Skip Navigation


Human Molecular Genetics Advance Access originally published online on July 28, 2008
Human Molecular Genetics 2008 17(20):3236-3246; doi:10.1093/hmg/ddn219
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
17/20/3236    most recent
ddn219v1
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 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 (3)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Linder, B.
Right arrow Articles by Fischer, U.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Linder, B.
Right arrow Articles by Fischer, U.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2008. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Tdrd3 is a novel stress granule-associated protein interacting with the Fragile-X syndrome protein FMRP

Bastian Linder1,{dagger},, Oliver Plöttner2,{dagger},{ddagger}, Matthias Kroiss1, Enno Hartmann3, Bernhard Laggerbauer1,§, Gunter Meister2, Eva Keidel2 and Utz Fischer1,*

1 Department of Biochemistry, Theodor Boveri Institute, Am Hubland, D-97074 Würzburg, Germany 2 Max Planck Institute of Biochemistry, Am Klopferspitz 18, D-82152 Martinsried, Germany 3 Institute for Biology, University of Lübeck, Ratzeburger Allee 160, D-23538 Lübeck, Germany

* To whom correspondence should be addressed. Tel: +49 9318884029; Fax: +49 9318884028; Email: utz.fischer{at}biozentrum.uni-wuerzburg.de

Received April 16, 2008; Revised July 4, 2008; Accepted July 24, 2008


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 REFERENCES
 
Tudor domains are widespread among proteins involved in RNA metabolism, but only in a few cases their cellular function has been analyzed in detail. Here, we report on the characterization of the ubiquitously expressed Tudor domain containing protein Tdrd3. Apart from its Tudor domain, we show that Tdrd3 possesses an oligosaccharide/nucleotide binding fold (OB-fold) and an ubiquitin associated domain capable of binding tetra-ubiquitin. A set of biochemical experiments revealed an interaction of Tdrd3 with FMRP, the product of the gene affected in Fragile X syndrome, and its autosomal homologs FXR1 and FXR2. FMRP has been implicated in the translational regulation of target mRNAs and shown to be a component of stress granules (SG). We demonstrate that overexpression of Tdrd3 in cells induces the formation of SGs and as a result leads to its co-localization with endogenous FMRP in these structures. Interestingly, the disease-associated FMRP missense mutation I304N identified in a Fragile X patient severely impairs the interaction with Tdrd3 in biochemical experiments. We propose a contribution of Tdrd3 to FMRP-mediated translational repression and suggest that the loss of the FMRP–Tdrd3 interaction caused by the I304N mutation might contribute to the pathogenesis of Fragile X syndrome.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 REFERENCES
 
Tudor domains are discernable sequence motifs of about 60 amino acids in numerous proteins and adopt a β barrel-like tertiary structure (1). An emerging functional theme for proteins containing this domain is their involvement in various aspects of the metabolism of RNA (2). This connection is easily inferred in the case of many Tudor domain containing (Tdrd) proteins, as these possess additional conserved sequence motifs predicted to bind RNA, for example in the case of Tdrd2 that bears a hnRNP K homology (KH) domain (3) or Tdrd10 that is predicted to contain an RNA recognition motif. Some members of this protein family contain domains in addition to the Tudor domain that do not directly point to an RNA related function, for example a myeloid–Nervy–DEAF-1 domain (MYND-domain; Tdrd1) (4) or a serine/threonine kinase domain (Tdrd8) (5). Again others like SMNrp/SPF30, which has experimentally been proven to function in pre-mRNA splicing (6), or the name-giving Drosophila Tudor-protein itself lack additional motifs. While the precise function of the latter protein and the identity of its interacting partners remain elusive, it has been shown to play a role in the organization of the pole plasm in oocytes and the nurse cell nuage (7). The oocyte pole plasm gives rise to the germ cell lineage and is characterized by a large abundance of ribonucleoprotein particles (RNPs) (8).

The Tudor domain containing protein studied thus far in most detail is the Survival Motor Neuron (SMN) protein. Reduced expression levels of this protein are the underlying cause for the neuronal disease spinal muscular atrophy (9). SMN is part of a macromolecular complex (termed SMN complex), which mediates assembly of Sm proteins onto U snRNA (10). During this essential step in the biogenesis of spliceosomal U snRNPs, the Tudor domain of SMN engages in important interactions, specifically in the binding of Sm proteins onto the SMN complex (1,11). Interestingly, this interaction appears to be stimulated by the modification of arginine residues within a subset of Sm proteins to symmetric di-methyl arginine (sDMA) (12,13). These observations along with structural studies have put forward the idea that Tudor domains are protein–protein interaction modules specifically designed to recognize sDMA-containing regions of proteins. In keeping with this notion, the Tudor domain of the splice factor SPF30/SMNrp and of the thus far uncharacterized Tdrd3 protein have been suggested to bind sDMA-containing sequences (14).

In the present study, we have performed experiments to gain insight into the cellular function of Tdrd3. This 83 kDa protein is unique in the Tdrd family, as it possesses an N-terminal oligosaccharide/nucleotide binding (OB) fold domain, a central ubiquitin binding associated (UBA) domain and a C-terminal Tudor domain. We find Tdrd3 to be expressed in a variety of tissues with a predominant cytosolic localization. The protein binds to tetraubiquitin by means of its UBA domain in vitro, suggesting a role for this protein in the proteolytic turnover of proteins mediated by the ubiquitin-proteasome system. Interestingly, a series of biochemical and genetic protein interaction studies suggest that the Tdrd3 protein associates with the Fragile X mental retardation protein (FMRP), and its autosomal homologs FXR1 and FXR2. Limiting amounts of functional FMRP cause Fragile X syndrome, the most common form of inherited mental retardation worldwide (MIM no. 300624 [OMIM] ) (15). Experimental evidence suggests that FMRP is engaged in post-transcriptional gene silencing events, in which mRNAs from actively translating polyribosomes are sequestered into translationally silent mRNPs. These mRNPs are generated as a consequence of various conditions of cellular stress, whereupon they accumulate in discrete cytoplasmic foci termed stress granules (SGs). Overexpression of Tdrd3 in HeLa cells results in the induction of such foci, where it co-localizes with endogenous FMRP. Strikingly, Tdrd3 interaction is abolished by the I304N missense mutation in FMRP associated with severe Fragile X syndrome. Taken together, our data reveal a novel FMRP-associated protein and suggest a role of Tdrd3 in FMRP-mediated translational repression.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 REFERENCES
 
