AR-42

Reolysin and Histone Deacetylase Inhibition in the Treatment of Head and Neck Squamous Cell Carcinoma

Oncolytic viruses (OVs) are emerging as powerful anti-cancer agents and are currently being tested for their safety and effi- cacy in patients. Reovirus (Reolysin), a naturally occurring non-pathogenic, double-stranded RNA virus, has natural onco- lytic activity and is being tested in phase I–III clinical trials in a variety of tumor types. With its recent US Food and Drug Administration (FDA) orphan drug designation for several tu- mor types, Reolysin is a potential therapeutic agent for various cancers, including head and neck squamous cell carcinomas (HNSCCs), which have a 5-year survival of 55%. Histone de- acetylase inhibitors (HDACis) comprise a structurally diverse class of compounds with targeted anti-cancer effects. The first FDA-approved HDACi, vorinostat (suberoylanilide hydroxa- mic acid [SAHA]), is currently being tested in patients with head and neck cancer. Recent findings indicate that HDAC inhibition in myeloma cells results in the upregulation of the Reolysin entry receptor, junctional adhesion molecule 1 (JAM-1), facilitating reovirus infection and tumor cell killing both in vitro and in vivo. In this study, we tested the anti-tumor efficacy of HDAC inhibitors AR-42 or SAHA in conjunction with Reolysin in HNSCCs. While HDAC inhibition increased JAM-1 and reovirus entry, the impact of this combination therapy was tested on the development of anti-tumor immune responses.

INTRODUCTION
Oncolytic viruses (OVs) are emerging as potentially powerful anti- cancer agents; talimogene laherparepvec (T-VEC) was approved for treatment of unresectable metastatic melanoma, and numerous trials are currently testing the safety and efficacy of a variety of OVs in pa- tients.1,2 Reovirus is a naturally occurring non-pathogenic, double- stranded RNA virus that was isolated from human respiratory and gastrointestinal tracts and has been extensively studied in a similar fashion to T-VEC.3 Reolysin is a type 3 Dearing reovirus (Oncolytics Biotech) and is currently being tested in phase I–III clinical trials in a variety of tumor types.3 With its recent orphan drug designation from the US Food and Drug Administration (FDA) for ovarian, gastric, peritoneal, pancreatic, and brain cancers, Reolysin is a potential ther- apeutic agent for several types of cancer, including head and neck squamous cell carcinomas (HNSCCs). Despite aggressive treatments, the diagnosis of locally advanced head and neck cancer carries a dismal prognosis, with fewer than 55% of patients predicted to survive longer than 5 years.4 Thus, there is a clear need for novel therapies with activity against these tumors.

Histone deacetylase inhibitors (HDACis) comprise a structurally diverse class of compounds that are targeted anti-cancer agents.5 The first FDA-approved HDACi, vorinostat (suberoylanilide hy- droxamic acid [SAHA]), is highly effective in the treatment of cuta- neous T cell lymphoma.6 Similar to Reolysin, SAHA is also being investigated for safety and efficacy in patients with head and neck cancer and preliminary results are promising.7 HDAC inhibition in myeloma cells has recently been reported to upregulate the Reolysin entry receptor, junctional adhesion molecule 1 (JAM-1), and allows for greater Reolysin infection and killing both in vitro and in vivo in myeloma-bearing nude mice.8 Importantly, the prevalence of JAM-1 in various cancer types has yet to be thoroughly explored. Moreover, the effect of HDAC inhibition on JAM-1 in HNSCCs and its immunological impact in immune-competent mice remains to be elucidated.Since HDAC inhibition is a promising approach for head and neck cancers, we tested the impact of HNSCC tumor cell treatment with the HDAC inhibitors AR-42 or SAHA on Reolysin entry and tumor cell killing. Using both immune-deficient and immune-competent mice, we explored the effects of this therapeutic strategy on tumor/ host interactions and anti-tumor immune responses. With growing evidence of the significance of immune-mediated mechanisms in oncolytic viral therapy, we sought to characterize the impact of combining Reolysin and HDACis in the treatment of HNSCCs in an immune-competent model. Here, we observed that HDAC inhibi- tion resulted in the significant enhancement of Reolysin replication and anti-tumor efficacy in vitro and in vivo, with enhanced im- mune-mediated anti-tumor responses.

