Revisiting autophagy addiction of tumor cells.

Inhibition of autophagy has been widely explored as a potential therapeutic intervention for cancer. Different factors such as tumor origin, tumor stage and genetic background can define a tumor's response to autophagy modulation. Notably, tumors with oncogenic mutations in KRAS were reported to depend on macroautophagy in order to cope with oncogene-induced metabolic stress. Our recent report details the unexpected finding that autophagy is dispensable for KRAS-driven tumor growth in vitro and in vivo. Additionally, we clarify that the antitumorigenic effects of chloroquine, a frequently used nonspecific inhibitor of autophagy, are not connected to the inhibition of macroautophagy. Our data suggest that caution should be exercised when using chloroquine and its analogs to decipher the roles of autophagy in cancer.

Macroautophagy, a catabolic cellular recycling process, assists in the maintenance of organelle and macromolecular homeostasis and intracellular nutrient availability in times of stress. Autophagy has been linked to many pathological conditions such as neurodegeneration, bacterial and viral infection, cancer, and Crohn disease. In cancer, haploinsufficiency of the autophagy core gene BECN1 has been shown to increase tumorigenesis; however, other studies have demonstrated a decrease in tumorigenesis upon autophagy inhibition in the context of oncogene-induced metabolic and genotoxic stress, illustrating the complexity of the relationship between autophagy and cancer.

The term “cancer” does not define a single disease, but represents a grouping of different genetic events that result in the transformation of normal host cells into cancer cells that exhibit inappropriate and uncontrolled growth. By focusing on the genetic landscape of a tumor, precision medicine approaches can be developed to exploit individual vulnerabilities. Such approaches have yielded success when the driving oncogene is directly targeted, as exemplified by inhibitors of tumor-promoting fusion proteins (such as BCR-ABL and EML4-ALK) or oncogenic mutant kinases (EGFR and BRAF). The situation is not as promising for cancers driven by oncogenic mutations in KRAS which are prevalent in a large percentage of tumors including pancreatic, colon and lung cancers, and often a poor predictor of response to targeted therapies. As there are no approved drugs to pharmacologically inhibit KRAS, identification of specific vulnerabilities in this tumor population could lead to therapeutics that would fulfill a high unmet medical need.

Autophagy addiction was reported to be one such vulnerability of KRAS mutant tumors, suggesting a therapeutic opportunity for autophagy inhibitors to target this tumor population. We endeavored to confirm this autophagy dependency by screening a large panel of human cancer cell lines for proliferative defects upon knockdown of the autophagy core components ULK1, ATG7 or PIK3C3. We used a deep coverage shRNA library comprised of 17 shRNA hairpins per gene to mitigate off-target activities of individual RNAi reagents. Neither KRAS mutated nor KRAS wild-type cancer lines were sensitive to the depletion of ULK1, ATG7 or PIK3C3, despite inhibition of macroautophagy by these shRNAs in a pooled screening format. Our cell culture and screening conditions maintained KRAS-dependency as demonstrated by the robust inhibition of KRAS-driven cell proliferation by shRNA reagents targeting KRAS itself. To complement the shRNA results, our labs independently used genome-editing tools to completely delete ATG7 function in several KRAS-mutated human cancer cell lines. These autophagy-deficient lines displayed sensitivity to nutrient deprivation in vitro but no impairment of growth under standard nutrient conditions or when grown as xenografts in vivo. Collectively, our results demonstrate that autophagy in tumors is not required to promote KRAS-driven cell growth.

Previous reports have described autophagy dependency of tumors in genetically engineered mouse models harboring deletions of Atg5 or Atg7. In most of these contexts, both tumor and normal host tissues are deficient for autophagy. As our studies conclude that tumor-autonomous autophagy does not contribute to tumor growth, the question arises whether autophagy in normal tissues drives tumorigenesis. There is accumulating evidence for a role of autophagy in immune or stromal cells and future studies will be needed to tackle possible mechanisms by which autophagy in such tissues can drive tumor growth.

Our work also explored the antitumorigenic effects of chloroquine, a weakly basic drug that accumulates in lysosomes, thereby increasing lysosomal pH and limiting their degradative capacity. Numerous studies rely on chloroquine, either as a single or combination agent, to assess the contribution of autophagy to various forms of stress, including that induced by cancer chemotherapeutic agents. We now demonstrate that chloroquine inhibits growth of autophagy-competent and autophagy-deficient cells to the same extent, clearly decoupling the antiproliferative activity of chloroquine from autophagy inhibition. We also confirmed that chloroquine synergizes with select targeted agents such as erlotinib and sunitib; however, this synergy is also independent of autophagy. Despite lack of specificity, chloroquine is still frequently used as a tool compound to demonstrate dependence on autophagy. Our data indicate that any phenotypes observed with chloroquine should not be automatically ascribed to the inhibition of autophagy, and that more specific methods of autophagy inhibition (such as genetic deletion of autophagy core components) are required to confirm the role of autophagy in a particular biological process. Chloroquine continues to be of interest as a potential oncology therapeutic due to its long history as a well-tolerated drug used to prevent and treat malaria. By focusing translational efforts on the autophagy-inhibitory property of chloroquine, researchers may be overlooking opportunities to understand and extend the potential therapeutic benefit of this drug. Determining the mechanisms by which chloroquine synergizes with targeted agents will be critical in identifying the combination partners likely to have the highest clinical impact.