Method for preparing pH-dependent antibodies

Abstract

A method for preparing engineered antibodies that exhibit improved pH-dependent antigen binding is disclosed. The method is based on the introduction of histidine residues at a subset of defined amino acid positions in antibody CDRs. The set of amino acid positions selected for histidine substitution is derived from a heat map of histidine occurrence in the CDRs of functional antibodies from natural antibody repertoires. The method provides a simpler and less time-consuming approach to identifying pH-dependent antibody variants. [Selected Figure] Figure 1

Description

[Technology Field]

[0001] (Field of invention) This disclosure relates to the field of antibody engineering, particularly to methods for manipulating antibody variants that exhibit pH-dependent binding to target antigens, and to antibodies prepared by such methods. [Background technology]

[0002] (background) There is growing interest in manipulating antibodies that exhibit pH-dependent binding to target antigens. In particular, there is interest in identifying antibodies that show decreased antigen binding at acidic pH compared to neutral pH.

[0003] By incorporating pH-sensitive antigen binding, the in vivo function of manipulated antibodies can be improved. Manipulated antibodies may retain high affinity antigen binding at neutral pH (e.g., pH 7.4) and show decreased binding at acidic pH (e.g., pH 4.5–6.0). Upon entering the endosomal pathway, pH-dependent antigen binding enables dissociation of the antibody-antigen complex in acidified endosomes (approximately pH 6.0) and recycling of the free antibody mediated by FcRn, thereby promoting enhanced antigen clearance, which can lead to lower antibody frequencies or reduced antibody doses.

[0004] Typically, pH-dependent antibody-antigen binding relies on the presence of ionic histidine residues that mediate structural changes in the binding or folding of the interacting antibodies. Changes in electrostatic interactions induced by histidine protonation at low pH values ​​can reduce binding affinity.

[0005] Various protein engineering approaches have been described that aim to incorporate pH sensitivity into proteins by using various strategies for histidine substitution. For example, rational design guided by structural modeling of the receptor interaction site of granulocyte colony-stimulating factor (GCSF) has made it possible to identify mutation sites that yield pH-sensitive variants when histidine is substituted (Sarkar et al., Nat Biotechnol, Vol. 20, pp908-13, 2002). However, evaluating individual histidine substitution sites that can mediate pH sensitivity in proteins is time-consuming and unpredictable.

[0006] As an alternative approach, combinatorial library screening using suitable high-throughput technologies, such as yeast surface display (YSD) or phage display, has been applied to manipulating pH-sensitive binding in various different protein scaffolds. Schroter et al., mAbs, vol. 7(1), 138-151, 2015, describe a general strategy for manipulating antibody heavy and light chain variable domains for reversible pH-sensitive antigen binding. This approach relies on the construction and screening of combinatorial histidine substitution libraries of heavy and light chains. In this approach, histidine mutations are randomly introduced into the complementarity-determining regions (CDRs) of the VH and VL domains. Histidine mutations can be introduced at any amino acid position in the CDRs of the VH and VL domains. The rate of histidine mutations is adjusted to achieve an average of three random mutations per variant in the library. Therefore, the complete library theoretically includes variants containing all possible combinations of the three random histidine substitutions in the CDR.

[0007] This combinatorial library approach enabled the successful identification of engineered variants of adalimumab that exhibit substantial pH-dependent binding to TNF. However, this strategy is relatively time-consuming because of the large library sizes required to sample all amino acid positions in the CDRs of the VH and VL domains due to the possible histidine substitutions. Furthermore, in this strategy, the overall antibody structure and stability due to the introduction of histidine residues at unfavorable positions are not considered. Summary of the Invention

[0008] (Summary of the Invention) There is a need for an alternative method for preparing engineered antibody variants that exhibit pH-dependent binding to a target antigen, which is simpler and less time-consuming to implement than prior art methods. Furthermore, there is a need for a method that enables the identification of engineered antibodies that exhibit pH-dependent antigen binding and do not unduly affect the antigen-binding ability at neutral pH.

[0009] To address these needs, the applicant conducted a systematic analysis of the natural occurrence rates of histidine residues in the complementarity-determining regions (CDRs) of a large repertoire of functional (i.e., antigen-binding) antibody variable domains. As a result of this analysis, the applicant was able to derive a "heat map" of the natural histidine occurrence rates in native antibodies. This "heat map" provides both a list of "hot spot" amino acid residues in the CDRs where histidine can naturally occur in the antibody repertoire, and also a list of "cold spot" amino acid residues in the CDRs where histidine typically does not naturally occur. These "hot spot" and "cold spot" lists can be applied, for example, to the engineering of pH-dependent antibody variants in rational design or combinatorial approaches.

[0010] In a first aspect, the present invention is a method for preparing an engineered antibody that exhibits pH-dependent binding to its antigen, comprising: preparing the engineered antibody by substituting at least one amino acid residue of a parent antibody with a histidine residue, and the following hot spot list: [Table 1] substituting at least one amino acid residue selected from with histidine, and the following cold spot list: [Table 2] wherein at least one amino acid residue from is not substituted with histidine, and the engineered antibody exhibits pH-dependent binding to its antigen, provided is the above method.

[0011] As a result of the substitution with the histidine, the engineered antibody exhibits an improvement in pH-dependent binding (i.e., higher pH-dependence) compared to the parent antibody.

[0012] The present invention is a method for preparing an engineered antibody that exhibits pH-dependent binding to its antigen, comprising: (a) providing a parent antibody that binds to an antigen; (b) preparing a panel of engineered variants of the parent antibody, wherein each of the engineered variants in the panel differs from the parent antibody by having at least one amino acid residue substituted with a histidine residue, and at least one amino acid residue selected from the following hot spot list: [Table 3] is substituted with histidine, and at least one amino acid residue from the following cold spot list: [Table 4] is not substituted with histidine, the above step; (c) screening the panel of engineered variants for pH-dependent binding to the antigen, thereby identifying an engineered antibody that exhibits pH-dependent binding to the antigen, further provided is the above method.

[0013] The manipulated antibody identified in step (c) exhibits improved pH-dependent binding (i.e., higher pH dependence) compared to the parent antibody as a result of the histidine substitution.

[0014] The present invention relates to a method for preparing an engineered antibody that exhibits pH-dependent binding to an antigen, and includes: (a) A step to identify the parent antibody that binds to the antigen; (b) A step of preparing a panel of manipulated variants of the parent antibody, wherein each of the manipulated variants in the panel differs from the parent antibody by having at least one amino acid residue substituted with a histidine residue, as shown in the following hotspot list: [Table 5] At least one amino acid residue selected from is substituted with histidine, and the following cold spot list: [Table 6] The step wherein at least one amino acid residue from is not substituted with histidine; (c) A step of screening a panel of the manipulated variants for pH-dependent binding to the target antigen, thereby identifying selected amino acid positions in which pH-dependent binding to the antigen is conferred by the presence of histidine; (d) the step of preparing one or more further modified variants of the parent antibody, wherein each of the variants contains histidine at two or more selected amino acid positions identified in step (c); and (e) The method further comprises the step of screening the further modified variant for pH-dependent binding to the antigen; thereby identifying a modified antibody that exhibits pH-dependent binding to the antigen.

[0015] The manipulated antibody identified in step (e) exhibits improved pH-dependent binding (i.e., higher pH dependence) compared to the parent antibody as a result of the histidine substitution.

[0016] In a second aspect, the present invention relates to an engineered antibody exhibiting pH-dependent binding to an antigen, wherein at least one amino acid in the CDR of the engineered antibody is a histidine residue, and the following hotspot list: [Table 7] At least one amino acid residue selected from is a histidine residue, and the following cold spot list: [Table 8] The invention provides the manipulated antibody characterized in that at least one amino acid residue selected from is not a histidine residue. [Modes for carrying out the invention]

[0017] (Detailed explanation) This disclosure provides a method for preparing an engineered antibody exhibiting pH-dependent binding to a target antigen, the method comprising substituting selected amino acids in the variable domain of a parent antibody (and preferably in its CDR) with histidine residues. It is generally known in the art that introducing ionic histidine residues into the VH and / or VL domains of an antibody molecule can alter the pH dependence of antigen binding. However, until now, methods for manipulating pH-dependent antigen binding have largely relied on screening random combinatorial libraries to be sampled for possible histidine substitutions of all amino acid residues in the CDR.

[0018] This method differs from conventional methods in that it guides the selection of specific amino acid residues to be substituted with histidine based on the natural occurrence rate of histidine residues at specific amino acid positions in the VH and VL domains of a functional antibody molecule, more specifically in its CDR, and the natural absence rate of histidine residues at other amino acid positions in the CDR. Accordingly, the method disclosed herein is based on histidine substitution at amino acid positions selected from a pre-selected subset of amino acid positions in the variable domain (and in particular the CDR), while other residues in the CDR are preferably pre-selected as not to be substituted with histidine.

[0019] In this specification, amino acid residues pre-selected as candidates for histidine substitution are referred to as "hotspot" residues (Table A). These are amino acid positions in the CDR or framework positions near the CDR, at which histidine is observed to appear naturally in the functional antibody repertoire.

[0020] Table A (Hotspot residue locations) [Table 9]

[0021] In this specification, amino acid residues that are preferably selected in advance not to be substituted with histidine are referred to as “cold spot” residues (Table B). These are amino acid positions in the CDR or framework positions near the CDR, where histidine is not observed in the functional antibody repertoire.

[0022] Table B (Cold Spot Residue Locations) [Table 10]

[0023] This disclosure provides a method for preparing a modified antibody exhibiting pH-dependent binding to an antigen, comprising the method of preparing the modified antibody by substituting at least one amino acid residue of a parent antibody with a histidine residue (which may be abbreviated herein as HIS). As a result of the histidine substitution, the modified antibody exhibits improved pH-dependent binding (i.e., higher pH dependence) compared to the parent antibody. The “parent antibody” can be any antibody capable of binding to the antigen of interest. The parent antibody is typically a conventional tetrachain immunoglobulin, in which antigen-binding specificity is conferred by paired VH and VL domains. However, this method is also applicable to the manipulation of pH-dependent binding in other parent antibodies, as well as in any other modified immunoglobulin configurations including antigen-binding fragments, e.g., Fab, F(ab'), F(ab')2, Fv, scFv, diabody, triabody, minibody, etc., and antigen-binding sites provided by paired VH and VL domains. In some embodiments, the parent antibody or its variable domain or CDR may be derived from a camelid animal species, including, but not limited to, llamas, camels, dromedaries, alpacas, vicuñas, or guanacos. The preparation of camelid monoclonal antibodies is described in detail in WO2010 / 001251, the contents of which are incorporated herein by reference. Other preferred characteristics of the parent antibody and the engineered antibodies derived therefrom are described elsewhere herein. The properties of the antigen to which the parent antibody binds are not particularly limited.

