Production of a recombinant monoclonal antibody to Herpes Simplex Virus glycoprotein D for immunoaffinity purification of tagged proteins.

We have developed a stable Chinese Hamster Ovary (CHO) cell line for the production of a recombinant monoclonal antibody (mAb) to a short protein sequence derived from the N-terminus of human herpes simplex virus type 1 glycoprotein D (HSV-1 gD). The antibody (designated r34.1) provides a useful tool for the immunoaffinity purification of HSV-1 gD tagged proteins, and provides a generic purification system by which various proteins and peptides can be purified. Recombinant 34.1 was assembled using cDNA derived from a HSV-1 gD specific murine hybridoma engineered to encode a full-length IgG molecule. Antibody expression cassettes were transfected into CHO-S cells, and a stable cell-line expressing up to 500 mg/L of antibody, isolated. Affinity purified r34.1 exhibited nanomolar affinity for its cognate ligand, and is stable throughout multiple cycles of immunoaffinity purification involving ligand binding at neutral pH, followed by acid elution. The HSV-1 gD tag expression and purification strategy has been used to enhance the secretion and purification of several vaccine immunogens including HIV envelope protein rgp120s, but the protocol has potential for generic application.

A novel antibody-TCR (AbTCR) platform combines Fab-based antigen recognition with gamma/delta-TCR signaling to facilitate T-cell cytotoxicity with low cytokine release.

The clinical use of genetically modified T-cell therapies has led to unprecedented response rates in leukemia and lymphoma patients treated with anti-CD19 chimeric antigen receptor (CAR)-T. Despite this clinical success, FDA-approved T-cell therapies are currently limited to B-cell malignancies, and challenges remain with managing cytokine-related toxicities. We have designed a novel antibody-T-cell receptor (AbTCR) platform where we combined the Fab domain of an antibody with the γ and δ chains of the TCR as the effector domain. We demonstrate the ability of anti-CD19-AbTCR-T cells to trigger antigen-specific cytokine production, degranulation, and killing of CD19-positive cancer cells in vitro and in xenograft mouse models. By using the same anti-CD19 binding moiety on an AbTCR compared to a CAR platform, we demonstrate that AbTCR activates cytotoxic T-cell responses with a similar dose-response as CD28/CD3ζ CAR, yet does so with less cytokine release and results in T cells with a less exhausted phenotype. Moreover, in comparative studies with the clinically validated CD137 (4-1BB)-based CAR, CTL019, our anti-CD19-AbTCR shows less cytokine release and comparable tumor inhibition in a patient-derived xenograft leukemia model.

Targeting Alpha-Fetoprotein (AFP)-MHC Complex with CAR T-Cell Therapy for Liver Cancer.

PURPOSE: The majority of tumor-specific antigens are intracellular and/or secreted and therefore inaccessible by conventional chimeric antigen receptor (CAR) T-cell therapy. Given that all intracellular/secreted proteins are processed into peptides and presented by class I MHC on the surface of tumor cells, we used alpha-fetoprotein (AFP), a specific liver cancer marker, as an example to determine whether peptide-MHC complexes can be targets for CAR T-cell therapy against solid tumors. EXPERIMENTAL DESIGN: We generated a fully human chimeric antigen receptor, ET1402L1-CAR (AFP-CAR), with exquisite selectivity and specificity for the AFP158-166 peptide complexed with human leukocyte antigen (HLA)-A*02:01. RESULTS: We report that T cells expressing AFP-CAR selectively degranulated, released cytokines, and lysed liver cancer cells that were HLA-A*02:01+/AFP+ while sparing cells from multiple tissue types that were negative for either expressed proteins. In vivo, intratumoral injection of AFP-CAR T cells significantly regressed both Hep G2 and AFP158-expressing SK-HEP-1 tumors in SCID-Beige mice (n = 8 for each). Moreover, intravenous administration of AFP-CAR T cells in Hep G2 tumor-bearing NSG mice lead to rapid and profound tumor growth inhibition (n = 6). Finally, in an established intraperitoneal liver cancer xenograft model, AFP-CAR T cells showed robust antitumor activity (n = 6). CONCLUSIONS: This study demonstrates that CAR T-cell immunotherapy targeting intracellular/secreted solid tumor antigens can elicit a potent antitumor response. Our approach expands the spectrum of antigens available for redirected T-cell therapy against solid malignancies and offers a promising new avenue for liver cancer immunotherapy. Clin Cancer Res; 23(2); 478-88. ©2016 AACR.

