and a second B29 or B28 Lys of a second insulin heterodimer comprising a second A chain polypeptide and a second polypeptide B chains conjugated together via a bifunctional linker selected from the group consisting of linker 1, linker 2, linker 3, linker 10, linker 11, linker 12, linker 13, linker 14, linker 15, linker 16, linker 17, linker 18, linker 19, linker 20, linker 21, linker 22, linker 23, linker 24, linker 25, linker 26, linker 27, linker 28, linker 29, linker 30, linker 31, linker 32, linker 33, linker 34, linker 35, linker 36, linker 37, linker 38, linker 39, linker 40, linker 41, linker 42, linker 43, linker 44, linker 45, linker 46, linker 47, linker 48, linker 49 and linker 50, provided that when the bifunctional lin the core is linker 10, linker 11, linker 12, linker 13, or linker 14, at least one of the first or second A-chain or B-chain polypeptides is conjugated with its N-terminal amino acid to a substituent, or at least N the α-terminal amino acids of the first insulin heterodimer molecule are conjugated to a substituent, or the N-terminal amino acids of both the first insulin heterodimer and the second insulin heterodimer are conjugated to a substituent.

In certain embodiments, the substituent contains an N-hydroxysuccinimide ester linked to a group having the general formula RC(O)-, where R may be R/CH2, R'NH, R'O, and R' may be H, linear alkyl chain, amino acid, peptide, polyethylene glycol (PEG), saccharides. In certain embodiments, the substituent is a carbamoyl group, an acetyl group, a glycine, a methyl group, a methoxy group, a dimethyl group, an isobutyl group, a PEGI group, an AEG group, an AEG-C6 alkyl group, or a PEG2 group.

The present invention further provides an insulin dimer comprising the first B29 or B28 Lys of a first insulin heterodimer molecule comprising a first A chain polypeptide and a first B chain polypeptide conjugated to a first linker selected from the group consisting of linker 5 and linker 7, and a second B29 or B28 Lys of a second insulin heterodimer comprising a second A chain polypeptide and a second B chain polypeptide conjugated to a second linker selected from the group consisting of linker 4, linker 6, linker 8 and linker 9 conjugated together via the first linker and second linker.

In another embodiment, the present invention provides an insulin analog dimer comprising a first B29 or B28 Lys molecule of a first insulin heterodimer comprising a first A chain polypeptide and a first B chain polypeptide conjugated to a first linker selected from the group consisting of linker 5 and linker 7 and a second B29 or B28 Lys of a second insulin heterodimer comprising a second A chain polypeptide and a second B chain polypeptide conjugated to a second linker selected from the group consisting of linker 5 and linker 7, the first and second linkers being conjugated together via bridging linker having the structure

----R---where R is a covalent bond, a carbon atom, a phenyl, a heteroatom, or an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. In certain aspects, R is a C2, C3, C4, C6, C7, C8, C9, or C10 acyl group, or PEG2, PEG3, PEG4, PEG5, PEG6, PEG7, PEG8, PEG9, PEG10, PEG11, PEG12, PEG13, or PEG25.

In certain embodiments, the substituent contains an N-hydroxysuccinimide ester linked to a group having the general formula RC(O)-, where R may be R'CH2, R'NH, R'O, and R' may be H, linear alkyl chain, amino acid, peptide, polyethylene glycol (PEG), saccharides. In certain embodiments, the substituent is a carbamoyl group, an acetyl group, a glycine, a methyl group, a methoxy group, a dimethyl group, an isobutyl group, a PEGI group, an AEG group, an AEG-C6 alkyl group, or a PEG2 group.

In another embodiment, the present invention provides an insulin analog dimer comprising a first B29 or B28 Lys molecule of a first insulin heterodimer comprising a first A chain polypeptide and a first B chain polypeptide conjugated to a first linker selected from the group consisting of linker 4, linker 6, linker 8 and linker 9, and a second B29 or B28 Lys of a second insulin heterodimer comprising a second A chain polypeptide and a second B chain polypeptide conjugated to a second linker selected from the group consisting of linker 4, linker 6, linker 8 and linker 9, wherein the first and second linkers are conjugated together via a bridged linker having the structure

N ₃ ---R---N ₃ where R is a covalent bond, a carbon atom, a phenyl, a heteroatom, or an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. In certain aspects, R is a C2, C3, C4, C6, C7, C8, C9, or C10 acyl group, or PEG2, PEG3, PEG4, PEG5, PEG6, PEG7, PEG8, PEG9, PEG10, PEG11, PEG12, PEG13, or PEG25.

In certain embodiments, the substituent contains an N-hydroxysuccinimide ester linked to a group having the general formula RC(O)-, where R may be R'CH2, R'NH, R'O, and R' may be H, linear alkyl chain, amino acid, peptide, polyethylene glycol (PEG), saccharides. In certain embodiments, the substituent is a carbamoyl group, an acetyl group, a glycine, a methyl group, a methoxy group, a dimethyl group, an isobutyl group, a PEGI group, an AEG group, an AEG-C6 alkyl group, or a PEG2 group.

The present invention further provides compositions comprising any of the insulin receptor partial agonists disclosed herein and a pharmaceutically acceptable salt.

The present invention provides a method of treating diabetes comprising administering to a diabetic subject a therapeutically effective amount of a composition comprising any of the aforementioned partial insulin receptor agonists. In some aspects, diabetes is

- 10 041034 type 1 diabetes, type 2 diabetes or gestational diabetes.

The present invention provides the use of a composition for the treatment of diabetes containing any of the aforementioned partial insulin receptor agonists. In certain aspects, the diabetes is type 1 diabetes, type 2 diabetes, or gestational diabetes.

The present invention provides the use of any of the insulin receptor partial agonists disclosed herein for the manufacture of a medicament for the treatment of diabetes. In certain aspects, the diabetes is type 1 diabetes, type 2 diabetes, or gestational diabetes.

Definitions.

Insulin - as used herein, this term refers to the active principle of the pancreas, which affects the metabolism of carbohydrates in the body of the animal, and which is significant in the treatment of diabetes. The term includes synthetic and biotechnologically produced products that are the same or similar to natural insulins in structure, use and intended effect and are of value in the treatment of diabetes mellitus. This term is a general term that refers to a 51 amino acid heterodimer comprising an A chain peptide having the amino acid sequence shown in SEQ ID NO: 1 and a B chain peptide having the amino acid sequence shown in SEQ ID NO: 2, moreover, the cysteine residues at positions 6 and 11 of the A chain are linked by a disulfide bond, the cysteine residues at position 7 of the A chain and position 7 of the B chain are linked by a disulfide bond, and the cysteine residues at position 20 of the A chain and 19 of the B chain are linked by a disulfide bond .

Insulin analog - this term, as used herein, includes any heterodimeric analog or single chain analog that contains one or more modifications of the native A chain peptide and/or B chain peptide. Modifications include, but are not limited to, substitution of any amino acid for a native amino acid at a position selected from A4, A5, A8, A9, A10, A12, A13, A14, A15, A16, A17, A18, A19, A21, B1 , B2, B3, B4, B5, B9, B10, B13, B14, B15, B16, B17, B18, B20, B21, B22, B23, B26, B27, B28, B29 and B30; removal of any or all of the provisions of B1-4 and B2 6-30; or conjugation, directly or via a polymeric or non-polymeric linker, of one or more acyl, polyethylglycine (PEG) or saccharide moieties; or any combination of them. As illustrated by the N-linked glycosylated insulin analogs disclosed herein, the term also includes any insulin heterodimer and single chain analog that has been modified such that it has at least one N-linked glycosylation site and, in particular, embodiments wherein the N-linked glycosylation site is linked to or occupied by an N-glycan. Examples of insulin analogs include, but are not limited to, the heterodimeric and single chain analogs disclosed in WO 20100080606, WO 2009/099763 and WO 2010080609, the disclosures of which are incorporated herein by reference. Examples of single -chain insulin analogues also include, but without restrictions disclosed in published international applications Wo 9634882, Wo 95516708, Wo 2005054291, Wo 2006097521, Wo 2007104734, Wo 2007104736, Wo 2007104738, Wo 200738, Wo 200738, Wo 200738, Wo 200738, Wo 200738, Wo 200738, Wo 200738, Wo 200738, Wo 200738, Wo 200738, Wo 200738, Wo 200738, WO US Pat. Nos. 5,304,473 and 6,630,348; and Kristensen et al., Biochem. J. 305: 981-986 (1995), the disclosures of which are incorporated herein by reference.

The term also includes single chain and heterodimeric polypeptide molecules that have little or no detectable activity at the insulin receptor but have been modified to include one or more amino acid modifications or substitutions so that they have insulin receptor activity that has at least 1%, 10%, 50%, 75%, or 90% of the insulin receptor activity of native insulin, and which also includes at least one N-linked glycosylation site. In certain aspects, the insulin analog is a partial agonist that has less than 80% (or 70%) of the insulin receptor activity of native insulin. These insulin analogs that have reduced growth hormone insulin receptor activity and increased insulin receptor activity include both heterodimers and single chain analogs.

Single chain insulin or single chain insulin analog as used herein, the term encompasses a group of structurally related proteins in which an Acepi peptide or functional analog and a B chain peptide or functional analog are covalently linked via a 2-35 amino acid peptide or polypeptide or non-peptide polymeric or non-polymeric linker, and which have at least 1, 10, 50, 75 or 90% of the activity of insulin against the insulin receptor compared to native insulin. Single-chain insulin or insulin analog also includes three disulfide bonds: the first disulfide bond is located between the cysteine residues at positions 6 and 11 of the A chain or its functional analog, the second disulfide bond is located between the cysteine residues at position 7 of the A chain or its functional analog and to position 7 of the B-chain or its functional analogue, and the third disulfide

- 11 041034 the bond is located between the cysteine residues at position 20 of the A chain or its functional analogue and at position 19 of the B chain or its functional analogue.

Linking peptide or C-peptide - as used herein, this term refers to the linking fragment C of the B-C-A polypeptide sequence of a single chain preproinsulin-like molecule. In particular, in the natural insulin chain, the C-peptide links the amino acid at position 30 of the B chain and the amino acid at position 1 of the A chain. The term may refer to native insulin C-peptide, monkey C-peptide, and any other 3 to 35 amino acid peptide that links the B chain to the A chain, and is thus intended to include any peptide that binds a peptide B-chains with an A-chain peptide in a single-chain insulin analog (see, for example, U.S. Publication Nos. 20090170750 and 20080057004 and WO 9634882) and in insulin precursor molecules such as those disclosed in WO 9516708 and U.S. Patent No. 7,105,314.

Amino acid modification - as used herein, the term refers to amino acid substitution or amino acid derivatization by adding to and/or removing chemical groups from an amino acid, and includes substitution with any of the 20 amino acids commonly found in human proteins, as well as atypical or unnatural amino acids. Commercial sources of atypical amino acids include Sigma-Aldrich (Milwaukee, WI), ChemPep Inc. (Miami, FL) and Genzyme Pharmaceuticals (Cambridge, MA). Atypical amino acids can be purchased from commercial suppliers, synthesized de novo, or chemically modified or derived from naturally occurring amino acids.

Amino acid substitution - as used herein, refers to the substitution of one amino acid residue with another amino acid residue.

Conservative amino acid substitution - as used herein, the term is defined herein as an exchange within one of the following five groups:

I. Small aliphatic non-polar or weakly polar residues: Ala, Ser, Thr, Pro, Gly.

II. Polar negatively charged residues and their amides: Asp, Asn, Glu, Gln, cysteic acid and homocysteine acid.

III. Polar positively charged residues: His, Arg, Lys; ornithine (Orn).

IV. Large aliphatic non-polar residues: Met, Leu, Ile, Val, Cys, norleucine (Nle), homocysteine.

V. Large aromatic residues: Phe, Tyr, Trp, acetylphenylalanine.

Treatment - as used herein, the term treat (or treat, undergo treatment, treat, etc.) refers to the administration of an IRPA of the present disclosure to a subject in need of it for the purpose of alleviating, alleviating, changing, ameliorating, improving, or effecting a condition (eg, diabetes), a symptom or symptoms of the condition (eg, hyperglycemia), or a predisposition to the condition. For example, as used herein, the term treatment of diabetes will generally refer to maintaining blood glucose levels near normal levels and may include raising or lowering blood glucose levels as appropriate.

Pharmaceutically acceptable carrier - as used herein, the term includes any of the standard pharmaceutical carriers such as phosphate buffered saline, water, emulsions such as oil/water or water/oil emulsions, and various types of wetting agents as appropriate. for administration to a needy individual or a needy individual. The term also includes any product approved by a U.S. federal government regulatory agency or listed in the USP for use in animals, including humans.

Pharmaceutically acceptable salt - as used herein, the term refers to salts of compounds that retain the biological activity of the parent compound, and which are not biologically or otherwise undesirable. Many of the compounds disclosed herein are capable of forming acidic and/or basic salts due to the presence of amino and/or carboxyl groups or the like.

Pharmaceutically acceptable base addition salts can be prepared from inorganic and organic bases. Salts derived from inorganic bases include, by way of example only, sodium, potassium, lithium, ammonium, calcium, zinc and magnesium salts. Salts derived from organic bases include, but are not limited to, salts of primary, secondary, and tertiary amines.

Pharmaceutically acceptable acid addition salts can be prepared from inorganic and organic acids. Salts derived from inorganic acids include salts of hydrochloric acid, hydrobromic acid, sulfuric acid, nitric acid, phosphoric acid, and the like. Salts derived from organic acids include salts of acetic acid, propionic acid, glycolic acid, pyruvic acid, oxalic acid, malic acid, malonic acid, succinic acid, maleic acid, fumaric acid, tartaric acid, citric acid, benzoic acid, cinnamic acid , mandelic acid, methanesulfonic acid,

- 12 041034 ethanesulfonic acid, p-toluenesulfonic acid, salicylic acid and the like.

An effective or therapeutically effective amount, as used herein, refers to a non-toxic but sufficient amount of an insulin analog to provide the desired effect. For example, one of the desired effects would be the prevention or treatment of hyperglycemia. The amount that is effective will vary from subject to subject, depending on the age and general condition of the individual, route of administration, and the like. Thus, it is not always possible to specify an exact effective amount. It is not always possible to determine the optimal effective amount prior to administration to an individual or an individual in need thereof. However, a suitable effective amount in any particular case can be determined by one of ordinary skill in the art using routine experimentation.

Parenteral - as used herein, this term does not mean through the alimentary canal, but by some other route, such as intranasal, inhalation, subcutaneous, intramuscular, intraspinal or intravenous.

Brief description of the drawings

Fig. 1 shows the glucose lowering effect of dimers 24, 18, and 40 compared to RHI when administered to diabetic minipigs at 0.69 nmol/kg.

Fig. 2A shows the results of an Insulin Tolerance Test (ITT) in mice comparing Compound A with RHI (humulin). Compound A was administered at a dose of 72 U/kg and at a dose of 300 U/kg, and humulin was administered at a dose of 18 U/kg and at a dose of 72 U/kg.

Fig. 2B shows the results of the Insulin Tolerance Test (ITT) in mice comparing Compound B with RHI (humulin). Compound B was administered at a dose of 72 U/kg and at a dose of 300 U/kg, and humulin was administered at a dose of 18 U/kg and at a dose of 72 U/kg.

Fig. 2C shows the results of the Insulin Tolerance Test (ITT) in mice comparing dimer 24 with RHI (humulin). Dimer 24 was administered at a dose of 72 U/kg and at a dose of 300 U/kg, and humulin was administered at a dose of 18 U/kg and at a dose of 72 U/kg.

Fig. 2D shows the results of dimer 55 in the insulin tolerance test (ITT) in mice. Dimer 55 was administered at a dose of 120 nmol/kg and at a dose of 300 nmol/kg.

Fig. 2E shows the results of dimer 58 in the Insulin Tolerance Test (ITT) in mice. Dimer 58 was administered at a dose of 60 nmol/kg and at a dose of 300 nmol/kg.

Fig. 2F shows the results of dimer 60 in the insulin tolerance test (ITT) in mice. Dimer 60 was administered at a dose of 72 nmol/kg and at a dose of 300 nmol/kg.

Fig. 2G shows the results of dimer 67 in the insulin tolerance test (ITT) in mice. Dimer 67 was administered at a dose of 120 nmol/kg and at a dose of 300 nmol/kg.

Fig. 3 shows that Compound A insulin dimer was degraded to insulin monomers upon 2 hour incubation with rat kidney cell membranes (RKCM) without glutathione (GSH). % of parent compound values are only semi-quantitative due to potential differences in ionization efficiency.

Fig. 4 shows that Compound A insulin dimer was degraded to insulin monomers upon 2 hour incubation with rat kidney cell membranes (RKCM) with glutathione (GSH). % of parent compound values are only semi-quantitative due to potential differences in ionization efficiency.

Fig. 5 shows that dimer 24 lost some A chain polypeptide but was not degraded to monomers upon 2 hour incubation with rat kidney cell membranes (RKCM) without glutathione (GSH). % of parent compound values are only semi-quantitative due to potential differences in ionization efficiency.

Fig. 6 shows that dimer 24 lost some A chain polypeptide but was not degraded to monomers upon 2 hour incubation with rat kidney cell membranes (RKCM) with glutathione (GSH). % of parent compound values are only semi-quantitative due to potential differences in ionization efficiency.

Fig. 7A shows the glucose lowering effect of dimers 4, 5, 7, 8 and 9 compared to RHI when administered to diabetic minipigs at 0.69 nmol/kg.

Fig. 7B shows the glucose lowering effect of dimers 18, 19, 20, 21 and 22 compared to RHI when administered to diabetic minipigs at 0.69 nmol/kg.

Fig. 7C shows the glucose lowering effect of dimers 23, 24, 26, 27 and 28 compared to RHI when administered to diabetic minipigs at 0.69 nmol/kg.

Fig. 7D shows the glucose lowering effect of dimers 29, 32, 37, 38 and 39 compared to RHI when administered to diabetic minipigs at 0.69 nmol/kg.

Fig. 7E shows the glucose lowering effect of dimers 40, 41, 43, 44 and 48 compared to RHI when administered to diabetic minipigs at 0.69 nmol/kg.

Fig. 7F shows the glucose lowering effect of dimers 55, 57, 58, 60 and 61 compared to RHI when administered to diabetic minipigs at 0.69 nmol/kg.

Fig. 7G shows the glucose lowering effect of dimers 62, 64, 67, 69 and 71 compared to

- 13 041034

RHI when administered to diabetic minipigs at 0.69 nmol/kg.

Fig. 7H shows the glucose lowering effect of dimers 72, 77 and 78 compared to RHI when administered to diabetic minipigs at 0.69 nmol/kg.

Detailed description of the invention

The present invention provides compounds containing two insulin molecules covalently linked to form a covalently linked insulin dimer that can activate an insulin receptor with regular insulin-like activity and reduced peak activity. These compounds are insulin receptor partial agonists (IRPA): they behave like other insulin analogs, effectively lowering glucose but with a lower risk of hypoglycemia.

Insulin dimers have been disclosed in Brandenburg et al. in US patent No. 3907763 (1973); Tatnell et al., Biochem J. 216: 687-694 (1983); Shuttler and Brandenburg, Hoppe-Seyler's Z. Physiol. Chem, 363, 317-330, 1982; Weiland et al., Proc Natl. Acad. sci. (USA) 87: 1154-1158 (1990); Deppe et al., Naunyn-Schmiedeberg's Arch Pharmacol (1994) 350: 213-217; Brandenburg and Havenith in US Pat. No. 6,908,897(B2) (2005); Knudsen et al., PLOS ONE 7: e51972 (2012); DiMarchi et al in WO 2011/159895; DiMarchi et al. in WO 2014/052451; and Herrera et al., WO 2014141165. More recently, insulin dimers have been described in Brant, Synthesis and Characterization of Insulin Receptor Partial Agonists as a Route to Improved Diabetes Therapy, Ph.D. Dissertation, Indiana University (April 2015) and Zaykov and DiMarchi, Poster P212, Exploration of the structural and mechanistic basis for partial agonism of insulin dimers, American Peptide Symposium, Orlando FL (June 20-25, 2015). However, the present inventors have found that the level of insulin activity and partial agonist activity of the dimers is a function of the dimeric structure, the sequence of the insulin analog, the length of the dimerizing linker, and the dimerization site that connects the two insulin polypeptides. The authors found that the insulin dimers of the present invention have a reduced risk of hypoglycemia when administered at high doses than native insulin or other insulin analogs when administered at high doses.

The present invention provides covalently bound insulin dimers-partial agonists formulated as new and capable of converting basal insulins (once daily administration) that exhibit an improved therapeutic index (TI) over current standard of care (SOC) basal insulins. ). These molecules can effectively lower glucose levels with a reduced risk of hypoglycemia in diabetic minipigs and have the properties of once-daily (QD) basal insulin. An improved TI may allow practitioners to more aggressively dose IRPA dimer to achieve fasting glucose control targets. Strict control of fasting glucose and HbA1c may allow these molecules to serve as 1) a long-acting single-acting insulin with an improved efficacy and safety profile in type 2 diabetes mellitus (T2DM) and 2) an improved basal basal insulin in type 1 diabetes mellitus (T1DM) (and sometimes T2DM) for use with additional doses of prandial rapid-acting insulin analogs (RAA).

The ideal long-acting insulin provides continuous fasting glucose control in diabetic patients with high PK/PD stability and reproducibility. However, currently available basal insulins, even with improved PK/PD stability and reproducibility, still have a narrow therapeutic index, and the incidence of hypoglycemia increases as glucose levels approach euglycemia target. This can often result in insufficient dosage to prevent hypoglycemia. Treatment with the IRPA of the present invention is expected to reverse this efficacy:hypoglycemia relationship by reducing the rate of change in glucose decline with increasing dosage.

