INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

RNA transport and localization provide an efficient way to distribute genetic information and to allow different portions of the cell to establish their own biochemical fates (12, 20, 32, 44). Examples of this process have been described in many different organisms and cell types. RNA localization and/or localized translation are linked to different biological processes such as asymmetric cell division (7, 28, 39, 45), long-term potentiation (30), synaptic transmission (40), cell motility (23), and axis formation in oocytes (44). It is now apparent that many determinants of RNA localization are conserved among these systems. The process of mRNA localization is initiated by association of RNA with one or more RNA-binding proteins (RBPs) through a targeting signal most commonly located in the 3' untranslated region (3' UTR) of the transcripts. This association results in the formation of large ribonucleoprotein complexes (RNPs) (1, 13, 24). These complexes then migrate along the cytoskeleton to their final destination, where they are anchored and translated. For the entire localization process, translation of localized mRNAs needs to be tightly regulated. Specific signals are important for repressing translation during mRNA transport and derepressing translation once RNPs are properly localized (15, 21, 22). The crucial role played by the cytoskeleton in many steps of transport, anchoring, and translation (2, 34, 41, 47) is another feature which appears to be conserved in numerous RNA localization systems. The multistep process of RNA localization is dependent on specific trans-acting proteins. Thus, it is essential to identify and characterize these proteins and to study their localization and regulation. Recent work has identified several classes of proteins as components of RNP involved in mRNA transport (20). These include members of the double-stranded RBP (dsRBP) family (25, 43), homologues of the zipcode-binding protein (10, 11, 16, 36), and members of the hnRNP family (3, 17, 33).

Mammalian Staufen, a member of the dsRBP family, contains four copies of the dsRBD consensus motif (29, 46), now designated dsRBD2 to dsRBD5 for consistency with the Staufen domains in Drosophila (43). In vitro, Staufen was shown to bind dsRNA without sequence specificity (29, 46). Molecular mapping of the functional domains related the RNA-binding activity mainly to dsRBD3, with a weaker activity mapped to dsRBD4 (46). Similarly, the spacer region between dsRBD4 and dsRBD5, which resembles the tubulin-binding domain of MAP1B, was shown to bind tubulin in vitro (46). In fibroblasts, Staufen is associated with polysomes and the rough endoplasmic reticulum (RER) (29, 46), whereas in neurons, it is associated with both the RER and microtubules (19). In expression studies in neurons, Staufen was demonstrated to be a component of RNA-containing granules migrating in both anterograde and retrograde manners along the dendrites (25). This was persuasive evidence that Staufen is involved in mRNA transport in mammals. However, challenging questions remain about the precise role of each Staufen isoform; the function of the multiple domains in Staufen protein; and the nature of its interaction with RNAs, protein cofactors, RER, and the cytoskeleton.

Recent studies in Drosophila have provided important clues about the function of Staufen. In Drosophila, Staufen is necessary for bicoid and oskar mRNA localization to the anterior and posterior poles of the oocyte, respectively (21, 42), and for prospero mRNA localization in neuroblasts (7, 27, 39). Out of the five copies of the dsRBD consensus sequences, dsRBD3 was shown to bind bicoid and prospero mRNAs in vitro (27, 43); dsRBD5 was shown to be involved in protein-protein interactions (39), demonstrating that Staufen is involved in both RNA-protein and protein-protein interactions. Although direct binding to bicoid RNA has not yet been shown in vivo, intermolecular bicoid RNA-RNA interactions are important for recruiting Staufen in the RNP complexes (14). In contrast, Staufen directly interacts with Oskar protein via its N-terminal domain in oocytes (5) and with Inscuteable and Miranda in neuroblasts through its C-terminal half and fifth dsRBD, respectively (27, 39). Staufen expression is also important for derepression of the localised oskar mRNA translation (5).

