Selective reduction of proteins

Abstract

To provide methods for producing uncapped cysteine protein preparations, for example, uncapped engineered cysteine antibody preparations.SOLUTION: The present invention provides, inter alia, a method for selectively reducing an engineered cysteine antibody. The method includes: contacting a reducing agent with engineered cysteine antibody molecules, each of the antibody molecules having at least one capped engineered cysteine residue and at least one interchain disulfide bond; and reacting the reducing agent with the antibody molecules under conditions sufficient to uncap engineered cysteine residues and form cap byproducts. The method also includes removing the cap byproduct during the reduction reaction. Substantially all of the interchain disulfide bonds present in the antibody molecules prior to reduction are retained following reduction. An antibody conjugate and method for preparing the antibody conjugate using an uncapped antibody preparation are also described.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority to U.S. Patent Application No. 61 / 938,378, filed February 11, 2014, the contents of which are incorporated herein by reference in their entirety. [Background technology]

[0002] Monoclonal antibodies in which selected amino acids have been mutated to cysteines (i.e., engineered cysteine ​​mAbs or ecmAbs) are particularly suitable for use in conjugates (e.g., antibody-drug conjugates (ADCs)) because conjugates derived from them can have favorable properties, such as homogeneity, good pharmacokinetics, stability, and solubility. The cysteine ​​mutations are placed at positions in the antibody amino acid sequence that do not normally form interchain disulfide bonds, and the expression machinery in the cells producing the mutant mAbs treats the cysteine ​​residue as an unpaired cysteine. Thus, the engineered cysteine ​​is typically expressed in the form of a mixed disulfide along with a non-encoded cysteine ​​molecule (i.e., the engineered cysteine ​​is "capped" with a capping agent, e.g., cysteine, cysteinylglycine, or glutathione).

[0003] Thus, to prepare conjugates (e.g., ADCs) from ecmAbs, it is generally necessary to subject the ecmAb to reducing conditions that convert the engineered cysteines from mixed disulfides to free thiols, and typically this uncapping or "activation" involves reduction of the ecmAb interchain disulfides. Although the interchain disulfides can be reformed by mild oxidation, such a reoxidation step adds complexity and expense to ADC preparation. Selective reduction methods have been elusive, as it has proven difficult to reduce the engineered cysteines without simultaneously reducing the interchain disulfides. Thus, a selective reduction method for uncapping engineered cysteine ​​residues is needed. The present invention addresses this and other needs. Summary of the Invention

[0004] The present invention provides, inter alia, a method for selectively reducing engineered cysteine ​​antibodies. The method includes contacting engineered cysteine ​​antibody molecules, each antibody molecule having at least one capped engineered cysteine ​​residue and at least one interchain disulfide bond, with a reducing agent; and reacting the antibody molecules with the reducing agent under conditions sufficient to uncap the engineered cysteine ​​residues and form a cap byproduct. The cap byproduct is removed during the reduction reaction. The method results in the formation of an uncapped engineered cysteine ​​antibody preparation. The uncapped engineered cysteine ​​antibody preparation can include both capped and uncapped engineered cysteine ​​antibodies. Substantially all of the interchain disulfide bonds present in the antibody molecules prior to the reduction reaction are retained in the uncapped engineered cysteine ​​antibody preparation. The method can further include supplementing the reduction reaction with additional reducing agent while removing the cap byproduct. During the reduction reaction, removal of the cap by-product from the reaction mixture can be performed, for example, by dialysis or diafiltration. According to a preferred embodiment, the concentration of the cap by-product is maintained in the reduction reaction mixture below the concentration at which re-capping prevents further activation of the engineered cysteine ​​residue.

[0005] In some embodiments, the method provides an uncapped engineered cysteine ​​antibody preparation. Typically, at least about 60% of the engineered cysteine ​​residues present in the antibody preparation are uncapped engineered cysteine ​​residues. In some embodiments, at least about 70%, at least about 75%, or at least about 80% of the engineered cysteine ​​residues present in the antibody preparation are uncapped engineered cysteine ​​residues. In some embodiments, using the methods of the invention, at least 85% of the interchain disulfide bonds present in the antibody molecules prior to the reduction reaction are retained in the uncapped engineered cysteine ​​antibody preparation. In some embodiments, the reducing agent and reduction conditions are selected such that less than about 20%, less than about 15%, less than about 10%, or less than about 5% of the interchain disulfide bonds present in the antibody molecules are converted to a pair of free thiols during the reduction reaction.

[0006] In related aspects, the invention provides antibody conjugates, including antibody drug conjugates, and methods of preparing antibody conjugates using uncapped antibodies. Residual reducing agent may be removed from the antibody preparation prior to preparation of the antibody drug conjugate.

[0007] The present invention also provides a method for selectively reducing non-antibody proteins having unpaired cysteine ​​residues. The method includes contacting a reducing agent with protein molecules, each protein molecule having at least one capped cysteine ​​residue and at least one interchain disulfide bond; and reacting the reducing agent with the protein molecules under conditions sufficient to uncap the cysteine ​​residues and form a cap byproduct. The cap byproduct is removed during the reduction reaction. The method results in the formation of an uncapped cysteine ​​protein preparation. The uncapped cysteine ​​protein preparation can include both capped and uncapped cysteine ​​protein molecules. Substantially all of the interchain disulfide bonds present in the protein molecules prior to the reduction reaction are retained in the uncapped cysteine ​​protein preparation. The method can further include supplementing the reduction reaction with additional reducing agent while removing the cap byproduct. Removal of the cap byproduct from the reaction mixture during the reduction reaction can be performed, for example, in dialysis or diafiltration. In preferred embodiments, the concentration of cap byproducts is maintained in the reduction reaction mixture at or below a concentration at which re-capping prevents further activation of the engineered cysteine ​​residues. The method provides an uncapped cysteine ​​protein preparation. Typically, at least about 60% of the capped cysteine ​​residues are uncapped (or, in other words, activated) using the method. At least about 70%, at least about 75%, or at least about 80% of the capped cysteine ​​residues are uncapped using the method. In some embodiments, using the method, at least 85% of the interchain disulfide bonds present in the protein molecule prior to the reduction reaction are retained in the uncapped engineered cysteine ​​antibody preparation. In some embodiments, the reducing agent and reduction conditions are selected such that less than about 20%, less than about 15%, less than about 10%, or less than about 5% of the interchain disulfide bonds present in the protein molecule are converted to a pair of free thiols during the reduction reaction. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 shows the results of rPLRP column analysis of samples generated by treating ecmAb (99 μM in solution) with three different monothiol reducing agents (cysteine, N-acetylcysteine ​​(NAC) and cysteamine) at low concentration (2 mM) under identical conditions. The Y-axis represents the % activation of total available engineered cysteines. None of the reducing agents approach 100% activation of available engineered cysteine ​​residues. There was no evidence of reduction of interchain disulfide bonds in these experiments. [Diagram 2] FIG. 2 shows the results of rPLRP column analysis of the H chain from the S239C ecmAb mcMMAF ADC. The S239C ecmAb was selectively reduced by the method of the present invention and conjugated to a mcMMAF drug-linker. The reducing agent was cysteine ​​at a concentration of 0.8 mM. Activation of the engineered cysteine ​​residue proceeded substantially to completion. The bar designated %H0 represents an H chain with no conjugated drug-linker. The bar designated %H1 represents an H chain with one conjugated drug-linker, and the bar designated %H2 represents an H chain with two conjugated drug-linkers. [Diagram 3] Figure 3 shows the results of rPLRP column analysis of the H chain from the S239C ecmAb mcMMAF ADC. S239C ecmAb was selectively reduced by the methods of the invention and conjugated to the mcMMAF drug-linker. The reducing agent was cysteine ​​at a concentration of 0.55 mM. A small percentage of the engineered cysteine ​​residues remained capped (%H0), however, the percentage of uncapped engineered cysteine ​​residues approached 100% (%H1) and the amount of non-specific reduction (%H2) remained low. [Figure 4]Figure 4 shows that activation of engineered cysteines was selectively achieved with S239C ecmAb despite the identity of the heavy and light chain variable regions. mAb1 is a humanized 2H12 ecMAb and mAb2 is a humanized h1F6 ecMAb (activation: %H0; selectivity: %H2). The percentage of decapping was similar for the two mAbs. Selective activation of engineered cysteines is shown for mAb1 at three different sites, S239, K326 and A327. The results suggest that engineered cysteine ​​activation can be selectively achieved at various sites. The experiments were performed with approximately 0.9 mM cysteine. [Diagram 5] Figure 5 shows the activation of engineered cysteines for the S239C mutant of mAb1. The mAb concentration in the reaction mixture was approximately the same as for the reaction in Figure 4, however the cysteine ​​concentration was low throughout the time course, approximately 0.5 mM. This experiment showed that maximum selectivity was obtained at low cysteine ​​concentrations, and given sufficient time, complete activation could be achieved with very good selectivity. The average drug loading of the conjugates prepared from the final time point in this experiment was 1.93, which is close to the nominal value of 2.0 and within the range of values ​​obtained by the conventional method of nonselective reduction followed by reoxidation. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] I. General The present invention provides, inter alia, a method for selectively removing sulfide caps from engineered cysteine ​​residues in antibodies, e.g., monoclonal antibodies. Although some selective removal is observed after prolonged exposure of antibodies to small molecule thiols at low concentrations, complete decap does not occur in static reaction mixtures. Advantageously, near complete activation of engineered cysteines can be achieved by using the methods of the present invention in which reducing agent concentration is maintained while removing reduction by-products. In embodiments in which it is not desired to activate substantially all of the engineered cysteines, the methods of the present invention can be used to adjust the level of activation. Surprisingly, uncapped antibodies are obtained with intact interchain disulfide bonds, which preserves the antibody structure and eliminates the need for a reoxidation step prior to conjugation of the antibody with a drug or other functional agent. Among other things, the methods of the present invention provide a significant simplification of current manufacturing standards for the preparation of antibody conjugates, e.g., antibody drug conjugates.

