Immune effectors required for the therapeutic activity of vorinostat.

Suberanilohydroxamic acid, a pan-histone deacetylase inhibitor (HDACi) best known as vorinostat (marketed under the name Zolinza® by Merck and Co., Inc.), was approved by the US Food and Drug Administration (FDA) for use in patients with cutaneous T-cell lymphoma (CTCL) in 2006.- Similar to other targeted anticancer agents, vorinostat has been developed based on the assumption that it would mediate direct cytotoxic effects on malignant cells. Indeed, vorinostat has been shown to trigger the apoptotic demise of cancer cells in vivo, in syngeneic models of murine adenocarcinoma and lymphoma as well as in mice bearing human acute lymphoblastic leukemia.

As for many other successful chemotherapeutics,, the antineoplastic activity of vorinostat does not originate from purely cancer cell-autonomous mechanisms, but is (at least partially) mediated by the immune system, as indicated by multiple observations. First, the antitumor effects of vorinostat can be improved by the co-administration of immunostimulatory antibodies targeting tumor necrosis factor receptor superfamily, member 4 (TNFRSF4, best known as OX40) or TNFRSF9 (best known as CD137),- as well as with α-galactosylceramide, a natural killer T (NKT) cell-activating agent that stimulates the production of interferon γ (IFNγ). Second, the antineoplastic activity of vorinostat against MC38 colon adenocarcinomas is lost in mice lacking immune effectors due to the combined knockout of recombination activating gene 2 (Rag2) and interleukin 2 receptor, γ chain (Il2rg, also known as γc). Along similar lines, the antitumor effects of vorinostat are abolished in Ifng hosts (which lack IFNγ) and in mice treated with IFNγ-neutralizing antibodies. Finally, malignant cells expressing a dominant-negative form of the IFNγ receptor also fail to respond to vorinostat.

Reminiscent of anthracyclines and oxaliplatin, which are immunogenic cell death (ICD) inducers,- vorinostat can stimulate a variant of apoptosis that is accompanied by all the major hallmarks of ICD, namely the exposure of calreticulin on the cell surface, the active secretion of ATP and the release of the non-histone chromatin-binding protein high mobility group box 1 (HMGB1). Accordingly, cancer cells succumbing to vorinostat are engulfed by dendritic cells in vitro. If injected subcutaneously into immunocompetent mice, vorinostat-treated cancer cells can induce an immune response that protects the animals against the inoculation of living cells of the same type.

In a transplantable mouse tumor model, namely, MC38 colon adenocarcinoma, the anticancer effects of vorinostat are abolished upon the injection of antibodies that deplete CD8+ T cells. Moreover, when tumors are implanted into mice lacking the gene coding for granzyme B, which is essential for the cytotoxic activity of T and NK cells, they fail to respond to vorinostat treatment. These observations suggest that the immunogenic demise of cancer cells induced by vorinostat elicits an antitumor immune response that is mediated by cytotoxic T lymphocytes. Thus, vorinostat resembles in its mode of action other ICD inducers, including anthracyclines and digoxin, which induce T cell-dependent but B cell-independent anticancer immune responses.-

However, vorinostat recently turned out to mediate therapeutic effects against B-cell lymphomas driven by the transgenic expression of v-myc avian myelocytomatosis viral oncogene homolog (Myc) in an unsuspected fashion. In this setting, the knockout of Rag2 or Ifng does abolish the therapeutic effects of vorinostat, yet the depletion of CD8+ T cells alone, CD4+ and CD8+ cells, NK cells alone, or all cytotoxic lymphocytes (NK and CD8+ T cells) fails to do so. Unexpectedly, in this model, the absence of B cells (due to the targeted disruption of the immunoglobulin µ chain-coding gene) fully abrogates the antineoplastic activity of vorinostat. These findings demonstrate that B cells (but neither T nor NK cells) play an essential role in the therapeutic response of lymphoma to vorinostat. At this stage, it is unknown whether tumor-specific antibodies and/or the production of IFNγ by B cells contribute to the therapeutic effects of vorinostat in this setting.

Vorinostat appears to exert immunomodulatory effects by acting on both cancer and immune cells. Schmudde and colleagues demonstrated that vorinostat sensitizes cancer cells to the cytotoxic effects of NK cells by promoting the expression of the NK-cell receptors killer cell lectin-like receptor subfamily K, member 1 (KLRK1, best known as NKG2D) and CD226 (also known as DNAM1). Conversely, Fiegler and collaborators showed that vorinostat limits the recognition of malignant cells by NK cells through the downregulation of natural killer cell cytotoxicity receptor 3 ligand 1 (NCR3LG1), a ligand for natural cytotoxicity triggering receptor 3 (NCR3) best known as B7-H6. Along similar lines, several groups reported that vorinostat exerts immunosuppressive side effects. In particular, vorinostat has been reported to stimulate the function of regulatory T cells,- to prevent the activation of tumor-specific NK and T cells,, and to stimulate the expression of the immunosuppressive enzyme indoleamine 2,3-dioxygenase 1 (IDO1) by dendritic cells. Although vorinostat can interfere with acute graft-vs.-host disease, it does not inhibit the graft-vs.-leukemia effects in mouse models, underscoring the complexity of its immunological effects., In the context of its clinical indication, CTCL, vorinostat has been found to inhibit the expression of the immunosuppressive cytokine interleukin-10 and to suppress the cytotoxic action of NK cells.

Which are the conclusions that we can draw from the aforementioned studies? First, vorinostat seems to differ from other anticancer agents that operate in a purely cancer cell-autonomous fashion, since its capacity to limit tumor growth fully depends on the immune system.,, Second, it appears that, at least in some tumor types (exemplified by Myc-driven lymphomas), B lymphocytes play a major positive role in the antineoplastic activity of vorinostat. This is in line with the fact that the accumulation of B cells within neoplastic lesions has a positive prognostic impact, at least in some cancers.-

Altogether, these results point to a surprising heterogeneity in the immune effectors mobilized by successful anticancer agents. After T and NK lymphocytes, B cells now enter the stage, asking for their share in the limelight of anticancer immunosurveillance.