Skip Navigation


Human Molecular Genetics Advance Access originally published online on August 27, 2003
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
12/21/2845    most recent
ddg297v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (39)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Scheel, H.
Right arrow Articles by Hofmann, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Scheel, H.
Right arrow Articles by Hofmann, K.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Human Molecular Genetics, 2003, Vol. 12, No. 21 2845-2852
DOI: 10.1093/hmg/ddg297
© 2003 Oxford University Press

Elucidation of ataxin-3 and ataxin-7 function by integrative bioinformatics

Hartmut Scheel, Stefan Tomiuk and Kay Hofmann*

Bioinformatics Group, Memorec Biotec GmbH, Stöckheimer Weg 1, D-50829 Köln, Germany

Received July 17, 2003; Revised August 6, 2003; Accepted August 18, 2003


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
The spinocerebellar ataxias (SCAs) are a class of hereditary neurodegenerative diseases, which are caused by the pathological expansion of unstable CAG triplet repeats found in a number of apparently unrelated genes. The proteins encoded by the SCA genes typically translate this expanded (CAG)n repeat into an expanded poly(Q) stretch. Several pathological features are common to all SCAs, irrespective of the gene harbouring the expansion. The specific contributions of the mutated genes are currently hard to assess, as the physiological role of most of the so-called ataxins is not known. By combining the results of profile-based sequence analysis with genome-wide functional data available for model organisms, we have derived detailed predictions of the physiological function of two SCA gene products. Ataxin-3, the protein mutated in Machado Joseph Disease (SCA3), belongs to a novel group of cysteine-proteases and is predicted to be active against ubiquitin chains or related substrates. The catalytic site of this enzyme class is similar to that found in UBP and UCH type ubiquitin proteases. For ataxin-7, the gene product of the SCA7 gene, we have identified an orthology relationship to the yeast open reading frame Ygl066c. Recently published evidence from genome-wide studies suggests that Ygl066c is a component of the SAGA histone acetyltransferase complex. By analogy, a similar role for the mammalian ataxin-7 can be expected. The functional predictions reported here are sufficiently precise to allow a direct experimental verification. Moreover, both findings have implications for the general pathogenesis of spinocerebellar ataxias by providing a direct connection of these diseases with ubiquitin metabolism and histone acetylation.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
It is a frequently occurring situation that genes causative for a disease state are identified by positional cloning approaches, yet the molecular function of the associated gene products—and thus the pathogenic mechanism—remains obscure. Contemporary bioinformatics approaches frequently allow a reliable prediction of disease gene function, ideally providing a prediction detailed enough for a straightforward experimental validation. In the past, this bioinformatical approach had to rely exclusively on the detection of homology to other proteins that are better characterized and allowed the prediction of some functional analogies. While this component continues to be of crucial importance, there is now a wealth of genomic information available that can be successfully tapped for the purpose of functional prediction. Useful data sources include yeast two-hybrid interaction data, protein complex data, genetic interaction data, structural data and co-expression data. Here, we demonstrate this integration of genomic information using two examples from the field of neurodegeneration.

Ataxin-3 (MJD1) and ataxin-7 are the protein products encoded by two genes mutated in different forms of spinocerebellar ataxia, SCA3 and SCA7 (1,2). Both proteins harbour an uninterrupted stretch of multiple glutamine residues, which is considerably longer in the disease state as a consequence of the pathological expansion of an unstable (CAG)n triplet repeat found in the coding portion of the SCA genes. The expanded poly(Q) region in the protein products is thought to contribute crucially to SCA pathogenesis, as these proteins induce the formation of intranuclear inclusion bodies. The mutated proteins are found within these inclusions, typically associated with other components such as ubiquitin and chaperones. Despite a lot of research into this topic, the pathological mechanism of this class of diseases is not entirely clear and several pathogenesis models exist, which are not mutually exclusive. Poly(Q) inclusions are known to sequester other glutamine-rich proteins, which might be required in soluble form for cell viability (3–5). In model organisms, poly(Q) toxicology can be rescued by overexpression of chaperones, suggesting an involvement of the cellular quality control mechanism in pathogenicity, an idea supported by the presence of ubiquitin and proteasome components in the inclusion bodies. Finally, a loss-of-function of the mutated protein might contribute substantially to the disease. In most cases, this latter contribution cannot really be assessed as the function of many ataxins is not known.

By combining the generalized sequence profile method with the generation of multiple alignments from structural superposition, we found that ataxin-3 (MJD1), the protein mutated in Machado Joseph Disease, is distantly related to a class of ubiquitin-specific proteases. The known active site residues of this protease class are perfectly conserved in ataxin-3 and its close relatives, suggesting that the MJD1 gene product encodes a catalytically active cysteine protease. Based on the domain context and the published interaction data, we predict that ataxin-3 is active against poly-ubiquitin chains or related substrates. For ataxin-7, which does not bear any resemblance to ataxin-3 except for the expandable poly(Q) stretch, we found a number of conserved homology domains, including a novel Zn-finger motif. The same domain architecture is shared by Ygl066w, a functionally uncharacterized protein from Saccharomyces cerevisiae. Based on the published interaction properties of this yeast protein, Ygl066w has been predicted to be a component of the general transcription machinery, more specifically of the SAGA histone acetyltransferase complex. By analogy, a similar role for the mammalian ataxin-7 is to be expected, which makes this protein an excellent candidate for the proposed link between poly(Q) pathology and histone acetylation.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Ataxin-3
In a previous attempt to derive hints about ataxin-3 function from its polypeptide sequence, we reported the presence of multiple copies of the UIM (ubiquitin interaction motif) in the C-terminal portion of the protein (6). UIMs are short conserved motifs of about 11 amino acids that occur in many proteins working in pathways of protein ubiquitination and ubiquitin-recognition (6,7). In the course of this published work it became obvious that the human genome encodes several ataxin-3-like proteins and that the N-terminal part of ataxin-3 is the most highly conserved portion of the protein. We were thus interested to learn more about the functional role of this protein domain.