[0010] II. Definition As used herein, the term "antibody" broadly refers to intact monoclonal antibodies, polyclonal antibodies, monospecific antibodies, multispecific antibodies (i.e., bispecific antibodies), and antibody fragments that exhibit the desired biological activity (i.e., specific binding to a target antigen) and have at least one native interchain disulfide bond. Intact antibodies are typically composed of four polypeptide chains (two heavy and two light chains), with each polypeptide having two main regions: a variable region and a constant region. The variable region specifically binds to and interacts with a target antigen. The variable region contains complementarity determining regions (CDRs) that recognize and bind to a specific binding site on a particular antigen. The constant region can be recognized by and interact with the immune system (see, e.g., Janeway et al., 2001, Immuno. Biology, 5th Ed., Garland Publishing, New York). The four polypeptide chains are covalently linked to each other through interchain disulfide bonds. Antibodies can be of any type (e.g., IgG, IgE, IgM, IgD and IgA), class (e.g., IgG1, IgG2, IgG3, IgG4, IgA1 and IgA2) or subclass. Antibodies can be derived from any suitable species. In some embodiments, antibodies are of human or murine origin. Monoclonal antibodies can be, for example, human, humanized, or chimeric. Depending on the attribute, the term "antibody" can refer to a single antibody molecule, or a collection of antibody molecules, e.g., an antibody solution.

[0011] As used herein, the term "inter-chain disulfide bond" refers to a covalent bond between two cysteine ​​residues on adjacent polypeptide chains in an antibody. A disulfide bond can be represented by the formula R 1 -S-SR 2 where the sulfur atom is in a cysteine ​​side chain, and R 1 and R 2represent the cysteine ​​residues and the remainder of the polypeptide chain in which they are found. Interchain disulfide bonds generally exist between a heavy and light chain or between two heavy chains in an antibody.

[0012] As used herein, the term "engineered cysteine ​​residue" refers to a cysteine ​​residue that is introduced into the peptide sequence of a protein (e.g., an antibody). A monoclonal antibody with an engineered cysteine ​​residue may be referred to as an "ecmAb." An engineered cysteine ​​residue is generally not present in the native (i.e., naturally occurring) peptide sequence of a protein. An engineered cysteine ​​residue can take the position of a naturally occurring amino acid at a given position in the peptide sequence and can be introduced into the peptide sequence through recombinant techniques, e.g., site-directed mutagenesis. An engineered cysteine ​​residue may be capped or decapped.

[0013] As used herein, the term "uncapped cysteine ​​residue" refers to a cysteine ​​residue whose alpha side chain has the formula R 1 R means a cysteine ​​residue containing a free thiol moiety having -SH. 1 represents the non-thiol portion of a cysteine ​​residue. The uncapped cysteine ​​residue may be an uncapped engineered cysteine ​​residue.

[0014] As used herein, the term "capped cysteine ​​residue" refers to a cysteine ​​residue whose alpha side chain has the formula R 1 -SSR 3 R 1 represents the non-thiol portion of a cysteine ​​residue, and R 3represents the non-thiol portion of the capping moiety having a molecular weight of less than or equal to about 500 Da. For example, the cap may be cysteine, cysteinylglycine, or glutathione (R represents the non-thiol portion of free cysteine ​​or glutathione, respectively). 3 (having the formula:

[0015] As used herein, the term "reduction reaction" refers to a reaction of formula R 1 -SSR 3 is reduced and a capped cysteine ​​residue having the structure R 1 R refers to a reaction that forms a thiol moiety having -SH. 1 and R 3 is defined as above.

[0016] As used herein, the term "reducing agent" refers to a compound capable of reducing disulfide bonds. Reducing agents include, but are not limited to, thiols, such as cysteine, cystamine, and β-mercaptoethanol.

[0017] As used herein, the term "oxidizing agent" refers to a compound that causes the conversion of a pair of free thiols to a disulfide bond. Examples of oxidizing agents include, but are not limited to, 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB), dehydroascorbic acid (DHAA), and copper sulfate (CuSO4). A "re-oxidation step" is an active step taken to cause the conversion of a pair of free thiols to a disulfide bond. An active step includes the introduction of an exogenous oxidizing agent and / or an intentional hold period to allow for autoxidation.

[0018] As used herein, "removing" a substance, e.g., a cap by-product, from a mixture containing an antibody and a substance means removing any portion of the substance from the mixture, including the entire substance. Removal of a substance also includes the transfer of an antibody from a first mixture that contains the substance and a second mixture that does not contain the substance. Removal of a substance from a mixture can include processes such as dialysis, diafiltration, chromatography, and similar means.

[0019] As used herein, the terms "antibody-drug conjugate" and "ADC" refer to an antibody conjugated, optionally through a linker, to a therapeutic agent (i.e., a drug).

[0020] As used herein, the term "drug-linker compound" or "drug-linker" means a molecule having a drug moiety and a linker attached thereto, where the linker contains a reactive moiety suitable for attachment to an amino acid residue (e.g., a cysteine ​​residue) in an antibody.

[0021] III. Description of the Preferred Embodiments In particular, the present invention provides a method for selectively reducing an engineered cysteine ​​antibody, thereby forming an uncapped antibody preparation. The method includes contacting engineered cysteine ​​antibody molecules, each antibody molecule having at least one capped engineered cysteine ​​residue and at least one interchain disulfide bond, with a reducing agent; and reacting the antibody molecules with the reducing agent under conditions sufficient to uncap the engineered cysteine ​​residues and form a cap byproduct. During the reduction reaction, the cap byproduct is removed, thereby preventing re-capping of the newly formed thiol. The reducing agent, which is removed along with the cap byproduct, is positioned to drive the reduction reaction forward. The reducing agent and conditions are selected such that the engineered cysteine ​​residues are selectively reduced (i.e., selectively activated). As a result, a separate reoxidation step is not required to reform the interchain disulfide bonds following the reduction step. Substantially all of the interchain disulfide bonds present in the capped antibody molecule with engineered cysteine ​​residues are retained in the uncapped antibody preparation. By use of the term "retain" it is not meant that the interchain disulfide bonds necessarily remain intact during the reduction reaction. Bonds may be broken during reduction, but they reform prior to conversion to a pair of free thiols. Bond reformation does not depend on a separate reoxidation step, however, it does occur during the reduction reaction, as shown in Scheme 2.

[0022] Scheme 1 shows a schematic diagram of cysteine ​​residue decapping according to the method of the invention, using a monoclonal antibody as an exemplary protein. As shown in reaction (i), reaction of an antibody having a capped engineered cysteine ​​residue (2) with a reducing agent (1) results in an antibody having an uncapped engineered cysteine ​​residue (3) and a cap by-product (4). Reaction (i) can proceed in both directions, and reaction of the uncapped engineered cysteine ​​residue with the cap by-product can result in recapping of the engineered cysteine ​​residue. The result of the reverse reaction is a composition in which only some of the antibodies in the composition have an uncapped engineered cysteine ​​residue. In accordance with the method of the invention, the reaction mixture is supplemented with additional reducing agent, and the cap by-product is removed from the reaction mixture, as shown in reaction (ii). Reducing agent supplementation and cap by-product removal drive the reaction forward and prevent the reverse reaction, facilitating the formation of the desired antibody having an uncapped engineered cysteine ​​residue (3).

[0023] Scheme 1 [ka]

[0024] Under normal (i.e., "static solution") conditions, the recapping reaction limits the extent of activation of engineered cysteines. The graph in Figure 1 shows the results of treatment of ecmAb with three different monothiols under identical conditions. The Y-axis represents the % activation of total available engineered cysteines. The graph shows that no reducing agent can approach 100% activation of the available engineered cysteine ​​residues, and this limitation is due to the disulfide by-product (4, Scheme 1), which recaps the generated thiol, as described above.

