Study of molecular mechanisms involved in pigmentation and melanoma using translational approaches

Using our previous results (Botton, T. et al., 2009, Botton, T. et al. 2011) and structure/activity relationship studies in collaboration with the team of Dr. Benhida (Nice Institute of Chemistry), we developed and selected candidates (Thiazole Benzensulfonamides) exhibiting strong death-promoting effects in melanoma cells with HA15 as the lead compound of this series. Interestingly, HA15 induced the death of all tested melanoma cells independently of their mutational status and was also effective on melanoma cells freshly isolated from patients sensitive or resistant to BRAF inhibitors. HA15 exhibited a strong efficacy in xenograft mouse models using melanoma cells sensitive and resistant to BRAF inhibitors without any sign of toxicity. We next performed pan-genomic, proteomic, and biochemical studies to decipher the action mechanism and identify the target of the best candidates. We identified BIP (also known as GRP-78 or HSPA5), an endoplasmic reticulum protein, as the specific target of our compound. We unequivocally demonstrated that the interaction between our compound and BIP increases endoplasmic reticulum (ER) stress and leads to melanoma cell death by concomitant induction of autophagy and apoptosis. Overexpression of BIP has been described in various cancers. It is thus not surprising that this molecule was also found to be active against other liquid and solid tumors. Taken together, our data suggest that our molecule has an important impact on inhibition of melanoma growth by targeting ER stress and may therefore be developed for treatment of patients with melanoma, and other cancers in general. This work led to the publication of more than 12 articles including one in Cancer Cell, and the filing of 4 licensed international patents. In July 2021, we co-founded a biotechnology start-up called Biper Therapeutics (https://www.biper-tx.com/) developing first-in-class small molecules to treat heavily resistant orphan cancers through modulation of the endoplasmic reticulum stress, a unique pathway to address cancer resistance.

We previously demonstrated that the antidiabetic drug metformin could exert antiproliferative and anti-invasive effects on melanoma in vitro and in vivo (Tomic, T. et al., Cell Death & Disease, 2011, Cerezo, M. et al., 2013). Following this proof of concept, we designed a pilot clinical study, in collaboration with Dermatology Department of Nice hospital, to evaluate the efficiency of metformin treatment in patients with advanced metastatic melanoma. Unfortunately, we did not observe a therapeutic effect of metformin as a single agent (Montaudié et al, PCMR, 2017). This can be explained, based on the recent literature, by the low bioavailability of metformin in the tumor. To overcome this limitation, we tested metformin derivative developed by a structure/function approach at the Nice Institute of Chemistry. This collaboration led us to identify the patented compound, CRO15, demonstrating an IC50 2000 times lower than metformin. The in vitro efficacy of CRO15 on melanoma cells, irrespective of their mutational context and resistance status to targeted therapies, and its innocuity to normal cells was confirmed using in vivo models. Mechanistically, low micromolar doses of CRO15 decrease OXPHOS, activate AMPK and inhibit mTOR, leading to the induction of autophagy and ultimately apoptosis. Interestingly, CRO15 also inhibits a limited number of kinases, such as MELK, involved in key oncogenic pathways (Jaune et al., CDDis, 2021). Thus, together with the apparent lack of toxicity, this combined action suggests the significant potential of CRO15 family members to fight therapy resistance in melanoma.

To continue to explore the mechanisms driving melanoma resistance, particularly in response to immune checkpoint inhibitors (ICI), we developed a new research program to decipher the reprogramming of metabolic pathway promoting melanoma resistance to anti-tumor immunity. Indeed, as immunotherapies are the first class of treatment that target immune cells and not directly the tumor cells itself, understanding the complex interplay among the various members of the tumor microenvironment (TME) is crucial to improve therapeutic response. Due to high resource consumption of cancer cells and vascularization impairments, the TME is frequently poor in nutrients and oxygen, driving a competition for nutriment access between cancer and stromal cells. Thus, manipulating tumor metabolism appears as the keystone of new strategies to improve response to immunotherapies and bypass resistance (Cerezo et al., Cell Death Dis., 2020).

As amino acids (AAs) are essential for protein synthesis and particularly needed to support the rapid growth of cancer cells, metabolic adaptations supporting AA synthesis appear to be attractive targets to improve melanoma response to immunotherapies. Interestingly, using metabolomic and fluxomic analysis we identified metabolic alterations, in melanomas resistant to ICI at the intersection of urea cycle and de novo pyrimidine synthesis with a particular upregulation of arginine synthesis. Interestingly, the rate limiting enzyme of arginine synthesis pathway, ASS1, is upregulated in samples from patients resistant to ICI as compared to responders. Arginine being known to be depleted in TME, our hypothesis is that this increase in arginine synthesis capacity make resistant cells less auxotrophic, giving them a proliferative advantage sufficient to overcome anti-tumor immunity. Based on these preliminary data, the aims of our project are to decipher how metabolic reprogramming drives resistance to ICI and determine if these metabolic alterations can serve as biomarkers and/or actionable targets to bypass resistance.

