Thermodynamic Consequences of Mutations in Vernier Zone Residues of a Humanized Anti-human Epidermal Growth Factor Receptor Murine Antibody, 528

Abstract

To investigate the role of Vernier zone residues, which are comprised in the framework regions and underlie the complementarity-determining regions (CDRs) of antibodies, in the specific, high affinity interactions of antibodies with their targets, we focused on the variable domain fragment of murine anti-human epidermal growth factor receptor antibody 528 (m528Fv). Grafting of the CDRs of m528Fv onto a selected framework region of human antibodies, referred to as humanization, reduced the antibody's affinity for its target by a factor of 1/40. The reduction in affinity was due to a substantial reduction in the negative enthalpy change associated with binding. Crystal structures of the ligand-free antibody fragments showed no noteworthy conformational changes due to humanization, and the loop structures of the CDRs of the humanized antibodies were identical to those of the parent antibodies. Several mutants of the CDR-grafted (humanized) variable domain fragment (h528Fv), in which some of the Vernier zone residues in the heavy chain were replaced with the parental murine residues, were constructed and prepared using a bacterial expression system. Thermodynamic analyses of the interactions between the mutants and the soluble extracellular domain of epidermal growth factor receptor showed that several single mutations and a double mutation increased the negative enthalpy and heat capacity changes. Combination of these mutations, however, led to somewhat reduced negative enthalpy and heat capacity changes. The affinity of each mutant for the target was within the range for the wild-type h528Fv, and this similarity was due to enthalpy-entropy compensation. These results suggest that Vernier zone residues make enthalpic contributions to antigen binding and that the regulation of conformational entropy changes upon humanization of murine antibodies must be carefully considered and optimized. To investigate the role of Vernier zone residues, which are comprised in the framework regions and underlie the complementarity-determining regions (CDRs) of antibodies, in the specific, high affinity interactions of antibodies with their targets, we focused on the variable domain fragment of murine anti-human epidermal growth factor receptor antibody 528 (m528Fv). Grafting of the CDRs of m528Fv onto a selected framework region of human antibodies, referred to as humanization, reduced the antibody's affinity for its target by a factor of 1/40. The reduction in affinity was due to a substantial reduction in the negative enthalpy change associated with binding. Crystal structures of the ligand-free antibody fragments showed no noteworthy conformational changes due to humanization, and the loop structures of the CDRs of the humanized antibodies were identical to those of the parent antibodies. Several mutants of the CDR-grafted (humanized) variable domain fragment (h528Fv), in which some of the Vernier zone residues in the heavy chain were replaced with the parental murine residues, were constructed and prepared using a bacterial expression system. Thermodynamic analyses of the interactions between the mutants and the soluble extracellular domain of epidermal growth factor receptor showed that several single mutations and a double mutation increased the negative enthalpy and heat capacity changes. Combination of these mutations, however, led to somewhat reduced negative enthalpy and heat capacity changes. The affinity of each mutant for the target was within the range for the wild-type h528Fv, and this similarity was due to enthalpy-entropy compensation. These results suggest that Vernier zone residues make enthalpic contributions to antigen binding and that the regulation of conformational entropy changes upon humanization of murine antibodies must be carefully considered and optimized. Structural and functional analyses of antigen-antibody protein-protein interactions have revealed that complementarity-determining regions (CDRs) 5The abbreviations used are:CDRcomplementarity-determining regionCDR-HCDR in heavy chainCDR-LCDR in light chainEGFRepidermal growth factor receptorsEGFRsoluble extracellular domain of human EGFRm528Fvvariable domain fragment of murine 528 antibodyh528Fvvariable domain of humanized 528 antibodym528FabFab fragment of murine antibody 528ITCisothermal titration calorimetryVHheavy chain of variable domainVLlight chain of variable domainMALDI-TOFmatrix-assisted laser desorption ionization time-of-flightFvvariable domain fragment of antibodyr.m.s.root mean squareSCshape complementarity. in the variable domains of antibodies play a critical role in the specificity and affinity of the antibodies for their targets by means of shape and charge complementarities (1Davies D.R. Cohen G.H. