Identification of CHMP2B as a Substrate Adaptor for the E3 Ubiquitin Ligase CHIP

Abstract

SCA48 causing mutations in the STUB1 gene which code for the protein CHIP cause the rare neurodegenerative diseases, SCAR16 and SCA48. Most of the SCA48 causing TPR mutations map to the interface between the first and second TPR motif of the CHIP TPR domain which is the binding site of CHIP and chaperone proteins. Chaperone proteins bind the CHIP TPR domain via their C-terminal tail, which truncates with an I/MEEVD sequence. The C-terminal aspartic acid of chaperones is critical for binding CHIP due to key asparagine and lysine residues in the CHIP TPR domain forming a carboxylate clamp around the -COOH of the aspartic acid. We were curious whether the CHIP TPR domain was important for CHIP binding to other C-terminal aspartic acid residue (CAspR) proteins that were cellularly relevant. Using biotin ligase proximity labeling with wild type CHIP and a SCA48 causing CHIP TPR mutant we identified several potential “non-canonical” interactors of CHIP. One of these CAspRs, the protein CHMP2B is a component of the ESCRT-III complex. Interestingly, C-terminal mutations that abrogate the -VD motif of CHMP2B cause frontotemporal dementia and amyotrophic lateral sclerosis (FTD/ALS). This led us to hypothesize that the CHIP TPR domain and the C-terminal -VD motif of CHMP2B are necessary for the CHIP/CHMP2B interaction to occur. I used an array of cellular, biochemical, and biophysical assays to validate that the CHIP/CHMP2B interaction is dependent on the CHIP TPR domain and the C-terminal of CHMP2B. Using these same techniques, I was able to show that there is a second highly charged region on CHMP2B that binds to the CHIP TPR domain, albeit with decreased affinity. This second binding site allows CHMP2B to compete with the chaperone, Hsp70 for CHIP binding. I was able to show through co-immunoprecipitation (Co-IP) experiments that CHMP2B acts as a substrate adaptor allowing CHIP to ubiquitinate of another ESCRT-III protein, CHMP4B. To our knowledge this is the first example of a non-canonical CAspR protein mediating substrate specificity of the E3 ligase CHIP. I was also able to show through Co-IP, that the addition of the wild type CHMP2B C-terminal -ALGVD sequence onto FTD/ALS causing CHMP2B mutations restores CHMP4B ubiquitination. Additionally, using ubiquitin linkage specific antibodies, I was able to show that the CHMP4B ubiquitination mediated by the CHIPCHMP2B complex is K63- linked. Interestingly, I found that the expression of SCA48 causing CHIP TPR mutations resulted in an aberrant endosomal phenotype, similar to what is seen with the expression of FTD/ALS causing C-terminal CHMP2B mutations. I showed that these enlarged endosomes were Rab7 and LAMP1 positive indicating a malfunction in late endosome processing consistent with a malfunction in the ESCRT-III machinery of which CHMP2B and CHMP4B are essential components. Furthermore, I was able to rescue the enlarged endosome phenotype with the addition of the wild type CHMP2B C-terminal -ALGVD sequence onto the C-terminal of the FTD/ALS causing CHMP2B mutant CHMP2B Q165X. This indicates that the -ALGVD sequence is necessary for the CHIP/CHMP2B interaction and that CHMP2B imparts substrate specificity allowing for K63-linked CHMP4B ubiquitination which is necessary for proper late endosomal trafficking. These results suggest a novel role for CHIP and CHMP2B in regulating endosomal trafficking. Developing mechanisms to ubiquitinate CHMP4B in the case of SCA48 causing CHIP mutations or FTD/ALS causing C-terminal CHMP2B truncation mutations may be a worthwhile strategy in the treatment of these two neurodegenerative diseases.