[0025] Reducing Agent Any suitable reducing agent can be used in the method of the present invention. Examples of reducing agents include, but are not limited to, monothiol reducing agents such as cysteine, N-acetylcysteine, cysteamine, β-mercaptoethanol, 2-mercaptoethanesulfonic acid sodium salt, and the like. Mild reducing agents such as cysteine ​​and the like are particularly suitable for removing the cap from capped cysteine ​​residues (e.g., capped and modified cysteine ​​residues) without reducing the interchain disulfide bonds. Stronger reducing agents such as dithiothreitol (DTT), dithioerythritol (DTE), and bis(2-mercaptoethylsulfone) often result in premature reduction of interchain disulfide bonds. Reducing agents such as TCEP and DTT that do not form or only temporarily form the mixed disulfide of intermediate 6 in Scheme 2 are unlikely to selectively uncap the capped cysteines. In some embodiments, the reducing agent is selected from cysteine, cysteamine, β-mercaptoethanol, 2-mercaptoethanesulfonic acid sodium salt, and mixtures thereof.

[0026] Reaction conditions Although many monothiols can be used for selective activation, the graph in Figure 1 shows that different thiols activate engineered cysteines at different rates and to different maximum levels. Thus, identifying appropriate conditions for activation of engineered cysteine ​​residues involves identifying an appropriate reducing agent and identifying an appropriate concentration of the reducing agent.

[0027] Reaction conditions will be determined in part by the location of the engineered cysteine ​​residue, the identity of the cap, or residue in the particular antibody, e.g., a solvent-exposed engineered cysteine ​​residue may be more easily uncapped than a buried, or partially buried, engineered cysteine ​​residue.

[0028] Conditions that favor selective activation of unpaired cysteine ​​residues in proteins can be understood based on the reactions in Scheme 2. If the reducing agent reacts with engineered cysteine ​​residues, as in reaction (a), i.e., much faster than the interchain disulfide bonds, as in reaction (b), selective activation can be relatively simple. This condition occurs only if the engineered cysteine ​​is highly exposed to the solvent on the antibody surface, if at all. However, frequently, less exposed positions are preferred for engineered cysteine ​​residues in antibody conjugates, such as ecmAbs used in ACD. The reducing agent can react with those less exposed engineered cysteine ​​residues and with disulfide bonds at comparable rates. Thus, it is often necessary to find selective activation conditions when reactions (a) and (b) occur at comparable rates.

[0029] If reactions (a) and (b) occur at comparable rates or reaction (b) is faster than reaction (a), selective activation of engineered cysteines is necessary to ensure that reaction (c) to reform the disulfide is faster than the second reduction step (d), which, if the interchain disulfide is attacked by a reducing agent, results in the undesired cleavage of the interchain disulfide. In general, it is not possible to influence the rate of reaction (c), since it is a unimolecular reaction. The thiol and the mixed disulfide shown in 6 are part of the same antibody. However, the rate of reaction (d) depends on the concentration of the reducing agent, and consequently reaction (d) can be made slower than reaction (c) by using a sufficiently low concentration of the reducing agent. Of course, reactions (a) and (b) also depend on the concentration of the reducing agent, and consequently, under those conditions, the activation is also slow. Therefore, selective activation of engineered cysteines is generally favored by maintenance of very mild reducing conditions (low concentrations of reducing agent, and relatively mild reducing agents) for extended periods of time.

[0030] Scheme 2 [ka]

[0031] The reduction reaction mixture can contain any suitable amount of protein. Typically, the concentration of the protein (antibody or non-antibody protein) in the reduction reaction mixture ranges from about 0.01 mg / ml to about 150 mg / ml, more typically from about 1 mg / ml to about 50 mg / ml. The reduction reaction mixture can contain, for example, about 0.01, 0.05, 0.1, 0.25, 0.5, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 12.5, 15, 17.5, 20, 22.5, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, 100, 110, 120, 130, 140, or 150 mg of protein (antibody or non-antibody protein) per ml of reduction reaction mixture. The concentration of protein (antibody or non-antibody protein) can be higher or lower depending on the specific reaction conditions used. Those skilled in the art will be able to convert mass-based concentrations (e.g., mg / ml) to molar concentrations (i.e., moles l / l).

[0032] Any suitable amount of reducing agent may be used in the methods of the invention. Generally, the concentration of reducing agent in the reduction reaction mixture is high enough for the reaction to proceed, but low enough so that interchain disulfide reduction is negligibly slow. The reduction reaction mixture is typically formed with an initial amount of reducing agent at an initial concentration. This initial concentration is substantially maintained by continuously or stepwise supplementing the reduction reaction mixture with additional amounts of reducing agent throughout the duration of the reduction reaction. According to certain embodiments, the reduction reaction mixture is supplemented in a continuous manner with additional amounts of reducing agent throughout the reduction reaction. Using the conditions described in the Examples, the optimal concentration of cysteine ​​was determined to be about 0.5 mM to about 1.5 mM. Depending on the strength of the reducing agent and the concentration of ecmAb, the concentration may be increased or decreased. According to some embodiments, the concentration of the reducing agent will be maintained at a concentration higher than the concentration of the total antibody. The optimal concentration ratio of reducing agent:total antibody will also depend on the strength of the reducing agent. According to some embodiments, the concentration of the reducing agent in the reduction reaction mixture will be about 5 to about 25 times, 5 to about 20 times, 5 to about 15 times, or 5 to about 10 times higher than the concentration of the total antibody in the reduction reaction mixture. For example, according to some embodiments, the concentration ratio of the reducing agent:total antibody in the reduction reaction mixture is about 5:1, 6:1, 7:1, 8:1, 9:1, 10:1, 11:1, 12:1, 13:1, 14:1, 15:1, 16:1, 17:1, 18:1, 19:1, or 20:1. According to some embodiments, for example, some embodiments where the reducing agent is cysteine, the concentration ratio of the reducing agent:total antibody in the reduction reaction mixture will be about 5:1 to about 12:1, about 7:1 to about 10:1, and preferably about 8:1. According to some such embodiments, the engineered cysteine ​​residue will be at position 239 of the H chain (numbered according to the EU index). Other concentration ratios may be used in the methods of the invention, depending on factors such as the particular reducing agent used or the location of the engineered cysteine ​​residue in the antibody.

[0033] The reduction reaction that decapps the cysteine ​​residues (engineered or native) is carried out such that it minimizes the reduction of interchain disulfide bonds and the subsequent generation of free thiols. Substantially all of the interchain disulfide bonds present in a protein having a capped cysteine ​​residue are retained in the uncapped protein preparation formed using the methods of the invention. Typically, at least about 80% of the interchain disulfide bonds are retained in the uncapped protein preparation. In certain embodiments, at least about 85% of the interchain disulfide bonds are retained in the uncapped protein preparation. In certain embodiments, at least about 90%, or about 95% of the interchain disulfide bonds are retained in the uncapped protein preparation. In certain embodiments, all of the interchain disulfide bonds are retained in the uncapped protein preparation. Or, in other words, in typical embodiments, less than about 20%, less than about 15%, less than about 10, or less than about 5% of the interchain disulfide bonds are converted to a pair of free thiols during the reduction reaction. When applied to antibodies, substantially all of the interchain disulfide bonds present in the capped antibody molecule having an engineered cysteine ​​residue (e.g., the antibody molecule prior to the reduction reaction) are retained in the uncapped antibody preparation formed using the methods of the invention. Typically, at least about 80% of the interchain disulfide bonds are retained in the uncapped antibody preparation. In certain embodiments, at least about 85% of the interchain disulfide bonds are retained in the uncapped antibody preparation. In certain embodiments, at least about 90%, or about 95% of the interchain disulfide bonds are retained in the uncapped antibody preparation. In certain embodiments, all of the interchain disulfide bonds are retained in the uncapped antibody preparation. Or, in other words, in exemplary embodiments, less than about 20%, less than about 15%, less than about 10, or less than about 5% of the interchain disulfide bonds are converted to a pair of free thiols during the reduction reaction.

[0034] The concentration of the reducing agent is maintained in the reduction reaction mixture, while the reduction by-products are removed from the reduction reaction mixture. The reducing agent addition and by-product removal can be performed continuously or stepwise. According to certain embodiments, the reducing agent addition and by-product removal are performed in a continuous manner. The reducing agent is added and the by-products can be removed by techniques such as, but not limited to, tangential flow filtration (e.g., diafiltration), size exclusion chromatography, solid phase immobilization, and dialysis. According to the diafiltration process, an aqueous reducing agent solution is typically added to the reduction reaction mixture at a predetermined flow rate while removing a portion of the mixture at approximately the same flow rate. A semi-permeable membrane can be used to retain the protein (e.g., antibody) in the reduction reaction mixture. Alternatively, the protein (e.g., antibody) can be immobilized on a solid support (e.g., protein A agarose beads), and the aqueous reducing agent solution can be passed through the immobilized protein (e.g., antibody). A reactive resin that reacts with the cap by-products can also be used. For example, cap by-products can be sequestered using a resin that has reactive moieties, such as disulfides, on the interior of pores small enough to exclude proteins (eg, antibodies).

[0035] Thus, some embodiments of the present invention provide a method as described above, wherein removing the cap by-product from the reduction reaction mixture comprises dialyzing or diafiltrating the reduction reaction mixture during the reduction reaction. According to some embodiments, removing the cap by-product from the reduction reaction mixture comprises diafiltrating the reduction reaction mixture during the reduction reaction.

