"Superhumanized" antibodies: reduction of immunogenic potential by complementarity-determining region grafting with human germline sequences: application to an anti-CD28

Humanized Abs are created by combining, at the genetic level, the complementarity-determining regions of a murine mAb with the framework sequences of a human Ab variable domain. This leads to a functional Ab with reduced immunogenic side effects in human therapy. In this study, we report a new approach to humanizing murine mAbs that may reduce immunogenicity even further. This method is applied to humanize the murine anti-human CD28 Ab, 9.3. The canonical structures of the hypervariable loops of murine 9.3 were matched to human genomic V gene sequences whose hypervariable loops had identical or similar canonical structures. Framework sequences for those human V genes were then used, unmodified, with the 9.3 complementarity-determining regions to construct a humanized version of 9.3. The humanized 9.3 and a chimeric 9.3 control were expressed in Escherichia coli as Fab. The humanized Fab showed a moderate loss in avidity in a direct binding ELISA with immobilized CD28-Ig fusion protein (CD28-Ig). Humanized 9.3 blocked ligation of CD28-Ig to cells expressing the CD28 receptor CD80. Lastly, the humanized 9.3 showed biological activity as an immunosuppressant by inhibiting a MLR.     

Therapy with monoclonal antibodies. An in vivo model for the assessment of therapeutic potential.

A set of rat-human and rat-rat chimeric mAb has been created, all possessing V regions identical in their specificity for the mouse CD8 Ag. In vitro all antibodies were able to block cell-mediated lysis but varied greatly in their capacity to utilize rabbit complement. We examined the ability of these chimeric antibodies to deplete in vivo and established a clear hierarchy. Of the human IgG subclasses, only IgG1, 2, and 3 could fix complement in vitro, yet all (IgG1-4) were remarkably potent at depleting CD8+ PBL in vivo. In contrast, human IgA2 and IgE were ineffective at clearing CD8+ PBL. The vector system used to create these antibodies together with the small doses of antibodies required to deplete in vivo make this a simple and rapid system for testing the effects of different antibody isotypes and mutants. We have shown that a mutant of human IgG1, which is incapable of fixing complement, depletes perfectly well in vivo, whereas an aglycosyl IgG1 mutant is rendered inactive. Our model provides a unique opportunity to study effector functions and motifs that are used by mAb in vivo and will help in the design of improved antibodies for human therapy.     

Humanization of an anti-CD34 monoclonal antibody by complementarity-determining region grafting based on computer-assisted molecular modelling

4C8 is a new mouse anti-human CD34 monoclonal antibody (mAb), which recognizes class II CD34 epitopes and can be used for clinical hematopoietic stem/progenitor cell selection. In an attempt to improve its safety profiles, we have developed a humanized antibody of 4C8 by complementarity-determining region (CDR) grafting method in this study. Using a molecular model of 4C8 built by computer-assisted homology modelling, framework region (FR) residues of potential importance to the antigen binding were identified. A humanized version of 4C8, denoted as h4C8, was generated by transferring these key murine FR residues onto a human antibody framework that was selected based on homology to the mouse antibody framework, together with the mouse CDR residues. The resultant humanized antibody was shown to possess antigen-binding affinity and specificity similar to that of the original murine antibody, suggesting that it might be an alternative to mouse anti-CD34 antibodies routinely used clinically.     

Stability of murine, chimeric and humanized antibodies against pre-S2 surface antigen of hepatitis B virus

We have constructed a humanized antibody with specificity for the pre-S2 surface antigen of hepatitis B virus (HBV) by grafting the complementarity determining regions (CDRs) of parental murine monoclonal antibody (mAb) into human anti-Sm antibody framework regions. The humanized antibody has a substitution at position 94 in a framework region of the heavy chain variable region, and exhibits the same antigen binding affinity as the parental murine monoclonal and chimeric antibodies. In order to assess the stability of these antibodies, thermal inactivation of the parental, chimeric and humanized antibodies was analyzed. Fifty percent inactivation of the chimeric and humanized antibodies was observed at 63.7 degrees C and 68.7 degrees C, respectively, compared to 55.0 degrees C for murine antibody. The humanized antibody also exhibited increased stability against denaturant. Guanidine-induced unfolding monitored by the changes in fluorescence intensity at 360 nm showed that midpoints of the transition of the chimeric and humanized antibodies were 2.47 M and 2.56 M, respectively, whereas that of the murine antibody was 1.36 M.     