A- and B-chains of insulin.

Disclosed herein are dimers of insulins or insulin analogs that have insulin receptor agonist activity. The level of insulin activity of the dimers is a function of the dimeric structure, the insulin analog sequence, the length of the dimerizing linker, and the dimerization site that connects the two insulin polypeptides. The insulin polypeptides of the present invention may contain sequences of native human insulin B and A chains (SEQ ID NOs: 1 and 2, respectively) or any of their known analogues or derivatives that exhibit insulin agonist activity when bound together in a heteroduplex. Such analogs include, for example, proteins that have an A chain and a B chain that differ from the A chain and B chain of human insulin in that they have one or more amino acid deletions, one or more amino acid substitutions, and/or one or more amino acid inserts that do not destroy the insulin activity of the insulin analogue.

One type of insulin analog, monomeric insulin analog, is well known in the art. These are fast-acting human insulin analogs, including, for example, insulin analogs in which:

- 14 041034 (a) the aminoacyl residue at position B28 is replaced by Asp, Lys, Leu, Val or Ala, and the aminoacyl residue at position B29 is Lys or Pro;

(b) aminoacyl residues at any of positions B27 and B30 are removed or substituted with a non-native amino acid.

In one embodiment, an insulin analog is provided containing Asp substituted at position B28 (e.g. insulin aspart (NOVOLOG); see SEQ ID NO: 9) or Lys substituted at position 28 and proline substituted at position B29 (e.g. , insulin lispro (HUMALOG) see SEQ ID NO: 6). Additional monomeric insulin analogs are disclosed in Chance, et al., US Pat. No. 5,514,646; Chance, et al., US Patent Application Serial No. 08/255297; Brems, et al., Protein Engineering, 5: 527-533 (1992); Brange, et al., EPO Publication No. 214826 (published March 18, 1987); and Brange, et al., Current Opinion in Structural Biology, 1: 934-940 (1991). These disclosures are expressly incorporated herein by reference to describe monomeric insulin analogues.

Insulin analogues may also contain amidated amino acid substitutions for acidic forms. For example, Asn can be replaced by Asp or Glu. Similarly, Gln can be replaced by Asp or Glu. In particular, Asn(A18), Asn(A21) or Asp(B3) or any combination of these residues may be replaced by Asp or Glu. In addition, Gln(A15) or Gln(B4) or both can be replaced with either Asp or Glu.

As disclosed herein, single chain insulin analogs are provided comprising the B chain and A chain of human insulin, or an analog or derivative thereof, wherein the carboxyl end of the B chain is linked to the amino end of the A chain via a linking moiety. In one embodiment, the A chain is the amino acid sequence GIVEQCCTSICSLYQLENYCN (SEQ ID NO: 1) and the B chain contains the amino acid sequence FVNQHLCGSH LVEALYLVCGERGFFYTPKT (SEQ ID NO: 2) or a carboxy-terminal truncated sequence thereof that has B30 removed, and analogs of these sequences, each sequence modified so that it contains from one to five amino acid substitutions in positions corresponding to the positions of native insulin selected from A5, A8, A9, A10, A14, A15, A17, A18, A21, B1, B2, B3, B4, B5, B9, B10, B13, B14, B20, B22, B23, B26, B27, B28, B29 and B30, provided that at least one of B28 or B29 is lysine. In one embodiment, the amino acid substitutions are conservative amino acid substitutions. Suitable amino acid substitutions at these positions that do not adversely affect the desired activity of insulin are known to those skilled in the art, as demonstrated, for example, in Mayer, et al., Insulin Structure and Function, Biopolymers. 2007; 88(5): 687-713, the disclosure of which is incorporated herein by reference.

In one embodiment, the insulin analog peptides may comprise an insulin A chain and an insulin B chain, or analogs thereof, wherein the A chain contains an amino acid sequence that has at least 70% sequence identity (e.g., 70, 75, 80, 85, 90.95%) over the length of the native peptide with GIVEQCCTSICSLYQLENYCN (SEQ ID NO: 1) and the B chain contains an amino acid sequence that has at least 60% sequence identity (e.g., 60, 65, 70, 75, 80, 85 , 90, 95%) by the length of the native peptide with FVNQHLCGSHLVEALYLVCGERGFFYTPKT (SEQ ID NO: 2) or its carboxyl-truncated sequence in which B30 has been removed.

Additional amino acid sequences may be added to the amino terminus of the B chain or to the carboxyl terminus of the A chain of the insulin polypeptides of the present invention.

For example, a number of negatively charged amino acids may be added to the amino terminus of the chain, including, for example, a peptide 1 to 12, 1 to 10, 1 to 8, or 1 to 6 amino acids in length containing one or more negatively charged amino acids, including, for example, glutamic acid and aspartic acid. In one embodiment, the amino terminal chain extension contains 1 to 6 charged amino acids. According to one embodiment, the disclosed insulin polypeptides contain a C-terminal amide or ester on the A-chain instead of a C-terminal carboxylate.

In various embodiments, the insulin analog has an isoelectric point that is offset from human insulin. In some embodiments, shifting the isoelectric point is achieved by attaching one or more arginine, lysine, or histidine residues to the N-terminus of the insulin A chain peptide and/or the C-terminus of the insulin B chain peptide. Examples of such insulin polypeptides include Arg ^A0 human insulin, Arg ^B31 Arg ^B32 human insulin, Gly ^A21 Arg ^B31 Arg ^B32 human insulin, Arg ^A0 Arg ^B31 Arg ^B32 human insulin, and Arg ^A0 Gly ^A21 Arg ^B31 Arg ^B32 human insulin. As a further example, insulin glargine (LANTUS; see SEQ ID NOs: 7 and 8) is an exemplary long-acting insulin analogue in which Asn ^A21 has been substituted with glycine and two arginine residues have been covalently linked to the C-terminus of B- peptide. The effect of these amino acid changes is to shift the isoelectric point of the molecule, resulting in a molecule that is soluble at acidic pH (eg pH 4-6.5) but insoluble at physiological pH. When insulin glargine solution is injected

- 15 041034 into muscle, the pH of the solution is neutralized and insulin glargine forms microprecipitates that slowly release insulin glargine over a 24-hour period after injection without a pronounced insulin peak and therefore with a reduced risk of inducing hypoglycemia. This profile makes it possible to provide the patient with basal insulin by dosing once a day. Thus, in some embodiments, the insulin analog comprises an A chain peptide wherein the amino acid at position A21 is glycine and a B chain peptide wherein the amino acids at positions B31 and B32 are arginine. This disclosure covers all single and multiple combinations of these mutations and any other mutations that are described herein (e.g., Gl y ^A21 - human insulin, Gl y ^A21 Arg ^B31 - human insulin, Arg ^B31 Arg ^B32 - human insulin, Arg ^B31 - human insulin).

In certain aspects of insulin receptor partial agonists, one or more amidated amino acids of the insulin analog are replaced with an acidic amino acid or another amino acid. For example, asparagine can be replaced by aspartic acid or glutamic acid or other residue. Similarly, glutamine can be replaced by aspartic acid or glutamic acid or other residue. In particular, AsnA ¹⁸ , Asn ^A21 or Asn ^B3 or any combination of these residues can be replaced by aspartic acid or glutamic acid or another residue. Gln ^A15 or Gln ^B4 or both can be replaced by aspartic acid or glutamic acid or another residue. In certain aspects of insulin receptor partial agonists, the insulin analogs contain aspartic acid or another residue at the A21 position, or aspartic acid or another residue at the B3 position, or both.

One of skill in the art will appreciate that other amino acids in the insulin analog can be substituted for other amino acids while maintaining the biological activity of the molecule. For example, without limitation, the following modifications are also widely accepted in the art: replacing the histidine residue at position B10 with aspartic acid (His ^B10 with Asp ^B10 ); replacement of the phenylalanine residue at position B1 with aspartic acid (PheB1 with AspB1); replacement of the threonine residue at position B30 with alanine (ThrB30 with AlaB30); replacement of the tyrosine residue at position B26 with alanine (TyrB26 for AlaB26); and replacing the serine residue at position B9 with aspartic acid (SerB9 to AspB9).

In various embodiments, the implementation of the insulin analog has a long profile of action. Thus, in some embodiments, the insulin analog may be acylated with a fatty acid. That is, an amide bond is formed between the amino group on the insulin analogue and the carboxylic acid group of the fatty acid. The amino group may be the alpha-amino group of the N-terminal amino acid of the insulin analog, or may be the epsilon-amino group of the lysine residue of the insulin analog. The insulin analog may be acylated at one or more of the three amino groups that are present in wild-type human insulin, may be acylated at a lysine residue that has been introduced into the wild-type human insulin sequence. In certain aspects of insulin receptor partial agonists, the insulin analog may be acylated at position A1, B1, or both A1 and B1. In some embodiments, the fatty acid is selected from myristic acid (C ₁₄ ), pentadecylic acid (C ₁₅ ), palmitic acid (C ₁₆ ), heptadecylic acid (C ₁₇ ), and stearic acid (C ₁₈ ).

Examples of insulin analogs can be found, for example, in published international applications WO 9634882, WO 95516708; WO 20100080606, WO 2009/099763 and WO 2010080609, US Pat. No. 6,630,348 and Kristensen et al., Biochem. J. 305: 981-986 (1995), the disclosures of which are incorporated herein by reference. In other embodiments, in vitro glycosylated or in vivo N-glycosylated insulin analogs may be acylated and/or pegylated.

According to one embodiment, an insulin analog is provided, wherein the insulin peptide A chain contains the sequence GIVEQCCX ₈ SICSLYQLX ₁₇ NX ₁₉ CX ₂₃ (SEQ ID NO: 3) and the B chain contains the sequence

X ₂₅ LCGX ₂₉ X ₃₀ LVEALYLVCGERGFFYTX ₃₁ X ₃₂ (SEQ ID NO: 4), and

X ₈ is threonine or histidine;

X ₁₇ is glutamic acid or glutamine;

X ₁₉ is tyrosine, 4-methoxyphenylalanine or 4-aminophenylalanine;

X ₂₃ is asparagine or glycine;

X ₂₅ is histidine or threonine;

X29 is alanine, glycine or serine;

X30 is histidine, aspartic acid, glutamic acid, homocysteine acid or cysteic acid;

X31 is proline or lysine; And

X ₃₂ is proline or lysine, provided that at least one of X ₃₁ or X ₃₂ is lysine.

In another embodiment, the B strand contains the sequence X22VNQX25LCGX29X30LVEALYLVCGERGFFYT-X31X32X33X34X35 (SEQ ID NO: 5), wherein

- 16 041034

X ₂₂ is phenylalanine or deaminophenylalanine;

X ₂₅ is histidine or threonine;

X29 is alanine, glycine or serine;

X ₃₀ is histidine, aspartic acid, glutamic acid, homocysteine acid or cysteic acid;

X ₃₁ is aspartic acid, proline or lysine;

X32 is lysine or proline;

X ₃₃ is threonine, alanine or absent;

X ₃₄ is arginine or absent; And

X ₃₅ is arginine or absent; with the proviso that at least one of X ₃₁ or X ₃₂ is a lysine.

Connecting snippet.

The insulin dimers disclosed herein are formed between the first and second insulin polypeptides, with each insulin polypeptide containing an A chain and a B chain. The first and second insulin polypeptides may be double-chain insulin analogs (i.e., those in which the A and B chains are linked only by inter-chain disulfide bonds between internal cysteine residues), wherein the first and second insulin polypeptides are linked to each other to form dimer via a covalent bond, a bifunctional linker, or copper(I) catalyzed azide-alkyne cycloaddition (CuAAC) click chemistry or copper-free click chemistry to link linking moieties on the respective B chains. In one embodiment, the first and second insulin polypeptides are linked to each other via a bifunctional linker connecting the B28 or B29 lysine side chain of the first insulin polypeptide chain to the B28 or B29 amino acid side chain of the B chain of the second insulin polypeptide.

In certain aspects of insulin receptor partial agonists, the binding moiety may be an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. The linking moiety may be a divalent, straight or branched, saturated or unsaturated, optionally substituted C1-C20 hydrocarbon chain in which one or more methylene units are optionally and independently substituted with -O-, -S-, -N(R)-, -C(O)-, C(O)O-, OC(O)-, -N(R)C(O)-, -C(O)N(R)-, -S(O)-, - S(O) ₂ -, -N(R)SO ₂ -, SO ₂ N(R)-, a heterocyclic group, an aryl group, or a heteroaryl group, where each occurrence of R is independently hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl fragment, aliphatic fragment, aryl fragment, heteroaryl fragment or heteroaliphatic fragment.

In one embodiment, the linking moiety contains a PEG linker, a short linear polymer of about 2-25 ethylene glycol units or 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 , 16 or 25 ethylene glycol units, and optionally one or more amino acids. In certain aspects of insulin receptor partial agonists, the PEG linker comprises the structure (PEG)2, (PEG)3, (PEG)4, (PEG)5, (PEG)6, (PEG)7, (PEG)8, (PEG) 9, (PEG)10, (PEG)11, (PEG)12, (PEG)13, (PEG)14, (PEG) ₁₅ , (PEG) _16, or (PEG) ₂₅ . The PEG linker may be a bifunctional linker that may be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides. The structure of a bifunctional PEGlinker conjugated to the epsilon-amino group of lysine groups at position B29 or B28 of the first and second insulin polypeptides can be represented by the following general formula

o where n=1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, or 25, and the wavy line denotes the bond between the linker and the epsilon-amino group . Methods for conjugating PEG to the epsilon-amino group of lysine are well known in the art, see, for example, Veronese, Biomaterials 22: 405-417 (2001).

In certain aspects of insulin receptor partial agonists, the PEG binding fragment conjugating the lysine epsilon amino group at position B29 or B28 of the first insulin polypeptide to the lysine epsilon amino acid at position B29 or B28 of the second insulin polypeptide is

- 17 041034 or

wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the linking moiety contains an acyl moiety containing 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or 16 carbon atoms. In certain aspects of insulin receptor partial agonists, the acyl moiety is a succinyl (4), adipoyl (C6), suberoyl (C8), or hexadecanedioyl (C16) moiety. The acyl moiety may contain a bifunctional linker that may be covalently conjugated to or linked to the epsilonamino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides. The structure of a bifunctional acyl linker conjugated to the epsilon-amino group of the lysine group at position B29 or B28 of the first and second insulin polypeptides can be represented by the following general formula

wherein n=0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, or 12, and the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In certain aspects of insulin receptor partial agonists, the acyl binding moiety conjugating the lysine epsilon amino group at position B29 or B28 of the first insulin polypeptide to the lysine epsilon amino acid at position B29 or B28 of the second insulin polypeptide is

or wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In certain aspects of insulin receptor partial agonists, the bifunctional acyl linker may further include one or two amino acids at one or both ends of the acyl linker. For example, in certain aspects of insulin receptor partial agonists, the amino acid at one or both ends of the linker is gamma-glutamic acid (yE), which can be represented by the following general formula

- 18 041034

wherein n=0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, or 12, and the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the linking moiety contains a bifunctional amide-containing alkyl chain linker that can be covalently conjugated to or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented by the following general formula

where n=1 or 2, o=1, 2, 3, 4 or 5, and the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides. In one particular embodiment, the linking fragment may have the structure

wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the linking moiety contains a bifunctional amide-containing alkyl chain linker that can be covalently conjugated to or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented by the following general formula

where n=1, 2, 3, 4 or 5, o=1 or 2, p=1, 2, 3, 4 or 5, and the wavy lines indicate the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides . In one particular embodiment, the linking fragment may have the structure

wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the linking moiety comprises a bifunctional amide-containing alkyl chain linker that can be covalently conjugated to or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, and which can be represented by the following general formula

wherein n=1 or 2, o=1, 2 or 3, p=1 or 2, and the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides. In one particular embodiment, the linking fragment may have the structure

- 19 041034 wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In some embodiments, the implementation of the connecting fragment contains a ring structure, which provides rigidity of the connecting fragment. In certain embodiments, the ring structure contains a benzyl group or a saturated or unsaturated alicyclic group containing 3, 4, 5, 6, 7, or 8 carbon atoms. In certain embodiments, the alicyclic group contains cyclopropyl, cyclobutyl, cyclopentyl, cyclohexyl, cyclopentyl, or cyclooctyl. In certain embodiments, the unsaturated alicyclic group (cycloalkane) contains a cyclopropenyl, cyclobutenyl, cyclopentenyl, cyclohexenyl, cycloheptenyl, or cyclooctenyl group. In certain embodiments, the ring structure may further comprise one or more saturated or unsaturated aliphatic side chains. In some embodiments, the implementation of the ring structure may further contain one or more aliphatic side chains containing one or more heteroatoms. In certain embodiments, the heteroatom is O, S, or N.

In some embodiments, the implementation of the ring structure contains a heteroatom. In certain embodiments, the heteroatom may be O, S, or N. In certain embodiments, the ring structure contains a benzyl group or a saturated or unsaturated alicyclic group containing 3, 4, 5, 6, 7, or 8 carbon atoms in which one or more carbon atoms are substituted with a heteroatom selected from N, O, and S. Examples of ring structures that include a heteroatom include, but are not limited to, ethylene oxide, ethyleneimine, trimethyl oxide, furan, tetrahydrofuran, thiophene, pyrrolidine, pyran, piperidine, imidazole, thiazole , dioxane, morpholine, pyrimidine, triazole, thietane, 1,3-diazetin, 2,3-dihydroazet, 1,2-oxathiolane, isoxazole, oxazole, silol, oxepan, thiepin, 3,4,5,6-tetrahydro-2H -azepine, 1,4-thiazepine, azocan and thiocan.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,1-diacylbenzene having the following general formula

moreover, R1 and R ₂ may be the same or different, and R1 and R ₂ independently represent a bond, a saturated or unsaturated C1-C ₂₀ - or C1-C ₆ -alkyl chain, with one or more methylene units optionally and independently substituted by - O-, -S-, -N(R)-, -C(O)-, C(O)O-, OS(O)-, -N(R)C(O)-, -C(O) N(R)-, -S(O)-, -S(O) ₂ -, -N(R)SO ₂ -, SO ₂ N(R)-, heterocyclic group, aryl group or heteroaryl group, each occurrence R is independently hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic moiety, an aryl moiety, a heteroaryl moiety or a heteroaliphatic moiety, a poly(ethylene glycol) (PEG) chain (PEG) ₂ , (PEG) ₃ , (PEG) ₄ , (PEG) ₅ , (PEG) ₆ , (PEG) ₇ , (PEG) ₈ , (PEG) ₉ , (PEG) ₁₀ , (PEG)n, (PEG) ₁₂ , (PEG) ₁₃ , (PEG) ₁₄ , (PEG) ₁₅ , (PEG) ₁₆ or (PEG) ₂ , and where the wavy lines represent the bond between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin lipids.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,1-diacylbenzene having the following general formula

wherein n and o are independently 0, 1, 2, 3, 4, or 5, with wavy lines representing the bond between the linker and the lysine epsilon amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,3-diacylbenzene having the following general formula

moreover, R1 and R2 may be the same or different, and R1 and R2 independently represent a bond, a saturated or unsaturated C ₁ -C ₂₀ - or C ₁ -C ₆ -alkyl chain, with one or more methylene units optionally and independently substituted by - O-, -S-, -N(R)-, -C(O)-, C(O)O-, OS(O)-, -N(R)C(O)-, -C(O) N(R)-, -S(O)-, -S(O)2-, -N(R)SO2-, SO2N(R)-, a heterocyclic group, an aryl group, or a heteroaryl group, where each occurrence of R independently represents

- 20 041034 is hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic moiety, an aryl moiety, a heteroaryl moiety or a heteroaliphatic moiety, a poly(ethylene glycol) (PEG) chain (PEG)2, (PEG)3, (PEG) ₄ , (PEG)5, (PEG) ₆ , (PEG) ₇ , (PEG)8, (PEG)9, (PEG) ₁₀ , (PEG) ₁₁ , (PEG) ₁₂ , (PEG) ₁₃ , (PEG) ₁₄ , (PEG) ₁₅ , (PEG) ₁₆ or (PEG) ₂ , and wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,3-diacylbenzene, having the following general formula: n and o are independently 0, 1, 2, 3, 4, or 5, with wavy lines representing the linkage between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,4-diacylbenzene having the following general formula

moreover, R1 and R2 may be the same or different, and R ₁ and R2 independently represent a bond, a saturated or unsaturated C1-C ₂₀ - or C ₁ -C ₆ -alkyl chain, with one or more methylene units optionally and independently substituted by - O-, -S-, -N(R)-, -C(O)-, C(O)O-, OS(O)-, -N(R)C(O)-, -C(O) N(R)-, -S(O)-, -S(O)2-, -N(R)SO2-, SO2N(R)-, a heterocyclic group, an aryl group, or a heteroaryl group, where each occurrence of R independently represents is hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic moiety, an aryl moiety, a heteroaryl moiety or a heteroaliphatic moiety, a poly(ethylene glycol) (PEG) chain (PEG)2, (PEG)3, (PEG)4, (PEG )5, (PEG)6, (PEG)7, (PEG)8, (PEG)9, (PEG)10, (PEG)11, (PEG)12, (PEG)13, (PEG)14, (PEG )15, (PEG)16 or (PEG)2, and where the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,4-diacylbenzene having the following general formula: n and o are independently 0, 1, 2, 3, 4, or 5, with wavy lines representing the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,3-diacylbenzene having the following general formula

wherein m, n, and o are independently 1 or 2; R 1 and R 2 may be the same or different, and R ₁ and R ₂ independently represent a bond, saturated or unsaturated C ₁ -C ₂₀ or C 1 -C ₆ -alkyl chain, and one or more methylene units are optionally and independently substituted by -O -, -S-, -N(R)-, -C(O)-, C(O)O-, OC(O)-, -N(R)C(O)-, -C(O)N (R)-, -S(O)-, -S(O) ₂ -, -N(R)SO ₂ -, SO ₂ N(R)-, heterocyclic group, aryl group or heteroaryl group, and each occurrence of R is independently hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic moiety, an aryl moiety, a heteroaryl moiety or a heteroaliphatic moiety, a poly(ethylene glycol) (PEG) chain (PEG) ₂ , (PEG) ₃ , (PEG) 4 , (PEG)5, (PEG)6, (PEG)7, (PEG)8, (PEG)9, (PEG)10, (PEG)11, (PEG)12, (PEG)13, (PEG)14, (PEG)15, (PEG) ₁₆ or (PEG) ₂ , and where the wavy lines represent the link between the linker and the lysine epsilonamino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the linking fragment contains a bifunctional linker,

