The outgrowth of primary melanoma, the deadliest skin cancer, and generation of metastasis is supported by the tumor microenvironment (TME) which includes non-cancerous cells. Since the TME plays an important role in melanoma pathogenesis, its targeting is a promising therapeutic approach. Thus, it is important to identify proteins in the melanoma TME that may serve as therapeutic targets. Here we show that the nicotinamide adenine dinucleotide glycohydrolase CD38 is a suitable target for this purpose. Loss of CD38 in the TME as well as inhibition of its enzymatic activity restrained outgrowth of primary melanoma generated by two transplantable models of melanoma, B16F10 and Ret-mCherry-sorted (RMS) melanoma cells. Pathological analysis indicated that loss of CD38 increased cell death and reduced the amount of cancer-associated fibroblasts (CAFs) and blood vessels. Importantly, in addition to inhibiting outgrowth of primary melanoma tumors, loss of CD38 also inhibited spontaneous occurrence of RMS pulmonary and brain metastasis. The underlying mechanism may involve, at least in the brain, inhibition of metastasis expansion, since loss of CD38 inhibited the outgrowth of B16F10 and RMS brain tumors that were generated by direct intracranial implantation. Collectively, our results suggest that targeting CD38 in the melanoma TME provides a new therapeutic approach for melanoma treatment.

Highlights d Changes in RNA editing levels are harmful to normal development d In C. elegans, editing affects genes at the RNA level, but not at the protein level d ADR-1 function is to promote editing by ADR-2 at the L4 larval stage d C. elegans ADR-1 has developmental functions that are unrelated to RNA editing

Primary and metastatic melanoma progression are supported by a local microenvironment comprising, inter alia , of cancer-associated fibroblasts (CAFs). We previously reported in orthotropic/syngeneic mouse models that the stromal ectoenzyme CD38 participates in melanoma growth and metastasis. The results presented here suggest that CD38 is a novel regulator of CAFs’ pro-tumorigenic functions. Orthotopic co-implantation of CD38 deficient fibroblasts and B16F10 melanoma cells limited tumor size, compared to CD38 expressing fibroblasts. Intrinsically, CAF-CD38 promoted migration of primary fibroblasts toward melanoma cells. Further, in vitro paracrine effects of CAF-CD38 fostered tumor cell migration and invasion as well as endothelial cell tube formation. Mechanistically, we report that CAF-CD38 drives the protein expression of an angiogenic/pro-metastatic signature, which includes VEGF-A, FGF-2, CXCL-12, MMP-9 and HGF. Data suggest that CAF-CD38 fosters tumorigenesis by enabling the production of pro-tumoral factors that promote cell invasion, migration and angiogenesis.

A-to-I RNA editing is widespread in eukaryotic transcriptomes and plays an essential role in the creation of proteomic and phenotypic diversity. Loss of ADARs, the proteins responsible for A-to-I editing, results in lethality in mammals. In C. elegans, knocking out both ADARs, ADR-1 and ADR-2, results in aberrant behavior and abnormal development. Studies have shown that ADR-2 can actively deaminate dsRNA while ADR-1 cannot. However, as most studies of C. elegans ADARs were performed on worms mutated in both ADAR genes, the effects observed cannot be attributed to a single ADAR or to the interactions between ADAR genes. Therefore, we set to study the effects of each C. elegans ADAR on RNA editing, gene expression, protein levels and the contribution of each of ADAR to the phenotypes observed in worms mutated in both genes, in order to elucidate their distinct functions. We found significant differences in the phenotypes observed in worms mutated in a single ADAR gene.Worms harboring adr-1 mutations have a significant reduction in their lifespan, while worms harboring adr-2 mutations have extended lifespan. We also observed severe abnormalities in vulva formation in adr-1 mutants, and we suggest that these phenotypes are a result of an RNA editing independent function of ADR-1. Mutations in each ADAR resulted in expressional changes in hundreds of genes, and a significant downregulation of edited genes.However, very few changes in the protein levels were observed. In addition, we found that ADR-1 binds many edited genes and primarily promotes editing at the L4 stage of development. While editing still occurs in the absence of ADR-1, most of the editing occurs in genes that are edited in wildtype worms, suggesting that ADR-1 does not prevent editing by binding to and protecting the RNA but rather enhances or promotes editing. Our results suggest that ADR-1 plays a significant role in the RNA editing process and by altering editing levels it causes the severe phenotypes that we observed. In contrast, a complete lack of RNA editing is less harmful to the worms. Furthermore, our results indicate that the effect of RNA editing on the protein content in the cell is minor and probably the main purpose of these modifications is to antagonize or enhance other gene regulatory mechanisms that act on RNA.