Recently we reported the molecular cloning and characterization of human (hStau) and mouse (mStau) Staufen (46). We now report that a novel endogenous Staufen isoform (mStauⁱ) containing a six-amino-acid insertion within its major dsRBD (dsRBD3) shows a severe reduction in its RNA-binding capacity in vitro. Our results show that this isoform, along with mStau, are components of RNA-protein complexes and that the ratio of the two isoforms is important for the proper subcellular localization of mStauⁱ. Finally, analyses of Staufen-containing complexes demonstrate that increasing the incorporation of mStauⁱ drastically reduces the amount of RNA in these complexes. These results provide new insights into the mechanism of regulation of mStau function in vivo.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Molecular cloning and sequencing of the cDNAs. The cloning of the short isoform of a mouse staufen homologue from a fetal total mouse cDNA library was previously described (46). DNAs from the isolated GT10 clones were subcloned into a Bluescript vector (Stratagene). Double-stranded DNAs were sequenced by the dideoxynucleotide method according to the Sequenase protocols (United States Biochemical Corp.).

RT-PCR. Reverse transcription-PCR (RT-PCR) was conducted according to the Perkin-Elmer GeneAmp RNA PCR protocol. This experiment was performed with 1 µg of total RNA isolated from different tissues and cell lines. Reverse transcription was carried out in the presence of oligo(dT)₁₆ primer, and PCR amplification incorporated the use of the mouse-specific sense and antisense primers 5'-GACCACCCGTGAAACACGATGCCCC-3' (positions 495 to 519; GenBank accession number AF061942) and 5'-TCCCTTCACCTTCCCCCACAAACTCCC-3' (positions 755 to 729; GenBank accession number AF061942), respectively. For human and monkey cDNAs, we used the human-specific sense primer 5'-GACAGGCTGCGAAACACGATGCTGC-3' (positions 478 to 502; GenBank accession number AF061940) and the mouse antisense primer. Aliquots of the amplified products were collected after 25, 28, 30, 32, and 35 PCR cycles and were run on agarose gels to test the linearity of the PCR reaction (not shown). Only PCR products obtained by 32 cycles of amplification are shown in Fig. 2.

Construction of fusion proteins. Maltose-binding protein (MBP)-dsRBD3 and MBP-dsRBD3ⁱ were constructed first by amplifying a fragment using the primer pair 5'-CAATGTATAAGCCCGTGGACCC-3' and 5'-AAAAAGCTTGTGCAAGTCTACTAATAGGATTCATCC-3' with Vent DNA polymerase (New England BioLabs). The resulting product was digested with HindIII and cloned into the EagI* and HindIII sites of a modified pMal-c vector (pMal-stop). To produce this vector, we introduced stop codons at the HindIII site of pMal-c by the ligation of the annealed complementary oligonucleotides 5'-AGCTTAATTAGCTGAC-3' and 5'-AGCTGTCAGCTAATTA-3'. EagI* was created by filling in the cohesive ends of EagI-digested pMal-c vector using the Klenow fragment of DNA polymerase I. This vector was used to purify the 398-bp PstI and HindIII fragment, which then was subcloned in the pMAL-stop vector to generate the mStau-RBD3 construct. These MBP-dsRBD3 fusion plasmids were introduced into Escherichia coli strain BL-21. The fusion proteins were obtained after induction with 1 mM isopropyl--D-thiogalactopyranoside (IPTG) for 2 to 3 h. Cells were lysed in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer for immediate use or frozen at 80°C for storage.

Antibody production and Western blotting. Polyclonal anti-Staufen antibodies were obtained by injection of purified His-hStau fusion protein into rabbits as previously described (46). For Western blotting, cells were lysed in 0.5% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, 1 µg of aprotinin per ml, and 1 µg of pepstatin A per ml in phosphate-buffered saline (PBS). Protein extracts were quantified by the Bradford method (Bio-Rad). Equal amounts of proteins were separated on SDS-12% polyacrylamide gels and transferred to nitrocellulose membranes. Membranes were blocked for 30 min in Tris-buffered saline (TBS) plus 5% dry milk and incubated with primary antibodies in Tris-buffered saline plus 0.05% Tween for 1 h at room temperature. Detection was performed by incubating the blots with peroxidase-conjugated anti-rabbit immunoglobulin G (IgG) antibodies (Dimension Labs) and with the Supersignal substrate (Pierce).