Comparison of GK1.5 and chimeric rat/mouse GK1.5 anti-CD4 antibodies for prolongation of skin allograft survival and suppression of alloantibody production in mice.

GK1.5 is a rat mAb that recognizes the mouse CD4 Ag. It has been shown to deplete CD4+ cells in vivo and to be immunosuppressive. To evaluate the effect of the C region of this antibody in achieving cell depletion, chimeric antibodies, each having the rat GK1.5 V regions and one of the four mouse IgG C region isotypes, were compared with the native rat antibody. The chimeric antibodies and the native antibody were tested for their ability to mediate in vitro C-dependent cytotoxicity, in vivo cell depletion, and prolongation of allogeneic skin graft survival and suppression of alloantibody production. In vitro C-dependent cytotoxicity assays revealed that rat IgG2b and the chimeric antibodies containing mouse IgG2a, mouse IgG2b, and mouse IgG3 were effective in lysing CD4+ lymphocytes whereas mouse IgG1 was ineffective. In vivo studies of CD4+ cell depletion showed that mouse IgG2a, rat IgG2b, and mouse IgG2b were effective isotypes, mouse IgG1 was less effective, and mouse IgG3 did not deplete CD4+ cells. A correlation was found between the ability of an isotype to deplete CD4+ cells in vivo and its ability to prolong the survival of skin allografts and to suppress alloantibody production. The nondepleting mouse IgG3 was ineffective in these assays. Overall the most effective mouse isotype was IgG2a which was as effective as rat IgG2b. These results indicate 1) that syngeneic isotypes of mAb can cause cell depletion and consequently the prolongation of allograft rejection and suppression of alloantibody production; 2) that not all isotypes are equally effective; and 3) that the ability of a given isotype to deplete cells in vivo does not correlate with its ability to mediate C-dependent lysis in vitro. Our results are consistent with the hypothesis that in vivo depletion of cells is mediated by opsonization and binding through the FcR.     

Alteration in H chain V region affects complement activation by chimeric antibodies

Chimeric antibodies to the synthetic polypeptide (Tyr, Glu)-Ala-Lys ((T,G)-A-L) were used to examine C activation by human IgG1. Two IgG1 antibodies, which contained mouse L chains and H chains with mouse V domains and human C domains, differed only in their VH domain. Ag binding and C activation by these antibodies were analyzed by ELISA. When limiting amounts of Ag were used in the assays, the antibodies required different quantities of Ag for optimal binding, suggesting that the antibodies bind to different epitopes on the (T,G)-A-L molecule. However, when competitive inhibition assays were performed with an optimal concentration of Ag, there were no differences in relative binding affinities for (T,G)-A-L or dissociation characteristics of the antibodies. C activation was examined at optimal Ag concentration to ensure equivalent binding of two IgG1 antibodies to Ag. After combination with immobilized Ag, these two antibodies bearing different V regions exhibited marked differences in the binding of C components C1q and C3d. When present in equal amounts in the assay, antibody 10B activated C and bound more C1q and C3d than antibody B11. These results indicate that V region differences can affect C activation by IgG.     

Comparison of rat and rat-mouse chimeric anti-murine CD4 antibodies in vitro. Chimeric antibodies lyse low-density CD4+ cells

GK1.5, a rat anti-mouse CD4 mAb, is effective in the treatment of several autoimmune syndromes, induces tolerance to co-administered Ag, and prolongs allograft survival. We have constructed a family of molecules with GK1.5 V regions and mouse gamma 1, gamma 2a, gamma 2b, or gamma 3 constant regions to investigate the mechanisms underlying the effectiveness of GK1.5. The rat-mouse chimeric antibodies are specific for murine CD4 and have identical binding curves as rat GK1.5 on CD4+ T cells. The chimeric GK1.5 gamma 2a, GK1.5 gamma 2b, and GK1.5 gamma 3 antibodies are more efficient than rat GK1.5 at C-mediated cytotoxicity. This is attributed to the enhanced capacity of the chimeric antibodies, compared to rat GK1.5, to lyse CD4+ cells with a low cell surface Ag density. This observation may have important implications for therapy.     