Tdrd3 is a conserved protein with a unique domain composition
We initially identified Tdrd3 in the ENSEMBL database by homology searches using the Tudor domain of SMN. Homology was restricted to the Tudor domain, but N- and C-terminal sequences were highly divergent to SMN. BLAST searches revealed a high degree of conservation in a wide range of species including insects, the nematode C. elegans and deuterostomians such as the sea urchin and vertebrates. No Tdrd3-ortholog was detected in fungi and only a distantly related ortholog was present in plants. Together with the absence from S. cerevisiae and S. pombe, these findings point towards a function of Tdrd3 that developed later in eukaryotic evolution. To functionally analyze Tdrd3, PCR was performed from a human brain cDNA library using primers derived from the annotated EST sequences. Subsequent 5'- and 3'-RACE resulted in a cDNA of 2578 nucleotides encoding an open reading frame of 744 amino acids. This is in contrast to the Tdrd3 entry in ENSEMBL, where an open reading frame of only 651 amino acids is predicted (ENSP00000196169). The difference most likely arises from different 5' and 3' splice sites for the first and second intron, respectively (data not shown), and results in the addition of 93 amino acids to the N-terminus of the protein. Importantly, this N-terminal sequence is annotated in Tdrd3 proteins from several other species such as rat, zebrafish or Drosophila (Fig. 1A, lower panel) and found in human EST databases (data not shown). The N-terminus of Tdrd3 shows homology to the OB-fold domain (16), a single-stranded nucleic acid binding domain (17). The structural hallmark of this motif is a β barrel consisting of five β sheets and an optional N-terminal {alpha} helix. Our full-length amino acid sequence is predicted to contain the complete five β-sheets typical of an OB-fold domain (Fig. 1A, lower panel). Homology searches identified in addition a UBA domain located in between the N-terminal OB-fold and the Tudor domain. The domain organization of Tdrd3 is depicted in Figure 1A.


Figure 1
View larger version (52K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1. Tdrd3 is widely expressed and localizes predominantly to the cytoplasm. (A) Upper panel: schematic view of the domain structure of SMN, the human Tdrd3 database entry and cloned human Tdrd3 sequence. Numbering corresponds to amino acid residues of cloned Tdrd3. Lower panel: alignment of the human Tdrd3 N-terminal domain annotated in the ENSEMBL database (H.s.[DB]), the Tdrd3 sequence identified in this study (H.s.) and Tdrd3 homologs from different model organisms. Shaded residues are >60% conserved. Four of the five β-strands characteristic of OB-fold domains (β2–β5) were inferred from Yin et al. (16), β1 was predicted with PSI-PRED. (B) A Northern blot membrane (Clontech) containing 2 µg of poly(A)-RNA each from various tissues was probed with [32P]-labeled Tdrd3 cDNA. A single band of 2.9 kb was detected in any tissue tested. (C) HeLa cytosolic extract (lane 1), HEK-293 total extract (lane 2) or total cytosolic and nuclear extract of Xenopus oocytes (lanes 3–5) were analyzed by SDS–PAGE and western blotting with an anti-Tdrd3 antibody. (D) Indirect immunofluorescence microscopy of HeLa cells with an anti-Tdrd3 antibody showed a cytosolic distribution (a). Nuclei were stained with DAPI (b), (c) and (d) show merged image and Nomarski optics, respectively. Scale bar is 20 µm.

 
Tdrd3 is a ubiquitously expressed protein with cytosolic localization
In a series of experiments, we analyzed the expression pattern of Tdrd3 among tissues and its intracellular localization. Northern blot analysis of eight different tissues revealed a single band at 2.9 kb which was detected in all tissues studied (Fig. 1B). The mRNA length corresponds to the size of the cloned cDNA with an additional poly(A) tail. Thus, Tdrd3 is expressed in a wide variety of tissues, which indicates a function common to a variety of cell types. To determine the sub-cellular localization of the protein in individual cells, a polyclonal antibody against human Tdrd3 was raised in rabbits, affinity purified and used for western blotting and indirect immunofluorescence experiments. In western blots of HeLa cytoplasmic extract and 293T total cell extract, a single band at ~83 kDa was observed, matching the predicted molecular weight of Tdrd3 protein (Fig. 1C). A corresponding signal was observed only in the cytosolic fraction of Xenopus laevis oocytes, but not in extracts derived from dissected nuclei (Fig. 1C). Consistently, indirect immunofluorescence microscopy revealed a strong cytosolic localization with perinuclear accumulation in HeLa (Fig. 1D) and COS-1 cells (data not shown). Similar results were obtained when HA–Tdrd3 transfected cells were analyzed (data not shown). Thus, Tdrd3 is a housekeeping gene with a predominant cytosolic localization.

The UBA domain of Tdrd3 specifically binds to Lys48-linked tetraubiquitin
Next we set out to identify interactors of Tdrd3. We initially turned our attention to the UBA domain. Although originally discovered in proteins involved in ubiquitin-mediated proteolysis (18), this domain has been shown more recently to be capable of acting as a general protein–protein interaction motif (19–21). Therefore, we immobilized GST fusions of full-length Tdrd3, Tdrd3 lacking the UBA domain or the isolated UBA domain on Glutathione–Sepahrose beads. The immobilized fusions were then incubated either with monoubiquitin, Lys48-linked tetraubiquitin (the signal for proteasomal degradation of proteins), Lys63-linked ubiquitin dimers (often found as non-proteolytic signals) or the small ubiquitin-related modifier 1 (SUMO-1), which is structurally similar to ubiquitin yet divergent in its cellular function (22). As shown in Figure 2A, the UBA domain of Tdrd3 is both necessary and sufficient for binding to Lys48 linked tetraubiquitin. In contrast, no binding occurred to monoubiquitin, Lys63-linked ubiquitin or the small ubiquitin related modifier SUMO-1 (Fig. 2B). As the major function of this type of modification is the targeting to proteasomes, these results suggest a role of Trdr3 in turnover of proteins mediated by the ubiquitin-proteasome system.


Figure 2
View larger version (54K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 2. Tdrd3 specifically binds Lys48 linked tetraubiquitin. (A) GST–Tdrd3 (lanes 1–3), GST–Tdrd3–{Delta}UBA (lanes 4–6), GST–UBA (lanes 7–9) or GST as a control (lanes 10–12) were immobilized on glutathione–sepharose and incubated with monomeric ubiquitin (lanes 1, 4, 7 and 10), Lys48-linked tetraubiquitin (lanes 2, 5, 8 and 11) or Lys63 linked ubiquitin (lanes 3, 6, 9 and 12). Bound proteins were eluted and ubiquitin was detected by western blotting. Lanes 13–15 show 10% of the input material. (B) Binding experiment as in (A) with recombinant SUMO-1. The bound fraction was detected by western blotting using an anti-SUMO-1 antibody.