- 21 041034 which can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,3-diacylbenzene having the following general formula, ^where m, n, and o is independently 1 or 2; wherein p and q are independently 0, 1, 2, 3, 4, or 5, with wavy lines representing the link between the linker and the lysine epsilonamino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,3-diacylbenzene, having the following general formula: m, n, and o are independently 1 or 2; wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,4-diacylcyclohexane having the following general formula

and wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,4-diacylcyclohexane having the following general formula

and wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,3-diacylbenzene having the following general formula

wherein m and n are independently 0, 1, or 2, provided that both m and n are not 0; R 1 and R 2 may be the same or different, and R ₁ and R 2 independently represent a bond, saturated or unsaturated C 1 -C ₂₀ - or C ₁ -C ₆ -alkyl chain, with one or more methylene units optionally and independently substituted by -O -, -S-, -N(R)-, -C(O)-, C(O)O-, OS(O)-, -N(R)C(O)-, -C(O)N (R)-, -S(O)-, -S(O) ₂ -, -N(R)SO2-, SO ₂ N(R)-, a heterocyclic group, an aryl group or a heteroaryl group, with each occurrence of R independently is hydrogen, a suitable protecting group, an acyl moiety, an arylalkyl moiety, an aliphatic moiety, an aryl moiety, a heteroaryl moiety or a heteroaliphatic moiety, a poly(ethylene glycol) (PEG) chain (PEG)2, (PEG) ₃ , (PEG) ₄ , ( PEG) ₅ , (PEG) ₆ , (PEG) ₇ , (PEG) ₈ , (PEG) ₉ , (PEG) ₁₀ , (PEG) _1b (PEG) 12 , (PEG) _{13 ,} (PEG) ₁₄ , (PEG ) ₁₅ , (PEG) ₁₆ or (PEG) 2 , and where the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,3-diacylbenzene having the following general formula

- 22 041034 wherein m and n are independently 1 or 2; wherein p and q are independently 0, 1, 2, 3, 4, or 5, with wavy lines representing the link between the linker and the lysine epsilonamino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,1-diacyl, having the following general formula

wherein n is 1, 2, 3, or 4, wherein the wavy lines represent the bond between the linker and the lysine epsilon amino group at position B29 or B28 of the insulin polypeptides. In specific embodiments, the 1,1-diacyl may have a structure selected from

(1,1-diacyl-C3; 1,2-diacyl-C4; 1,1-diacyl-C5 and 1,1-diacyl-C6, respectively), with wavy lines indicating the bond between the linker and the lysine epsilon-amino group at position B29 or B28 insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,2-diacyl having the following general formula

wherein n is 1, 2, 3, or 4, wherein the wavy lines represent the bond between the linker and the lysine epsilon amino group at position B29 or B28 of the insulin polypeptides. In specific embodiments, the 1,2-diacyl may have a structure selected from

(1,2-diacyl-C3; 1,2-diacyl-C4; 1,2-diacyl-C5 and 1,2-diacyl-C6, respectively), with wavy lines indicating the bond between the linker and the lysine epsilon-amino group at position B29 or B28 insulin polypeptides.

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,3-diacyl having the following general formula

wherein n is 1, 2, or 3, wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides. In specific embodiments, the 1,3-diacyl may have a structure selected from

(1,3-diacyl-C4; 1,3-diacyl-C5 and 1,3-diacyl-C6, respectively), with wavy lines representing the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another embodiment, the binding fragment contains a bifunctional linker that can be covalently conjugated to or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,4-diacyl having the following general formula

(1,4-diacyl-C6), with wavy lines representing the link between the linker and the lysine epsilonamino group at position B29 or B28 of the insulin polypeptides.

- 23 041034

In another embodiment, the binding moiety contains a bifunctional linker that can be covalently conjugated or linked to the epsilon-amino group of lysine residues at position B29 or B28 of the first and second insulin polypeptides, which can be represented as 1,3-diacylcyclobutyl, having the following general formula

and wherein the wavy lines represent the link between the linker and the lysine epsilon-amino group at position B29 or B28 of the insulin polypeptides.

In another aspect of the present invention, the first and second insulin polypeptides can be conjugated together using a copper-catalyzed Huisgen azide-alkyne cycloaddition (CuAAc), in particular CuAAC click chemistry. In this aspect, the epsilon-amino group of lysine B28 or B29 of the first insulin polypeptide is conjugated to a linker fragment having a proximal end and a distal end, wherein the proximal end of the linker fragment is conjugated to the epsilon-amino group, and the distal end contains an azide group. In this aspect, the epsilonamino group of lysine B28 or B29 of the second insulin polypeptide is conjugated to a linker fragment having a proximal end and a distal end, wherein the proximal end of the linker fragment is conjugated to an epsilon amino group and the distal end contains an alkyne group. In the presence of Cu ²⁺ and a reducing agent, the azide and alkyne groups will form a continuous linking moiety containing a triazole moiety. For a description of the click chemistry of CuAAC, see US Pat. No. 8,129,542, which is incorporated herein in its entirety.

In certain aspects of insulin receptor partial agonists, the first insulin polypeptide may have a lysine B28 or B29 epsilon-amino conjugated linker having the formula

and the second insulin polypeptide may have a B28 or B29 lysine epsilon-amino conjugated linker having the formula

In the presence of Cu ²⁺ and a reducing agent, the linkers combine to form a linking fragment having the structure

In certain aspects of insulin receptor partial agonists, the first insulin polypeptide may have a lysine B28 or B29 epsilon-amino conjugated linker having the formula

where n=1, 2, 3, 4, 5, 6, 7, 8, 9 or 10, and the second insulin polypeptide may have a linker conjugated with the epsilon-amino group of lysine B28 or B29 having the formula

wherein n=1, 2, 3, 4, 5, 6, 7, 8, 9, or 10. In the presence of Cu ²⁺ and a reducing agent, the linkers combine to form a linking moiety having the structure

- 24 041034

wherein each n is independently 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10.

In another aspect, both the first insulin polypeptide and the second insulin polypeptide may have a linker conjugated to their respective B28 or B29 lysine epsilon-amino group having the formula wherein each n is independently 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10. Conjugation of linkers to form a binding moiety can be achieved by providing a molecule (intermediate or bridging linker) having the structure

----r-- wherein R is a covalent bond, a carbon atom, a phenyl, a heteroatom, or an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. In certain aspects, R is a C2, C3, C4, C6, C7, C8, C9, or C10 acyl group, or PEG2, PEG3, PEG4, PEG5, PEG6, PEG7, PEG8, PEG9, PEG10, PEG11, PEG12, PEG13, or PEG25.

In another aspect, both the first insulin polypeptide and the second insulin polypeptide may have a linker conjugated to their respective lysine B28 or B29 epsilon-amino group having the formula o

'1 V .

where each n is independently 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10. Conjugation of linkers to form a linking moiety can be achieved by providing a molecule (intermediate or bridging linker) having the structure

---R--wherein R represents a covalent bond, a carbon atom, a phenyl, a heteroatom, or an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. In certain aspects, R is a C2, C3, C4, C6, C7, C8, C9, or C10 acyl group, or PEG2, PEG3, PEG4, PEG5, PEG6, PEG7, PEG8, PEG9, PEG10, PEG11, PEG12, PEG13, or PEG25.

In certain aspects, the first insulin polymer is conjugated at the epsilon amino group of lysine B28 or B29 to an azide terminating linker as described above, and the second insulin polypeptide is conjugated at the epsilon amino group of lysine B28 or B29 to a linker terminating in a cyclooctyne moiety, and the linkers are conjugated to form a linker fragment using cycloaddition click chemistry without copper. See, for example, US Pat. No. 7,807,619, which is incorporated herein in its entirety.

The following table shows examples of linkers that can be used to construct the dimers of the present invention. The dimers shown contain 2,5-dioxopyrrolidin-1yl groups for conjugation to the epsilon-amino group of lysine B29 or B29.

- 25 041034

2 LJ z—4 o o ^z ° )=° .. C6N+C6+NC 6 3 Clu _? ιζ \=o O=( ΖΙ P L z—A o o / \ yE-S8-yE 4 F ^ o ^5? Click-1 5 _x o /Ύ II GxA—Ν ₃ 0 Click-2 6 Z—V 0 0 ^F yV 1 [1 ¹ A _Ν - ' * 0 A ³ p ^H F Click-3 7 O Click-4

- 26 041034

8 O ( 1 o \-N Ji .., G ABOUT * Click-5 9 Click-6 10 Q°V\ ( About 1 / </ o^^pp b °0 C2 eleven z—4 o' ° / o /° /—' ^z C4 12 ^o< x C6 13 z?---'2 o \ p ° \ )=o °\ __zP Y C8

- 27 041034

- 28 041034

- 29 041034

- 30 041034

25 </ ° ζι 0 \ R Sat-alanine 26 o \ o o=^ / IZ \ O z° G ^Z UL^O Sbisoleucine 27 about ⁷ ° \ ? ^ζ - /^° o \ ζ ^Ο o^yj Sat-leucine 28 o=yy nU s/ ° ΖΙ —\ r ⁰ \ o \ /° oUU Sat-valine 29 YY==O Y^ o ° ^o= \ o _x i ^ ^z Dipropylph enol thirty 0 O /—, 0 0 l >Ol k 1 N-0 ^{4/ 0-N} \J 0 0 Transcyclohex an-1,4diacid 31 О 0 /—Ч 0 0 L-oN— 0 C) Ciscyclohex an-1,4diacid

- 31 041034

32 /O °. pnV Sr O'O tertbutylpipe ridintric arboxyl t 33 Cl 0 0 C6H-chloro- 1,3,5- triazine- NC6 34 if'A o Ο Γ ~¥ nA A ⁰ A 0 Terephthal t 35 ο ίί^Ί O X^yJUysA VA o AV Ό 0 ^z isophthalate 36 _G BosN/C^ O^o O^ ^N y^O Heptandio at 37 <? ? ? 0 4 eleven- diacyl-SZ 38 Oh oh 1.1- diacyl-C4

- 32 041034

39 η OO Oh Oh 1.1- diacyl-C5 40 O ? oh oh ° o n=1, 2, 3 or 4 1.1- diacyl-Sb 41 oX 1 w o o=/ T 1,2- diacyl-SZ 42 0 Υχ, ° Ol 1 °r 1,2- diacyl-C4 43 o=\ 1 W oW °\L! about t 1,2- diacyl-C5 44 4°% θW 1,2- diacyl-Sb 45 Oh. 1 W o o=/ ^o= \ o Z-^O o=/ T 1.3- diacyl-C4 46 oX 1 W o o \u003d / ° x o Z-W o=/ T 1.3- diacyl-C5 47 o=\ ί WO o zW o=/ Y 1.3- diacyl-Sb 48 J 0=0 oW - o zW o=/ T 1.4- diacylcyclobuty l-C1 49 CT^^'V ъ ° 1.4- cyclohex IL-C1 50 oh wtw 1.4- cyclohex IL-C2

- 33 041034

Conjugation of the bifunctional linker with the epsilon-amino group of the lysine residue at position

B29 or B28 of the B chain polypeptide of two molecules of insulins or insulin analogues to form an insulin dimer linked via a binding moiety can be schematically shown as

moreover, the insulin 1 and insulin 2 molecules may be the same or different, and the bifunctional linker and the resulting binding fragment after conjugation may have the structure of any linker and the resulting binding fragment disclosed herein.

Modification of insulin polypeptides.

In some embodiments, at least one of the A chain polypeptides or B chain polypeptides of a partial insulin receptor agonist is modified to contain an acyl group. The acyl group may be covalently linked to the amino acid of the insulin polypeptide directly or indirectly to the amino acid of the insulin polypeptide via a spacer, the spacer being located between the amino acid of the insulin polypeptide and the acyl group. The insulin polypeptide may be acylated at the same amino acid position at which the hydrophilic moiety is bound, or at a different amino acid position. For example, acylation can occur at any position, including any amino acid of the A- or B-chain polypeptides, as well as a position within the binding moiety, provided that the activity exhibited by the non-acylated insulin polypeptide is retained after acylation. Non-limiting examples include acylation at the A1 position of the A chain and the B1 position of the B chain.

In one specific aspect of the present invention, the first and/or second insulin polypeptide (or derivative or conjugate thereof) is modified to contain an acyl group by direct acylation of the amine, hydroxyl, or thiol of the amino acid side chain of the insulin polypeptide. In some embodiments, the first and/or second insulin polypeptide is directly acylated at the amine, hydroxyl, or thiol of the amino acid side chain. In this regard, an insulin polypeptide can be provided that is modified by one or more amino acid substitutions in the A or B chain polypeptide sequence, including, for example, at positions A1, A14, A15, B1, B10 or B22, or at any position of the binding moiety, to an amino acid containing an amine, hydroxyl or thiol in the side chain.

In some embodiments, the spacer between the first and/or second insulin polypeptide and the acyl group is an amino acid containing an amine, hydroxyl, or thiol in the side chain (or a dipeptide or tripeptide containing an amino acid containing an amine, hydroxyl, or thiol in the side chain). In some embodiments, the spacer comprises a hydrophilic bifunctional spacer. In one particular embodiment, the spacer contains an amino poly(alkyloxy)carboxylate. In this regard, the spacer may contain, for example, NH ₂ (CH ₂ CH ₂ O) _n (CH ₂ ) _m COOH, where m is any integer from 1 to 6 and n is any number from 2 to 12, including, for example, 8-amino-3,6-dioxooctanoic acid, which is commercially available from Peptides International, Inc. (Louisville, KY). In one embodiment, the hydrophilic bifunctional spacer contains two or more reactive groups, such as an amino group, hydroxyl, thiol and carboxyl group, or any combination thereof. In some embodiments, the hydrophilic bifunctional spacer contains a hydroxyl group and a carboxylate. In other embodiments, the hydrophilic bifunctional spacer contains an amino group and a carboxylate. In other embodiments, the hydrophilic bifunctional spacer contains a thiol group and a carboxylate.

In some embodiments, the spacer between the first and/or second insulin polypeptide and the acyl group is a hydrophobic bifunctional spacer. Hydrophobic bifunctional spacers are known in the art. See, for example, Bioconjugate Techniques, G. T. Hermanson (Academic Press, San Diego, CA, 1996), which is incorporated by reference in its entirety. In some embodiments, the implementation of the hydrophobic bifunctional spacer contains two or more reactive groups, such as amino group, hydroxyl, thiol and carboxyl group, or any combination thereof. In some embodiments, the hydrophobic bifunctional spacer contains a hydroxyl group and a carboxylate. In other embodiments, the hydrophobic bifunctional spacer contains an amino group and a carboxylate. In other embodiments, the hydrophobic bifunctional spacer contains a thiol group and a carboxylate. Suitable hydrophobic bifunctional spacers containing a carboxylate and a hydroxyl group or thiol group are known in the art and include, for example, 8hydroxyoctanoic acid and 8-mercaptooctanoic acid.

In some embodiments, the bifunctional spacer may be a synthetic or naturally occurring amino acid containing an amino acid backbone that is 3 to 10 atoms in length (e.g., 6-aminocaproic acid, 5-aminovaleric acid, 7-aminoheptanoic acid, and 8-aminooctanoic acid ). Alternatively, the spacer may be a dipeptide or tripeptide spacer having a peptide backbone that is 3 to 10 atoms long (eg, 6 to 10 atoms). Each amino acid of the dipeptide or tripeptide spacer attached to the insulin polypeptide may be independently selected from the group consisting of: natural and/or non-natural amino acids, including, for example, any of the D or L isomers of naturally occurring amino acids (Ala, Cys, Asp, Glu, Phe, Gly, His, Ile, Lys, Leu, Met, Asn, Pro, Arg, Ser, Thr, Val, Trp, Tyr) or any D or L isomer of unnatural amino acids selected from the group consisting of: β-alanine ( β-Ala), N-a-methylalanine (Me-Ala), aminobutyric acid (Abu), αaminobutyric acid (γ-Abu), aminocaproic acid (ε-Ahx), aminoisobutyric acid (Aib), aminomethylpyrrolecarboxylic acid, aminopiperidinecarboxylic acid, aminoserine ( Ams), aminotetrahydropyran-4-carboxylic acid, N-methoxy-N-methylamide arginine, β-aspartic acid (β-Asp), azetidinecarboxylic acid, 3-(2-benzothiazolyl)alanine, α-tert-butylglycine, 2-amino -5ureido-n-valeric acid (citrulline, Cit), β-cyclohexylalanine (Cha), acetamidomethyl cyste ine, diaminobutyric acid (Dab), diaminopropionic acid (Dpr), dihydroxyphenylalanine (DOPA), dimethylthiazolidine (DMTA), γ-glutamic acid (γ-Glu), homoserine (Hse), hydroxyproline (Hyp), N-methoxy-N- isoleucine methylamide, methylisoleucine (MeIle), isonipecotic acid (Isn), methylleucine (MeLeu), methyllysine, dimethyllysine, trimethyllysine, methanoproline, methionine sulfoxide (Met(O)), methionine sulfone (Met(O2)), norleucine (Nle), methylnorleucine ( Me-Nle), norvaline (Nva), ornithine (Orn), para-aminobenzoic acid (PABA), penicillamine (Pen), methylphenylalanine (MePhe), 4-chlorophenylalanine (Phe(4-Cl)), 4-fluorophenylalanine (Phe(4 -F)), 4-nitrophenylalanine (Phe(4-NO2)), 4-cyanophenylalanine ((Phe(4-CN)), phenylglycine (Phg), piperidinylalanine, piperidinylglycine, 3,4-dehydroproline, pyrrolidinylalanine, sarcosine (Sar), selenocysteine (Sec), O-benzylphosphoserine, 4-amino-3-hydroxy-6-methylheptanoic acid (Sta), 4-amino-5-cyclohexyl-3-hydroxypentanoic acid (ACHPA), 4-amine o-3-hydroxy-5-phenylpentanoic acid (AHPPA), 1,2,3,4-tetrahydroisoquinoline-3-carboxylic acid (Tic), tetrahydropyranglycine, thienylalanine (Thi), O-benzylphosphotyrosine, O-phosphotyrosine, methoxytyrosine, ethoxytyrosine , O-(bisdimethylaminophosphono)tyrosine, tetrabutylamine tyrosine sulfate, methylvaline (MeVal), 1-amino-1cyclohexanecarboxylic acid (Acx), aminovaleric acid, beta-cyclopropylalanine (Cpa), propargylglycine (Prg), allylglycine (Alg), 2-amino- 2-cyclohexylpropanoic acid (2-Cha), tert-butylglycine (Tbg), vinylglycine (Vg), 1-amino-1-cyclopropanecarboxylic acid (Acp), 1-amino-1-cyclopentanecarboxylic acid (Acpe), alkylated 3-mercaptopropionic acid, 1- amino-1-cyclobutanecarboxylic acid (Acb). In some embodiments, the dipeptide spacer is selected from the group consisting of: Ala-Ala, β-Ala-e-Ala, Leu-Leu, Pro-Pro, γ-aminobutyric acid-γaminobutyric acid, and γ-Glu-Y-Glu.

The first and/or second insulin polypeptide may be modified to contain an acyl group by acylation of the long chain alkane. In specific aspects, the long chain alkane contains an amino, hydroxyl, or thiol group (eg, octadecylamine, tetradecanol, and hexadecanethiol) that is reactive with the carboxyl group of the insulin polypeptide, or an activated form thereof. The carboxyl group of the insulin polypeptide, or an activated form thereof, may be part of the amino acid (eg, glutamic acid, aspartic acid) side chain of the insulin polypeptide, or may be part of the peptide backbone.

In some embodiments, the first and/or second insulin polypeptide is modified to contain an acyl group by acylation of a long chain alkane with a spacer that is attached to the insulin polypeptide. In specific aspects, the long chain alkane contains an amino group, a hydroxyl group, or a thiol group that is reactive with the carboxyl group of the spacer, or an activated form thereof. Suitable spacers containing a carboxyl group or an activated form thereof are described herein and include, for example, bifunctional spacers such as amino acids, dipeptides, tripeptides, hydrophilic bifunctional spacers, and hydrophobic bifunctional spacers. As used herein, the term activated form of a carboxyl group refers to a carboxyl group with the general formula R(C=O)X, where X is a leaving group and R is an insulin polypeptide or spacer. For example, activated forms of carboxyl groups may include, but are not limited to, acyl chlorides, anhydrides, and esters. In some embodiments, the activated carboxyl group is an N-35 041034 hydroxysuccinimide (NHS) leaving group ester.

