Patent Application
20230272030
This application claims the priority of Chinese Patent Application No. 202010723972.9, filed with the China National Intellectual Property Administration on Jul. 24, 2020, and titled with “INSULIN-FC FUSION PROTEIN AND APPLICATION THEREOF”, which is hereby incorporated by reference in its entirety.
The present disclosure relates to the field of polypeptide drugs, in particular to an insulin-Fc fusion protein with enhanced insulin activity and prolonged in vivo half-life after being cleaved by site-specific protease, a preparation method thereof and an application thereof.
In recent years, the incidence of diabetes has been increasing year by year. For type I diabetes, blood glucose is controlled mainly by exogenous insulin; and for type 2 diabetes, insulin has become the main drug for blood glucose control as the disease progresses. Therefore, the use of insulin to treat diabetes has become an effective way.
Insulin therapy is necessary for patients with abnormal insulin secretion (type I) or insulin resistance (type II), and blood glucose levels can be normally regulated by insulin administration. However, like other protein and peptide hormones, insulin has a very short in vivo half-life and thus suffers from the disadvantage of repeated administration. Such frequent administration causes severe pain and discomfort to the patient. For this reason, many studies have been carried out on protein formulations and chemical conjugation (fatty acid conjugates, polyethylene polymer conjugates) in order to improve the quality of life by prolonging the in vivo half-life of proteins and reducing the frequency of administration. Commercially available long-acting insulins include insulin glargine (lantus, lasting about 20 hours to 22 hours) manufactured by Sanofi Aventis, and insulin detemir (levemir, lasting about 18 hours to 22 hours) and tresiba (insulin degludec, lasting about 40 hours) manufactured by Novo Nordisk. These long-acting insulin formulations do not produce peaks in blood insulin concentration, which makes them suitable as basal insulins. However, since these formulations do not have a sufficiently long half-life, there is still the disadvantage of being injected once a day or every two to three days. Thus, there are limitations in achieving the intended goal of once-weekly dosing frequency to improve convenience for diabetic patients requiring long-term insulin administration.
Patent publication CN103509118B discloses a single-chain insulin fused to the Fc region of an antibody. Although this insulin-Fc fusion protein has showed an improved half-life in in vitro experiments, it has low in vivo hypoglycemic activity and is not suitable for clinical use.
The success in controlling diabetes is highly correlated with the compliance of the patient being treated, and it is desirable to reduce the frequency of injections required. However, these existing modified insulin molecules are either very inactive and not suitable for clinical use, or very active and have a rapid hypoglycemic effect after administration to patients, resulting in the side effect of hypoglycemia. Therefore, there is an urgent need in the field for a novel long-acting insulin suitable for clinical use.
After extensive research, the inventors provide an insulin-Fc fusion protein, which can obtain enhanced insulin activity and prolonged in vivo half-life after being cleaved by site-specific protease, and it is surprisingly found that the fusion protein has steady and stable in vivo hypoglycemic effect, which can improve the safety of clinical medication and patient compliance, thereby better achieving blood glucose management and providing a better quality of life.
In a first aspect, the present disclosure provides an insulin-Fc fusion protein with enhanced insulin activity and prolonged in vivo half-life after being cleaved by site-specific protease, having the structure of formula (I):
X-E1-Y-E2-Z-L-Fc (I),
One or both of E1 and E2 are present and are an amino acid fragment comprising a site-specific protease cleavage site; E1 and E2 each may comprise 1-10 or more amino acids in length, such as 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 amino acids; if present at the same time, E1 and E2 may be cleaved by the same or different site-specific proteases, such as by the same site-specific protease; if Y is present, preferably both E1 and E2 are present; if Y is absent, preferably one of E1 and E2 is present; the site-specific protease cleavage site may be a cleavage site of Kex2 and/or Furin protease, such as a cleavage site of Kex2 protease.
L is a linker linking Z and Fc; L may be a polypeptide fragment, for example, L comprises a flexible unit (also referred to as a flexible peptide fragment herein) of one, two or more amino acids selected from Ala, Thr, Gly and Ser, such as a flexible unit consisting of G and S; L may also be a polypeptide fragment comprising a rigid unit (also referred to as a rigid peptide fragment herein).
In some embodiments, the rigid unit comprises or consists essentially of rigid amino acids, the rigid amino acids including but not limited to V, P, I, K and L.
In some embodiments, the rigid unit comprises one or more PPPX1LP (SEQ ID NO: 125), wherein X1 is any amino acid;
In other embodiments, the rigid unit comprises one or more X2APPPX1LP (SEQ ID NO: 126), wherein X1 is any amino acid and X2 is K or V.
In other embodiments, the rigid unit comprises a polypeptide fragment selected from the group consisting of:
In other embodiments, the rigid unit comprises a polypeptide fragment selected from the group consisting of:
In some embodiments, L comprises both rigid and flexible units, and may be more than two units.
Fc is the Fc region of an immunoglobulin; Fc may be derived from a human immunoglobulin; the Fc region may be an Fc region derived from IgG, IgA, IgD, IgE or IgM; preferably, the Fc region is an Fc region derived from IgG, such as an Fc region derived from IgG1, IgG2, IgG3 or IgG4; further preferably, the Fc region is an Fc region derived from IgG2; or compared to the sequence from which it is derived, the Fc region may have one or more substitutions, additions and/or deletions while still retains the ability to prolong half-life, for example, the Fc region is derived from human IgG and has a mutation that reduces or eliminates the binding to FcγR and/or a mutation that enhances the binding to FcRn, the mutation may be selected from the group consisting of: N297A, G236R/L328R, L234A/L235A, N434A, M252Y/S254T/T256E, M428L/N434S, T250R/M428L and a combination thereof; and the Fc region may be glycosylated or unglycosylated.
In some embodiments, for the fusion protein of the present disclosure, the insulin is selected from human insulin, bovine insulin or porcine insulin, preferably human insulin; for example, the A and B chains of insulin are derived from human insulin.
In some embodiments, Y, E1 and E2 are all present, or wherein Y is absent and one of E1 and E2 is present.
In other embodiments, the fusion protein comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 47-72.
In a second aspect, the present disclosure provides an insulin-Fc fusion protein with a structure of Ins-L-Fc. In some embodiments, the C-peptide may be removed from the fusion protein of the first aspect of the present disclosure by a specific protease to produce the fusion protein of the second aspect of the present disclosure. In some embodiments, the insulin-Fc fusion protein exists in the form of a homodimer, the structural diagram of which is shown in
Wherein, Ins is an insulin moiety providing insulin activity and comprises A and B chains of insulin linked by a covalent bond and located in different peptide chains; the covalent bond is preferably a disulfide bond.
L is a linker linking Z and Fc; L may be a polypeptide fragment (also referred to as a linking peptide in some embodiments herein), for example, L comprises a flexible unit of one, two or more amino acids selected from Ala, Thr, Gly and Ser; L may also be a polypeptide fragment comprising a rigid unit.
In some embodiments, L comprises one or more rigid units comprising or consisting essentially of rigid amino acids, the rigid amino acids including but not limited to V, P, I, K and L.
In other embodiments, the rigid unit comprises one or more PPPX1LP (SEQ ID NO: 125), wherein X1 is any amino acid.
In other embodiments, the rigid unit comprises one or more X2APPPX1LP (SEQ ID NO: 126), wherein X1 is any amino acid and X2 is K or V.
In other embodiments, the rigid unit comprises a polypeptide fragment selected from the group consisting of:
In other embodiments, the rigid unit comprises a polypeptide fragment selected from the group consisting of:
Fc is the Fc region of an immunoglobulin; Fc may be derived from a human immunoglobulin; the Fc region may be an Fc region derived from IgG, IgA, IgD, IgE or IgM; preferably, the Fc region is an Fc region derived from IgG, such as an Fc region derived from IgG1, IgG2, IgG3 or IgG4; further preferably, the Fc region is an Fc region derived from IgG2; or compared to the sequence from which it is derived, the Fc region may have one or more substitutions, additions and/or deletions while still retains the ability to prolong half-life, for example, the Fc region is derived from human IgG and has a mutation that reduces or eliminates the binding to FcγR and/or a mutation that enhances the binding to FcRn, the mutation is selected from the group consisting of: N297A, G236R/L328R, L234A/L235A, N434A, M252Y/S254T/T256E, M428L/N434S, T250R/M428L and a combination thereof; and the Fc region may be glycosylated or unglycosylated.