RNA binding assay. Bacterial extracts from IPTG-induced cultures or affinity-purified fusion proteins were separated on SDS-7.5% polyacrylamide gels, and the proteins were transferred to nitrocellulose membranes. Increasing amounts of bacterial extracts and of affinity-purified fusion proteins were separated by SDS-PAGE. Amounts of mStau and mStauⁱ were quantitated by Western blotting. Equal amounts were then loaded on the gel for the Northwestern assay. RNA interaction was detected by Northwestern assay using ³²P-labeled 3' UTR of bicoid RNA as previously described (46).

Immunofluorescence. mStau tagged with hemagglutinin epitope (HA) and mStau tagged with green fluorescent protein (GFP) were constructed by PCR amplification of the full-length cDNA using the primer pair 5'-TCTGGATCCGAAAGTATAGCTTCTACCATTG-3' and 5'-TACAATCTAGATTATCAGCGGCCGCACCTCCCGCACGCTGAAAC-3'. The 3' primer was synthesized with a NotI site just upstream from the stop codon, allowing ligation of a NotI cassette containing either the GFP sequence or three copies of the HA tag. After digestion with BamHI and XbaI, the resulting fragment was cloned in Bluescript. The BamHI/XbaI fragment was then subcloned in the pCDNA3/RSV vector (18), and a NotI cassette was introduced at the NotI site. Mammalian cells were transfected transiently with the cDNAs by the calcium phosphate precipitation technique, fixed in 4% paraformaldehyde in PBS for 25 min at room temperature, and permeabilized with 0.3% Triton X-100 in PBS containing 0.1% bovine serum albumin (BSA). Nonspecific sites were then blocked with 1% BSA in PBS-0.3% Triton X-100 and incubated with mouse anti-HA for 1 h at room temperature as indicated. Cells were washed in permeabilization buffer and incubated with Texas red-conjugated secondary antibodies (Jackson Immunoresearch Laboratories, West Grove, Pa.) in blocking buffer for 1 h. GFP and GFP fusion proteins were detected by autofluorescence.

Immunoprecipitation. COS-1 cells were transfected with 6 µg of DNA using the FUGENE-6 reagent (Roche Biochemicals), washed in PBS 36 h posttransfection, and then harvested in 1 ml of lysis buffer (0.5% Triton X-100-50 mM Tris-Cl [pH 7.5], 15 mM EGTA, 1 mM dithiothreitol, 10 µg of phenylmethylsulfonyl fluoride per ml, standard protease inhibitor cocktail). The lysate was incubated at 4°C for 1 h and centrifuged at 14,000 × g for 15 min. Proteins in the supernatant were quantified, and aliquots were analyzed by SDS-PAGE and Western blotting to confirm expression of each protein. For immunoprecipitation, 1 mg of total extract was preincubated with anti-HA ascites fluid (1/250) for at least 6 h at 4°C and centrifuged at 14,000 × g for 15 min. The supernatant was incubated for 4 h at 4°C in the presence of 50 µl of a ~60% protein A-Sepharose slurry equilibrated in lysis buffer and then centrifuged at 500 × g for 30 s at 4°C. The pellet was washed four times in lysis buffer, resuspended in reducing sample buffer, and analyzed by SDS-PAGE and Western blotting using polyclonal anti-GFP (1/100; Clontech) or anti-Stau (1/500) antibodies.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

mStauⁱ is a novel mStau isoform produced by alternative splicing. We previously reported the cloning of the short isoform of mStau from a mouse total embryonic cDNA library (46). The encoded mStau protein is 91% identical to its human counterpart, and its genetic organization is identical. We showed that mStau binds dsRNA and tubulin in vitro and that the RNA-binding activity maps mainly to dsRBD3 but also weakly to dsRBD4 (46). Analysis of the mouse staufen cDNAs allowed us to identify a novel endogenous transcript containing an 18-bp insertion within the sequence coding for dsRBD3 (Fig. 1). We designated this isoform mStauⁱ. Except for the sequence of the insertion, the cDNAs coding for each isoform are 100% identical. In addition, hybridization of mouse genomic DNA with a 191-bp fragment of staufen cDNA (nucleotides 57 to 248) revealed a single band on a Southern blot, independent of the restriction enzyme used (Fig. 2A). The staufen gene is therefore present in the mouse genome as a single copy. Characterization of the corresponding genomic sequence further revealed that the transcripts are generated by differential splicing, choosing either of two splicing acceptor sites (Fig. 2B). Similar genomic organization is observed in the human genome, suggesting that this phenomenon is conserved among mammals (6).