RESULTS
A previous study found a marked upregulation of the reovirus entry receptor, JAM-1, after treatment of myeloma cells with HDAC inhib- itors.8 To evaluate the effect of HDAC inhibition on reovirus suscep- tibility of patients with head and neck cancer, we tested the impact of treating SCC74A (human) and mouse tonsil epithelial (MTE) (mu- rine) squamous carcinoma cells with HDAC inhibitors (AR-42 or SAHA) on the reovirus entry receptor (JAM-1). Flow cytometry of treated cells revealed a significant increase in JAM-1 cell surface expression after treatment with AR-42 or SAHA (p < 0.001) (Fig- ure 1A). Reduced JAM-1 expression on the cell surface after reovirus infection and HDAC inhibitor treatment is consistent with receptor internalization after binding to reovirus, as indicated by western blot (WB) analysis for the reovirus capsid protein (s-NS) (Figure S1). These data indicate that JAM-1 is upregulated after HDACi treatment and that Reolysin results in receptor internalization in head and neck cancer cells.HDACi Enhances Reolysin Replication Consistent with the increased cell surface receptor, HDACi treatment markedly enhanced reovirus capsid protein (s-NS) after infection, suggesting an increase in viral entry or progeny (Figures 1A and S1). We therefore examined the effects of combinatorial treatment on Re- olysin viral replication. An increase in the levels of reovirus capsid pro- tein (represented by green fluorescence) was also observed in red fluo- rescent protein (RFP)-expressing SCC-74A cells treated with either AR-42 or SAHA (Figure 1B). Quantification of virus titers revealed a significant increase in viral replication (as measured in plaque-form- ing units [PFUs] in cells treated with AR-42 or SAHA; p < 0.0001) versus Reolysin treatment alone (Figure 1C). Collectively, these results indicate that HDAC inhibition of head and neck squamous carcinoma cells resulted in increased reovirus entry and replication.

Next, we determined the impact of combinatorial therapy on humanand murine head and neck squamous cancer cell killing. In the SCC-2 and SCC-74A human cancer cell lines, there was a significant and synergistic increase in tumor cell killing following HDACi plus Reo- lysin combination treatment, when compared to individual treatmentgroups (p % 0.003 in all cell lines tested) (Figures S1B, S1C, and 2A). The combination enhancement of head and neck tumor cell killing was also observed in a wide panel of cell lines (SCC-1, Cal27, SCC- 11, and SCC-47) (Figures S1B and S1C). Cell viability was measured at 48 hr after treatment with HDAC inhibitors (AR-42 or SAHA) and Reolysin with concentrations at 0.0625, 0.125. 0.25, 0.5, 1, 2, and 4 times their respective IC50 values as described.9 There was a synergis- tic interaction between AR-42 or SAHA when combined with Reoly- sin, indicative of a combination index (CI) value of less than one as determined by a Chou-Talalay analysis in both the SCC-2 and SCC-74A cell lines (Figure 2A). Consistent with increased virus repli- cation, human SCC74A tumor cells treated with the combination of Reolysin and HDACis displayed significantly more dead cells, as measured by flow cytometric analysis for Annexin and propidium io- dide (PI) dual-positive cells, when compared to each treatment alone (p < 0.001) (Figure 2B).

Moreover, a significant increase in overall cell killing (Figure 2C) and apoptotic cell death (Figure S2A) was also observed in the MTE murine squamous carcinoma cell line treated with the combination of HDAC inhibitors and Reolysin versus each individual treatment group alone (p % 0.01). Taken together, these findings illustrate the ability of HDACis to synergistically interact with Reolysin and enhance HNSCC killing in vitro.HDAC inhibition has also been shown to inhibit anti-viral immunity by blocking the cell-intrinsic type I interferon (IFN) responses.10–12 It is also known that reovirus individual therapy induces a robust immune response, with inflammatory cytokine production and the activation of both innate and adaptive immune responses in various tumor types.13–19 To further elucidate the role of HDACi and Reolysin combinatorial therapy on tumor cells, the pro-inflammatory cytokine expression of IFN-a and MCP-1 was assessed as previously described20 following 24 hr of treatment in both the SCC74A and MTE cell lines. There was enhanced IFN-a expression in both the SCC74A and MTE cancer cell lines following combination treatment when compared with monotherapies. Interestingly, we observed a sig- nificant enhancement of MCP-1 expression following combination HDACi and Reolysin therapy when compared to each individual treatment group (Figure S2B) (p < 0.05). Collectively, these data indicate that HDACi plus Reolysin combination therapy results in a pro-inflammatory tumor cell response.Based on in vitro findings, two murine tumor models were employedto assess the combinatorial efficacy of Reolysin and SAHA in vivo. In humans, HNSCC recurs both locally and at distant sites. Reolysin is a versatile oncolytic agent, as it can be used both intravenously or intra- tumorally, depending on the clinical scenario. It has been tested in clinical trials in both formats, with evidence of local effects for systemic therapy and systemic effects for local therapy.21–24

