Herpesvirus Entry Mediator As a Novel Therapeutic Target in Multiple Myeloma

The use of immune checkpoint inhibitors has markedly improved the survival outcomes of many patients with solid tumors. However, in multiple myeloma (MM), the use of such inhibitors has failed to achieve survival benefit. We previously showed that oncolytic herpes simplex virus 1 (oHSV-1) can effectively infect MM cells and induce apoptosis through the receptors for oHSV-1, NECTIN-1 and HVEM (herpes virus entry mediator). HVEM, a member of the TNF receptor superfamily (TNFRSF14), is an immune checkpoint present on T cells, B cells, NK cells and macrophages. Depending on HVEM's interaction with its major inhibitory ligand, B- and T-lymphocyte attenuator (BTLA) or its major activating ligand, tumor necrosis factor superfamily member 14 (TNFSF14, LIGHT), it is pivotal in downmodulating or stimulating the immune cells. Accordingly, targeting HVEM to block its interaction with BTLA is an area of therapeutic interest.

Flow cytometry analysis showed that HVEM was consistently expressed on the surface of all tested MM cell lines (n=9). Mass cytometry analysis of bone marrow (BM) from multi-relapsing MM patients revealed that HVEM was also expressed in different immune populations and that its expression was increased in T lymphocytes upon relapse from the anti-CD38 antibody daratumumab. Loss of function experiments by silencing RNA (siRNA) showed that, after 48hrs, HVEM knockdown significantly reduced the proliferation of the MM cell lines MM.1S (48% decrease vs control si-RNA; P˂0.0001) and H929 (35% decrease vs control si-RNA; P˂0.0001), as indicated by the MTS assay. To this end, we developed several anti-HVEM monoclonal antibodies directed against the extracellular domain of human HVEM and used surface plasmon resonance experiments and CyTOF to validate their binding. Among all antibodies tested, HVEM5 had one of the highest binding affinities, with minimal cross-reactivity with other non-cancer tissues as assessed on tissue array by imaging mass cytometry.

An ELISA-based BTLA:HVEM inhibitor screening assay showed that HVEM5, but not a non-binding control (HVEM4), resulted in a 100% blockade of the interaction of HVEM with recombinant BTLA protein. Moreover, HVEM5 elicited potent antibody-dependent cellular cytotoxicity by primary NK cells against MM cell lines (MM.1S and H929), in contrast to a control human IgG (p=0.0022 and 0.0061, respectively), and was comparable to that of daratumumab (Fig.1A). Flow cytometry data showed that ex vivo HVEM5 treatment of tumor-associated macrophages (TAMs) polarized them from a M2/repair-type to a M1/kill-type, as indicated by their decreased CD163 (69% increase vs. control IgG; P=0.0011; n=4) and increased CD80 (30% increase vs control IgG; P=0.0436; n=4) expression. Furthermore, HVEM5 (10 µg/mL) treatment of peripheral blood mononuclear cells almost doubled CD8+ T lymphocytes proliferation ex vivo.

To assess whether HVEM5 exerts an agonistic effect and activates HVEM-NF-κB downstream activation signaling, we used a luciferase reporter assay in a Jurkat cell line overexpressing HVEM. HVEM5 binding induced a 1.5-fold increase in NF-κB promoter activity (1.50±0.07) compared to a control human IgG (1.02±0.02) or the non-binding antibody HVEM4 (1.09±0.07) (p=0.0005). This increase was equivalent to that observed using the HVEM immune activator, LIGHT (Fig.1B). In line with these results, immunofluorescence staining of primary monocytes showed that HVEM5 treatment increased the nuclear translocation of the NF-kB p65 transcriptional subunit, which was greater than that of LIGHT-treated cells (Fig.1B). Furthermore, HVEM5 generated with a mouse-reactive Fc fragment induced a significantly longer overall survival (median 22 days) and tumor reduction in an NCr nude mouse model when injected intravenously twice a week (3 mg/kg, Mon-Fri), whereas the median survival was 15 days in the IgG-treated control animals (p=0.01) (Fig.1A; right trace). Further in vivo studies in HVEM/BTLA double humanized immune competent animal models and in HVEM knock-out mouse models are ongoing, and the results will be presented at the meeting. To our knowledge, the present work is the first to report that MM cells ubiquitously express HVEM, which could be exploited as a novel therapeutic opportunity to not only target it as an immune checkpoint, but also as a direct MM-targeting strategy.