[0036] Frequently, the methods of the invention are performed to maximize uncapping of cysteine ​​residues. However, any one protein molecule can have capped and uncapped cysteine ​​residues. Frequently, the uncapped antibody preparation will include a plurality of the following antibody molecules: an antibody molecule having at least two uncapped engineered cysteine ​​residues and an antibody molecule having no capped engineered cysteine ​​residues; an antibody molecule having at least two capped engineered cysteine ​​residues and an antibody molecule having no uncapped engineered cysteine ​​residues; and an antibody molecule having at least one capped engineered cysteine ​​residue and at least one uncapped engineered cysteine ​​residue. Advantageously, the methods of the invention provide high levels of selectively uncapped engineered cysteine ​​residues. For example, the methods can be performed such that at least 40% of the engineered cysteine ​​residues are uncapped. According to some embodiments, the method can be performed such that at least about 40%, at least about 45%, at least about 50%, at least about 55%, at least about 60%, at least about 65%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, or at least about 90% of the engineered cysteine ​​residues are uncapped. According to certain embodiments, at least 60% of the engineered cysteine ​​residues are uncapped. Higher or lower levels of uncapping can also occur, depending in part on the particular antibody and the location of the engineered cysteine ​​residue within the antibody. Although the method of the invention is generally performed to maximize uncapping of cysteine ​​residues, there are cases where maximizing uncapping of cysteine ​​residues is not desired. For example, for an antibody with two engineered cysteine ​​residues, it may be desired to simply conjugate one of the cysteine ​​residues to a functional agent. According to some such embodiments, the method of the invention can be used to tightly regulate the activation of the engineered cysteine ​​residues. An advantageous aspect of the method of the present invention is that intermediate levels of uncapping can be easily achieved simply by stopping activation at the desired level and then conjugating.Such intermediate complexes are difficult to obtain using the method of complete reduction followed by reoxidation.

[0037] The reaction mixture can include additional reagents or components. By way of non-limiting example, the reaction mixture can include buffers (e.g., 2-(N-morpholino)ethanesulfonic acid (MES), 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES), 3-morpholinopropane-1-sulfonic acid (MOPS), 2-amino-2-hydroxymethyl-propane-1,3-diol (TRIS), potassium phosphate, sodium phosphate, phosphate buffered saline, sodium citrate, sodium acetate, and sodium borate), co-solvents (e.g., dimethylsulfoxide, dimethylformamide, ethanol, methanol, isopropanol, glycerol, tetrahydrofuran, acetone, acetonitrile, and acetic acid), salts (e.g., NaCl, KCl, CaCl2, and Mn 2+ and Mg 2+The reaction mixture may include, for example, a salt of ethylene glycol-bis(2-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA), 2-({2-[bis(carboxymethyl)amino]ethyl}(carboxymethyl)amino)acetic acid (EDTA), and 1,2-bis(o-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA). The buffers, co-solvents, salts, denaturants, surfactants, and chelating agents may be used at any suitable concentration, which may be readily determined by one of skill in the art. Generally, the buffers, co-solvents, salts, denaturants, surfactants, and chelating agents, if present, are included in the reaction mixture at a concentration ranging from about 1 μM to about 1 M. For example, buffers, co-solvents, salts, denaturants, detergents and chelators can be included in the reaction mixture at a concentration of about 1 μM, or about 10 μM, or about 100 μM, or about 1 mM, or about 10 mM, or about 25 mM, or about 50 mM, or about 100 mM, or about 250 mM, or about 500 mM, or about 1 M. In particular, co-solvents can be included in the reaction mixture in an amount ranging from about 1% v / v to about 75% v / v, or higher, for example. Co-solvents can be included in the reaction mixture in an amount of about 5, 10, 20, 30, 40, or 50% v / v.

[0038] The reaction is carried out under conditions sufficient to form uncapped cysteine ​​residues. The reaction may be carried out at any suitable temperature. Generally, the reaction is carried out at a temperature of about 4° C. to about 40° C. The reaction may be carried out, for example, at about 25° C. or about 37° C. The reaction may be carried out at any suitable pH. Generally, the reaction is carried out at a pH of about 6.5 to about 10. In some cases, the pH is about 7.0 to about 8.5. The reaction may be carried out for any suitable time. The length of time selected will depend on the strength of the reducing agent. Because mild reducing conditions (low concentration of reducing agent and relatively mild reducing agent) are used in the methods of the invention, the reduction reaction will typically be extended longer than expected for reduction of interchain disulfide bonds. According to some preferred embodiments, the reduction reaction will proceed until at least about 60%, at least about 70%, at least about 80%, or at least about 85% of the capped engineered cysteine ​​residues are uncapped. According to some embodiments, the reduction reaction will be incubated under appropriate conditions for at least 1 hour, at least 2 hours, at least 3 hours, at least 4 hours, at least 5 hours, or at least 6 hours. The reaction may be carried out under an inert atmosphere, such as a nitrogen or argon atmosphere. Other reaction conditions may be used in the methods of the invention, depending on the identity of the particular antibody or reducing agent.

[0039] In some embodiments, the pH of the reaction mixture ranges from about 6.5 to about 8.5. In some embodiments, the methods of the invention include maintaining the reaction mixture at a temperature ranging from about 4° C. to about 37° C. In some embodiments, the reaction mixture is maintained at the desired temperature for a period ranging from about 1 to about 8 hours.

[0040] Many known purification techniques can be used at various points during the methods of the invention. Such techniques can be used to remove excess reducing agent, exchange buffers or other components from the reaction mixture, and, if necessary, concentrate or dilute the antibody composition. Purification techniques useful in the methods of the invention include, but are not limited to, tangential flow filtration (TFF), gel filtration, immunoprecipitation, affinity chromatography, and the like. Preferably, purification time is minimized to prevent undesired oxidation or re-capping of uncapped engineered cysteines.

[0041] antibody complex The uncapped engineered cysteine ​​residues on the antibodies serve as useful means for the installation of various functional agents, such as imaging agents (e.g., chromophores and fluorophores), diagnostic agents (e.g., MRI contrast agents and radioisotopes), stabilizers (e.g., polyethylene glycol polymers), and therapeutic agents. Antibodies with uncapped cysteine ​​residues can be conjugated to functional agents to form antibody-functional agent-conjugates. The functional agents (e.g., drugs, detection agents, stabilizers) are conjugated (covalently attached) to the antibody at the engineered cysteine ​​residue. The functional agents can be attached indirectly through a linker or directly through a thiol-reactive group on the functional agent.

[0042] An antibody with an uncapped cysteine ​​residue can be conjugated to a drug to form an antibody-drug conjugate (ADC). Typically, an ADC includes a linker between the drug and the antibody. The linker can be a cleavable or non-cleavable linker. A cleavable linker is typically susceptible to cleavage under intracellular conditions such that cleavage of the linker releases the drug from the antibody at the target site. Suitable cleavable linkers include, for example, enzyme-cleavable linkers, including peptidyl-containing linkers that can be cleaved by intracellular proteases, such as lysosomal proteases, or endosomal proteases, or sugar linkers, such as glucuronide-containing linkers that can be cleaved by glucuronidase. Peptidyl linkers include, for example, dipeptides, such as valine-citrulline (Val-cit), phenylalanine-lysine (phe-lys), or valine-alanine (val-ala). Other suitable fragmentable linkers include, for example, pH-sensitive linkers (e.g., linkers that are hydrolyzable at a pH below 5.5, e.g., hydrazone linkers) and linkers that can be cleaved under reducing conditions (e.g., disulfide linkers). Non-cleavable linkers typically release the drug upon proteolysis of the antibody.

[0043] Prior to conjugation to the antibody, the linker will have a group reactive with the uncapped engineered cysteine ​​residue and attachment will be through that reactive group. Thiol-specific reactive groups are preferred and include, for example, maleimides; haloacetamides (e.g., iodo, bromo, or chloro); haloesters (e.g., iodo, bromo, or chloro); halomethylketones (e.g., iodo, bromo, or chloro); benzyl halides (e.g., iodide, bromide, or chloride); vinylsulfones; (pyridyl)disulfides; disulfide dioxide derivatives; mercury derivatives such as 3,6-bis(mercurymethyl)dioxane with acetate, chloride, or nitrate counterions; and polymethylenebismethanethiolphonates. The linker can include, for example, maleimides that are coupled to the antibody via a thio-succinimide linkage.

[0044] The drug can be any cytotoxic, cytostatic, or immunosuppressant drug. In embodiments where a linker connects the antibody and the drug, the drug has a functional group capable of forming a bond with the linker. For example, the drug can have an amine, a carboxylic acid, a thiol, a hydroxyl group, or a ketone capable of forming a bond with the linker. In embodiments where the drug is directly attached to the linker, the drug will have a group reactive with the uncapped engineered cysteine ​​prior to attachment to the antibody.