[0024] The engineered antibody is prepared by substituting one or more selected amino acid residues of the parent antibody with histidine. The engineered antibody may contain a total of 1, 2, 3, 4, 5, or more histidine substitutions compared to the parent antibody. The selected amino acid residues substituted with histidine are typically located in or near the CDR of the antibody variable domain. In the case of an antibody molecule containing paired VH and VL domains, the residue selected for histidine substitution may be located in the CDR of the VH domain, the CDR of the VL domain, or in the CDR of both the VH and VL domains.

[0025] As described herein, at least one amino acid position to be substituted with histidine must be selected from the “hotspot” list (Table A). In one embodiment, at least two, at least three, at least four, or at least five amino acid positions to be substituted with histidine are selected from the “hotspot” list. In one embodiment, all histidine substitutions in the manipulated antibody are made at amino acid positions selected from the hotspot list (Table A). That is, only residues in the hotspot list may be selected for histidine substitution. However, in another embodiment, histidine substitutions may be included at one or more positions not on the hotspot list, provided that at least one histidine substitution is made at a position selected from the hotspot list (Table A). For example, as will be discussed below, in one embodiment, one or more histidine substitutions may be included at “coldspot” positions in VH CDR3 and / or VL CDR3.

[0026] As described herein, the parent antibody preferably has certain amino acid positions that are not selected for histidine substitution. These amino acid positions are those shown in the “cold spot” list (Table B). The manipulated antibody prepared according to the method herein must contain at least one amino acid position in the cold spot list (Table B) that is not substituted with histidine. In one embodiment, at least two, at least three, at least four, or at least five amino acid positions in the cold spot list (Table B) are not substituted with histidine. In a further embodiment, none of the amino acid positions in the cold spot list (Table B) are substituted with histidine.

[0027] In the case of VH CDR3 and VL CDR3, it is permissible (and in some embodiments may be beneficial) to introduce histidine residues at one or more cold spot amino acid residue positions listed in Table B, and to test the resulting antibody variant(s) for pH-dependent binding. Given the significant changes in the properties of VH CDR3 and VL CDR3, introducing histidine residues at cold spot positions in these CDRs (in addition to or instead of hot spot positions) is more acceptable in the antibody structure than introducing them at cold spot positions in CDR1 and CDR2, and can make a significant contribution to pH-dependent antigen binding. Therefore, histidine substitution at the listed cold spot positions in VH CDR3 and VL CDR3 is permissible in all embodiments of the present invention.

[0028] In one specific embodiment, all amino acid residues substituted with histidine are selected from the hotspot list (Table A), provided that one or more of the following residues: H100g, H100k, H100m, H100n, L95, L95d, L95e, L95f, and L97 can be further substituted with histidine.

[0029] (pH dependent binding) Modified antibodies into which one or more histidine substitutions have been introduced according to the method described herein may exhibit pH-dependent binding to the target antigen.

[0030] As used herein in relation to antibody-antigen binding interactions, the term "pH-dependent binding" means that the antigen-binding activity of an antibody at an acidic pH differs from that of an antibody at a neutral pH.

[0031] Various measures of antigen-binding activity can be used as indicators of the difference between antigen-binding activity at acidic pH and antigen-binding activity at neutral pH.

[0032] In one embodiment, the "affinity" of an antibody for its antigen can be used as an indicator of antigen-binding activity. In a preferred embodiment, if the affinity of the engineered antibody for its antigen at acidic pH is lower than the affinity of the engineered antibody for its antigen at neutral pH, the engineered antibody exhibits "pH-dependent binding".

[0033] In one embodiment, the dissociation rate constant (k d )(also referred to as the off-rate k off ) of the antibody-antigen interaction can be used as an indicator of antigen-binding activity. In a preferred embodiment, if the dissociation rate constant (k d ) of the engineered antibody-antigen interaction at acidic pH is higher than the dissociation rate constant (k d ) of the engineered antibody-antigen interaction at neutral pH, the engineered antibody exhibits "pH-dependent binding".

[0034] In embodiments where the dissociation rate constant (k d ) is used as an indicator of pH-dependent binding, the ratio of the dissociation rate constant (k d ) of the engineered antibody-antigen interaction at acidic pH to the dissociation rate constant (k d ) of the engineered antibody-antigen interaction at neutral pH can be at least 1.5, or at least 2, or at least 5, or at least 10.

[0035] In other embodiments, the dissociation rate constant (k d ) of the antibody engineered at acidic pH (e.g., pH 5.5) is higher than the dissociation rate constant (k d ) of the parental antibody at the same acidic pH. In such embodiments, k d at acidic pH (e.g., pH 5.5) is high as a result of histidine substitution compared to the parental antibody. In certain embodiments, the dissociation rate constant (k d ) of the engineered antibody-antigen interaction at acidic pH to the dissociation rate constant (k dThe ratio to ) can be at least 1.5, or at least 2, or at least 5, or at least 10. Considering the antibody recycling pathway in vivo, the dissociation rate (k) at acidic pH (e.g., pH 5.5) d ) is a particularly valid measure of pH-dependent binding. Faster dissociation of the antigen-antibody complex at acidic pH (pH 5.5–6.0) allows for the release of the antigen in acidified endosomes.

[0036] In one embodiment, the equilibrium dissociation constant (K D ) can be used as an indicator of antigen-binding activity. In a preferred embodiment, the equilibrium dissociation constant (K) of the manipulated antibody-antigen interaction at an acidic pH is used. D ) is the equilibrium dissociation constant (K) of the manipulated antibody-antigen interaction at neutral pH. D If the value is higher than ), the manipulated antibody exhibits "pH-dependent binding."

[0037] Equilibrium dissociation constant (K D In embodiments where ) is used as an indicator of pH-dependent binding, the equilibrium dissociation constant (K) of the manipulated antibody-antigen interaction at acidic pH is used. D The equilibrium dissociation constant (K) of the manipulated antibody-antigen interaction at neutral pH. D The ratio to ) can be greater than 1.5, greater than 2, greater than 5, or greater than 10. In a particularly preferred embodiment, the manipulated antibody may exhibit antigen binding 20 to 40 times stronger at neutral pH than at acidic pH.

[0038] Equilibrium dissociation constant (K D In a non-limiting embodiment using ) as an indicator of antigen-binding activity, the engineered antibody has an equilibrium dissociation constant (K) for its antigen in the range of 10–20 nM at an acidic pH (e.g., pH 5.5). D ) may show; in contrast, the K of the manipulated antibody against its antigen at a neutral pH (e.g., pH 7.4) D This can be approximately 0.5 nM or less.

[0039] Other measures of antigen-binding activity that can be used to indicate pH-dependent binding include affinity constants or association rate constants (k a ) or on speed (k on ) is possible.

[0040] For antagonist antibodies, the measure of antigen-binding activity can be the blocking ability, i.e., the IC50 in a suitable assay.

[0041] In a preferred embodiment, a modified antibody containing one or more histidine substitutions exhibits a higher pH dependence of its binding to the antigen compared to the parent antibody. "Higher pH dependence of binding" means that the difference between antigen-binding activity at acidic pH and antigen-binding activity at neutral pH is significantly greater for the modified antibody than for the parent antibody from which it is derived.

[0042] In some embodiments, the parent antibody may not exhibit significant pH dependence in binding to its target antigen, and this pH dependence may be introduced in variants modified as a result of one or more histidine substitutions. In other embodiments, the parent antibody may exhibit measurable pH dependence, which is significantly higher in variants modified as a result of one or more histidine substitutions.

[0043] In some embodiments, it may be possible to introduce pH dependence to a manipulated antibody without significantly impairing its antigen-antibody binding affinity at neutral pH. In such embodiments, the manipulated antibody may exhibit significantly lower antigen-binding activity at acidic pH compared to its parent antibody, while the antigen-binding activity of the manipulated antibody at neutral pH may be equivalent to (i.e., not significantly different from) that of its parent antibody at neutral pH.

[0044] One advantage of the method described herein for selecting candidate amino acids for histidine substitution according to a "heatmap" of the natural occurrence rate of histidine in the functional antibody repertoire is that histidine residues are introduced only at "hotspot" amino acid positions where they typically occur naturally in the antibody repertoire, and are therefore acceptable in the antibody structure. It has been shown that histidine substitution at hotspot positions can confer pH dependence of antigen binding without significantly affecting the strength of antigen binding activity at neutral pH.

[0045] In one embodiment, the engineered antibody exhibits a higher pH-dependent binding to its antigen compared to the parent antibody, while retaining at least 80% of the parent antibody's antigen-binding activity at neutral pH. In a further embodiment, the engineered antibody may retain at least 80%, at least 85%, at least 90%, at least 95%, or 100% of the parent antibody's antigen-binding activity at neutral pH, while still exhibiting pH-dependent binding.

[0046] In a preferred embodiment, the manipulated antibody exhibits a higher pH-dependent binding to its antigen compared to the parent antibody, while retaining at least 80%, at least 85%, at least 90%, at least 95%, or 100% of the parent antibody's affinity for the antigen at neutral pH.

[0047] Equilibrium dissociation constant (K D In this embodiment, where ) is used as an indicator of antigen-binding activity, the manipulated antibody is the K of the parent antibody against the target antigen at neutral pH. D The equilibrium dissociation constant (K) for the target antigen at a neutral pH that is 20% or less, 15% or less, 10% or less, or 5% or less greater than the value of the target antigen. D ) may exhibit. In other embodiments, the manipulated antibody exhibits the K of the parent antibody against the target antigen at neutral pH. D The equilibrium dissociation constant (K) for the target antigen at a neutral pH that is substantially equal to the equilibrium dissociation constant (K) D ) may indicate.

