Deubiquitinating function of ataxin-3: Insights from the solution structure of the Josephin domain Yuxin Mao*, Francesca Senic-Matuglia†, Pier Paolo Di Fiore†‡§, Simona Polo†‡, Michael E. Hodsdon¶ʈ, and Pietro De Camilli*ʈ *Howard Hughes Medical Institute and Department of Cell Biology, Boyer Center for Molecular Medicine, Yale University School of Medicine, New Haven, CT 06510; ¶Department of Laboratory Medicine, Yale University School of Medicine, New Haven, CT 06510; and †Istituto Fondazione Italiana per la Ricerca sul Cancro di Oncologia Molecolare, ‡Istituto Oncologico Europeo, and §Dipartimento di Medicina, Chirurgia ed Odontoiatria, Universita’ degli Studi di Milano, 20122 Milan, Italy Contributed by Pietro De Camilli, July 25, 2005 Spinocerebellar ataxia type 3 is a human neurodegenerative components of the quality control and protein degradation disease resulting from polyglutamine tract expansion. The af- system. Ataxin-3 interacts with the valosin-containing protein fected protein, ataxin-3, which contains an N-terminal Josephin (VCP͞p97) (18, 19), a type II AAA ATPase that is associated domain followed by tandem ubiquitin (Ub)-interacting motifs with variety of cellular functions, including membrane fusion, (UIMs) and a polyglutamine stretch, has been implicated in the endoplasmic reticulum-associated degradation, and regulation function of the Ub proteasome system. NMR-based structural of the NF-␬B pathway (20). Ataxin-3 also interacts with HHR23 analysis has now revealed that the Josephin domain binds Ub (RAD23 in yeast) (19, 21), a ‘‘shuttle factor’’ responsible for and has a papain-like fold that is reminiscent of that of other translocating ubiquitinated proteins to the proteasome, and with deubiquitinases, despite primary sequence divergence but con- the proteasome itself (19). sistent with its deubiqutinating activity. Mutation of the cata- The polyQ stretch is variable in length and causes disease when lytic Cys enhances the stability of a complex between ataxin-3 expanded beyond a critical threshold. The polyQ stretch com- and polyubiquitinated proteins. This effect depends on the prises 10–40 Glu residues in unaffected individuals and 55–84 integrity of the UIM region, suggesting that the UIMs are bound Glu residues in Machado–Joseph disease patients (22). Similar to the substrate polyubiquitin during catalysis. We propose that expansions of polyQ stretches are responsible for at least nine ataxin-3 functions as a polyubiquitin chain-editing enzyme. inherited human neurodegenerative disorders, including Hun- tington’s disease. Proteins containing such expansions have an ͉ ͉ ͉ ͉ ataxia polyglutamine ubiquitin ubiquitin interaction motif increased propensity to misfold and aggregate (23–25), often valosin-containing protein forming intranuclear inclusions in affected cells. Accordingly, there is evidence that proteins involved in the control of protein pinocerebellar ataxia type 3, also known as Machado–Joseph folding and degradation can affect disease progression in animal Sdisease, is one of several hereditary autosomal dominant models (26, 27). The functional link of ataxin-3 to the ubiquitin- neurodegenerative disorders caused by expansion of a polyglu- proteasome system and to VCP͞p97 suggests a potential inter- tamine (polyQ) stretch in the affected gene product (1, 2). The play between the properties of these proteins and the pathogenic gene responsible for spinocerebellar ataxia type 3 encodes a role of its polyQ expansion. In fact, a very recent study demon- 42-kDa protein named ataxin-3. Ataxin-3 contains an N- strated that functions of ataxin-3 directly related to its property terminal Josephin domain (JD), a conserved module named to bind and process polyubiquitin suppress polyQ-induced de- after the Machado–Joseph disease, two ubiquitin (Ub)- generation in Drosophila (28). interacting motifs (UIMs), a polyQ stretch, and a short variable The proposed role of ataxin-3 as a deubiquitinating enzyme tail. In one splice variant, the C-terminal region ends with a third raises the question of a potential coordination between such putative UIM (3). catalytic activity and polyubiquitin binding by the UIMs. For The JD is present in at least 30 predicted proteins, including example, UIMs may help to recruit ataxin-3 to polyubiquitinated two human proteins comprising the JD alone. Computational substrates, orient such substrates relative to the JD, or mediate analyses resulted in two different predictions about the JD function. One study noted a distant structural homology to the a polyubiquitin editing function of ataxin-3. In addition, given epsin N-terminal homology͞AP180 N-terminal homology do- the lack of primary sequence similarity of ataxin-3 to other mains (4–6) present in adaptor