[0045] Useful classes of drugs include, for example, antitubulin agents, DNA minor groove binders, DNA replication inhibitors, alkylating agents, antibiotics, antifolates, antimetabolites, chemotherapy sensitizers, topoisomerase inhibitors, vinca alkaloids, or the like. Particularly useful classes of cytotoxic agents include, for example, DNA minor groove binders, DNA alkylating agents, and tubulin inhibitors. Exemplary cytotoxic agents include, for example, auristatins, camptothecins, duocarmycins, etoposide, maytansine, and maytansinoids (e.g., DM1 and DM4), taxanes, benzodiazepines, or benzodiazepine-containing drugs (e.g., pyrrolo[1,4]-benzodiazepines (PBDs), indolinobenzodiazepines, and oxazolidinobenzodiazepines), and vinca alkaloids. Benzodiazepine-containing drugs are described in WO 2010 / 091150, WO 2012 / 112798, WO 2007 / 085930, and WO 2011 / 023883.

[0046] According to some exemplary embodiments, suitable cytotoxic agents include, for example, DNA minor groove binders (e.g., enediynes and lexitropsins, CBI compounds; see also U.S. Patent No. 6,130,237), duocarmycins (see U.S. Publication No. 20060024317), taxanes (e.g., paclitaxel and docetaxel), puromycin, vinca alkaloids, CC-1065, SN-38, topotecan, morpholino-doxorubicin, rhizoxin, cyanomorpholino-doxorubicin, echinomycin, combretastatin, netropsin, epothilone A and B, estramustine, crypto, cemadotin, maytansinoids, discodermolide, eleutherobin, and mitoxantrone.

[0047] The drug may be an antitubulin agent. Examples of antitubulin agents include, but are not limited to, taxanes (e.g., Taxol® (paclitaxel), Taxotere® (docetaxel)), T67 (Tularik), and vinca alkaloids (e.g., vincristine, vinblastine, vindesine, and vinorelbine). Other antitubulin agents include, for example, baccatin derivatives, taxane analogs (e.g., epothilone A and B), codazole, colchicine and colcemid, estramustine, crypto, cemadotin, maytansinoids, combretastatins, discodermolide, auristatins, and eleutherobin.

[0048] The drug can be maytansine or maytansinoids, another group of antitubulin agents (see also Chari et al., 1992, Cancer Res. 52:127-131 and US Pat. No. 8,163,888).

[0049] The drug may be an acristatin. Auristatins include, but are not limited to, AE, AFP, AEB, AEVB, MMAF, and MMAE. The synthesis and structure of auristatins are described in U.S. Patent Application Publication Nos. 2003-0083263 and 2009-0111756; International Patent Publication No. 04 / 010957; International Patent Publication No. 02 / 088172; U.S. Patent No. 6,884,869; U.S. Patent No. 7,659,241; U.S. Patent No. 7,498,298; U.S. Patent No. 8,343,928; and U.S. Patent No. 8,609,105 (each of which is incorporated herein by reference in its entirety).

[0050] In some embodiments, the drug moiety is selected from the group consisting of antitubulin agents, DNA binding agents, and DNA alkylating agents. In some embodiments, the drug is selected from the group consisting of auristatins, pyrrolobenzodiazepines, duocarmycins, maytansinoids, taxanes, calicheamicins, and anthracyclines.

[0051] Drug-linkers can be used to form ADCs in a single step. According to other embodiments, bifunctional linker compounds can be used to form ADCs in a two-step or multi-step method. According to one example, an uncapped engineered cysteine ​​residue is reacted with a reactive moiety of a linker in a first step, and a functional group on the linker is reacted with a drug to form an ADC in a subsequent step.

[0052] In general, the functional group on the linker is selected for specific reaction with an appropriate reactive group on the drug moiety. As a non-limiting example, an azide-based moiety can be used for specific reaction with a reactive alkyne group on the drug moiety. The drug is covalently attached to the linker through 1,3-dipolar cycloaddition of the azide and alkyne. Other useful functional groups include, for example, ketones and aldehydes (suitable for reaction with hydrazides and alkoxyamines); phosphines (suitable for reaction with azides); isocyanates and isothiocyanates (suitable for reaction with amines and alcohols); and activated esters, such as N-hydroxysuccinimidyl esters (suitable for reaction with amines and alcohols). See, for example, Bioconjugate Techniques, 2000, pp. 211-215, 1999. nd These and other linking means are well known to those of skill in the art, as described in, for example, U.S. Pat. No. 6,393,363, and U.S. Pat. No. 6,393,363. Ed. (Elsevier). One of skill in the art will understand that when complementary pairs of reactive functional groups are selected for selective reaction of a drug moiety to a linker, each member of the pair can be used on either the linker or the drug.

[0053] Accordingly, some embodiments of the invention provide methods for preparing an uncapped antibody preparation, further comprising combining the uncapped antibody (e.g., including an uncapped antibody preparation) with a drug-linker compound under conditions sufficient to form an antibody-drug conjugate (ADC).

[0054] In some embodiments, the method includes combining an uncapped antibody with a bifunctional linker compound under conditions sufficient to form an antibody-linker conjugate. In such embodiments, the method of the invention can further include combining the antibody-linker conjugate with a drug moiety under conditions sufficient to covalently attach the drug moiety to the antibody through the linker.

[0055] In some embodiments, the ADC has the formula:

[0056] [ka] [wherein Ab is an antibody, LU is a linker, D is a drug, and the subscript p is 1 to 8].

[0057] In the above formula, the linker LU is conjugated to the antibody through an uncapped engineered cysteine. The value of the subscript p depends on the number of uncapped engineered cysteines available for conjugation. For example, for an antibody with two terminally capped engineered cysteines (one site on each H chain, or one site on each L chain), the value of p can be 2. Similarly, for an antibody with four uncapped engineered cysteines (e.g., two sites on each H chain, or two sites on each L chain, or one site on each H chain and one site on each L chain), the value of p can be 4. According to some preferred embodiments, p has a value of 1 to 4.

[0058] Drug burden The average number of drug-linker molecules per antibody (or average drug loading) is an important characteristic of the ADC composition, since it is the main determinant of the amount of drug that can be delivered to the target cell. The average drug loading includes the drug conjugated to the engineered cysteine ​​residue, and the drug conjugated to the site other than the intended engineered cysteine ​​residue, and the amount of unconjugated antibody in the composition. If an average drug loading of about 2 drugs per antibody is targeted, an antibody with two engineered cysteine ​​residues (e.g., one site on each H chain or one site on each L chain) can be used to prepare the ADC composition. If an average drug loading of about 4 drugs per antibody is targeted, an antibody with four engineered cysteine ​​residues (e.g., two sites on each H chain, or two sites on the L chain, or one site on the H chain and one site on the L chain) can be used to prepare the ADC composition. Those skilled in the art will appreciate that other levels of drug loading (including, for example, drug loading levels of 2 or less and drug loading levels of 4 or more) may be therapeutically useful depending on the particular antibody or the particular drug. Sites for drug conjugation may be introduced into the antibody by placing engineered cysteines at multiple sites or more than one site in the H chain, or by placing engineered cysteines at the sites in the L chain, or both. Importantly, the levels of decapping of engineered cysteine ​​residues described above allow for the preparation of ADC compositions with useful drug loading.

[0059] Typically, ADC compositions prepared with antibodies having two engineered cysteine ​​residues have an average drug loading of about 1.5 to 2.5 drugs per antibody. The average number of drug moieties per antibody can be, for example, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.2, 2.3, 2.4, or 2.5. In some embodiments, the average drug loading for ADC compositions prepared with antibodies having two engineered cysteine ​​residues is about 1.5 to about 2.2 or about 1.8 to about 2 drug moieties per antibody. Typically, ADC compositions prepared with antibodies having four engineered cysteine ​​residues have an average drug loading of about 3.4 to 4.5 drug moieties per antibody. The average number of drug moieties per antibody can be about 3.3, 3.4, 3.5, 3.6, 3.7, 3.8, 3.9, or 4.0. The average drug loading for ADC compositions prepared with antibodies having four engineered cysteine ​​residues is about 3.6 to about 4.2, or about 3.8 to about 4, drug moieties per antibody.

[0060] The methods of the invention are typically performed to selectively modify uncapped engineered cysteine ​​residues, however, varying degrees of unengineered (i.e., native) cysteine ​​residue modification are usually observed. Reaction conditions can be adjusted, if necessary, to limit modification of unengineered cysteine ​​residues. In certain cases, the methods can be performed to eliminate or minimize modification of unengineered cysteine ​​residues. For example, the methods can be performed such that less than about 20%, less than about 15%, less than about 10%, or less than about 5% of the antibody molecules in the ADC composition have a drug moiety covalently attached to an unengineered cysteine ​​residue. The methods can also be used to prepare ADC compositions in which the number of antibody molecules with modified unengineered cysteine ​​residues ranges from less than about 5%, 10%, 20%, 25%, or 30% of the total antibody molecules.

[0061] Various analytical methods can be used to determine the yield and isomeric mixture of the conjugate. Following conjugation of the drug to the antibody, the conjugated drug-antibody species can be separated. In some embodiments, the conjugated antibody species can be separated based on the characteristics of the antibody, drug and / or conjugate. Other techniques useful for the analysis of ADC compositions include, but are not limited to, reverse phase chromatography, capillary electrophoresis and mass spectrometry. For example, ADC compositions can be analyzed by LC / MS coupled with proteolytic digestion to determine the location of the drug moiety in the ADC.

