Overexpression of REIC/Dkk-3 in Normal Fibroblasts Suppresses Tumor Growth via Induction of Interleukin-7

We previously showed that the tumor suppressor gene REIC/Dkk-3, when overexpressed by an adenovirus (Ad-REIC), exhibited a dramatic therapeutic effect on human cancers through a mechanism triggered by endoplasmic reticulum stress. Adenovirus vectors show no target cell specificity and thus may elicit unfavorable side effects through infection of normal cells even upon intra-tumoral injection. In this study, we examined possible effects of Ad-REIC on normal cells. We found that infection of normal human fibroblasts (NHF) did not cause apoptosis but induced production of interleukin (IL)-7. The induction was triggered by endoplasmic reticulum stress and mediated through IRE1α, ASK1, p38, and IRF-1. When Ad-REIC-infected NHF were transplanted in a mixture with untreated human prostate cancer cells, the growth of the cancer cells was significantly suppressed. Injection of an IL-7 antibody partially abrogated the suppressive effect of Ad-REIC-infected NHF. These results indicate that Ad-REIC has another arm against human cancer, an indirect host-mediated effect because of overproduction of IL-7 by mis-targeted NHF, in addition to its direct effect on cancer cells.Cancer cells, like normal cells, cannot be free from regulation by other cells in the body (1). The microenvironment can exert both promotive and suppressive effects on malignant cells (2). The embryonic environment has been shown to suppress malignant phenotypes (3, 4), and this was recently indicated to be due to suppression of Nodal function by Lefty (5). Cells comprising cancer stroma in adult tissues are also involved in tumor suppression (6, 7). Mobilization of such potential tumor-suppressive effects of the microenvironment would provide an additional arm for cancer therapy (8).Adenovirus vectors combined with appropriate cargo genes have great potential in cancer gene therapy because of their high infection efficiency and marginal genotoxicity (9). However, they show no target cell specificity and thus may also infect normal cells present in the surroundings of cancer cells. Provided that the interaction between cancer cells and normal cells is relevant to progression/suppression of cancer, it is critically important to understand not only cell autonomous phenomena in individual cell types infected by a therapeutic virus vector but also potential effects of the therapeutic virus vector on the composite system of interacting cell populations.We have been studying the possible utility of an adenovirus vector carrying the tumor suppressor gene REIC/Dkk-3 (Ad-REIC) for gene therapy against human cancer. REIC/Dkk-3 was first identified as a gene that was down-regulated in association with immortalization of normal human fibroblasts (NHF)² (10). Expression of REIC/Dkk-3 gene was shown to be reduced in many human cancer cells and tissues, including prostate cancer, renal clear cell carcinoma, testicular cancer, and non-small cell lung cancer (11–14), probably due to hypermethylation of the promoter (15). A single injection of Ad-REIC into tumors formed by transplantation of human prostate cancer cells (PC3 cells) into mice resulted in 4 of 5 mice becoming tumor-free (13). Subsequently, we found that Ad-REIC was effective also for human cancers derived from the testis, pleura, and breast (14, 16, 17). The potent multitargeting anti-cancer function of Ad-REIC shows great promise for clinical application, which will be shortly initiated.REIC/Dkk-3 is a highly glycosylated secretory protein and is considered to physiologically act on cells via a yet-unidentified receptor. However, we found that the induction of apoptosis in cancer cells by Ad-REIC was because of endoplasmic reticulum (ER) stress loaded by overproduction of the REIC/Dkk-3 protein and that exogenously applied REIC/Dkk-3 protein showed no apoptosis inducing activity for cancer cells (13, 14). Activation of c-Jun N-terminal kinase (JNK) was shown to be an essential step for the induction of apoptosis by Ad-REIC. ER stress is evoked by overload of unfolded/misfolded proteins in the ER, and eukaryotic cells respond to the threat by activating an unfolded protein response, i.e. attenuating de novo protein synthesis, promoting protein degradation by proteasomes, and inducing chaperone proteins to help proper folding of proteins (18). When ER stress remains at a level manageable by the unfolded protein response, cells can survive. On the other hand, overload of unfolded/misfolded protein beyond the cellular adoptive response leads to apoptotic cell death. Although Ad-REIC strongly induces apoptosis in many types of cancer cells, normal cells thus far examined are resistant to Ad-REIC-induced apoptosis despite expression of REIC/Dkk-3 at a level similar to that in cancer cells (13). The aim of this study was to determine the mechanisms of differential response of normal cells and cancer cells to Ad-REIC and to reveal the possible effect of Ad-REIC on a composite interacting system of normal cells and cancer cells. We found that Ad-REIC induced NHF to produce IL-7 via ER stress-triggered activation of p38. Furthermore, Ad-REIC-infected NHF significantly suppressed tumor growth of untreated PC3 cells transplanted in a mixture in vivo. These results mean that, in addition to its direct cancer cell-killing activity, Ad-REIC has another mechanism of action against human cancer, an indirect host-mediated effect because of overproduction of IL-7 by mis-targeted NHF.