In those aspects of the present invention in which the long chain alkane is acylated with a peptide, insulin polypeptide, or spacer, the long chain alkane may be of any size and may contain a carbon chain of any length. The long chain alkane may be linear or branched. In some aspects, the long chain alkane is a C ₄ -C ₃₀ alkane. For example, the long chain alkane can be any of C ₄ alkane, C ₆ alkane, C ₈ alkane, C ₁₀ alkane, C ₁₂ alkane, C ₁₄ alkane, C ₁₆ alkane, C ₁₈ alkane, C ₂₀ alkane, C ₂₂ alkane, C ₂₄ alkane, C ₂₆ alkane, C ₂₈ alkane, or C ₃₀ alkane. In some embodiments, the long chain alkane comprises a C ₈ C ₂₀ alkane, such as a C ₁₄ alkane, a C ₁₆ alkane, or a C ₁₈ alkane.

In some embodiments, the amino, hydroxy or thiol group of the first and/or second insulin polypeptide is acylated with cholesterol acid. In one specific embodiment, the peptide is linked to cholesterol acid via an alkylated desamino Cys spacer, i. spacer on the conic alkylated 3-mercaptopropionic acid. Suitable methods for acylation of peptides at amines, hydroxyls and thiols are known in the art. See, for example, Miller, Biochem. Biophys. Res. Commun 218: 377-382 (1996); Shimohigashi and Stammer, Int. J. Pept. Protein Res. 19:54-62 (1982); and Previero et al., Biochim. Biophys. Acta 263: 7-13 (1972) (with respect to hydroxyl acylation methods); and San and Silvius, J. Pept. Res. 66: 169-180 (2005) (with respect to thiol acylation methods); Bioconjugate Chem. Chemical Modifications of Proteins: History and Applications, p. 1, 2-12 (1990); Hashimoto et al., Pharmaceutical Res. Synthesis of Palmitoyl Derivatives of Insulin and their Biological Activity Vol. 6, NO: 2, p. 171-176 (1989).

The acyl group of the acylated peptide of the first and/or second insulin polypeptide may be of any size, such as any carbon chain length, and may be linear or branched. In some specific embodiments of the present invention, the acyl group is a C4-C30 fatty acid. For example, the acyl group may be any of C ₄ fatty acids, C ₆ fatty acids, C ₈ fatty acids, C 10 fatty acids, C ₁₂ fatty acids, C ₁₄ fatty acids, C ₁₆ fatty acids, C ₁₈ fatty acids, C ₂₀ fatty acid, C ₂₂ fatty acid, C ₂₄ fatty acid, C ₂₆ fatty acid, C ₂₈ fatty acid, or C ₃₀ fatty acid. In some embodiments, the acyl group is a C ₈ -C ₂₀ fatty acid, such as a C ₁₄ fatty acid or a C ₁₆ fatty acid. In some embodiments, the acyl group is urea.

In an alternative embodiment, the acyl group is a bile acid. The bile acid can be any suitable bile acid, including, but not limited to, cholic acid, chenodeoxycholic acid, deoxycholic acid, lithocholic acid, taurocholic acid, glycocholic acid, and cholesterol acid.

The acylated first and/or second insulin polypeptide described herein may be further modified such that it contains a hydrophilic moiety. In some specific embodiments, the implementation of the hydrophilic fragment may contain polyethylene glycol (PEG) chain. The incorporation of the hydrophilic moiety can be accomplished by any suitable means, such as any of the methods described herein. In some embodiments, the acylated single chain analog contains an amino acid selected from the group consisting of Cys, Lys, Orn, Homo-Cys, or Ac-Phe and the amino acid side chain is covalently linked to a hydrophilic moiety (eg, PEG). In one embodiment, the acyl group is attached to position A1, A14, A15, B1, B2, B10 or B22 (according to the amino acid numbering of the A and B chains of native insulin), optionally via a spacer containing Cys, Lys, Orn, homo -Cys or Ac-Phe.

Alternatively, the acylated first and/or second insulin polypeptide contains a spacer, the spacer being both acylated and modified to contain a hydrophilic moiety. Non-limiting examples of suitable spacers include a spacer containing one or more amino acids selected from the group consisting of Cys, Lys, Orn, Homo-Cys, and Ac-Phe.

In some embodiments, the amino terminus of at least one N-terminal amino acid of at least one of the A chain polypeptides and B chain polypeptides of the partial insulin receptor agonist is modified to contain a substituent. The substituent may be covalently linked to the amino group of the N-terminal amino acid directly or to the amino group indirectly through a spacer, the spacer being located between the amino group of the N-terminal amino acid of the insulin polypeptide and the substituent. The substituent may be an acyl moiety as discussed above. The substituent may have the general formula RC(O)- where R may be R'CH ₂ , R'NH, R'O, and R' may be H, a linear alkyl chain, an amino acid, a peptide, polyethylene glycol (PEG), saccharides, i.e. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is a carbamoyl group, an acetyl group, a glycine, a methyl group, a methoxy group, a dimethyl group, an isobutyl group, a PEG1 group, or

- 36 041034

PEG2-rpynny (for substituent structures, see examples herein). Insulin carbamoylation has been disclosed by Oimoni et al., Nephron 46: 63-66 (1987), and insulin dimers containing carbamoyl groups at the N-terminus have been disclosed in PCT Publication No. WO 2014052451 (eg, MIU-90).

In certain embodiments, at least one N-terminal amino acid is conjugated at the N2 nitrogen to a substituent containing an N-hydroxysuccinimide ester linked to a group having the general formula RC(O)-, where R may be R'CH ₂ , R 'NH, R'O, and R' may be H, linear alkyl chain, amino acid, peptide, polyethylene glycol (PEG), saccharides, i.e. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is a carbamoyl group, an acetyl group, a glycine, a methyl group, a methoxy group, a dimethyl group, an isobutyl group, a PEGI group, or a PEO2 group.

In certain embodiments, the saccharide covalently linked to one or more amino termini of the first and second insulin polypeptides may be a monosaccharide, see for example dimer 51. In some embodiments, the saccharide contains one or more amino groups. In some embodiments, the saccharide and amino group are separated by a C ₁ -C ₆ alkyl group, such as a C ₁ -C ₃ alkyl group. In some embodiments, the saccharide is aminoethylglucose (AEG). In some embodiments, the saccharide ligand is in the D configuration. In other embodiments, the saccharide ligand is in the L configuration. The structures of these exemplary saccharides are shown below. Other exemplary saccharides will be apparent to those skilled in the art.

Typically, the saccharides can be directly or indirectly conjugated via an amino-terminal linker of one or more of the first and second insulin polypeptides. In certain aspects, the linker is an alkyldioyl, -C(O)(CH ₂ ) _n C(O)-, where n=0-45, 0-20, 0-10, or 05.

Representative substituents conjugated to the N-terminal amino group may be

wherein the wavy line denotes the bond between the substituent and the N-terminal amino group. The substituent may also be (Me2; N-dimethyl), with the wavy line representing the bond between Me2 and the carbon-alpha of the N-terminal amino acid.

Examples of insulin dimers.

In certain embodiments, the present invention provides insulin dimers wherein the first B29 or B28 Lys of a first insulin heterodimer molecule comprising a first A-chain polypeptide and a first B-chain polypeptide and a second B29 or B28 Lys of a second insulin heterodimer comprising a second A-chain polypeptide chains and a second B chain polypeptide are conjugated together via a bifunctional linker selected from the group consisting of linker 1, linker 2, linker 3, linker 10, linker 11, linker 12, linker 13, linker 14, linker 15, linker 16, linker 17, linker 18, linker 19, linker 20, linker 21, linker 22, linker 23, linker 24, linker 25, linker 26, linker 27, linker 28, linker 29, linker 30, linker 31, linker 32, linker 33 , Linker 34, Linker 35, Linker 36, Linker 37, Linker 38, Linker 39, Linker 40, Linker 41, Linker 42, Linker 43, Linker 44, Linker 45, Linker 46, Linker 47, Linker 48, Linker 49 and Linker 50, provided that when the bifunctional linker is a 10 linker, a 11 linker, a 12 linker, a 13 linker, or a 14 linker, at least one of the first or second A chain or B chain polypeptides is conjugated at its N-terminal amino acid to a substituent disclosed herein, or at least the N-terminal amino acids of the first insulin heterodimer molecule are conjugated to a substituent disclosed herein, or the N-terminal amino acids of both the first insulin heterodimer and the second insulin heterodimer are conjugated to a substituent. In certain embodiments, the substituent contains an N-hydroxysuccinimide ester linked to a group having the general formula RC(O)-, where R may be R'CH2, R'NH, R'O, and R' may be H, linear alkyl chain, amino acid, peptide, polyethylene glycol (PEG), saccharides, i.e. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is a carbamoyl group, an acetyl group, a glycine, a methyl group, a methoxy group, a dimethyl group, an isobutyl group, a PEGI group, an AEG group, an AEG-C6 alkyl group, or a PEG2 group.

In certain embodiments, the present invention provides insulin dimers wherein the first B29 or B28 Lys of a first insulin heterodimer molecule comprising a first A chain polypeptide and a first B chain polypeptide conjugated to a first linker selected from the group consisting of linker 5 and linker 7 and a second B29 or B28 Lys of a second insulin heterodimer comprising a second A chain polypeptide and a second B chain polypeptide conjugated to a second linker selected from the group consisting of linker 4, linker 6, linker 8 and linker 9 conjugated together through the first linker and the second linker. In certain embodiments, at least one of the first or second A chain or B chain polypeptides is conjugated at its N-terminal amino acid to a substituent disclosed herein, or at least the N-terminal amino acids of the first insulin heterodimer molecule are conjugated to a substituent, disclosed herein, or the N-terminal amino acids of both the first insulin heterodimer and the second insulin heterodimer are conjugated to a substituent. In certain embodiments, the substituent contains an N-hydroxysuccinimide ester linked to a group having the general formula RC(O)-, where R may be R'CH ₂ , R'NH, R'O, and R' may be H , linear alkyl chain, amino acid, peptide, polyethylene glycol (PEG), saccharides, i. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is a carbamoyl group, an acetyl group, a glycine, a methyl group, a methoxy group, a dimethyl group, an isobutyl group, a PEGI group, an AEG group, an AEG-C6 alkyl group, or a PEG2 group.

In certain embodiments, the present invention provides insulin dimers wherein the first B29 or B28 Lys of a first insulin heterodimer molecule comprising a first A chain polypeptide and a first B chain polypeptide is conjugated to a first linker selected from the group consisting of linker 5 and linker 7 and a second B29 or B28 Lys of a second insulin heterodimer comprising a second A chain polypeptide and a second B chain polypeptide is conjugated to a second linker selected from the group consisting of linker 5 and linker 7, the first and second linkers being conjugated together via bridging linker having the structure

--------r------where R is a covalent bond, a carbon atom, a phenyl, a heteroatom, or an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic . In certain aspects, R is a C2, C3, C4, C6, C7, C8, C9, or C10 acyl group, or PEG2, PEG3, PEG4, PEG5, PEG6, PEG7, PEG8, PEG9, PEG10, PEG11, PEG12, PEG13, or PEG25. In certain embodiments, at least one of the first or second A chain or B chain polypeptides is conjugated at its N-terminal amino acid to a substituent disclosed herein, or at least the N-terminal amino acids of the first insulin heterodimer molecule are conjugated to a substituent, disclosed herein, or the N-terminal amino acids of both the first insulin heterodimer and the second insulin heterodimer are conjugated to a substituent. In certain embodiments, the substituent contains an N-hydroxysuccinimide ester linked to a group having the general formula RC(O)-, where R may be R'CH ₂ , R'NH, R'O, and R' may be H , linear alkyl chain, amino acid, peptide, polyethylene glycol (PEG), saccharides, i. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is a carbamoyl group, an acetyl group, a glycine, a methyl group, a methoxy group, a dimethyl group, an isobutyl group, a PEGI group, an AEG group, an AEG-C6 alkyl group, or a PEG2 group.

In certain embodiments, the present invention provides insulin dimers wherein the first B29 or B28 Lys of a first insulin heterodimer molecule comprising a first A chain polypeptide and a first B chain polypeptide is conjugated to a first linker selected from the group consisting of linker 4, linker 6, linker 8 and linker 9, and a second B29 or B28 Lys of a second insulin heterodimer comprising a second A chain polypeptide and a second B chain polypeptide is conjugated to a second linker selected from the group consisting of linker 4, linker 6, linker 8 and linker 9, wherein the first and second linkers are conjugated together via a bridging link

- 38 041034 ra having the structure

N ₃ ---R---N ₃ where R is a covalent bond, a carbon atom, a phenyl, a heteroatom, or an optionally substituted group selected from the group consisting of acyl, aliphatic, heteroaliphatic, aryl, heteroaryl, and heterocyclic. In certain aspects, R is a C2, C3, C4, C6, C7, C8, C9, or C10 acyl group, or PEG2, PEG3, PEG4, PEG5, PEG6, PEG7, PEG8, PEG9, PEG10, PEG11, PEG12, PEG13, or PEG25. In certain embodiments, at least one of the first or second A chain or B chain polypeptides is conjugated at its N-terminal amino acid to a substituent disclosed herein, or at least the N-terminal amino acids of the first insulin heterodimer molecule are conjugated to a substituent, disclosed herein, or the N-terminal amino acids of both the first insulin heterodimer and the second insulin heterodimer are conjugated to a substituent. In certain embodiments, the substituent contains an N-hydroxysuccinimide ester linked to a group having the general formula RC(O)-, where R may be R'CH ₂ , R'NH, R'O, and R' may be H , linear alkyl chain, amino acid, peptide, polyethylene glycol (PEG), saccharides, i. in certain aspects, RC(O)- may be acetyl, phenylacetyl, carbamoyl, N-alkylcarbamoyl, or alkoxycarbonyl. In certain aspects, the substituent is a carbamoyl group, an acetyl group, a glycine, a methyl group, a methoxy group, a dimethyl group, an isobutyl group, a PEGI group, an AEG group, an AEG-C6 alkyl group, or a PEG2 group.

In other embodiments, the first and second insulin heterodimers may comprise any of the insulin molecules or insulin analogs disclosed herein.

The present invention also provides insulin dimers selected from

- 39 041034

dimera 8

- 40 041034

dimera 13

- 41 041034

dimera 17

- 42 041034

dimera 20

- 43 041034

- 44 041034 dimer 24

dimera 30

- 45 041034

dimera 36

- 46 041034

dimera 39

- 47 041034

- 48 041034

dimera 45

- 49 041034

- 50 041034

- 51 041034

- 52 041034

dimera 57

- 53 041034

dimera 60

- 54 041034

- 55 041034

- 56 041034

- 57 041034

- 58 041034

dimera 75

- 59 041034

- 60 041034

- 61 041034

dimera 86

- 62 041034

dimera 89

- 63 041034

dimera 94

Moreover, the disulfide bonds between the residues Cys6 and Cys11 of the A-chain polypeptide and the disulfide bonds between Cys ₇ and Cys ₂₀ of the A-chain with Cys ₇ and Cys ₁₉ of the B-chain polypeptide, respectively, are represented by a solid line between them; moreover, the binding fragments are covalently linked to the epsilon-amino acid of the shown lysine residue, and the A-chain polypeptide for dimers 1-40, 42-52, 5486 and 88-94 has the amino acid sequence shown in SEQ ID NO: 1; the A chain polypeptide for dimer 56 has the amino acid sequence shown in SEQ ID NO: 11; the chain polypeptide for dimers 1-17, 21-27, 36, 37, 39-40 and 42-52, 54-82, 84-86 and 88-94 has the amino acid sequence shown in SEQ ID NO: 2; the B chain polypeptide for dimers 18 and 32-35 has the amino acid sequence shown in SEQ ID NO: 6; the B chain polypeptide for dimers 19 and 83 has the amino acid sequence shown in SEQ ID NO: 9; the B chain polypeptide for dimers 20, 28-31 and 38 has the amino acid sequence shown in SEQ ID NO: 10; and the Acepi polypeptide and the B chain polypeptide for dimers 53 and 87 are SEQ ID NO: 7 and SEQ ID NO: 8, respectively.

pharmaceutical compositions.

In accordance with one embodiment, a pharmaceutical composition is provided comprising any of the novel insulin dimers disclosed herein, preferably with a purity level of at least 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99 %, and pharmacists

- 64 041034 a theoretically acceptable diluent, carrier or excipient. Such compositions may contain the insulin dimer disclosed herein at a concentration of at least 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 , 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25 mg/ml or more. In one embodiment, the pharmaceutical compositions contain aqueous solutions that are sterilized and optionally stored in various packaging containers. In other embodiments, the pharmaceutical compositions comprise a lyophilized powder. Pharmaceutical compositions may also be packaged as part of a kit that includes a disposable device for administering the composition to a patient. Containers or kits may be labeled for storage at room temperature or refrigerated storage.

The disclosed insulin dimers are believed to be suitable for any application previously described for insulin peptides. Accordingly, the insulin dimers disclosed herein can be used to treat hyperglycemia or other metabolic diseases that result from high blood glucose levels. Accordingly, the present invention encompasses pharmaceutical compositions comprising the insulin dimers disclosed herein and a pharmaceutically acceptable carrier for use in the treatment of a patient suffering from high blood glucose levels. In one embodiment, the patient to be treated with the insulin dimer disclosed herein is a domesticated animal, and in another embodiment, the patient to be treated is a human.

One method of treating hyperglycemia according to the present disclosure comprises the steps of administering the insulin dimers disclosed herein to a patient via any standard route of administration, including parenteral, including intravenous, intraperitoneal, subcutaneous or intramuscular, intrathecal, transdermal, rectal, oral, nasal or by inhalation. In one embodiment, the composition is administered subcutaneously or intramuscularly. In one embodiment, the composition is administered parenterally and the insulin polypeptide or prodrug derivative thereof is prefilled into a syringe.

The insulin dimers disclosed herein may be administered alone or in combination with other antidiabetic agents. Anti-diabetic agents known in the art or under development include native insulin, native glucagon and their functional analogs, sulfonylureas such as tolbutamide (Orinase), acetohexamide (Dymelor), tolazamide (Tolinase), chlorpropamide (Diabinese), glipizide (Glucotrol), glyburide (Diabeta, Micronase, Glynase), glimepiride (Amaryl), or gliclazide (Diamicron); meglitinides such as repaglinide (Prandin) or nateglinide (Starlix); biguanides such as metformin (Glucophage) or phenformin; thiazolidinediones such as rosiglitazone (Avandia), pioglitazone (Actos) or troglitazone (Rezulin), or other PPARy inhibitors; alpha-glucosidase inhibitors that inhibit the breakdown of carbohydrates such as miglitol (Glyset), acarbose (Precose/Glucobay); exenatide (Byetta) or pramlintide; dipeptidyl peptidase-4 (DPP-4) inhibitors such as vildagliptin or sitagliptin; SGLT inhibitors (sodium dependent glucose transporter 1); or FBF-ase (fructose-1,6 bisphosphatase) inhibitors.

Pharmaceutical compositions containing the insulin dimers disclosed herein may be formulated and administered to patients using standard pharmaceutically acceptable carriers and routes of administration known to those skilled in the art. Accordingly, the present disclosure also encompasses pharmaceutical compositions comprising one or more of the insulin dimers disclosed herein, or a pharmaceutically acceptable salt thereof, in combination with a pharmaceutically acceptable carrier. For example, pharmaceutical compositions containing insulin dimers disclosed herein may optionally contain zinc ions, preservatives (eg, phenol, cresol, parabens), isotonic agents (eg, mannitol, sorbitol, lactose, dextrose, trehalose, sodium chloride, glycerin), buffer substances, salts, acids and alkalis, as well as other excipients. These substances may in each case be present singly or, alternatively, as mixtures. Glycerin, dextrose, lactose, sorbitol and mannitol are typically present in a pharmaceutical preparation at a concentration of 100-250 mM, NaCl at a concentration of up to 150 mM. Buffers such as, for example, phosphate, acetate, citrate, arginine, glycylglycine or TRIS (i.e. 2-amino-2-hydroxymethyl-1,3-propanediol) buffer and the corresponding salts are present at a concentration of 5-250 mm, usually from about 10-100 mm. Other excipients may also be salts or arginine.

In one embodiment, the pharmaceutical composition contains insulin dimer at a concentration of 1 mg/ml at a pH of from about 4.0 to about 7.0 in a phosphate buffer system. Pharmaceutical compositions may contain an insulin dimer as the sole pharmaceutically active ingredient, or the insulin dimer may be combined with one or more additional active agents.

All therapeutic methods, pharmaceutical compositions, kits, and other similar embodiments described herein provide that insulin dimers include all pharmaceutically acceptable salts thereof.

In one embodiment, a kit is provided with a device for administering an insulin dimer composition to a patient. The kit may further include a plurality of containers, such as vials, test tubes, bottles, and the like. Preferably, the kits will also include instructions for use. According to one embodiment, the device of the kit is an aerosol dispenser, wherein the composition is loaded into the aerosol device. In another embodiment, the kit contains a syringe and a needle, and in one embodiment, the insulin dimer composition is prefilled into a syringe.

Compounds of the present invention can be obtained using standard methods of synthesis, recombinant DNA methods or any other methods for obtaining peptides and fusion proteins. Although some non-natural amino acids cannot be expressed using standard recombinant DNA methods, methods for their preparation are known in the art. Compounds of the present invention that include non-peptide moieties may be synthesized by standard organic chemistry reactions in addition to standard peptide chemistry reactions, when applicable.

