AMBRA1 and FAK1: crosstalking for improved targeted therapy in melanoma

ABSTRACT

Through genetically engineered mouse models of melanoma, we identified Autophagy/beclin 1 regulator 1 (Ambra1) as novel tumor-suppressor in melanoma. In these settings, loss of Ambra1 associated with the hyperactivation of focal adhesion kinase 1 (Fak1) signaling, the inhibition of which resulted in reduced tumor growth and invasiveness. We therefore propose FAK1 inhibition for current melanoma therapy in AMBRA1-low tumors.

Abbreviations

AKT, serine/threonine kinase 1; AMBRA1, autophagy/beclin 1 regulator 1; BRAF, v-raf murine sarcoma viral oncogene homolog; BRAFi, BRAF inhibitor; CCLE, Cancer Cell Line Encyclopedia;g ESTDAB, European Searchable Tumor Line Database; FAK1, focal adhesion kinase 1; FAKi, FAK1 inhibitor; LMC, Leeds Melanoma Cohort; MEK, MAPK/ERK kinase; PP2A, protein phosphatase 2A; PTEN, phosphatase and tensin homolog; TCGA-SKCM, The Cancer Genome Atlas - Skin Cutaneous Melanoma; YAP, yes-associated protein 1.

Commentary

Melanoma is the most deadly and aggressive skin cancer, with high metastatic potential, elevated heterogeneity and mutational burden being the major contributors of melanoma biology. Over the last decade, molecular players of melanoma have been characterized, among which is v-raf murine sarcoma viral oncogene homolog (BRAF), with the V600E substitution found in >50% of patients.¹ The increased knowledge about melanoma molecular mechanisms has substantially contributed to the improvement of patient treatment and survival, for instance through the use of BRAF-targeted therapy (BRAF inhibitor, BRAFi).² Despite such advances, however, BRAFi monotherapy may fail over time, mainly as a consequence of acquired resistance to therapeutic agents and activation of downstream targets, among which is MAPK/ERK kinase (MEK).² Such complex scenario therefore highlights the need for new targets in melanoma patient treatment.

Recently, we have proposed the Autophagy/beclin 1 regulator 1 (Ambra1) as a novel tumor suppressor in melanoma.³ By means of the pre-clinical Braf^V600E/+;Pten^−/−(phosphatase and tensin homolog) mouse model of melanoma, we have shed light on new phenomena orchestrated by Ambra1 in melanoma, particularly cell migration, extracellular matrix remodeling, epithelial-to-mesenchymal transition-like process, invasiveness and metastatic potential. Indeed, tumors in which Ambra1 was genetically ablated not only displayed faster kinetics of growth, but were also more likely to colonize local and distal organs, as demonstrated in different murine models of melanoma. Similarly, human melanoma cellular models in which AMBRA1 expression was manipulated, along with human melanoma cells from the European Searchable Tumor Line Database (ESTDAB) and the Cancer Cell Line Encyclopedia (CCLE) or patients from The Cancer Genome Atlas – Skin Cutaneous Melanoma (TCGA-SKCM) and the Leeds Melanoma Cohort (LMC) ranked in terms of AMBRA1 expression levels, clearly showed a negative correlation between AMBRA1 expression and invasive/migratory capacity.