[0048] In other embodiments, introducing pH dependence may reduce the affinity of the engineered antibody to the antigen at neutral pH. However, provided that the desired pH dependence of binding is achieved, a decrease in affinity at neutral pH may be acceptable depending on the properties of the target antigen, the intended use of the antibody, and the characteristics of the parent antibody. For example, if the parent antibody exhibits particularly high affinity for the antigen at neutral pH, a decrease in the affinity of the engineered antibody to the antigen at neutral pH may be acceptable if this is accompanied by a significant increase in pH dependence (i.e., provided that the desired "multiple difference" between binding at neutral pH and binding at acidic pH is achieved). In some embodiments, provided that the desired pH dependence of binding is also achieved, the affinity of the engineered antibody to the target antigen at neutral pH may be up to 10 times, or even 20 times, lower than the affinity of the parent antibody to the antigen at neutral pH.

[0049] In each of the above embodiments, "acidic pH" refers to a pH in the range of pH 4.0 to pH 6.5, preferably pH 4.5 to pH 6.0, more preferably pH 5.5 to pH 6.0, and more preferably pH 5.5 to pH 5.8.

[0050] In each of the above embodiments, "neutral pH" refers to a pH in the range of pH 6.7 to pH 10.0, preferably pH 7.0 to pH 8.0, and more preferably pH 7.4.

[0051] In certain embodiments, pH-dependent binding can be evaluated by comparing the antigen-binding activity of an antibody at an acidic pH of 5.5 with that of the same antibody at a neutral pH of 7.4.

[0052] In another embodiment, pH-dependent binding can be evaluated by comparing the antigen-binding activity of the antibody at an acidic pH of 6.0 with the antigen-binding activity of the same antibody at a neutral pH of 7.4.

[0053] Appropriate methods and conditions for determining the antigen-binding activity of an antibody at a specified pH are generally known in the art. For example, the antigen-binding activity of an antibody at a specified pH can be measured by ELISA using Biacore® (GE Healthcare) or the MSD platform, as described in the attached experimental examples.

[0054] (Heatmap of histidine appearance rate) The amino acid residues of the parent antibodies described herein may be selected for histidine substitution or not for histidine substitution based on a "heatmap" of the occurrence rate of histidine residues at specific amino acid positions in the CDR of functional antibodies in the natural antibody repertoire. This "heatmap" provides both a list of "hotspot" amino acid residue positions in the CDR and framework positions near the CDR in which histidine may naturally occur in the antibody repertoire, and a list of "coldspot" amino acid residue positions in the CDR and framework positions near the CDR in which histidine typically does not naturally occur.

[0055] The term "antibody repertoire" refers to a population of antibodies that exhibit antigen-binding activity. The "natural antibody repertoire" is typically a population of antibodies produced by immunization of a host using a target antigen. These repertoires may include antibodies against two or more target antigens. In this disclosure, the histidine appearance rate "heatmap" itself originates from a diverse repertoire of llama antibodies induced by immunization of host animals using multiple different target antigens. Furthermore, antibodies isolated from these immunizations were screened by specific phage display selection, which enabled the enrichment of antibodies exhibiting pH-dependent antigen binding.

[0056] (Selection of residues to substitute) For any given parent antibody, the selection of candidate amino acid residues for histidine substitution is guided by a heatmap of histidine occurrence rates.

[0057] The terms “substitution” and “replacement” of amino acid residues are used interchangeably herein. In the context of an engineered antibody, which is an engineered variant of a parent antibody, “substitution” requires that an amino acid residue at a specified amino acid position in the parent antibody is replaced with a different amino acid residue, i.e., a histidine residue that does not naturally occur at that position in the parent antibody.

[0058] The histidine occurrence “heatmap” described herein provides a list of hotspot amino acid positions where histidine is observed to occur naturally. The target parent antibody from which the engineered antibody is derived may naturally contain histidine residues at one or more amino acid positions on the hotspot list. In these situations, the naturally occurring histidine residues will generally be retained in the engineered antibody because they can contribute to pH-dependent antigen binding. However, in the methods described herein, this naturally occurring histidine is not counted as a histidine substitution. Therefore, the methods described herein require that the engineered antibody contains at least one additional histidine residue at one or more hotspot amino acid positions compared to the parent antibody from which it is derived. The introduction of one or more additional histidine residues into the engineered antibody can improve the pH-dependent binding of the engineered antibody (i.e., result in higher pH dependence of binding to the target antigen), even for parent antibodies that already contain hotspot histidine residues.

[0059] Candidate amino acid residues to be substituted with histidine are selected from a "hotspot" list (Table A) identified according to the Kabat numbering system (Kabat et al., "Sequences of Proteins of Immunological Interest," 5th edition, Public Health Service, National Institutes of Health, Bethesda, MD. (1991)) for amino acid positions in the CDRs of the VH and VL domains and framework positions near the CDRs. The manipulated antibody must contain at least one histidine substitution at an amino acid position selected from Table A.

[0060] Various approaches can be used to select the most appropriate histidine substitution to confer pH-dependent antigen binding to the target parent antibody. If structural information about the binding interaction of the parent antibody with its target antigen is available, it may be appropriate to utilize a “rational design” approach to select candidate amino acids from a hotspot list for histidine substitution. Hotspot residues that are likely to be directly involved in the antigen-antibody binding interaction or that affect the overall structure of the antigen-binding contact surface of the parent antibody can be selected based on such structural information. Subsequently, engineered variants of the parent antibody in which one or more selected hotspot residues or different combinations thereof are substituted with histidine can be synthesized and screened for pH-dependent antigen binding, for example, using the screening method described herein.

[0061] Furthermore, candidate residues for histidine substitution can be selected based on the family or subtype of the variable domain of the parent antibody. This specification also provides lists of various subtype-specific hotspot residues derived from heatmaps of histidine occurrence rates in the antibody repertoire. In one embodiment, if the family or subtype of the variable domain of the parent antibody's VH and / or VL domains is known, candidate residues for histidine substitution can be selected from the following subtype-specific or family-specific hotspot lists. These lists are not intended to be limiting, but merely to provide examples of amino acid positions that may be most appropriate for histidine substitution for a particular family or subtype of variable domain. For any given variable domain, all possible amino acid positions for histidine substitution, regardless of the family or subtype of the variable domain, can be selected from the complete hotspot list (Table A).

[0062] Regarding the heavy chain variable domains of the VH3 family: Table VH3A (Hotspot residues of the VH3 family) [Table 11] Table VH3B (Cold spot residues of the VH3 family) [Table 12]

[0063] Regarding the heavy chain variable domains of the VH1 family: Table VH1A (Hotspot residues of the VH1 family) [Table 13] Table VH1B (Cold spot residues of the VH1 family) [Table 14]

[0064] Regarding the heavy chain variable domains of the VH4 family: Table VH4A (Hotspot residues of the VH4 family) [Table 15] Table VH4B (Cold spot residues of the VH4 family) [Table 16]

[0065] Regarding λ-type light chain variable domains: Table VλA (Vλ hotspot residues) [Table 17] Table VλB (Cold spot residues of Vλ) [Table 18]

[0066] Regarding the light chain variable domains of the Vλ1 family: Table Vλ1A (Hotspot residues of the Vλ1 family) [Table 19] Table Vλ1B (Cold spot residues of the Vλ1 family) [Table 20]

[0067] Regarding the light chain variable domains of the Vλ2 family: Table Vλ2A (Hotspot residues of the Vλ2 family) [Table 21] Table Vλ2B (Cold spot residues of the Vλ2 family) [Table 22]

[0068] Regarding the light chain variable domains of the Vλ3 family: Table Vλ3A (Hotspot residues of the Vλ3 family) [Table 23] Table Vλ3B (Cold spot residues of the Vλ3 family) [Table 24]

[0069] Regarding the light chain variable domains of the Vλ5 family: Table Vλ5A (Hotspot residues of the Vλ5 family) [Table 25] Table Vλ5B (Cold spot residues of the Vλ5 family) [Table 26]

[0070] Regarding the light chain variable domains of the Vλ8 family: Table Vλ8A (Hotspot residues of the Vλ8 family) [Table 27] Table Vλ8B (Cold spot residues of the Vλ8 family) [Table 28]

[0071] Regarding κ-type light chain variable domains: Table VKA (VK hotspot residues) [Table 29] Table VKB (cold spot residues of VK) [Table 30]

[0072] Regarding the light chain variable domains of the VK1 family: Table VK1A (Hotspot residues of the VK1 family) [Table 31] Table VK1B (Cold spot residues of the VK1 family) [Table 32]

[0073] Regarding the VK2 family of light chain variable domains: Table VK2A (Hotspot residues of the VK2 family) [Table 33] Table VK2B (Cold spot residues of the VK2 family) [Table 34]

[0074] Regarding the light chain variable domains of the VG4 family: Table VK4A (Hotspot residues of the VK4 family) [Table 35] Table VK4B (Cold spot residues of the VK4 family) [Table 36]

[0075] (Screening of variant panels) A particular embodiment of the method of the present invention may be based on constructing a panel of manipulated variants of the parent antibody, followed by screening the panel to identify one or more manipulated variants exhibiting a desired pH-dependent binding. Constructing a panel of variants allows for systematic mutational analysis of numerous possible histidine substitutions at amino acid positions defined by a heatmap. This method differs from the combinatorial method in that it guides the construction of the variant panel using a histidine occurrence heatmap, thus eliminating the need, or making it inappropriate, to sample all possible amino acid positions and their combinations in the CDR as candidates for histidine substitutions.

[0076] The systematic mutation analysis methods described herein are generally: (a) A step of providing a parent antibody that binds to an antigen; (b) The step of preparing a panel of manipulated variants of the parent antibody, wherein each of the manipulated variants in the panel differs from the parent antibody by having at least one amino acid residue substituted with a histidine residue, wherein at least one amino acid residue selected from the hotspot list (Table A) is substituted with histidine, and at least one amino acid residue selected from the coldspot list (Table B) is not substituted with histidine; (c) The process includes screening a panel of the manipulated variants for pH-dependent binding to the antigen, thereby identifying the manipulated antibody that exhibits pH-dependent binding to the antigen.

[0077] This method requires the preparation of a panel of manipulated variants of the target parent antibody, each of which differs from the parent antibody by having at least one amino acid substituted with histidine. The selection of the amino acid site to be substituted with histidine is based on a heatmap of histidine occurrence rates. Each of the manipulated variants in the panel must include at least one histidine substitution at an amino acid site selected from the hotspot list (Table A), while at least one amino acid residue selected from the coldspot list (Table B) is not substituted with histidine. Preferably, each of the manipulated variants in the panel includes different patterns of histidine substitution. This allows for parallel analysis of the effects of different histidine substitutions on pH-dependent antigen binding.