 
FMRP and its paralogs interact with Tdrd3
In order to identify additional binding partners of Tdrd3, 2x106 cDNA clones were tested in a yeast two hybrid (Y2H) interaction screen using Tdrd3 as bait. This approach revealed a strong interaction of Tdrd3 with the FXR1 protein. FXR1 is encoded by one of the two autosomal paralogs of the FMR1 gene, which is affected in patients suffering from Fragile X syndrome (23). To verify this interaction and to elucidate whether Tdrd3 also interacts with the other FXR1 homologs, FXR2 and FMRP, we conducted in vitro binding studies with immobilized ZZ-tagged Tdrd3 and [35S]-labeled FMRP, FXR1 and FXR2 translated in reticulocyte lysate. Efficient binding of all three FMR paralogs to ZZ–Tdrd3 was observed but not to a control column containing ZZ tag only (Fig. 3A). Of note, the same results were obtained when experiments were conducted with immobilized GST–Tdrd3 and in vitro translated FMRP, FXR1 and FXR2 (data not shown). As binding of the investigated two proteins in these assays may be mediated by an unknown factor present in the reticulocyte lysate, direct interaction studies were performed with purified recombinant proteins. FMRP, FXR1 and FXR2 were expressed as ZZ-tagged fusion proteins, immobilized onto IgG-Sepharose and incubated with GST–Tdrd3 protein. Tdrd3 bound to all ZZ-fusion proteins tested but not to the ZZ control (Fig. 3B). Thus, Tdrd3 directly binds to FMRP and its homologs FXR 1 and 2 in vitro.


Figure 3
View larger version (57K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 3. Tdrd3 interacts with FMRP, the product of the disease gene for Fragile X syndrome, and its paralogs FXR1 and FXR2 both in vitro and in vivo. (A) In vitro translated, [35S]-labeled FXR1 (lanes 1 and 4), FXR2 (lanes 2 and 5) or FMRP (lanes 3 and 6) efficiently bound to immobilized ZZ–Tdrd3 (lanes 1–3) but not ZZ-domain alone (lanes 4–6). Lanes 7–9 show 30% of the translated proteins used in the binding experiments. (B) ZZ–FMRP (lane 1), ZZ–FXR1 (lane 2), ZZ–FXR2 (lane 3) or ZZ-domain alone (lane 4) was immobilized on IgG-Sepharose and incubated with GST–Tdrd3. Bound proteins were analyzed by SDS–PAGE and Coomassie-blue staining. Lanes 5–9 show 10% of the protein inputs. (C) Anti-Tdrd3 and anti-FMRP immunoprecipitates from HEK-293 whole cell extracts were analyzed by western blotting for the presence of co-precipitated FMRP (lanes 1–5) and Tdrd3 (lanes 6–10), respectively. FMRP and Tdrd3 are marked by arrows. Unspecific degradation products are indicated by asterisks in (B) and (C).

 
Next we analyzed whether the observed interaction between FMRP and Tdrd3 also occurs in the context of a living cell. To address this question, co-immunoprecipitation experiments of endogenous Tdrd3 and FMRP from cellular extracts were performed. As shown in Figure 3C, immunoprecipitation of FMRP leads to the co-precipitation of Tdrd3, and FMRP can be co-precipitated with Tdrd3 (Fig. 3C). Taken together, by means of yeast two-hybrid assays, protein–protein interaction experiments and co-immunoprecipitation studies from cellular extracts, we establish Tdrd3 as a novel interactor of proteins of the FMRP/FXR family.

Tdrd3 is a novel component of stress granules
Previous studies have revealed a role of FMRP and its paralogs in posttranscriptional gene regulation. FXR1 has been shown to upregulate translation of certain mRNAs as a consequence of binding to AU-rich elements within the 3'-UTR of these RNAs (24). In contrast, FMRP negatively regulates mRNA translation, most likely by interfering with ribosome subunit joining on mRNAs and thus formation of 80S ribosomes (25). More recently, FMRP has also been connected to the microRNA pathway, but its precise role in this process remains elusive (26,27). Translational silencing is often linked to the emergence of SGs in cells. These are cytosolic foci that contain stalled 48S pre-initiation complexes, key components of microRNPs like Argonaute 2 (AGO2) and other RNA binding proteins (28,29). Indeed, FMRP-mediated translational regulation coincides with SG formation, a process that can also be induced by FMRP overexpression (Fig. 4A–C) (30). Given that FMRP and Tdrd3 interact in vitro and in vivo, we asked whether overexpression of Tdrd3 can likewise induce SG formation. Indeed, upon overexpression, we observed a strong accumulation of both HA-tagged (Fig. 4D–F) and GFP-tagged Tdrd3 (Fig. 4G–I) into discrete cytosolic foci. To confirm that these foci induced by Tdrd3 are indeed functional SGs, we investigated whether they are sensitive to cycloheximide or emetine. Both drugs are translational inhibitors that act by stalling ribosomes on mRNA, thus leading to the stabilization of polysomes and a decrease in 48S pre-initiation complexes. As the pool of translationally silent mRNPs in SGs is in equilibrium with the mRNAs in the polysomal fraction (31), this results in reduced number and size of SGs (32). Indeed, treatment of Tdrd3-transfected cells with either drug significantly reduced the amount of cytosolic granules in GFP–Tdrd3 positive cells (Fig. 4J–O), suggesting that these granules are functional storage compartments of mRNPs. To ensure a direct link between emetine treatment and granule dispersion, time-lapse microscopy of individual cells was performed. Although under these experimental conditions, GFP–Tdrd3 containing foci were not dissociated completely, dissemination of distinct granules was clearly observed (Supplementary Material, Fig. S1), suggesting a direct influence of polysome stabilization on the turnover of these structures. Moreover, treatment of cells with arsenite, a potent inducer of cellular stress known to disrupt polysomes and induce SGs, led to the formation of Tdrd3-containing foci even in cells expressing only low levels of GFP–Tdrd3 (Figs 4P–R and Supplementary Material, Fig. S1). Next, we tested whether known SG markers co-localize to these foci. Therefore, immunofluorescence analysis was performed of GFP–Tdrd3-transfected cells with antibodies against the known SG proteins T-cell intracellular antigen-related protein (TIAR) (33), death associated protein 5 (DAP5) (34) and FMRP. Figure 5A–D shows TIAR-positive SGs induced by GFP–Tdrd3 overexpression. Notably, even upon SG induction by arsenite treatment (Fig. 5E–H), when untransfected cells also showed TIAR-positive SGs (arrows), all SGs visible in transfected cells contained GFP–Tdrd3 (arrowheads). These results could be confirmed using DAP5 as SG marker (Fig. 5I–P). Finally, we tested the presence of FMRP in Tdrd3 containing SGs. Endogenous FMRP was efficiently recruited to GFP–Tdrd3 containing SGs in untreated and arsenite treated cells (Fig. 5Q–X). These results were confirmed by analyzing endogenous FMRP and Tdrd3 in untransfected cells (Fig. 6). While under non-stress conditions, both proteins were distributed throughout the cytosol (Fig. 6A–D), arsenite treatment resulted in the induction of SGs containing both FMRP and Tdrd3 (Fig. 6E and F). Hence, these data confirm Tdrd3 as a novel SG component and suggest a functional link to FMRP-mediated translational regulation.


