Mechanism and chain specificity of RNF216/TRIAD3, the ubiquitin ligase mutated in Gordon Holmes syndrome. (24th April 2019)

• Record Type: Journal Article
• Authors: Seenivasan, Ramkumar; Hermanns, Thomas; Blyszcz, Tamara; Lammers, Michael; Praefcke, Gerrit J K; Hofmann, Kay
• Is Part Of: Human molecular genetics. Volume 28:Number 17(2019)
• Page Start: 2862
• Page End: 2873
• Publication Date: 2019-04-24
• Subjects: Human molecular genetics -- Periodicals; Human chromosome abnormalities -- Periodicals; 572.8
• DOI: 10.1093/hmg/ddz098
• Languages: English
• ISSNs: 0964-6906

Abstract

Abstract: Gordon Holmes syndrome (GDHS) is an adult-onset neurodegenerative disorder characterized by ataxia and hypogonadotropic hypogonadism. GDHS is caused by mutations in the gene encoding the RING-between-RING (RBR)-type ubiquitin ligase RNF216, also known as TRIAD3. The molecular pathology of GDHS is not understood, although RNF216 has been reported to modify several substrates with K48-linked ubiquitin chains, thereby targeting them for proteasomal degradation. We identified RNF216 in a bioinformatical screen for putative SUMO-targeted ubiquitin ligases and confirmed that a cluster of predicted SUMO-interaction motifs (SIMs) indeed recognizes SUMO2 chains without targeting them for ubiquitination. Surprisingly, purified RNF216 turned out to be a highly active ubiquitin ligase that exclusively forms K63-linked ubiquitin chains, suggesting that the previously reported increase of K48-linked chains after RNF216 overexpression is an indirect effect. The linkage-determining region of RNF216 was mapped to a narrow window encompassing the last two Zn-fingers of the RBR triad, including a short C-terminal extension. Neither the SIMs nor a newly discovered ubiquitin-binding domain in the central portion of RNF216 contributes to chain specificity. Both missense mutations reported in GDHS patients completely abrogate the ubiquitin ligase activity. For the R660C mutation, ligase activity could be restored by using a chemical ubiquitin loading protocol that circumvents the requirement for ubiquitin-conjugating (E2) enzymes. This result suggests Arg-660 to be required for the ubiquitin transfer from the E2 to the catalytic cysteine. Our findings necessitate a re-evaluation of the previously assumed degradative role of RNF216 and rather argue for a non-degradative K63 ubiquitination, potentially acting on SUMOylated substrates.