In the first step, current versions of the protein database were searched for proteins with similarity to the ataxin-3 N-terminal domain. In order to obtain the full complement of mammalian ataxin-3-like proteins, the Ensembl version of the human genome database (8) and translated versions of the current mammalian EST databases were also searched. As shown in Figure 1, it was possible to identify four ataxin-3-like proteins in mammals. Besides ataxin-3 there is one closely related protein, which we call ataxin-3-like 1 (Atx3L1), sharing the same domain architecture comprising the conserved N-terminal domain followed by multiple UIMs and a poly(Q) stretch. Two shorter proteins that have been published before as josephin-1 and josephin-2 consist of not much more than the N-terminal domain (9). Nematodes and plants seem to possess two ataxin-3-like proteins, one of them corresponding to the long form, the other to the short josephin-form, while insects only have a copy of the latter. There is no evidence for ataxin-3-like proteins in yeasts, but there are two proteins of that class found in the parasite Plasmodium falciparum. In the longer plasmodium protein, the conserved ataxin-3 N-terminal domain is followed by a ubiquitin-like UBX domain instead of the UIM motifs (Fig. 1A). In the hope of finding more distantly related members of this family with some degree of functional annotation, we constructed generalized profiles from the multiple alignment shown in Figure 1B. Unfortunately, no related proteins could be identified in this screen, using the stringent criterion of P<0.01 (data not shown).



View larger version (63K):
[in this window]
[in a new window]
  		Figure 1. (A) Multiple alignment of the ataxin-3 N-terminal domain. Positions invariant or conservatively replaced in at least 50% of the sequences are shown on black and grey background, respectively. The first two columns indicate the gene name and the species abbreviation (Hs, Homo sapiens; Ce, Caenorhabditis elegans; At, Arabidopsis thaliana; Dm, Drosophila melanogaster; Pf, Plasmodium falciparum). The Swissprot/TrEMBL accession numbers of the sequences are shown in the last column. The top line contains the PsiPred secondary structure prediction for ataxin-3, H denoting alpha-helices and E extended structures. The PredictProtein/PROF prediction largely agrees with PsiPred and is not shown. (B) Domain organisation of representative ataxin-3 related proteins. The N-terminal box labelled ‘catalytic’ refers to the putative protease domain discussed in this manuscript. The dark-coloured ovals in the C-terminal region of ataxin-3 denote multiple copies of the UIM motif. The third copy is only present in selected splice forms and is shown in lighter colour. The box labelled ‘UBX’ denotes a ubiquitin-like UBX domain in the C-terminus of the Plasmodium version of ataxin-3. (C) Evolutionary relationship of the ataxin-3/Josephin family, as determined by neighbour-joining dendrogram analysis.

 
However, we serendipitously obtained useful similarity information on the ataxin-3 N-terminal domain family by using a completely different screen that was originally targeted for identifying new activating proteases for ubiquitin and related modifiers. After exhaustive application of conventional generalized profile searches starting from multiple alignments of the UCH protease family, we tried to make the profile searches more sensitive to very distantly related outlier sequences by incorporating information from three-dimensional structures into the profile construction process. According to the FSSP database (10), which holds information on structural relationships calculated by the Dali algorithm, the closest known structural neighbour of the UCH-L3 protease (PDB:1UCH) is a leader protease (PDB:1QMY) from the foot and mouth disease virus. The corresponding sequences do not show any recognizable similarity, although the general fold and the active site geometry of the two enzyme classes are clearly related. The spdbv program (11) was used to calculate a rigid-body superposition of the two structures. From this superposition, a structurally valid two-sequence alignment was derived; a number of manual adjustments were required to overcome the rigid-body limitations. Starting from this, generalized profile searches were performed resulting in the incorporation of other UCH proteases and members of the viral protease class into the structure-based alignment.

In this database search, several high-significance matches to members of the UBP family of ubiquitin-specific proteases were obtained. The significant sequence relationship between the UCH and UBP classes of ubiquitin proteases was somewhat surprising. However, the recently published structure of HAUSP (12), the first structurally characterized UBP protease, clearly demonstrates the structural relationship and active site correspondence between these two protease classes. Thus, the significant profile scores obtained in our screen are biologically meaningful and underscore the suitability of our approach. We further refined the structure-based profile searches by also including members of the UBP family into the training set. The resulting profile, consisting of members of the three protease classes, found significant matches to several members of the ataxin-3 family described above. While not all of the structural features of the UCH/leader protease/UBP profile can be reliably mapped to the ataxin-3 sequence, the catalytically most important regions can be aligned with high confidence (Fig. 2). The catalytic cysteine residue corresponds to C14 of human ataxin-3, while the proton-donating histidine residue corresponds to the H119 position. As the third residue of the catalytic triad, an aspartate is observed in all UCH and viral proteases, while in UBP proteases either aspartate or asparagine can be found in that position. Members of the ataxin-3 family share this variability: ataxin-3 itself carries an asparagine (N134) while most other ataxin-3 like proteins use an aspartate instead.



View larger version (37K):
[in this window]
[in a new window]
  		Figure 2. Multiple alignment of representative members of the protease families discussed in the manuscript. Shading of conserved residues and species abbreviations are analogous to Figure 1A. The topmost block contains human ataxin-3 and a plant josephin sequence, the second block contains leader proteases from foot and mouth disease virus (FMDV) and equine rhinitis virus (ERV). The third block contains two UCH type proteases (yeast Yuh1 and human UCH-L3). The bottom block contains a divergent set of UBP-type proteases. Residues important for catalysis are indicated by their consensus symbols below the alignment.