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 7-12Crossref PubMed Scopus (484) Google Scholar, 2Braden B.C. Poljak R.J. FASEB J. 1995; 9: 9-16Crossref PubMed Scopus (184) Google Scholar, 3Padlan E.A. Adv. Protein Chem. 1996; 49: 57-133Crossref PubMed Google Scholar, 4Mariuzza R.A. Pojak R.J. Curr. Opin. Immunol. 1993; 5: 50-55Crossref PubMed Scopus (78) Google Scholar). Antibodies have a common fold (the immunoglobulin fold), and hypervariable CDRs are located on one edge of the framework region (1Davies D.R. Cohen G.H. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 7-12Crossref PubMed Scopus (484) Google Scholar, 2Braden B.C. Poljak R.J. FASEB J. 1995; 9: 9-16Crossref PubMed Scopus (184) Google Scholar, 3Padlan E.A. Adv. Protein Chem. 1996; 49: 57-133Crossref PubMed Google Scholar, 4Mariuzza R.A. Pojak R.J. Curr. Opin. Immunol. 1993; 5: 50-55Crossref PubMed Scopus (78) Google Scholar). Grafting of the CDRs of antibodies onto the frameworks of other antibodies has been attempted (5Ewert S. Honegger A. Plueckthun A. Methods. 2004; 34: 184-199Crossref PubMed Scopus (182) Google Scholar, 6Nishibori N. Horiuchi H. Furusawa S. Matsuda H. Mol. Immunol. 2006; 43: 634-642Crossref PubMed Scopus (48) Google Scholar, 7Soderlind E. Strandberg L. Jirholt P. Kobayashi N. Alexeiva V. Aberg A.M. Nilsson A. Jansson B. Ohlin M. Wingren C. Danielsson L. Carlsson R. Borrebaeck C.A. Nat. Biotechnol. 2000; 18: 852-856Crossref PubMed Scopus (276) Google Scholar, 8Jirholt P. Ohlin M. Borrebaeck C.A. Soderlind E. Gene (Amst.). 1998; 215: 471-476Crossref PubMed Scopus (68) Google Scholar, 9Nicaise M. Valerio-Lepiniec M. Minard P. Desmadril M. Protein Sci. 2004; 13: 1882-1891Crossref PubMed Scopus (48) Google Scholar). Especially interesting from diagnostic and therapeutic viewpoints is the humanization, or reshaping, of murine antibodies, whereby a set of CDRs from murine antibodies are transplanted to appropriate scaffolds of human antibodies to reduce immune responses against murine antibodies in human hosts (10Verhoeyen M. Milstein C. Winter G. Science. 1988; 239: 1534-1536Crossref PubMed Scopus (289) Google Scholar, 11Clark M. Immunol. Today. 2000; 21: 397-402Abstract Full Text Full Text PDF PubMed Scopus (245) Google Scholar, 12Stochwin L.H. Holmes S. Biochem. Soc. Trans. 2003; 31: 433-436Crossref PubMed Google Scholar, 13Stern M. Herrmann R. Crit. Rev. Oncol. Hematol. 2005; 54: 11-29Crossref PubMed Scopus (213) Google Scholar, 14Jones P.T. Dear P.H. Foote J. Neuberger M.S. Winter G. Nature. 1986; 321: 522-525Crossref PubMed Scopus (1134) Google Scholar, 15Riechmann L. Clark M. Waldmann H. Winter G. Nature. 1988; 332: 323-327Crossref PubMed Scopus (1324) Google Scholar, 16Yoon S.O. Lee T.S. Kim S.J. Jang M.H. Kang Y.J. Park J.H. Kim K.S. Lee H.S. Ryu C.J. Gonzales N.R. Kashmiri S.V. Lim S.M. Choi C.W. Hong H.J. J. Biol. Chem. 2006; 281: 6985-6992Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar, 17Queen C. Schneider W.P. Selick H.E. Payne P.W. Landolfi N.F. Duncan J.F. Avdalovic N.M. Levitt M. Junghans R.P. Waldmann T.A. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 10029-10033Crossref PubMed Scopus (570) Google Scholar, 18Co M.S. Deschamps M. Whitley R.J. Queen C. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 2869-2873Crossref PubMed Scopus (114) Google Scholar, 19Co M.S. Avdalovic N.M. Caron P.C. Avdalovic M.V. Scheinberg D.A. Queen C. J. Immunol. 1992; 48: 1149-1154Google Scholar, 20Ohtomo M. S. Mol. Immunol. 1995; PubMed Scopus Google Scholar). complementarity-determining region in heavy chain in light chain epidermal growth factor receptor soluble extracellular domain of human variable domain fragment of murine 528 antibody variable domain of humanized 528 antibody fragment of murine antibody 528 titration heavy chain of variable domain light chain of variable domain laser desorption ionization variable domain fragment of antibody mean shape complementarity. of the CDRs of murine antibodies onto appropriate frameworks of human antibodies results in reduced affinity or specificity for the target antigen (10Verhoeyen M. Milstein C. Winter G. Science. 