A new mode of DNA binding distinguishes Capicua from other HMG-box factors and explains its mutation patterns in cancer.

Marta Forés,Leiore Ajuria,Lucía Simón-Carrasco,Matthias Drosten,Núria Samper,Gerardo Jiménez,Sergio González-Crespo,Mariano Barbacid,Ken M Cadigan

doi:10.1371/journal.pgen.1006622

Abstract

HMG-box proteins, including Sox/SRY (Sox) and TCF/LEF1 (TCF) family members, bind DNA via their HMG-box. This binding, however, is relatively weak and both Sox and TCF factors employ distinct mechanisms for enhancing their affinity and specificity for DNA. Here we report that Capicua (CIC), an HMG-box transcriptional repressor involved in Ras/MAPK signaling and cancer progression, employs an additional distinct mode of DNA binding that enables selective recognition of its targets. We find that, contrary to previous assumptions, the HMG-box of CIC does not bind DNA alone but instead requires a distant motif (referred to as C1) present at the C-terminus of all CIC proteins. The HMG-box and C1 domains are both necessary for binding specific TGAATGAA-like sites, do not function via dimerization, and are active in the absence of cofactors, suggesting that they form a bipartite structure for sequence-specific binding to DNA. We demonstrate that this binding mechanism operates throughout Drosophila development and in human cells, ensuring specific regulation of multiple CIC targets. It thus appears that HMG-box proteins generally depend on auxiliary DNA binding mechanisms for regulating their appropriate genomic targets, but that each sub-family has evolved unique strategies for this purpose. Finally, the key role of C1 in DNA binding also explains the fact that this domain is a hotspot for inactivating mutations in oligodendroglioma and other tumors, while being preserved in oncogenic CIC-DUX4 fusion chimeras associated to Ewing-like sarcomas.

Highlights

HMG-box factors are abundant nuclear proteins with highly diverse functions in the cell
We have focused on another HMG-box protein, Capicua (CIC), which has recently emerged as an important mediator of Ras/MAPK signaling in both Drosophila and mammals
The C1 and C-clamp domains are not related in sequence, we considered the possibility that C1 might recognize a specific conserved motif adjacent to the consensus CIC binding sites (CBSs)

Summary

Introduction

HMG-box factors are abundant nuclear proteins with highly diverse functions in the cell. They contain one or more HMG-box domains that bind the minor groove of DNA, bending the duplex away from the interaction site. Proteins with tandem HMG-box domains usually function as architectural and chromatin factors and do not exhibit DNA sequence specificity. Proteins with a single HMG-box domain, including Sox/SRY (Sox) and TCF/LEF1 (TCF) transcription factors, function as developmental regulators and bind specific AT-rich motifs in enhancers and promoters Sox proteins rarely act on their own and are often assisted by partner factors that bind next to the Sox sites, thereby stabilizing the complex and providing the specificity needed for in vivo function [3]. Once tethered to DNA, HMG-box proteins can exert their transcriptional effects through additional interactions with co-activators or co-repressors

Methods

Results

Conclusion