A systems biology approach uncovers the core gene regulatory network governing iridophore fate choice from the neural crest

Abstract

Multipotent neural crest (NC) progenitors generate an astonishing array of derivatives, including neuronal, skeletal components and pigment cells (chromatophores), but the molecular mechanisms allowing balanced selection of each fate remain unknown. In zebrafish, melanocytes, iridophores and xanthophores, the three chromatophore lineages, are thought to share progenitors and so lend themselves to investigating the complex gene regulatory networks (GRNs) underlying fate segregation of NC progenitors. Although the core GRN governing melanocyte specification has been previously established, those guiding iridophore and xanthophore development remain elusive. Here we focus on the iridophore GRN, where mutant phenotypes identify the transcription factors Sox10, Tfec and Mitfa and the receptor tyrosine kinase, Ltk, as key players. We present expression data, as well as loss and gain of function results, guiding the derivation of an initial iridophore specification GRN. Moreover, we use an iterative process of mathematical modelling, supplemented with a novel, Monte Carlo screening algorithm suited to the qualitative nature of the experimental data, to allow for rigorous predictive exploration of the GRN dynamics. Predictions were experimentally evaluated and testable hypotheses were derived to construct an improved version of the GRN, which we showed produced outputs consistent with experimentally observed gene expression dynamics. Our study reveals multiple important regulatory features, notably a sox10-dependent positive feedback loop between tfec and ltk driving iridophore specification; the molecular basis of sox10 maintenance throughout iridophore development; and the cooperation between sox10 and tfec in driving expression of pnp4a, a key differentiation gene. We also assess a candidate repressor of mitfa, a melanocyte-specific target of sox10. Surprisingly, our data challenge the reported role of Foxd3, an established mitfa repressor, in iridophore regulation. Our study builds upon our previous systems biology approach, by incorporating physiologically-relevant parameter values and rigorous evaluation of parameter values within a qualitative data framework, to establish for the first time the core GRN guiding specification of the iridophore lineage. Author SummaryMultipotent neural crest (NC) progenitors generate an astonishing array of derivatives, including neuronal, skeletal components and pigment cells, but the molecular mechanisms allowing balanced selection of each fate remain unknown. In zebrafish, melanocytes, iridophores and xanthophores, the three chromatophore lineages, are thought to share progenitors and so lend themselves to investigating the complex gene regulatory networks (GRNs) underlying fate segregation of NC progenitors. Although the core GRN governing melanocyte specification has been previously established, those guiding iridophore and xanthophore development remain elusive. Here we present expression data, as well as loss and gain of function results, guiding the derivation of a core irid...

TheMITFparalogtfecis required in neural crest development for fate specification of the iridophore lineage from a multipotent pigment cell progenitor

“…In a recent detailed study, we demonstrated that, indeed, tfec serves as a robust marker of the iridophore lineage, allowing us to define the major stages of iridophore development [14].…”

“…Characterisation of the phenotypes of mutants affecting multiple NC derivatives has been widely used to infer the identities and potencies of these progenitors. In the pigment cell field, a progressive fate restriction model has been developed, with both a multipotent chromatoblast and a bipotent melano-iridoblast as identified partially-restricted intermediates ( [10] [11]- [14]). Studies of fate-specification mutants further permits elucidation of the gene regulatory networks (GRNs) governing diversification of these precursors ( [14], [15]).…”

“…In terms of their differentiation, heightened purine synthesis is central to the development of the guanine crystals that form the reflecting platelets. Thus, purine nucleoside phosphorylase 4a (pnp4a), which encodes an enzyme important in the biosynthesis of guanine, is a robust marker of mature iridophores ( [13], [14]). Additionally, mutations of several enzymes specific to differentiated iridophores have been shown to disrupt purine biosynthesis [30], while disruption of other proteins was found to impair iridophore survival ([31]- [33]).…”

“…Nevertheless, Ltk signalling alone is not sufficient for iridophore specification, since iridophores are eliminated in sox10 mutants, even though ltk expression is strongly detectable by WISH ( [11], [14]). The function of Ltk signalling in specification of the iridophore lineage is, thus, likely analogous to that of Wnt signalling in generating melanocytes.…”

“…Focussing on cells in the posterior trunk, we distinguished a subset of tfec-positive cells within the premigratory domain that have downregulated the early NCC marker, foxd3, but do not express definitive pigment cell markers; we refer to these as early chromatoblasts (Cbl early). At approximately 22 hpf, ltk and mitfa are upregulated in the tfec+; foxd3premigratory cells of the trunk ( [11], [14], [19]), indicating that they correspond to partially restricted progenitors, capable of generating pigment cells (Cbl late). By 24 hpf, tfec labels migrating progenitors which co-express mitfa and ltk markers, and which we consider fatespecified iridoblasts, but which likely retain at least melanocyte potential too (Ib(sp); [14]).…”