Restricted variable region gene usage and possible rheumatoid factor relationship among human monoclonal antibodies specific for the AD-1 epitope on cytomegalovirus glycoprotein B

Chimeric Antigen Receptors Targeting Human Cytomegalovirus

Human cytomegalovirus (CMV) is a ubiquitous pathogen that causes significant morbidity in some vulnerable populations. Individualized adoptive transfer of ex vivo expanded CMV-specific CD8+ T cells has provided proof-of-concept that immunotherapy can be highly effective, but a chimeric antigen receptor (CAR) approach would provide a feasible method for broad application. We created 8 novel CARs using anti-CMV neutralizing antibody sequences, which were transduced via lentiviral vector into primary CD8+ T cells. All CARs were expressed. Activity against CMV-infected target cells was assessed by release of cytokines (interferon-γ and tumor necrosis factor-α), upregulation of surface CD107a, proliferation, cytolysis of infected cells, and suppression of viral replication. While some CARs showed varying functional activity across these assays, 1 CAR based on antibody 21E9 was consistently superior in all measures. These results support development of a CMV-specific CAR for therapeutic use against CMV and potentially other applications harnessing CMV-driven immunotherapies.

The human IgE-encoding transcriptome to assess antibody repertoires and repertoire evolution

Upon encounter with antigen, the B lymphocyte population responds by producing a diverse set of antigen-specific antibodies of various isotypes. The vast size of the responding populations makes it very difficult to study clonal evolution and repertoire composition occurring during these processes in humans. Here, we have explored an approach utilizing the H-EPSILON-encoding transcriptome to investigate aspects of repertoire diversity during the season of antigen exposure. We show through sequencing of randomly picked transcripts that the sizes of patients' repertoires are relatively small. This specific aspect of the transcriptome allows us to construct evolutionary trees pinpointing features of somatic hypermutation as it occurs in humans. Despite the small size of the repertoires, they are highly diverse with respect to VDJ gene usage, suggesting that the H-EPSILON-encoding transcriptome is a faithful mimic of other class-switched isotypes. Importantly, it is possible to use antibody library and selection technologies to define the specificity of clonotypes identified by random sequencing. The small size of the H-EPSILON-encoding transcriptome of peripheral blood B cells, the simple identification of clonally related sets of genes in this population, and the power of library and selection technologies ensure that this approach will allow us to investigate antibody evolution in human B lymphocytes of known specificity. As H-EPSILON repertoires show many of the hallmarks of repertoires encoding other isotypes, we suggest that studies of this type will have an impact on our understanding of human antibody evolution even beyond that occurring in the IgE-producing B cell population.

Acquisition of potential N-glycosylation sites in the immunoglobulin variable region by somatic mutation is a distinctive feature of follicular lymphoma

Most patients with follicular lymphoma (FL) have somatically mutated V genes with intraclonal variation, consistent with location in the germinal center site. Using our own and published sequences, we have investigated the frequency of potential N-glycosylation sites introduced into functional V(H) genes as a consequence of somatic mutation. FL cells were compared with normal memory B cells or plasma cells matched for similar levels of mutation. Strikingly, novel sites were detected in 55 of 70 (79%) patients with FL, compared to 7 of 75 (9%) in the normal B-cell population (P <.001). Diffuse large B-cell lymphoma (DLCL) showed an intermediate frequency (13 of 32 [41%] patients). Myeloma and the mutated subset of chronic lymphocytic leukemia showed frequencies similar to those of normal cells in 5 of 64 (8%) patients and 5 of 40 (13%) patients, respectively. In 3 of 3 random patients with FL, immunoglobulin was expressed as recombinant single-chain Fv in Pichia pastoris, and glycosylation was demonstrated. These findings indicate that N-glycosylation of the variable region may be common in FL and in a subset of DLCL. Most novel sites are located in the complementarity-determining regions. V(H) sequences of nonfunctional V(H) genes contained few sites, arguing for positive selection in FL. One possibility is that the added carbohydrate in the variable region contributes to interaction with elements in the germinal center environment. This common feature of FL may be critical for tumor behavior.