[0062] antibody Many suitable antibodies can be used in the methods of the present invention. The antibodies used in the methods of the present invention are useful for many applications, including in vitro or in vivo diagnosis, in vivo imaging, and treatment of diseases and conditions associated with characteristic antigens. Five human antibody classes (IgG, IgA, IgM, IgD, and IgE), and various subclasses within those classes (e.g., IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2) are recognized based on structural differences, such as the number of immunoglobulin units in a single antibody molecule, the disulfide bridge structure of individual units, and differences in chain length and sequence. Antibody classes and subclasses are referred to as antibody isotypes.

[0063] The antibody can be an intact antibody or an antigen-binding antibody fragment, provided that the antibody fragment contains at least one interchain disulfide bond.

[0064] Typically, the antibodies are human, rodent (e.g., mouse and rat), donkey, sheep, rabbit, goat, guinea pig, camel, horse, or chicken. The antibodies can be, for example, murine, chimeric, humanized, or fully human antibodies, produced by techniques well known to those of skill in the art. Recombinant antibodies containing both human and non-human portions, e.g., chimeric and humanized monoclonal antibodies, which can be produced using standard recombinant DNA techniques, are useful antibodies. Chimeric antibodies are molecules in which different portions are derived from different animal species, e.g., those having a variable region derived from a murine monoclonal and a human immunoglobulin constant region (see, e.g., Cabilly et al., U.S. Pat. No. 4,816,567; and Boss et al., U.S. Pat. No. 4,816,397; which are incorporated herein by reference in their entireties). A humanized antibody is an antibody molecule from a non-human species having one or more complementarity determining regions from the non-human species and a framework region from a human immunoglobulin molecule (see, e.g., Queen, U.S. Pat. No. 5,585,089, which is incorporated herein by reference in its entirety). Such chimeric and humanized monoclonal antibodies can be produced by recombinant DNA techniques known in the art. As used herein, a "human" antibody includes antibodies having a human immunoglobulin amino acid sequence, e.g., as described in U.S. Pat. Nos. 5,939,598 and 6,111,166, and includes antibodies isolated from human immunoglobulins, human B cells, or animals transgenic for one or more human immunoglobulins.

[0065] Antibodies can be monospecific, bispecific, trispecific, or of greater multispecificity.

[0066] In certain cases, the constant domain has an effector function. The term antibody effector function, as used herein, refers to a function contributed by the Fc domain of an Ig. Such a function may be brought about, for example, by binding of the Fc effector domain to an Fc receptor on a phagocyte or an immune cell with lytic activity, or by binding of the Fc effector domain to a component of the complement system. The effector function may be, for example, "antibody-dependent cellular cytotoxicity" or ADCC, "antibody-dependent cellular phagocytosis" or ADCP, "complement-dependent cytotoxicity" or CDC. In certain cases, the constant domain lacks one or more effector functions. Conjugation of a drug linker compound to an engineered cysteine ​​residue located in the effector function binding domain can modulate the effector function.

[0067] The antibodies may be directed against any antigen of medical and / or therapeutic interest. For example, the antigen may be associated with a pathogen (such as, but not limited to, viruses, bacteria, fungi, and protozoa), a parasite, a tumor cell, or a particular medical condition. In the case of tumor-associated antigens (TAA), the cancer may be of the immune system, lung, colon, rectum, breast, ovary, prostate, head, neck, bone, or any other anatomical location. Antigens of interest include, but are not limited to, CD30, CD40, Lewis Y, CD70, CD2, CD20, CD22, CD33, CD38, CD40, CD52, HER2, EGFR, VEGF, CEA, HLA-DR, HLA-DR10, CA125, CA15-3, CA19-9, L6, Lewis X, alpha-fetoprotein, CA242, placental alkaline phosphatase, prostate specific antigen, prostatic acid phosphatase, epidermal growth factor, MAGE-1, MAGE-2, MAGE-3, MAGE-4, anti-transferrin receptor, P97, MUC1-KLH, gp100, MART1, IL-2 receptor, human chorionic gonadotropin, mucin, P21, MPG, and Neu oncogene product.

[0068] Some particularly useful antibodies include, but are not limited to, the following: antibodies against the CD33 antigen (e.g., the humanized 2H12 antibody as described in International Application No. WO 2013 / 173496), antibodies against the CD70 antigen (e.g., the humanized 1F6 antibody as described in International Application No. WO 2006 / 113909), antibodies against the CD30 antigen (e.g., the humanized AC10 antibody as described in International Application No. WO 2008 / 025020), antibodies against the CD19 antigen (e.g., the humanized BU12 antibody as described in International Application No. WO 2009 / 052431), antibodies against IIV-1, NTBA, or αV β6. Many other internalizing antibodies that bind tumor-specific antigens can be used and have been reviewed (see, e.g., Franke et al. (2000), Cancer Biother Radiopharm. 15:459-76; Murray (2000), Semin Oncol. 27:64-70; Breitling et al., Recombinant Antibodies, John Wiley, and Sons, New York, 1998). The disclosures of those references and international applications are incorporated herein by reference.

[0069] In some embodiments, the present disclosure provides a method for preparing an antibody having an uncapped engineered cysteine ​​residue as described above, wherein the antibody comprises at least three interchain disulfide bonds. In some embodiments, the antibody comprises at least four interchain disulfide bonds. In some embodiments, the antibody comprises one, two, three, four or five interchain disulfides. In some embodiments, the engineered cysteine ​​residue is present in the heavy or light chain constant region of the antibody.

[0070] Engineered cysteine ​​residues The site of the engineered cysteine ​​can have an effect on the properties of the ADC. For example, an engineered cysteine ​​that is completely buried in the protein's structure is difficult to conjugate due to poor access to solvent, whereas an engineered cysteine ​​on the outer surface of the antibody can result in an ADC with impaired stability due to prolonged exposure to substances in plasma. Also, ADCs prepared from ecomAbs with highly surface-exposed engineered cysteines are sensitive to the hydrophobicity of the drug, whereas engineered cysteines at more protected positions are less sensitive to the properties of the drug due to limited access to other materials in solution. The location of the engineered cysteine ​​residue can also be used to modulate the effector function as desired for a particular ACD. For example, conjugation of a drug-linker to an engineered cysteine ​​residue in an effector function binding domain can be used to block binding to an effector function-mediated receptor.

[0071] In some embodiments, the engineered cysteine ​​is located in the heavy chain constant region, the heavy chain variable region, the light chain variable region, the light chain constant region, or a combination thereof. Preferred engineered cysteine ​​residues are those located in moieties that can be conjugated and provide stable linkages. By conjugated, it is meant that the engineered cysteine ​​residue can be attached to a functional agent (e.g., an imaging agent, a diagnostic agent, a stabilizing agent, or a therapeutic agent) without first denaturing the antibody. Methods for selecting sites for introducing cysteine ​​residues that can then be conjugated to a functional agent are known in the art (see, for example, Junutula et al., 2008, Nature Biotechnology, 26(8), 925-932).

[0072] In some embodiments, the modified cysteine ​​residue is a cysteine ​​residue having a fractionation solvent accessibility of 10% or more, 20% or more, 30% or more, 40% or more, or 50% or more. In some embodiments, the cysteine ​​residue is a cysteine ​​residue having a fractionation solvent accessibility of about 10% to about 95%, about 10% to about 85%, about 10% to about 75%, about 10% to about 60%, about 20% to about 95%, about 20% to about 85%, about 20% to about 75%, about 20% to about 60%, about 40% to about 95%, about 40% to about 85%, about 40% to about 75%, or about 40% to about 60%. Methods for determining the fractional solvent accessibility of residues at specific sites are known in the art and can be determined, for example, using the online server getarea using the methods described in Fraczkiewicz and Braun, 1998, J. Comp. Chem., 19, 319-333 (see http: / / curie.utmb.edu / getarea.html). Exemplary residues include those residues at positions 15, 114, 121, 127, 168, 205 (numbered according to Kabat) on the L chain, or positions 112, 114 or 116 (numbered according to Kabat) on the H chain. Exemplary residues include those residues in the Fc region of an IgG1 antibody, for example those residues at positions 239, 326 or 269 in the Fc region (numbered according to the EU index). The fractional solvent accessibility of residues at positions 239, 326 and 327 is approximately 50%, approximately 94% and approximately 23%, respectively.

[0073] One of skill in the art will appreciate that the conditions required for selective activation of engineered cysteines will depend on the site of the engineered cysteine ​​in the antibody. The formula in Scheme 2 provides a framework for selecting reduction conditions that allow for selective activation of the engineered cysteines wherever they occur within the protein sequence. In some embodiments, the antibodies have 1-8, or 2-8, or 2-4 engineered cysteine ​​residues.