We chose intratumoral injections for our in vivo models to mirror the clinical scenario of recurrent locoregional disease that is best treated with intratumoral injections in the clinic. This is one of the most common scenarios for many HNSCC recurrences.To determine the direct anti-tumoral effects of treatment, female athymic nude mice bearing human HNSCC SCC74A xenografts were treated with DMSO, SAHA, Reolysin, or combination therapy. There was a significant reduction in the rate of tumor growth in com- bination-treated mice versus all other treatment groups (Figure 3A, left panel) (p < 0.01). Kaplan-Meir survival curves of mice treated with the SAHA plus Reolysin combination displayed a significant survival advantage over mice that received the individual treatment (p < 0.001) (Figure 3A, right panel). Immunohistochemical (IHC) H&E analysis of tumor sections at the time of euthanasia (when tumors reached 1,500 mm3 after the indicated treatment) revealed an obvious immune cell infiltrate in tumors from mice treated with both SAHA and Reolysin (Figure 3B, top panels). Further character- ization of these immune cells revealed a robust macrophage (CD68+) infiltrate (Figure 3B, bottom panels), as well as enhanced natural killer (NK) cell presence (natural cytotoxicity triggering receptor [NCR1]+) (Figure S3A, top panels).To assess the effects of combinatorial treatment on the immune response, immune-competent wild-type mice bearing MTE syngeneic squamous tumors were utilized. Mice treated with the SAHA plus Re- olysin combination had a significantly slower tumor growth rate over time (p < 0.0001) (Figure 4A, left panel) and a significant enhance- ment in murine survival when compared to each individual treatment (p % 0.01) (Figure 4A, right panel). To assess immune cell responses against the tumor, IHC analysis was performed. As with the SCC74A tumor model, a robust macrophage (CD68+) response was observed (Figure 4B, top panels), with an enhanced NK cell (NCR+) and CD4 T cell presence (Figure S3A, middle and bottom panels) in MTE tumors following HDACi plus Reolysin treatment. Interest- ingly, we also observed a marked increase in CD8 T cells following combinatorial therapy (Figure 4B, bottom panels).To assess MTE-specific immune responses, total splenocytes were harvested from three mice per group at 7 days after study recruitment and co-cultured ex vivo with MTE tumor cells. Splenocytes from mice treated with SAHA plus Reolysin exhibited significantly greater MTE tumor cell-killing capacity compared to each individual treatment group, as determined by live and dead flow cytometric analysis (p < 0.001) (Figure 4B). Phenotypic analysis of these splenoctyes did not reveal any differences in the immune cell subsets assessed, which included activated NK cells (CD49b+CD335+), myeloid- derived suppressor cells (CD11b+Gr1+), macrophages (CD11b+F4/80+), CD4+ T cells, CD8+ T cells, effector CD4+ or CD8+ T cells (CD4+ or CD8+ and CD3+CD44+CD62L+), or dendritic cells (CD11c+ and CD80+ or MHCII+) (Figure S3B). Data obtained from both immunodeficient and immunocompetent mice suggest that HDAC inhibition improved virus entry and also augmented anti- tumor immune responses in vivo.

DISCUSSION
Treatment of head and neck cancers with HDAC inhibitors and Re- olysin separately is currently being tested in patients.7 Understanding how the mechanisms of each therapeutic, or the combination thereof, can affect tumors and the anti-tumor immune response is critical to enhancing patient outcomes. Recent reports show that HDAC inhibi- tion in myeloma cells results in the upregulation of the Reolysin entry receptor, JAM-1, thus allowing for greater Reolysin infection and killing both in vitro and in vivo.8 Consistent with these findings, we observed JAM-1 expression and sensitivity to Reolysin killing in HNSCC tumor cells following treatment with the HDAC inhibitors AR-42 or SAHA. Using both immune-deficient and immune-compe- tent mice, as well as male and female mice, we explored the effects of this therapeutic strategy on the tumor as well as the anti-tumor immune response. HDAC inhibition resulted in a significant upregulation of the JAM-1 reovirus surface receptor on HNSCC cells. Moreover, this inhibition resulted in the significant enhancement of Reolysin replication and both in vitro and in vivo anti-tumor efficacy with enhanced anti-tumor immune responses.In accordance with the epigenetic regulation of JAM-1 and its induction following treatment with HDAC inhibition in multiple myeloma,8 we also observed JAM-1 receptor upregulation and in- creases in reovirus replication in HNSCCs. The synergistic killing activity and pro-inflammatory responses following HDAC inhibition in combination with Reolysin highlight the therapeutic potential of combination therapy with the induction of potentially immuno- genic25 cell death in both human and murine cell lines. While apoptotic cell death is often dysregulated in cancer cells,26,27 combi- nation therapy resulted in a robust and significant increase in tumor cell apoptosis.