 
Based on the statistically significant profile comparison scores and the conservation of the catalytically important residues, it is predicted that all members of the ataxin-3 family are cysteine proteases assuming the same fold as UCH, UBP and FMDV leader proteases, i.e. the papain fold (13). The UBP-type proteases contain a well-conserved asparagine or glutamine residue shortly upstream of the catalytic cysteine, which balances the oxyanion hole of the transition state, probably in a concerted manner with another hydrogen bond providing group from either a backbone imino- or side chain amino-group (12); this feature is also shared by papain and most other proteases belonging to this fold (13). Ataxin-3 and its relatives carry a highly conserved glutamine residue at the corresponding position, which is likely to assist in the catalytic reaction. The substrate specificity of ataxin-3 cannot be derived from this kind of bioinformatical analysis. However, the presence of a UIM domain in ataxin-3 and the presence of a UBX domain in the plasmodium homolog suggest a role of this protein family in ubiquitin-dependent pathways.

Ataxin-7
Ataxin-7, the protein product of the SCA7 gene, is a protein of 892 amino acids, which carries an expandable poly(Q) region close to the N-terminus. No homologs of ataxin-7 in species other than human have been described and no functional information on that protein is available. Sequence database searches with BLAST (14) identified a number of closely related sequences from vertebrates, including two human paralogues, plus a number of dubious matches of borderline significance, several of which seem to be caused by the extended compositionally biased regions of the ataxin-7 protein sequence. Comparison of ataxin-7 with its ortho- and paralogues from various organisms revealed the existence of three conserved sequence blocks indicated in Figure 3. In an attempt to find more distantly related, but functionally characterized, homologues of ataxin-7, we used generalized profile searches starting from multiple alignments of the three conserved regions. Two of those regions, spanning ataxin-7 residues from 132–176 and 341–387, respectively, yielded highly significant scores to ORFs from both fission yeast and budding yeast (Fig. 3B). In particular, the second block was found to be highly conserved and reached significance values P<10-6. The third block conserved in the human ataxin-7 paralogue family was absent in the group of yeast proteins. A closer inspection of the conservation patterns in blocks 1 and 2 revealed a high number of invariant residues with the potential to coordinate metal ions. Block 1 contains five totally conserved positions, two of them occupied by cysteine and one by histidine. The better conserved block ‘2’ has 18 invariant residues, among them three cysteines, two aspartates and one histidine. A conservation pattern like this is typical of metal-binding proteins, including metallo-enzymes.



View larger version (44K):
[in this window]
[in a new window]
  		Figure 3. (A) Domain architecture of the three human members of the ataxin-7 family (ataxin-7, KIAA1218 and an unnamed protein with Swissprot accession number Q8IV05) and the related yeast protein Ygl066w. Conserved blocks labelled ‘1’ and ‘2’ are found in all members of the family, block ‘3’ is missing in the yeast protein. The S. pombe orthologue (ORF-name C126.04c) shares the Ygl066w architecture. (B) Multiple alignment of the conserved blocks 1 and 2. Shading of conserved residues is analogous to Figure 1A. The putative metal coordinating residues are indicated below the alignment.

 
The identification of a Saccharomyces homolog of ataxin-7 opens new avenues to the elucidation of ataxin-7 function, as a large body of genome-wide functional data is available for yeast (15). Although Ygl066w, the budding yeast orthologue of ataxin-7, has not been addressed by targeted research so far, it has been covered by two informative systematic studies: first, the published Cellzome screen for multi-protein complexes in yeast (16) has identified Ygl066w as a component of two complexes with widely overlapping subunit composition. Complex 420 using Spt7 as a bait, and complex 442 using Taf90 as bait, were found to contain Ygl066w together with a number of other proteins shown in Table 1.


View this table:
[in this window]
[in a new window]
  		Table 1. Yeast complexes containing Yg1066w
 
Most proteins found in the same complexes as Ygl066w are well-known components of both the TBP-associated complex and of the SAGA histone acetyltransferase complex. Even more informative was a recently published study by Sanders et al. (17) using a similar approach but focussing on TBP-associated proteins. According to their data, Ygl066w interacts exclusively with the TAF proteins of the SAGA complex (which are also found in the TBP complex), but not with those occurring in the TBP complex alone. The obvious explanation, that Ygl066w is a new component of the yeast SAGA complex, was corroborated by additional interaction studies (17). The SAGA complex is a multi-subunit assembly conserved from yeast to human, which regulates transcription by histone acetylation. Although no functional information on ataxin-7 and its paralogues is available, there is a high probability that all of these proteins are components of SAGA-like histone acetyltransferase complexes, maybe in a tissue-specific manner.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
Ataxin-3
Ataxin-3 is the protein mutated in Machado Joseph Disease (MJD, SCA3), the most common form of hereditary spinocerebellar ataxias. Based on the results of our sequence analysis, we predict ataxin-3 to be a cysteine protease assuming the papain fold, that is structurally and functionally related to two quite divergent subfamilies of ubiquitin-specific proteases, the UCH and the UBP enzymes. Several other pieces of evidence provide additional links between ataxin-3 and the protein ubiquitination pathway: (i) the inclusion bodies of SCA3 contain tightly bound ubiquitin and components of the proteasome (18); (ii) wild-type ataxin-3 contains two or three copies of the ubiquitin-binding UIM motif (6) and has been found to physically interact with Rad23, a ubiquitin-domain containing protein; (iii) the plasmodium homolog of ataxin-3 contains a C-terminal UBX domain. UBX domains were originally identified as conserved regions of unknown function that frequently associated with ubiquitin-binding UBA domains (19). Nowadays, UBX domains are known to be structurally related to ubiquitin; they also exhibit a similar interaction pattern (20).