Antibody-mediated neutralization of cytomegalovirus: modulation of efficacy induced through the IgG constant region

Antibodies can neutralize the infectious properties of human cytomegalovirus (CMV). In vivo, the major neutralization determinants are located on glycoprotein B (gB). Recombinant human antibodies, that carry different constant regions (IgG1, IgG3 and the synthetic variant IgG3mA) against two of these epitopes were investigated for their ability to recruit the complement cascade for destruction of the virus. It was shown that all variants of an antibody against the antigenic domain (AD)-2 epitope displayed a similar neutralization activity despite the fact that improved C1q binding was observed for IgG3 and IgG3mA over the IgG1 variant. In contrast, an antibody against the AD-1 epitope carrying the normal IgG3 constant region, was less efficient than its IgG1 counterpart in neutralizing the virus in the absence of complement. However, it restored its activity in the presence of complement to the level of the naturally occurring IgG1 version. The same antibody was substantially more potent in neutralizing the virus in the presence of complement if it carried the IgG3mA constant region. This demonstrates the importance of the constant domain for the biological activity of AD-1 specific antibodies, a factor that should be taken into account when using antibody-based therapeutics or when inducing antibodies by vaccination.

Lack of autoantibody production associated with cytomegalovirus infection

To confirm an association between cytomegalovirus (CMV) infection and the presence of antibodies to Smith (Sm), to ribonucleoprotein (RNP), and to a component of the U1 ribonucleoproteins (U1-70 kD), we measured antibodies to these protein antigens using an enzyme immunoassay and an immunoblot. The antibodies were measured in the sera of 80 healthy subjects, one-half of whom were naturally CMV seropositive and one-half were CMV seronegative, and in eight subjects immunized with a live attenuated strain of CMV. None of the vaccinees developed antibodies to Sm, to RNP, or to U1-70 kD at either 4 or 12 months after immunization. Additionally, there was no statistically significant association between levels of antibodies to Sm or to RNP and between sera obtained from vaccinees, natural CMV seropositive individuals, and CMV seronegative individuals. One CMV seropositive serum and one CMV seronegative serum tested positive for antibodies to U1-70 kD. These data indicate that neither wild-type infection nor the live-attenuated Towne vaccine frequently induce autoantibody production.

A central core structure in an antibody variable domain determines antigen specificity

Antibody binding sites provide an adaptable surface capable of interacting with essentially any molecular target. Using CDR shuffling, residues important for the assembly of mucin-1 specific paratopes were defined by random recombination of the complementarity determining regions derived from a set of mucin-1 specific clones, previously selected from an antibody fragment library. It was found that positions 33 and 50 in the heavy chain and 32, 34, 90, 91 and 96 in the light chain were conserved in many of the clones. These particular residues seem to be located centrally in the binding site as indicated by a structure model analysis. The importance of several of these conserved residues was supported by their presence in a mouse monoclonal antibody with a known structure and the same epitope specificity. Several of these corresponding residues in the mouse monoclonal antibody are known to interact with the antigen. In conclusion, critical residues important for maintaining a human antigen-specific binding site during the process of in vitro antibody evolution were defined. Furthermore, an explanation for the observed restricted germline gene usage in certain antibody responses against protein epitopes is provided.

Recombinant cytomegalovirus glycoprotein gB (UL55) induces an autoantibody response to the U1-70 kDa small nuclear ribonucleoprotein

Human cytomegalovirus (CMV) infection can be life threatening in the immune compromised and is associated with congenital defects and / or mental retardation in the neonate. The demonstrated association between CMV infection and rheumatoid factor (RF) raised the possibility of an induction of an autoimmune response upon vaccination with a candidate CMV vaccine, glycoprotein gB (UL55). The antibody responses generated after injections of an adenovirus-gB construct (Ad-gB) were studied in autoimmune-prone (MRL/mpj) and normal (BALB.k, C3H, and BALB/c) mice. Enzyme-linked immunosorbent assay and immunoblot analyses were done to identify the autoantibodies produced following immunization. Immunization with Ad-gB induced a significant IgG anti-viral response in all strains tested (p < 0.0001) compared to phosphate-buffered saline or HeLa controls. Ad-gB induced a significant IgG autoantibody response (p > 0.005) to the U1-70 kDa spliceosome protein in both autoimmune and normal strains whereas immunization with recombinant human La/SS-B did not. Autoantibodies to U1-70 kDa are part of the anti-ribonucleoprotein response seen in systemic lupus erythematosus and mixed connective tissue disease. Low levels of IgG RF and anti-double-stranded DNA antibodies were also induced. This study raises concern that immunization with CMV gB in individuals genetically predisposed to autoimmunity could trigger the development or acceleration of an autoimmune disease.