The following examples are intended to facilitate further understanding of the present invention.

Examples

General procedures.

All chemicals were purchased from commercial sources unless otherwise noted. Reactions were generally carried out at ambient temperature or at room temperature unless otherwise noted. Moisture or air sensitive reactions were carried out under nitrogen or argon using anhydrous solvents and reagents. The progress of the reactions was monitored by analytical thin layer chromatography (TLC) and ultra high performance liquid chromatography-mass spectrometry (UHPLC-MS). TLC was performed on E. Merck TLC plates pre-coated with silica gel 60F-254, layer thickness 0.25 mm. The plates were visualized using UV 254 nm and/or exposed to cerium ammonium molybdate (CAM) or p-anisaldehyde staining solutions followed by charring. Ultra high performance liquid chromatography (UHPLC) was performed on a Waters Acquity™ UPLC® system.

UHPLC-MS Method A: Waters Acquity™ UPLC® WEN C18 1.7 µm 1.0x50 mm column with a 10:90-95:5 v/v gradient. CH ₃ CN/H ₂ O+0.05 vol% TFA for 2.0 min; flow rate 0.3 ml/min, UV wavelength 215 nm; UHPLC-MS; method B: Waters Acquity™ UPLC® WEN C18 1.7 µm 2.1x100 mm column with a 60:40-100:0 v/v gradient. CH ₃ CN/H ₂ O+0.05 vol.% TFA for 4.0 min and 100:0-95:5 v/v. CH ₃ CN/H ₂ O+0.05 vol% TFA for 40 sec; flow rate 0.3 ml/min, UV wavelength 200-300 nm; UHPLC-MS; Method C: Waters Acquity™ UPLC® WEN C18 1.7 µm 2.1x100 mm column with a 20:80-90:10 v/v gradient. CH ₃ CN/H ₂ O+0.05 vol.% TFA for 4.0 min and 90:10-95:5 v/v. CH ₃ CN/H ₂ O+0.05 vol% TFA for 0.5 min; flow rate 0.3 ml/min, UV wavelength 200-300 nm; UHPLC-MS; method D: Waters Acquity™ UPLC® WEN C8 1.7 µm 2.1x100 mm column with a gradient of 10:90-55:45 v/v. CH ₃ CN/H ₂ O+0.05 vol.% TFA for 4.0 min and 55:45-95:5 v/v. CH ₃ CN/H2O+0.05 vol% TFA for 40 sec; flow rate 0.3 ml/min, UV wavelength 200-300 nm; UHPLC-MS; Method E: Waters Acquity™ UPLC® BEH300 C4 1.7 µm 2.1x100 mm column with 10:90-50:50 v/v gradient. CH ₃ CN/H ₂ O+0.05 vol.% TFA for 4.3 min and 50:50-70:30 v/v. CH ₃ Cn/H ₂ O+0.05 vol% TFA for 0.5 min; flow rate 0.3 ml/min, UV wavelength 200-300 nm; UHPLC-MS; Method F: Waters Acquity™ UPLC® WEN C8 1.7 µm 2.1x100 mm column with a gradient of 20:80-72.5:27.5 v/v CH ₃ CN/H ₂ O+0.05 vol% TFA for 4.3 min and 72.5:27.5-95:5 v/v. CH ₃ CN/H ₂ O+0.05 vol% TFA for 0.5 min; flow rate 0.3 ml/min, UV wavelength 200-300 nm and UHPLC-MS; Method G: Waters Acquity™ UPLC® WEN C8 1.7 µm 2.1x100 mm column with a 20:80-90:10 v/v gradient. CH ₃ CN/H ₂ O+0.1 vol.% TFA for 4.0 min and 90:10-95:5 v/v. CH ₃ CN/H ₂ O+0.1 vol% TFA for 0.4 min; flow rate 0.3 ml/min, UV wavelength 200-300 nm.

Mass analysis was performed on a Waters SQ Detector with electrospray ionization in positive ion detection mode, with a mass-to-charge ratio scan range of 170-900, or on a Waters Micromass® LCT Premier™ XE with electrospray ionization in positive ion detection mode, with a scan range of the ratio of mass to charge was 300-2000. The identification of the resulting insulin conjugates or IRPA was confirmed by comparing the theoretical molecular weight with the experimental value, which was measured using UPLC-MS. In order to determine, in particular, the positions of the bonds, the insulin dimers were subjected to DTT treatment (for the a/b chain) or Glu-C cleavage (with or without reduction and alkylation), and then the resulting peptides were analyzed by LC-MS. Based on the measured masses, the positions of the bonds were determined.

Flash chromatography was performed using either Biotage Flash Chromatography

- 66 041034 (Dyax Corp.), or CombiFlash® Rf (Teledyne Isco). Chromatography with normal phases was carried out on silica gel (20-70 μm, pore size 60 A) in pre-filled cartridges of the specified size. Ion exchange chromatography was carried out on a silica gel-based material bound with a hydrophilic anionic poly(2-sulfoethylaspartamide) coating (PolySULFOETHYL A column, PolyLC Inc., 250x21 mm, 5 μm, pore size 1000 A). Reverse phase chromatography was performed on C18 bound silica gel (20-60 µm, pore size 60-100 A) in prefilled cartridges of the specified size. Preparative HPLC was performed on a Gilson 333334 binary system using a Waters DELTA PAK C4 15 µm, 300 A, 50x250 mm column or KROMASIL® C8 10 µm, 100 A, 50x250 mm column, flow rate 85 ml/min, with the indicated gradient. The solutions were concentrated on a rotary evaporator under reduced pressure or freeze dried on a VirTis Freezemobile Freeze Dryer (SP Scientific).

Abbreviations: acetonitrile (AcCN), aq. dimethylaminopropyl)-N'-ethylcarbodiimide (EDC), gram(s) (g), 1-hydroxybenzotriazole hydrate (HOBt), h(s) (h or h), mass spectrum (ms or MS), microgram(s) (µg ), microliter(s) (µl), micromoles (µmoles), milligram(s) (mg), milliliter(s) (ml), millimoles (mmol), minute(s) (min), retention time (R _t ) , room temperature (rt), saturated (saturated or saturated), saturated aq. sodium chloride solution (brine), triethylamine (TEA), trifluoroacetic acid (TFA), and N,N,N',N'-tetramethyl-O-(N-succinimidyl)uronium tetrafluoroborate (TSTU).

The term RHI refers to recombinant human insulin and is used to indicate that the insulin has an amino acid sequence that is characteristic of native wild-type human insulin. As used herein in the tables, this term indicates that the amino acid sequence of the insulin contained in the dimer is that of native wild-type human insulin. EXAMPLE 1

The synthesis of 2,5-dioxopyrrolidin-1-yl-6-((6-((2,5-dioxopyrrolidin-1-yl)oxy)-6oxyhexyl)amino)-6-oxohexanoate (linker 1; C6+NC6) is described.

Step 1. Benzyl 6-((6-(benzyloxy)-6-hydroxyhexyl)amino)-6-oxohexanoate.

To a mixture of adipic acid monobenzyl ester (600 mg, 2.54 mmol) and 6(benzyloxy)-6-oxohexane-1-aminium 4-methylbenzenesulfonate (1.0 g, 2.54 mmol) in DMF (12.71 ml ) HOBt (584 mg, 3.81 mmol), Hunig's base (888 μl, 5.08 mmol) and EDC (731 mg, 3.81 mmol) were added. After stirring overnight, the reaction mixture was partitioned between sat. _NaHCO3 and EtOAc. The organic phase was separated, washed with 1.0 M HCl and brine, dried over Na ₂ SO ₄ and concentrated to give the title compound as a semi-solid and used in the next step without further purification. UPLC-MS method A: Rt=1.26 min, m/z=440.3 [M+1].

Step 2. 6-((5-Carboxypentyl)amino)-6-oxohexanoic acid.

A suspension of the product of step 1 (1.08 g, 2.457 mmol) and Pearlman's catalyst (20 wt.% on carbon, 173 mg, 0.246 mmol) in MeOH (50 ml) was stirred at 50 psi H ₂ overnight. The catalyst was filtered off and the filtrate was subjected to C8 reverse phase chromatography (Kromasil, C8 10 μm 100 A, 250x50 mm; solvent A=water/0.05% TFA, solvent B=AcCN/0.05% TFA), flow rate \u003d 85 ml / min, gradient B to A 5-30% in 30 min. UPLC-MS method A: Rt=0.40 min, m/z=260.15 [M+1].

Step 3. 2,5-dioxopyrrolidin-1-yl-6-((6-((2,5-dioxopyrrolidin-1-yl)oxy)-6-hydroxyhexyl)amino)6-oxohexanoate.

To a solution of the product of step 2 (50 mg, 0.193 mmol) in DMF (964 μl) was added TSTU (116 mg, 0.386 mmol). After cooling to 0°C, triethylamine (53.8 μl, 0.386 mmol) was added to the mixture. After stirring for 45 min, the formation of the desired compound was observed: UPLC-MS method A: Rt=0.71 min, m/z=453.4 [M+1]. The resulting 2,5-dioxopyrrolidin-1-yl-6-((6-((2,5-dioxopyrrolidin-1yl)oxy)-6-hydroxyhexyl)amino)-6-oxohexanoate was used as a 0.2 M solution in DMF without cleaning.

Example 2

The synthesis of bis(2,5-dioxopyrrolidin-1-yl)-6,6'-(adipoylbis(azanediyl))dihexanoate (linker 2; C6N+C6+NC6) is described.

Step 1. Dibenzyl-6,6'-(adipoylbis(azanediyl))dihexanoate.

To a solution of 6-(benzyloxy)-6-oxohexane-1-aminium 4-methylbenzenesulfonate (2.693 g, 6.84 mmol) and adipic acid (500 mg, 3.42 mmol) in DMF (17.1 ml) was added Hunig's base (1.793 ml, 10.26

- 67041034 mmol), HOBt (1.572 g, 10.26 mmol) and EDC (1.968 g, 10.26 mmol). After stirring overnight, the reaction mixture was poured into water (500 ml) and stirred for 30 minutes. The title compound was collected by filtration as a solid and dried by suction. UPLC-MS method A: Rt=1.23 min, m/z=553.5 [M+1].

Step 2. Bis(2,5-dioxopyrrolidin-1-yl)-6,6'-(adipoylbis(azanediyl))dihexanoate.

The title compound was prepared using a procedure similar to that described in Example 1, replacing benzyl-6-((6-(benzyloxy)-6-hydroxyhexyl)amino)-6-oxohexanoate with dibenzyl-6,6'(adipoylbis(azanediyl) )dihexanoate in step 2. UHPLC-MS method A: Rt=0.74 min, m/z=567.4 [M+1].

Example 3

The synthesis of (2S,2'S)-2,2'-(oxtanedioylbis(azanediyl))bis(5-((2,5-dioxopyrrolidin-1-yl)oxy)5-oxopentanoic acid) (linker 3; gamma-Glu- suberic-gamma-Glu).

Step 1. (S)-5-(benzyloxy)-4-(8-(((S)-1-(benzyloxy)-4-carboxy-1-oxobutan-2-yl)αmino)-8oxooctanamido)-5- oxopentanoic acid.

To a solution of H-GLU-OBZL (1.00 g, 4.21 mmol) in DMF (10.5 ml) was added triethylamine (5.875 ml, 42.1 mmol) followed by disuccinimidyl suberate (776 mg, 2.107 mmol). After stirring for 1 hour, the reaction mixture was concentrated and the resulting residue was purified on a C18 column (ISCO, 44 g), flow=37 ml/min; AcCN gradient in water with 0.05% TFA: 2%-20% over 20 min followed by retention. After lyophilization, the intermediate bis-carboxylic acid was obtained. UPLC-MS method B: Rt=2.66 min, m/z=613.3 [M+1].

Step 2 Bis-N-hydroxysuccinimide ester (S)-5-(benzyloxy)-4-(8-(((S)-1(benzyloxy)-4-carboxy-1-oxobutan-2-yl)amino) -8-oxooctanamido)-5-oxopentanoic acid.

To a suspension of the product of step 1 (455 mg, 0.743 mmol) in acetonitrile (7.4 ml) was added TSTU (492 mg, 1.634 mmol) as a solid followed by triethylamine (228 μl, 1.634 mmol), at which point the suspension dissolved . The reaction mixture was stirred for 1.5 h and concentrated on a rotary evaporator at room temperature. The product was purified by C-8 reverse phase chromatography (Kromasil column, C8 10 µm, 100 A, size 250x50 mm; solvent A=water/0.05% TFA, solvent B=AcCN/0.05% TFA) , flow=85 ml/min, gradient B to A 10-80% in 30 min. After lyophilization of the fractions, an ester with bis-NHS was obtained. UPLC-MS method B: Rt=2.77 min, m/z=807 [M+1].

Step 3. (2S,2'S)-2,2'-(oxtanedioylbis(αzanediyl))bis(5-((2,5-dioxon-pyrrolidin-1-yl)oxy)-5oxopentanoic acid).

The product of step 2 (250 mg, 0.310 mmol) was hydrogenated using palladium on carbon (66.0 mg, 0.031 mmol) as catalyst and acetone containing 0.1% TFA as solvent (6.2 ml) at 1 atm hydrogen overnight. The catalyst was filtered off and the filtrate was concentrated to give the title compound. Was pumped to high vacuum during the night. UPLC-MS method C: Rt=3.61 min, m/z=627.3 [M+1].

Example 4

General method A: synthesis of K ^6,B29 insulin conjugates (analogues).

In an appropriately sized container, insulin or insulin analog was dissolved with gentle agitation at room temperature in a mixed solvent: 2:3 v/v. 0.1 M Na ₂ CO ₃ :AcCN. After the mixture became transparent, the pH was adjusted to a value of 10.5-10.8 using an alkaline solution, for example 0.1N. NaOH. In a separate vial, the activated ester intermediate (binding moiety) was dissolved in an organic solvent such as DMSO at room temperature. Aliquots of the activated ester (linker) solution were added over a period of time to the solution containing insulin until the UHPLC chromatogram indicated that most of the unmodified insulin had reacted and that a significant portion of the reaction mixture had converted to B29-conjugated insulin. The reaction was stopped by adding an amine nucleophile, eg 2-aminoethanol. The reaction solution was stirred at room temperature for 30 minutes. The resulting solution was carefully diluted with cold H ₂ O (20x) at 0°C, and brought its pH to a final pH of 2.5 using 1N. HCl (and 0.1N NaOH if necessary). The solution was first concentrated by ultrafiltration or with a tangential flow filtration (TFF) system or using Amicon Ultra-15 Centrifugal Units with a 1K, 3K or 10K MWCO membrane. The concentrated solution was usually first subjected to ion exchange chromatography (PolySULFOETHYL A column, PolyLC Inc., 250x21 mm, 5 µm, 1000 A; buffer A: 0.1% (v/v) H3PO4 25 ¹ % AcCN; buffer B: 0. 1% (v/v) H3PO4/25% AcCN/0.5 M NaCl). Fractions containing the desired purity B29 conjugate were pooled and concentrated using a TFF or Amicon Ultra-15 system. The resulting solution was then further purified using

- 68 041034 reverse phase HPLC (Waters C4 250x50 mm, 10 µm, 1000 A column or Kromasil column

C8 250x50 mm, 10 µm, 100 A; buffer A: 0.05-0.1% TFA in water; buffer B: 0.05-0.1% TFA in AcCN). Fractions containing the title conjugate were pooled and freeze dried or buffer exchanged using the TFF system and/or Amicon Ultra-15 to give the title product.

Example 5

The synthesis of N ^6,B29 -5-azidopentanoyl-desB30-insulin (A:Y19A) (analogue 1) is described.

In a 20 ml scintillation vial, desB30 A:Y19A insulin (112 mg, 0.020 mmol) was dissolved with gentle agitation at room temperature in a mixed solvent (2 ml, 2:3 v/v 0.1 M Na ₂ CO ₃ : AcCN ). After the mixture became transparent, the pH was adjusted to a value of 10.5-10.8 using an alkaline solution, for example 0.1N. NaOH. In a separate 8 ml scintillation vial, 2,5-dioxopyrrolidin-1-yl 5-azidopentanoate (linker 5; see example 6) (4.79 mg, 0.020 mmol) was dissolved in DMSO (500 μl) at room temperature. Aliquots of the activated ester solution were added over a period of time to the solution containing insulin until the UPLC chromatogram indicated that most of the unmodified insulin had reacted and that a significant portion of the reaction mixture had converted to B29 conjugated insulin. The reaction was stopped by adding an amine nucleophile, eg 2-aminoethanol. The reaction solution was stirred at room temperature for 30 minutes. The resulting solution was carefully diluted with cold H ₂ O (20x) at 0°C and its pH was adjusted to a final pH of 2.5 using 1.0N. HCl (and 0.1N NaOH if necessary). The solution was first concentrated by ultrafiltration using Amicon Ultra-15 Centrifugal Units with a 3K or 10K MWCO membrane. The concentrated solution was subjected to reverse phase HPLC (KROMASIL C8 column 250x50 mm, 10 µm, 100 A, 25-35% buffer B in buffer A for 20 min; buffer A: 0.05% TFA in water; buffer B: 0. 05% TFA in AcCN). Fractions containing analog 1 were pooled and then freeze dried. UPLC-MS method D: Rt=3.91 min, m/z=1435.86 [(M+4)/4].

Example 6

N ^6,B29 -acylated analogue 2, analogue 3 and analogue 4.

The RHIs were generated for use in the construction of dimers using clickhimy, and they were generated using General Method A or a procedure similar to that described in Example 4 but replaced with recombinant human insulin and

O (2,5-dioxopyrrolidin-1-ylpent-4-inoate; linker 4);

.o /Ύ II V o (2,5-dioxopyrrolidin-1-yl-azidopentanoate; linker 5); or

(perfluorophenyl-1-(bicyclo[6.1.0]non-4-yn-9-yl)-3-oxo-2,7,10,13,16-pentaoxa-4-azanonedecane-19oate) (linker 6) for creating analogue 2, analogue 3 or analogue 4, respectively. The analogs were characterized using UPLC-MS method D, with the exception of analog 5, which was characterized using UPLC-MS method F.

Example 7

The synthesis of N ^2,1A ,N ^2,1B -bis(carbamoyl)-human insulin (analogue 5) is described.

- 69 041034

To a suspension of RHI (1 g, 0.172 mmol) in water (50 ml) was added a solution of dibasic potassium phosphate, (0.249 g, 1.429 mmol) in water (5.0 ml). After stirring at room temperature for 30 minutes, potassium cyanate (0.279 g, 3.44 mmol) was added to the resulting mixture. The reaction mixture was allowed to stir for 16 hours. To stop the reaction, unreacted potassium cyanate was removed with TFF using an MWCO 3K diafiltration device and the product was isolated as a solid by lyophilization. The product contained approximately 10-35% A1/B1/B29-tris-urea-KSh, which optionally could be removed by C8 reverse phase chromatography (KROMASIL column, C8 10 µm 100 A, 250x50 mm; solvent A =water/0.05% TFA, solvent B=AcCN/0.05% TFA), flow rate=85 ml/min, gradient B to A 2634% over 30 min). UPLC-MS method D: Rt=4.29 min, m/z=1474.6 (z=4). N-terminal substituent o

A has the structure (carbamoyl), with the wavy line representing the bond between the substituent and the N2 nitrogen of the N-terminal amino acid.

Example 8

The synthesis of N ² ^ ^1A ,N ² ^ ^1B -bis(carbamoyl)-desB30-human insulin (analogue 6) is described.

The title compound was prepared using a procedure similar to that described in Example 7, replacing RHI with desB30 insulin. UPLC-MS method D: Rt=4.10 min, m/z=1448.9 (z=4). No

And the terminal substituent has the structure (carbamoyl), with the wavy line representing the bond between the substituent and the N2 nitrogen of the N-terminal amino acid.

Example 9

The synthesis of N ² ^ ^1A ,N ² ^ ^1B -bis(carbamoyl)-insulin lispro (analogue 7) is described.

The title compound was prepared using a procedure similar to that described in Example 7, replacing RHI with insulin lispro. UPLC-MS method D: Rt=4.07 min, m/z=1473.6 (z=4). No

The Λα terminal substituent has the structure ^ngs (carbamoyl), with the wavy line representing the bond between the ^substituent and the N2 nitrogen of the N-terminal amino acid.

Example 10

The synthesis of N ² ^ ^1A -acetyl-human insulin (analogue 8) is described.

To a solution of RHI (400 mg, 0.069 mmol) in DMSO (4.6 ml) was added dropwise a solution of 2,5-dioxopyrrolidin-1-yl acetate o/η o f o (10.82 mg, 0.069 mmol) in 100 μl DMSO. After stirring for 3 h, the reaction mixture was diluted with water (95 ml), acidified until the pH was approximately 3, and then diafiltered through Amicon Ultra-15 Centrifugal Units with a 3 or 10K MWCO membrane to remove most of the DMSO. The resulting solution was first subjected to ion exchange chromatography (PolySULFOETHYL A column, PolyLC Inc., 250x21 mm, 5 μm, 1000 A, flow rate 15 ml/min; buffer A: 0.1% (v/v) H ₃ PO4/25 % AcCN; buffer B: 0.1% (v/v) H ₃ PO4/25% AcCN/0.5 M NaCl) using a gradient of 10-40% buffer B in buffer A over 24 min. Fractions containing the desired ^N2 ^ ^1A -acetylRHI were pooled and concentrated and then subjected to reverse phase chromatography on (KROMASIL, C8 10 µm 100 A, 250x50 mm; eluent A=water/0.05% TFA, eluent B=AcCN /0.05% TFA, gradient 26-30% B to A). The position of the modification was confirmed using DTT analysis. UPLC-MS method D: Rt=3.5 min and m/z=1463.5 (z=4). The N-terminal substituent has an o structure (acetyl), with the wavy line representing the bond between the substituent and the N2 nitrogen of the N-terminal amino acid.

