Regulation of ATG5 and the ATG12–ATG5-ATG16L1 complex in prostate cancer

Autophagy is a highly conserved pathway in which an autophagosome envelops cytoplasmic cargo and delivers it to the lysosome for degradation in order to maintain cellular homeostasis or survival in response to stress. The ATG12–ATG5-ATG16L1 complex functions as an essential regulator of autophagosome formation. We have discovered that DU145 prostate cancer (PCa) cells have a splice donor-site mutation that triggers aberrant splicing of ATG5 and leads to the proteasomal degradation of ATG12 and ATG16L1, thus completely inactivating autophagy. We demonstrate that ATG5, ATG12, and ATG16L1 are coordinately degraded when not associated with the complex and that the ATG5-ATG16L1 interaction is essential for preventing ubiquitination and turnover, thereby facilitating ATG12 conjugation. We also show that this interaction can be disrupted through alternative ATG5 splicing and by ATG5 genetic mutations that have been identified in human tumors. Meta-analysis of available mRNA expression data indicates that ATG5 is significantly downregulated in PCa. We confirmed previous reports that found prostate cancers have frequent deletions of the 6q21 locus containing ATG5. However, mRNA expression of neighboring genes is largely unaffected, indicating ATG5 can also be selectively downregulated though other mechanisms. Together, this suggests that ATG5 functions as a tumor suppressor gene that can be inactivated by a variety of different mechanisms. ATG5 is more significantly underexpressed than many established PCa tumor suppressor genes and is also underexpressed in PCa metastases compared to primary tumors. This implies that ATG5 is also a tumor suppressor in advanced PCa. ATG5 re-expression in ATG5-deficient DU145 PCa cells resulted in dramatic suppression of xenograft tumor growth, indicating that ATG5 is a functional PCa tumor suppressor gene. Therefore, autophagy may actually be tumor suppressive at both early and late stages of prostate tumorigenesis, which suggests that autophagy inhibition may be counterproductive for the treatment of advanced prostate cancers.