proteins implicated in endocy- deubiquitinating enzymes, it would be of interest to elucidate the tosis (7). Another study identified signature motifs for Cys structural basis of its enzymatic activity. To begin addressing proteases in the JD and suggested that it may function as a these questions, we have performed NMR studies and biochem- deubiquitinating enzyme (8). Indeed, ataxin-3 has been shown ical analyses. We report that the tertiary structure of the ataxin-3 recently to have deubiquitination activity, which was abolished JD is strikingly similar to other deubiquitinating enzymes, such by mutation of a Cys (C14) in the putative catalytic site (9, 10). as UCH-L3. Our results also suggest cooperation between the A recent classification defined the ataxin-3͞Josephin family as JD and the UIMs during catalysis and are consistent with an one of five distinct families of deubiquitinating enzymes (11). editing function for this enzyme. Collectively, our results define The UIM is a conserved 15-aa motif first discovered as a polyubiquitin binding domain in the S5a proteasome subunit (12). This motif was subsequently identified in proteins involved Freely available online through the PNAS open access option. in endocytic traffic, as well as in other components of the Ub Abbreviations: JD, Josephin domain; NOE, nuclear Overhauser effect; polyQ, polyglu- tamine; Ub, ubiquitin; UIM, Ub-interacting motif; VCP, valosin-containing protein. proteasome system, such as deubiquitinating enzymes and Ub Data deposition: The JD chemical shift assignments have been deposited in the BioMagRes- ligases (13–15). Through its UIMs, ataxin-3 binds to polyubiq- Bank (BMRB accession no. 6742). Coordinates for the ensemble of the JD structures have uitin chains but not to diubiquitin or monoubiquitin (16, 17), been deposited in the Protein Data Bank (PDB ID code 2AGA). suggesting that it may function in the protein surveillance ʈTo whom correspondence may be addressed. E-mail: [email protected] or pathway leading to the proteasome. This possibility is further [email protected]. supported by the reported interaction of ataxin-3 with other © 2005 by The National Academy of Sciences of the USA 12700–12705 ͉ PNAS ͉ September 6, 2005 ͉ vol. 102 ͉ no. 36 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0506344102 Downloaded by guest on September 26, 2021 ataxin-3 as a critical component of the polyubiquitin-dependent Restraints on the backbone dihedral angles were derived from pathways that control protein folding and stability. an analysis of backbone chemical shifts with the TALOS program (31). Hydrogen bonds were identified during the later stages of Materials and Methods structure determination based on the consistent proximity of Antibodies and Reagents. Antibodies against ataxin-3 were gen- hydrogen bonding partners in the calculated ensembles. Nuclear erated in rabbits by using recombinant GST-JD (amino acids Overhauser effect (NOE) correlations between nearby protons 1–198) and affinity-purified. Monoclonal anti-Ub antibodies were identified in 3D 15N-NOESYHSQC, 13C-NOESYHSQC (P4D1) were from Santa Cruz Biotechnology, and monoclonal (aromatic), and 13C-NOESYHSQC (aliphatic) NMR spectra. anti-Flag M2 antibodies were from Sigma. Structure calculations were ultimately performed with the CYANA software package (32) based on NOEs that were inter- Cloning, Mutagenesis, and Protein Preparation. A cDNA encoding preted and calibrated by the program CANDID (33). Manually full-length human ataxin-3 (Machado–Joseph disease 1-1 splice interpreted NOE restraints based primarily on the use of sym- variant carrying a polyQ sequence of 22 residues) was amplified metry-related 3D NOE crosspeaks during the iterative cycles of by PCR with Integrated Molecular Analysis of Genomes and NOE interpretation and structure calculation were added to Their Expression (I.M.A.G.E.) clone ID 4393766 as a template. assist structural convergence. The final series of structure cal- The cDNA was subcloned in pGEX-6P-1 and pET43 for expres- culations used the CANDID-derived NOE restraints and the sion in bacteria and in pcDNA-FLAG vectors for mammalian previously described hydrogen bond and backbone torsion angle expression. The JD domain from the same proteins (amino acids restraints. After an initial brief minimization of randomized 1–193) was obtained in a similar way. All of the mutants were conformations, simulated annealing began with 5,000 steps of generated by site-directed mutagenesis, and the constructs were molecular dynamics at high temperature, followed by 35,000 verified by sequencing. dynamics steps for cooling and a final 10,000 steps of conjugate GST and His-6 fusion proteins were produced according to gradient minimization. Generally, 50 structures were