Example 11.

The synthesis of N ² ^ ^1A ,N ² ^ ^1B -bis(carbamoyl)-N ⁶ ^ ^29B -acylated RHI is described.

Analog 5 conjugated to either 2,5-dioxopyrrolidin-1-yl-azidopentanoate (linker 5) to create analog 9, or 2,5-dioxopyrrolidin-1-ylpent-4-inoate (linker 4) to create analog 10, was prepared using General Method A or a procedure similar to that described in Example 4.

Example 12.

The following ^N6 ^ ^29B -acylated RHI analogs (analog 11, analog 12, and analog 13) were prepared for use in the construction of dimers using click chemistry. Analogues were obtained using General method A or a procedure similar to that described in example 4, but with replacement for recombinant human insulin (RHI) and the appropriate binding fragment selected from

- 70 041034

GP

8), or (linker 9) to create analog 11, analog 12 or analog 13, respectively. The analogs were characterized using UPLC-MS method D, with the exception of analog 12, which was characterized using UPLC-MS method F.

Example 13

General Method B: Synthesis of N ^6,29A ,N ^6,29B -insulin dimers under organic base conditions.

In an appropriately sized container, the insulin or insulin analog is suspended at room temperature in an organic solvent or mixed aqueous (aqueous)/organic solvents, eg DMSO, in the presence of a base, eg TEA. The mixture is left to stir gently until the insulin is completely dissolved. To the resulting solution is added an activated ester intermediate (linker) in a solution of organic solvents such as DMSO or DMF. After that, the chromatogram of UHPLC shows that a significant part of the reaction mixture turned into dimer N ^6,29B ,N ^6,29B -insulins (or dimer N ^6,28B ,N ^6,28B -insulin lispro). The reaction mixture can be directly purified by reverse phase HPLC (Waters C4 250x50 mm, 10 µm, 1000 A column or KROMASIL C8 250x50 mm, 10 µm, 100 A column; buffer A: 0.05-0.1% TFA in deionized water; buffer B: 0.05-0.1% TFA in AcCN), or the reaction can be stopped by careful dilution with cold acidic H ₂ O (20x, pH approx. 3.0) at 0°C, and its pH adjusted to a final pH of 2.5 using 1N. HCl (and 0.1N NaOH if necessary). The solution can first be concentrated using ultrafiltration or a tangential flow filtration (TFF) system, or using Amicon Ultra-15 Centrifugal Units with a 1K, 3K or 10K MWCO membrane. The concentrated solution is usually first subjected to ion exchange chromatography (PolySULFOETHYL A column, PolyLC Inc., 250x21 mm, 5 µm, 1000 A; buffer A: 0.1% (v/v) H3PO ₄ /25% AcCN; buffer B: 0 .1% (v/v) H3PO ₄ /25% AcCN/0.5 M NaCl). Fractions containing B29 conjugate of the desired purity are pooled and concentrated using a TFF or Amicon Ultra-15 system. The concentrated solution is then purified by reverse phase HPLC (Waters C4 250x50 mm, 10 µm, 1000 A column or KROMASIL C8 250x50 mm, 10 µm, 100 A column; Buffer A: 0.05-0.1% TFA in deionized water; buffer B: 0.05-0.1% TFA in AcCN). Fractions containing the desired insulin dimer are pooled and freeze-dried or buffer exchanged using the TFF system and/or Amicon Ultra-15 to give ^N6.29B , ^N6.29B insulin dimers.

Example 14

General Method C: Synthesis of N ^6,29B ,N ^6,29B -insulin dimers under aqueous base conditions.

In an appropriately sized container, dissolve the insulin or insulin analogue with gentle agitation at room temperature in a mixed solvent: 2:3 v/v. 0.1 M Na ₂ CO ₃ :AcCN. After the mixture becomes clear, the pH is adjusted to a value of 10.5-10.8 using an alkaline solution, for example 0.1N. NaOH. In a separate vial, the activated ester intermediate (linker) is dissolved in an organic solvent such as DMSO at room temperature. Aliquots of the activated ester solution are added over a period of time to the solution containing insulin until the HPLC chromatogram shows that most of the unmodified insulin has reacted and that a significant part of the reaction mixture has converted to the dimer N ^6,B29 ,N ^{6 ,B29} - insulins (or dimer n ^6,28b ,N ^6,28b - insulin lispro). OS reaction

- 71 041034 tanavlyayut by adding an amine nucleophile, such as 2-aminoethanol. The reaction solution was stirred at rt for 30 minutes. The resulting solution is carefully diluted with cold H ₂ O (20x) at 0°C, and its pH is adjusted to a final pH of 2.5 using 1N. HCl (and 0.1N NaOH if necessary). The solution is first concentrated by ultrafiltration or by a tangential flow filtration (TFF) system or by using Amicon Ultra-15 Centrifugal Units with a 1K, 3K or 10K MWCO membrane. The concentrated solution is usually first subjected to ion exchange chromatography (PolySULFOETHYL A column, PolyLC Inc., 250x21 mm, 5 µm, 1000 A; buffer A: 0.1% (v/v) H ₃ PO4/25% AcCN; buffer B: 0.1% (v/v) H ₃ PO4/25% AcCN/0.5 M NaCl). Fractions containing B29 conjugate of the desired purity are pooled and concentrated using a TFF or Amicon Ultra-15 system. The resulting solution is then further purified by reverse phase HPLC (Waters C4 250x50 mm, 10 µm, 1000 A column or KROMASIL C8 250x50 mm, 10 µm, 100 A column; Buffer A: 0.05-0.1% TFA in water; buffer B: 0.05-0.1% TFA in AcCN). Fractions containing the title insulin dimer are pooled and freeze dried or buffer exchanged using a TFF system and/or Amicon Ultra-15 to give ^N6,29B , ^N6,29B -insulin dimers.

Example 15

This example illustrates the synthesis of N ^6,B29 ,N ^6,B29 -(2,2'-(ethane-1,2-diylbis(oxy))diacetyl)bis[human insulin] (dimer 1).

Dissolve RHI (2.6 g, 0.448 mmol) in a mixture of Na ₂ CO ₃ (0.1 M) (15.8 ml) and AcCN (10.5 ml) and add 0.895 ml (0.179 mmol) of a 0.2 M solution of bis (2,5-dioxopyrrolidin-1-yl)-2,2'-(ethane-1,2diylbis(oxy))diacetate (linker 8) in DMF. The reaction mixture was stirred for 30 min, an additional portion of 0.895 ml (0.179 mmol) of a 0.2 M solution of bis(2,5-dioxopyrrolidin-1-yl)2,2'-(ethane-1,2-diylbis(oxy)) was added diacetate in DMF and stirred the reaction mixture for an additional 30 min. Pour the reaction mixture into 60 ml of 20% AcCN/0.1% TFA/water, adjust the pH to 2.5 and diafilter using an Amicon Ultra-15 with a 10K MWCO membrane for concentration until the resulting volume is approximately 10 ml . The resulting solution was subjected to ion exchange chromatography (PolySULFOETHYL A column, 250x21 mm, 5 μm, 1000 A, gradient 10-80% buffer B in buffer A over 30 min; buffer A: 0.1% (v/v) H ₃ PO4/25% AcCN Buffer B: 0.1% (v/v) H ₃ PO ₄ /25% AcCN/0.5M NaCl). Fractions containing the title compound were pooled and concentrated. The resulting solution was then subjected to reverse phase chromatography (KROMASIL C8 column 250x50 mm, 10 μm, 100 A; gradient 27-35% of AcCN with 0.05% TFA in water with 0.05% TFA). UPLC-MS method E: Rt=2.75 min, m/z=1960.4 (z=6), 1680.4 (z=7).______________

No. dimer a Dimer structure showing the binding fragment between lysine residues B2 9 and B2 9' type of insulin; N ends Rt (min) (M+6)/6 or (M+7)/7 insulin 1 ABOUT ABOUT RHI; A1,A1', B1,B1'=H 2.75 1960.4 The wavy line indicates the bond with the epsilon-amino group B29 Lys and B29' Lys, respectively.

Example 16

This example illustrates the synthesis of N ^2,1A ,N ^2,1A ,N ^2,1B ,N ^2,1B -tetrakis(carbamoyl)-N ^6,B29 ,N ^6,B29 (hexanedioyl)bis[human insulin] (dimer 2) .

Dissolve N ^2,1A ,N ^2,1B -bis(carbamoyl)-RHI (150 mg, 0.025 mmol) in DMSO (1 mL) and add triethylamine (0.106 mL, 0.764 mmol) followed by di(N -succinimidyl) adipate (linker 12) (4.33 mg, 0.013 mmol) dissolved in 100 µl DMSO. Stirred for 1 hour, and poured the reaction mixture into 20 ml of water. Acidified to pH=2 and diafiltered using 10K Amicon Ultra 15. The product was purified by ion exchange chromatography using a gradient of 10-40% solvent B in solvent A over 24 min and repurified by reverse phase chromatography on phase C-8 , gradient B to A 26-36% over 30 min. UPLC-MS method E: Rt=3.75 min, m/z=1983.9, (z=6).

- 72 041034

No. dimer Dimer structure showing the binding fragment between lysine residues B2 9 and B2 9' type of insulin; N-termini of insulin Rt (min) (M+6) /6 or (M+7)/7 2 ⁰ about RHI; ΑΙ, ΑΓ, Bl, Bl'=carbamoyl 3.75 1983.9 The wavy line indicates the bond with the epsilon-amino group B29 Lys and B29' Lys, respectively.

Examples 17.

Tab. 3 shows dimers that were prepared using the appropriate intermediates (linkers) according to either General Method B or General Method C as indicated using RHI, DesB30 RHI, insulin lispro, insulin aspart, insulin glargine, or the corresponding analogue. For example, for dimers with carbamoylated N-terminals analog 5 or analog 6 (DesB30) was used; for dimers with A1 acetylated N-terminals analog 8 was used. [(M+6)/6], (or semi-charged, ie [(M+7)/7]) particles of the parent compound at a certain retention time (Rt). Insulin and insulin molecules linked together by a binding moiety are the same for each of the dimers shown in Table 1. 3.

showing a binding fragment between lysine residues B29 and B29' of insulin; N ends of insulin RHI;

ΑΙ,ΒΙ,ΑΙ',B1'=carbamoyl

RHI;

ΑΙ, ΒΙ, ΑΙ', B1

- 73 041034

- 74 041034

13 0 0 RHI; Al, AG=acetyl , Β1,ΒΓ=Η IN 3.83 1973.8 14 0 0 ₄ A^ RHI; Al, Bl, Al', B1 '=carbamoyl IN 4.10 1740.5 15 Ο ο RHI; Al,Bl,Al',B1 '=H 3.97 1716.0 16 0 \A..^ ₀ ^..^ 0 RHI; Al, Bl, Al', Bl '=carbamoyl IN 3.80 1752.4 17 ο Ο RHI; Al, Bl, Al', B1 '=carbamoyl IN 4.13 1778.7 18 Ο insulin lispro; Al,Bl,Al', Bl'=H WITH 3.74 1960.54 19 0 /γχ^-νχ-χΑγ 0 insulin aspart; Al,Bl,Al', Bl'=H WITH 3.74 1966.13 20 ο ο RHI desB30; Al,Bl,Al',B1 '=H WITH 4.57 1926.04 21 ϊ ^Η vV-^N^^ 0 RHI; Al, Bl, Al', B1 '=carbamoyl IN 4.48 1716.7

- 75 041034

- 76 041034

- 77 041034

- 78 041034

- 79 041034

83 ABOUT O insulin aspart; Al,Bl,Al' ,B1 '=carbamoyl IN 3.50 1994.39 85 L '-.--'-Cl ----0 ~~ With C RHI; Al, Bl, Al', Bl '=H D 3.46 1829.31 53 soo o 4 Insulin glargine; Al, Bl, Al', Bl '=H D 3.37 1788.87 87 To ς C ο insulin glargine; Al,Bl,Al',B1 '=H D 3.43 1902.16 The wavy line indicates the bond with the epsilon-amino group B29 Lys and B29' Lys, respectively.

Example 18.

General Method D: Synthesis of N ^6,29B ,N ^6,29B -insulin dimers using Cu ²⁺ catalyzed click chemistry.

In an appropriately sized container, the appropriate acetylene-containing insulin intermediate (analogue) was dissolved with gentle stirring at room temperature in a mixed solvent of DMSO and aq. triethylammonium acetate buffer (pH 7.0, final concentration 0.2 mM). In another suitably sized container, the appropriate azido-containing insulin intermediate (analogue) was dissolved with gentle agitation at rt in a mixed solvent of DMSO and water. Both solutions were combined, thoroughly mixed, degassed by gentle sparging with N ₂ . To the resulting solution was added freshly prepared sodium ascorbate or ascorbic acid solution (final concentration is 0.5 mM) and, after thorough mixing, a solution of 10 mM CUSO4 and tris[(1-benzyl-1H-1,2,3-triazole-4- yl)methyl]amine (i.e. TBTA ligand) in 55% DMSO. After degassing by gentle sparging with N ₂ and thorough mixing, the mixture was stored at rt with occasional stirring overnight. The reaction mixture was carefully diluted with a mixed solvent (v/v 7:3 AcCN/water with 0.05% TFA) at 0°C and the pH was adjusted to 2.50 using 0.1, 1.0N. HCl (and 0.1N NaOH if necessary). The solution was first concentrated by ultrafiltration or with a tangential flow filtration (TFF) system or using Amicon Ultra-15 Centrifugal Units with a 1K membrane,

- 80 041034

3K or 10K MWCO. The concentrated solution was usually first subjected to ion exchange chromatography (PolySULFOETHYL A column, PolyLC Inc., 250x21 mm, 5 µm, 1000 A; buffer A: 0.1% (v/v) H ₃ PO4/25% AcCN; buffer B: 0.1% (v/v) H ₃ PO ₄ /25% AcCN/0.5 M NaCl). Fractions containing the desired product at the desired purity were pooled and concentrated using a TFF or Amicon Ultra-15 system. The resulting solution was then further purified by reverse phase HPLC (Waters C4 250x50 mm, 10 µm, 1000 A column or KROMASIL C8 250x50 mm, 10 µm, 100 A column; Buffer A: 0.05-0.1% TFA in water; buffer B: 0.05-0.1% TFA in AcCN). Fractions containing the desired product at the desired purity were pooled and freeze dried or buffer exchanged using the TFF system and/or Amicon Ultra-15 to give insulin dimers.

Tab. 4 lists dimers 40, 41, 45, 46, 47, 59, 57, 79, 80, 82 and 84 which were prepared using the respective intermediates according to General Method D. These dimers were characterized using UPLC-MS, Method D, or UHPLC-MS method E, or UHPLC-MS method G, and they demonstrated six-character, i. [(M+6)/6] (or seven-character, ie [(M+7)/7]) particles of the parent compound at a certain retention time (Rt).

Table 4 No. dimer a Skeleton of the first insulin skeleton second insulin A (') Dimer structure showing the binding fragment between lysine residues B2 9 and B2 9' Rt (min) (M+6)/6 or (M+7)/7 40 Analog 3 Analog 2 O 0 3.83 1970.38 41 Analog 1 Analog 2 N'4 Oh P a O 3.88 1938.84 45 Analog 9 Analog 2 ABOUT ABOUT 4.28 1985.30 46 Analog 3 Analog 10 O O 4.50 1985.43 47 Analog 9 Analog 10 O ABOUT 4.59 1714.34 75 Analog 3 Analog 2 O 3.40 1696.35 57 Analog 3 Analog 12 ^Ν Λ II 0 3.57 1975.90 79 Analog eleven Analog 13 .о _ о |f N - % ⁰ I 0 ⁰ 3.38 1993.39

- 81 041034

Example 19.

General method E: Synthesis of ^ ^,29B ^ ^6,29B -insulin dimers using Cu ²⁺ catalyzed double click chemistry.

In an appropriately sized container, the appropriate azido-containing insulin intermediate (analogue) was dissolved with gentle stirring at room temperature in a mixed solvent of DMSO and aq. triethylammonium acetate buffer (pH 7.0, final concentration 0.2 mM). In another appropriately sized container, the appropriate bis-acetylene-containing bridging or intermediate linker was dissolved with gentle agitation at room temperature in a mixed solvent of DMSO and water. Both solutions were combined, thoroughly mixed, degassed by gentle sparging with N ₂ . To the resulting solution was added freshly prepared sodium ascorbate or ascorbic acid solution (final concentration is 0.5 mm) and, after thorough mixing, a solution of 10 mm CuSO ₄ and tris[(1-benzyl-1H-1,2,3triazol-4-yl )methyl]amine (i.e. TBTA ligand) in 55% DMSO. After degassing by gentle sparging with N2 and thorough mixing, the mixture was stored at room temperature with occasional stirring overnight. The reaction mixture was carefully diluted with a mixed solvent (v/v 7:3 AcCN/water with 0.05% TFA) at 0°C and the pH was adjusted to 2.50 using 0.1, 1.0N. HCl (and 0.1N NaOH if necessary). The solution was first concentrated by ultrafiltration or with a tangential flow filtration (TFF) system or using Amicon Ultra-15 Centrifugal Units with a 1K, 3K or 10K MWCO membrane. The concentrated solution was usually first subjected to ion exchange chromatography (PolySULFOETHYL A column, PolyLC Inc., 250x21 mm, 5 μm, 1000 A; buffer A: 0.1% (v/v) H ₃ PO ₄ /25% AcCN; buffer B : 0.1% (v/v) H ₃ PO ₄ /25% AcCN/0.5 M NaCl). Fractions containing the desired product at the desired purity were pooled and concentrated using a TFF or Amicon Ultra-15 system. The resulting solution was then further purified by reverse phase HPLC (Waters C4 250x50 mm, 10 µm, 1000 A column or KROMASIL C8 250x50 mm, 10 µm, 100 A column; Buffer A: 0.05-0.1% TFA in water; buffer B: 0.05-0.1% TFA in AcCN). Fractions containing the desired product at the desired purity were pooled and freeze dried or buffer exchanged using the TFF system and/or Amicon Ultra-15 to give insulin dimers.

Tab. 5 lists dimers 42-44 and 54 which were prepared using the respective intermediates according to General Method E. The bis-acetylene bridged or intermediate linkers were

These dimers were characterized using UPLC-MS method D, or UPLC-MS method E, and they showed six-character, ie. [(M+6)/6], (or semi-charged, ie [(M+7)/7]) particles of the parent compound at a certain retention time (Rt).

- 82 041034

Example 20.

This example illustrates the synthesis of N ^2,1A ,N ^2,1A ,N ^2,1B ,N ^2,1B -tetrakis(acetyl or PEG1 or methoxyacetyl) dimers (dimer 48, 55, 56, 69 and 70).

To a solution of dimer 40, 19, or 4 (21 mg, 1.777 µmol) in DMSO (2 mL) was added TEA (3.96 µL, 0.028 mmol) at room temperature, followed by a solution of 2,5-dioxopyrrolidin-1-ylacetate (2 .23 mg, 0.014 mmol) in DMSO (100 µl) or other appropriate N-hydroxysuccinimide activated ester (2,5-dioxopyrrolidin-1-ylmethoxyacetate or 2,5-dioxopyrrolidin-1-yl-PEG1-acetate) in DMSO (100 µl). After 3 hours the reaction mixture was diluted with 12 ml water/AcCN=7/3 with 0.1% TFA and the pH was adjusted until it was 2.5. The resulting clear solution was concentrated using Amicon Ultra 15 Centrifuge Filters with a 10K MWCO membrane. The resulting solution was first subjected to ion exchange chromatography (PolySULFOETHYL A, 250x21 mm, 5 µm, 1000 A, 15 ml/min, gradient from 5% to 45% over 30 min; buffer A: 0.1% (v/v ) H ₃ PO ₄ /25% acetonitrile in water Buffer B: 0.1% (v/v) H ₃ PO ₄ /25% acetonitrile/0.5 M NaCl in water). Fractions containing the desired product at the desired purity were pooled and concentrated using an Amicon Ultra-15 with a 10K MWCO membrane. The resulting solution was then subjected to reverse phase HPLC (KROMASIL C8 column 250x50 mm, 10 μm, 100 A; buffer A: 0.05% TFA in AcCN/H ₂ O; buffer B: 0.05% AcCN; flow rate 85 ml/ min). The desired fractions were combined and freeze dried to give dimer 48, 55, 56, 69, or 70 as shown in Table 1. 6. UHPLC-MS method F or G was used. N-terminal substituents have the structure (acetyl), ⁰ (PEG1) or (methoxyacetyl), with the wavy line representing the bond between the substituent and the N2 nitrogen

- 83 041034

N-terminal amino acid.

Example 21.

Tab. 7 shows dimers 49, 50 and 51 and shows acyl groups attached to amino groups N ^2,A1 , N ^2,B1 , N ^2,A1 and n ^2,b1 . These dimers were prepared from dimer 40 using procedures similar to those described for dimer 48, but replacing 2,5-dioxopyrrolidin-1-ylacetate with the corresponding N-hydroxysuccinimide activated esters to give dimers 49, 50 and 51. The activated esters were yourself

2,5-dioxopyrrolidin-1-yl-Emoc-glycine acetate

2,5-dioxopyrrolidin-1-yl-PEG2-acetate

2,5-dioxopyrrolidin-1-yl-AEG-C6 acetate, where AEG is aminoethylglucose

These dimers were characterized using either UPLC-MS method F (dimers 50 and 51) or UPLC-MS method G (dimer 49) and showed six-character, ie. [(M+6)/6], (or semi-charged, ie [(M+7)/7]) particles of the parent compound at a certain retention time (Rt). Dimers are given in table. 7.