The immunogenicity of humanized and fully human antibodies

Monoclonal antibodies represent an attractive therapeutic tool as they are highly specific for their targets, convey effector functions and enjoy robust manufacturing procedures. Humanization of murine monoclonal antibodies has vastly improved their in vivo tolerability. Humanization, the replacement of mouse constant regions and V framework regions for human sequences, results in a significantly less immunogenic product. However, some humanized and even fully human sequence-derived antibody molecules still carry immunological risk. To more fully understand the immunologic potential of humanized and human antibodies, we analyzed CD4+ helper T cell epitopes in a set of eight humanized antibodies. The antibodies studied represented a number of different VH and VL family members carrying unique CDR regions. In spite of these differences, CD4+ T cell epitopes were found only in CDR-sequence containing regions. We were able to incorporate up to two amino acid modifications in a single epitope that reduced the immunogenic potential while retaining full biologic function. We propose that immunogenicity will always be present in some antibody molecules due to the nature of the antigen-specific combining sites. A consequence of this result is modifications to reduce immunogenicity will be centered on the affinity-determining regions. Modifications to CDR regions can be designed that reduce the immunogenic potential while maintaining the bioactivity of the antibody molecule.     

TP-PCR法构建抗人CD28嵌合抗体双启动子昆虫杆状病毒重组转移载体

为构建能同时表达抗人CD28嵌合重链和嵌合轻链基因的双启动子杆状病毒转移载体,利用PCR法扩增出抗人CD28鼠源性单抗的可变区基因和人IgG1的恒定区基因,再采用一种不需要限制性核酸内切酶和连接酶的新方法——三引物PCR（TP-PCR）法将两者拼接后分别插入杆状病毒转移载体pAcUW3的ph和p10启动子后,并通过酶切、电泳、PCR扩增和测序对重组体进行了鉴定。研究结果表明,TP-PCR法快速、方便、准确,无需设计外源的DNA序列,就能构建完好的融合表达基因。该转移载体的构建成功为嵌合抗体基因在昆虫细胞中的表达奠定了基础。     

Humanization of a mouse monoclonal antibody by CDR-grafting: the importance of framework residues on loop conformation.

A mouse monoclonal antibody (mAb 425) with therapeutic potential was 'humanized' in two ways. Firstly the mouse variable regions from mAb 425 were spliced onto human constant regions to create a chimeric 425 antibody. Secondly, the mouse complementarity-determining regions (CDRs) from mAb 425 were grafted into human variable regions, which were then joined to human constant regions, to create a reshaped human 425 antibody. Using a molecular model of the mouse mAb 425 variable regions, framework residues (FRs) that might be critical for antigen-binding were identified. To test the importance of these residues, nine versions of the reshaped human 425 heavy chain variable (VH) regions and two versions of the reshaped human 425 light chain variable (VL) regions were designed and constructed. The recombinant DNAs coding for the chimeric and reshaped human light and heavy chains were co-expressed transiently in COS cells. In antigen-binding assays and competition-binding assays, the reshaped human antibodies were compared with mouse 425 antibody and to chimeric 425 antibody. The different versions of 425-reshaped human antibody showed a wide range of avidities for antigen, indicating that substitutions at certain positions in the human FRs significantly influenced binding to antigen. Why certain individual FR residues influence antigen-binding is discussed. One version of reshaped human 425 antibody bound to antigen with an avidity approaching that of the mouse 425 antibody.     

Construction, expression and characterization of humanized antibodies directed against the human alpha/beta T cell receptor.

Completely humanized antibodies with specificity for the human alpha/beta TCR have been produced by genetic engineering. The L and H chain V region exons encoding the murine mAb BMA 031 CD regions and human EU framework regions were synthesized and replaced into previously isolated genomic fragments. These fragments were inserted into mammalian expression vectors containing the human kappa and gamma 1 C region exons. Two variants were constructed each containing selected BMA 031 amino acids within the human frameworks. The humanized genes were transfected into Sp2/0 hybridoma cells by electroporation and transfectomas secreting humanized antibody were isolated. Levels of antibody expression up to 7 pg/cell/24 h were obtained. The humanized antibody, BMA 031-EUCIV2, competed poorly with murine BMA 031 for binding to T cells. BMA 031-EUCIV3, however, bound specifically to T cells and competed effectively with both the murine BMA 031 antibody and a previously constructed chimeric BMA 031 antibody for binding to these cells. The relative affinity of BMA 031-EUCIV3 was about 2.5 times lower than BMA 031. The ability to promote antibody dependent cell-mediated cytolysis was significantly enhanced with the engineered antibodies as compared to murine BMA 031. Humanized BMA 031 is a clinically relevant, genetically engineered antibody with potential uses in transplantation, graft vs host disease, and autoimmunity.     