[0078] In one embodiment of the method, each manipulated variant in the panel may contain a single histidine substitution at one of the listed hotspot amino acid positions (Table A). Preferably, each manipulated variant in the panel contains a different histidine substitution at one of the listed hotspot amino acid positions (Table A). Such a panel makes it possible to analyze the effect of histidine substitutions at individual hotspot amino acid positions on pH-dependent binding. The panel of manipulated variants may represent the entire hotspot list; that is, it may include variants having a single histidine substitution at each of the amino acid positions listed in Table A, or it may represent a selected subset of the hotspot list, such as a subset of substitutions associated with one or more specific variable domain families or subtypes. Furthermore, the panel may also include variants containing histidine substitutions at one or more "cold spot" positions in VH CDR3 or VL CDR3.

[0079] In other embodiments, the panel prepared in step (b) may include manipulated variants containing various combinations of two or more, three or more, four or more, or five or more histidine substitutions, provided that at least one of these substitutions is at a hotspot amino acid position selected from Table A, while at least one amino acid from Table B is not histidine-substituted. Preferably, all of the histidine substitutions are made at hotspot amino acid positions selected from Table A. This panel makes it possible to screen combinations of histidine residues (and in particular combinations of hotspot histidine substitutions) that confer pH-dependent binding. Furthermore, the panel may also include variants containing histidine substitutions at one or more "cold spot" positions in the VH CDR3 or VL CDR3, optionally combined with histidine substitutions at one or more hotspot amino acid positions.

[0080] In one embodiment, an initial screening can be performed on a panel of manipulated variants containing single histidine substitutions as described above, in order to identify a single amino acid position that can confer pH-dependent binding by substitution with histidine. Based on the results of this initial screening, one or more further manipulated variants can be constructed by combining histidine substitutions that have already been identified as conferring pH-dependent binding. Such methods include: (a) A step to identify the parent antibody that binds to the antigen; (b) A step of preparing a panel of manipulated variants of the parent antibody, wherein each of the manipulated variants in the panel differs from the parent antibody by having at least one amino acid residue substituted with a histidine residue, having at least one amino acid residue selected from the hotspot list (Table A) substituted with histidine, and having at least one amino acid residue selected from the coldspot list (Table B) not substituted with histidine (where preferably each of the manipulated variants includes a different pattern of histidine substitution); (c) A step of screening a panel of the manipulated variants for pH-dependent binding to the target antigen, thereby identifying selected amino acid positions in which pH-dependent binding to the antigen is conferred by the presence of histidine; (d) the step of preparing one or more further modified variants of the parent antibody, wherein each of the variants contains histidine at two or more selected amino acid positions identified in step (c); and (e) The process may include screening the further modified variant for pH-dependent binding to the antigen; thereby identifying the modified antibody that exhibits pH-dependent binding to the antigen.

[0081] In certain embodiments of the method, histidine substitutions in the VH domain and VL domain can be analyzed individually for their effects on pH-dependent binding. Accordingly, in one embodiment, step (b) is a step of preparing a panel of manipulated variants of the parent antibody, wherein each of the manipulated variants in the panel differs from the parent antibody in that at least one amino acid residue in the VH domain is substituted with a histidine residue, at least one amino acid residue selected from the heavy chain hotspot list (Table HA) is substituted with histidine, and at least one amino acid residue selected from the heavy chain coldspot list (Table HB) is not substituted with histidine. Furthermore, the panel may also include variants with histidine substitutions at the "coldspot" positions in the VH CDR3.

[0082] Typically, the manipulated variant contains a mutant VH domain with one or more histidine substitutions paired with the VL domain of the parent antibody. Using such a library, it is possible to screen for histidine substitutions that confer pH-dependent binding in the VH domain.

[0083] Table HA (VH hotspot residue location) [Table 37] Table HB (VH cold spot residue location) [Table 38]

[0084] In a further embodiment, step (b) is a step of preparing a panel of manipulated variants of the parent antibody, wherein each of the manipulated variants in the panel differs from the parent antibody in that at least one amino acid residue in the VL domain is substituted with a histidine residue, at least one amino acid residue selected from the light chain hotspot list (Table LA) is substituted with histidine, and at least one amino acid residue selected from the light chain hotspot list (Table LB) is not substituted with histidine. Furthermore, the panel may also include variants having a histidine substitution at a “cold spot” position in the VL CDR3.

[0085] Typically, the manipulated variant contains a mutant VL domain with one or more histidine substitutions paired with the VH domain of the parent antibody. Using such a library, it is possible to screen for histidine substitutions that confer pH-dependent binding in the VL domain.

[0086] Table LA (VL hotspot residue location) [Table 39] Table LB (VL cold spot residue location) [Table 40]

[0087] For λ-class light chain variants, the "hot spot" and "cold spot" residue positions can be selected from the following list: Table VλA (Vλ hotspot residues) [Table 41] Table VλB (Vλ cold spot residues) [Table 42]

[0088] For κ-class light chain variants, the "hot spot" and "cold spot" residue positions can be selected from the following list: Table VKA (VKA hotspot residue locations) [Table 43] Table VKAB (VKA cold spot residue location) [Table 44]

[0089] In one embodiment, histidine substitutions of VH and VL domains identified by individual initial VH and VL domain screenings can be combined in a single manipulated variant. Subsequently, further screening can be performed to identify VH and VL combinations that exhibit pH-dependent binding. Such a method is: (a) A step to identify the parent antibody that binds to the antigen; (b) The step of preparing a first panel of manipulated variants of the parent antibody, wherein each of the manipulated variants in the panel differs from the parent antibody in that at least one amino acid residue in the VH domain is substituted with a histidine residue, at least one amino acid residue selected from the heavy chain hotspot list (Table HA) is substituted with histidine, and at least one amino acid residue selected from the heavy chain coldspot list (Table HB) is not substituted with histidine; (c) Screening a first panel of the manipulated variant for pH-dependent binding to the antigen; thereby identifying one or more selected amino acid positions in the VH domain to which pH-dependent binding is conferred by the presence of histidine; (d) A step of preparing a second panel of manipulated variants of the parent antibody, wherein each of the manipulated variants in the panel differs from the parent antibody in that at least one amino acid residue in the VL domain is substituted with a histidine residue, at least one amino acid residue selected from the light chain hotspot list (Table LA) is substituted with histidine, and at least one amino acid residue selected from the light chain hotspot list (Table LB) is not substituted with histidine; (e) Screening a second panel of the manipulated variant for pH-dependent binding to the antigen; thereby identifying one or more selected amino acid positions in the VL domain where pH-dependent binding is conferred by the presence of histidine; (f) A step of preparing one or more further modified variants of the parent antibody, wherein each of the variants differs from the parent antibody by having one or more selected amino acids at the VH domain identified in step (c) substituted with histidine, and by having one or more selected amino acids at the VL domain identified in step (e) substituted with histidine; and (g) The process may include screening the further modified variant for pH-dependent binding to the antigen; thereby identifying the modified antibody that exhibits pH-dependent binding to the antigen.

[0090] This approach makes it possible to identify synergistic effects between histidine mutations identified in the initial (individual) screening of VH and VL domains. For example, to identify synergistic effects between different combinations of VH and VL domain mutations, the highest-performing VH mutation (i.e., the mutation that confers the greatest pH dependence) can be combined with many different VL mutations, or the highest-performing VL mutation can be combined with a different VH mutation. In this embodiment, the "cold spot" position in the VH CDR3 or VL CDR3 and their combinations may also include histidine substitutions.

[0091] (Screening method) Various screening methods can be used to identify engineered antibodies that exhibit pH-dependent binding to target antigens. Such screening methods include: (a) A step to determine the antigen-binding activity of an antibody manipulated at an acidic pH; (b) A step to determine the antigen-binding activity of the manipulated antibody at a neutral pH; (c) The process includes selecting an engineered antibody whose antigen-binding activity at an acidic pH differs from its antigen-binding activity at a neutral pH.

[0092] Appropriate methods and conditions for determining antigen-binding activity at a specified pH are generally known in the art. For example, the antigen-binding activity of an antibody at a specified pH can be determined by Biacore® (GE Healthcare) as described in the attached experimental examples. Other preferred techniques include ELISA or MSD, which can be used to determine antigen binding under different pH conditions (e.g., pH 7.4 and pH 5.5) and at specific antibody concentrations (e.g., EC50).

[0093] Methods described herein may include the steps of measuring the antigen-binding activity of an engineered antibody at neutral and acidic pH, and comparing one or more binding parameters at neutral and / or acidic pH to determine whether the engineered antibody exhibits pH-dependent binding.

[0094] In one embodiment, the antigen-binding activity of the manipulated antibody can be measured by dual pH ELISA as described in the attached examples. In this setup, the initial binding of the antibody to the antigen is evaluated at a neutral pH (e.g., pH 7.4), followed by a washing step at an acidic pH (e.g., pH 5.5). This makes it possible to identify the effect of pH on antibody association, dissociation, or both, and can better represent the in vivo situation in which the binding of the antibody to the target antigen occurs at a neutral pH and the release of the antigen occurs at an acidic endosomal pH.

[0095] Typically, selected manipulated antibodies exhibit lower affinity for their antigen at acidic pH than at neutral pH. Even a slight significant difference in affinity at acidic pH compared to neutral pH during initial screening can indicate pH-dependent binding.

[0096] The dissociation rate constant (k) of the manipulated antibody-antigen interaction at an acidic pH. d ), that is, the off rate is the dissociation rate constant (k) of the manipulated antibody-antigen interaction at neutral pH. d) may be higher than ). In initial screening, (k at acidic pH d (k) at a neutral pH d If even a slight significant difference is observed in ), pH-dependent binding can be observed. In another embodiment, the dissociation rate constant (k) of the antibody-antigen interaction manipulated at acidic pH is observed. d The dissociation rate constant (k) of the manipulated antibody-antigen interaction at neutral pH. d The ratio to ) can be at least 1.5, or at least 2, or at least 5, or at least 10. Combining multiple histidine mutations may increase the overall degree of pH dependence.

