Beyond MITF: Multiple transcription factors directly regulate the cellular phenotype in melanocytes and melanoma

Abstract

MITF governs multiple steps in the development of melanocytes, including specification from neural crest, growth, survival, and terminal differentiation. In addition, the level of MITF activity determines the phenotype adopted by melanoma cells, whether invasive, proliferative, or differentiated. However, MITF does not act alone. Here, we review literature on the transcription factors that co-regulate MITF-dependent genes. ChIP-seq studies have indicated that the transcription factors SOX10, YY1, and TFAP2A co-occupy subsets of regulatory elements bound by MITF in melanocytes. Analyses at single loci also support roles for LEF1, RB1, IRF4, and PAX3 acting in combination with MITF, while sequence motif analyses suggest that additional transcription factors colocalize with MITF at many melanocyte-specific regulatory elements. However, the precise biochemical functions of each of these MITF collaborators and their contributions to gene expression remain to be elucidated. Analogous to the transcriptional networks in morphogen-patterned tissues during embryogenesis, we anticipate that the level of MITF activity is controlled not only by the concentration of activated MITF, but also by additional transcription factors that either quantitatively or qualitatively influence the expression of MITF-target genes.

Keywords: MITF; SOX10; TFAP2A; YY1; melanocytes.

FIGURE 1

Protein domains of MITF, SOX10, YY1, and TFAP2A. (a–d) Protein domains and consensus binding motifs for human (a) MITF-M, (b) SOX10, (c) YY1, and (d) TFAP2A. Highlighted domains include those involved in transcriptional activation (red), DNA-binding (blue), dimerization (green), and repression (yellow). A red bar denotes the conserved PY motif in TFAP2A. Domain information is modified from MITF-M (Goding, 2000; Hartman & Czyz, 2015), SOX10 (Mollaaghababa & Pavan, 2003), YY1 (Atchison, 2014; Thomas & Seto, 1999), and TFAP2A (Eckert et al., 2005; Williams & Tjian, 1991). Consensus binding motif for MITF-M is modified from (Cheli et al., 2010), and all other motifs are modified from JASPAR (Sandelin, Alkema, Engstrom, Wasserman, & Lenhard, 2004)

FIGURE 2

The MITF rheostat model. Increasing levels of MITF activity (red), which is the product of both expression and post-translational modifications, drive the rheostat toward differentiation

FIGURE 3

Co-occupancy of MITF, SOX10, YY1, and TFAP2A at regulatory elements in melanocytes. (a) Overlap between peaks of MITF, SOX10, YY1, and TFAP2A binding in human melanocyte or melanoma cell lines. ChIP-seq data from MITF (Webster et al., 2014), SOX10 (Laurette et al., 2015), YY1 (Li et al., 2012), TFAP2A (Seberg et al., 2017). (b) Overlap between peaks of SOX10 and TFAP2A binding in mouse melan-a cells. ChIP-seq data from SOX10 (Fufa et al., 2015), TFAP2A (Seberg et al., 2017). (c) A model for regulatory elements occupied by MITF and multiple collaborators. These elements are frequently flanked by chromatin modifier BRG1 and H3K27ac marks (modified from Laurette et al., 2015)

FIGURE 4

Effects of MITF collaborators on the MITF rheostat. (a) Quantitative model: The presence of uniform activator transcription factors (blue), acting alongside or in combination with MITF (red), raises the level of the MITF rheostat at any given step. (b) Qualitative model: MITF has distinct collaborators for the regulation of genes specific to the proliferative/differentiating phenotype (e.g., SOX10, TFAP2A) and genes specific to the invasive phenotype (e.g., AP1, TEAD1)