Infections with DNA tumor viruses, including members of the polyomavirus family, often result in tumor formation in immune-deficient hosts. The complex control involved in antiviral and antitumor immune responses during these infections can be studied in murine polyomavirus (PyV)-infected mice as a model. We found that NK cells efficiently kill cells derived from PyV-induced salivary gland tumors in vitro in an NKG2D (effector cell)-RAE-1 (target cell)-dependent manner; but in T cell-deficient mice, NK cells only delay but do not prevent the development of PyV-induced tumors. In this article, we show that the PyV-induced tumors have infiltrating functional NK cells. The freshly removed tumors, however, lack surface RAE-1 expression, and the tumor tissues produce soluble factors that downregulate RAE-1. These factors include the proinflammatory cytokines IL-1alpha, IL-1beta, IL-33, and TNF. Each of these cytokines downregulates RAE-1 expression and susceptibility to NK cell-mediated cytotoxicity. CD11b(+)F4/80(+) macrophages infiltrating the PyV-induced tumors produce high amounts of IL-1beta and TNF. Thus, our data suggest a new mechanism whereby inflammatory cytokines generated in the tumor environment lead to evasion of NK cell-mediated control of virus-induced tumors.

Vorinostat, an oral histone deacetylase inhibitor with antitumor activity, is in clinical trials for hematologic and solid tumors that metastasize and compromise bone structure. Consequently, there is a requirement to establish the effects of vorinostat on tumor growth within bone. Breast (MDA-231) and prostate (PC3) cancer cells were injected into tibias of SCID/NCr mice and the effects of vorinostat on tumor growth and osteolytic disease were assessed by radiography, micro-computed tomography, and histologic and molecular analyses. Vorinostat-treated and control mice without tumors were also examined. Tumor growth in bone was reduced approximately 33% by vorinostat with inhibited osteolysis in the first few weeks of the experiment. However, osteolysis became more severe in both the vehicle and vorinostat-treated groups. Vorinostat increased the expression of tumor-derived factors promoting bone resorption, including PTHrP, IL-8, and osteopontin. After 4 weeks of vorinostat therapy, the non-tumor-bearing contralateral femurs and limbs from vorinostat-treated tumor-free SCID mice showed significant bone loss (50% volume density of controls). Thus, our studies indicate that vorinostat effectively inhibits tumor growth in bone, but has a negative systemic effect reducing normal trabecular bone mass. Vorinostat treatment reduces tumor growth in bone and accompanying osteolytic disease as a result of decreased tumor burden in bone. However, vorinostat can promote osteopenia throughout the skeleton independent of tumor cell activity.