The conservation pattern of the ataxin-3 family does not allow the prediction of the enzymatic substrates. Even the sequence relationship to the ubiquitin-specific UCH and UBP families does not provide a direct link to substrate preference, as this group also includes the viral leader peptidases that are not active against ubiquitin. Nevertheless, the published biochemical properties of ataxin-3, combined with the domain architecture of ataxin-3 and its close relatives suggest an involvement in the ubiquitin pathway. An obvious hypothesis for ataxin-3 function would thus predict a proteolytic activity against components of the ubiquitin pathway, possibly against ubiquitin itself. Several ubiquitin proteases contain ubiquitin-binding domains in addition to the catalytic portion of the enzyme. The role of these ancillary domains has not been firmly established so far. We favour the hypothesis that the non-catalytic domains of UBPs have a role in substrate recognition. The presence of a non-catalytic ubiquitin-binding domain would, according to this hypothesis, suggest a specificity for polyubiquitin chains, potentially with preference of one chain topology over the others. Some support for this idea can be derived from the chain-preferring properties of isopeptidase T, a UBP-type ubiquitin protease with a ubiquitin-binding UBA domain in its non-catalytic portion (21).

If the prediction holds true, ataxin-3 would join the ranks of several other deubiquitinases associated with neurodegenerative diseases: (i) the ubiquitin C-terminal hydrolase UCH-L1 is mutated in PARK5, a juvenile form of Parkinson disease (22); (ii) in the mouse, mutations of UCH-L1 are the cause of a naturally occurring neurodegenerative condition called ‘gad’ ataxia (23); the ‘gad’ disease phenotype can be acerbated by UCH-L3 deletion, suggesting that the disease is due to a proteolytic defect (24); (iii) another spontaneously occurring ataxia model of the mouse, the ax(J) ataxia, is caused by defects in the UBP-type protease USP14 (25); and (iv) the UBP type protease USP7/HAUSP has been found to bind to wild-type, but not to poly(Q) expanded ataxin-1 (26). Currently, it is not clear if the cause of the disease is a pathological ubiquitin depletion due to insufficient turnover, or if the deubiquitination has a signalling role.

Recently, ataxin-3 has been proposed to be related to an all-alpha helical domain found in the N-terminus of adaptins, epsins, and other proteins involved in endocytosis (9). This proposed relationship was mainly based on a similar content and arrangement of secondary structure elements and was refined by a threading analysis using the adaptin180 structure as a template. We agree with the authors of this study that the N-terminal domain of ataxin-3 (also called ‘josephin-domain’) consists mainly of alpha helices. However, the prediction of an evolutionary relationship between adaptins and the ataxin-3 N-terminus is in contradiction with our findings. A comparison of the three-dimensional structures of various protease families adopting the papain fold shows a large degree of variability outside of the catalytically important region (13). The catalytic cysteine residue is invariably part of an alpha-helical structure, while the catalytic histidine and the orienting glutamate/glutamine residue are found in a beta-strand context (compare Fig. 2A). While these beta-strands can be fairly short, they are invariably present in this protease class, thus making our prediction also structurally incompatible with the all-alpha fold of the adaptin/epsin domain.

We consider the significant profile matches, in combination with the observed invariance of the catalytic residues, a more convincing indicator of protein relationships than a threading analysis. Additionally, there is little experimental support for a role of ataxin-3 in endocytosis. The presence of UIM motifs has been used to that end (9), although UIMs are not really a hallmark of endocytosis proteins but rather indicators of ubiquitin binding (6,27,28). Ataxin-3 has been reported to localize to the nucleus and to have an effect on transcriptional activation. Both findings are much easier to reconcile with a function of ataxin-3 as a ubiquitin protease rather than as an endocytosis protein.

Ataxin-7
By using profile-based sequence analysis, we found that the yeast protein Ygl066w is a likely fungal orthologue of a human paralogy group containing ataxin-7 and two more uncharacterized proteins. As Ygl066w is a bona fide subunit of the SAGA histone acetyltransferase complex, the observed relationship strongly suggests that ataxin-7 is involved in histone acetylation, too. The exact role of the ataxin-7 like proteins in this complex is not clear, although the observed pattern of residue invariance suggests a role involving two or more metal ions, potentially in an enzymatic context. If ataxin-7 has an enzymatic activity, its exact nature cannot be predicted from our data.

Nevertheless, this newly found direct connection between a SCA-mutated gene and histone acetylation is intriguing. On several occasions, neurodegenerative diseases arising by triplet instability have been found connected to histone modifications. In a genetic screen for modifiers of ataxin-1 toxicity, the acetylation regulators Sin3a, Rpd3, CtBP and Sir2 have been identified (29). Furthermore, inhibitors of histone deacetylation have been found to reduce poly(Q) mediated toxicity in a Drosophila model system (30). Both results are compatible with a model where histone acetylase components required for cell viability are sequestered to poly(Q) inclusion bodies. So far, transcriptional cofactors like CtBP have been considered prime candidates (31), as they have been found to bind to glutamine-expanded fragments of huntingtin or ataxin-1. Based on our analysis, a role of ataxin-7 in poly(Q)-mediated histone acetylase depletion is an attractive alternative hypothesis—at least in the case of SCA7, where ataxin-7 is known to be sequestered in inclusion bodies. As poly(Q) inclusions are known to sequester other poly(Q) proteins as well, even in their wild-type form, a role of ataxin-7 in other poly(Q) diseases is also conceivable.