Fragments of the V1/V2 domain of HIV-1 glycoprotein 120 engineered for improved binding to the broadly neutralizing PG9 antibody.

The V1/V2 domain of the HIV-1 envelope protein gp120 possesses two important epitopes: a glycan-dependent epitope recognized by the prototypic broadly neutralizing monoclonal antibody (bN-mAb), PG9, as well as an epitope recognized by non-neutralizing antibodies that has been associated with protection from HIV infection in the RV144 HIV vaccine trial. Because both of these epitopes are poorly immunogenic in the context of full length envelope proteins, immunization with properly folded and glycosylated fragments (scaffolds) represents a potential way to enhance the immune response to these specific epitopes. Previous studies showed that V1/V2 domain scaffolds could be produced from a few selected isolates, but not from many of the isolates that would be advantageous in a multivalent vaccine. In this paper, we used a protein engineering approach to improve the conformational stability and antibody binding activity of V1/V2 domain scaffolds from multiple diverse isolates, including several that were initially unable to bind the prototypic PG9 bN-mAb. Significantly, this effort required replicating both the correct glycan structure as well as the ß-sheet structure required for PG9 binding. Although scaffolds incorporating the glycans required for PG9 binding (e.g., mannose-5) can be produced using glycosylation inhibitors (e.g., swainsonine), or mutant cell lines (e.g. GnTI(-) 293 HEK), these are not practical for biopharmaceutical production of proteins intended for clinical trials. In this report, we describe engineered glycopeptide scaffolds from three different clades of HIV-1 that bind PG9 with high affinity when expressed in a wildtype cell line suitable for biopharmaceutical production. The mutations that improved PG9 binding to scaffolds produced in normal cells included amino acid positions outside of the antibody contact region designed to stabilize the ß-sheet and turn structures. The scaffolds produced address three major problems in HIV vaccine development: (1) improving antibody responses to poorly immunogenic epitopes in the V1/V2 domain; (2) eliminating antibody responses to highly immunogenic (decoy) epitopes outside the V1/V2 domain; and (3) enabling the production of V1/V2 scaffolds in a cell line suitable for biopharmaceutical production.

Characterization of a monoclonal antibody to a novel glycan-dependent epitope in the V1/V2 domain of the HIV-1 envelope protein, gp120.

Recent studies have described several broadly neutralizing monoclonal antibodies (bN-mAbs) that recognize glycan-dependent epitopes (GDEs) in the HIV-1 envelope protein, gp120. These were recovered from HIV-1 infected subjects, and several (e.g., PG9, PG16, CH01, CH03) target glycans in the first and second variable (V1/V2) domain of gp120. The V1/V2 domain is thought to play an important role in conformational masking, and antibodies to the V1/V2 domain were recently identified as the only immune response that correlated with protection in the RV144 HIV-1 vaccine trial. While the importance of antibodies to polymeric glycans is well established for vaccines targeting bacterial diseases, the importance of antibodies to glycans in vaccines targeting HIV has only recently been recognized. Antibodies to GDEs may be particularly significant in HIV vaccines based on gp120, where 50% of the molecular mass of the envelope protein is contributed by N-linked carbohydrate. However, few studies have reported antibodies to GDEs in humans or animals immunized with candidate HIV-1 vaccines. In this report, we describe the isolation of a mouse mAb, 4B6, after immunization with the extracellular domain of the HIV-1 envelope protein, gp140. Epitope mapping using glycopeptide fragments and in vitro mutagenesis showed that binding of this antibody depends on N-linked glycosylation at asparagine N130 (HXB2 numbering) in the gp120 V1/V2 domain. Our results demonstrate that, in addition to natural HIV-1 infection, immunization with recombinant proteins can elicit antibodies to the GDEs in the V1/V2 domain of gp120. Although little is known regarding conditions that favor antibody responses to GDEs, our studies demonstrate that these antibodies can arise from a short-term immunization regimen. Our results suggest that antibodies to GDEs are more common than previously suspected, and that further analysis of antibody responses to the HIV-1 envelope protein will lead to the discovery of additional antibodies to GDEs.

