Battle at the p15 promoter

The company we keep often affects our behaviour, and the same is true of the oncoprotein Myc: in the company of Max, Myc can activate the transcription of genes that generally stimulate proliferation; but Myc can also behave as a trancriptional repressor. The proteins that bring out this side of Myc's personality have remained elusive, but two papers in the April issue of Nature Cell Biology identify them.

Miz-1 is a zinc-finger-containing protein that induces cell-cycle arrest and activates the transcription of several genes that are repressed by Myc. Staller and colleagues reasoned that Miz-1 might activate transcription of a cyclin-dependent kinase inhibitor, and used PCR and western blots to identify p15^Ink4b as a target of Miz-1. Deletion mapping identified a Miz-1-binding region in the promoter of Cdkn2b, the gene that encodes p15^Ink4b. Co-expression of Myc blocked transactivation of Cdkn2b by Miz-1, and co-immunoprecipitations identified a complex of Miz-1, Myc and Max. But how does Myc block the ability of Miz-1 to activate transcription of Cdkn2b? The authors wondered whether Myc might prevent Miz-1 from recruiting a co-activator and their speculations proved correct: inactivation of the co-activator protein p300 inhibited Miz-1's ability to activate Cdkn2b transcription, and Miz-1 co-immunoprecipitated with p300. Mapping of the p300-binding site on Miz-1 showed that Myc and p300 both bind to overlapping sites on Miz-1, explaining how Myc blocks reruitment of p300 to the Cdkn2b promoter.

But does Myc use this mechanism to make cells proliferate? To find out, the authors made chimeras of Myc and Mad-1 that have Myc's transcriptional activator activity, but can't bind Miz-1. Unlike wild-type Myc, the chimeras couldn't block accumulation of p15^Ink4b, and couldn't transform p53^−/− cells. Myc's oncogenic abilities threfore depend on its ability to block Cdkn2b transcription.

Transforming growth factor-β ( TGF-β) can also induce expression of p15^Ink4b by activating Smad transcription factors, but why this response is blocked by overexpression of c-Myc has remained a mystery. Joan Seoane and colleagues now explain why. Treatment of keratinocytes with TGF-β downregulated Myc and decreased levels of the Myc–Miz-1 complex. To see whether this upregulates transcription of Cdkn2b, they created a dominant-negative form of Miz-1, Miz-dZF, that binds Myc but not the Cdkn2b promoter. Overexpression of Miz-dZF increased expression from a Cdkn2b reporter contruct, but not as well as treatment with TGF-β. So TGF-β must activate transcription of Cdkn2b, as well as relieving repression of its transcription by Myc. Reasoning that Smads were the most likely candidates for this bipartite effect of TGF-β, the authors located a Smad-binding region in the Cdkn2b promoter, and immunoprecipitated a Smad-containing complex from the region. Forced expression of Myc didn't prevent formation of this complex, but it did block its ability to transactivate Cdkn2b, so Myc's ability to repress transcription dominates the ability of Smads to activate it. The authors identified a ternary complex of Myc, Miz-1 and Smad4, so Myc and Smads don't battle it out by binding to the same site on Miz. The most likely explanation is that Myc's ability to represss transcription — at least of Cdkn2b — is due to its ability to block the interaction between Miz-1 and the co-activator p300. Whether this is a general mechanism by which Myc represses transcription is an exciting possibility.