General implications for poly(Q) pathogenesis
One of the major questions in understanding the pathology of poly(Q) expansion diseases is the specific role of the mutated protein. Would any protein with a poly(Q) stretch exceeding a certain threshold and having a sufficiently high neuronal expression level be able to cause a neurodegenerative disease? This idea is supported by a number of observations, discussed in recent review literature (32,33). Some phenotypic aspects of poly(Q) diseases caused by mutations in different genes are strikingly similar, including the formation of intranuclear inclusion bodies. It has been demonstrated that these inclusions sequester several cellular proteins, including ubiquitin, proteasome components, chaperones, transcriptional activators and poly(Q) containing proteins in general. Much less clear, however, is the question if this sequestration is part of the pathogenesis mechanism (by depleting the cells of these essential factors) or rather a part of a rescue mechanism (by storing harmful mutant proteins in an inaccessible form). The recruitment of chaperones and components of the protein degradation mechanism can be interpreted as an attempt of the cell to get rid of the mutant proteins.

A number of recent studies have highlighted the importance of the mutated protein and its interaction partners for the disease mechanism (26,34,35). During the last years, several interaction partners of ataxins and other poly(Q) expanded proteins have been identified, and quite frequently the interaction propensity was sharply modulated by the length of the polyQ stretch. Both among genes causing neurodegeneration and among their interaction partners, proteins of the ubiquitin pathway are clearly over-represented: genes mutated in Parkinson disease include the ubiquitin ligase parkin and the ubiquitin protease UCH-L1; ataxin-1 interacts with the ubiquitin protease HAUSP (26); the huntingtin interacting protein HIP2 is a ubiquitin conjugating enzyme with an additional UBA domain (36); the androgen receptor (polyQ-expanded in SBMA) interacts with SARIP (TrEMBL-Acc.:Q99ND9), a UEV domain protein, and with ARA54, a RING-finger ubiquitin ligase with an additional UEV domain (37), to name just a few examples. Ataxin-3, which interacts with ubiquitin and with the UBA protein Rad23 (38), seems to follow this trend. Notably, most of the interactions mentioned here occur with wild-type proteins, suggesting that ubiquitin recruitment to the inclusion bodies might not only be due to the poly(Q) expansion, but rather are caused by intrinsic properties of the affected genes.

One of the most intriguing results from the study of invertebrate models of poly(Q) diseases was the modification of the disease severity by regulators of histone acetylation and the rescue by inhibitors of histone deacetylation (30,39). These observations point strongly towards a role of the general transcription regulation machinery in poly(Q) pathology. Many transcriptional regulators contain oligo(Q) stretches and their recruitment to poly(Q) inclusions by virtue of this feature is conceivable. The depletion of ataxin-7, which we predict to be a component of the SAGA complex, could be another factor contributing to an overall decrease of histone acetylation. It will be interesting to test if overexpression of wild-type ataxin-7 is able to rescue the ataxia phenotype in poly(Q) diseases other than SCA7. Finally, it should be noted that the same studies identifying the yeast orthologue of ataxin-7 as a SAGA subunit also found the ubiquitin protease Ubp8 as an additional subunit of this complex (16,17). This unexpected link between histone acetylation and the ubiquitin pathway might offer a first glimpse of a unified theory of spinocerebellar ataxia pathogenesis.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
All sequence database searches were performed with a nonredundant data set constructed from current releases of SwissProt, TrEMBL, and GenPept (40,41). Generalized profile construction (42) and searches were run locally using the pftools package, version 2.1. (program available from the URL ftp://ftp.isrec.isb-sib.ch/sib-isrec/pftools/). Generalized profiles were constructed using the BLOSUM45 substitution matrix (43) and default penalties of 2.1 for gap opening and 0.2 for gap extension. The statistical significance of profile matches was derived from the analysis of the score distribution of a randomized database (44,45). Database randomization was performed by individually inverting each protein sequence, using SwissProt 34 as the data source. Only sequence matches found with a probability of P<0.01 were included into subsequent rounds of iterative profile refinement. As discussed in Hofmann (45), the probability values obtained by randomization scaling of profiles are much more reliable than those obtained from theoretical scaling methods using assumptions of uniformity of sequence composition. Sequences were assembled from expressed sequence tag and genomic data provided by the NCBI and the Ensembl project (8). Structural superposition and generation of structurally correct multiple alignments were performed using the spdbv program (11). Information on structural neighbours was obtained from the FSSP database (10). Dendrogram analysis was performed using the ‘neighbour joining’ algorithm (46). Secondary structure prediction were performed by the Psipred server (47) and by the PredictProtein network service running the PROF method (48).


   FOOTNOTES
 
* To whom correspondence should be addressed. Email: kay.hofmann{at}memorec.com Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. David, G., Abbas, N., Stevanin, G., Durr, A., Yvert, G., Cancel, G., Wood, N., Ruberg, M., Agid, Y., Mandel, J.L. et al. (1997) Cloning of the SCA7 gene reveals a highly unstable CAG repeat expansion. Nat. Genet., 17, 65–70.[CrossRef][ISI][Medline]

  2. Kawaguchi, Y., Okamoto, T., Taniwaki, M., Aizawa, M., Inoue, M., Katayama, S., Kawakami, H., Nakamura, S., Nishimura, M. and Akiguchi, I. (1994) CAG expansions in a novel gene for Machado–Joseph disease at chromosome 14q32.1. Nat. Genet., 8, 221–228.[CrossRef][ISI][Medline]

  3. Gusella, J.F. and MacDonald, M.E. (2000) Molecular genetics: unmasking polyglutamine triggers in neurodegenerative disease. Nat. Rev. Neurosci., 1, 109–115.[ISI][Medline]

  4. Margolis, R.L. and Ross, C.A. (2001) Expansion explosion: new clues to the pathogenesis of repeat expansion neurodegenerative diseases. Trends Mol. Med., 7, 479–482.[CrossRef][ISI][Medline]

  5. Uchihara, T., Fujigasaki, H., Koyano, S., Nakamura, A., Yagishita, S. and Iwabuchi, K. (2001) Non-expanded polyglutamine proteins in intranuclear inclusions of hereditary ataxias—triple-labeling immunofluorescence study. Acta Neuropathol. (Berl.), 102, 149–152.[Medline]