[0097] The equilibrium dissociation constant (K) of the manipulated antibody-antigen interaction at an acidic pH. D ) is the equilibrium dissociation constant (K) of the manipulated antibody-antigen interaction at neutral pH. D ) may be higher than ). In initial screening, K at acidic pH D K at neutral pH D Even a slight significant difference in this regard indicates pH-dependent binding. The equilibrium dissociation constant (K) of the manipulated antibody-antigen interaction at acidic pH. D The equilibrium dissociation constant (K) of the manipulated antibody-antigen interaction at neutral pH. D The ratio to ) can be greater than 1.5, greater than 2, greater than 5, or greater than 10.

[0098] Combining multiple histidine mutations can increase the overall degree of pH dependence. In a particularly preferred embodiment, combining histidine mutations can achieve antigen binding that is approximately 20 to 40 times stronger at neutral pH than at acidic pH.

[0099] As described elsewhere in this specification, pH-dependent binding improvement can be achieved with or without a significant effect on affinity for the target antigen at neutral pH.

[0100] Equilibrium dissociation constant (K DIn a non-limiting embodiment using ) as an indicator of antigen-binding activity, the engineered antibody has an equilibrium dissociation constant (K) for its antigen in the range of 10–20 nM at an acidic pH (e.g., pH 5.5). D ) may show; in contrast, the K of the manipulated antibody against its antigen at a neutral pH (e.g., 7.4) D This can be approximately 0.5 nM or less.

[0101] In each of the above embodiments, "acidic pH" refers to a pH in the range of pH 4.0 to pH 6.5, preferably pH 4.5 to pH 6.0, more preferably pH 5.5 to pH 6.0, and more preferably pH 5.5 to pH 5.8.

[0102] In each of the above embodiments, "neutral pH" refers to a pH in the range of pH 6.7 to pH 10.0, preferably pH 7.0 to pH 8.0, and more preferably pH 7.4.

[0103] In certain embodiments, pH-dependent binding can be evaluated by comparing the antigen-binding activity of an antibody at an acidic pH of 5.5 with that of the same antibody at a neutral pH of 7.4.

[0104] In another embodiment, pH-dependent binding can be evaluated by comparing the antigen-binding activity of the antibody at an acidic pH of 6.0 with the antigen-binding activity of the same antibody at a neutral pH of 7.4.

[0105] The methods described herein may further or alternatively include the steps of measuring the antigen-binding activity of a parent antibody at acidic pH and / or neutral pH, and comparing the parameters of this binding interaction with the corresponding binding parameters of an engineered antibody-antigen interaction.

[0106] The dissociation rate constant (k) of an antibody manipulated at an acidic pH (e.g., pH 5.5). d ) is the dissociation rate constant (k) of the parent antibody at the same acidic pH. d) can be higher than . In such an embodiment, k at an acidic pH (e.g., pH 5.5) d This is higher compared to the parent antibody as a result of histidine substitution. Dissociation rate constant (k) of manipulated antibody-antigen interactions at acidic pH d The dissociation rate constant (k) of the parent-antibody-antigen interaction at the same acidic pH. d The ratio to ) can be at least 1.5, or at least 2, or at least 5, or at least 10.

[0107] (Manipulated antibodies) Furthermore, this specification provides an engineered antibody that exhibits pH-dependent binding to the antigen, characterized in that at least one amino acid in the CDR of the engineered antibody is a histidine residue, at least one amino acid residue selected from the hotspot list (Table A) is a histidine residue, and at least one amino acid residue from the coldspot list (Table B) is not a histidine residue.

[0108] "Manipulated antibody" means an antibody whose amino acid sequence has been intentionally altered or mutated in vitro. In this disclosure, "manipulated antibody" means an antibody variant in which the amino acid sequence of the antibody's CDR has been intentionally altered or mutated to introduce one or more additional histidine residues.

[0109] The CDRs of the antibody VL and VH domains can typically be defined as containing the following amino acids: residues 24–34 (CDRL1), 50–56 (CDRL2), and 89–97 (CDRL3) in the light chain variable domain, and residues 31–35 or 31–35c (CDRH1), 50–65 (CDRH2), and 95–102 (CDRH3) in the heavy chain variable domain (Kabat et al., "Sequences of Proteins of Immunological Interest," 5th edition, Public Health Service, National Institutes of Health, Bethesda, MD. (1991)).

[0110] The manipulated antibody may contain a total of 1, 2, 3, 4, 5, or more histidine residues in the VH domain, the VL domain, or both the VH and VL domains of the CDR.

[0111] At least one histidine residue in the CDR of the manipulated antibody must appear at an amino acid position selected from the “hotspot” list (Table A). In some embodiments, at least two, at least three, at least four, or at least five histidine residues must appear at amino acid positions selected from the hotspot list. In some embodiments, all histidine residues in the manipulated antibody appear at amino acid positions selected from the hotspot list (Table A). As described herein, there are certain amino acid positions in the CDR of the manipulated antibody that are preferably not occupied by histidine residues. These amino acid positions are the amino acid positions shown in the “coldspot” list (Table B). The manipulated antibody must contain at least one amino acid position that is not occupied by a histidine residue in the coldspot list (Table B). In some embodiments, at least two, at least three, at least four, or at least five amino acid positions in the coldspot list (Table B) are not histidine. In further embodiments, none of the amino acid positions in the coldspot list (Table B) are histidine. As an exception, in some embodiments, histidine substitution may also be included at the "cold spot" position in VH CDR3 or VL CDR3 and combinations thereof.

[0112] Manipulated antibodies exhibit pH-dependent binding to their antigens. This means that the antigen-binding activity of an antibody at an acidic pH differs from that of an antibody at a neutral pH.

[0113] In one embodiment, the manipulated antibody has a lower affinity for its antigen at acidic pH than at neutral pH.

[0114] In one embodiment, the dissociation rate constant (k) of the manipulated antibody-antigen interaction at an acidic pH is described. d ) is the dissociation rate constant (k) of the manipulated antibody-antigen interaction at neutral pH. d It is higher than ).

[0115] In one embodiment, the equilibrium dissociation constant (K) of the manipulated antibody-antigen interaction at an acidic pH is D ) is the equilibrium dissociation constant (K) of the manipulated antibody-antigen interaction at neutral pH. D It is higher than ).

[0116] In a particularly preferred embodiment, the manipulated antibody may exhibit antigen binding 20 to 40 times stronger at neutral pH than at acidic pH.

[0117] Equilibrium dissociation constant (K D In a non-limiting embodiment using ) as an indicator of antigen-binding activity, the engineered antibody has an equilibrium dissociation constant (K) for its antigen in the range of 10–20 nM at an acidic pH (e.g., pH 5.5). D ) may show; in contrast, the K of the manipulated antibody against its antigen at a neutral pH (e.g., 7.4) D This can be approximately 0.5 nM or less.

[0118] (Structure of the manipulated antibody) The engineered antibodies described herein are typically conventional quadrivalent immunoglobulins, where antigen-binding specificity is provided by paired VH and VL domains, with six CDRs contributing to the antigen-binding site. However, the term “engineered antibody” also encompasses, but is not limited to, antigen-binding fragments and engineered antigen-binding constructs of conventional immunoglobulins, including Fab, F(ab'), F(ab')2, Fv, scFv, diabody, triabody, minibody, etc., and any other modified immunoglobulin constructs, including antigen-binding sites provided by paired VH and VL domains.

[0119] The manipulated antibody may contain a constant region comprising one or more or all of the CH1 domain, hinge region, CH2 domain, and CH3 domain. In particular, the manipulated antibody may contain an Fc region (composed of CH2 and CH3 domains) that can confer one or more antibody effector functions. The Fc region of the manipulated antibody may be manipulated or modified to confer useful functional properties of its own.

[0120] The manipulated antibody may be any antibody class, including IgA, IgD, IgE, IgG, or IgM. In a preferred embodiment, the manipulated antibody is IgG. The manipulated antibody may be any subclass (isotype) of IgG, including IgG1, IgG2, IgG3, or IgG4.

[0121] In a preferred embodiment, the engineered antibody may contain an Fc region capable of binding to the neonatal Fc receptor (FcRn). In particular, it may be desirable for the Fc region to exhibit a strong binding affinity to FcRn at neutral pH. The Fc region of the engineered antibody can be modified to enhance binding to the neonatal Fc receptor, FcRn. An example of Fc modification that confers enhanced binding to FcRn is amino acid substitution at one or more amino acid positions in the CH2 and / or CH3 domains of the Fc region.

[0122] A suitable Fc mutation is the Abdeg® mutation described by Vaccaro et al. in Nature Biotechnology, Vol. 23, 1283-1288 (2005). Abdeg® mutants are engineered variants of human IgG, including, but are not limited to, human IgG1, which includes mutations that change Met252, Ser254, Thr256, His 433, and Asn434 to Tyr252, Thr254, Glu256, Lys433, and Phe434 (EU numbering). Engineered antibodies with Abdeg® mutations (e.g., engineered human IgG1) exhibit enhanced FcRn binding affinity compared to their wild-type counterparts and bind more stably to FcRn during exocytosis events on the cell surface. The preferred manipulated antibodies as defined herein combine one or more histidine substitutions in the CDR of the variable domain with an Abdeg mutation in the Fc region.

[0123] Further preferred Fc mutations include the NHance® mutation described in US 8,163,881. NHance® mutants are engineered variants of human IgG, including, but are not limited to, human IgG1, which includes mutations that replace His 433 and Asn434 with Lys433 and Phe434 (EU numbering). Engineered antibodies with NHance® mutations (e.g., engineered human IgG1) exhibit enhanced binding affinity to FcRn compared to their wild-type counterparts. Preferred engineered antibodies as defined herein combine one or more histidine substitutions in the CDR of the variable domain with the NHance® mutation in the Fc region.

[0124] In a preferred embodiment, the manipulated antibody may include variable domains (VH and VL) which are manipulated variants of the variable domain of a camelid antibody (e.g., llama antibody) containing one or more histidine substitutions in the CDR, as well as the Fc region of human IgG (e.g., human IgG1) containing Abdeg® mutations Tyr252, Thr254, Glu256, Lys433, and Phe434 (EU numbering).

[0125] In a more preferred embodiment, the manipulated antibody may include variable domains (VH and VL) which are manipulated variants of the variable domain of a camelid antibody (e.g., llama antibody) containing one or more histidine substitutions in the CDR, as well as the Fc region of human IgG (e.g., human IgG1) containing NHance® mutations Lys433 and Phe434 (EU numbering).

