Cellular interactions during zebrafish adult stripe formation and implications for pigment pattern evolution

Abstract

Pigment patterns are among the most striking of traits, yet we know very little about the interactions required for organizing pigment cells during adult pattern formation or how evolutionary changes in pigment cell development generate different patterns. As an adult the zebrafish, <italic>Danio rerio<italic> displays a pattern of dark horizontal stripes composed of black melanophores alternating with light interstripes composed of yellow xanthophores. Previous work showed that melanophore-xanthophore interactions are critical for stripe organization, but did not explain why stripes develop where they do. Here I investigate the role of a third pigment cell, the iridescent iridophore, in zebrafish stripe development and pigment pattern evolution. I show that in <italic>D. rerio<italic>, iridophores are the first pigment cell type to develop during adult stripe formation and that they do so precisely at the location of the first interstripe. By ablating iridophores in wild-type and mutants I found that iridophores play a role in organizing both xanthophores and melanophores into stripes. Additionally, I found that iridophores express <italic>colony stimulating factor 1 (csf1)<italic> and that localized Csf1 expression is sufficient to direct localized xanthophore development. Next, I asked whether differences in these interactions could play a role in the evolutionary loss of stripes in the closely related species, <italic>D. albolineatus<italic>. I found that iridophores are also required for stripe termination in <italic>D. rerio<italic>, but that <italic>D. albolineatus<italic> fails to develop iridophores at times and locations critical for stripe development. Instead, I found that <italic>D. albolineatus<italic> developed precocious xanthophores. Early, widespread xanthophore development in <italic>D. albolineatus<italic> was associated with similarly widespread and early expression of <italic>csf1<italic>, a change associated with <italic>cis<italic>-regulatory evolution at the Colony stimulating factor-1a (Csf1a) locus. Finally, to test if these changes in xanthophore and iridophore development are sufficient to explain the loss of stripes in <italic>D. albolineatus<italic>, I expressed <italic>csf1<italic> similarly to <italic>D. albolineatus<italic> in <italic>D. rerio<italic>. Under these conditions, xanthophores and melanophores were intermingled, iridophore development was repressed and stripes were absent, as in <italic>D. albolineatus<italic>. These results suggest that changes in the timing of pigment cell differentiation can have downstream effects on pattern formation and likely contribute to evolutionary diversification in this group.