In some embodiments, the insulin is selected from human insulin, bovine insulin or porcine insulin, preferably human insulin; for example, the A and B chains of the insulin are derived from human insulin.
In other embodiments, L comprises CTP, for example, 1, 2, 3 or more CTPs.
In a third aspect, the present disclosure provides a method for producing an insulin-Fc fusion protein with enhanced insulin activity and prolonged half-life, comprising contacting the fusion protein described in the first aspect of the present disclosure with a site-specific protease capable of cleaving the site-specific protease cleavage site, preferably the site-specific protease is Kex2 and/or Furin protease.
In some embodiments, the insulin-Fc fusion protein with enhanced insulin activity and prolonged in vivo half-life of the present disclosure is obtained by the above method.
In a fourth aspect, the present disclosure provides a polynucleotide encoding the fusion protein, preferably the polynucleotide is an expression vector capable of expressing the fusion protein.
In a fifth aspect, the present disclosure provides a cell capable of expressing an insulin-Fc fusion protein, comprising the above-described polynucleotide.
In a sixth aspect, the present disclosure provides a method for producing an insulin-Fc fusion protein, comprising culturing the cells described in the fifth aspect of the present disclosure under conditions for expressing the insulin-Fc fusion protein; preferably further comprising contacting the insulin-Fc fusion protein with a site-specific protease capable of cleaving the site-specific protease cleavage site, wherein the culturing and the contacting may be performed simultaneously or separately. The method may also comprise a protein purification step to obtain the target fusion protein.
In a seventh aspect, the present disclosure provides a method for characterizing the structure of an insulin-Fc fusion protein, comprising detecting the deglycosylated molecular weight of the fusion protein and characterizing disulfide bonds.
In an eighth aspect, the present disclosure provides a pharmaceutical composition comprising the fusion protein described in the first and third aspects, the polynucleotide described in the fourth aspect or the cell described in the fifth aspect.
In a ninth aspect, the present disclosure provides a method for lowering blood glucose and/or treating diabetes, comprising administering the fusion protein described in the first and second aspects, the polynucleotide described in the fourth aspect or the cell described in the fifth aspect to a subject in need thereof, preferably the diabetes is type I or type II diabetes. Furthermore, when administering the fusion protein described in the first aspect of the present disclosure, additional administration of appropriate site-specific protease, or utilization of appropriate site-specific proteases present in the body, may also be considered.
Corresponding to the above methods for lowering blood glucose and/or treating diabetes, the present disclosure also provides use of the fusion protein, polynucleotide or cell in the manufacture of a medicament for lowering blood glucose and/or treating diabetes. The present disclosure also provides the fusion protein, polynucleotide or cell for lowering blood glucose and/or treating diabetes.
Next, the present disclosure will be described in more detail in conjunction with the embodiments. It is apparent that the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present disclosure.
Insulin is a hormone secreted by pancreatic β cells to promote glucose uptake and inhibit fat degradation, thus acting to control blood glucose levels. In the nucleus of β cells, the DNA of the insulin gene region on the shorter arm of Chromosome 11 is transcribed into mRNA, and the mRNA moves from the nucleus to the endoplasmic reticulum in the cytoplasm, and is translated into preproinsulin, which consists of 106 amino acid residues and contains a signal peptide of about 20 residues at the N-terminal. When preproinsulin passes through the endoplasmic reticulum membrane, the signal peptide is removed by signal peptidase to form a long peptide chain, proinsulin, consisting of 86 amino acids. Proinsulin is cleaved by proteolytic enzymes in the Golgi apparatus to cut off two arginine residues at positions 31 and 32, a lysine residue at position 64 and an arginine residue at position 65. The cleaved chain is called the C-peptide serving as a linking moiety, and the simultaneously produced insulin is secreted out of β cells into the blood circulation. A small part of proinsulin that has not been hydrolyzed by protease enters the blood circulation along with insulin. Proinsulin has almost no biological activity, only 5%-10% of insulin.
The “insulin” of the present disclosure includes not only naturally occurring insulin, but also functional variants of insulin. The functional variant refers to a polypeptide that is obtained by modifications, such as additions, deletions and/or substitution of one or more amino acids, to the native sequence and/or structure of insulin and still has insulin activity (regulating blood glucose levels in the body). The substitution, addition or deletion of an amino acid may be a naturally occurring mutant form or an artificially modified mutant form for specific purposes. It is well known to those of ordinary skill in the art that in practice, functional variants of insulin are often also referred to as insulin. Another example is the insulin analogs disclosed in CN105636979 B and CN 201480006998. With reference to this specification, this practice is also covered herein.
From another aspect, a functional variant of insulin refers to a polypeptide that has at least 80% (preferably 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100%) amino acid sequence homology to natural insulin, and still has insulin activity. For some functional variants, chemical substitutions (e.g., α-methylation, α-hydroxylation), deletions (e.g., deamination), or modifications (e.g., N-methylation) are possible at some groups of specific amino acid residues.
Those skilled in the art are familiar with the methods of preparing functional variants of insulin and methods of testing their effects, and the insulin analogs that have been marketed include, for example, insulin lispro (Eli Lilly), insulin aspart (Novo Nordisk), insulin glulisine (Aventis), insulin glargine (Sanofi), insulin detemir (Novo Nordisk), and insulin degludec (Novo Nordisk).
For insulin lispro, proline at position 28 and lysine at position 29 on the B chain of human insulin are reversed, and the other amino acid sequence and structure remain unchanged. As a result of the reversal of the two amino acids, the function of insulin has not been changed, but the insulin, which used to form dimers and hexamers easily, no longer aggregates easily into dimers and hexamers, but exists in the form of monomers. Therefore, it will be easily absorbed after subcutaneous injection, resulting in a rapid onset of action.
Insulin aspart is also a fast-acting insulin, in which the proline at position B28 of human insulin is substituted by aspartic acid, so that this insulin analog is less prone to aggregate as a hexamer, which makes it easily absorbed subcutaneously for rapid action.
Insulin glulisine uses lysine instead of asparagine at position B3 and glutamic acid instead of lysine at position B29 to achieve a rapid onset of action.
Insulin glargine (Lantus) differs from human insulin in that 1) the aspartic acid at position 21 of the A chain is substituted by glycine; 2) two arginine residues are added to the C-terminal of the B chain. The result of such changes are as follows: the substitution at position A21 by glycine leads to a more stable binding of hexamer, and in the neutral environment of the subcutaneous tissue, the solubility decreases to form precipitate, resulting in slow absorption, similar to the peakless secretion of basal insulin, which is suitable for long-acting treatment, and its action time will be further prolonged if a small amount of zinc is added; the addition of two arginine residues to the C-terminal of the B chain changes the isoelectric point of insulin, rising from the original pH 4.5 to pH 6.7, which allows the formation of micro-precipitates in the neutral environment of subcutaneous tissues and prolongs the decomposition, absorption and action time of insulin.
For insulin detemir (Levermir), which is developed and produced by Novo Nordisk, structurally, the amino acid at position B30 is deleted, and a 14-carbon free fatty acid chain of N-16-alkanoic acid group is linked at the lysine at position B29. In the drug solution with zinc ions, the insulin molecule still exists in the form of hexamer. The modification of the fatty acid chain leads to slow subcutaneous absorption, and the insulin detemir in the plasma will bind to the albumin in the plasma due to the presence of the fatty acid, while only free insulin detemir can play a hypoglycemic effect, which also prolongs the action time of insulin.
For insulin degludec, the threonine at position B30 is deleted, and a 16-carbon fatty diacid side chain is linked at the lysine at position B29 via a glutamic acid linker. Under the action of phenol and zinc ions, insulin degludec aggregates into double hexamers in the preparation. After subcutaneous injection, with the diffusion of phenol and the slow release of zinc ions, insulin degludec monomer can be slowly and continuously released, and then absorbed into the blood. Based on the above characteristics, insulin degludec has an ultra-long action time in diabetic patients with a half-life of about 25 hours.