Development of new versions of anti-human CD34 monoclonal antibodies with potentially reduced immunogenicity

Despite the widespread clinical use of CD34 antibodies for the purification of human hematopoietic stem/progenitor cells, all the current anti-human CD34 monoclonal antibodies (mAbs) are murine, which have the potential to elicit human antimouse antibody (HAMA) immune response. In the present study, we developed three new mouse anti-human CD34 mAbs which, respectively, belonged to class I, class II and class III CD34 epitope antibodies. In an attempt to reduce the immunogenicity of these three murine mAbs, their chimeric antibodies, which consisted of mouse antibody variable regions fused genetically to human antibody constant regions, were constructed and characterized. The anti-CD34 chimeric antibodies were shown to possess affinity and specificity similar to that of their respective parental murine antibodies. Due to the potentially better safety profiles, these chimeric antibodies might become alternatives to mouse anti-CD34 antibodies routinely used for clinical application.     

The immunogenicity of humanized and fully human antibodies: Residual immunogenicity resides in the CDR regions

Monoclonal antibodies represent an attractive therapeutic tool as they are highly specific for their targets, convey effector functions and enjoy robust manufacturing procedures. Humanization of murine monoclonal antibodies has vastly improved their in vivo tolerability. Humanization, the replacement of mouse constant regions and V framework regions for human sequences, results in a significantly less immunogenic product. However, some humanized and even fully human sequence-derived antibody molecules still carry immunological risk. to more fully understand the immunologic potential of humanized and human antibodies, we analyzed CD4⁺ helper T cell epitopes in a set of eight humanized antibodies. the antibodies studied represented a number of different VH and VL family members carrying unique CDR regions. In spite of these differences, CD4⁺ T cell epitopes were found only in CDR-sequence containing regions. We were able to incorporate up to two amino acid modifications in a single epitope that reduced the immunogenic potential while retaining full biologic function. We propose that immunogenicity will always be present in some antibody molecules due to the nature of the antigen-specific combining sites. A consequence of this result is modifications to reduce immunogenicity will be centered on the affinity-determining regions. Modifications to CDR regions can be designed that reduce the immunogenic potential while maintaining the bioactivity of the antibody molecule.Key words: therapeutic, antibody, immunogenicity, deimmunizing, epitope     

Trimeric gp120-specific bovine monoclonal antibodies require cysteine and aromatic residues in CDRH3 for high affinity binding to HIV Env

We isolated HIV-1 Envelope (Env)-specific memory B cells from a cow that had developed high titer polyclonal immunoglobulin G (IgG) with broad neutralizing activity after a long duration vaccination with HIV-1_AD8 Env gp140 trimers. We cloned the bovine IgG matched heavy (H) and light (L) chain variable (V) genes from these memory B cells and constructed IgG monoclonal antibodies (mAbs) with either a human constant (C)-region/bovine V-region chimeric or fully bovine C and V regions. Among 42 selected Ig+ memory B cells, two mAbs (6A and 8C) showed high affinity binding to gp140 Env. Characterization of both the fully bovine and human chimeric isoforms of these two mAbs revealed them as highly type-specific and capable of binding only to soluble AD8 uncleaved gp140 trimers and covalently stabilized AD8 SOSIP gp140 cleaved trimers, but not monomeric gp120. Genomic sequence analysis of the V genes showed the third heavy complementarity-determining region (CDRH3) of 6A mAb was 21 amino acids in length while 8C CDRH3 was 14 amino acids long. The entire V heavy (VH) region was 27% and 25% diverged for 6A and 8C, respectively, from the best matched germline V genes available, and the CDRH3 regions of 6A and 8C were 47.62% and 78.57% somatically mutated, respectively, suggesting a high level of somatic hypermutation compared with CDRH3 of other species. Alanine mutagenesis of the VH genes of 6A and 8C, showed that CDRH3 cysteine and tryptophan amino acids were crucial for antigen binding. Therefore, these bovine vaccine-induced anti-HIV antibodies shared some of the notable structural features of elite human broadly neutralizing antibodies, such as CDRH3 size and somatic mutation during affinity-maturation. However, while the 6A and 8C mAbs inhibited soluble CD4 binding to gp140 Env, they did not recapitulate the neutralizing activity of the polyclonal antibodies against HIV infection.     

