Inhibitors of the stromal protease fibroblast activation protein attenuate tumor growth in vivo

Abstract

Fibroblast activation protein (FAP) is a serine protease that is selectively expressed on tumor stromal fibroblasts in a vast majority of epithelial cancers. We have previously shown that FAP overexpression enhances tumor growth and suggested a possible therapeutic role for functional inhibition of FAP activity (Cancer Res 62:4767-4772, 2002). PT-100 and PT-630 are boronic dipeptides, which competitively inhibit dipeptidyl peptidases including FAP and DPP-IV/CD26 with Ki in the nanomolar range. PT-100 treatment has been shown to decrease tumor size and cause tumor regression and rejection in a variety of animal tumor models (Cancer Res 64: 5471-5480, 2004). Its mechanism of action is complex, as it not only inhibits FAP enzymatic activity but also has been shown to have immunomodulatory effects including upregulation of cytokine and chemokine expression. PT-630 inhibits FAP enzymatic activity to a similar degree as PT-100, but without invoking a cytokine response. Therefore, we investigated the effects of PT-630 and PT-100 in animal models that have strong FAP expression. HT-29 xenografts induce stromal FAP expression in the tumor fibroblasts without any detectable expression on the colorectal cancer cells themselves. In addition HEK293 cells transfected with FAP were also tested as an epithelial FAP overexpression animal model. PT-100 and PT-630 when administered orally to mice inoculated subcutaneously with either HT-29 or HEK-FAP cells, significantly inhibited tumor growth when compared to saline treated controls. Both PT-100 and PT-630 decreased tumor volume in the HT-29 and HEK-FAP xenografts by approximately 50%. Immunocapture of FAP and DPP-IV from the tumors showed significantly decreased enzymatic activity in PT 100 and PT 630 treated xenografts indicating that FAP activity is suppressed in vivo. A dose related response for PT-630 was also suggested. We therefore demonstrate that PT-100 and PT-630 inhibit tumor growth of HT-29 and HEK-FAP xenografts. In addition, PT-100 and PT-630 attenuate enzymatic activity in vivo in a dose dependent manner. Given that PT-630 has minimal immunomodulatory effects, this suggests that the antitumor activity seen in these animal models may be mediated by the inhibition of FAP at the tumor microenvironment.