Seeking the molecular determinants of such events, we identified the focal adhesion kinase 1 (FAK1) signaling as the core of AMBRA1-mediated invasive capacity (Figure 1), as shown by both transcriptomics and biochemistry analyses of murine tumors and in vitro approaches. Indeed, in all experimental conditions, we documented increased phosphorylation of FAK1 (pFAK1) at Y397 and Y576 residues, as well of SRC (another component of the FAK1 signaling) at Y416. This was also accompanied with AMBRA1-silenced cells showing more active and elongated pFAK1-containing focal adhesions. The analysis of FAK1 signaling hyperactivation was also implemented by correlative evaluations in data from publicly available datasets, including the TCGA-SKCM, the CCLE and the LMC. Additionally, we demonstrated that the AMBRA1-mediated regulation of FAK1 signaling relies on the interaction between the two molecular players (Figure 1), rather than on other processes governed by AMBRA1, e.g. autophagy or interaction with protein phosphatase 2A (PP2A). Although an interplay between AMBRA1 and FAK1 had already been documented,⁴ no functional outcomes for such interaction had been provided so far in melanoma. Such gap has now been filled, as demonstrated by the increased invasive capacity of melanoma cells upon disruption of AMBRA1-FAK1 interaction (Figure 1). The significance of this molecular axis has also been highlighted by either in vitro genetic manipulation and ex vivo administration of FAK1 inhibitor (FAKi) in AMBRA1-silenced or -null cells (Figure 1). In such conditions, indeed, the higher cell migratory capacity typical of AMBRA1-silenced melanoma cells was severely abolished. Interestingly, we also defined a dramatic increase in sensitivity to FAKi in ESTDAB cells with low AMBRA1 expression. Notably, this was the case not only for human melanoma cells, as the kinetics of tumor growth was considerably affected in vivo as well, specifically in mice bearing Ambra1-depleted tumors (Figure 1). Shedding light not only on new essential molecular mechanisms governed by Ambra1, our results also underline oncogenic functions of FAK1 signaling in melanoma. This is interestingly in line with previous reports, in which the activation of FAK1 signaling has been described in aggressive uveal and cutaneous melanoma,^5–7 with several molecular players, including serine/threonine kinase 1 (AKT)⁸ and yes-associated protein 1 (YAP)⁷ also being involved. This growing body of evidence about the role(s) played by FAK1 in melanoma has therefore raised interest toward the possible use of FAKi in melanoma therapy. Disappointingly, however, FAKi monotherapy has revealed limited efficacy in clinical studies with advanced solid tumors.⁹ Although this may raise caution in regards to the monotherapy uses of FAKi, synergistic inhibition of FAK1 and other key molecular players of melanoma, such as BRAF⁶ and MEK¹⁰ has proved to increase efficacy upon acquired therapy resistance. In both cases, indeed, cytotoxic or, more generally, growth inhibitory effects have been described. Thus, it is not surprising that such strategy is currently under investigation in synergistic melanoma clinical trials of FAKi (defactinib or IN10018) in combination with MEK inhibitors (VS-6766 or cobimetinib).⁹ Overall, this line of evidence may hence imply that FAK1 targeting may be most effective in combination with other therapies. It is however noteworthy to mention that our results also reveal that FAKi therapy alone was effective in counteracting melanoma growth upon specific genetic conditions. This raises further interesting points about the use of FAKi therapy in melanoma. Indeed, we not only underlined how dependent Ambra1-deficient tumors are on the hyperactivation of Fak1 signaling for tumor progression, but also intimated that such a genetic background specifically leads to enhanced sensitivity to Fak1 targeting in pre-clinical models. Comprehensively, these results may pave the way for the identification of classes of genetic biomarkers in melanoma as – and other than – AMBRA1, which may help define sub-groups of patients highly responsive to FAKi monotherapy.

Figure 1.

Hyperactivation of FAK1 signaling upon AMBRA1 deficiency promotes melanoma development. Autophagy/beclin 1 regulator 1 (Ambra1) ablation in a v-raf murine sarcoma viral oncogene homolog (Braf)-mutated and phosphatase and tensin homolog (Pten)-deleted mouse model of melanoma promotes phosphoactivation of the focal adhesion kinase 1 (Fak1), resulting in increased motility and invasiveness of melanoma cells, eventually leading to higher metastatic burden. Administration of FAK1 inhibitor (FAKi) in Ambra1-depleted tumors reduces melanoma cell motility and invasion, as well as tumor growth

Funding Statement

This work has been supported by the Danish Cancer Society [KBVU R204-A12424]; LEO Fondet [LF-OC-19-000004]; Melanoma Research Alliance young investigator grant [MRA 620385].

Disclosure statement

No potential conflicts of interest were disclosed.