[0074] Non-antibody proteins Although the process described herein is illustrated with respect to antibodies, it will be understood by those skilled in the art that it can be advantageously used for any protein that has an unpaired cysteine ​​(a cysteine ​​that does not generally form inter- or intra-chain bonds within a protein), modified or native, that is capped with a thiol during expression or production. Proteins for which this process is particularly useful are those that contain native cysteines that, in addition to containing an unpaired cysteine, form interchain disulfide bonds, particularly bonds that can be cleaved without immediately resulting in unfolding of the protein. When referring to non-antibody proteins, the term interchain disulfide bond refers to a covalent bond between two cysteine ​​residues on adjacent polypeptide chains. Candidate non-antibody proteins include those proteins that contain a solvent-exposed disulfide bond whose stability in the native folded conformation is comparable to that of a capped thiol. A modified cysteine ​​protein, as used herein, is a protein in which a selected amino acid in the protein is mutated to a cysteine. Exemplary proteins also include Fc-fusion proteins, eg, proteins comprising the Fc region of an antibody covalently linked to a protein that provides specificity for a desired target. EXAMPLES

[0075] IV. Working Examples Example 1: Exemplary method for carrying out selective reduction The reaction vessel is connected to a tangential flow filtration (TFF) apparatus and an ultrafiltration membrane is installed (e.g., 88 cm 2, Millipore Pelicon 3, regenerated cellulose). (Many different membrane types with appropriate size can be used for this antibody. The flux rate and sieving coefficient required for calculation of the addition rate for the reducing agent should be determined with the permeate flux and with the membrane type and surface area that will be used in this step). The membrane is flushed and equilibrated according to the manufacturer's instructions. A diafiltration buffer reservoir is set up to contain a buffer (e.g., 50 mM Tris / 5 mM EDTA, pH 8.0). It is connected to the reaction vessel via tubing, and the flow through the tubing is controlled by a peristaltic pump (e.g., diafiltration buffer pump). The capped, modified cysteine-containing mAb is placed in the reaction vessel. The reaction is carried out at room temperature and the reaction mixture is continuously pumped through the ultrafiltration membrane with a retentate line that is constricted to maintain a transmembrane pressure of approximately 20 psi.

[0076] A reservoir containing a solution of reducing agent is connected via tubing to the diafiltration buffer line or other location in the flow path or reaction vessel. The flow rate is calculated at which point a stock solution containing reducing agent should be added to the reaction mixture to maintain the initial (desired) concentration of reducing agent in the reaction mixture. (The optimal concentration of reducing agent in the reaction mixture is determined experimentally using the techniques described herein and is different for different thiols due to their different reducing strengths, as described in FIG. 1). The concentration of reducing agent should remain mostly constant over the entire reaction time. The diafiltration buffer is also pumped into the reaction vessel at a rate that is adjusted so that the total reaction volume remains constant (as in constant-volume-diafiltration), i.e., the introduction rate of reducing agent and diafiltration buffer into the reactor is consistent with the volume loss rate via permeation.

[0077] Example 2: Decapping of engineered cysteine ​​residues using cysteine, N-acetylcysteine, and cysteamine as reducing agents without removal of the cap by-product 20 mg (135 nmoles at 15 mg / ml) of S239C engineered cysteine ​​antibody (IgG1 antibody with an engineered cysteine ​​residue at position 239; numbered according to the EU index) were treated with 33 μl of 100 mM (3.3 μmoles) cysteine, N-acetylcysteine, or cysteamine for 8 hours at pH 8.0 and room temperature. Samples were removed at 1 hour intervals, purified, conjugated with excess SGD-1269, and stored frozen. Purification and conjugation quenched the reduction reaction and preserved the thiols generated by the reduction step. Conjugated samples were analyzed by reversed-phase HPLC under denaturing conditions ([rPLRP]), from which the extent of engineered cysteine ​​decapping was estimated (conversion of H0 to H1). The extent of interchain disulfide cleavage was determined from the amount of H chains with one or more mcMMAF conjugated (mcMMAF is a maleimidocaproic acid linker attached to the drug monomethyl auristatin F). No H chains with two or more mcMMAF molecules conjugated were observed. Results from a static solution reduction experiment are shown in FIG. 1. The results show that each of the three thiols can react with ecmAb to decapture engineered cysteines, however, the three thiols have very different behaviors. NAC behaves in a manner most directly described by the first reaction in Scheme 1: the reaction proceeds until a sufficient concentration of disulfide 4 accumulates, and then stops, as forward and reverse reactions occur at the same rate. Cysteine ​​behaves as a stronger reducing agent than NAC, however, instead of stopping with a partially activated ecmAb, the reaction reverses and regenerates ecmAb. This reversal suggests that an additional reaction is involved with cysteine, namely autoxidation of the reducing agent.

[0078] [ka]

[0079] Thus, the cysteine ​​produced by the autoxidation of cysteine ​​also recaps the activated, modified cysteine ​​and reverses the initial reduction. Examination of Figure 1 shows that this same phenomenon, i.e., initial reduction followed by recapping, also occurs with cysteamine; however, cysteamine is a weaker reducing agent, and as a result, the initial extent of reduction was not as high as with cysteine ​​or N-acetylcysteine.

[0080] Example 3: Selective activation of engineered cysteine ​​residues in ecmAb The reaction vessel was connected to a tangential flow filtration (TFF) apparatus and an ultrafiltration membrane was installed (e.g., 88 cm 2 The membrane was flushed and equilibrated according to the manufacturer's instructions. The diafiltration buffer reservoir was set to contain 50 mM Tris / 5 mM EDTA, pH 8.0. It was connected to the reaction vessel via tubing, and the flow through the tubing was controlled by a peristaltic pump (e.g., diafiltration buffer pump). A capped engineered cysteine-containing engineered cysteine ​​mAb (h2H12 S239C ecMab, h1F6 239 ecMab, h2H12 K326C ecMab, or h2H12 A327C ecMab; numbered according to the EU index) was placed in the reaction vessel at a concentration of about 15 mg / ml, and the pH was adjusted to 8.0. The reaction was carried out at room temperature and the reaction mixture was continuously pumped through the ultrafiltration membrane with the retentate line constricted to maintain a transmembrane pressure of approximately 20 psi.

[0081] A reservoir containing the cysteine ​​solution was connected via tubing to the diafiltration buffer line. The flow rate was calculated at which point the cysteine ​​stock solution should be added to the reaction mixture to maintain the initial (desired) concentration of cysteine ​​in the reaction mixture. The concentration of the cysteine ​​stock provided in the reaction mixture and the rate of addition were calculated as follows, so that the concentration of cysteine ​​remained mostly constant over the entire reaction time. In the examples shown in Figures 2-5, the reaction concentrations of cysteine ​​ranged from 0.5 to 0.9 mM (as described in the figure legends); the stock concentrations of cysteine ​​were either 100 mM, 10 mM, or 5 mM, and the flow rate of cysteine ​​injected into the reactor was adjusted to maintain the experimental cysteine ​​concentration according to the following formula: The sieving coefficient for cysteine ​​was measured at 0.8 to 0.9. Cysteine ​​concentrations in the reaction mixtures were determined periodically throughout the experiment using a DTNB assay to ensure that cysteine ​​levels remained at the desired concentrations (not shown).

[0082] Diafiltration buffer is also pumped into the reaction vessel at a rate adjusted so that the total reaction volume remains constant (as in constant-volume-diafiltration), i.e., the rate of introduction of cysteine ​​and diafiltration buffer into the reactor matches the rate of volume loss via permeation.

[0083] Samples were removed at the intervals indicated in the figure, purified by elution through a PD-10 column, and conjugated with SGD-1269. The conjugated samples were then analyzed by rPLRP. Analysis of the rPLRP chromatograms provided the percentage of H chains that remained capped (%H0), selectively reduced at the engineered cysteine ​​(%H1), or reduced at the engineered cysteine ​​and also at the interchain disulfide sites (%H2; non-selective reduction).

[0084] Calculation of cysteine ​​addition rate Cysteine ​​was added to the reaction mixture at the same rate that cysteine ​​was lost through the permeate line. The initial rate of analyte loss through the permeate line was calculated from Equation 2, which can be derived from the theoretical equation for clearance by constant volume diafiltration, i.e., Equation 1. The rate at which cysteine ​​stock solution was pumped into the reaction mixture to maintain the initial cysteine ​​concentration is given by Equation 3.

[0085] Equation 1: C / C0=e(-NS). C is the concentration at any time t; C0 is the initial concentration; N is the number of diavolumes at time t; and S is the sieving coefficient for the analyte (determined empirically). The number of diavolumes N is r*t / Vd, where r is the permeate flow rate and Vd is the batch volume. Making this substitution, obtaining the derivative, and evaluating the derivative at t=0 gives Equation 2.

[0086] Equation 2: dC / dt = -C0 * (r / Vd) * S, where dC / dt is the rate of loss of cysteine ​​at t0. The rate at which cysteine ​​solution should be added to reach a given concentration in the reaction mixture is given by Equation 3.

[0087] Equation 3: R = -dC / dt * Vd / [Cys], where [Cys] is the concentration of the cysteine ​​stock solution.

[0088] Thus, substituting the expression for dC / dt in Equation 2 gives Equation 4 for the addition rate. Equation 4: R = C 0- * r * S / [Cys] = (0.8 mM) * (720 mL / hr) * (0.8) / (100 mM) = 4.61 mL / hr.