The fusion protein described herein refers to both a protein formed by amino acids linked by peptide bonds and a protein formed from two or more peptide chains linked by disulfide bonds.
The “insulin-Fc fusion protein” in the present disclosure refers to a fusion protein formed by insulin (including functional variants thereof) and the Fc region of an immunoglobulin, and is sometimes simply referred to as “fusion protein” herein. In addition, in order to distinguish between the insulin-Fc fusion proteins before and after the cleavage of the linking peptide moiety by enzyme, the fusion protein before the cleavage by enzyme is sometimes referred to as “insulin precursor-Fc fusion protein”, and the corresponding “insulin-Fc fusion protein” used refers to the fusion protein after the cleavage of the linking peptide moiety by enzyme. However, it is more common herein to not specifically distinguish between the fusion proteins before and after the cleavage by enzyme, in which case the fusion protein or insulin-Fc fusion protein encompasses its forms both before and after the cleavage by enzyme. In addition, when it is clear from the context which form the fusion protein refers to, “fusion protein” or “insulin-Fc fusion protein” is often used directly to refer to this form.
The sequence of A chain in natural human insulin is:
The sequence of B chain in natural human insulin is:
The fusion protein described herein may also comprise an additional sequence that prolongs in vivo half-life, and for example, the additional sequence is selected from one or more of Fc, CTP (C-terminal peptide), XTEN, SABA (serum albumin binding adnectin) and PAS. The additional sequence may be located at the terminal, linker or other positions in the fusion protein. For simplicity, the structural formulae X-E1-Y-E2-Z-L-Fc and Ins-L-Fc used herein encompass also these cases where the additional sequence is located at other positions.
During the in vivo processing of insulin, the linking peptide linking the A and B chains of insulin is C-peptide. C-peptide includes both its naturally-occurring sequence and a variant form with the same function formed by substitution, deletion or addition of one or more amino acids based on the naturally-occurring sequence.
As a reference, the sequence of C-peptide of human insulin in its natural form is:
In the insulin-Fc fusion protein of the present disclosure, the linking peptide is not limited to the C-peptide of natural insulin or the variant/fragment thereof, but can also be any other suitable polypeptide linking the A and B chains of insulin. In some embodiments, the linking peptide may comprise 1-100 or more amino acids in length, such as 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, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 50, 60, 70, 80, 90, 100 amino acids, or a value between any two of the values above.
In some embodiments, the sequence of the linking peptide is:
In other embodiments, the linking peptide may be in the form of a polypeptide of any length:
Human immunoglobulin IgG is composed of four polypeptides (two identical copies of light chain and heavy chain) covalently linked by disulfide bonds. The proteolysis of IgG molecules by papain produces two Fab fragments and one Fc fragment. The Fc fragment is composed of two polypeptides linked together by disulfide bonds. Each polypeptide, from N- to C-terminal, consists of hinge region, CH2 domain and CH3 domain. The structure of the Fc fragment is almost the same in all subtypes of human immunoglobulin. IgG is one of the most abundant proteins in human blood, which constitutes 70% to 75% of total immunoglobulin in human serum.
The Fc region of immunoglobulin is safe to be used as a pharmaceutical carrier because it is a biodegradable polypeptide that can be metabolized in the body. In addition, compared with the entire immunoglobulin molecule, the Fc region of immunoglobulin has a relatively low molecular weight, which is beneficial to the preparation, purification and production of fusion proteins. Since the immunoglobulin Fc region does not contain Fab fragment (its amino acid sequence varies according to the antibody subclass and is therefore highly heterogeneous), it is expected that the immunoglobulin Fc region can greatly increase the homogeneity of the substance and have low antigenicity
It is generally understood by those of ordinary skill in the art that the term “Fc region of an immunoglobulin” refers to a protein fragment comprising heavy chain constant region 2 (CH2) and heavy chain constant region 3 (CH3) of an immunoglobulin but not comprising the variable regions of the heavy and light chains of an immunoglobulin. It may also contain the hinge region in the heavy chain constant region. Furthermore, the Fc fragment used in the present disclosure may contain part or all of the Fc region containing heavy chain constant region 1 (CH1) and/or the light chain constant region 1 (CL1) without variable regions of heavy chain and light chain, as long as it has a physiological function that is basically similar to or better than that of natural protein. Moreover, it may be an Fc fragment with a relatively long deletion in the amino acid sequences of CH2 and/or CH3. For example, the immunoglobulin Fc region used in the present disclosure may comprise 1) CH1 domain, CH2 domain and CH3 domain; 2) CH1 domain and CH2 domain; 3) CH1 domain and CH3 domain; 4) CH2 domain and CH3 domain; 5) CH1 domain, CH2 domain, CH3 or CL domain; 6) the combination of one or more constant region domains with (part or all of) the immunoglobulin hinge region; or 7) the dimer of any domains of heavy chain constant region and light chain constant region. In summary, the Fc region of an immunoglobulin in the present disclosure refers to any form of Fc or variants/derivatives thereof comprising one or more constant region domains of heavy/light chain or variants thereof and capable of imparting a function of prolonging in vivo half-life to the fusion protein, such as a single chain Fc, a monomeric Fc.
Besides, the immunoglobulin Fc region of the present disclosure comprises natural amino acid sequence and sequence variants (mutants) thereof. Owing to one or more deletions, insertions, non-conservative or conservative substitutions, or combination thereof of amino acid residues, the amino acid sequence derivative may have a sequence different from the natural amino acid sequence. For example, for IgG Fc, amino acid residues at positions 214 to 238, 297 to 299, 318 to 322, or 327 to 331 that are known to be critical to binding can be used as suitable targets for modification. The immunoglobulin Fc region of the present disclosure may also comprise a variety of other derivatives, including those without the region capable of forming disulfide bonds, those having several amino acid residues deletion at the N-terminal of the natural Fc, or those having additional methionine residues to the N-terminal of the natural Fc. In addition, to get rid of effector functions, deletion may be designed at complement binding site, such as C1q binding site and ADCC site. Techniques for preparing such derivatives of the immunoglobulin Fc region are disclosed in WO 97/34631 and WO 96/32478, which are incorporated herein by reference in their entirety. In addition, it is well known to those of ordinary skill in the art that the mutation of one or more amino acids in the Fc region can enhance the affinity of Fc to FcRn and prolong half-life in serum, such as the T250Q/M428L mutation (CN 1798767 B), and these mutant forms of Fc regions are also within the meaning of the Fc region of the present disclosure.
For proteins and peptides, amino acid substitutions that generally do not change the molecular activity are known in the art (H. Neurath, R. L. Hill, The Proteins, Academic Press, New York, 1979). The most common substitutions are Ala/Ser, Val/Ile, Asp/Glu, Thr/Ser, Ala/Gly, Ala/Thr, Ser/Asn, Ala/Val, Ser/Gly, Thy/Phe, Ala/Pro, Lys/Arg, Asp/Asn, Leu/Ile, Leu/Val, Ala/Glu and Asp/Gly, in either way.
If necessary, the Fc region is allowed to be modified, such as phosphorylation, sulfation, acrylate, glycosylation, methylation, farnesylation, acetylation, and amidation.
The Fc derivatives have the same biological activity as the Fc region in the present disclosure or have improved structural stability (such as structural stability to heat, pH, etc.) than the corresponding Fc region thereof.
In addition, these Fc regions can be derived from natural forms isolated from human and other animals including cattle, goat, pig, mouse, rabbit, hamster, rat and guinea pig, or derived from recombinant or derivative of transformed animal cells or microorganisms. Herein, the Fc region can be obtained from natural immunoglobulin by separating intact immunoglobulin from human or animal organisms and treating them with proteolytic enzymes. Papain digests natural immunoglobulin into Fab and Fc regions, while pepsin treatment results in the production of pFc′ and F(ab′)2 fragments. Fc or pFc′ fragments can be isolated, e.g., by size exclusion chromatography.