  6. Hofmann, K. and Falquet, L. (2001) A ubiquitin-interacting motif conserved in components of the proteasomal and lysosomal protein degradation systems. Trends Biochem. Sci., 26, 347–350.[CrossRef][ISI][Medline]

  7. Di Fiore, P.P., Polo, S. and Hofmann, K. (2003) Opinion: when ubiquitin meets ubiquitin receptors: a signalling connection. Nat. Rev. Mol. Cell Biol., 4, 491–497.[CrossRef][ISI][Medline]

  8. Hubbard, T., Barker, D., Birney, E., Cameron, G., Chen, Y., Clark, L., Cox, T., Stupka, E., Ureta-Vidal, A., Vastrik, I. et al. (2002) The Ensembl genome database project. Nucl. Acids Res., 30, 38–41.[Abstract/Free Full Text]

  9. Albrecht, M., Hoffmann, D., Evert, B.O., Schmitt, I., Wullner, U. and Lengauer, T. (2003) Structural modeling of ataxin-3 reveals distant homology to adaptins. Proteins, 50, 355–370.[CrossRef][ISI][Medline]

  10. Holm, L. and Sander, C. (1997) Dali/FSSP classification of three-dimensional protein folds. Nucl. Acids Res., 25, 231–234.[Abstract/Free Full Text]

  11. Guex, N. and Peitsch, M.C. (1997) SWISS-MODEL and the Swiss-PdbViewer: an environment for comparative protein modeling. Electrophoresis, 18, 2714–2723.[CrossRef][ISI][Medline]

  12. Hu, M., Li, P., Li, M., Li, W., Yao, T., Wu, J.W., Gu, W., Cohen, R.E. and Shi, Y. (2002) Crystal structure of a UBP-family deubiquitinating enzyme in isolation and in complex with ubiquitin aldehyde. Cell, 111, 1041–1054.[CrossRef][ISI][Medline]

  13. Anantharaman, V. and Aravind, L. (2003) Evolutionary history, structural features and biochemical diversity of the NlpC/P60 superfamily of enzymes. Genome Biol., 4, R11.[CrossRef][Medline]

  14. Altschul, S.F., Gish, W., Miller, W., Myers, E.W. and Lipman, D.J. (1990) Basic local alignment search tool. J. Mol. Biol., 215, 403–410.[CrossRef][ISI][Medline]

  15. Oliver, S.G. (2002) Functional genomics: lessons from yeast. Phil. Trans. R. Soc. Lond. B Biol. Sci., 357, 17–23.[CrossRef][ISI][Medline]

  16. Gavin, A.C., Bosche, M., Krause, R., Merino, A., Drewes, G., Bouwmeester, T., Bork, P., Seraphin, B., Kuster, B., Neubauer, G. et al. (2002) Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature, 415, 141–147.[CrossRef][Medline]

  17. Sanders, S.L., Jennings, J., Canutescu, A., Link, A.J. and Weil, P.A. (2002) Proteomics of the eukaryotic transcription machinery: identification of proteins associated with components of yeast TFIID by multidimensional mass spectrometry. Mol. Cell. Biol., 22, 4723–4738.[Abstract/Free Full Text]

  18. Chai, Y., Koppenhafer, S.L., Shoesmith, S.J., Perez, M.K. and Paulson, H.L. (1999) Evidence for proteasome involvement in polyglutamine disease: localization to nuclear inclusions in SCA3/MJD and suppression of polyglutamine aggregation in vitro. Hum. Mol. Genet., 8, 673–682.[Abstract/Free Full Text]

  19. Hofmann, K. and Bucher, P. (1996) The UBA domain: a sequence motif present in multiple enzyme classes of the ubiquitination pathway. Trends Biochem. Sci., 21, 172–173.[CrossRef][ISI][Medline]

  20. Buchberger, A., Howard, M.J., Proctor, M. and Bycroft, M. (2001) The UBX domain: a widespread ubiquitin-like module. J. Mol. Biol., 307, 17–24.[CrossRef][ISI][Medline]

  21. Wilkinson, K.D., Tashayev, V.L., O'Connor, L.B., Larsen, C.N., Kasperek, E. and Pickart, C.M. (1995) Metabolism of the polyubiquitin degradation signal: structure, mechanism, and role of isopeptidase T. Biochemistry, 34, 14535–14546.[CrossRef][Medline]

  22. Lansbury, P.T., Jr and Brice, A. (2002) Genetics of Parkinson's disease and biochemical studies of implicated gene products. Curr. Opin. Cell Biol., 14, 653–660.[CrossRef][ISI][Medline]

  23. Saigoh, K., Wang, Y.L., Suh, J.G., Yamanishi, T., Sakai, Y., Kiyosawa, H., Harada, T., Ichihara, N., Wakana, S., Kikuchi, T. et al. (1999) Intragenic deletion in the gene encoding ubiquitin carboxy-terminal hydrolase in gad mice. Nat. Genet., 23, 47–51.[ISI][Medline]

  24. Kurihara, L.J., Kikuchi, T., Wada, K. and Tilghman, S.M. (2001) Loss of Uch-L1 and Uch-L3 leads to neurodegeneration, posterior paralysis and dysphagia. Hum. Mol. Genet., 10, 1963–1970.[Abstract/Free Full Text]

  25. Wilson, S.M., Bhattacharyya, B., Rachel, R.A., Coppola, V., Tessarollo, L., Householder, D.B., Fletcher, C.F., Miller, R.J., Copeland, N.G. and Jenkins, N.A. (2002) Synaptic defects in ataxia mice result from a mutation in Usp14, encoding a ubiquitin-specific protease. Nat. Genet., 32, 420–425.[CrossRef][ISI][Medline]

  26. Hong, S., Kim, S.J., Ka, S., Choi, I. and Kang, S. (2002) USP7, a ubiquitin-specific protease, interacts with ataxin-1, the SCA1 gene product. Mol. Cell. Neurosci., 20, 298–306.[CrossRef][ISI][Medline]

  27. Fisher, R.D., Wang, B., Alam, S.L., Higginson, D.S., Robinson, H., Sundquist, W.I. and Hill, C.P. (2003) Structure and ubiquitin binding of the ubiquitin interacting motif. J. Biol. Chem., 14, 28976–28984.