View larger version (11K): [in this window] [in a new window]   FIG. 1.   Sequence comparison of Drosophila (Dm; GenBank accession number M69111) and mouse (Mm) Staufen dsRBD3 domains. The position and sequence of the six-amino-acid insert in mStaui are indicated. Colons and dots represent identical and conserved amino acids. Amino acids included in the consensus described by Krovat and Jantsch (26) are identified in bold. Topological motifs adopted by the Drosophila domain (8) are presented schematically above the sequences.

View larger version (31K): [in this window] [in a new window]   FIG. 2.   Characterization of the mStau gene and transcripts. (A) Southern blot analysis. Mouse genomic DNA was digested with BamHI (lane B), EcoRI (lane E), and SacI (lane S), transferred to a nitrocellulose membrane, and hybridized with a 191-nucleotide cDNA fragment that covers a single exon (nucleotides 57 to 248). DNA molecular weight markers in kilobase pairs are indicated on the left. (B) Genomic characterization of the splicing sites. Genomic and cDNA sequences are aligned, and the positions of splicing consensus sequences are indicated. (C) Differential splicing of the mStau gene. RT-PCR amplification of mRNAs isolated from mouse tissues (lanes 2 to 9 and 13), human HeLa cells (lane 11), and monkey COS-1 cells (lane 12). Lane 2, brain; lane 3, heart; lane 4, liver; lane 5, lungs; lane 6, spleen; lane 7, kidney; lane 8, male genitals; lane 9, female genitals, lane 11, HeLa cells; lane 12, COS-1 cells; lane 13, mouse NIH/3T3; lanes 1 and 10, negative controls.

mStauⁱ shows impaired dsRNA-binding activity. Comparison of the mStauⁱ amino acid sequence with the known nuclear magnetic resonance structure of dsRBD3 of Drosophila Staufen and of other dsRBDs (8, 26) showed that the six-amino-acid insert is localized within the first beta strand of the consensus ---- motif (Fig. 1) (8). To test whether the six-amino-acid insertion modifies the dsRNA-binding capacity of mStauⁱ, fusions of the full-length mStau and mStauⁱ proteins (His-mStau and His-mStauⁱ) and of their isolated dsRBD3s (MBP-dsRBD3 and MBP-dsRBD3ⁱ) were expressed in bacteria. Their RNA-binding capacity was analyzed by the Northwestern assay (46). While the MBP-dsRBD3 strongly bound the dsRNA probe, the binding of MBP-dsRBD3ⁱ was strongly impaired and detected only after an extended exposure time (Fig. 3). This difference in binding was observed both with the affinity-purified fusion proteins and the bacterial extracts. Impaired binding was also visible with the full-length protein, although the difference was less dramatic (Fig. 3). This can be explained by the presence of dsRBD4, a weak dsRBD that contributes to the dsRNA-binding activity of full-length proteins (46). Finally, under these conditions, the dsRNA probe did not bind other proteins in the bacterial extracts, overexpressed MBP, or BSA (Fig. 3). These results demonstrate that the alternative splicing event produces a protein with impaired RNA-binding activity.

View larger version (31K): [in this window] [in a new window]   FIG. 3.   RNA binding assay. Bacterially expressed dsRBD3 and full-length fusion proteins after affinity purification or in the crude bacterial extracts (as indicated) were electrophoresed on a polyacrylamide gel, transferred to nitrocellulose, and incubated with 32P-labeled 3' UTR bicoid RNA. After extensive washing, bound RNA was detected by autoradiography. Controls included bacterial crude extract overexpressing MBP and 5 µg of BSA.