In addition, the immunoglobulin Fc region of the present disclosure may be a form having natural sugar chains, or increased or reduced sugar chains compared to the natural form, or may be a deglycosylated form. The increase, decrease or removal of immunoglobulin Fc sugar chains can be accomplished by methods commonly used in the art, such as chemical methods, enzymatic methods, genetic engineering methods or methods of mutating the N297 glycosylation site. Removal of sugar chains from the Fc fragment results in a significant reduction in binding affinity to complement (C1q) and reduction or loss of antibody-dependent cell-mediated cytotoxicity or complement-dependent cytotoxicity, and thereby unnecessary in vivo immune responses will not be induced. In view of this, the immunoglobulin Fc region in deglycosylated or unglycosylated form may be more suitable for the purpose of the present disclosure for use as a medicament.
The term “deglycosylation” as used in the present disclosure means the enzymatic removal of carbohydrate moiety from the Fc region, and the term “unglycosylation” means that the Fc region is produced in an aglycosylated form by prokaryotes (preferably E. coli), or by a method of mutating the N297 glycosylation site to G, A or any other amino acid.
In addition, the immunoglobulin Fc region may be an Fc region derived from IgG, IgA, IgD, IgE, and IgM, or prepared by a combination or hybrid thereof. Preferably, it is derived from IgG or IgM (two of the most abundant proteins in human blood), most preferably IgG (which is known to extend the half-life of ligand-binding protein)
The term “combination” as used in the present disclosure means a dimer or a multimer formed by two or more single-chain polypeptides which are linked together, where the single-chain polypeptides can be derived from the same or different immunoglobulin Fc region. That is, the dimer or the multimer may be formed by two or more fragments selected from the group consisting of IgG Fc fragment, IgA Fc fragment, IgM Fc fragment, IgD Fc fragment, and IgE Fc fragment.
Proinsulin is inactive or very low in activity, and the conventional process for preparing recombinant insulin in the prior art is to express protein by Escherichia coli or yeast, and then process the expressed protein into an active molecule with trypsin or trypsin plus carboxypeptidase B. However, when the Fc region of the antibody is used to form a conjugate with insulin, the conventional preparation process cannot be used because there are many trypsin cleavage sites on the Fc, which will be cleaved and become inactive during processing proinsulin into an active molecule. In the prior art, in order to avoid this problem, single-chain insulin is directly conjugated with the Fc region. However, the inventors have found through research that such insulin has very low in vivo activity.
The inventors unexpectedly found that if the mature mechanism of insulin in vivo is simulated and the insulin conjugate is prepared which has a more similar structure to natural insulin (the A and B chain in the mature molecule are linked by disulfide bonds) and is linked to an Fc region, the activity of insulin can be greatly improved. After extensive screening, the inventors found that an active long-acting insulin conjugate molecule can be obtained by preparing the fusion polypeptide with the structure of the present disclosure, introducing a protease cleavage site of Kex2 or Furin protease, and then processing with the protease.
The Kex2 protease described in the present disclosure is a calcium ion-dependent protease, which can specifically recognize and cleave the carboxyl-terminal peptide bond of bibasic amino acids such as Arg-Arg and Lys-Arg. Unlike trypsin, Kex2 cannot recognize and cleave the carboxy-terminal peptide bond of a single basic amino acid, namely arginine or lysine. The Kex2 protease is responsible for processing precursors of killer toxin and α-factor in yeast. The activity of Kex2 protease is not inhibited by conventional serine protease inhibitors such as aprotinin, PMSF and TPCK.
Furin described in the present disclosure is an important endoprotease in eukaryotic cells. It is located in the network outside the Golgi apparatus and is a major protein convertase in the exocrine pathway, which can recognize specific amino acid sequences, and cleaves and processes the precursors of many important polypeptides and proteins in the secretory pathway to make them biologically active after activated by two times of self-cleavage in the endoplasmic reticulum-Golgi apparatus. It is named because its encoding gene (fur) is located upstream of the proto-oncogene fes/feps. Specifically, furin catalyzes and cleaves the carboxy-terminal peptide bond of Arg-Xaa-Yaa-Arg (Xaa is any amino acid and Yaa is Arg or Lys) in the proprotein to produce a mature protein.
After the fusion polypeptide of the present disclosure is processed with protease, the linking peptide between the A chain and the B chain is removed, so that disulfide bonds are formed between the A chain and the B chain in a manner similar to natural insulin. For example, two disulfide bonds are formed by the sulfhydryl groups in four cysteines, A7 (Cys)-B7(Cys) and A20 (Cys)-B19 (Cys), to link the two chains A and B. In addition, a disulfide bond is also preferably formed by A6 (Cys) and A11 (Cys) inside the A chain. The inventors surprisingly found that even if the A chain or the B chain is linked to the Fc region, it does not affect the formation of correct disulfide bond linking and spatial folding between the A chain and the B chain to form the insulin fused to the Fc region, thereby accomplishing the present disclosure.
In the fusion protein of the present disclosure, the function of the linker L is to link the A chain or B chain of insulin with the Fc region. The linker L may be a polypeptide or a chemical structure other than a peptide chain.
In some embodiments, the linker is a polypeptide comprising a flexible unit (flexible peptide fragment) consisting essentially of A, T, G and/or S, such as a flexible unit consisting of G and S; the flexible unit may comprise 2-50 or more amino acids in length, such as 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 35, 40, 45 or 50 amino acids.
In other embodiments, the linker is a polypeptide comprising a rigid unit (rigid peptide fragment) consisting essentially of rigid amino acids including but not limited to V, P, I, K, and L.
The insulin-Fc fusion protein is fermented and secreted by CHO cells. After transcription and translation in CHO cells, the fusion protein undergoes a series of processing comprising post-translational modifications such as proline hydroxylation, O-glycosylation, N-glycosylation, deletion of lysine at C-terminal and the like, and such modifications occur on sequences other than the B and A chains of insulin. Besides, the insulin-Fc fusion protein also forms disulfide bonds in the organelles of CHO cells to stabilize its structure.
The disulfide bond of the insulin-Fc fusion protein is formed between two cysteine (Cys) residues. Its disulfide bonds can be divided into two parts according to the position with some in insulin and others in Fc. The disulfide bonds of insulin are located in the B and A chains, and the amino acids of the B and A chains are represented by position (X) in order from the N-terminal to the C-terminal, which are BX and AX, respectively. In some embodiments, the disulfide bonds are CysA7-CysB7, CysA20-CysB19 and CysA6-CysA11. The Fc region consists of two single chains with the same amino acid sequence, and in some embodiments, there are two disulfide bonds in each single chain and two interchain disulfide bonds between the two single chains, meaning that there are 6 disulfide bonds in Fc.
UPLC-QTOF is a conventional instrument for analyzing the structure of biological macromolecules. Its main functional modules are UPLC and QTOF. After being separated by UPLC, the sample to be tested enters the ion source in the state of solution to be ionized and becomes charged ions, which enter the mass analyzer QTOF under the action of an accelerating electric field. Under the action of electric and magnetic fields, the m/z of various ions are captured by two mass spectrometers of triple quadrupole (Q) and time-of-flight mass spectrometry (TOF). The software calculates the precise molecular weight, and finally realizes the structure analysis of complex biological macromolecular proteins. The present disclosure adopts UPLC-QTOF, a commonly used instrument with high resolution and high sensitivity, as an ideal method for analyzing fusion proteins, and mainly analyzes and characterizes the deglycosylation of the fusion protein, its molecular weight after deglycosylation reduction, disulfide bonds and disulfide bond mismatch rate.
The results show that in some embodiments, the insulin-Fc fusion protein has a molecular weight and disulfide bonds consistent with the theory, a low mismatch rate, and post-translational modifications such as proline hydroxylation, O-glycosylation, N-glycosylation, deletion of lysine at C-terminal and the like.
In this example, the construction method of the insulin precursor fusion protein is mainly described. Herein, the insulin precursor fusion protein is sometimes also referred to as insulin fusion protein and has a molecular form of proINS-L-Fc. It may be secreted and expressed in yeast or eukaryotic cells (such as CHO, HEK293, etc.), and the expressed protein exists in the form of homodimer. In order to assist protein to be secreted and expressed, a signal peptide and/or propeptide can be added to the N-terminal of the protein. The signal peptide includes but is not limited to the sequences shown in Table 1 below.
proINS refers to a natural insulin precursor or an analog thereof derived from human or otherwise. The analog includes inserted, deleted, truncated or mutated insulin precursors, such as A14E\B16E\B25H\desB30 variant, A14E\B16H\B25H\desB30 variant or A14E\desB30 variant. The analog may reduce the immunogenicity of insulin, or reduce proteolysis to improve the stability of insulin, or reduce the affinity of insulin to insulin receptor (IR) to prolong the in vivo half-life and the like. It can also be used for any other purpose.