  28. Shekhtman, A. and Cowburn, D. (2002) A ubiquitin-interacting motif from Hrs binds to and occludes the ubiquitin surface necessary for polyubiquitination in monoubiquitinated proteins. Biochem. Biophys. Res. Commun., 296, 1222–1227.[CrossRef][ISI][Medline]

  29. Fernandez-Funez, P., Nino-Rosales, M.L., de Gouyon, B., She, W.C., Luchak, J.M., Martinez, P., Turiegano, E., Canal, I., Orr, H.T., Zoghbi, H.Y. et al. (2000) Identification of genes that modify ataxin-1-induced neurodegeneration. Nature, 408, 101–106.[CrossRef][Medline]

  30. Steffan, J.S., Bodai, L., Pallos, J., Poelman, M., McCampbell, A., Apostol, B.L., Housman, D.E., Jackson, G.R., Marsh, J.L. and Thompson, L.M. (2001) Histone deacetylase inhibitors arrest polyglutamine-dependent neurodegeneration in Drosophila. Nature, 413, 739–743.[CrossRef][Medline]

  31. Nucifora, F.C., Jr, Sasaki, M., Peters, M.F., Huang, H., Cooper, J.K., Yamada, M., Takahashi, H., Tsuji, S., Troncoso, J., Dawson, V.L. et al. (2001) Interference by huntingtin and atrophin-1 with cbp-mediated transcription leading to cellular toxicity. Science, 291, 2423–2428.[Abstract/Free Full Text]

  32. Temussi, P.A., Masino, L. and Pastore, A. (2003) From Alzheimer to Huntington: why is a structural understanding so difficult? EMBO J., 22, 355–361.[CrossRef][ISI][Medline]

  33. Zoghbi, H.Y. and Orr, H.T. (2000) Glutamine repeats and neurodegeneration. A. Rev. Neurosci., 23, 217–247.[CrossRef][ISI][Medline]

  34. La Spada, A.R. and Taylor, J.P. (2003) Polyglutamines placed into context. Neuron, 38, 681–684.[CrossRef][ISI][Medline]

  35. Chen, H.K., Fernandez-Funez, P., Acevedo, S.F., Lam, Y.C., Kaytor, M.D., Fernandez, M.H., Aitken, A., Skoulakis, E.M., Orr, H.T., Botas, J. et al. (2003) Interaction of akt-phosphorylated ataxin-1 with 14-3-3 mediates neurodegeneration in spinocerebellar ataxia type 1. Cell, 113, 457–468.[CrossRef][ISI][Medline]

  36. Kalchman, M.A., Graham, R.K., Xia, G., Koide, H.B., Hodgson, J.G., Graham, K.C., Goldberg, Y.P., Gietz, R.D., Pickart, C.M. and Hayden, M.R. (1996) Huntingtin is ubiquitinated and interacts with a specific ubiquitin-conjugating enzyme. J. Biol. Chem., 271, 19385–19394.[Abstract/Free Full Text]

  37. Kang, H.Y., Yeh, S., Fujimoto, N. and Chang, C. (1999) Cloning and characterization of human prostate coactivator ARA54, a novel protein that associates with the androgen receptor. J. Biol. Chem., 274, 8570–8576.[Abstract/Free Full Text]

  38. Wang, G., Sawai, N., Kotliarova, S., Kanazawa, I. and Nukina, N. (2000) Ataxin-3, the MJD1 gene product, interacts with the two human homologs of yeast DNA repair protein RAD23, HHR23A and HHR23B. Hum. Mol. Genet., 9, 1795–1803.[Abstract/Free Full Text]

  39. McCampbell, A., Taye, A.A., Whitty, L., Penney, E., Steffan, J.S. and Fischbeck, K.H. (2001) Histone deacetylase inhibitors reduce polyglutamine toxicity. Proc. Natl Acad. Sci. USA, 98, 15179–15184.[Abstract/Free Full Text]

  40. Bairoch, A. and Apweiler, R. (1997) The Swiss-Prot Protein Sequence Data Bank and its supplement Trembl. Nucl. Acids Res., 25, 31–36.[Abstract/Free Full Text]

  41. Benton, D. (1990) Recent changes in the GenBank on-line service. Nucl. Acids Res., 18, 1517–1520.[Abstract/Free Full Text]

  42. Bucher, P., Karplus, K., Moeri, N. and Hofmann, K. (1996) A flexible motif search technique based on generalized profiles. Comput. Chem., 20, 3–23.[CrossRef][ISI][Medline]

  43. Henikoff, S. and Henikoff, J.G. (1992) Amino acid substitution matrices from protein blocks. Proc. Natl Acad. Sci. USA, 89, 10915–10919.[Abstract/Free Full Text]

  44. Hofmann, K. and Bucher, P. (1995) The FHA domain: a putative nuclear signalling domain found in protein kinases and transcription factors. Trends Biochem. Sci., 20, 347–349.[CrossRef][ISI][Medline]

  45. Hofmann, K. (2000) Sensitive protein comparisons with profiles and hidden Markov models. Brief. Bioinform., 1, 167–178.[Abstract/Free Full Text]

  46. Saitou, N. and Nei, M. (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol., 4, 406–425.[Abstract]

  47. McGuffin, L.J., Bryson, K. and Jones, D.T. (2000) The PSIPRED protein structure prediction server. Bioinformatics, 16, 404–405.[Abstract/Free Full Text]