The next major subject of our work involving CLEC12B, a C-type lectin-like receptor family 12 member B, perfectly illustrates how understanding the biology of normal melanocytes can have direct implications in the study of melanoma. There is very little known about it in the literature except that it is an inhibitory receptor on myeloid cells, and its function is dependent on recruitment of phosphates SHP-1 and SHP-2. Since our team’s identification of CLEC12B as a top melanocytic gene in transcriptome analysis (Regazzetti et al., 2015), two separate PhD projects have been established to study its role in pigmentation (Laura Sormani) and in melanoma (Henri Montaudie). These were technically challenging projects as few scientific tools were available to study this molecule. The first of these projects demonstrated that, in the skin, CLEC12B is selectively expressed in pigment-producing melanocytes and its expression is decreased in highly pigmented skin compared to white skin (Sormani et al., JID 2021). CLEC12B directly controls melanin production and pigmentation via recruitment and activation of SHP-1 and SHP-2 through its immunoreceptor tyrosine-based inhibitory motif (ITIM).  Mechanism also involves degradation of CREB and downregulation of the MITF pathway (the master regulator of melanocyte function). This discovery of a novel gene which regulates our skin colour is an exciting prospect for development of melanogenic agents in the clinical and cosmetic fields and in parallel provides us with novel insights in the understanding of melanocyte biology and regulation of melanogenesis.

The second project focused on the role of CLEC12B in melanoma. These results have shown reduced expression of CLEC12B in melanoma metastases compared to benign melanocytic lesions (Montaudie et al., JID 2021). CLEC12B correlated with poor patient prognosis and mice xenografted with melanoma cells overexpressing CLEC12B had decreased tumor growth. These results suggest that CLEC12B may be a novel tumor-suppressor gene in melanoma. The mechanisms central to its effect are again recruitment of SHP-2 through its ITIM domain with parallel inactivation of STAT-1, 3 and -5 but activation of the p53/p21/p27 pathway. Results obtained from these projects are protected by a patent from INSERM Transfert (EP/08.06.17/EPA17305681).

A- CXCR3B-mediated apoptosis driving T-cell reactivity in vitiligo

In this translational, multidisciplinary project involving collaborations from within our Centre, our on-site hospital, and clinical researchers from several different hospital departments, we aimed to find out how vitiligo is initiated, with a larger vision of being able to prevent its early development and subsequently long-term disease. We focussed on the innate immune system as previous transcriptome data from our laboratory had suggested that the genes associated with activation of innate immunity were upregulated in lesional skin of subjects with nonsegmental or generalised vitiligo compared to healthy skin (Regazzetti et al, JID 2015). In this study we have shown, for the first time, activation of innate immune cells (natural killer and innate type I lymphocytes) in normal-appearing skin to be an early trigger in initiation of vitiligo. This process is initiated by endogenous PAMPs and DAMPs as well as extracellular (oxidative) stress, which drives innate cells to produce a large quantity of IFNg which directly drives upregulation of CXCR3B on human melanocytes. Previously, CXCR3 was only thought to be expressed by T cells (CXCR3A) and here we have shown melanocytes themselves to express this chemokine receptor but of the B isoform. Importantly, vitiligo patients with increased circulating IFNg have strong basal expression of CXCR3B. Activation of CXCR3B with CXCL10 directly triggers apoptosis of some melanocytes and, in others, increases their expression of costimulatory markers (CD40 and HLA-DR) and adhesion molecules (ICAM-1), which induces T cell proliferation and subsequent adaptive immune-mediated destruction of melanocytes. Together, these results have unravelled the mechanisms involved in triggering the initial chain of events which ultimately leads to pathogenesis of the disease that is vitiligo (Tulic et al., Nat Com 2019). We have recently extended these studies to patients with segmental vitiligo (Passeron et al., JID 2021). These novel results are protected by INSERM Transfert in the form of two patents (EP18305161 and PCT/EP2019/053767), the manuscripts have received a lot of attention (as evidenced by local and international invitations) and have received a Prize as the Best Paper of 2019 (Fonds de Dotation ‘Pour sa Peau, Pour sa Vie’). Together, these data indicate that reducing local stress and targeting CXCR3B directly could provide effective approaches for halting the initiation of the disease. In parallel, better understanding of how the immune system targets normal melanocytes might pave new avenues for the improvement of immunotherapies against melanoma.

B- Dysbiosis in skin microbiota as a contributor to vitiligo 

Although genetic studies have identified over 50 vitiligo susceptibility loci, the delay in vitiligo age-of-onset over the past 30 years emphasizes the key role of environmental factors in triggering vitiligo. As discussed above in our most recent publications, innate immune cells from vitiligo subjects have greater response to stress which can be triggered by PAMPs. Bacteria are among the top producers of PAMPs and may induce an innate immune response in vitiligo. We know that commensal cutaneous or gut dysbiosis have been linked to various dermatological disorders. This 3-year project (funded by ISISPharma) is the subject of our PhD student’s research (Ms Hanene Bzioueche). For this work we have studied matched stool (gut), skin and blood samples from vitiligo patients and healthy controls, sampling both surface (epidermis, swab) and deep (dermis, biopsies) skin layers. Swabs and biopsies were taken from both lesional and non-lesional sites of vitiligo subjects. In this most recent study, we have been able to identify bacterial strains that are enriched and those that are depleted in vitiligo skin (Bzioueche et al., JID 2021); we plan to target the latter for clinical intervention. Importantly, we have shown that biopsy microbiota in lesional skin have a distinct composition and are associated with mitochondrial stress and changes in innate immune function detected in the blood of the same subjects. Finally, we demonstrated that changes in the skin are also seen in the gut of the same individuals showing an increased Firmicutes/Bacteroidetes ratio, similar to what has been described for other autoimmune diseases and thus giving support for the gut-skin axis in vitiligo, however this remains to be formally tested.