Structural Basis of the gp120 Superantigen-Binding Site on Human Immunoglobulins

B cell superantigens (SAg) interact with normal human nonimmune Igs (Igs), independently of the light chain isotype, and activate a large proportion of the B cell repertoire. Recently, the major envelope protein of HIV-1, gp120, was found to exhibit SAg-like properties for B cells with potential pathologic consequences for the infected host. This unconventional mode of interaction contrasts with its binding to immunization-induced Abs, which requires the tertiary structure of the heavy and light chain variable regions. In this report, we have examined the structural basis of the interaction between human Igs and gp120. We found that gp120 binding is restricted to Igs from the VH3 gene family and that the two VH genes 3-23 and 3-30, known to be overutilized during all stages of B cell development, frequently impart gp120 binding. We also provide evidence that the viral gp120 SAg can interact with only a subset of the human VH3+ Igs that can convey binding to the prototypic bacterial B cell SAg protein A from Staphylococcus aureus. Finally, we have identified amino acid positions present primarily in the first and third framework regions of the Ig heavy chain variable region, outside the conventional hypervariable loops, which correlate with gp120 binding. In a three-dimensional sequence-homology model, these residues partially overlap with the predicted SAg protein A binding site for VH3+ Igs.

Insertions and deletions in hypervariable loops of antibody heavy chains contribute to molecular diversity

Antibody diversity, a molecular feature which allows these proteins to specifically interact with a diverse set of targets, is created at the genetic level by a variety of means. These include germline gene segment recombination, junctional diversity and single basepair (bp) substitution. We here demonstrate that a human high affinity antibody specific for an exogenous protein antigen carry three amino acid residues immediately adjacent to the first hypervariable loop of the heavy chain. These additional residues are shown not to be encoded by the germline repertoire. We also describe the characteristics of insertions and deletions, not found in any known germline sequence, within the first and second hypervariable loops of other previously described antibody-encoding genes. These findings demonstrate that insertions or deletions of entire codons provide yet another approach by which the human antibody repertoire is diversified in vivo. Since these major genetic modifications occur within or immediately adjacent to loops contributing to the antigen-binding site, they are likely to affect the binding properties of the mutated antibodies.

Characteristics of human antibody repertoires following active immune responses in vivo

Possibilities to develop human monoclonal antibody specificities have recently been much facilitated by improvements of human hybridoma technology but even more so by the emerging phage-display technique. However, until recently very little has been known about the characteristics at the molecular level of the induced, T cell-dependent human antibody response, frequently targeted by these techniques. Rather, the major part of available sequence information has been related to tumor-derived or autoreactive antibodies. We have now investigated high affinity, monospecific, human antibody repertoires as developed by hybridoma technology. The VH region gene usage among such in vivo-induced repertoires is in only some respects similar to that found in the total B cell population. A limited number of heavy-chain variable segment loci account for the majority of all induced antibodies. A particular VH gene locus (4-34) frequently employed by peripheral B cells and associated with autoreactive antibodies was rarely used by the induced repertoire. Furthermore, in particular antigen systems, V region usage differs from the total available repertoire, and heavy-chain CDR3 is generally longer among antibodies induced against foreign protein antigens than in the average B cell population. Light-chain gene usage is often restricted to just a few dominant genes frequently found among B cells in general. In comparison, variable regions derived by phage-display technology in some antigen systems display even longer heavy-chain CDR3 than hybridoma-derived antibodies. This technique also appears to select a different set of germline genes preferentially (both with respect to VH and JH) as compared to hybridoma technology. In summary, the T cell-dependent antibody response against foreign antigens appears to differ from the average circulating B cell in several ways, and thus does not seem to represent a random selection of the available repertoire.

