Oncolytic adenoviruses seem like ideal candidates to target

Oncolytic adenoviruses seem like ideal candidates to target cancer initiating cells because they are cytotoxic and are not subject to the typical mechanisms of drug resistance, such as drug efflux pumps and defective apoptotic signaling. Indeed, studies with oncolytic adenoviruses in breast and 4 methylumbelliferone tumors suggest that they may be effective against CSC (Alonso et al., 2012; Bauerschmitz et al., 2008; Jiang et al., 2007). These viruses are often generated on the basis of its tumor selectivity to improve control of viral replication, resulting in diminished toxicity. The use of tissue specific promoters preventing the expression of E1A in non-target tissues is a useful strategy. Here we show that uPAR promoter controlled adenoviruses are able to kill neoplastic-sphere forming cells in vitro with a similar or even with higher efficacy than wild type adenoviruses. This is consistent with the observation that these cells endogenously express the uPAR gene, suggesting that the transcription factors acting on the regulatory sequences of the uPAR gene are present in the target cells and can activate the uPAR promoter. Importantly in vivo, AduPARE1A treated tumors were much smaller than mock-treated tumors, and had a similar proportion of CD133+ cells, and a smaller capacity to generate tumorspheres. This is in contrast to what was observed in gemcitabine treated tumors, where we observed an enrichment in the CD133 population and an increase in tumorsphere number and size. The fact that the in vivo effects of AduPARE1A treatment on the CP15 and CP13 PCSC population showed decreased sphere formation but no significant changes in the percentage of CD133+ cells may indicate that additional PCSC present in tumorspheres, not detected by CD133 positivity, would be sensitive to AduPARE1A treatment. This highlights the need to investigate for additional PCSC cellular markers as better predictors of AduPARE1A response. Nevertheless, our data suggest that AduPARE1A might be able to kill PCSC in vivo, similarly to what was seen in vitro and this can be appealing for pancreatic cancer treatment. Interestingly, a subpopulation of uPAR positive cells has been identified in small cell lung cancer cell lines, derived from lung and from bone marrow and brain metastasis, which possess multi-drug resistance and clonogenic activity (Gutova et al., 2007). uPAR positive cells identified in the small cell lung cancer H466 were capable of forming spheres and efficiently formed transplantable tumors (Qiu et al., 2012). uPAR signaling has also been reported to induce cancer stem cell like properties in breast cancer cells (Jo et al., 2010). Recently, it has been shown that uPAR is important in the maintenance of stem cells and is highly expressed in glioma initiating cells. The regulation of malignant stem-cell renewal was proposed to be through the activation of several components of the hedgehog pathway (Gopinath et al., 2013). Moreover, a recent study provides evidences of mutual regulation mechanisms among uPAR and beta-catenin signaling in cancer stemness in medulloblastoma (Asuthkar et al., 2012). Although there are no specific data on uPAR signaling in pancreatic cancer stem cells, in the present work we show that tumorspheres from both CP15 and CP13 tumors highly express the uPAR gene, supporting a role of uPAR in pancreatic CSC biology as well as the activity of the uPAR promoter in the PCSC population. Taken together, these results indicate that uPAR promoter is active in pancreatic cancer stem cells and uPAR-controlled oncolytic adenoviruses, on top of eliminating pancreatic epithelial cancer cells, as we have previously reported (Huch et al., 2009) are able to kill neoplastic cells with PCSC properties in vitro. Moreover the antitumor effects of AduPARE1A and the lack of enrichment in PCSC features of treated tumors support the in vivo killing of pancreatic cancer stem cells. The following are the supplementary data related to this article.