Figure 4
View larger version (71K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 4. Expression of Tdrd3 in HeLa cells induces SGs. HeLa cells were transiently transfected with HA–FMRP (A–C), HA–Tdrd3 (D–F) or GFP–Tdrd3 (G–R). Visualization of HA-tagged FMRP revealed a cytosolic staining in transfected cells. In some cells, expression of HA-FMRP resulted in cytosolic SGs (arrowheads in A). The same staining pattern was observed when cells were transfected with HA–Tdrd3 (arrowheads in D) or GFP–Tdrd3 (G). To test whether GFP–Tdrd3 containing SGs can be dispersed by cycloheximide or emetine, transfected HeLa cells were treated with either drug 1 h prior to fixation. While untreated (G–I) cells showed a high amount of SGs, granule formation was significantly reduced in cycloheximide (J–L) or emetine (M–O) treated cells. Upon treatment with arsenite (P–R), all transfected cells contained GFP–Tdrd3 positive SGs and no cells with a diffuse GFP–Tdrd3 distribution could be observed. Scale bars in (F) and (R) are 20 µm and apply to images (A–F) and (G–R), respectively. DAPI staining of nuclei is shown in the merged images.

 


Figure 5
View larger version (91K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 5. The SG components TIAR, DAP5 and FMRP co-localize with GFP–Tdrd3 in SGs. GFP–Tdrd3 transfected cells were analyzed by immunofluorescence with anti-TIAR (A–H), anti-DAP5 (I–P) or anti-FMRP (Q–X) antibodies. Both proteins showed a strong co-localization with GFP–Tdrd3 positive SGs. Arsenite treatment (E–H, M–P and U–X) led to the induction of SGs containing TIAR, DAP5 or FMRP in untransfected cells (arrows). SGs in transfected cells showed a strong GFP–Tdrd3 signal (arrowheads). Scale bar is 20 µm and applies to all images. DAPI staining of nuclei is shown in the merged images.

 


Figure 6
View larger version (55K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 6. Endogenous Tdrd3 and FMRP co-localize in arsenite-induced SGs. HeLa cells were left untreated (A–D) or treated with arsenite (E–H). Immunofluorescence analysis with anti-FMRP (A and E) or anti-Tdrd3 antibodies (B, F) revealed a diffuse cytosolic localization of both proteins in untreated cells. Upon arsenite treatment, FMRP-positive SGs were induced and Tdrd3 co-localized in these structures (arrowheads). Scale bar is 20 µm and applies to all images. DAPI staining of nuclei is shown in the merged images.

 
Mapping of the interaction between FMRP and Tdrd3
FMRP is a protein of 632 amino acids in length with three conserved RNA binding motifs. Two KH domains (KH1 and KH2) are found in the central part of the protein and an arginine–glycine rich domain in the C-terminal region. Having established an interaction between Tdrd3 and FMRP, we were next interested in narrowing down the sites in both proteins required for their mutual interaction. In a set of experiments, in vitro translated truncations of FMRP were incubated with immobilized full-length Tdrd3. Interestingly, two different fragments in the primary structure of FMRP are essential for Tdrd3 binding. First, a region spanning residues 430–486 participate in the interaction. This is shown by the fact that binding of FMRP1–430 is not significantly higher than the background (Fig. 7B, compare lanes 4 and 15), while FMRP1–486 shows strong binding (Fig. 7B, compare lanes 5 and 16). Second, a region encompassing the two KH domains (amino acids 216–332) is necessary for interaction with Tdrd3, as FMRP216–632 is efficiently bound (Fig. 7B, lane 7) while FMRP332–632 fails to do so (Fig. 7B, lane 8). Fragments lacking any one of these binding sites lose the ability to interact with Tdrd3, suggesting that both regions of the protein constituting the Tdrd3 binding site are brought into proximity of each other in the folded protein.


Figure 7
View larger version (62K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 7. Mapping of the interaction domains of FMRP and Tdrd3. (A and C) Schematic drawings of the FMRP and Tdrd3 fragments used for interaction mapping. The tables summarize the results of the interaction studies shown in (B) and (D). (B) Recombinant full-length ZZ–Tdrd3 (lanes 1–11) or ZZ-domain alone (lanes 12–22) were immobilized on IgG-sepharose and incubated with FMRP deletion constructs that had been translated in vitro in the presence of L-[35S]-methionine. The only deletion constructs that produced a signal significantly over background were those containing the KH domains and a stretch of amino acids ranging from residue 430–486 (lanes 5 and 7). (D) To map the FMRP binding domain of Tdrd3, immobilized full-length FMRP was used to precipitate fragments of Tdrd3. Interestingly, although binding could be confined to residues 558–744 (lane 4), the Tudor domain itself was not essential for interaction (lane 6). In contrast, 20 amino acids directly C-terminal to the Tudor domain seem to be necessary for interaction (compare lanes 7 and 8). (E) Schematic summary of the Tdrd3/FMRP interaction domains. The location of the pathogenic I304N missense mutation is indicated by a star.

 
To map the corresponding FMRP-binding site in the Tdrd3 primary sequence, N-terminal and C-terminal truncated proteins were translated in vitro and tested for FMRP-binding activity. As shown in Figure 7D, the C-terminal region spanning amino acids 558–744 and encompassing the Tudor domain was sufficient for the interaction with Tdrd3 (Fig. 7D, lane 4). Surprisingly, the deletion of the Tudor domain failed to affect binding to FMRP (Fig. 7D, lane 5). However, analysis of further truncations revealed that the 20 amino acids directly adjacent to the C-terminal boundary of the Tudor domain are necessary for binding to FMRP. Taken together, these results imply a model where a small stretch neighboring the Tudor domain of Tdrd3 interacts with a binding site in the middle of FMRP comprised of the KH domains and a region from residues 430–486 (Fig. 7E).

A pathogenic amino acid substitution of FMRP prevents Tdrd3 binding
Although the majority of patients affected by Fragile-X syndrome exhibit an expansion of a CGG-repeat adjacent to the FMR gene leading to transcriptional silencing of FMRP, a missense mutation causing a single amino acid substitution (I304N) has been described in one patient (35). Since the site of this missense mutation locates within FMRPs binding site for Tdrd3 (Fig. 7E), we asked whether interaction between both proteins was affected. Therefore, ZZ-tagged wild-type or I304N mutant FMRP (FMRPI304N) was immobilized onto IgG-Sepharose and binding of in vitro translated Tdrd3 was assessed (Fig. 8A). Whereas wild-type FMRP bound Tdrd3 also stably under stringent conditions (750 mM NaCl, Fig. 8A, lane 4), only low levels of Tdrd3 binding to FMRPI304N were detected even at physiological salt concentrations (Fig. 8A, lane 5). This indicates that the I304N mutation abolishes the binding capacity of the Tdrd3 protein to FMRP.