[0126] Engineered antibodies exhibiting two characteristics—pH-dependent antigen binding and enhanced binding affinity to FcRn—can exert active antigen scavenging effects, making them useful for removing soluble antigens from plasma in vivo. The concept of "active antigen scavenging" involves using engineered antibodies to mimic the activity of cell surface endocytosis receptors, thereby enabling the selective removal of target antigens from plasma. Briefly, engineered antibodies bind to the corresponding antigen present in plasma (neutral pH) to form an antigen-antibody complex. This complex is randomly absorbed into endosomes (usually by iontophoresis), where, when the pH becomes acidic, the antibody Fc portion binds to FcRn. Due to the pH-dependent antigen-binding properties of the engineered antibody, the bound antigen is released into the acidic endosome and is degraded via the lysosomal pathway. The released antibody (still binding to FcRn but no longer binding to its antigen) is then recycled back to the cell surface, where it can participate in further rounds of antigen binding and internal migration. This "active antigen removal" process can be enhanced by Fc mutations that promote binding to FcRn at acidic pH without substantially affecting binding at pH 7.4.

[0127] In addition to the specific Fc mutations listed above, the manipulated antibodies may also include any other mutations or modifications that improve antibody uptake into endosomes, such as changes in pI, changes in charge, or changes in FcRn affinity at pH 7.4 and / or pH 6.0; any mutations or modifications that promote increased binding to any other "high recycling" Fc receptors, such as Fcγ receptors, particularly FcγRIIb; and any mutations that increase immune complex formation (e.g., complement binding sites).

[0128] The engineered antibodies may be engineered variants or chimeric antibodies of parental antibodies from humans, mice, rats, rabbits, camelids, or other mammalian species. In certain embodiments, the engineered antibodies may be engineered variants of camelid antibodies isolated from camelid species such as llamas, camels, dromedaries, guanacos, and vicuñas, or engineered variants of camelid-human chimeric antibodies (e.g., llama-human chimeric) in which the variable domain (or its CDR) is derived from a camelid species (e.g., llama) and the constant domain is derived from a human. Techniques for producing camelid antibodies and camelid-human chimeric antibodies are described in WO 2010 / 001251, the content of which is incorporated herein by reference.

[0129] The antigen-binding specificity of the manipulated antibody, i.e., the properties of the target antigen to which it binds, is not particularly limited. The applicant has demonstrated that the methodology provided herein is applicable to the manipulation of pH-dependent variants of numerous different parent antibodies with varying antigen-binding properties. Therefore, the described techniques for manipulating pH-dependent antigen binding are principles of general applicability and are not limited to antibodies that bind to a single specific target antigen.

[0130] (Example) The present invention is further illustrated by the following embodiments, but should not be construed as limiting. All references, patents, and published patent applications cited throughout this application are incorporated herein by express reference. [Brief explanation of the drawing]

[0131] (Brief explanation of the drawing) [Figure 1] We illustrate pH-dependent antigen binding in engineered variants of the parent antibody 18E2, as measured using functional assays. The IC50 of ligand-receptor interaction inhibition was measured for both the parent antibody 18E2 and engineered variants of 18E2 containing one or two defined histidine mutations at both pH 7.4 and pH 5.5. [Figure 2] This study illustrates the pH-dependent binding of the parent antibody 5D1 to the histidine mutant, as measured using MSD technology. [Figure 3] This study illustrates the pH-dependent binding of the parent antibody 5D1 to the histidine mutant, as measured using MSD technology. [Figure 4] This study illustrates the pH-dependent binding of the parent antibody 5D1 to the histidine mutant, as measured using MSD technology. [Examples]

[0132] (Example 1: Building a Heatmap) To identify the amino acid residue positions in the V domain where histidine residues are naturally tolerated within a large antibody repertoire, sequences of the VH, Vλ, and Vκ domains or corresponding Fab fragments from numerous functional (antigen-binding) antibodies obtained from several target antigen projects were pooled and analyzed.

[0133] Each targeted antigen project included immunization of active llamas with purified antigen or DNA expressing the target (DNA vaccination), construction of Fab or scFv phage display libraries, and selection and screening of specific binding factors by phage display. Phage display was performed according to a standard procedure, involving up to three rounds of selection using trypsin as the elution agent, followed by expression of Fab (or scFv) in periplasmic extracts, and identification of Fab (or scFv) binding factors by ELISA, MSD, or SPR (Biacore). In addition to trypsin elution, a parallel phage display process was subsequently performed to further enrich antibodies with pH-dependent binding. This process involved up to three rounds of selection using an eluate at pH 5.5 in TBS (50 mM Tris pH 5.5 and 150 mM NaCl). All sequences were then pooled and subdivided by type (VH, Vλ, or Vκ) and subfamily (e.g., VH1, VH2, and VH3). By removing any strictly duplicate sequences and using only unique sequences, the number of histidine residues at each position in the V domain was counted. This created a "heatmap" showing which amino acid positions in the V domain of antibodies derived from immunized llamas accept histidine. Furthermore, this dataset can be made even more complete using sequences identified by high-throughput sequencing using RNA from immunized llamas.

[0134] (Example 2: Use of a heatmap to enhance pH dependence) Parental antibodies or Fab fragments possessing the required characteristics in terms of high affinity, functionality (e.g., blockade of ligand-receptor interactions), and preferably some degree of pH dependence were selected for histidine engineering using heatmaps.

[0135] First, a small number of histidine mutations were tested on a single parent antibody. Mutations were introduced at position H35 in the VH domain and position L34 in the VL domain of antibody 18E2. Since the target antigen of this antibody cannot be matched with Biacore® due to the presence of conformational epitopes, the pH-dependent effect measured the antibody's ability to block ligand-receptor interactions. IC50 was measured when all proteins were incubated at pH 7.4 or pH 5.5. The results are shown in Table 1 and Figure 1. The data suggest that introducing a single mutation increased the pH dependence of this antibody fourfold. Interestingly, combining the mutated VH with the mutated Vλ further improved the pH dependence without the excessive effect at pH 7.4.

[0136] Table:1 [Table 45]

[0137] These data confirmed that histidine amino acid substitutions at the positions defined by the heatmap are functional, meaning they can enhance pH dependence. Since histidine mutations at other positions could potentially produce better (or worse) effects, a systematic mutation analysis was conducted.

[0138] (Example 3: Complete mutation analysis using a heat map to enhance pH dependence) The pH dependence of antibody 5D1 directed to a second antigen target was introduced using a heatmap. The heatmap revealed that 18 positions in VH of 5D1 were susceptible to histidine substitution, and 12 positions in Vλ of 5D1 were susceptible to histidine substitution (see Table 2). Note that histidine residues are naturally present at positions 55 and 91 in VL, further reducing the number of mutants to be tested. Since the nearest germline to VL is VL8, position L49 in FR2 was not mutated to histidine during initial screening. However, based on the structural conservation of the CDR2 loop, this position could also have been mutated according to the heatmap.

[0139] Table 2 [Table 46]

[0140] Synthetic genes were ordered for each mutation (Geneart, Thermo Scientific). The lyophilized DNA was diluted with H2O and re-cloned into mammalian expression vectors containing human constant domains (CH1-CH2-CH3 for VH, Cλ for Vλ, or Cκ for Vκ) using standard DNA re-cloning methods (restriction enzymes and T4 ligase). The constructed constructs were sequenced and used for transient transfection into 10 ml of suspension HEK293 cells to create a panel of modified variant antibodies, each containing a single HIS substitution.

[0141] Each single VH 5D1 mutant was individually transfected with a wild-type light chain, while each single VL 5D1 mutant was individually transfected with a wild-type heavy chain. HEK293 cells were grown in Freestyle 293 expression medium (Gibco) at 37°C and 5% CO2, at an optimal density of 0.5–0.8 × 10⁶. 6Transfection was performed under cell / mL conditions. The transfection mixture contained 40 μL of OPTI-MEM (Gibco), 1.5 μg of polyethyleneimine (PEI) (Polysciences), and 0.5 μg of DNA per 1 mL of HEK293 culture. The ratio applied to the heavy and light chains was 1–3.

[0142] Four hours after transfection of cells, 9% Primatone HS / UF (Kerry BioScience / Sheffield) was added to reach 10% of the total culture volume. The culture was incubated for 6 days with shaking at 100 rpm, and then collected by centrifugation at 1,000 × g for 10 minutes. 10 μL of a 50% slurry of Protein A-Sepharose beads (GE Healthcare) in 1 × PBS was added per 1 mL of cell culture supernatant. The supernatant containing the beads was incubated on a rotor at 4°C and 25 rpm for at least 2 hours. The beads were pelletized by centrifugation at 610 × g for 2 minutes at a low deceleration rate. The beads bound to functional mAB were transferred to a 96-well 0.45 μm filter plate (PALL) and subsequently washed with 1 × PBS, 1 × PBS containing 0.5 M NaCl (ChemLab), and 0.5 × PBS containing 0.15 M NaCl (ChemLab). Vacuum was applied between each washing step. The bound mAB was eluted using 50 mM sodium citrate (Sigma-Aldrich) at pH 3 containing 0.3 M NaCl (ChemLab). The eluted fraction was neutralized with 1 M KHPO4 / KH2PO4 (VWR) at pH 8, and the protein concentration was measured using NanoDrop 2000 (Thermo Scientific). Screening was performed using Biacore or the MSD platform.

[0143] Several settings can be used for screening with Biacore. In the "direct" setting, the binding of the antigen coated on the CM5 chip (GE Healthcare) to the purified antibody was measured at 25°C in HBSEP buffer at pH 7.4 (0.01 M 4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid (HEPES) pH 7.4, 0.15 M NaCl (ChemLab), 3 mM EDTA, 0.005% (v / v) Surfactant P20), according to the manufacturer's instructions, at pH 7.4 or pH 5.5, or by adjusting the pH to 5.5 using the same buffer but by adding 1 M HCl. In the different "capture" settings, mutant antibodies are captured on the chip coated with polyclonal anti-human IgG in all channels. mAB was injected into HBSEP to bring the RU to 150–300, followed by injection of ligands (samples) at various concentrations in HBSEP buffer at pH 7.4 or adjusted to pH 5.5. After each cycle, the chip surface was regenerated by injecting 10 mM glycine at pH 1.5 for 10 seconds. All kinetic analyses were performed using appropriate software provided by GE Healthcare.