[0089] result Selective activation of engineered cysteines in ecmAb was achieved by continuously diafiltering ecmAb during the reaction, replenishing cysteine ​​in the diafiltration buffer throughout the process. Diafiltration continuously depleted reaction by-products. The rate of cysteine ​​addition was calculated based on the theoretical rate of cysteine ​​clearance by diafiltration to provide a constant concentration of cysteine ​​in the reaction mixture. Figure 2 shows approximately 80% activation of available engineered cysteines after 4.5 hours (%H1), with negligible cleavage of interchain disulfide bonds (as indicated by %H2). This experiment was repeated with 0.6 mM cysteine ​​for an extended period of time, with similar results (Figure 3). At higher concentrations, the reaction was faster and somewhat less specific. It will be appreciated that the reaction mixture is in a state of flux throughout the process, with the cysteine ​​solution entering the reaction mixture via the cysteine ​​feed pump and exiting the reaction mixture via the TFF membrane (permeating), and as a result, its concentration may vary somewhat during the reaction, but activation conditions may nevertheless generally be maintained near the nominal concentration. Cysteine ​​concentrations of 1.5-2 mM and higher resulted in enhanced interchain disulfide cleavage. Figure 4 shows that engineered cysteine ​​activation can be selectively achieved at various sites, regardless of the identity of the antibody. Over the course of activation time, the K326 mutant suggests that activation is faster at this site than at the S239 or A327 sites. This is likely due to the fact that the K326 site is more solvent accessible, however, the method nevertheless achieves the desired selective activation. FIG. 5 shows that the reaction can proceed to very close to 100% activation of the engineered cysteines with very good selectivity by using low concentrations of reducing agent for extended periods of time.

[0090] Although the foregoing has been described in some detail by way of illustration and example, for purposes of clarity and understanding, those skilled in the art will understand that certain changes and modifications may be practiced within the scope of the appended claims. Further, each reference provided herein is incorporated by reference in its entirety into this specification to the same extent as if each reference was individually incorporated by reference. [1] 1. An uncapped antibody preparation prepared by a process for selectively reducing one or more capped engineered cysteine ​​residues in an intact antibody, the process comprising the steps of: a) contacting antibody molecules, each antibody molecule having at least one capped engineered cysteine ​​residue, with a reducing agent; b) reacting the antibody molecule with the reducing agent under conditions sufficient to uncap the engineered cysteine ​​residues and form cap by-products; and c) removing said cap by-product during step b); whereby an uncapped engineered cysteine ​​antibody preparation is formed, and substantially all of the heavy-light and heavy-heavy interchain disulfide bonds are retained in the uncapped engineered cysteine ​​antibody preparation. [2] 2. The uncapped antibody preparation of item 1, wherein the method comprises supplementing the reduction reaction with additional reducing agent while removing cap by-products. [3] The preparation comprises: an antibody molecule having at least two uncapped engineered cysteine ​​residues and no capped engineered cysteine ​​residues; an antibody molecule having at least two capped engineered cysteine ​​residues and no uncapped engineered cysteine ​​residues; and Antibody molecules having at least one capped engineered cysteine ​​residue and at least one uncapped engineered cysteine ​​residue. 3. The uncapped antibody preparation of item 1 or 2, comprising two or more members selected from the group consisting of: [4] (a) at least 90% of the free thiols present in the antibody preparation are derived from engineered cysteine ​​residues; and / or (b) removing the cap by-product during the reduction reaction comprises dialysis or diafiltration; and / or (c) The uncapped antibody preparation of any one of items 1 to 3, wherein prior to the reduction reaction, each antibody molecule comprises at least three (preferably four) interchain disulfide bonds. [5] 5. The uncapped antibody preparation of any one of items 1 to 4, wherein each antibody molecule has at least two engineered cysteine ​​residues, preferably each antibody molecule has four engineered cysteine ​​residues. [6] 5. The uncapped antibody preparation of any one of items 1 to 4, wherein the engineered cysteine ​​residue is present in a heavy chain constant region of the antibody molecule; or the engineered cysteine ​​residue is present in a heavy or light chain variable region of the antibody molecule. [7] 7. The uncapped antibody preparation of any one of items 1 to 6, wherein the reducing agent is selected from the group consisting of cysteine, cysteamine, beta-mercaptoethanol, 2-mercaptoethanesulfonic acid sodium salt, and mixtures thereof. [8] 8. The uncapped antibody preparation of item 7, wherein the method comprises maintaining a reducing agent at a concentration about 5-fold to about 15-fold higher, preferably about 5-fold to about 10-fold higher than a concentration of the antibody during the reduction reaction. [9] 9. The uncapped antibody preparation of item 7 or 8, wherein the reducing agent is cysteine; preferably the concentration of cysteine ​​is maintained at a concentration of 0.5 mM to about 1.5 mM during the reduction reaction.

[10] (a) the concentration of the cap by-product is maintained below a concentration that prevents further activation of the engineered cysteine ​​residue by re-capping; (b) The uncapped antibody preparation of any one of items 1 to 9, wherein the uncapped engineered cysteine ​​antibody preparation is a monoclonal antibody preparation.

[11] 11. The uncapped antibody preparation of any one of items 1 to 10, wherein the method further comprises the step of removing residual reducing agent from the uncapped antibody preparation and / or further comprises the step of purifying the antibody preparation.

[12] 12. An antibody-drug conjugate (ADC) composition prepared by a method comprising combining the uncapped antibody preparation of any one of items 1 to 11 with a drug-linker compound under conditions sufficient to form an antibody-drug conjugate, The composition, wherein the average drug loading of the antibody-drug conjugate is 4 drug moieties per antibody, or about 3.6-4.2 drug moieties per antibody, or about 2 drug moieties per antibody, or 1.5-2.2 drug moieties per antibody; and wherein the antibody-drug conjugate is optionally in solution.

[13] 1. A method for producing an uncapped antibody composition, the method comprising: a) contacting antibody molecules, wherein each antibody molecule has at least one capped engineered cysteine ​​residue, with a reducing agent; b) reacting the reducing agent with the antibody molecule under conditions sufficient to uncap the engineered cysteine ​​residues and form cap by-products; and c) removing said cap by-product during step b). wherein an uncapped antibody composition is formed and substantially all of the heavy-light and heavy-heavy interchain disulfide bonds are retained in the uncapped antibody.

[14] Item 14. The method of claim 13, wherein the method comprises supplementing the reduction reaction with additional reducing agent while removing the cap by-product.

[15] The uncapped antibody composition comprises: an antibody molecule having at least two uncapped engineered cysteine ​​residues and no capped engineered cysteine ​​residues; an antibody molecule having at least two capped engineered cysteine ​​residues and no uncapped engineered cysteine ​​residues; and Antibody molecules having at least one capped engineered cysteine ​​residue and at least one uncapped engineered cysteine ​​residue. 15. The method according to item 13 or 14, comprising two or more members selected from the group consisting of:

[16] (a) at least 90% of the free thiols present in the uncapped antibody composition are derived from engineered cysteine ​​residues; and / or (b) removing the cap by-product during the reduction reaction comprises dialysis or diafiltration; and / or (c) The method according to any one of items 13 to 15, wherein prior to the reduction reaction, each antibody molecule in the composition comprises at least three (preferably four) interchain disulfide bonds.

[17] 17. The method according to any one of items 13 to 16, wherein each antibody molecule in the composition has at least two engineered cysteine ​​residues, preferably each antibody molecule has four engineered cysteine ​​residues.

[18] 17. The method of any one of items 13 to 16, wherein the engineered cysteine ​​residue is present in a heavy chain constant region of the antibody molecule in the composition; or the engineered cysteine ​​residue is present in a heavy or light chain variable region of the antibody molecule in the composition.

[19] 19. The method according to any one of items 13 to 18, wherein the reducing agent is selected from the group consisting of cysteine, cysteamine, β-mercaptoethanol, 2-mercaptoethanesulfonic acid sodium salt, and mixtures thereof.

[20] 20. The method according to item 19, wherein the method comprises maintaining a reducing agent at a concentration about 5 to about 15 times higher, preferably about 5 to about 10 times higher, than a concentration of the antibody during the reduction reaction. [twenty one] 21. The method according to item 19 or 20, wherein the reducing agent is cysteine; preferably the concentration of cysteine ​​is maintained at a concentration of 0.5 mM to about 1.5 mM during the reduction reaction. [twenty two] (a) the concentration of the cap by-product is maintained below a concentration at which re-capping prevents further activation of the engineered cysteine ​​residue; and / or (b) the uncapped antibody is an uncapped monoclonal antibody; The method according to any one of Items 13 to 21. [twenty three] 23. The method according to any one of items 13 to 22, wherein the method further comprises the step of removing residual reducing agent from the method and / or further comprises the step of purifying the uncapped antibody. [twenty four] 24. A method for producing an antibody-drug conjugate (ADC) composition, the method comprising combining an uncapped antibody composition of any one of items 13 to 23 with a drug-linker compound under conditions sufficient to form an antibody-drug conjugate; optionally, the average drug loading of the antibody-drug conjugate is 4 drug moieties per antibody, or about 3.6-4.2 drug moieties per antibody, or about 2 drug moieties per antibody, or 1.5-2.2 drug moieties per antibody; wherein the antibody-drug conjugate is optionally in solution.

Claims

[Claim 1] The invention described in the specification.

Images