N-terminal substituents have the structure

(glycine),

(PEG2) or (AEG-C6).

- 84 041034

Table 7

Dimer structure showing

Type binding fragment (M+6) /6 insulin;

between lysine residues

N-termini or (M+7)/7 insulin

1722.06

B1, B1 -PEG2

RHI;

ΑΙ, ΑΓ

Bl,Bl'=AEMC6 epsilon-amino group

B29 Lys and B29' Lys, respectively.

Example 22.

This example illustrates the synthesis of dimer 52 using click chemistry without copper.

To a solution of analogue 3 (10 mg, 1.686 µmol) in 1.0 ml 3:2 v/v. H ₂ O/AcCN at room temperature was added a solution of analogue 4 (10.5 mg, 1.686 μmol) in 1.0 ml 3:2 v/v. _H2O /AcCN. After stirring at room temperature for 2 hours, the reaction mixture was first subjected to ion exchange chromatography (PolySULFOETHYL A, 250x21 mm, 5 µm, 1000 A, 15 ml/min, gradient from 5% to 45% in 30 min; buffer A: 0.1 % (v/v) H ₃ PO ₄ /25% acetonitrile in water Buffer B: 0.1% (v/v) H ₃ PO ₄ /25% acetonitrile/0.5 M NaCl in water) . Fractions containing the desired product at the desired purity were pooled and concentrated using an Amicon Ultra-15 with a 3K or 10K MWCO membrane. The resulting solution was then subjected to reverse phase HPLC (KROMASIL C8 column 250x50 mm, 10 μm, 100 A; buffer A: 0.05% TFA in AcCN/H ₂ O; buffer B: 0.05% AcCN; flow rate 85 ml/ min). The desired fractions were combined and freeze dried to give dimer 52. UPLC-MS method F:

Rt=3.73 min, m/z=1738.59 [(M+7)/7+1]. The results are shown in table. 8.

B29 Lys and B29' Lys, respectively.

Example 23.

This example illustrates the synthesis of N ^2,1A ,N ^2,1A ,N ^2,1B ,N ^2,1B -tetrakis(dimethyl or isobutyl) dimers (dimers 60, 58, 65 and 67). Dimer 40, 19 or 4 (100 mg, 8.46 µmol) was dissolved (suspended) in water (10 ml) and the pH was adjusted to 4.0 with acetic acid solution, then formaldehyde (0.013 ml, 0.169 mmol) or isobutyraldehyde was added (0.025 ml, 0.272 mmol) followed by the addition of a freshly prepared solution of sodium cyanoborohydride (10.63 mg, 0.169 mmol) in water (500 µl). A precipitate formed. The mixture was gently stirred. Approximately 1 hour after completion of the reaction, the mixture was carefully acidified by adding 1N. HCl dropwise to pH 2.9. The suspension became a clear solution. The mixture was purified by reverse phase preparative HPLC (column C-8, 50x250 cm, 85 ml/min, gradient from 29% to 36% over 25 min). (Water with 0.1% TFA and MeCN with 0.05% TFA). The desired fractions were lyophilized to give dimers (19.9 mg, 1.506 µmol, 17.80% yield). UPLC-MS method D: Rt=3.31 min, m/z=1989.44 [(M+6)/6+1].

- 85 041034

N-terminal substituents have the structure I (isobutyl), with the wavy line representing the bond between the substituent and the N2 nitrogen of the N-terminal amino acid, or I (N-dimethyl; Me2), the wavy line representing the bond between the N2 nitrogen and the C2 carbon of the N-terminal amino acids. Dimers are given in table. 9.

tabli No. dime ra Dimer structure showing the binding fragment between lysine residues B2 9 and B2 9' 60 O P ABOUT 58 ABOUT 65 ABOUT O 67 ABOUT 1 Wavy line denote B29 Lys and B29' Lys, respectively

ma 9 type of insulin; N-termini of insulin Rt (min) (M+6)/6 or (M+7)/7 RHI; ΑΙ, ΒΙ, ΑΙ', B'=Me2 3.31 1989.44 RHI; ΑΙ, ΒΙ, ΑΙ', B'=Me2 3.42 1978.48 RHI; ΑΙ, ΒΙ, ΑΙ', B' = isobutyl 4.13 1997.04 RHI; ΑΙ, ΒΙ, ΑΙ', B'=Me2 3.42 1719.39 d bond with epsilon-amino group ΙΗΟ .

Example 24.

The synthesis of dimers 61, 62, 63, 64, and 66 was carried out as follows.

The synthesis of 2,5-dioxopyrrolidin-1-yl-6-((2-((2,5-dioxopyrrolidin-1-yl)oxy)-2 oxoethyl)amino)-6-oxohexanoate (C6-glycine linker; linker 24 ).

Step 1. Benzyl (2,5-dioxopyrrolidin-1-yl) adipate.

To a solution of 6-(benzyloxy)-6-oxohexanoic acid (5 g, 21.16 mmol) in DMF (10 ml) was added N-ethyl-N-isopropylpropan-2-amine (4.44 ml, 25.4 mmol) followed by TSTU (7.01 g, 23.28 mmol).

The reaction was stirred at 0° C. for 1 h and at room temperature for 1 h. The mixture was poured into ice water/ethyl ether (1/1, 100 ml). The mixture was extracted with ethyl ether (3x50 ml), washed with water (2x10 ml) and brine (10 ml). The organic layer was dried over MgSO ₄ , filtered through a Celite pad and concentrated to give the title compound as a colorless syrup (5.2 g, 15.6 mmol, 74%). LC-MS 2 min: Rt=1.05 min, m/z=334.1 [M+1].

Step 2. ((Carboxymethyl)amino)-6-oxohexanoic acid.

To a solution of glycine (225 mg, 3.0 mmol) in DMF (2.5 ml) was added dropwise the product of step 1 (1.0 g, 3.0 mmol) in DMF (2.5 ml) followed by TEA ( 418 µl, 3.0 mmol). The reaction was stirred at room temperature for 18 hours. DMF was removed under reduced pressure. The crude product was purified by C18 reverse phase chromatography (eluted with 0-40% AcCN/water in 16 column volumes (CV)). Fractions containing the desired product were combined, concentrated and lyophilized to give the intermediate (6-(benzyloxy)-6-oxohexanoyl)glycine. Pd/C (10%, 160 mg, 0.15 mmol) was added to the above intermediate in water (3 ml). The reaction was stirred at room temperature under a balloon of hydrogen for 18 hours. The mixture was filtered through a pad of celite, washed with MeOH/water (1/1, 10 ml). The filtrate was concentrated and lyophilized to give the title compound (400 mg, 2.2 mmol, 66%). LC-MS 2 min: Rt=0.28 min, m/z=204.03 [M+1].

- 86 041034

Step 3. 2,5-Dioxopyrrolidip-1-yl-6-((2-((2,5-dioxopyrrolidin-1-yl)oxy)-2-oxoethyl)amino)-6oxohexanoate.

To the product of step 2 (10 mg, 0.049 mmol) in DMF (0.5 ml) at 0°C was added TEA (0.015 ml, 0.108 mmol) followed by TSTU (31.1 mg, 0.103 mmol). The reaction was warmed to room temperature and stirred at this temperature for 1 h. TLC (EtOAc/MeOH/water/AcCN: 2:1:1:1 (v:v:v:v)) showed formation of the desired product (Rf: 0 ,25), and no starting material remained. The crude material was used to construct dimers without purification.

Linker 25 (C6-alanine), linker 26 (C6-isoleucine), linker 27 (C6-leucine) and linker 28 (C6-valine), in which the amino acid containing the C6-amino acid linker is alanine, isoleucine, leucine and valine, respectively, was synthesized similarly to the method shown above. Dimers were designed using the above linkers using preparative method D. The results are shown in table. 10.

Dimer structure showing the binding fragment between lysine residues B29 and B29'

1995.62

type of insulin;

N-termini of insulin (M+6)/6 or (M+7)/7 1993.24

Al, Bl, Al , Bl = carbamoes

Example 25.

The synthesis of dimers 73, 89, 90, 91, 92, and 93 was carried out as follows.

The synthesis of bis-2,5-dioxopyrrolidin-1-yl-3,3'-(1,3-phenylene) dipropionate (dipropylphenyl; linker 29) is described.

Step 1. Bis-2,5-dioxopyrrolidin-1 -yl-3,3'-(1,3-phenylene) dipropionate.

To a solution of 3,3'-(1,3-phenylene)dipropionic acid (21.8 mg, 0.098 mmol) in DMF (0.6 ml) at 0°C was added TEA (29 ml, 0.206 mmol) followed by TSTU (62.0 mg, 0.206 mmol). The reaction was warmed to room temperature and stirred at this temperature for 1 h. TLC (EtOAc/MeOH/water/AcCN: 2/1/1/1) showed formation of the desired product (Rf: 0.25) with no starting material remaining. . UPLC-MS method B: Rt=3.47 min, m/z=417.19 [M+1]. The product was used without further purification to construct dimer 73 using analog 5 using method D.

The synthesis of bis(2,5-dioxopyrrolidin-1-yl)benzene-1,3-dicarboxylate (terephthalate; linker 34) is described.

- 87 041034

Step 1. Bis(2,5-dioxopyrrolidin-1-yl)benzene-1,3-dicarboxylate.

At 0°C, to a solution of terephthalic acid (100 mg, 0.602 mmol) in THF (2 ml) was added 2-(2,5dioxopyrrolidin-1-yl)-1,1,3,3-tetramethylisouronium tetrafluoroborate (371 mg, 1.234 mmol ), followed by Nethyl-N-isopropylpropan-2-amine (0.222 mL, 1.234 mmol)). The ice bath was removed after 30 min. The solution was stirred at room temperature for 1 hour. An additional 25 ml of THF was added and the reaction was left overnight at room temperature. The product was concentrated down to approximately 5 ml and a portion was used as is without further purification to construct dimer 89 using RHI using method D. The remaining material was diluted with ethyl acetate (200 ml) and washed with brine (10 ml), the organic layer was dried with using Na ₂ SO ₄ , filtered and concentrated.

The synthesis of bis(2,5-dioxopyrrolidin-1-yl)isophthalate (isophthalate; linker 35) is described.

Step 1. Bis(2,5-dioxopyrrolidin-1-yl)isophthalate.

To isophthalic acid (54 mg, 0.325 mmol) in DMSO (1 mL) was added TSTU (215 mg, 0.715 mmol) followed by TEA (0.137 mL, 0.975 mmol). LC-MS 2 min: Rt=0.79 min, m/z=721.28 [2M+1]. The product was used without purification to construct dimer 90 using analog 5 and dimer 91 using RHI in method E.

The synthesis of bis(2,5-dioxopyrrolidin-1-yl)-4-((tert-butoxycarbonyl)amino)heptanedioate (heptanedioate; linker 36) is described.

Step 1. Bis(2,5-dioxopyrrolidin-1-yl)-4-((tert-butoxycarbonyl)amino)heptanedioate.

To 4-((tert-butoxycarbonyl)amino)heptanedioic acid (16.5 mg, 0.06 mmol) in DMSO (0.5 ml) was added TSTU (39.7 mg, 0.132 mmol) followed by TEA (0.025 ml , 0.180 mmol). LC-MS 2 min: Rt=0.90 min, m/z=470.34 [M+1]. The product was used without purification to construct dimer 92 using analog 5 and dimer 93 using RHI in method E.

The results are shown in table. eleven.

Table 11 No. dimer Structure of a dimer showing a binding fragment between lysine residues B29 and B29' type of insulin; N-termini of insulin Rt (min) (M+6) /6 or (M+7)/7 73 - ABOUT RHI; Al, Bl, A1 ', B1' = carbamoyl 3.55 1711.57 89 O O RHI; ΑΙ, ΒΙ, ΑΙ', B1'=H 3.43 1678.64 90 6%% Oh Oh RHI; Al, Bl, A1', B1' = carbamoyl 3.67 1703.20

- 88 041034

Example 26.

The synthesis of dimers 71, 72, 77, 78, 81, and 87 was carried out as follows. The synthesis of bis(2,5dioxopyrrolidin-1-yl)-( 1 S,4S)-cyclohexane-1,4-dicarboxylate (linker 30; trans-cyclohexane-1,4-diacids) is described.

TSTU (734 mg, 2.439 mmol) and DIPEA (0.5 ml, 2.86 mmol). The resulting reaction mixture was stirred at room temperature for 1 hour. The product was ground in the reaction solution as a white solid; it was filtered and washed with DCM (2x5 ml); and dried under vacuum to give the title compound. UHPLC-MS calculated for C ₁₆ H ₁₈ N ₂ O ₈ , 366.11, observed m/e: 367.16 (M+H)+, (Rt: 3.20/5.00 min). UHPLC-MS method A.

1H NMR (500 MHz, DMSO): δ 2.81-2.89 (m; 2H); 2.80 (s; 8H); 2.02-2.10 (m; 4H); 1.57-1.63 (m; 4H).

The synthesis of bis(2,5-dioxopyrrolidin-1-yl)-(1R, 4R)-cyclohexane-1,4-dicarboxylate (linker 31; cis-cyclohexane-1,4-diacid) is described.

O O /—ξ o o } -N

Oh oh

TSTU (734 mg, 2.439 mmol) and DIPEA (0.5 ml, 2.86 mmol). The resulting reaction mixture was stirred at room temperature for 1 hour. The residue was purified by silica gel chromatography (0-100% EtOAc/hexanes) to give the title compound. UHPLC-MS calculated for C ₁₆ H ₁₈ N ₂ O ₈ , 366.11, observed m/z: 367.17 (M+H) ⁺ , (Rt: 3.17/5.00 min). UHPLC-MS method A.

1 H NMR (500 MHz, DMSO): δ 3.02-3.08 (m; 2H); 2.80 (s; 8H); 1.80-1.90 (m; 8H).

The synthesis of 1-(tert-butyl)-3,5-bis(2,5-dioxopyrrolidin-1-yl)-(3R,5S)-piperidine-1,3,5tricarboxylate (linker 32) is described.

.0 o.

N

O^O

To a solution of (3R,5S)-1-(tert-butoxycarbonyl)piperidine-3,5-dicarboxylic acid (200 mg, 0.734 mmol) in DMF (7 ml) at 0°C was added TSTU (485 mg, 1.611 mmol) and DIPEA (0.3 ml, 1.718 mmol). The resulting reaction mixture was stirred at room temperature for 2 hours. The residue was purified by silica gel chromatography (0-100% EtOAc/hexanes) to give the title compound. UHPLC-MS calculated for C ₂₀ H ₂₅ N ₃ O ₁₀ , 467.15, observed m/e: 468.30 (M+H) ⁺ , (Rt: 0.98/2.00 min). UHPLC-MS method A.

General Method F: Synthesis of N ^6.29 B,N ^6.29 B'-insulin dimers under organic base conditions.

Suspend the insulin or insulin analogue at room temperature in an appropriately sized container in an organic solvent or mixed aqueous/organic solvents, eg DMSO, in the presence of a base, eg TEA or 1,1,3,3-tetramethylguanidine (TMG). The mixture is left to stir gently until the insulin is completely dissolved. K received

- 89 041034 To this solution is added the activated ester intermediate in a solution of organic solvents such as DMSO or DMF. After UPLC, the chromatogram shows that a significant part of the reaction mixture turned into a dimer of N ⁶ ^ ²⁹ B,N ⁶ ^ ^B29B -insulins (or dimer n ^6.28b , n ^6.28b of insulin lispro), the reaction solution was transferred using an automatic pipette to a centrifuge 50 ml tube containing IPAc/MTBE (v/v 4:1) (45 ml). The addition was carried out dropwise. The resulting white suspension was centrifuged (3000 rpm, 15 min at 4°C) to obtain a clear supernatant and a white precipitate. The supernatant was discarded and the white precipitate was dried under vacuum. The white precipitate containing the crude intermediate was then dissolved in 2 ml of TFA at 0°C and stirred for 10 min at the same temperature. After completion of the boc removal reaction, the reaction solution was transferred using an automatic pipette into a 50 ml centrifuge tube containing MTBE (45 ml). The addition was carried out dropwise. The resulting white suspension was centrifuged (3000 rpm, 15 min at 4°C) to obtain a clear supernatant and a white precipitate. The supernatant was discarded and the white precipitate was dried under vacuum and redissolved in a solution of CH3CN/H2O (v/v 1:4). The reaction mixture can be directly purified by reverse phase HPLC (Waters C4 250x50 mm, 10 µm, 1000 A column or KROMASIL C8 250x50 mm, 10 µm, 100 A column; buffer A: 0.05-0.1% TFA in deionized water; buffer B: 0.05-0.1% TFA in AcCN), or the reaction can be stopped by careful dilution with cold acidic H ₂ O (20x, pH ~ 3.0) at 0°C, and its pH adjusted to a final pH of 2.5 using 1N. HCl (and 0.1N NaOH if necessary). The solution can first be concentrated using ultrafiltration or a tangential flow filtration (TFF) system, or using Amicon Ultra-15 Centrifugal Units with a 1K, 3K or 10K MWCO membrane. The concentrated solution is usually first subjected to ion exchange chromatography (PolySULFOETHYL A column, PolyLC Inc., 250x21 mm, 5 µm, 1000 A; buffer A: 0.1% (v/v) H3PO4/25% AcCN; buffer B: 0. 1% (v/v) H3PO4/25% AcCN/0.5 M NaCl). Fractions containing B29 conjugate of the desired purity are pooled and concentrated using a TFF or Amicon Ultra-15 system.

The concentrated solution is then purified by reverse phase HPLC (Waters C4 250x50 mm, 10 µm, 1000 A column or KROMASIL C8 250x50 mm, 10 µm, 100 A column; Buffer A: 0.05-0.1% TFA in deionized water; buffer B: 0.05-0.1% TFA in AcCN). Fractions containing the desired insulin dimer are pooled and freeze dried or buffer exchanged using a TFF system and/or Amicon Ultra-15 to give ^N6,29b , ^N6,29b -insulin dimers.

Tab. 12 lists dimers 71, 72, 77, 78, 81 and 87 which were prepared using the appropriate linker according to General Method B (dimers 77 and 78), General Method C (dimers 71 and 72), or General Method F (dimers 81 and 87). These dimers were characterized using UPLC-MS method D, or UPLC-MS method E, and they showed six-character, ie. [(M+6)/6], (or semi-charged, ie [(M+7)/7]) particles of the parent compound at a certain retention time (Rt).

- 90 041034

Table 12 No. dimer Structure of a dimer showing a binding fragment between lysine residues B29 and B29' type of insulin; N-termini of insulin Rt (min ) (M+6) /6 or (M+7)/7 71 ¹ s+U RHI; ΑΙ, ΒΙ, ΑΙ', B1'=H 3, 05 1959.33 1679.86 72 RHI; ΑΙ, ΒΙ, ΑΙ', B1'=H 3, 07 1959.33 1679.86 81 Oh Oh n RHI; ΑΙ, ΒΙ, ΑΙ', B1'=H 3.37 1959.48 1679.74 77 G RHI; ΑΙ,ΒΙ,ΑΙ',B1'= carbamoyl 3.50 1988.07 1704.29 78 Oh, — K o u ... / u.u RHI; ΑΙ,ΒΙ,ΑΙ',B1'= carbamoyl 3.52 1988.08 1704.35 86 Oh Oh xAqV. Chg n RHI: ΑΙ,ΒΙ,ΑΙ',B1'= carbamoyl 3.48 1988.18 1704.24 The wavy line indicates the bond with the epsilon-amino group B29 Lys and B29' Lys, respectively.

Example 26.

The synthesis of dimer 68 and dimer 74 was carried out as follows.

The synthesis of bis(2,5-dioxopyrrolidin-1-yl)-6,6'-((6-chloro-1,3,5-triazine-2,4-diyl)bis(azanediyl))dihexanoate (linker 33; C6N-chloro-1,3,5-triazine-NC6).

A solution of 2,4,6-trichloro-1,3,5-triazine (80 mg, 0.434 mmol) and methyl-6-aminohexanoate (129 mg, 0.889 mmol), HCl salt, in CH ₂ Cl ₂ (1 ml) was cooled up to -30°С. A solution of DIPEA (0.379 ml, 2.169 mmol) in CH ₂ Cl ₂ (1 ml) was added dropwise. The mixture was stirred at -30° C.-room temperature for 5 hours. Then CH ₂ Cl ₂ (20 ml) was added and washed with aqueous HCl (1M) (2x10 ml), aqueous NaHCO ₃ (10 ml) and brine (10 ml). The organic layer was dried over sodium sulfate and filtered, concentrated in vacuo to give dimethyl 6,6'-((6-chloro-1,3,5-triazine-2,4diyl)bis(azanediyl))dihexanoate (135 mg, 0.336 mmol).

To a solution of dimethyl 6,6'-((6-chloro-1,3,5-triazine-2,4-diyl)bis(azanediyl))dihexanoate (135 mg, 0.336 mmol) in THF (0.5 ml) and methanol (0.5 ml) was added aqueous (2M) LiOH (504 μl, 1.008 mmol). The mixture was stirred at room temperature for 1 h and then concentrated under vacuum to obtain a dry residue. Dissolve the residue in water and neutralize with aq. HCl. The precipitate was collected by filtration and washed with water. Dry the solid under vacuum to give 6,6'-((6-chloro-1,3,5-triazine-2,4-diyl)bis(azanediyl))dihexanoic acid.