The insulin precursor of this example can be processed into mature insulin by proteases such as Kex2, Furin, trypsin and the like. The insulin precursor of this example can also promote the correct folding and processing of the protein through the optimized C-peptide. The analog of the insulin precursor used in this example includes but is not limited to those shown in Table 3 below.
L represents the linker between proINS and Fc and can consist of amino acids of 0 to any number in length. It can be either a flexible polypeptide or a rigid polypeptide. L can assist the two insulin molecules linked to the Fc homodimer to form correct spatial structures, respectively. Preferably, L has a sequence including but not limited to the sequences shown in Tables 4 and 5.
Fc is preferably derived from human IgG; more preferably human IgG and variants thereof without ADCC and CDC activities, such as IgG2 and IgG4; more preferably mutated human IgG with prolonged half-life. Fc may also be a fragment of Fc or a fusion of Fc with other proteins/protein fragments. The Fc used in the present disclosure includes but is not limited to the following sequences.
The amino acid sequence features of some insulin precursor fusion proteins constructed in the present disclosure are shown in Table 6.
The insulin precursor fusion protein can be converted into a mature insulin fusion protein after processed by proteases such as Kex2, Furin, trypsin, etc. to remove sequences such as C-peptide and the like. In all the examples of the present patent, the protein cleaved and processed by enzyme is named by adding the suffix M (mature) to the name of the precursor protein. For example, after the insulin precursor fusion protein SS302-002 is processed by the protease Kex2, the mature protein is named as SS302-002M. The amino acid sequences of the mature insulin fusion proteins obtained by some insulin precursor fusion proteins of the present disclosure processed by protease are as follows.
According to the method described in the “Molecular Cloning: A Laboratory Manual (Third Edition)”, the expression vector of insulin precursor fusion protein was constructed.
The sequence of each insulin precursor fusion protein was optimized based on the codon preference of CHO cells.
After gene synthesis of the optimized DNA sequence, it was cloned into a eukaryotic expression vector pFRL3.0 or pTS1 by virtue of HindIII and EcoRI sites. The pFRL3.0 vector comprises the dihydrofolatereductase (DHFR) gene and can achieve high-level protein expression through the co-amplification of DHFR and the target gene. The CHO cells transfected with the vector was screened under MTX to establish a stably expressed cell line. pTS1 is a transient transfection plasmid without screening marker, and can quickly obtain a small amount of insulin precursor fusion protein for early molecular identification. The schematic diagrams of the expression vectors of the insulin precursor fusion protein are shown in
The plasmids expressing the insulin precursor-Fc fusion protein prepared in Example 1 were transfected into human embryonic kidney cell HEK-293 to transiently express the target protein. HEK-293 cells were thawed and cultured in cell culture shaker flasks for passage culture at a density of 1.0×106 cells/mL with a culture medium of OPM-293 CD05 Medium (Shanghai OPM Biosciences Co., Ltd.) under the culture conditions of 37° C., 120 rpm and CO2. The cells were passaged every two days, and could be used for transient transfection after one week of culture. The cell density was adjusted before transfection to make the cell density of about 4.0×106 cells/ml on the day of transfection. The plasmid was transiently transfected into HEK-293 cells using the FectoPRO kit (Polyplus Transfection), with a ratio of DNA to FectoPRO® Reagent of 1:1 (μg/μL), that is, 1 μg of DNA transfected per milliliter of cells corresponding to 1 μL of FectoPRO® Reagent. The plasmid was diluted with Opti-MEM (Gibco) at room temperature in an amount of 10% of the total volume of the transient transfection system, and mixed well by shaking. The diluted plasmid was added to the centrifuge tube of FectoPRO® Reagent at one time, mixed well immediately, and incubated at room temperature for 10 min. The prepared plasmid and transfection reagent mixture were added to the density-adjusted HEK-293 cell suspension at one time and mixed well. Then the cell culture shaker flask was placed in an incubator under the culture conditions of 37° C., 5% CO2, and a shaker speed of 120 rpm. After the cells were transfected and cultured for 4 hours, Volume of FectoPRO® Booster was added at 0.5 μL per milliliter of cells. After 24 hours of culture, the culture conditions were changed to 31° C., 5% CO2 and 120 rpm for fermentation. After 3-5 days of culture, when the cell viability was less than 90%, the supernatant was harvested by centrifugation (3000 rpm), detected for expression level, and then purified to obtain the target protein.
The plasmids partially expressing the insulin precursor-Fc fusion protein prepared in Example 1 were transfected into Chinese hamster ovary cells (CHO DG44) (Invitrogen) to construct stably expressing cell lines, from which high-yielding cell lines were selected for fed-batch culture to prepare the target protein.
The host cell DG44 was thawed and cultured with complete medium containing CDM1N (Shanghai OPM Biosciences Co., Ltd.) plus 1% HT (Invitrogen) under the culture conditions of 37° C., 5% CO2, and a shaker speed of 120 rpm. A certain amount of cell suspension was taken up aseptically with a pipette every day for counting. When the cell density reached 3×106-4×106 cells/mL, cells were passaged, and the initial density of the passaged cell was maintained at about 1×106 cells/mL. When the total amount of cells met the transfection requirements, cells were harvested for electroporation. The host cells (CHO DG44) were transfected by electroporation using a Bio-Rad electroporator. A 4 mm electroporation cup was used for electroporation, and the specific electroporation parameters were as follows: voltage of 290V, pulse length of 20 milliseconds, and the number of electroporation of 1 time. 1×107 cells were subjected to electroporation at a time, and 40 μg of plasmid was used at a total volume of 0.8 mL. After electroporation, cells were transferred into 15 mL of recovery medium (CDM1N+1% HT), and cultured statically in a cell culture dish for 48 hours. After 48 hours, cells were centrifuged, resuspended in screening medium (CDM1N+100 nM MTX), and diluted to about 1×104 cells/mL. Then the diluted cells were inoculated in a 96-well plate at 100 μL/well, and placed in an incubator for static culture at 37° C. and 5% CO2. After 5 days of culture, cells were supplemented with 50 μL of the screening medium. When the clone confluence rate reached 80% or more, the expression level was analyzed by dot blotting, in which the antibody was HRP-labeled goat anti-human IgG antibody. The clones with high expression level were screened out, transferred from 96-well plates to 24-well plates for continuous culture, and supplemented with 1 mL of the screening medium. The screening and amplification of high-yielding clones in 12-well plates and 10 cm cell culture dishes were carried out using the same method.
To increase the yield of the fusion protein, cells were cultured with increasing MTX concentrations. The co-amplification of DHFR gene and the fusion protein gene was achieved through the inhibition of DHFR gene by MTX. In the screening process, methods known to those of ordinary skill in the art were used. For example, the details can be referred to: 1. Yang Wei, Wang Di, Chen Keqing, et al. Selection of electroporation transfection conditions of plasmid [J]. Journal of Huazhong University of Science and Technology, 2009, 38 (6): 858-860.; 2. Gu Xin, Li Yan. Discussion on the method of electroporation of mammalian cells DG44-CHO [J]. Biotechnology Letters, 2008, 19(1):87; 3. Jun, Kim, Baik, Hwang, Lee: Selection strategies for the establishment of recombinant Chinese hamster ovary cell line with dihydrofolate reductase-mediated gene amplification. Appl Microbiol Biotechnol. 2005, 69 (2): 162-169. 10.1007/s00253-005-1972-8.