Figure 8
View larger version (60K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 8. The pathogenic FMRP I304N mutation identified in a patient with Fragile X syndrome strongly interferes with Tdrd3 interaction. (A) To assess binding of Tdrd3, recombinant wild-type (lanes 1–4) and I304N mutant (lanes 5–8), ZZ–FMRP was immobilized on IgG sepharose. ZZ domain alone served as a control (lane 9). After incubation with in vitro translated Tdrd3, the beads were washed with buffers containing either 150 mM (lanes 1, 5 and 9), 300 mM (lanes 2 and 6), 500 mM (lanes 3 and 7) or 750 mM (lanes 4 and 8) NaCl. Eluted proteins were resolved by SDS–PAGE and analyzed by Coomassie-blue staining (lower panel) and autoradiography (upper panel). Binding of Tdrd3 to immobilized FMRPI304N was severely reduced even at physiological salt concentration (lane 5). Thirty percent of in vitro translated Tdrd3 are depicted as input (lane 10). Asterisks indicate unspecific degradation products. (B) To assess whether the reduced Tdrd3 binding observed in (A) is a consequence of the known defect in homooligomerization of FMRPI304N, immobilized ZZ–FMRPwt (lanes 1 and 2) or ZZ–FMRPI304N (lanes 3 and 4) were incubated with purified recombinant His-FMRPwt (lanes 1 and 3) or His-FMRPI304N (lanes 2 and 4). As a control, ZZ domain was incubated with a mix of His-FMRPwt and His-FMRPI304N (lane 5). Coomassie-blue staining (upper panel) revealed unaffected formation of FMRPwt/FMRPwt homooligomers (lane 1) and FMRPwt/FMRPI304N heterooligomers (lanes 2 and 3) while only residual FMRPI304N/FMRPI304N homooligomerization could be observed (lane 4). Upon incubation with [35S]-labeled in vitro translated Tdrd3, Tdrd3 bound to wild-type FMRPwt/FMRPwt homooligomers (lanes 1–3) and FMRPwt/FMRPI304N hetero-oligomers, but failed to associate with FMRPI304N/FMRPI304N homooligomers (lane 4). Thirty percent of in vitro translated Tdrd3 is depicted as input (lane 6). IgG heavy and light chain are indicated on the right and unspecific degradation products are marked by an asterisk. (C–J) The FMRPwt/FMRPI304N hetero-oligomers observed in (B) were tested for their ability to induce SGs. Accordingly, HeLa cells endogenously expressing FMRPwt were transfected with HA–FMRPwt (C–F) or HA–FMRPI304N (G–J). SG formation could be observed in both cases (arrowheads).

 
Previous studies had shown that wild-type FMRP forms homo-oligomers in vitro, and that homo-oligomerization of FMRPI304N is severely compromised. Hetero-oligomerization between wild-type and mutant FMRP, in contrast, was not affected (25). Hence, we tested whether formation of FMRP oligomers is a pre-requisite for Tdrd3 binding (Fig. 8B). For this, ZZ-tagged FMRPwt (Fig. 8B, lanes 1 and 2) and FMRPI304N (lanes 3 and 4) were immobilized onto IgG-Sepharose. Oligomerization was initiated by the addition of His-tagged FMRPwt (lanes 1 and 3) or FMRPI304N (lanes 2 and 4). As expected, roughly stoichiometric FMRPwt/FMRPwt homooligomers (lane 1) and FMRPwt/FMRPI304N hetero-oligomers (lanes 2 and 3) were formed while FMRPI304N/FMRPI304N homooligomerization was markedly reduced (lane 4). After washing, the immobilized oligomers were incubated with in vitro translated [35S]-labeled Tdrd3. The residual FMRPI304N/FMRPI304N homo-oligomer failed entirely to pull down radiolabeled Tdrd3 (Fig. 8B, lane 4, lower panel). Therefore, although the loss of Tdrd3 binding seems to correlate with reduced homooligomerization, we propose that binding deficiency is a direct effect of the I304N mutation on FMRP/Tdrd3 interaction rather than a consequence of impaired oligomerization. Surprisingly, FMRPwt/FMRPI304N hetero-oligomers bound Tdrd3 as efficient as wild-type homooligomers (Fig. 8B, lower panel, lanes 1–3). As a consequence, hetero-oligomer formation by FMRPI304N expression in an FMRPwt background should be sufficient to induce SGs in cells. This was tested by the transfection of HeLa cells with HA-tagged FMRPI304N, which indeed resulted in the formation of SGs comparable to wild-type FMRP transfection (Fig. 8C–J). Also, no significant deficiency in the recruitment of endogenous Tdrd3 (Fig. 8B and F) or co-transfected GFP–Tdrd3 (Supplementary Material, Fig. S2) to FMRPI304N-induced SGs could be observed, suggesting that FMRPwt/FMRPI304N hetero-oligomers are functional in vivo.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 REFERENCES
 
Tdrd3 contains motifs implicated in a variety of different cellular processes
In this study, we aimed at providing a first biochemical and functional characterization of the Tudor domain containing protein Tdrd3. While the structure of murine Tdrd3’s Tudor domain has recently been solved (PDB ID: 2D9T) and has been suggested to be a binding platform for proteins containing symmetrical dimethylarginine (14), functional implications for Tdrd3 remained elusive. Homology-based sequence alignment suggests that the N-terminal part of Tdrd3 constitutes an OB-fold domain. This domain is frequently found in proteins with affinity to single-stranded DNA and/or RNA. Interestingly, while only four of the five β-sheets characteristic for an OB-fold domain could be identified in the current database entries of Tdrd3, our 5'-RACE identified an additional sequence with an in-frame upstream initiator methionine codon. The open reading frame encoded by this additional sequence is predicted by PSI-PRED to contain a kinked β-strand typical for the OB-fold β1 strand (17). In the middle part of the protein, a UBA domain is located whose structure has recently been solved by NMR spectroscopy (PDB ID: 1WJI). Complementary to this structural information, we provide evidence that the UBA domain binds to Lys48-linked tetraubiquitin, i.e. the signal for the proteosomal degradation of proteins (Fig. 2). Remarkably, this interaction is highly specific as neither monoubiquitin, Lys63-linked ubiquitin dimers, nor the ubiquitin-related protein SUMO bound to detectable levels to Tdrd3’s UBA domain in vitro. It is, therefore, a possibility that Tdrd3 associates with proteins designated for proteasomal degradation (36).

Direct interaction of Tdrd3 with FMRP
By Northern blot analysis, we find the Tdrd3-mRNA to be expressed in a large spectrum of tissues, indicating a function in basal cellular processes rather than in tissue-specific pathways. In cultured cells and in Xenopus oocytes, the Tdrd3 protein was found to be predominantly cytosolic. Indirect immunofluorescence in cultured cells also revealed an occasional faint nuclear staining. Whether this points to an additional function of Tdrd3 in the nucleus is currently unclear. Hence, Tdrd3 is likely to perform its function in the cytoplasm of cells.

