Functional characterization of the tumor antagonizing gene RNASET2.

Several types of cancer including solid tumors and blood malignancies are associated with alterations in different regions of chromosome 6. In particular, the 6q region has been frequently found to be deleted in a wide range of solid tumors, including ovarian carcinoma. Therefore, this chromosomal region has been extensively investigated in order to identifying putative tumor suppressor genes. In specific, a minimal region of deletion for ovarian cancer has been defined in the peritelomeric 6q27 region, and our group has recently isolated from this region the RNASET2 gene, coding for a secreted ribonuclease belonging to the Rh/T2/S acid ribonuclease family. The RNASET2 gene was found to be down-regulated at the transcriptional level in several primary ovarian tumors and ovarian cancer cell lines and its potential for tumor growth inhibition has been subsequently tested in vivo in a highly tumorigenic and metastatic Hey3Met2 ovarian cancer cell line. To this aim, athymic mice were injected with Hey3Met2 cell clones stably transfected with expression vectors encoding wild-type or a catalytically inactive mutant RNASET2 protein. Strikingly, wild-type RNASET2-expressing cells formed smaller tumors in mice and gave rise to a lower number of metastasis when compared to control cells transfected with the vector only. Moreover, mutant RNASET2-expressing cells also displayed a similar or even more increased inhibition of tumorigenicity when compared to wild-type RNASET2, suggesting that catalytic activity is not involved in control of tumor suppression and metastasis. With the purpose of further defining the biological role of RNASET2, we have undertaken a functional characterization of this gene, by comparing RNASET2-overespressing cells behavior both in vitro and in vivo. In sharp contrast with previous in vivo observations, several in vitro tests such as proliferation, clonogenicity, growth in soft agar and apoptosis assays failed to reveal any change in cancer-related phenotypes in the same RNASET2-overexpressing cell line, thus suggesting a non-cell autonomous activity for RNASET2 as a tumor suppressor. This “asymmetric” in vivo/in vitro property has been recently ascribed to a novel class of tumor suppressor genes, called tumor antagonizing genes (TAGs), whose oncosuppressive activity is carried out only in vivo by induction of microenvironmental remodeling. To confirm this hypothesis we turned to a xenograft assay and, as expected, Hey3Met2 cells expressing wild-type or mutant RNASET2 were not able to support tumor growth when compared to control mice. Strikingly, a detailed histological analysis of tumor sections revealed a different architectural pattern between control mice and RNASET2-expressing tumors. The former showed a solid and diffuse growth with strong component of human tumor cells, whereas in the latter a more organized structure and a 2 massive infiltration of host stromal cells mainly derived from monocyte/macrophage lineage has been found. In order to establish whether these cells could be functionally responsible for RNASET2-mediated tumor suppression, a further in vivo xenograft assay with Rag-2-/- γc-/- mice was performed, since these mice lack both lymphocytes and NK cells, thus retaining innate immunological response only through macrophages. As previously observed in nude mice, RNASET2-expressing cells were again clearly suppressed in tumorigenicity, further suggesting a functional role for macrophages recruitment in tumor suppression mediated by RNASET2. To definitively confirm the involvement of this cell lineage, in an additional in vivo experiment Rag-2-/- γc-/- mice were pretreated with the macrophages-depleting agent clodronate before inoculation of Hey3Met2 cells. Untreated mice showed once again a clear effect of tumor suppression exerted by RNASET2 whereas in clodronate-treated mice a very similar tumor development was noticed in both control and RNASET2-expressing cells. Finally, immunohistochemical analysis showed that recruited immune cells mainly belong to M1 subtype of anti-tumorigenic macrophages, suggesting a possible mechanism of tumor suppression exerted by this secreted ribonuclease. In order to extend RNASET2 functional characterization, beside the experimental model described above a complementary research approach based on gene silencing was carried out, aimed at studying how the lack of RNASET2 could affect tumor development and providing a powerful tool for further investigations. Our working hypothesis considers RNASET2 inactivation as a critical step in ovarian cancer progression, therefore cancer cell lines endowed with a low in vivo tumorigenic potential and expressing high endogenous RNASET2 levels might represent an ideal model system to challenge this hypothesis. In such a context, RNAi-mediated gene knockdown could provide additional information on this topic. RNAi-mediated silencing of the RNASET2 gene has therefore been attempted in the human ovarian carcinoma cell line OVCAR3, which displays the above mentioned features. Complete knockdown of RNASET2 in OVCAR3 cell line was successfully achieved and preliminary in vitro tests are in keeping with our previous data, showing no significant differences in proliferation and clonogenic potential in vitro in RNASET2-silenced when compared to control clones, and further supporting a “non-cell autonomous” behavior of RNASET2. Gene silencing of RNASET2 also allowed us to studying both the intracellular routing and likely the mechanism(s) by which extracellular RNASET2 affects in vivo cell behavior, demonstrating that providing an external source of RNASET2 to silenced OVCAR3 cells resulted in the binding of the ribonuclease to the cell surface and its subsequently internalization.