1988; 239: 1534-1536Crossref PubMed Scopus (289) Google Scholar, 14Jones P.T. Dear P.H. Foote J. Neuberger M.S. Winter G. Nature. 1986; 321: 522-525Crossref PubMed Scopus (1134) Google Scholar, 16Yoon S.O. Lee T.S. Kim S.J. Jang M.H. Kang Y.J. Park J.H. Kim K.S. Lee H.S. Ryu C.J. Gonzales N.R. Kashmiri S.V. Lim S.M. Choi C.W. Hong H.J. J. Biol. Chem. 2006; 281: 6985-6992Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). The of Foote and Winter J. Winter G. J. Mol. Biol. 1992; PubMed Scopus Google has that antibody residues in the framework the CDRs play a critical role in of the loop structures of the these residues, referred to as Vernier zone residues, with the of these residues for the of loop in antibodies, and of these have a role in the of structures in the and a role in affinity J. Winter G. J. Mol. Biol. 1992; PubMed Scopus Google Scholar). The epidermal growth factor receptor is a with domain J. 2000; Full Text Full Text PDF PubMed Scopus Google Scholar, H.S. C. C.W. S. Mol. 2003; Full Text Full Text PDF PubMed Scopus Google Scholar, J. Oncol. Biol. 2004; Full Text Full Text PDF PubMed Scopus Google Scholar). of the has been in and is considered to be with critical in the of A. G. R. M. N. 1988; PubMed Scopus Google Scholar, R. N. Crit. Rev. Oncol. Hematol. 1995; PubMed Scopus Google Scholar, Rev. 1992; 13: PubMed Scopus Google Scholar, A. Adv. 2000; PubMed Google Scholar). To several regulation of the have been attempted for some of which are in J. J. Oncol. 2005; PubMed Scopus Google Scholar, Oncol. 2005; PubMed Scopus Google Scholar, G. Google Scholar, J. J. Oncol. PubMed Scopus Google Scholar, L. B. A. C. J. M. A. P. Google Scholar). we focused on M. M. S. N. H. S. 1996; Google Scholar, M. M. N. H. H. S. 1998; Google Scholar, H. M. M. N. S. H. H. R. M. S. Immunol. PubMed Scopus Google Scholar, S. R. M. N. H. H. S. H. S. Immunol. PubMed Scopus Google Scholar, R. H. M. J. Biochem. PubMed Scopus Google Scholar). we constructed a of the variable region of antibody and antibody H. R. M. M. H. S. H. S. Immunol. 2004; PubMed Scopus Google Scholar, R. H. M. 2006; PubMed Scopus Google Scholar). that humanization of a murine antibody for the human A. J. G.H. J. Proc. Natl. Acad. Sci. U. S. A. PubMed Scopus Google led to a substantial reduction of the antibody's affinity for its this we focused on Vernier zone residues of 528 to is for the reduction of affinity that upon Several mutants of the variable domains of humanized 528 (h528Fv), in which some of the Vernier zone residues in the heavy chain were replaced with the parental murine residues, were constructed and prepared using a bacterial expression system. The interactions between mutants and the soluble extracellular domain of were using titration the of results we the role that Vernier zone residues play in the high affinity of antibodies for their of CDR-grafted 528 murine antibody 528 was humanized by means of the of human immunoglobulin variable regions were from the on the by using the the of human variable light and variable heavy with the with the 528 framework of in and in were selected for the framework and the the of the variable domain fragment of anti-human murine antibody 528 and of humanized The were by a on as R. H. M. J. Biochem. PubMed Scopus Google Scholar, R. PubMed Scopus Google Scholar). The were M. H. M. J. Immunol. Methods. 1998; PubMed Scopus Google Scholar). The and are in was as A. M. H. J. Biol. Chem. 2003; Full Text Full Text PDF PubMed Scopus Google Scholar). The used for are in The mutants constructed in this are in constructed in this humanized murine in a and of and the and of and its mutants were the expression for bacterial of the H. R. M. M. H. S. H. S. Immunol. 2004; PubMed Scopus Google Scholar, R. H. M. 2006; PubMed Scopus Google Scholar). was with the constructed and in with of of the a of was to the a of to expression of the target and the were The was with and the were by for The were in and against the for The were by for The was onto a affinity with the The was with and and the was with and was by by on was by which were by Protein were in a of and were in for was to the of the and to for and to the with and the and was for of the of was by on were on a with a laser was as a and to in from each were with the in a and of the was onto the on the the were with of to was in and with of and were of fragment of was by of with The