To identify proteins interacting with the Tdrd3 protein, we performed an Y2H screen using a human brain cDNA library. Interestingly, FXR1, the autosomal paralog of the Fragile X mental retardation protein FMRP, emerged as a robust interacting protein out of this screen. Although not identified in the two hybrid screen, we also found strong binding of Tdrd3 with FXR2, the second protein closely related to FMRP, and with FMRP itself. The association with the latter is of particular interest, as this protein is known to be absent or mutated in the Fragile X-mental retardation syndrome. This interaction has been confirmed by two independent and complementary techniques. First, direct contact of FMRP with Tdrd3 was observed when recombinantly expressed proteins were analyzed by in vitro binding assays. Second, and most importantly, we also detected binding of FMRP and Tdrd3 to each other by co-immunoprecipitation from cellular extracts, making it likely that one direct interaction partner of Tdrd3 in vivo is FMRP.

Interaction mapping has revealed binding of FMRP to a stretch of 20 amino acids adjacent to the C-terminal boundary of the Tudor domain of Tdrd3 (amino acids 704–723). FMRP in turn contacts Tdrd3 via two sequence elements, the second KH domain and a stretch comprising amino acids 430–486. As both elements are necessary but individually not competent in binding, we hypothesize that these two regions come into close proximity in the tertiary structure of the protein to form the binding platform for Tdrd3.

Most Fragile X patients exhibit mutations in the FMR1 gene that lead to transcriptional silencing and hence to the absence of FMRP. In a single case, however, also the substitution of an isoleucine to an asparagine in position 304 (I304N) elicits the disease. This mutation has been shown to interfere with several functions of FMRP, including mRNA binding, homodimerization and translational regulation. Remarkably, when tested in in vitro binding assays, FMRPI304N failed to interact with Tdrd3. These data are consistent with our finding that a region surrounding this residue is important for Tdrd3 binding.

It has recently been proposed that SMN interacts with FMRP and that this interaction is dependent on the amino acids 470–485 of FMRP (37). It will be interesting to analyze whether SMN and Tdrd3 compete for FMRP binding. However, SMN interaction is unaffected by the FMRP I304N mutation (37) and hence the binding sites appear to differ at least partially.

Tdrd3 as a component of stress granules
It is known from previous work that FMRP inhibits translation of mRNAs, most likely at the level of translation initiation (25). Current research focuses on the role of FMRP after stress stimuli such as heat and osmotic shock or oxidative stress. Under these conditions, FMRP concentrates in SGs, where translational silencing of mRNAs occurs (30). It is known that overexpression of FMRP and other components involved in mRNA metabolism likewise induces formation of SGs. Several lines of evidence suggest that the cytosolic foci induced by overexpression of Tdrd3 are equivalent to these structures. First, they disperse upon treatment of cells with either cycloheximide or emetine, showing them to be in equilibrium with polyribosomes bound to translated mRNAs. Second, arsenite treatment induces Tdrd3 containing granules even in cells expressing only low levels of Tdrd3. Tdrd3 in these cells has no aggregates in the absence of this stress-inducing agent. Last, Tdrd3 co-localizes with three known SG marker proteins, namely TIAR, DAP5 and FMRP, in these foci. Based on these results, we conclude that Tdrd3 is a novel component in SGs where it interacts with FMRP.

Our data raise the possibility that Tdrd3 functions in FMRP-mediated translational silencing. Consistent with this is our preliminary observation that Tdrd3 itself represses translation in an in vitro system (data not shown). Since Tdrd3 binds to Lys48-linked tetraubiquitin, it is tempting to speculate that Tdrd3 provides a link between translational regulation in SGs and the turnover of proteins mediated by the ubiquitin-proteasome system. Alternatively and not mutually exclusive, Tdrd3 may aid the removal of proteins co-recruited with the silenced RNP by proteasomal degradation. It is worth to mention that HDAC6, another protein containing an ubiquitin binding motif has recently been shown to be necessary for SG formation (38) and that interfering with the ubiquitin-proteasome system can induce SGs (39). Clearly, it will be interesting to identify proteins whose stability and/or function is influenced by Tdrd3.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
SUPPLEMENTARY MATERIAL
 FUNDING
 REFERENCES
 
Cloning of Tdrd3-cDNA
Tdrd3-cDNA was amplified using the ‘Marathon human brain cDNA’ library (Clontech) with transcript-specific primers. 5'- and 3'-RACE-PCR was carried out using the ‘Marathon RACE Kit’ (Clontech) according to the manufacturer’s instructions. Tdrd3-cDNA and fragments thereof, respectively, were subcloned into pET21a-ZZ for expression with a ZZ tag and pHA, an N-terminal HA tag containing derivative of pcDNA3.1 (Invitrogen). FMRP, FXR1 and FXR2 constructs were generated as described (25).

Preparation of recombinant proteins and antibodies
Recombinant FMRP, FXR1 and FXR2 proteins were expressed and purified as described previously (25). pET21a–ZZ–Tdrd3 and pGEX5X1–Tdrd3 were transformed in E. coli BL21(DE3) and after induction with 1 mM IPTG expression was allowed for 5 h at 26°C. Pelleted bacteria were resuspended in 500 mM NaCl, 50 mM Tris–HCl (pH 7.4), 5 mM MgCl2, 0.02% (v/v) Igepal CA630 (P500) supplemented with 0.5 mM PMSF. Suspensions were lysed by sonication and cleared by centrifugation. GST- or His-tagged fusion protein containing supernatants were incubated with Glutathione–Sepharose (GE Life Sciences) or Ni-NTA (Qiagen), respectively, washed with the same buffer except for 300 mM NaCl (P300) and bound proteins were eluted with 10 mM glutathione in P300 for GST-fusion proteins or 150 mM imidazole in P300 for His-tagged proteins. After elution, proteins were dialysed against P300. In vitro translated proteins were generated by coupled transcription and translation in reticulocyte lysate (Promega, Madison, WI, USA). Rabbits were immunized with His-tagged Tdrd3 protein as described (6) and the serum affinity purified using His-Tdrd3 covalently bound to NHS-activated Sepharose (GE Life Sciences). Commercial primary antibodies were purchased from Santa Cruz (goat polyclonal anti-DAP5: sc-13736), Covance (mouse monoclonal anti-HA antibody: HA.11) and abcam (rabbit polyclonal anti-TIAR; ab26257).

Protein binding assays
For protein binding assays, lysates containing 1–2 µg of ZZ-tagged fusion proteins were incubated on 30 µl of IgG-Sepharose (GE Life Sciences). For GST-fusion proteins, 1–2 µg of the purified protein were incubated with 30 µl Glutathione–Sepharose. After washing, with P300, 1–5 µl of an in vitro translation reaction or 1–2 µg of purified recombinant protein were added and binding was allowed for 2 h at 4°C. After washing, the affinity matrix was transferred to a new tube and bound proteins were recovered with SDS-loading buffer and analyzed by SDS–PAGE. For ubiquitin binding assays, GST-fusion proteins were immobilized on Glutathione–Sepharose and incubated with monomeric ubiquitin, Lys48- and Lys63-linked tetraubiquitin, respectively. After washing with PBS containing 0.5% Tween 20, proteins were eluted with SDS-loading buffer and immunoblot was performed using anti-ubiquitin antibody.

