Functions and interactions of the carboxyl terminus of the *Myc antagonist Mxi1

Abstract

The Mxi1 protein, a member of the Mad/Mxi1 sub-family of the greater Myc network, has been characterized as a Myc antagonist, transcriptional repressor, and candidate tumor suppressor protein. On the molecular level, Mxi1 can compete with Myc for the common DNA binding partner Max, as well as for a subset of genetic targets bearing the Myc/Max (Mxi1/Max) E-box consensus. Through the recruitment of the Sin3/HDAC complex by its amino terminal domain, Mxi1 can effect transcriptional repression at these genetic targets that are alternatively transactivated by Myc/Max. Findings from our and other groups have supported the existence of a second repression domain encoded by the Mxi1 carboxyl terminus; this domain has been shown to effect repression at E-box sites as well as at the c-myc promoter, and appears to do so in a Sin3-independent fashion.;In an attempt to gain insight into the mediators of the transcriptional repression function of the Mxi1 carboxyl terminus, we performed yeast two hybrid screens using this region as bait and isolated a previously described protein called Yaf-2. Both in vitro and in vivo, Mxi1 can recruit Yaf-2 and its related family member RYBP. While the precise function of the RYBP and Yaf-2 proteins remains to be fully elucidated, it is notable that RYBP has been implicated as a putative adapter protein that can recruit Polycomb group (PcG) transcriptional repressors to DNA binding proteins. To determine if one function of RYBP's association with Mxi1 could be for the targeting of PcG proteins, we employed yeast three-hybrid assays. Our results indicate that RYBP can serve as a bridging factor between the Mxi1 carboxyl terminus and MPc2, a mammalian homolog of Drosophila Polycomb. This finding prompted us to assess whether Mxi1 was also capable of existing in complexes with PcG proteins in vivo. Indeed, co-immunoprecipitation studies showed that Mxi1 could associate with various members of the PRC1 Polycomb complex including hPc2, M33, Mel-18, Bmi-1 and Ring1A. We pursued further the interactions between Mxi1 and hPc2, and Mxi1 and M33. Importantly, hPc2 and M33 can enhance transcriptional repression by Mxi1 but not Mxi1 deleted for its carboxyl terminus as assessed in transient transfection reporter assays.;In conclusion, we have uncovered another mode of repression mediated by the carboxyl terminal domain in Mxi1, in addition to the Sin3/HDAC type of repression brought about by Mxi1's amino terminus. It is especially intriguing that Mxi1 engages PcG proteins, since the Myc proteins that Mxi1 antagonizes can recruit members of the trithorax group of proteins that antagonize PcG proteins. This scenario closely resembles that which has already been described wherein Mxi1 and Myc can respectively recruit the Sin3/HDAC and TRRAP/HAT complexes that also work in opposition. Beyond this, our findings contribute to an emerging list of cancer/chromatin connections that involve proteins, like Polycomb group members, that modify chromatin non-covalently. Ultimately, these connections are likely to expand the spectrum of epigenetic changes that are seen in the evolution of human cancers and could suggest new targets for therapeutic intervention.