[0144] For antibody testing using the MSD platform, plates (MSD) were coated overnight with the target in 1×PBS at 4°C. Subsequently, the plates were washed three times with 200 μL of 1×PBS while shaking at 600 rpm, and blocked at room temperature for 1–2 hours with 150 μL of 1% (m / v) casein (Sigma-Aldrich) in 1×PBS. Subsequently, the plates were washed three times with 200 μL of 0.05% (m / v) Tween (Merck) (pH 7.4) in 1×PBS or with citrate buffer (0.05 M citrate (Sigma-Aldrich), 0.14 M sodium citrate (Sigma-Aldrich), 0.15 M NaCl (ChemLab)) containing 0.05% (m / v) Tween (Merck) (pH 5.5). To calculate the dose-response curve, 25 μL of mAB added at various concentrations to wells in 1×PBST (PBS containing 0.05% (m / v) Tween) (pH 7.4) or citrate buffer containing 0.05% (m / v) Tween (Merck) (pH 5.5) at pH 7.4 or 5.5. After incubation at room temperature for 1 hour with shaking at 600 rpm, the wells were washed four times with 1×PBST at pH 7.4 or pH 5.5. For the final wash, an additional incubation of 5–10 minutes was performed at room temperature with shaking at 600 rpm using PBST pH 7.4 or citrate buffer pH 5.5. All wells were washed with 1×PBS (pH 7.4) as a final step, and 25 μL of 1:2000 goat anti-human Fc Sulfo (MSD) in 0.1% casein (Sigma-Aldrich)-1×PBS (pH 7.4) was added. The mixture was then allowed to stand at room temperature for 1 hour with shaking at 600 rpm. After the washing step, 200 μL of 1×PBST (pH 7.4) and 150 μL of MSD reading buffer (MSD) were added to each well, and the plate was read using a QuickPlex SQ 120 instrument (MSD).

[0145] In addition to a single pH setting (antibody binding and washing at the same pH of 7.4 or 5.5), the inventors also tested a two-point pH ELISA in which antibody binding was performed at pH 7.4, but washing was performed in citrate buffer at pH 5.5. This setting makes it possible to distinguish the effect of pH on association, dissociation, or both, and better represents the in vivo situation in which binding must occur at pH 7.4 and antigen release must occur in endosomes at pH 5.5–6.0.

[0146] After screening all single mutants shown in Table 2, several single mutants appeared to have an effect on pH-dependent binding (Table 3). For example, 5D1VLm3, 5D1VHm6, and 5D1VHm14 showed significant pH-dependent binding compared to 5D1wt. Furthermore, 5D1VHm13 showed some improvement in pH dependence with favorable kinetics.

[0147] Table:3 [Table 47] [Table 48]

[0148] Furthermore, pH dependence was observed using MSD technology, as shown in Figures 2 and 3, and Table 4. The effects of these mutants are visualized by the curve shown in Figure 2. In Figure 2, pH 5.5 is applied during washing (resulting in a decrease in maximum binding), and pH 5.5 is applied during both binding and washing (resulting in a loss of maximum binding for mutants 5D1VLm3 and 5D1VHm6). In contrast, EC50 measured at pH 7.4 remained similar to that of the wild-type parental antibody.

[0149] Table: 4 [Table 49]

[0150] These data suggest that, even when pH dependence is introduced, its effect is more pronounced on the association phase than on the dissociation phase. Therefore, introducing further histidine mutations to enhance pH dependence may be beneficial.

[0151] (Example: 4 combinations of histidine mutations that further increase pH dependence) Since VLm3, VHm6, and VHm13 all exhibit pH-dependent effects, these mutations were combined in the same antibody. The double HIS mutants VLm3 / VHm6 and VLm3 / VHm13 were generated by simply combining the plasmids encoding each strand during transfection of HEK293 cells. The pH dependence of the resulting antibodies was tested using MSD (see Figure 4).

[0152] The mutation combinations had a very positive effect on pH dependence, as shown in Table 5. The binding shown by the double mutant at pH 5.5 was very poor (pH 5.5-pH 5.5 setting). In contrast, most antibody binding at pH 7.4 was released during washing at pH 5.5 (pH 7.4-pH 5.5 setting).

[0153] Table:5 [Table 50] nd: EC50 cannot be measured due to low binding.

[0154] (Example 5: Complete mutagenesis analysis using a heatmap to enhance the pH dependence of antibody 3D6) The pH dependence of antibody 3D6 directed to a second antigen target was introduced using a heatmap. According to the heatmap and CDR length, 25 positions in the VH region of 3D6 were susceptible to histidine substitution, and 16 positions in the Vκ region of 3D6 were susceptible to histidine substitution (see Table 6). Of these 25 positions, 21 were tested in the initial screening.

[0155] Table:6 [Table 51]

[0156] For each mutation, synthetic genes were ordered (Geneart, Thermo Scientific), cloned into appropriate expression vectors, and antibodies were prepared as shown in Example 3. The pH dependence of the antibodies was tested using a capture setup with a Biacore T200 chip coated with a high volume of anti-human Fc antibody (8000 RU). Subsequently, wild-type or mutant 3D6 antibody (2 μg / ml) in HBS-EP at pH 7.4 was injected to preferably reach 100–200 RU. After stabilizing the signal, the target was injected at 2.5 or 10 nM and washed with HBS-EP buffer at pH 7.4 or 5.5. The equilibrium dissociation constant K at pH 7.4 was also examined. D (M) was measured using T200 software. Since the association was performed at pH 7.4, only the dissociation constant (kd 1 / s) was extracted from the fitting at pH 5.5. The results are shown in Tables 7 and 8.

[0157] Table:7 [Table 52]

[0158] Table:8 [Table 53]

[0159] These data indicate that pH dependence was introduced to 3D6 while minimizing the loss of binding at pH 7.4. Combining single histidine mutations may be beneficial to enhance pH dependence.

[0160] (Example: 6. Combinations of histidine mutations that further increase pH dependence in antibody 3D6) Based on the results in Table 7 or 8, new mutants VHm23, VHm24, VHm25, VHm26, VKm16, VKm17, VKm18, and VKm19 (Table 9) were created by combining mutations within the same domain that introduced significant pH dependence and / or improved binding at pH 7.4. For example, 3D6VHm23 has mutations VHm9 and VHm13, and 3D6VHm24 has mutations VHm0 and VHm16.

[0161] Table:9 [Table 54]

[0162] Novel VH and VK mutants were further combined by simply combining the plasmids encoding each strand during transfection of HEK293 cells. For example, the antibody 3D6VHm23 / VKm11 was combined by co-transfecting VHm23 mutant DNA with mutant VKm11 DNA. After purification, the pH dependence of the resulting antibodies was tested using Biacore® as described in Example 5. Antibodies exhibiting excellent pH dependence and excellent affinity at pH 7.4 were further characterized, with the only difference being the higher concentration of the target added (0.04 nM; 0.1 nM; 0.37 nM; 1.1 nM; 3.3 nM, and 10 nM). D (M) was measured using T200 software. Only the dissociation constant (kd s-1) was extracted from the fitting at pH 5.5. The top five antibodies are shown in Table 10.

[0163] Table: 10 [Table 55]

[0164] The results from Biacore clearly demonstrate that the method presented herein allows for the rapid identification of highly pH-dependent antibodies with minimal binding loss at pH 7.4.

Claims

1. 1. A method for preparing an engineered antibody that exhibits pH-dependent binding to its antigen, comprising: preparing the engineered antibody by substituting at least one amino acid residue of a parent antibody with a histidine residue, the engineered antibody comprising the following hotspot list: [Table 1] wherein at least one amino acid residue selected from the following cold spot list: [Table 2] wherein at least one amino acid residue from is not substituted with histidine, and the engineered antibody exhibits pH-dependent binding to its antigen.

2. 10. The method of claim 1, wherein the engineered antibody has a lower affinity for its antigen at acidic pH than at neutral pH.

3. The dissociation rate constant (k d ) is the dissociation rate constant (k d 2. The method of claim 1, wherein the .alpha.-to-.alpha.

4. The dissociation rate constant (k d ) is the dissociation rate constant (k d 2. The method of claim 1, wherein the .alpha.-to-.alpha.

5. The equilibrium dissociation constant (K) of the engineered antibody-antigen interaction at acidic pH D ) is the equilibrium dissociation constant (K) of the engineered antibody-antigen interaction at neutral pH. D 3. The method of claim 1, wherein the .alpha.-hydroxybenzoate concentration is higher than 100 ppm.

6. 6. The method of claim 1, wherein at least two amino acid residues selected from the hot spot list are substituted with histidine.

7. 6. The method of claim 1, wherein at least three amino acid residues selected from the hot spot list are substituted with histidine.

8. 6. The method of claim 1, wherein at least four amino acid residues selected from the hotspot list are substituted with histidine.

9. 9. The method of any one of claims 1 to 8, wherein one or more of the following residues: H100g, H100k, H100m, H100n, L95, L95d, L95e, L95f, L97 are also substituted with histidine.

10. 10. The method of claim 1, wherein at least two amino acid residues selected from said cold spot list are not substituted with histidine.

11. 10. The method of any one of claims 1 to 9, wherein at least three amino acid residues selected from said cold spot list are not substituted with histidine.

12. 10. The method of claim 1, wherein none of the amino acid residues selected from the cold spot list of VHCDR1, VHCDR2, VLCDR1, and VLCDR2 are substituted with histidine.

13. The method of any one of claims 1 to 12, wherein the parent antibody or its CDR is derived from a camelid species.

14. 1. A method for preparing an engineered antibody that exhibits pH-dependent binding to its antigen, comprising: (a) providing a parent antibody that binds to an antigen; (b) preparing a panel of engineered variants of the parent antibody, each of the engineered variants in the panel differing from the parent antibody by substitution of at least one amino acid residue with a histidine residue, and comprising the following hotspot list: [Table 3] wherein at least one amino acid residue selected from the following cold spot list: [Table 4] wherein at least one amino acid residue from is not substituted with histidine; (c) screening the panel of engineered variants for pH-dependent binding to the antigen, thereby identifying engineered antibodies that exhibit pH-dependent binding to the antigen.

15. 15. The method of claim 14, wherein the engineered antibody identified in part (c) has a lower affinity for its antigen at acidic pH than at neutral pH.

