Revisiting the role of autophagy in melanoma

ABSTRACT

Although alterations of the macroautophagy/autophagy-lysosome pathway have been observed in cancer for many years, the mechanisms underlying these changes and the importance of autophagic and lysosomal reprogramming by cancer have yet to be well identified. Our recent study demonstrates that oncogenic BRAF signaling promotes melanoma growth and resistance to BRAF-targeted therapy through phosphorylation and functional inactivation of TFEB (transcription factor EB) and consequent suppression of the autophagy-lysosome gene network. This is by no means the first time that this pathway has been directly linked to oncogenic BRAF-driven melanoma. The key observations revealed in this study also leads to a complex but growing convergence of our understanding of the biology of the autophagy-lysosome pathway and the mechanisms underlying cancer prevention and treatment.

KEYWORDS:

• Autophagy
• BRAF
• lysosome
• melanoma
• TFEB

Autophagy constitutes a barrier to malignant transformation but also can serve as a platform that tumors use for their adaptation to energy stress and various cancer therapies. In the case of melanoma, particularly those harboring BRAF^V600E, tumor resistance to BRAF inhibitors has been reported to engage ER stress-associated autophagy induction, and it has been proposed that BRAF inhibitor-induced autophagy serves as a target for melanoma therapy. Less clear are the precise mechanisms by which autophagy responds to BRAF signaling and its impacts on disease progression and therapy response. Although autophagy induction occurs primarily in the cytoplasm, in some cases, including the physiologically relevant condition of starvation, autophagy is virtually initiated in the nucleus as part of a transcriptional program controlling lysosome biogenesis/function, mediated by the MiT/TFE transcription factors. Our study indicates that a particular member of the MiT/TFE family, TFEB, serves an important function in connecting BRAF signaling to autophagy-lysosome-mediated catabolism in melanoma [1]. We present evidence for a direct interaction, phosphorylation, and inactivation of TFEB by the BRAF^V600E downstream effector MAPK/ERK and provide a compelling model to explain the suppressive role that TFEB and resultant autophagy-lysosome activation plays in BRAF^V600E-driven melanoma. In a broader context, the proposed model suggests a novel mechanism by which loss of signaling through TFEB can fuel tumor progression, dissemination, and chemoresistance. This finding also underscores the importance of the autophagy-lysosome pathway in tumor suppression.

Any interpretation of the potential role of autophagy in cancer therapy requires a pre-understanding of how autophagy is regulated and how it abnormally functions in cancer cells. Approximately 40–60% of melanomas harbor BRAF mutations that promote RAF-MAP2K/MEK-MAPK/ERK pathway activation and melanoma proliferation. We discovered that, upon exposure to BRAF inhibitors, BRAF^V600E melanoma cells set autophagy in motion, surprisingly not through induction of ER stress as previously proposed, but by activation of TFEB as an integrated response that upregulates the lysosome biogenesis/function. The TFEB program controls a plethora of homeostatic functions, most notably the regulation of autophagy and lysosome biogenesis/function. Removing TFEB, but not its family members TFE3 and MITF, erase the autophagy-promoting effect of BRAF inhibitors. In fact, TFEB moves to the nucleus soon after BRAF inhibition – a characteristic shared by most MiT/TFE factors that respond to stress, yet one that had never been reported to occur during a targeted therapy.

To address this mechanism, we demonstrated a critical role of constitutively activated MAPK/ERK, which lies downstream of BRAF^V600E, in the regulation of cytoplasmic localization of TFEB, which influences the outcome of the autophagy-lysosomal response to BRAF inhibitors. Further analyses revealed that TFEB acts as a target for BRAF^V600E through MAPK/ERK-induced phosphorylation on serine 142, which keeps TFEB in check in the cytoplasm in oncogenically primed melanoma. The forced cytoplasmic localization of TFEB using a mutant that mimics S142 phosphorylation (S142E) causes increased TFEB lysosome association, assembly of the inactive TFEB-YWHA/14–3-3 complex, and protection from BRAF inhibitor-induced autophagy activation, suggesting that this treatment response hinges on TFEB dephosphorylation. Conversely, alanine substitution for S142 generates a non-phosphorylatable TFEB mutant (S142A) that functions as a ‘universal inducer’ of the autophagy-lysosome program irresponsive to BRAF-mediated inhibition. Remarkably, this regulation is independent of MTORC1 activation, a cannonical mechanism of TFEB suppression. As if the discovery of the TFEB response for BRAF-targeted therapy was not striking enough, we further show that the role of TFEB in autophagy-lysosomal activation is amplified by the phosphorylation and suppression of the TFEB antagonist ZKSCAN3 through a MAPK9/JNK2/p38 MAPK-dependent mechanism. These findings confirm the coordination between TFEB and ZKSCAN3 that had previously been observed in the regulation of the autophagy-lysosome gene network. Thus, the enhanced autophagic effect of BRAF-inhibiting agents might not be an inherent survival response, but could be due, in whole or in part, to the loss of the BRAF-MAP2K/MEK-MAPK/ERK signaling and an unleashed TFEB-ZKSCAN3 pathway in melanoma.

Our study also raised a mechanistic possibility that BRAF^V600E-mediated oncogenic growth may be, at least in part, through TFEB inhibition. When TFEB is activated, BRAF^V600E melanoma cells are compromised in the number and size of tumors formed. In other words, for BRAF-driven tumor progression to occur, a forced decrease in the autophagy-lysosome-promoting mechanism must also occur. Intriguingly, TFEB activation in BRAF^V600E melanoma cells results not only in tumor suppression but also in decreased metastasis in a syngeneic mouse model. By contrast, increased TFEB inactivation translates into elevated proliferation, epithelial-mesenchymal transition (EMT), and metastasis of tumor cells. Although it is thought that a tumor with higher autophagic response would respond worse to chemotherapy, in this case of BRAF-targeted therapy, we found that TFEB-activated tumors are more sensitive to BRAF inhibitor as measured by both clonogenic survival and regrowth of the tumors following treatment. Consistently, TFEB-disrupted clones show higher resistance to BRAF inhibitor treatment, suggesting that impaired autophagic-lysosomal response is causally related to drug resistance in melanoma.

How does TFEB and by extension the autophagy-lysosomal pathway exert its role in tumor growth and response to anticancer drugs at a molecular level? To this end, we performed RNA-Seq of melanoma xenografts seeking the target(s) that is necessary for the acquired tumor growth in TFEB-inactivated cells. This led to the discovery of aberrant accrual of TGFB/TGF-β (transforming growth factor beta) and consequently increased activation of TGFB signaling in melanoma cells with disrupted TFEB function. Blockade of TGFB activation reverts EMT and restores drug sensitivity of TFEB-deficient cells. Imaging and biochemical data suggest that the autophagy apparatus binds immature TGFB and governs it for lysosomal degradation. Decline of the autophagosome-lysosomal functions results in an exacerbated generation of TGFB even in TFEB-activated melanoma cells. More details of this regulation merits further investigation, especially where the interaction occurs and how this is selectively affected by TFEB. It also remains to be shown whether BRAF^V600E targets TFEB in other cellular processes. Our work on the interaction between BRAF, TFEB, and TGFB reveals an integrated regulation of the autophagy-lysosome pathway for oncogenic adaptation and calls for more caution regarding the usage of proximal inhibitors of autophagy in patients receiving BRAF-targeted therapy for melanoma and other relevant cancers.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the National Institutes of Health [ES029092]; National Institutes of Health [CA140964].