Yeast two-hybrid screen
Tdrd3 was cloned into the pGBKT7 vector and transformed in the PJ69-2A bait strain. After selection on SDC medium without tryptophane, mating with the Y187 prey strain that was pre-transformed with a human brain cDNA-bank in the pACT2 vector (Matchmaker, Clontech) was performed. Colonies growing on SDC-Medium without leucine, tryptophane, adenine and histidine were assayed for β-galactosidase activity and positive clones recovered for plasmid preparation.

Immunocytochemistry and stress treatment
HeLa cells were grown on cover slips in DMEM medium containing 10% fetal calf serum at 37°C and 5% CO2. One day after transfection with Nanofectin (PAA), cells were fixed either with 2% formaldehyde, permeabilized with 0.2% Triton X 100 and blocked with 1% bovine serum albumin (BSA) or in case of drug treatment fixed with ice-cold methanol/acetone (1:1) and blocked with 1% BSA. Treatment with cycloheximide (50 µg/ml), emetine (20 µg/ml) or arsenite (500 µM) was performed for 1 h immediately before fixation. After incubation with primary antibody, cells were washed and incubated with an appropriate secondary antibody conjugated to AlexaFluor594 or AlexaFluor488 (Invitrogen). After washing, cells were embedded in DAPI containing mounting medium and analyzed with a Zeiss Axiovert 200M equipped with a Plan Apochromat 63x/1.40 Oil immersion objective and Nomarski optics. Images were acquired with AxioVision software and processed using Adobe Photoshop CS3.

Northern blotting
A Northern blot membrane with 2 µg polyA+mRNA from eight different human tissues was purchased from Clontech. Tdrd3 cDNA was radioactively labeled using Klenow fragment with the ‘Megaprime DNA labeling kit’ (GE Life Sciences) according to the manufacturer’s description. The membrane was blocked with 900 mM NaCl, 50 mM NaH2PO4, 5 mM EDTA, 50% (v/v) formamide, 0.1% (w/v) SDS, 0.04% (w/v) Ficoll 400, 0.04% (w/v) polyvinylpyrrolidone, 0.04% (w/v) BSA, 100 µg/ml herring sperm DNA followed by hybridization for 3 h at 42°C. Washing was carried out once with 2x SSC+buffer at 42°C, and consecutively with decreasing concentration and increasing temperature until 0.1x SSC at 60°C (1x SSC+ is 150 mM NaCl, 15 mM sodium citrate, pH 7.0, 0.1% SDS). Finally, the membrane was exposed on X-ray film.


   SUPPLEMENTARY MATERIAL
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
FUNDING
 REFERENCES
 
Supplementary Material is available at HMG Online.


   FUNDING
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
REFERENCES
 
This work was supported by the DFG grants FOR855 and SFB581 to U.F.


   ACKNOWLEDGEMENTS
 
We thank F. Melchior and S. Jentsch for reagents and A. Chari for critically reading this manuscript.

Conflict of Interest statement. None declared.


   FOOTNOTES
 
{dagger} The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors. Back

Sequence information has been submitted to GenBank with the accession number EU643838. Back

{ddagger} Present address: Roche Diagnostics GmbH, Nonnenwald 2, D-82377 Penzberg, Germany. Back

§ Present address: TRION Pharma GmbH, Frankfurter Ring 193a, D-80807 München, Germany. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 SUPPLEMENTARY MATERIAL
 FUNDING
 REFERENCES
 

  1. Selenko P., Sprangers R., Stier G., Buhler D., Fischer U., Sattler M. SMN tudor domain structure and its interaction with the Sm proteins. Nat. Struct. Biol. (2001) 8:27–31.[CrossRef][Web of Science][Medline]

  2. Ponting C.P. Tudor domains in proteins that interact with RNA. Trends Biochem. Sci. (1997) 22:51–52.[CrossRef][Web of Science][Medline]

  3. Lamb F.S., Barna T.J., Goud C., Marenholz I., Mischke D., Schutte B.C. Complex RNA processing of TDRKH, a novel gene encoding the putative RNA-binding tudor and KH domains. Gene (2000) 246:209–218.[CrossRef][Web of Science][Medline]

  4. Chuma S., Hiyoshi M., Yamamoto A., Hosokawa M., Takamune K., Nakatsuji N. Mouse Tudor Repeat-1 (MTR-1) is a novel component of chromatoid bodies/nuages in male germ cells and forms a complex with snRNPs. Mech. Dev. (2003) 120:979–990.[CrossRef][Web of Science][Medline]

  5. Sabeur K., Ball B.A., Corbin C.J., Conley A. Characterization of a novel, testis-specific equine serine/threonine kinase. Mol. Reprod. Dev. (2008) 75:867–873.[CrossRef][Web of Science][Medline]

  6. Meister G., Hannus S., Plottner O., Baars T., Hartmann E., Fakan S., Laggerbauer B., Fischer U. SMNrp is an essential pre-mRNA splicing factor required for the formation of the mature spliceosome. EMBO J. (2001) 20:2304–2314.[CrossRef][Web of Science][Medline]

  7. Boswell R.E., Mahowald A.P. Tudor, a gene required for assembly of the germ plasm in Drosophila melanogaster. Cell (1985) 43:97–104.[CrossRef][Web of Science][Medline]

  8. Breitwieser W., Markussen F.H., Horstmann H., Ephrussi A. Oskar protein interaction with Vasa represents an essential step in polar granule assembly. Genes Dev. (1996) 10:2179–2188.[Abstract/Free Full Text]

  9. Lefebvre S., Bürglen L., Reboullet S., Clermont O., Burlet P., Viollet L., Benichou B., Cruaud C., Millasseau P., Zeviani M., et al. Identification and characterization of a spinal muscular atrophy-determining gene. Cell (1995) 80:155–165.[CrossRef][Web of Science][Medline]

  10. Meister G., Buhler D., Pillai R., Lottspeich F., Fischer U. A multiprotein complex mediates the ATP-dependent assembly of spliceosomal U snRNPs. Nat. Cell Biol. (2001) 3:945–949.[CrossRef][Web of Science][Medline]

  11. Buhler D., Raker V., Luhrmann R., Fischer U. Essential role for the tudor domain of SMN in spliceosomal U snRNP assembly: implications for spinal muscular atrophy. Hum. Mol. Genet. (1999) 8:2351–2357.[Abstract/Free Full Text]

  12. Brahms H., Meheus L., de Brabandere V., Fischer U., Lührmann R. Symmetrical dimethylation of arginine residues in spliceosomal Sm protein B/B' and the Sm-like protein LSm4, and their interaction with the SMN protein. RNA (2001) 7:1531–1542.[Abstract]

  13. Friesen W.J., Massenet S., Paushkin S., Wyce A., Dreyfuss G. SMN, the product of the spinal muscular atrophy gene, binds preferentially to dimethylarginine-containing protein targets. Mol. Cell (2001) 7:1111–1117.[CrossRef][Web of Science][Medline]