After screening the clone pool in the 10 cm cell culture dish, the high-yielding clones were transferred to cell culture shaker flasks for culture at 37° C., 5% CO2 and a shaker speed of 120 rpm. After the high-yielding cell clones grew to a certain number, a part of the cells were collected for cryopreservation, and the remaining cells were subjected to fed-batch culture, during which cells were inoculated at a density of 1×106 cells/ml and placed in cell culture shaker flasks for culture at 37° C., 5% CO2 and a shaker speed of 120 rpm. After inoculation, cells were taken every day for counting to record the cell density and viability. Feeding was started from the 3rd day of culture, once a day. On the 3rd to 8th day, the feeding amount was 2%, 3%, 4%, 3%, 3% and 3% of the initial volume, respectively, and from the 9th day, the feeding amount was 2%, with the total feeding ratio of 20%˜30%. Glucose was supplemented once a day to maintain the glucose concentration in the culture system at 3-4 g/L. The culture period was 12-14 days. After the culture, the supernatant was harvested by centrifugation (3000 rpm), detected for expression level, and then purified to obtain the target protein.
Each insulin precursor-Fc fusion protein (SS302-002, SS302-004, SS302-005, SS302-008, SS302-012, SS302-014, SS302-015, SS302-019, SS302-029, SS302-030 and SS302-035) expressed in Example 2 of the cell fermentation solution was captured by affinity chromatography after removing cell debris by centrifugation and filtration through a 0.22 μm filter membrane. Bestchrom's protein A was used as an affinity medium. The protein A chromatography column was equilibrated using 3-5 times the column volume of buffer (20 mM Na2HPO4-citric acid, pH 7.5) to elute to a stable baseline, and then the treated supernatant of the fermentation solution was loaded on the column (loading capacity of 3-8 g/L). After the loading was completed, the impurity protein was washed to the baseline with washing buffer (20 mM Tris, 1.5 M NaCl, 2 M Urea, pH 7.5), and finally the column was eluted using elution buffer of 20 mM Na2HPO4-citric acid and 0.4M Arg with pH 3.5. The samples were collected separately according to the reading of the UV detector, starting from when the absorption value at UV280 nm was higher than 0.15 AU and stopping lower than 0.20 AU again. The collected samples were immediately added with 2.0 mol/L Tris-HCl buffer and stirred slowly to adjust the pH of the samples to 6.5-7.0. Then the samples were stored at −80° C. for subsequent SDS-PAGE analysis (
The SDS-PAGE results are shown in
The protein captured in step 1 was subjected to buffer exchange with G25 using a buffer of 50 mM Tris, 150 mM NaCl, pH 8.0. After the buffer exchange, each protein was cleaved with Kex2 to remove the C-peptide to obtain insulin-Fc fusion proteins. The cleavage conditions of SS302-002 and SS302-004 were as follows: the final protein concentration of 1 mg/mL, the feeding ratio (mass ratio) of 200:1 (precursor: Kex2), the final concentration of CaCl2) of 20 mM/L, and the total reaction volumes of 5 mL and 3 mL, respectively, and the cleavage was performed in a water bath at 37° C. for 6 h. The cleavage conditions of the three proteins SS302-008 and SS302-012 were as follows: the final protein concentration of 1 mg/mL, the feeding ratio (mass ratio) of 50:1 (precursor: Kex2), the final concentration of CaCl2) of 20 mM/L, and the total reaction volume of 190 mL, and the cleavage was performed in a water bath at 37° C. for 6 h. The cleavage conditions of SS302-014, SS302-015, SS302-019, SS302-029, SS302-030 and SS302-035 were as follows: the final protein concentration of 1 mg/mL, the feeding ratio (mass ratio) of 1:25 (Kex2: precursor), the final concentration of CaCl2) of 20 mM/L, and the total reaction volume of 60-180 mL (varying slightly for different proteins), and the cleavage was performed in a water bath at 37° C. for 6 h. The insulin-Fc fusion proteins after cleavage of each insulin precursor-Fc fusion by protease were named as S302-002M, SS302-004M, SS302-005M, SS302-008M, SS302-012M, SS302-014M, SS302-015M, SS302-019M, SS302-029M, SS302-030M and SS302-035M.
In order to remove protease and impurities after the reaction and obtain the correctly folded insulin-Fc fusion protein with high purity, cleaved SS302-004M and SS302-005M were filtered with 10 KD ultrafiltration tube to remove protease and other impurities, so as to obtain the insulin-Fc fusion protein with high purity. Cleaved SS302-008M, SS302-012M, SS302-014M, SS302-015M, SS302-029M and SS302-030M were subjected to hydrophobic chromatography to remove impurities. The medium for hydrophobic chromatography, Butyl HP (Bestchrom), was equilibrated using 3-5 column volume of buffer of 20 mM Tris, 1M (NH4)2SO4, pH 7.5. After the equilibration was completed, the sample was loaded (loading capacity of 3-8 g/L). After the loading was completed, the linear gradient elution was performed with a buffer of 20 mM Tris, pH 7.5 (0-100%, 20 column volume). The samples were collected separately according to the reading of the UV detector and detected. The impurities of cleaved SS302-019M and SS302-035M were removed in two steps. The first step was anion chromatography. The medium for anion chromatography, Q HP (Bestchrom), was equilibrated using 3-5 column volume of buffer of 20 mM Tris, pH 8.5. After the equilibration was completed, the sample was loaded (loading capacity of 5 g/L). After the loading was completed, the linear gradient elution was performed with a buffer of 20 mM Tris, 0.5M NaCl, pH 8.5 at a flow rate of 3 ml/min (0-60% B, 15 CV). The samples were collected separately according to the reading of the UV detector (by the same method as above) and detected. The samples with high purity were combined for hydrophobic chromatography of the next step. The medium for hydrophobic chromatography, Butyl HP (Bestchrom), was equilibrated using 3-5 column volume of buffer of 20 mM Tris, 1M NaCl, pH 8.0. After the equilibration was completed, the sample was loaded with a loading capacity of 3-8 g/L. After the loading was completed, the linear gradient elution was performed with a buffer of 20 mM Tris, pH 8.0 at a flow rate of 1 ml/min (0-100% B, 15 CV). The samples were collected separately according to the reading of the UV detector (by the same method as above) and detected for structural analysis, in which the molecular weight and disulfide bonds were characterized by UPLC-QTOF, see Example 4 for details.
The insulin fusion protein precursor has a structure of proINS-L-Fc, with proINS being a human insulin precursor (comprising B-C-A) and L a linker, and its schematic diagram is shown in
The spatial structure of the insulin-Fc fusion protein was supported and stabilized by the disulfide bonds formed between the sulfhydryl groups of two Cys residues. The disulfide bonds are divided into two parts, with some in insulin and others in Fc. The disulfide bonds of insulin are located in the B and A chains, and the amino acids of the B and A chains are respectively named by BX and AX in order from the N-terminal to the C-terminal, wherein X is the position of the amino acid in the sequence, and the disulfide bonds are CysA7-CysB7, CysA20-CysB19, and CysA6-CysA11. Fc consists of two polypeptide chains with the same sequence, and there are two disulfide bonds in each polypeptide chain, i.e., four disulfide bonds in two polypeptide chains, and two interchain disulfide bonds between the two polypeptide chains, meaning that there are 6 disulfide bonds in Fc. Theoretically, the disulfide bonds of the insulin-Fc fusion protein is not affected by the kex2 proteolysis. The structural analysis of the disulfide bonds of the insulin-Fc fusion protein was accomplished by buffer exchange of non-reducing denaturation, cleavage by restriction enzyme and analysis by the software UNIFI. There were two pretreatment methods. When analyzed by UNIFI, the two chains of the insulin-Fc fusion protein precursor were named as chain 1 and chain 2, respectively, of which the peptide fragments formed through proteolysis by Glu-C in pretreatment method 1 were named as 1:VN and 2:VN by UNIFI (see Tables 8-11), and the peptide fragments formed through proteolysis by Glu-C and trypsin in pretreatment method 2 were named as 1:VTN and 2:VTN by UNIFI (see Table 15); the two B chains of the mature insulin-Fc fusion protein were named as chain 1 and chain 3, and the two A+Fc chains were named as chain 2 and chain 4, respectively, of which the peptide fragments formed through proteolysis by Glu-C and trypsin in pretreatment method 2 were named as 1:VTN, 2:VTN, 3:VTN and 4:VTN by UNIFI (see Tables 12-14 and 16), where N represents the software number of the peptide fragment after proteolysis, which was sequentially numbered as 1, 2, 3, . . . and so on from the N-terminal to the C-terminal. Moreover, the disulfide bond in UNIFI was represented by “=”, the interchain disulfide bond was located between the two peptide fragments, and the intrachain disulfide bond was located on the right side of the peptide fragment.