  48. Rost, B. and Liu, J. (2003) The PredictProtein server. Nucl. Acids Res., 31, 3300–3304.[Abstract/Free Full Text]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Hum Mol GenetHome page
S. Alves, E. Regulier, I. Nascimento-Ferreira, R. Hassig, N. Dufour, A. Koeppen, A. L. Carvalho, S. Simoes, M. C. P. de Lima, E. Brouillet, et al.
Striatal and nigral pathology in a lentiviral rat model of Machado-Joseph disease
Hum. Mol. Genet., July 15, 2008; 17(14): 2071 - 2083.
[Abstract] [Full Text] [PDF]

Home page
 NeuroscientistHome page
J. Gardiner, D. Barton, J. May Vanslambrouck, F. Braet, D. Hall, J. Marc, and R. Overall
Defects in Tongue Papillae and Taste Sensation Indicate a Problem with Neurotrophic Support in Various Neurological Diseases
Neuroscientist, June 1, 2008; 14(3): 240 - 250.
[Abstract] [PDF]

Home page
 Eukaryot CellHome page
E. L. Ponder and M. Bogyo
Ubiquitin-Like Modifiers and Their Deconjugating Enzymes in Medically Important Parasitic Protozoa
Eukaryot. Cell, November 1, 2007; 6(11): 1943 - 1952.
[Full Text] [PDF]

Home page
 J. Neurosci.Home page
B. O. Evert, J. Araujo, A. M. Vieira-Saecker, R. A. I. de Vos, S. Harendza, T. Klockgether, and U. Wullner
Ataxin-3 Represses Transcription via Chromatin Binding, Interaction with Histone Deacetylase 3, and Histone Deacetylation.
J. Neurosci., November 1, 2006; 26(44): 11474 - 11486.
[Abstract] [Full Text] [PDF]

Home page
 Hum Mol GenetHome page
X. Zhong and R. N. Pittman
Ataxin-3 binds VCP/p97 and regulates retrotranslocation of ERAD substrates
Hum. Mol. Genet., August 15, 2006; 15(16): 2409 - 2420.
[Abstract] [Full Text] [PDF]

Home page
 Genes Dev.Home page
B. E. Riley and H. T Orr
Polyglutamine neurodegenerative diseases and regulation of transcription: assembling the puzzle.
Genes & Dev., August 15, 2006; 20(16): 2183 - 2192.
[Abstract] [Full Text] [PDF]

Home page
 Hum Mol GenetHome page
P. S. Thomas Jr, G. S. Fraley, V. Damien, L. B. Woodke, F. Zapata, B. L. Sopher, S. R. Plymate, and A. R. La Spada
Loss of endogenous androgen receptor protein accelerates motor neuron degeneration and accentuates androgen insensitivity in a mouse model of X-linked spinal and bulbar muscular atrophy
Hum. Mol. Genet., July 15, 2006; 15(14): 2225 - 2238.
[Abstract] [Full Text] [PDF]

Home page
 J. Cell Biol.Home page
A. Janer, E. Martin, M.-P. Muriel, M. Latouche, H. Fujigasaki, M. Ruberg, A. Brice, Y. Trottier, and A. Sittler
PML clastosomes prevent nuclear accumulation of mutant ataxin-7 and other polyglutamine proteins
J. Cell Biol., July 3, 2006; 174(1): 65 - 76.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
A. M. Ellisdon, B. Thomas, and S. P. Bottomley
The Two-stage Pathway of Ataxin-3 Fibrillogenesis Involves a Polyglutamine-independent Step
J. Biol. Chem., June 23, 2006; 281(25): 16888 - 16896.
[Abstract] [Full Text] [PDF]

Home page
 Hum Mol GenetHome page
A. Haacke, S. A. Broadley, R. Boteva, N. Tzvetkov, F. U. Hartl, and P. Breuer
Proteolytic cleavage of polyglutamine-expanded ataxin-3 is critical for aggregation and sequestration of non-expanded ataxin-3
Hum. Mol. Genet., February 15, 2006; 15(4): 555 - 568.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
Y. Mao, F. Senic-Matuglia, P. P. Di Fiore, S. Polo, M. E. Hodsdon, and P. De Camilli
Deubiquitinating function of ataxin-3: Insights from the solution structure of the Josephin domain
PNAS, September 6, 2005; 102(36): 12700 - 12705.
[Abstract] [Full Text] [PDF]

Home page
 Genome ResHome page
K. K. Kojima and H. Fujiwara
An extraordinary retrotransposon family encoding dual endonucleases
Genome Res., August 1, 2005; 15(8): 1106 - 1117.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
G. Nicastro, R. P. Menon, L. Masino, P. P. Knowles, N. Q. McDonald, and A. Pastore
The solution structure of the Josephin domain of ataxin-3: Structural determinants for molecular recognition
PNAS, July 26, 2005; 102(30): 10493 - 10498.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
V. B. Palhan, S. Chen, G.-H. Peng, A. Tjernberg, A. M. Gamper, Y. Fan, B. T. Chait, A. R. La Spada, and R. G. Roeder
Polyglutamine-expanded ataxin-7 inhibits STAGA histone acetyltransferase activity to produce retinal degeneration
PNAS, June 14, 2005; 102(24): 8472 - 8477.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
S. J. McMahon, M. G. Pray-Grant, D. Schieltz, J. R. Yates III, and P. A. Grant
Polyglutamine-expanded spinocerebellar ataxia-7 protein disrupts normal SAGA and SLIK histone acetyltransferase activity
PNAS, June 14, 2005; 102(24): 8478 - 8482.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
B. G. Burnett and R. N. Pittman
The polyglutamine neurodegenerative protein ataxin 3 regulates aggresome formation
PNAS, March 22, 2005; 102(12): 4330 - 4335.
[Abstract] [Full