  14. Cote J., Richard S. Tudor domains bind symmetrical dimethylated arginines. J. Biol. Chem. (2005) 280:28476–28483.[Abstract/Free Full Text]

  15. Penagarikano O., Mulle J.G., Warren S.T. The pathophysiology of fragile x syndrome. Annu. Rev. Genomics Hum. Genet. (2007) 8:109–129.[CrossRef][Web of Science][Medline]

  16. Yin J., Sobeck A., Xu C., Meetei A.R., Hoatlin M., Li L., Wang W. BLAP75, an essential component of Bloom’s syndrome protein complexes that maintain genome integrity. EMBO J. (2005) 24:1465–1476.[CrossRef][Web of Science][Medline]

  17. Theobald D.L., Mitton-Fry R.M., Wuttke D.S. Nucleic acid recognition by OB-fold proteins. Annu. Rev. Biophys. Biomol. Struct. (2003) 32:115–133.[CrossRef][Web of Science][Medline]

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

  19. Dieckmann T., Withers-Ward E.S., Jarosinski M.A., Liu C.F., Chen I.S., Feigon J. Structure of a human DNA repair protein UBA domain that interacts with HIV-1 Vpr. Nat. Struct. Biol. (1998) 5:1042–1047.[CrossRef][Web of Science][Medline]

  20. Bertolaet B.L., Clarke D.J., Wolff M., Watson M.H., Henze M., Divita G., Reed S.I. UBA domains mediate protein–protein interactions between two DNA damage-inducible proteins. J. Mol. Biol. (2001) 313:955–963.[CrossRef][Web of Science][Medline]

  21. Grant R.P., Hurt E., Neuhaus D., Stewart M. Structure of the C-terminal FG-nucleoporin binding domain of Tap/NXF1. Nat. Struct. Biol. (2002) 9:247–251.[CrossRef][Web of Science][Medline]

  22. Geiss-Friedlander R., Melchior F. Concepts in sumoylation: a decade on. Nat. Rev. Mol. Cell. Biol. (2007) 8:947–956.[CrossRef][Web of Science][Medline]

  23. Siomi M.C., Siomi H., Sauer W.H., Srinivasan S., Nussbaum R.L., Dreyfuss G. FXR1, an autosomal homolog of the fragile X mental retardation gene. EMBO J. (1995) 14:2401–2408.[Web of Science][Medline]

  24. Vasudevan S., Steitz J.A. AU-rich-element-mediated upregulation of translation by FXR1 and Argonaute 2. Cell (2007) 128:1105–1118.[CrossRef][Medline]

  25. Laggerbauer B., Ostareck D., Keidel E.M., Ostareck-Lederer A., Fischer U. Evidence that fragile X mental retardation protein is a negative regulator of translation. Hum. Mol. Genet. (2001) 10:329–338.[Abstract/Free Full Text]

  26. Jin P., Alisch R.S., Warren S.T. RNA and microRNAs in fragile X mental retardation. Nat. Cell Biol. (2004) 6:1048–1053.[CrossRef][Web of Science][Medline]

  27. Jin P., Zarnescu D.C., Ceman S., Nakamoto M., Mowrey J., Jongens T.A., Nelson D.L., Moses K., Warren S.T. Biochemical and genetic interaction between the fragile X mental retardation protein and the microRNA pathway. Nat. Neurosci. (2004) 7:113–117.[CrossRef][Web of Science][Medline]

  28. Anderson P., Kedersha N. RNA granules. J. Cell Biol. (2006) 172:803–808.[Abstract/Free Full Text]

  29. Anderson P., Kedersha N. Stress granules: the Tao of RNA triage. Trends Biochem. Sci. (2008) 33:141–150.[Web of Science][Medline]

  30. Mazroui R., Huot M., Tremblay S., Filion C., Labelle Y., Khandjian E.W. Trapping of messenger RNA by Fragile X Mental Retardation protein into cytoplasmic granules induces translation repression. Hum. Mol. Genet. (2002) 11:3007–3017.[Abstract/Free Full Text]

  31. Kedersha N., Stoecklin G., Ayodele M., Yacono P., Lykke-Andersen J., Fritzler M.J., Scheuner D., Kaufman R.J., Golan D.E., Anderson P. Stress granules and processing bodies are dynamically linked sites of mRNP remodeling. J. Cell Biol. (2005) 169:871–884.[Abstract/Free Full Text]

  32. Kedersha N., Anderson P. Mammalian stress granules and processing bodies. Methods Enzymol. (2007) 431:61–81.[CrossRef][Web of Science][Medline]

  33. Kedersha N.L., Gupta M., Li W., Miller I., Anderson P. RNA-binding proteins TIA-1 and TIAR link the phosphorylation of eIF-2 alpha to the assembly of mammalian stress granules. J. Cell Biol. (1999) 147:1431–1442.[Abstract/Free Full Text]

  34. Nousch M., Reed V., Bryson-Richardson R.J., Currie P.D., Preiss T. The eIF4G-homolog p97 can activate translation independent of caspase cleavage. RNA (2007) 13:374–384.[Abstract/Free Full Text]

  35. De Boulle K., Verkerk A.J., Reyniers E., Vits L., Hendrickx J., Van Roy B., Van den Bos F., de Graaff E., Oostra B.A., Willems P.J. A point mutation in the FMR-1 gene associated with fragile X mental retardation. Nat. Genet. (1993) 3:31–35.[CrossRef][Web of Science][Medline]

  36. Pickart C.M., Fushman D. Polyubiquitin chains: polymeric protein signals. Curr. Opin. Chem. Biol. (2004) 8:610–616.[CrossRef][Web of Science][Medline]

  37. Piazzon N., Rage F., Schlotter F., Moine H., Branlant C., Massenet S. In vitro and in cellulo evidences for association of the survival of motor neuron complex with the fragile X mental retardation protein. J. Biol. Chem. (2008) 283:5598–5610.[Abstract/Free Full Text]

  38. Kwon S., Zhang Y., Matthias P. The deacetylase HDAC6 is a novel critical component of stress granules involved in the stress response. Genes. Dev. (2007) 21:3381–3394.[Abstract/Free Full Text]

  39. Mazroui R., Di Marco S., Kaufman R.J., Gallouzi I. Inhibition of the ubiquitin-proteasome system induces stress granule formation. Mol. Biol. Cell (2007) 18:2603–2618.[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
 RNAHome page
N. G. Farny, N. L. Kedersha, and P. A. Silver
Metazoan stress granule assembly is mediated by P-eIF2{alpha}-dependent and -independent mechanisms
RNA, October 1, 2009; 15(10): 1814 - 1821.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Supplementary Data
Right arrow All Versions of this Article:
17/20/3236    most recent
ddn219v1
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 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 (3)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Linder, B.
Right arrow Articles by Fischer, U.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Linder, B.
Right arrow Articles by Fischer, U.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?