Microenvironmental effects limit efficacy of thymoquinone treatment in a mouse model of ovarian cancer

Andrew J. Wilson, Jeanette Saskowski, Whitney Barham, Dineo Khabele, Fiona Yull

Ovarian cancer is the most lethal gynecologic malignancy, with limited treatment options for chemoresistant disease. An important link between inflammation and peritoneal spread of ovarian cancer is NF-κB signaling. Thymoquinone (TQ) exerts multiple anti-tumorigenic cellular effects, including NF-κB inhibition. We aimed to investigate the therapeutic potential of TQ in an established murine syngeneic model of ovarian cancer. ID8-NGL mouse ovarian cancer cells stably expressing an NF-κB reporter transgene were injected intra-peritoneally into C57BL/6 mice, and mice were treated with TQ or vehicle for 10 or 30 days. TQ was combined with the macrophage depleting drug, liposomal clodronate, in selected experiments. Effects on peritoneal tumor burden were measured by volume of ascites, number of peritoneal implants and mesenteric tumor mass. NF-κB reporter activity and markers of proliferation and apoptosis were measured in tumors and in confirmatory in vitro experiments. Protein or mRNA expression of M1 (anti-tumor) and M2 (pro-tumor) macrophage markers, and soluble cytokine profiles, were examined from harvested ascites fluid, peritoneal lavages and/or tumor sections. 2-tailed Mann-Whitney tests were used for measuring differences between groups in in vivo experiments. Consistent with its effects in vitro, TQ reduced proliferation and increased apoptosis in ID8-NGL tumors after 10 and 30 day treatment. Prolonged TQ treatment did not significantly alter tumor number or mass compared to vehicle, but rather exerted an overall deleterious effect by stimulating ascites formation. Increased ascites was accompanied by elevated NF-κB activity in tumors and macrophages, increased pro-tumor M2 macrophages and expression of pro-tumorigenic soluble factors such as VEGF in ascites fluid, and increased tumor infiltration of M2 macrophages. In contrast, a 10 day exposure to TQ produced no ascites, and reduced tumor NF-κB activity, M2 macrophages and soluble VEGF levels. Peritoneal macrophage depletion by clodronate significantly reduced tumor burden. However, TQ-stimulated ascites was further enhanced by co-treatment with clodronate, with macrophages present overwhelmingly of the M2 phenotype. Our findings show that pro-tumorigenic microenvironmental effects limited the efficacy of TQ in a syngeneic mouse model of ovarian cancer, and provide caution regarding its potential use in clinical trials in ovarian cancer patients.