Although HDACi-induced cell death can be highly immuno- genic,28,29 HDAC inhibition has been shown to inhibit anti-viral im- munity via blocking both cell-intrinsic type 1 IFN responses as well as NK cell function.10–12,30 HDAC inhibition is also thought to have a negative impact on antigen-presenting cells as well as T lympho- cytes,31 suggesting that it could potentially impact the develop- ment of anti-tumor immunity. On the contrary, our findings using SAHA did not reveal an inhibition of either IFN responses in tumor cells or T cell ability to traffic to tumors. Moreover, enhanced IFN responses and robust T cell infiltration were observed following combinatorial therapy. An elevated macrophage infiltrate also sug- gests enhanced antigen presentation, but this has not yet been fully explored.Reovirus individual therapy induces a robust immune response, with inflammatory cytokine production and the activation of both innate and adaptive immune responses in various tumor types.13–19 This has also been corroborated in patients, wherein treatment with Reolysin resulted in the activation of both NK and cytotoxic T lymphocytes.32 To determine the impact of this therapeutic strategy on mounting a successful anti-tumor immune response against HNSCCs, we utilized both immune- deficient and immune-competent murine tumor models. In female athymic nude mice lacking an intact T cell response, a significant in- crease in the survival of combination-treated mice was observed. IHC analysis of tumor sec- tions from mice treated with SAHA and Reoly- sin also revealed a robust immune cell infiltrate, specifically macrophages, when compared to either agent alone. These data indicate that HDACi and reovirus treatment results in enhanced innate immune responses, even in the absence of adaptive immunity. Evaluation of the anti-tumor efficacy of this combination in a syngeneic mouse model of squamous carci-
noma also revealed a significant macrophage and CD8 T cell tumor infiltration and enhancement of murine survival. This was accompa- nied by enhanced ex vivo anti-tumor “memory” responses, as elicited by a significant increase in tumor cell killing by splenocytes derived from animals treated with both agents. These data indicate a substan- tial therapeutic benefit following combination treatment for tumor- specific (most likely) T cell responses in both male and female mice. Taken together, combinatorial therapy enhanced both innate and adaptive immunity when compared to individual treatment groups.

Collectively our data demonstrate that treatment of Reolysin-infected animals with HDAC inhibitors increased both reovirus cytotoxic effects and anti-tumor immunity. Future studies will focus on ascertaining the specific cell type(s) that are responsible for the ther- apeutic efficacy of combination therapy. Moreover, because murine studies were conducted in both male and female mice using a treat- ment schema that mirrors those used in patients, clinical studies may provide novel correlates to enhance ongoing and future clinical trials. This study provides the rationale for further investigation of this combination strategy in patients with HNSCCs.Six human squamous cell carcinoma (SCC) cell lines (UM-SCC-74A, UM-SCC-1, UD-SCC-2, UM-SCC-11A, UM-SCC-47, and ATCC Cal27) as well as murine (MTE) squamous carcinoma cells (shPTP- BL-Ras) and murine fibroblasts (L929) were cultured as previously described.33,34 All cells were propagated in DMEM (Life Technolo- gies) supplemented with penicillin (100 U/mL), streptomycin (100 mg/mL), and 10% fetal bovine serum (FBS). Reolysin was kindly provided by Oncolytics Biotech. AR-42 was a gift from Arno Thera- peutics and SAHA was purchased from Selleckchem.All flow cytometric analyses were conducting using a Becton Dickin- son fluorescence-activated cell sorter (FACS) LSRII and analyzed using FloJo software as previously described.9,35 JAM-1 surface levels were assessed using an anti-human or anti-mouse JAM-1 antibody (BD Biosciences Pharmingen) as compared to their respective isotype controls. The relative mean fluorescence intensity (MFI) was used to determine JAM-1 receptor surface levels. Cellular apoptosis was assessed using Annexin V-450 and propidium iodide staining (BD Biosciences Pharmingen) in accordance with the manufacturer’s in- structions. The percentage of dead cells was quantified using a Live/ Dead Fixable Dead Cell Stain Kit (Invitrogen) per the manufacturer’s instructions. The following antibodies were utilized for immune cell phenotypic analysis: CD49b-V450, F4/80-BV421, CD8-V450, CD44-PeCy7, CD62L-allophycocyanin (APC)-Cy7, CD11c-APC- Cy7, CD80-V450 (BD Biosciences Pharmingen), CD335-APC, CD11b-APC-Vio770, Gr1-APC, CD3-APC, CD4-fluorescein isothio- cyanate (FITC), and major histocompatibility complex class II AR-42 (MHCII)-FITC (Miltenyi Biotec).