SS302-002 (about 130 KD), SS302-002 (between 95-130 KD), SS302-004 (between 95-130 KD), SS302-004 (about 95 KD) and SS302-005 were treated by the pretreatment method 1 to analyze their disulfide bonds. The steps of the pretreatment method 1 are as follows. The sample of protein SS302 was placed into a 0.5 mL 10 kD ultrafiltration tube and concentrated to 5 mg/mL under a condition of 4° C. and 12000 rpm. 30 μL of the concentrated sample was added with 18 μL of 8M guanidine hydrochloride (pH7.5) and 0.48 μL of 1M IAA (iodoacetamide), mixed well by vortex, and incubated at room temperature in the dark for 40 min. 1.8 μL of the above sample was diluted with 23 μL of 50 mM Tris-HCl (pH8) buffer, added with 2.25 μL of 0.1 mg/mL Glu-C at a ratio of protein:enzyme=25:1 (μg:μg), water-bathed at 37° C. overnight, and added with 3 μL of 10% FA (formic acid) the next day to stop the reaction for UPLC-QTOF detection. Due to the incomplete denaturation by the pretreatment method 1, the linker region was difficult to be enzymatically cleaved, so that the disulfide bonds on the insulin and the disulfide bonds in the hinge region were linked together by the linking peptide. The large molecular weight makes matching difficult, so this method results in the loss of key disulfide bond information and was mainly used to compare the difference in disulfide bond mismatches between the two bands SS302-002 (about 130 KD) and SS302-002 (between 95-130 KD), and between the two bands SS302-004 (between 95-130 KD) and SS302-004 (about 95 KD).
SS302-008, SS302-012, SS302-012M, SS302-014, SS302-014M, SS302-015, SS302-015M, SS302-019M, SS302-029M, SS302-030M, SS302-035 and SS302-035M were treated by the pretreatment method 2 to analyze their disulfide bonds. The steps of the pretreatment method 2 are as follows. 40 μL of the sample of protein SS302 was added with 120 μL of 8M guanidine hydrochloride, water-bathed at 60° C. for 1 h, cooled to room temperature, added with 3.2 μL of 1M IAA, incubated at room temperature in the dark for 45 min, and subjected to buffer exchange for 3 times into 50 mM Tris-HCl buffer (pH 8) using an 0.5 mL 10 kD ultrafiltration tube under a condition of 12000 rpm and 4° C., so that the sample concentration after the buffer exchange was about 0.62 mg/mL. 40 μL of the above sample was added with 2 μL of Glu-C(0.5 mg/mL) and 2 μL of trypsin (0.5 mg/mL) at a ratio of protein:enzyme=25:1 (μg:μg), water-bathed at 37° C. overnight, and added with 5 μL of 10% FA the next day to stop the reaction for UPLC-QTOF detection. In the pretreatment method 2, trypsin and Glu-C were used together for enzymatic cleavage to realize the enzymatic cleavage of the linker region and the correct matching of disulfide bonds in the above SS302 molecules, and this method obtains more realistic calculation results of mismatched disulfide bonds.
The detection results of disulfide bonds obtained by UPLC-QTOF were analyzed combined with UNIFI software to analyze the correct disulfide bonds and mismatched disulfide bonds, and the disulfide bond mismatch is reflected by the total mismatch rate and insulin mismatch rate, where the total mismatch rate is the ratio of the total XIC peak area of the mismatched disulfide bond peptides to the total XIC peak area of all disulfide bond peptides, and the insulin mismatch rate is the ratio of the total XIC peak area of the mismatched disulfide bonds in the insulin moiety to the total XIC peak area of all disulfide bond peptides. The mismatch rates of SS302-002, SS302-004, SS302-005, SS302-008, SS302-012, SS302-012M, SS302-014, SS302-014M, SS302-015, SS302-015M, SS302-019M, SS302-029M, SS302-030M, SS302-035, and SS302-035M are shown in Table 7.
For fusion proteins comprising a flexible linker (SS302-004 and SS302-005), SS302-005 had the highest mismatch rate among all molecules, and the target band of SS302-004 (between 95-130 KD) had a relatively low mismatch rate, but a not high yield due to the fact that it was not easily separated from the components with high mismatch rate. For fusion proteins comprising both flexible and rigid moieties in the linker (SS302-002), the target band had a comparable total mismatch rate and insulin mismatch rate to fusion proteins comprising a flexible linker (SS302-004), both of which had components with high total mismatch rate and insulin mismatch rate and are not easily purified and separated. However, the precursor proteins and mature proteins comprising a rigid linker (SS302-008, SS302-012, SS302-012M, SS302-014, SS302-014M, SS302-015, SS302-015M, SS302-019M, SS302-029M, SS302-030M, SS302-035, SS302-035M) had a total mismatch rate and insulin mismatch rate of less than 8%. The disulfide bond results of SS302-002, SS302-004, SS302-012M, SS302-019M, SS302-030M, SS302-035 and SS302-035M in Example 4 are described in detail, and the results are shown in Tables 8-16.
In conclusion, a rigid linker had a great positive effect on the accuracy of the structural expression of the insulin fusion protein in CHO cells, and the stronger the rigidity, the higher the accuracy of its molecular structural expression.
arepresents that the fusion protein contains a flexible linker, and brepresents that the fusion protein contains a rigid linker;
Combined with SDS-PAGE technology, this molecule can be purified to obtain a band of about 130 KD and a band between 95-130 KD. The two bands were subjected to disulfide bond identification respectively to estimate the total mismatch rate and insulin mismatch rate of disulfide bonds. The results showed that total mismatch rate and insulin mismatch rate were both 9% for the band of about 130 KD, and the total mismatch rate and insulin mismatch rate were both 29% for the band between 95-130 KD. The results of the disulfide bonds of the band of about 130 KD are shown in Table 8, and the results of the disulfide bonds of the band between 95-130 KD are shown in Table 9. The mismatched disulfide bonds were mainly presented as the self-linking of the B chain of insulin and the mismatch between the two B chains of insulin.
1:V2-2:V2
1:V1-1:V2
1:V2-1:V8
1:V1-2:V1
1:V8-2:V8
This molecule was purified to obtain a band between 95-130 KD (P1-4 combined sample) and a band of about 95 KD (P13-15 combined sample). The two bands were subjected to disulfide bond identification, respectively. The results showed that total mismatch rate and insulin mismatch rate were both 4% for the band between 95-130 KD, and the total mismatch rate and insulin mismatch rate were both 37% for the band of about 95 KD. The results of the disulfide bonds of the band between 95-130 KD are shown in Table 10, and the results of the disulfide bonds of the band of about 95 KD are shown in Table 11. The mismatched disulfide bonds were mainly presented as the self-linking of the B chain of insulin and the mismatch between the two B chains of insulin.
1:V2-2:V2
1:V1-1:V2
1:V2-1:V8
1:V1-2:V1
1:V2-2:V2
1:V1-1:V2
1:V2-1:V8
1:V1-2:V1
1:V8-2:V8
This molecule had disulfide bonds consistent with the theory, a total mismatch rate of 2.9% and an insulin mismatch rate of 2.2%. The results of the disulfide bonds are shown in Table 12.
2:VT2-2:V
T20-2:VT31
1:VT3x2
1:VT3-1:V
T4
This molecule had disulfide bonds consistent with the theory, a total mismatch rate of 2.8% and an insulin mismatch rate of 1.2%. The results of the disulfide bonds are shown in Table 13.
2:VT24-2:
VT43
1:VT3-2:V
T4
1:VT5-2:V
T43
2:VT24-2:
VT34
1:VT5-2:V
T2
1:VT5-2:V
T34
1:VT3-2:V
T43
1:VT3-1:V
T5
This molecule had disulfide bonds consistent with the theory, a total mismatch rate of 1.7% and an insulin mismatch rate of 0%. The results of the disulfide bonds are shown in Table 14.