fragment was from and the domain fragment by means of a Protein The was and used for of and were using the Crystal was by the using Crystal and and and were in against a of and for and and for The of and for and for in and for were the of were to by using as a The in the with the a and The was were with and of the PubMed Scopus Google Scholar). The of was by with the of Biol. Scholar). The used were the domain of humanized antibody and the domain of antibody with the Biol. PubMed Scopus Google was with the and in the P. J. M. Biol. 1998; 54: PubMed Scopus Google Scholar). The was using the in J. Biol. PubMed Scopus Google against a in were using J. Mol. Biol. 1993; PubMed Scopus Google as in the with a were using on the for were the of were to by using as a The in the with the a and The was The was as for The used were the and domains of and the and domains of antibody and are in for and factor of factor in a for the interactions between the domain and h528Fv, and several mutants were by using a from S. J.F. Biochem. 1989; PubMed Scopus Google Scholar). The were as a of in in a was with a of the in the The was in a of and were in of a binding using the The enthalpy changes and binding for the antigen-antibody interactions were from the titration The and the entropy for the were from the and The heat capacity changes were from the of the enthalpy changes. of Protein of was by using The of m528Fv was by using and the of was by using for the and mutants for those a was used for the J. Biol. Chem. 1995; Full Text Full Text PDF PubMed Scopus Google Scholar). murine antibody 528 was humanized by the selected human antibody frameworks with the to the framework of and regions by E.A. K.S. C. of of of were onto the frameworks and human heavy chain are Immunol. PubMed Scopus Google Scholar). of parental murine antibody 528 and selected human antibody framework to J. 1992; PubMed Scopus Google Scholar). The were as and The fragments were using E. expression and were from the by affinity and of m528Fv and were of functional of the was to the of analyses using some and antibody that fragments binding identical to that of and of the binding of the to the antigen by the was S. R. M. N. H. H. S. H. S. Immunol. PubMed Scopus Google Scholar, R. H. M. J. Biochem. PubMed Scopus Google Scholar, H. R. M. M. H. S. H. S. Immunol. 2004; PubMed Scopus Google Scholar, A. J. G.H. J. Proc. Natl. Acad. Sci. U. S. A. PubMed Scopus Google Scholar). These results suggest that m528Fv and the of as the parental 528 To investigate the interactions between and the antibody fragments we analyses by means of expression was used for of The of the was using the chain of is to be the and a of which that was with those in J. Biol. Chem. Full Text PDF PubMed Google Scholar, A.M. G. J. Biol. Chem. Full Text PDF PubMed Google Scholar). the Thermodynamic and from the titration are in and the of the enthalpy changes due to binding is in The affinity of for was that of the m528Fv fragment the negative enthalpy change of the was that for m528Fv and the negative entropy change was The heat capacity changes of the interactions of m528Fv and h528Fv, from the of the enthalpy were and of antibody interactions and in a of the enthalpy change of the interactions between and results are The structures of and were in of and by the The one in the and in the and are in The and of the of were and and those of were and The structures of and are in is on the was on the of The mean of the between the of and and of the the CDRs were are in and The results that the of is to that of the structures CDRs changes in the loop structures were the CDRs were The structures of these CDRs be the B. A.M. C. J. Mol. Biol. PubMed Scopus Google Scholar). has a and to is in the loop and to and to structures and has one and a in the of each chain in a in the of each loop or chain CDRs in a loop structures of the light and heavy of and The of the are and changes in the of and were for m528Fv and the of the and of were and the and of were on those of m528Fv and the of the and of were on the of the of the were and for the and and The of the for m528Fv and were and The in the from the interactions residues and in the chain To the role of the Vernier zone residues in the high affinity of antibodies for their targets, we constructed mutants of in which some of the Vernier zone residues were to the residues of the parental murine The Vernier zone residues of the of were identical to those of the parental