All roads lead to melanocytes

Abstract

Dear Sir, In a recent article published in Cell and highlighted in a News and Views piece in PCMR (Hwang and Pavan, 2010), Adameyko et al. have challenged the old notion that trunk neural crest-derived melanocytes only originate from cells migrating on a dorsolateral pathway (Adameyko et al., 2009). They found that nerve-associated Schwann cell precursors, which originate from ventro-medially migrating neural crest cells, also give rise to large numbers of melanocytes. As these studies were conducted in chicken and mouse, they raise the question as to whether ventro-medially derived melanocytes represent a novel pathway that originated in these two species or whether this pathway might be common to many more vertebrates. Interestingly, in our early work in a reptilian model (soft-shell turtle, Trionyx sinensis japonicus), we found melanocytes not only in skin but also in various internal organs and tissues, such as lung, skeletal muscle, dorsal aorta, peritoneum, blood vessels, bone marrow, and fat (Hou and Takeuchi, 1991, 1994). These extracutaneous melanocytes clearly originated from precursors in the ventral pathway (Hou and Takeuchi, 1991, 1994). We verified this finding by another series of experiments involving the culture of isolated trunk somites and proposed that in reptiles, somite-associated neural crest cells in the ventral pathway also serve as a source for melanocytes (Hou, 1999). Whether any of these ventrally derived melanocytes also end up in the skin, however, is not yet known. Nevertheless, together with the new findings by Adameyko et al., we can now assume that the derivation of melanocytes from ventrally migrating neural crest cells is a feature common to all amniotes. The spatial distribution of these melanocytes, however, differs in the different species. As mentioned, in turtles, trunk melanocytes are found in various internal organs, but in chicken and mice, their internal distribution is much more restricted. Why this difference? It is conceivable that ventrally migrating neural crest cells are multipotent in all amniotes, and that it is the tissue microenvironment in the developing embryo that is key to their ultimate location. Indeed, we found that extracts of lungs from embryonic turtle but not embryonic chicken are capable of inducing neural crest cells to differentiate into melanocytes (Hou and Kwon, 1995). Whether this difference is due to distinctions in the expression of inhibitory or of attractive substances, however, is not clear. Interestingly, Adameyko et al. found that in vitro, neuregulin-1 (NRG1)-ERBB3 signaling inhibits melanoblast differentiation (also see Buac et al., 2009) while insulinlike growth factor-1 (IGF1) and platelet-derived growth factor (PDGF) act to increase melanoblast numbers. Hence, in addition to Wnt, stem cell factor (SCF), and endothelins, whose positive roles in inducing melanocyte differentiation have been widely studied, we may count IGF1 and PDGF and their receptors and signaling molecules among additional factors that are, conceivably, expressed differentially in internal tissues in turtles, birds and mammals. It is equally plausible, of course, that internal organs of turtles, in contrast to those of birds and mammals, lack factors unfavorable to melanocyte development. Such a lack of negatively acting factors has been suggested to explain the widespread internal pigmentation in chicken with the Silkie mutation (Faraco et al., 2001). Evidently, these considerations make it important now to identify the factors associated with the differential accumulation of internal melanocytes among the different evolutionary lineages of vertebrates.