16. The engineered antibody identified in part (c) satisfies the dissociation rate constant (k d ) is the dissociation rate constant (k d 15. The method of claim 14, wherein the temperature is higher than

17. The engineered antibody identified in part (c) satisfies the dissociation rate constant (k d ) is the dissociation rate constant (k d 15. The method of claim 14, wherein the temperature is higher than

18. The engineered antibody identified in part (c) satisfies the equilibrium dissociation constant (K D ) is the equilibrium dissociation constant (K) of the engineered antibody-antigen interaction at neutral pH. D 16. The method according to claim 14 or 15, characterized in that the temperature is higher than

19. and step (b) comprising preparing a panel of engineered variants of said parent antibody, each of said engineered variants in said panel differing from said parent antibody by substitution of at least one amino acid residue in the VH domain with a histidine residue, and comprising the following hotspot list: [Table 5] wherein at least one amino acid residue selected from the following cold spot list: [Table 6] 19. The method of any one of claims 14 to 18, wherein at least one amino acid residue from is not substituted with histidine.

20. and step (b) comprising preparing a panel of engineered variants of said parent antibody, each of said engineered variants in said panel differing from said parent antibody by substitution of at least one amino acid residue in the VL domain with a histidine residue, and comprising the following hotspot list: [Table 7] wherein at least one amino acid residue selected from the following cold spot list: [Table 8] 19. The method of any one of claims 14 to 18, wherein at least one amino acid residue selected from is not substituted with histidine.

21. the VL domain of the parent antibody belongs to the λ class and has the following hotspot list: [Table 9] wherein at least one amino acid residue selected from the following cold spot list: [Table 10] 21. The method of claim 20, wherein at least one amino acid residue from is not substituted with histidine.

22. wherein the VL domain of the parent antibody belongs to the kappa class and has the following hotspot list: [Table 11] wherein at least one amino acid residue selected from the following cold spot list: [Table 12] 21. The method of claim 20, wherein at least one amino acid residue from is not substituted with histidine.

23. 1. A method for preparing an engineered antibody that exhibits pH-dependent binding to its antigen, comprising: (a) identifying a parent antibody that binds to the antigen; (b) preparing a panel of engineered variants of the parent antibody, each of the engineered variants in the panel differing from the parent antibody by substitution of at least one amino acid residue with a histidine residue, and comprising the following hotspot list: [Table 13] wherein at least one amino acid residue selected from the following cold spot list: [Table 14] wherein at least one amino acid residue from is not substituted with histidine; (c) screening the panel of engineered variants for pH-dependent binding to the target antigen, thereby identifying selected amino acid positions where the presence of a histidine confers pH-dependent binding to the antigen; (d) preparing one or more further engineered variants of the parent antibody, each of the variants comprising a histidine at two or more of the selected amino acid positions identified in step (c); and (e) screening the further engineered variants for pH-dependent binding to the antigen; thereby identifying engineered antibodies that exhibit pH-dependent binding to the antigen.

24. 24. The method of claim 23, wherein the engineered antibody identified in part (e) has a lower affinity for its antigen at acidic pH than at neutral pH.

25. The engineered antibody identified in part (e) satisfies the dissociation rate constant (k d ) is the dissociation rate constant (k d 24. The method of claim 23, wherein the temperature is higher than

26. The engineered antibody identified in part (e) satisfies the dissociation rate constant (k d ) is the dissociation rate constant (k d 24. The method of claim 23, wherein the temperature is higher than

27. The engineered antibody identified in part (e) satisfies the equilibrium dissociation constant (K D ) is the equilibrium dissociation constant (K) of the engineered antibody-antigen interaction at neutral pH. D 25. The method according to claim 23 or 24, characterized in that the temperature is higher than

28. 28. The method of any one of claims 23 to 27, wherein each of the engineered variants in the panel of step (b) comprises a single histidine substitution at an amino acid position selected from the hotspot list.

29. 29. The method of any one of claims 23 to 28, wherein the further engineered variants of step (d) comprise histidine residues at three or more of the selected amino acid positions identified in step (c).

30. 29. The method of any one of claims 23 to 28, wherein the further engineered variants of step (d) comprise histidine residues at four or more of the selected amino acid positions identified in step (c).

31. 29. The method of any one of claims 23 to 28, wherein the further engineered variant of step (d) further comprises a histidine residue at one or more of the following amino acid positions: H100g, H100k, H100m, H100n, L95, L95d, L95e, L95f, L97.

32. 1. A method for preparing an engineered antibody that exhibits pH-dependent binding to its antigen, comprising: (a) identifying a parent antibody that binds to the antigen; (b) preparing a first panel of engineered variants of the parent antibody, each of the engineered variants in the panel differing from the parent antibody by substitution of at least one amino acid residue in the VH domain with a histidine residue, and having the following hotspot list: [Table 15] wherein at least one amino acid residue selected from the following cold spot list: [Table 16] wherein at least one amino acid residue from is not substituted with histidine; (c) screening the first panel of engineered variants for pH-dependent binding to the antigen; thereby identifying one or more selected amino acid positions in the VH domain where the presence of a histidine confers pH-dependent binding; (d) preparing a second panel of engineered variants of the parent antibody, each of the engineered variants in the panel differing from the parent antibody by substitution of at least one amino acid residue in the VL domain with a histidine residue, and having the following hotspot list: [Table 17] wherein at least one amino acid residue selected from the following cold spot list: [Table 18] wherein at least one amino acid residue selected from the group consisting of: (e) screening the second panel of engineered variants for pH-dependent binding to the antigen; thereby identifying one or more selected amino acid positions in the VL domain where the presence of a histidine confers pH-dependent binding; (f) preparing one or more further engineered variants of the parent antibody, each of which differs from the parent antibody by a substitution of an amino acid at one or more selected amino acid positions in the VH domain identified in step (c) with histidine, and by a substitution of an amino acid at one or more selected amino acid positions in the VL domain identified in step (e) with histidine; (g) screening the further engineered variants for pH-dependent binding to the antigen; thereby identifying engineered antibodies that exhibit pH-dependent binding to the antigen.

33. the VL domain of the parent antibody belongs to the λ class and has the following hotspot list: [Table 19] wherein at least one amino acid residue selected from the following cold spot list: [Table 20] 33. The method of claim 32, wherein at least one amino acid residue from is not substituted with histidine.

34. wherein the VL domain of the parent antibody belongs to the kappa class and has the following hotspot list: [Table 21] wherein at least one amino acid residue selected from the following cold spot list: [Table 22] 33. The method of claim 32, wherein at least one amino acid residue from is not substituted with histidine.

35. 35. The method of any one of claims 32-34, wherein the engineered antibody identified in part (g) has a lower affinity for its antigen at acidic pH than at neutral pH.

36. The engineered antibody identified in part (g) satisfies the dissociation rate constant (k d ) is the dissociation rate constant (k d 35. The method according to any one of claims 32 to 34, characterized in that the concentration of hydroxybenzoates is higher than

37. The engineered antibody identified in part (g) satisfies the dissociation rate constant (k d ) is the dissociation rate constant (k d 35. The method according to any one of claims 32 to 34, characterized in that the concentration of hydroxybenzoates is higher than

38. The engineered antibody identified in part (g) satisfies the equilibrium dissociation constant (K D ) is the equilibrium dissociation constant (K) of the engineered antibody-antigen interaction at neutral pH. D 38. The method according to any one of claims 32 to 37, characterized in that the concentration is higher than

39. 39. The method of any one of claims 32-38, wherein the further engineered variants of step (f) comprise histidines at all three or more of the selected amino acid positions identified in steps (c) and (e).

40. 39. The method of any one of claims 32-38, wherein the further engineered variants of step (f) comprise histidines at all four or more of the selected amino acid positions identified in steps (c) and (e).

41. 39. The method of any one of claims 32 to 38, wherein the further engineered variant of step (f) further comprises a histidine residue at one or more of the following amino acid positions: H100g, H100k, H100m, H100n, L95, L95d, L95e, L95f, L97.

42. 42. The method of any one of claims 14 to 41, wherein the parent antibody or its CDRs are derived from a camelid species.

43. 1. An engineered antibody that exhibits pH-dependent binding to its antigen, wherein at least one amino acid in a CDR of the engineered antibody is a histidine residue and has the following hot spot list: [Table 23] wherein at least one amino acid residue selected from is a histidine residue, and the following cold spot list: [Table 24] wherein at least one amino acid residue selected from the group consisting of: is not a histidine residue.

44. 44. The engineered antibody of claim 43, wherein the engineered antibody has a lower affinity for its antigen at acidic pH than at neutral pH.

45. The dissociation rate constant (k d ) is the dissociation rate constant (k d 44. The engineered antibody of claim 43, wherein the antibody has a IgG antibody activity greater than IgG.

46. The equilibrium dissociation constant (K) of the engineered antibody-antigen interaction at acidic pH D ) is the equilibrium dissociation constant (K) of the engineered antibody-antigen interaction at neutral pH. D 45. The engineered antibody of claim 43 or 44, wherein the antibody has a IgG antibody activity greater than IgG.

47. 47. The engineered antibody of any one of claims 43-46, wherein at least two amino acid residues selected from said hot spot list are histidines.

48. 47. The engineered antibody of any one of claims 43-46, wherein at least three amino acid residues selected from said hot spot list are histidines.

49. 47. The engineered antibody of any one of claims 43-46, wherein at least four amino acid residues selected from said hot spot list are histidines.

50. 50. The engineered antibody of any one of claims 43-49, further comprising a histidine residue at one or more of the following amino acid positions: H100g, H100k, H100m, H100n, L95, L95d, L95e, L95f, L97.

51. 51. The engineered antibody of any one of claims 43-50, wherein at least two amino acid residues at the positions listed in Table B are not histidine.

52. 51. The engineered antibody of any one of claims 43-50, wherein at least three amino acid residues at positions selected from said cold spot list are not histidine.

53. 51. The engineered antibody of any one of claims 43-50, wherein none of the amino acid residues at the positions listed in the cold spot list is histidine, with the proviso that one or more of the following amino acid positions may contain histidine: H100g, H100k, H100m, H100n, L95, L95d, L95e, L95f, L97.

54. 54. The engineered antibody of any one of claims 43 to 53, which is an engineered variant of a camelid antibody.

55. 55. The engineered antibody of any one of claims 43 to 54, which comprises an Fc region and binds to the neonatal Fc receptor (FcRn).

Images