murine one and we focused on the Vernier zone residues of the heavy chain the target and which the and have on the of the was selected as the target The residues these of were with those of and the mutants were prepared as using a bacterial expression system. The mutants constructed are in Thermodynamic analyses of the interactions between and the mutant were by and the results are in of to led to a substantial in the negative enthalpy change of the with the was the as that of of the entropy to h528Fv, the affinity of for the target was by the The heat capacity change was which was negative that of of to led to a substantial in the negative enthalpy change of the with and the was the as that of the parental of the entropy the affinity of for the target was by the The heat capacity change was which was negative that of of to led to a substantial in the negative enthalpy change of the with of the entropy the affinity of for the target was by the The heat capacity change was which was negative that of and with the residues in These are located in the framework region the of the of the The enthalpy change for and was for by the entropy of the between and was the as that of with of with led to reduction of the enthalpy change due to the mutation in The heat capacity change was which was negative that of with The negative enthalpy change was increased by with that of The enthalpy change was for by the in entropy and the affinity for the target was by mutation of The heat capacity change was which was negative that of with The enthalpic due to the mutations was reduced by of with and the negative of the heat capacity change was reduced with that of with The mutation led to substantial reduction of the enthalpic and of were with the murine The mutations increased the negative enthalpy change by however, the affinity for the target was The heat capacity change was of was with The enthalpy changes and entropy were reduced by the and no change in the was The heat capacity change was identical to that for Thermodynamic of of parental murine 528 the binding of m528Fv and to on in a that the of the were identical to that of the parental The analyses of the binding of m528Fv or to that humanization led to reduction of the due to enthalpy changes to the of as and interactions binding S. Scopus Google Scholar). the in the negative enthalpy change due to humanization of 528 was by in the of with upon Crystal of and humanization conformational changes in the antibody we the structures of and were to a of m528Fv appropriate for we prepared a fragment of from the parental murine 528 and Crystal structures of and that the of of the and Vernier zone residues were humanization of The residues between and were in from m528Fv to h528Fv, for the regions and of between the and of m528Fv and were and which that the shape complementarities in the were These results suggest that no conformational changes were by humanization, in the or the of the the of the and of were and and of were on those of m528Fv and the of the and of were on the of the of the were and for and and on analyses of antibody fragments have that of and fragments are within G. Poljak R.J. Nature. PubMed Scopus Google Scholar, H. M. M. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google we critical changes in the of and Thermodynamic of in Vernier of of the affinity or specificity of humanized antibodies, Vernier zone residues of humanized antibodies be with the residues of murine antibodies. To the role of Vernier zone residues of humanized antibodies, we the target and we mutants and these have been in other antibody M. J. M. Mol. Immunol. 31: PubMed Scopus (46) Google Scholar, C.A. J. C.J. M. Protein 1991; PubMed Scopus Google Scholar, Foote J. J. Immunol. PubMed Scopus Google Scholar). The of the interactions between the mutants and that these mutations increased the negative enthalpy changes. the mutations affinity for the target which that entropy changes were some however, we no of the affinity for the enthalpic on the of These results suggest that the of mutations on the affinity were of mutations with of the region of the that of and with the murine residues led to in the negative of the heat capacity heat capacity changes in protein-protein are and are to from of the E. Adv. Protein Chem. 1992; 43: PubMed Google Scholar, A. Curr. Opin. Chem. Biol. PubMed Scopus Google Scholar, 1991; PubMed Scopus Google Scholar, M. Mol. Biol. 1993; PubMed Scopus Google