HIV-1 envelope proteins and V1/V2 domain scaffolds with mannose-5 to improve the magnitude and quality of protective antibody responses to HIV-1.

Two lines of investigation have highlighted the importance of antibodies to the V1/V2 domain of gp120 in providing protection from HIV-1 infection. First, the recent RV144 HIV-1 vaccine trial documented a correlation between non-neutralizing antibodies to the V2 domain and protection. Second, multiple broadly neutralizing monoclonal antibodies to the V1/V2 domain (e.g. PG9) have been isolated from rare infected individuals, termed elite neutralizers. Interestingly, the binding of both types of antibodies appears to depend on the same cluster of amino acids (positions 167­171) adjacent to the junction of the B and C strands of the four-stranded V1/V2 domain ß-sheet structure. However, the broadly neutralizing mAb, PG9, additionally depends on mannose-5 glycans at positions 156 and 160 for binding. Because the gp120 vaccine immunogens used in previous HIV-1 vaccine trials were enriched for complex sialic acid-containing glycans, and lacked the high mannose structures required for the binding of PG9-like mAbs, we wondered if these immunogens could be improved by limiting glycosylation to mannose-5 glycans. Here, we describe the PG9 binding activity of monomeric gp120s from multiple strains of HIV-1 produced with mannose-5 glycans. We also describe the properties of glycopeptide scaffolds from the V1/V2 domain also expressed with mannose-5 glycans. The V1/V2 scaffold from the A244 isolate was able to bind the PG9, CH01, and CH03 mAbs with high affinity provided that the proper glycans were present. We further show that immunization with A244 V1/V2 fragments alone, or in a prime/boost regimen with gp120, enhanced the antibody response to sequences in the V1/V2 domain associated with protection in the RV144 trial.

Glycoform and net charge heterogeneity in gp120 immunogens used in HIV vaccine trials.

BACKGROUND: The RV144 clinical trial showed for the first time that vaccination could provide modest but significant protection from HIV-1 infection. To understand the protective response, and to improve upon the vaccine's efficacy, it is important to define the structure of the immunogens used in the prime/boost regimen. Here we examined the heterogeneity in net charge, attributable to glycoform variation, of the gp120 immunogens contained in the AIDSVAX B/E vaccine. METHODOLOGY/PRINCIPAL FINDINGS: Isoelectric focusing and glycosidase digestion were used to assess variation in net charge of the gp120s contained in the AIDSVAX B/E vaccine used in the RV144 trial. We observed 16 variants of MN-rgp120 and 24 variants of A244-rgp120. Glycoform variation in gp120 produced in Chinese hamster ovary cells was compared to glycoform variation in gp120 produced in the 293F human embryonic kidney cell line, often used for neutralization assays. We found that gp120 variants produced in CHO cells were distinctly more acidic than gp120 variants produced in 293 cells. The effect of glycoform heterogeneity on antigenicity was assessed using monoclonal antibodies. The broadly neutralizing PG9 MAb bound to A244-rgp120, but not to MN-rgp120, whether produced in CHO or in 293. However, PG9 was able to bind with high affinity to MN-rgp120 and A244-rgp120 produced in 293 cells deficient in N-acetylglucosaminyltransferase I. CONCLUSIONS/SIGNIFICANCE: MN- and A244-rgp120 used in the RV144 trial exhibited extensive heterogeneity in net charge due to variation in sialic acid-containing glycoforms. These differences were cell line-dependent, affected the antigenicity of recombinant envelope proteins, and may affect assays used to measure neutralization. These studies, together with recent reports documenting broadly neutralizing antibodies directed against carbohydrate epitopes of gp120, suggest that glycoform variation is a key variable to be considered in the production and evaluation of subunit vaccines designed to prevent HIV infection.