Construction of a humanized antibody to hepatitis B surface antigen by specificity-determining residues (SDR)-grafting and de-immunization

We previously constructed a humanized antibody, HuS10, by grafting the complementarity-determining regions (CDRs) of a parental murine monoclonal antibody into the homologous human antibody sequences. This process is termed CDR grafting. Some residues that were thought to affect the CDR loops and stabilize the structure of the variable regions were retained in the framework region. HuS10 exhibited in vivo virus-neutralizing activity, but its murine content had the potential to elicit immune responses in patients. In this study, to minimize the immunogenic potential of HuS10, we replaced 17 mouse residues in HuS10 with the comparable human residues using specificity-determining residue (SDR)-grafting and de-immunization methods. The resultant humanized antibody, HzS-III, had the same affinity and epitope specificity as HuS10 and had reduced immunogenic potential, as assessed by T-cell epitope analysis. Thus, SDR grafting in combination with de-immunization may be a useful strategy for minimizing the immunogenicity of humanized antibodies. In addition, HzS-III may be a good candidate for immunoprophylaxis of HBV infection.     

Characterization of murine and humanized anti-CD33, gelonin immunotoxins reactive against myeloid leukemias

M195 antibodies recognize CD33, an antigen present on acute myeloid leukemia blasts as well as some myeloid progenitor cells, but not on the ultimate hematopoietic progenitor stem cell. Immunotoxins (IT) reactive with human myeloid leukemias were constructed by conjugating gelonin, a single-chain ribosome-inactivating protein, to murine and genetically engineered, humanized M195 antibodies via anN-succinimidyl-3-(2-pyridyldithio)-propionate linkage. No losses of gelonin cytotoxic activity or M195 binding activity were observed after conjugation of up to two toxin molecules per antibody. Toxin conjugates displayed specific, potent toxicity for CD33⁺ cells. The murine and humanized IT were not toxic to CD33^– cells and were 600 and 4500 times more potent, respectively, than free gelonin in inhibiting CD33⁺ HL60 cells. Treatment of HL60 cells with 1 g/ml HuM195-gelonin resulted in more than 1000 times lower colony formation; normal bone marrow mononuclear cell colonyforming units treated with HuM195-IT were reduced by a factor of 10. HL60 leukemia cells could be effectively purged from an excess of normal bone marrow cells. Exposure of target cells to IT for as little as 30 min was as effective as continuous exposure of IT for up to 6 days. However, measures of the efficacy of the immunotoxin were directly related to the length of time of observation after IT exposure and were inversely related to cell concentration. M195-gelonin immunoconjugates are potential candidates for therapeutic use in in vivo or ex vivo bone marrow purging of myeloid leukemias.These studies were supported in part by the Lucille P. Markey Charitable Trust, ACS Grant No. IM551, NIH PO1CA33049, NIH RO1CA55349. Research conducted, in part, by the Clayton Foundation for Research. David A Scheinberg is a Lucille P. Markey Scholar     

Characterization and humanization of a monoclonal antibody that neutralizes human leukocyte interferon: a candidate therapeutic for IDDM and SLE

We have developed a panel of murine monoclonal antibodies that recognize human interferon alpha. One of these mononclonal antibodies binds and neutralizes, with high affinity, all of seven tested recombinant human interferon alphas. This mononclonal antibody also neutralizes the interferon activity present in two independent pools of interferon alphas prepared following stimulation of human peripheral blood leukocytes. The complementary determining regions from this murine mononclonal antibody were transferred to a human IgG2 heavy chain and to a human kappa1 light chain. In addition, six (heavy chain) and two (light chain) amino acids were transferred from the framework regions. This generated a humanized mononclonal antibody that retained the specificity of the mouse parent. The humanized anti-interferon alpha antibody is a candidate therapeutic for those diseases, such as insulin-dependent diabetes, systemic lupus erythematosis, psoriasis and Crohn's disease, which are all characterized by pathological expression of interferon alpha.     

