page 35, Loercher et al. show that, in melanocytes, the link is the MITF transcription factor.
MITF was already known for its differentiation-inducing activity. Now it is also shown to slow cell growth by activating transcription of a cell cycle inhibitor gene (in addition to pigment and melanocyte survival genes). This mitotic inhibitor, p16Ink4a, arrests cells in G1 by blocking phosphorylation of Rb. Hypophosphorylated Rb binds to E2F and thus prevents it from activating cell cycle progression genes.
The cell cycle arrest is needed for differentiation, as precursor cells lacking p16Ink4a did not show features of melanocyte differentiation in response to MITF. The authors speculate that free E2F, which accumulates when p16Ink4a is inactive and Rb is phosphorylated, may transcribe genes that repress differentiation as well as genes that promote cell cycle progression. Alternatively, differentiation may be jumpstarted (or made possible) by long-lasting chromatin remodeling (e.g., via the recruitment of polycomb group proteins) that occurs when cells permanently exit from the cell cycle.
Later, MITF is needed to maintain the quiescent state. RNAi-induced loss of MITF inhibited expression of INK4A (the gene that encodes p16Ink4a) and sent differentiated melanocytes back into the cell cycle. Cultured melanocytes occasionally escaped from the cell cycle block on their own by inactivating p16Ink4a. Many natural melanomas are also deficient in INK4A expression. The selective pressure to proliferate probably favors mutation of INK4A over MITF, as the latter is needed for transcription of survival genes.