Scholar). by and C.J. Protein Sci. 1996; 5: PubMed Scopus Google and Science. PubMed Scopus Google and E. Adv. Protein Chem. 1992; 43: PubMed Google have the of other as of the and changes due to Science. PubMed Scopus Google Scholar, P. M. R. J. Proc. Natl. Acad. Sci. U. S. A. 2000; PubMed Scopus Google with the negative of the heat capacity The heat capacity changes due to antigen binding of the mutants and within the range for antigen-antibody which changes of the of or the binding from upon of with to the of E. 1995; 21: PubMed Scopus Google the of binding is by the change in entropy from upon is the change in the entropy from conformational changes due to of the antigen-antibody and is the entropy is by is the which the entropy change is considered to be and be considered to have a E. 1995; 21: PubMed Scopus Google Scholar). has been that the conformational entropy to a for by the the be as the between and is to be and from is to be the the entropy change is and is to be from the between and h528Fv, is to be which is the for that the by is reduced by The entropy change is to be and is to be which was the for which that the negative entropy change by the conformational changes was These that humanization of m528Fv led to reduction of the entropy due to the conformational changes. and of the interactions between and the constructed mutants mutants for and and These results suggest that conformational changes were increased by the mutations and that the entropy due to the conformational changes was for by the increased entropy that mutation of Vernier zone residues led to a substantial in the conformational entropy change and that of these mutations led to reduction of the conformational entropy and mutations increased the conformational entropy and however, of these mutations led to a in the conformational changes of the between and 528 antibody fragments in a of on of for the binding of and mutants to are against the in was with a of The for m528Fv with a of referred to as has been in enthalpy change of a mutant with is for by entropy and enthalpy change is for by a entropy which the in the change of antigen binding the we the for the of antigen binding for and its mutants from to and these were from the for m528Fv The of and its mutants showed and the for m528Fv is from the These results suggest the of A. H. H. Full Text Full Text PDF PubMed Scopus Google that from enthalpy-entropy and be by mutations in Vernier zone The results of analyses suggest that mutations in the framework regions of the antibody structures of the in the Vernier zone residues to of the regions upon however, enthalpy-entropy with the of the antibody affinity for its has been in of the of affinity of from the between enthalpy and entropy changes is by in negative enthalpy a in negative entropy changes the affinity or H. M. A. J. Biol. Chem. 1995; Full Text Full Text PDF PubMed Scopus (46) Google Scholar). that affinity of antibodies to a in the changes of the antibody upon binding L.H. P. Science. PubMed Scopus Google Scholar). The results that some mutations in the Vernier residues increased the negative heat capacity and this be with changes in the structures of the results that of Vernier zone residues of the humanized antibody 528 with the murine residues the affinity of antigen and analyses suggest that the of changes due to mutations on the of the antibody with the target were by compensation. These results suggest that Vernier zone residues enthalpic contributions to antigen binding and that the regulation of conformational entropy changes upon humanization of murine antibodies must be carefully considered and optimized. that the in this were in framework residues and in the of antigen the residues of m528Fv were in Crystal structures of antibody fragments that no conformational changes were the loop structures by These suggest that the that the antibody the antigen a critical to the high affinity of the antibody for its that the of and in m528Fv was by critical of variable domain interactions to high specificity and affinity of antibodies for targets has been G. Poljak R.J. Nature. PubMed Scopus Google Scholar, H. M. M. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar, H. J. Biol. Chem. 2003; Full Text Full Text PDF PubMed Scopus Google Scholar, H. H. N. J. Mol. Biol. PubMed Scopus Google Scholar, Foote J. J. Immunol. Google Scholar). Structural analyses of the for of the from and N. S. M. and N. of the for with the with