Regulation of RUVBL1-RUVBL2 AAA-ATPases by the nonsense-mediated mRNA decay factor DHX34, as evidenced by Cryo-EM

Abstract

Nonsense-mediated mRNA decay (NMD) is a surveillance pathway that degrades aberrant mRNAs and also regulates the expression of a wide range of physiological transcripts. RUVBL1 and RUVBL2 AAA-ATPases form an hetero-hexameric ring that is part of several macromolecular complexes such as INO80, SWR1, and R2TP. Interestingly, RUVBL1-RUVBL2 ATPase activity is required for NMD activation by an unknown mechanism. Here, we show that DHX34, an RNA helicase regulating NMD initiation, directly interacts with RUVBL1-RUVBL2 in vitro and in cells. Cryo-EM reveals that DHX34 induces extensive changes in the N-termini of every RUVBL2 subunit in the complex, stabilizing a conformation that does not bind nucleotide and thereby down-regulates ATP hydrolysis of the complex. Using ATPase-deficient mutants, we find that DHX34 acts exclusively on the RUVBL2 subunits. We propose a model, where DHX34 acts to couple RUVBL1-RUVBL2 ATPase activity to the assembly of factors required to initiate the NMD response.

CryoEM of RUVBL1–RUVBL2–ZNHIT2, a complex that interacts with pre-mRNA-processing-splicing factor 8

“…We verified that all 3 RUVBL2 subunits in each RUVBL1–RUVBL2 hexamer were affected by these conformational changes by performing several analyzes that discarded an artefact of image processing ( Supplementary Figure S4 ). Curiously, and as in the case previously described for DHX34 helicase in complex with RUVBL1–RUVBL2 ( 12 ), all RUVBL1 subunits were unaffected by these large changes and still bound to nucleotide, suggesting that RUVBL1 probably does not perform the same function than RUVBL2 in the complex.…”

“…In addition, we cloned and expressed an N-terminal fragment (ZNHIT2-N; residues 1–123) (Figure 1A – C ) that included the zinc finger HIT domain (residues 12–48) whose structure has been determined by NMR ( 23 ), and a C-terminal fragment (ZNHIT2-C; residues 205–375) (Figure 1A – C ) that accumulates a large number of mutations associated to several types of cancer ( Supplementary Figure S1A ). We also purified several versions of the RUVBL1 and RUVBL2 ATPases: the individual RUVBL1 and RUVBL2 subunits, the preassembled RUVBL1–RUVBL2 complex, RUVBL1–RUVBL2 where the OB-fold region of each DII domain in both subunits was removed (RUVBL1–RUVBL2ΔDII), and the RUVBL1 E303Q –RUVBL2 E318Q ATPase deficient mutant, all of these following methods described before ( 12 ).…”

“…Human RUVBL1–RUVBL2 ΔDII as well as isolated RUVBL1, isolated RUVBL2 and RUVBL1 E303Q –RUVBL2 E300Q catalytic mutants were expressed and purified as reported ( 12 , 17 , 29 ).…”

“…RUVBL1 and RUVBL2 are closely related AAA-ATPases that form an hexameric RUVBL1–RUVBL2 ring of alternating subunits. Both subunits contain a domain, known as domain II (DII), that protrudes from each subunit in the RUVBL1–RUVBL2 hexamer and which has been found to interact with several proteins ( 8 , 11 , 12 ). The ATPase activity of RUVBL1 and RUVBL2, which resides in the AAA+ ring, is needed for the functions of R2TP ( 13 ), but what is the exact role of ATP hydrolysis and how this is regulated is mostly unknown.…”

“…Recent work has found that it is an N-terminal region of both RUVBL1 and RUVBL2 that blocks the exit route for nucleotides ( 8 , 13 , 17 , 19 ). DHX34, an RNA helicase that participates in nonsense-mediated mRNA decay (NMD) ( 12 ) and PIH1D1, one of the components of R2TP ( 8 ) can destabilize these N-terminal regions in RUVBL2 after binding to the DII domains. These conformational changes promote the release of the bound ADP from the nucleotide binding pocket ( 8 , 12 ).…”