Quantitative analysis of multiple V-region interactions among normal human IgG

There is to date no quantitative method for scoring putative V-region interactions among serum antibodies, and the available qualitative techniques are not amenable to routine utilization. This deficit may explain the paucity of observations on the characteristics of the immune network, in contrast with the multiplicity of phenomenological descriptions on idiotype regulation. We describe here a novel methodology that uses isoelectric focusing (IEF) to resolve human F(ab')2 preparations from large pools of normal serum IgG into multiple bands, and computer-aided data processing to analyze interactions between the resulting blotted proteins and normal serum IgG from individual donors. Our results show that in all normal human sera tested, there are IgG-mediated interactions with a large number of IEF fractions of human F(ab')2. These interactions are V region specific, as assessed by inhibition experiments and by lack of binding of IgG monoclonal antibodies, and are characterized by average affinities that are in the micromolar range, as measured by surface plasmon resonance.

Light chain shuffling of a high affinity antibody results in a drift in epitope recognition

Human polyclonal and monoclonal antibodies against pathogens and toxins are potentially useful in the treatment of various diseases. A number of human monoclonal antibodies with protective capacity in vitro have been established by conventional hybridoma technology. However, with the development of phage-display technology, the possibility of specifically tailoring antigen-binding properties has improved substantially. We show here that the reactivity of a high affinity, virus-neutralizing human antibody against the AD-2 epitope of cytomegalovirus gB can be modified by introducing other Vkappa sequences together with the original VH sequence. The fine specificity, as determined by the requirement of particular amino acid residues in the epitope, is shifted in these new antibody fragments. It was also evident that the VH/Vkappa pairing was not promiscuous, since antibody fragments selected by phage display retained light chain sequences very similar to the original hybridoma-derived light chain, proving that a high affinity interaction was very dependent on a co-operativity between both variable domains. These findings show that phage display technology might modify the binding properties of pre-existing, high affinity antibodies.

Human monoclonal antibodies to cytomegalovirus. Characterization and recombinant expression of a glycoprotein-B-specific antibody

Human monoclonal antibodies (mAb) to human cytomegalovirus (HCMV) were established from spleen cells of a HCMV-positive donor. The antibodies (gamma 3, lambda) secreted from a stable heterohybridoma cell line were further characterized by immunoprecipitation and immune-fluorescence microscopy using HCMV infected cells and recombinant cell lines expressing HCMV glycoprotein B. The antibody reacted with the entire glycoprotein B or the extracellular domain expressed as glycoprotein-B--beta-galactosidase fusion protein in the native state, but the antibody was not neutralizing HCMV. Denatured and reduced forms of glycoprotein B were not recognized by this antibody, however, native glycoprotein B on the surface of infected cells was detected efficiently. The genes encoding the Fab part of the antibody were cloned and expressed in Escherichia coli. Recombinant Fab fragments specifically binding the extracellular domain of glycoprotein B could easily be isolated from the periplasmic space. Recombinant antibodies provide the opportunity to modify effector functions and to add tags to diagnostic antibodies for more efficient detection of CMV-infected cells.

Molecular characterization of human monoclonal antibodies specific for several HIV proteins: analysis of the VH3 family expression

We have analyzed the heavy-chain variable (VH) region genes expressed by a panel of human monoclonal antibodies derived from an immunized volunteer, an AIDS patient and seropositive asymptomatic donors, and specific for HIV-1 env, pol and gag gene products. The third complementarity-determining regions show a high complexity with unconventional gene recombination events. Most of the VH genes utilized are also frequently encountered in other immune responses. Their sequences are, in general, typical of an antigen-driven immune response. Molecular mechanisms that generate high-affinity antibodies are then effective during HIV infection. Remarkably, VH3 family, which dominates the human antibody repertoire, is barely encountered among anti-HIV antibodies.