2:VT43x2
1:VT5-2:V
T43
1:VT5-2:V
T34
2:VT34×2
2:VT24-2:
VT34
1:VT3-2:V
T43
2:VT4-2:V
T43
2:VT4-2:V
T34
This molecule had disulfide bonds consistent with the theory, a total mismatch rate of 4.3% and an insulin mismatch rate of 2.2%. The results of the disulfide bonds are shown in Table 15.
1:VT17-1:
VT36
1:VT3-1:V
T36
1:VT3-1:V
T46
1:VT36-1:
VT46
1:VT15-1:
VT17
1:VT3-1:V
T17
This molecule had disulfide bonds consistent with the theory, a total mismatch rate of 2.5% and an insulin mismatch rate of 2.0%. The results of the disulfide bonds are shown in Table 16.
2:VT2-2:V
T4
1:VT3-2:V
T33
1:VT3x2
T4
24 healthy male Kunming mice (22-28 g) were randomly divided into 4 groups, 6 mice/group: (1) SS302-002M—24 nmol/kg; (2) SS302-002-24 nmol/kg; (3) insulin glargine −48 nmol/kg; and (4) negative control group. The administration was performed by subcutaneous injection in the neck. The blood glucose level was detected at 0, 1, 2, 4, 6, 8, 10, 12, 24, 36, 48, 60, 72, and 96 h, respectively. During the experiment, the mice were not fasted, and were given sufficient water and food.
As shown in
50 healthy male C57 mice aged 8-10 weeks and weighing 22-28 g were randomly divided into 10 groups, 5 mice/group, including SS302-008M, SS302-012M, SS302-014M, SS302-015M, SS302-019M, SS302-029M, SS302-030M, SS302-035M, insulin degludec and control group. The samples to be tested were administered subcutaneously at the neck at 15 nmol/kg and insulin degludec at 30 nmol/kg. The blood glucose level was detected at different time points before and after administration. During the experiment, the mice were not fasted. The experimental data were plotted using Graphpad prism 7.0, and the difference was statistically analyzed by Mann-Whitney test.
As shown in
25 healthy male C57 mice aged 8-10 weeks and weighing 22-28 g were randomly divided into 5 groups, 5 mice/group. SS302-035M was administered subcutaneously in the neck at 5, 7.5, 10, and 12.5 nmol/kg, respectively, and the blood glucose level was detected at 0, 4, 24, 48, 72, 96, and 120 h. During the experiment, the mice were not fasted. The experimental data were plotted using Graphpad prism 7.0, and the difference was statistically analyzed by Mann-Whitney test.
As shown in
C57BL/6j mice (8 weeks old, body weight of 22-28 g) were intraperitoneally injected with 0.4% streptozotocin (STZ) solution prepared in citric acid-sodium citrate buffer at 40 mg/kg for five consecutive days, once a day, and the fasting blood glucose level was detected on the 7th to 10th day after the last administration. A fasting blood glucose level >13.8 mmol/L (fasting time of 8:00 a.m-14:00 p.m) was considered as successful modeling.
35 STZ-induced type I diabetic mice were randomly divided into 7 groups according to their blood glucose level: 1-2: high and low dose groups of SS302-002M; 3-4: high and low dose groups of SS302-004M; 5-6: high and low dose groups of insulin glargine; and (7) control group (20 mM Tris+300 mM NaCl). Among them, the high and low dose groups of SS302-002M and SS302-004M were respectively administered at 12.5 nmol/kg and 6.25 nmol/kg by subcutaneous injection in the neck, and the high and low dose groups of insulin glargine were respectively administered at 25 nmol/kg and 12.5 nmol/kg by subcutaneous injection in the neck. Changes in blood glucose levels were monitored at different time points before and after administration. During the experiment, the mice were not fasted, and were given sufficient water and food.
The results are shown in
It is worth noting that at the same moles of insulin, i.e., at a dose of 25 nmol/kg, the blood glucose level decreased and recovered more rapidly in the insulin glargine group than in the SS302-002M and SS302-004M groups, dropped to the lowest blood glucose level (about 5 mmol/L) about 1 hour after administration (lower than the normal C57 blood glucose level of about 8 mmol/L), then quickly rose again, and returned to the initial blood glucose level at 6 h. This suggests that SS302-002M and SS302-004M had a more steady and stable PD profile and higher clinical safety.
C57BL/6j mice (12 weeks old, body weight of 22-28 g) were intraperitoneally injected with 0.4% streptozotocin (STZ) solution prepared in citric acid-sodium citrate buffer at 40 mg/kg for five consecutive days, once a day, and a fasting blood glucose level detected on the 7th to 10th day after the last administration >13.8 mmol/L (fasting time of 8:00 a.m-14:00 p.m) was considered as successful modeling.
40 successfully STZ-modeled type I diabetic mice were randomly divided into 8 groups according to their blood glucose level: (1) SS302-008M—7.5 nmol/kg group; (2) SS302-012M—7.5 nmol/kg group; (3) SS302-035M—7.5 nmol/kg group; (4) SS302-008M—15 nmol/kg group; (5) SS302-012M—15 nmol/kg group; (6) SS302-035M—15 nmol/kg group; (7) insulin degludec—30 nmol/kg; and (8) buffer control group (20 mM Tris+150 mM NaCl). The blood glucose level was detected at different time points before and after administration. During the experiment, the mice were not fasted. The experimental results were plotted using Graphpad prism 7.0, and the difference was statistically analyzed by Mann-Whitney test.
As shown in
10 SD rats (8-10 weeks old, body weight of 250-350 g) were randomly divided into 2 groups with 3♂2♀ in each group, and SS302-008M or SS302-012M were administered subcutaneously in the neck at 20 nmol/kg, respectively. The blood glucose level was detected at different time points before and after administration, and whole blood was collected to separate serum for PK detection. During the experiment, the mice were not fasted, and were given sufficient water and food. All data were plotted with Graphpad prism 7.0, and the difference was statistically analyzed by Mann-Whitney test.
Mouse anti-insulin monoclonal antibody (abcam, ab8302) was diluted with PBS to 1 μg/mL, added to a microplate at 100 μL/well, and placed at 4° C. overnight for coating. After the removal of the coating solution, the plate was washed with PBST 4 times, then added with 4% BSA at 250 μl/well, and blocked at 37° C. for 2 h. After the removal of the blocking solution, the plate was washed with PBST 4 times. The SS302-008M/SS302-012M standard was serially diluted with 2% BSA to obtain a total of 8 gradients starting from 200 ng/ml to establish a standard curve. Rat serum was diluted to various gradients with 2% BSA. The negative control was normal rat serum. The above samples were added to a microplate at 100 μl/well and incubated at 37° C. for 1 h. The plate was then washed 4 times with PBST, added with a secondary antibody (Mouse monoclonal Anti-Human IgG2 Fc (HRP), 1:3000) (abcam, ab99779) diluted with 2% BSA at 100 μL/well and incubated at 37° C. for 1 h. The plate was then washed 4 times with PBST, added with TMB chromogen solution at 100 μl/well to develop color at 37° C. in dark for 10 min, and then added with 2M H2SO4 at 50 μL/well to stop the reaction. The OD450/630 value was detected by a microplate reader.
As shown in
The pharmacokinetic results of SS302-008M and SS302-012M in SD rats are shown in
4 male healthy general-grade beagle dogs weighing 8-12 kg were evaluated for pharmacodynamic and pharmacokinetic parameters after a single subcutaneous administration of 2.5 nmol/kg SS302-035M. Blood samples were collected at different time points before and after administration, and the sampling sites were peripheral veins of four limbs. About 1 mL of whole blood was collected at each time point, put into an anticoagulant tube containing EDTA-K2, and then centrifuged at 3000 g/min for 10 min at 4° C. to collect plasma. A drop of whole blood at time points 0 h before administration and 1, 2, 3, 4, 6, 24, 48, 72, 96, 120, 144 and 168 h after administration was taken to detect the blood glucose level of the animal using a blood glucose meter (Roche's ACCU-CHEK Performa) and blood glucose test strips (Roche's ACCU-CHEK Performa). The pharmacodynamic (PD) results are shown in
The full-length sequences of the fusion protein precursors constructed in the examples of the present disclosure are as follows:
| Number | Date | Country | Kind |
|---|---|---|---|
| 202010723972.9 | Jul 2020 | CN | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/CN2021/107040 | 7/19/2021 | WO |