TSTU ( 174 mg, 0.578 mmol) followed by triethylamine (81 µl,

- 91 041034

0.578 mmol). The reaction was stirred for 1 h. UHPLC indicates the formation of the desired material,

UPLC-MS method C: Rt=0.99 min, m/z=568.2 [M+1]. This reagent (0.1M/DMF) was used without further purification.

Dimers were prepared using either General Method B (Dimer 74) or General Method C (Dimer 68). These dimers were characterized using UHPLC-MS method D or UHPLC-MS method E, and they showed six-character, ie. [(M+6)/6], (or semi-charged, ie [(M+7)/7]) particles of the parent compound at a certain retention time (Rt). Dimer 68 was constructed using linker 33 and RHI. Dimer 74 was constructed using linker 33 and analogue 5. The results are shown in table. 13.

Table 13

Dimer structure showing (M+6)/6 linking fragment between lysine residues or

R29 and R29' (M+7)/7

1993.09

Al, B1 '=carb amoil

B29 Lys and B29' Lys

-amino group

Example 27.

Synthesis of dimer 54 was carried out as follows.

Dissolve ΑΙ-ΤΦΥ-RHI (D. Liu et al., Journal of Peptide Sci., 2012, 18, 336-341) (100 mg, 0.017 mmol) in a premix containing water (5 ml) and dibasic potassium phosphate ( 24.49 mg, 0.141 mmol) (the pH of the resulting solution is approximately 4). Potassium cyanate (27.5 mg, 0.339 mmol) was added and stirred overnight. The product was purified by reverse phase C-8 chromatography (column KROMASIL, C8 10 μM, 100 A, size 250x50 mm; solvent A=water/0.05% TFA, solvent B=AcCN/0.05% TFA) , flow=85 ml/min, gradient B to A 26-34% over 30 min. UPLC-MS method F: Rt=4.48 min, m/z=1486.9 [(M+4)/4+1].

Dissolve the above product (60 mg, 10.09 µmol) in DMSO (594 µl) and add triethylamine (28.1 µl, 0.202 mmol) followed by a solution of the linker disuccinimidyl suberate (1.858 mg, 5.04 µmol) dissolved in 100 µl DMSO. Stirred for 3 hours. HPLC indicates completion of the reaction. The entire reaction mixture was added to ammonium hydroxide (2105 μl, 15.13 mmol) (dropwise, exotherm expected). Stir gently for 2 hours and confirm deprotection of the TFA group. Dilute the mixture with 20 ml of water and remove most of the ammonium hydroxide by diafiltration using 10K Amicon tubes. The pH was adjusted to about 3 and the salts were removed by diafiltration. The product was purified by ion exchange chromatography (PolySULFOETHYL A column, PolyLC Inc., 250x21 mm, 5 μm, 1000 A; buffer A: 0.1% (v/v) H ₃ PO ₄ /25% AcCN; buffer B: 0.1% (v/v) H ₃ PO ₄ /25% AcCN/0.5 M NaCl). The product was repurified by C-8 reverse phase chromatography (KROMASIL column, C8 10 µM, 100 A, size 250x50 mm; solvent A=water/0.05% TFA, solvent B=AcCN/0.05% TFU). The result is shown below. The results are shown in table. 14.

- 92 041034

Table 14 No. dimer Dimer structure showing the binding fragment between lysine residues B2 9 and B2 9' type of insulin; N-termini of insulin Rt (min) (M+6) /6 or (M+7)/7 54 ABOUT 0 RHI; A1,AG=H, Bl,B1'=carbamoyl 4.57 1974.45 The wavy line indicates the bond with the epsilon-amino group B29 Lys and B29' Lys, respectively.

Example 28.

Insulin receptor binding assays were performed as follows.

IR binding assay was performed in scintillation proximity assay (SPA) in 384-well format using cell membranes derived from CHO cells overexpressing human IR(B) grown in F12 medium containing 10% PBS and antibiotics (G418, penicillin/ streptavidin). Cell membranes were obtained in 50 mm Tris buffer, pH 7.8 containing 5 mm MgCl ₂ . The assay buffer contained 50 mM Tris buffer, pH 7.5, 150 mM NaCl, 1 mM CaCl ₂ , 5 mM MgCl ₂ , 0.1% BSA, and protease inhibitors (Complete-Mini-Roche). Cell membranes were added to WGA PVT PEI SPA beads (final concentration 5 mg/ml) followed by the addition of insulin dimer molecules at appropriate concentrations. After 5-15 min incubation at room temperature, ¹²⁵ [1]-insulin was added at a final concentration of 0.015 nM for a final total volume of 50 μl. The mixture was incubated with shaking at room temperature for 1 to 12 h followed by scintillation counting to determine the binding of ¹²⁵ [I]-insulin to IR and the effect of titration of insulin dimer molecules on this interaction.

Example 29.

Insulin receptor (IR) AKT phosphorylation assays were performed as follows.

IR AKT Phosphorylation Assay: Insulin receptor activation can be assessed by measuring Akt protein phosphorylation, a key step in the insulin receptor signaling cascade. CHO cell lines overexpressing human IR were used in an HTRF sandwich ELISA kit (Cisbio, Phospho-AKT(Ser473) and Phospho-AKT(Thr308) Cellular Assay Kits). Cells were grown in F12 medium supplemented with 10% FBS, 400 μg/ml G418 and 10 mM HEPES. Prior to analysis, cells were incubated in serum-free medium for 2 to 4 hours. Alternatively, cells can be pre-frozen and aliquoted in medium containing 20% DMSO and used in the assay after thawing, settling, and resuspension. Cells were seeded at 10,000 cells per well in 20 μl of serum-free F12 medium in 384-well plates. Each test compound plate was monitored with humulin and insulin glargine. Titratable compounds were added to cells (2 μl/well, final concentrations=1000 nM, titrated up to 0.512 pM in 1:5 dilutions) and incubated at 37° C. for 30 minutes. Cells were lysed with 8 µl of the prepared lysis buffer provided in the CisBio kit and incubated at 25°C for 1 h. Diluted antibody reagents (anti-AKT-d2 and anti-pAKT-Eu3/cryptate) were prepared according to with kit instructions, and then 10 µl was added to each well of the cell lysate, followed by incubation at 25°C for 3.5 to 5 hours. The plate was read using an Envision plate reader (excitation=320 nm; emission=665 nm) to determine the agonist activity of IR pAkt in terms of both potency and maximal response for each compound. Alternatively, the compounds were tested in the same manner in the presence of 1.6 nM humulin in order to determine whether each compound was able to compete with the full agonist activity of insulin.

Example 30.

Tab. 15 shows the in vitro biological activity of insulin dimers at the insulin receptor (IR). Activity was measured either by ligand competition assays as described in Example 28 or by Akt-phosphorylation functional assays as described in Example 29.

- 93 041034

Table 15 dimer number Binding to IR 1C ₅₀ (nM) IR pAkt %max dimer number Binding to IR 1C ₅₀ (nM) IR pAkt %max 1 1.50 50.5 43 2.02 35 2 3.77 33 44 1, 18 33 3 1.38 59.5 45 4.00 38 4 1, 62 47.5 46 0.38 25 5 2.42 43 47 4.56 thirty 6 1.00 83 48 5, 09 35 7 3.19 60 49 5.16 58 8 7.15 47.5 50 3, 61 39 9 3.29 42 51 0.59 27 10 3.59 34 52 3, 93 67 eleven 1.37 32 54 0.49 26 12 13, 1 41 55 2.02 28 13 2.51 36 56 2.94 33 14 4.42 38 57 2.02 26 15 2.53 49 58 0.61 34 16 4.76 59 0.97 53 17 5.27 62 60 1.81 60 18 5.23 36 61 20.3 33 19 1.40 25 62 5, 01 32 20 0.62 36 63 26, 7 41 21 1.88 41 64 15, 7 34 22 1.85 41.5 65 3, 05 42 23 2.48 39 66 34.5 33 24 4.29 29, 3 67 0.54 24 25 1.82 46 68 4.81 34 26 3.74 44 69 4.99 18 27 3.98 31 70 14.8 21 28 1.96 25 71 2.28 45 29 1.18 thirty 72 2.35 42 thirty 1.79 thirty 73 37.8 35 31 1.36 31 74 73, 8 33 32 21, 6 33 75 1.85 52 33 2.41 38 76 1.87 35 34 0.57 67 77 42.1 23 35 1.17 66 78 167 22 36 0.53 28 79 1.47 42 37 3.52 42 80 5.48 39 38 2.18 45 81 0.3 40 39 3, 17 45 82 1.73 50 40 2.03 41 83 558 21 41 0.64 35 84 2.19 50 42 2.97 49

Example 31.

In this example, the in vivo effects of several insulin receptor partial agonists of the present invention were compared with Compound A (MIU-90 insulin dimer disclosed in PCT Publication No. WO 2014052451) and Compound B (B29, B29'-suberoil-(insulin)2), disclosed in Deppe et al., Nauyn-Schmiedeberg's Arch. Pharmacol. 350: 213-217 (1994), but using RHI instead of bovine insulin in an intraperitoneal insulin tolerance test (IP-ITT) performed on adult male lean C57BL/6NTac mice.

Groups of N=6-8 animals per group were randomized by weight (approximately 30 grams on average). Two days before the study, mice were prepared for dosing with an intraperitoneal injection of 0.9% sodium chloride solution at a dosing volume of 5 ml/kg. On the morning of the study, food was removed 2 or 4 hours before the study. The concentration of glucose in the blood was determined at T=0 min (baseline) using a glucometer. Mice were then dosed with vehicle, dimer 24, dimer 55, dimer 58, dimer 60, dimer 67, compound A, compound B, or humulin (RHI) at 5 ml/kg via intraperitoneal injection (see Table 16 for doses used). Blood glucose levels were determined from tail phlebotomies 30 to 360 minutes after dosing.

- 94 041034

Table 16 Compound Doses A 72 nmol/kg 300 nmol/kg IN 72 nmol/kg 300 nmol/kg Dimer 2 4 72 nmol/kg 300 nmol/kg Dimer 55 120 nmol/kg 300 nmol/kg Dimer 58 60 nmol/kg 300 nmol/kg Dimer 60 120 nmol/kg 300 nmol/kg Dimer 67 60 nmol/kg 300 nmol/kg

Humulin 18 nmol/kg 72 nmol/kg

The results are shown in FIG. 2A, 2B, 2C, 2D, 2E, 2F and 2G. The results demonstrate that the glucose profile for dimer 24, dimer 55, dimer 58, dimer 60 and dimer 67 was essentially the same at both doses tested, while increasing the dosage of compounds A and B caused an increase in glucose-lowering activity, indicating a lower potential risk. hyperglycemia for dimers compared to RHI or Compounds A and B.

Example 32.

The glucose lowering effect of dimers 24, 18 and 40 was compared to RHI in diabetic Yucatan minipigs (minipigi D) as follows.

Yucatan minipigs were made type 1 diabetic using alloxan injections according to a proprietary protocol developed by the Sinclair Research Center (Auxvasse, MO). Induction was considered successful if basal glucose levels exceeded 150 mg/dl. Minipigs D were used in these experiments with plasma glucose levels of approximately 300 mg/dL.

These studies used male Yucatan minipigs equipped with two vascular access ports (VAPs) in the jugular vein. On the day of the study, after an overnight fast, the minipigs were placed in suspensions and accessed to the VAP for infusion and sampling. At t=0 min and after collecting two baseline blood samples for plasma glucose measurement (t=-30 min and t=0 min), minipigs were administered humulin (recombinant human insulin, RHI) or IRPA as a single IV bolus in the amount of 0.69 nmol/kg. Humulin and IRPA were 0.69 nmol/ml in buffer containing glycerol, 16 mg/ml; metacresol, 1.6 mg/ml; phenol, 0.65 mg/ml; sodium phosphate anhydrous, dibasic, 3.8 mg/ml; The pH was adjusted to 7.4 with HCl. After dosing, sampling was continued for 480 minutes; time points for sample collection were -30 min, 0 min, 8 min, 15 min, 30 min, 45 min, 60 min, 90 min, 120 min, 150 min, 180 min, 210 min, 240 min, 270 min, 300 min, 330 min, 360 min, 420 min, 480 min. Blood was collected in 3-EDTA tubes supplemented with 10 μg/ml aprotinin and stored on ice until processing, which was performed within 30 minutes of collection. After centrifugation at 3000 rpm, 4° C. for 8 minutes, plasma was collected and aliquoted for glucose measurement using a Beckman Coulter AU480 chemistry analyzer and for compound levels.

Fig. 1 shows that at a concentration of 0.69 nmol/kg, RHI lowered serum glucose levels below 50 mg/dl, whereas insulin dimers did not. This result indicates that insulin dimers pose a lower risk of hypoglycemia than RHI.

Example 33.

Glucose lowering effect of dimers 4, 5, 7, 8, 9, 18-29, 32, 37-41, 43, 44, 48, 55, 57, 58, 60, 61, 62, 64, 67, 69, 71 , 72, 77 and 78 were compared to RHI in diabetic miniature Yucatan pigs (minipigi D) as follows.

Yucatan minipigs were made type 1 diabetic using alloxan injections according to a proprietary protocol developed by the Sinclair Research Center (Auxvasse, MO). Induction was considered successful if basal glucose levels exceeded 150 mg/dl. These studies used D minipigs with plasma glucose levels of approximately 300-400 mg/dl and equipped with two vascular access ports (VAPs) in the jugular vein.

On the day of the study, after an overnight fast, the minipigs were placed in suspensions and accessed to the VAP for infusion and sampling. At t=0 min and after collecting two baseline blood samples for plasma glucose measurement (t=-30 min and t=0 min), minipigs were given humulin (recombinant human insulin, RHI) or another dimer as a single IV bolus per day. amount of 0.69 nmol/kg (0.35 nmol/kg for compound #78). Humulin and dimers were 69 nmol/ml in buffer containing glycerol, 16 mg/ml; metacresol, 1.6 mg/ml; phenol, 0.65 mg/ml; anhydrous sodium phosphate,

- 95 041034 dibasic, 3.8 mg/ml, pH adjusted to 7.4 with HCl. After dosing, sampling was continued for 480 minutes; the sampling time points were -30, 0, 8, 15, 30, 45, 60, 90, 120, 150, 180, 210, 240, 270, 300, 330, 360, 420, and 480 min. Blood was collected in 3-EDTA tubes supplemented with 10 μg/ml aprotinin and stored on ice until processing, which was performed within 30 minutes of collection. After centrifugation at 3000 rpm, 4° C. for 8 minutes, plasma was collected and aliquoted for glucose measurement using a Beckman Coulter AU480 chemistry analyzer. The results are shown in figures 7A-7H. These figures show that at a concentration of 0.69 nmol/kg, RHI reduced serum glucose levels below 50 mg/dl, whereas insulin dimers did not. This result indicates that insulin dimers pose a lower risk of hypoglycemia than RHI.

Example 34.

This experiment compared the stability of the disulfide binding moiety of compound A with the suberoyl (C8) binding moiety of dimer 24.

μM compound A and dimer 24 were separately incubated in rat kidney cell membranes (RKCM) with or without 5 mM glutathione (GSH). Samples were taken at time 0 and time 2 h and the reaction was stopped with 1 volume of 10% MeOH in AcCN with 0.1% formic acid. Stopped samples were then centrifuged and frozen until analysis. The samples were then thawed and analyzed using the Thermo Orbi Velos system. Targeted MetID analysis was performed on extracted ion chromatograms (XIC) using 3 isotopes from 2 charge states at a 10 ppm window.

Compound A metabolites were detected in RKCM both without GSH and with GSH. As shown in FIG. 3, the monomer was approximately 1% of the parent compound (compound A stock solution) by 2 hours of incubation. As shown in FIG. 4, the monomer was approximately 6.5% of the parent compound (compound A stock solution) by 2 hours of incubation. These results show that the disulfide bond breaks down over time. Metabolites were not observed in the control at 0 h for Compound A or in the stock solutions.

Dimer 24 produced metabolites that were detected in RKCM, however, although loss of the A chain polypeptide was observed due to the breaking of disulfide bonds between the A chain polypeptide and the B chain polypeptide, no monomers were detected. Fig. 5 shows that without GSH, the A chain polypeptide loss was less than 1% of the parent compound (dimer 24 stock solution). Fig. 6 shows that with GSH the A chain polypeptide loss was less than 1% of the parent compound (dimer stock solution 24). Metabolites were not observed in the control at 0 h for dimer 24 or in stock solutions. The new procedure with acidic conditions did a good job of stopping the disulfide exchange.

Sequence table SEQ ID NO: Description Subsequence 1 Insulin Homo sapiens, A-chain GIVEQCCTSICSLYQLENYCN 2 Insulin Homo sapiens, B-chain FVNQHLCGSHLVEALYLVCGERGFFYTPKT 3 An artificial sequence, the A-chain of insulin X ₂ is isoleucine or threonine; X ₃ is valine, glycine or leucine; X ₈ is threonine or histidine; X ₁₇ is glutamic acid or glutamine; X ₁₉ is tyrosine, 4-methoxyphenylalanine, alanine or 4-aminophenylalanine; X ₂₃ is asparagine or glycine; GX ₂ X ₃ EQCCX ₈ SICSLYQLX ₁₇ NX ₁₉ CX ₂₃ 4 artificial sequence, GX ₂ X ₃ EQCCX ₈ SICSLYQLX ₁₇ NX ₁₉ CX ₂₃

- 96 041034

B-chain of insulin X ₂₅ is histidine or threonine; X ₂₉ is alanine, glycine or serine; X ₃₀ is histidine, aspartic acid, glutamic acid, homocysteine acid or cysteic acid; X ₃₁ is proline or lysine; and X ₃₂ is proline or lysine, provided that at least one of X ₃₁ or X ₃₂ is lysine 5 X ₂ 2 is phenylalanine X22VNQX25X26CGX29X30LVEALYLVCGERG or deaminophenylalanine; X ₂ 5 is histidine or threonine; X ₂₆ is glycine or leucine; X ₂₇ is phenylalanine or aspartic acid; X ₂₉ is alanine, glycine or serine; X ₃₀ is histidine, aspartic acid, glutamic acid, homocysteine acid or cysteic acid; X ₃₁ is aspartic acid, proline or lysine; X ₃₂ is lysine or proline; X ₃₃ is threonine, alanine or absent; X ₃₄ is arginine or absent; and X ₃₅ is arginine or absent; provided that at least one of X ₃₁ or X ₃₂ is a lysine FX27YTX ₃₁ X ₃₂ X ₃₃ X ₃ 4X ₃ 5

6 Artificial sequence, Insulin lispro B-chain FVNQHLCGSHLVEALYLVCGERGFFYTKPT 7 artificial sequence, A-chain of insulin glargine GIVEQCCTSICSLYQLENYCG 8 Artificial sequence, Insulin glargine B-chain FVNQHLCGSHLVEALYLVCGERGFFYTPKT RR 9 Artificial sequence, B-chain of insulin aspart FVNQHLCGSHLVEALYLVCGERGFFYTDKT 10 artificial sequence, B:des30 FVNQHLCGSHLVEALYLVCGERGFFYTPK eleven artificial sequence, A:Y19A GIVEQCCTSICSLYQLENACN

While the present invention has been described herein with reference to the illustrated embodiments, it should be understood that the present invention is not limited thereto. Those of ordinary skill in the art and having access to the teachings of this document will recognize additional modifications and embodiments within the scope of the present invention. Thus, the present invention is only limited by the claims appended herein.

Claims (9)

1. A compound selected from the group consisting of: where the disulfide bonds between the residues Cys 6 and Cys 11 of the A chain polypeptide and the disulfide bonds between Cys 7 and Cys 20 of the A chain with Cys 7 and Cys 19 of the B chain polypeptide, respectively, are represented by a solid line between them; where the binding fragments are covalently linked to the epsilonamino acid of the specified lysine residue, where the A-chain polypeptide for dimers 71, 72, 77 and 78 has the amino acid sequence shown in SEQ ID NO: 1; and the B chain polypeptide for dimers 71, 72, 77 and 78 has the amino acid sequence shown in SEQ ID NO: 2, respectively. 2. Composition for the treatment of diabetes, containing one or more compounds according to claim 1 and a pharmaceutically acceptable carrier. 3. A method of treating diabetes, comprising administering to a diabetic individual a therapeutically effective amount of a composition comprising an insulin receptor partial agonist according to claim 2. 4. The method of claim 3, wherein the diabetes is type 1 diabetes, type 2 diabetes, or gestational diabetes. 5. An insulin dimer according to claim 1, comprising a first B29 or B28 Lys of a first insulin heterodimer molecule having a first A-chain polypeptide and a first B-chain polypeptide, and a second B29 or B28 Lys of a second insulin heterodimer having a second A-chain polypeptide. chains and a second B chain polypeptide conjugated together with a bifunctional linker selected from the group consisting of: o □ ,—h o o o /--x o o A A<, )A A A* 7'^ A [ N-0 '---' ON JLN O '--- ^z ON I to LL b O or 0O 6. Compound having the following formula - 98 041034 7. A composition for the treatment of diabetes comprising one or more of the compounds of claim 6 and a pharmaceutically acceptable carrier. 8. A method of treating diabetes, comprising administering to an individual with diabetes a therapeutically effective amount of a composition comprising an insulin receptor partial agonist according to claim 7. 9. The method of claim 8, wherein the diabetes is type 1 diabetes, type 2 diabetes, or gestational diabetes.