Characterization of IgG FcR-mediated proliferation of human T cells induced by mouse and human anti-CD3 monoclonal antibodies. Identification of a functional polymorphism to human IgG2 anti-CD3.

T cell activation induced by mouse anti-CD3 mAb has shown to be dependent on the Ig isotype of these antibodies. A study of isotype dependency of human antibodies, however, seems more relevant to human effector systems, especially in view of the availability of humanized antibodies for clinical applications. We constructed a panel of mouse and mouse/human chimeric anti-CD3 mAb, which differ only in their CH region and hence have identical binding sites and affinity. By using these antibodies, we now studied their ability to induce T cell proliferation in human PBMC and analyzed the classes of IgG FcR involved in these responses. The human (h)IgG1, hIgG3, and hIgG4, as well as mouse (m)IgG2a and mIgG3 anti-CD3 mAb induced an Fc gamma RI (CD64)-dependent T cell proliferation in all donors. Activation with hIgG2 and mIgG1 anti-CD3 mAb was observed to be mediated via the low affinity Fc gamma RII (CD32). It was found that leukocytes in a normal donor population display a functional polymorphism with respect to hIgG2 anti-CD3 responsiveness. This polymorphism was found to be inversely related to the previously defined Fc gamma RII-polymorphism to mIgG1 anti-CD3 mAb. Monocytes expressing the Fc gamma RII mIgG1 low responder (LR) allele support hIgG2 anti-CD3 induced T cell proliferation efficiently, whereas cells homozygous for the Fc gamma RII mIgG1 high responder (HR) allele do not. This observation could be confirmed in T cell activation studies using hFc gamma RIIa-transfected mouse fibroblasts, expressing either the mIgG1 anti-CD3 HR or LR Fc gamma RII-encoding cDNA.     

Construction of humanized anti-ganglioside monoclonal antibodies with potent immune effector functions

 Gangliosides GD3, GD2 and GM2, which are the major gangliosides expressed on most human cancers of neuroectodermal and epithelial origin, have been focused on as effective targets for passive immunotherapy with monoclonal antibodies. We previously developed a chimeric anti-GD3 mAb, KM871, and a humanized anti-GM2 mAb, KM8969, which specifically bound to the respective antigen with high affinity and showed potent immune effector functions. Humanization of anti-ganglioside antibody is expected to enhance its use for human cancer therapy. In the present study, we generated a chimeric anti-GD2 mAb, KM1138, and further developed the humanized form of anti-GD2 and anti-GD3 mAbs by the complementarity-determining regions grafting method. The resultant humanized anti-GD2 mAb, KM8138, and anti-GD3 mAb, KM8871, showed binding affinity and specificity similar to those of their chimeric counterparts. In addition, both humanized mAbs had functional potency comparable to the chimeric mAbs in mediating the immune effector functions, consisting of antibody-dependent cellular cytotoxicity and complement-dependent cytotoxicity. The production of these humanized anti-ganglioside mAbs, with potent effector functions and low immunogenicity, precedes the evaluation of the therapeutic value of anti-ganglioside mAbs in passive immunotherapy and the target validation for ganglioside-based vaccine therapy. Received: 30 November 2000 / Accepted: 30 January 2001     

Antibody engineering by codon-based mutagenesis in a filamentous phage vector system.

A novel codon-based mutagenesis procedure is described that allows rapid and efficient modification of antibody amino acid sequences expressed as F(ab) fragments in M13. The procedure succeeded in generating a library of mutations in the complementarity-determining regions of chimeric L6, an antibody against a tumor-associated Ag. A set of anti-Id antibodies (anti-Id 1, 3, and 7) that bind near the L6 Ag-binding site served as model Ag. The goal was to select mutant antibody sequences that altered the L6 reactivity with the anti-Id in subtle ways, i.e., to eliminate the binding to one anti-Id while preserving other reactivities or to identify mutants with increased binding. A high frequency of variant M13 phage clones exhibiting altered specificity for the anti-Id were identified. Codon-based mutagenesis in conjunction with the M13 antibody expression and screening system should provide an efficient and general approach for redirecting the specificity and potentially improving the affinity of antibodies in vitro.