Overexpressed mStauⁱ localizes in discrete RER-containing granules. We and others have reported that hStau localizes to the RER (29, 46). To determine the subcellular localization of mStau and mStauⁱ, we constructed fusion proteins with a GFP or HA tag fused to the C-terminal end and transfected these constructs into COS-1 cells. We first observed by Western blotting that the fusion proteins were not degraded and that they were expressed at about the same levels (not shown). Their subcellular distribution was then observed by fluorescence microscopy. As observed for hStau-GFP, mStau-GFP also exhibited a distribution typical of the RER with an abundant tubulovesicular distribution (Fig. 4A). A different distribution was observed when COS-1 cells were transfected with the cDNA coding for mStauⁱ-GFP (Fig. 4C). The protein was found in large clusters throughout the cytoplasm. Treatment of the cells with Triton X-100 prior to fixation did not wash away either the mStau-GFP (Fig. 4B) or mStauⁱ-GFP (Fig. 4D) signal, indicating that both proteins are associated with cytoskeletal elements resistant to the detergent extraction. As reported previously (46), GFP alone showed a diffuse cytoplasmic distribution and was completely extractable by prior treatment with Triton X-100 (not shown). Quantifying the percentage of transfected cells exhibiting a granular or reticular distribution revealed that the vast majority of mStauⁱ-transfected cells showed a granular distribution, while only a minority of mStau-transfected cells exhibited a granular distribution (Table 1). This indicates that overexpressed mStau and mStauⁱ are different in subcellular distribution but that both can exhibit a granular and a reticular ER distribution.

View larger version (66K): [in this window] [in a new window]   FIG. 4.   Subcellular localization of the mStau and mStaui proteins. COS-1 cells were transfected with cDNAs coding for either mStau-GFP (A and B) or mStaui-GFP (C and D). Untreated cells (A and C) or Triton X-100-treated cells (B and D) were fixed, and GFP autofluorescence was visualized. Bar = 10 µm.

View larger version (56K): [in this window] [in a new window]   FIG. 5.   Colocalization of mStau and mStaui with markers of the RER by confocal microscopy. cDNAs coding for mStau-GFP (A to C) and for mStaui-GFP (D to F) fusion proteins were transfected into COS-1 cells. Triton X-100-treated cells were fixed and labeled with anticalnexin (A and D). GFP was detected by autofluorescence using the FITC channel (B and E), whereas anticalnexin was detected with Texas red-conjugated anti-rabbit IgG antibodies using the rhodamine channel. C and F are superpositions of A and B and D and E, respectively. mStau colocalizes with calnexin (CNX)-labeled ER, and calnexin is associated with mStaui-labeled granules. Cells expressing mStaui-GFP, but not labeled with anticalnexin antibodies, and untransfected cells labeled for calnexin did not present a signal in the rhodamine and in the fluorescein channel, respectively, showing that the observed signals are specific. Bar = 10 µm.

View larger version (60K): [in this window] [in a new window]   FIG. 6.   Rescue of mStaui phenotype by coexpression of mStau. COS-1 cells were cotransfected with cDNAs coding for mStaui-GFP and mStau-HA fusion proteins. Triton X-100-treated cells were fixed and labeled with anti-calnexin (A) or anti-HA (D) antibodies. mStaui-GFP was detected by autofluorescence using the FITC channel (B and E), whereas anticalnexin and anti-HA were detected with Texas red-conjugated anti-rabbit IgG antibodies using the rhodamine channel. C and F are superpositions of A and B and D and E, respectively. Controls (as described in the legend to Fig. 5) demonstrated that the observed signals are specific. Bar = 10 µm.

mStauⁱ is part of a multiple-isoform complex. To determine whether mStau and mStauⁱ are present in the same complexes, we performed coimmunoprecipitation assays from cells cotransfected with mStau and mStauⁱ. COS-1 cells were cotransfected with mStau-HA and mStauⁱ-GFP, and protein extracts were immunoprecipitated with anti-HA antibodies. The immunoprecipitated proteins were visualized by Western blotting using anti-hStau (Fig. 7A) and anti-GFP (Fig. 7B) antibodies. mStauⁱ-GFP was present in the immunoprecipitated mStau-HA pellets, demonstrating that the two proteins are components of a common complex (lane 5). Symmetrical results were obtained when the tags on the proteins were interchanged: mStau-GFP was also detected in immunoprecipitates of mStauⁱ-HA (lane 6). The controls that included the transfection of COS-1 cells with either mStau-GFP alone (lanes 1 and 2) or cotransfection of mStau-HA with GFP (lanes 3 and 4) were uniformly negative. The equal level of expression of each protein in the loading extracts was confirmed with anti-Staufen and anti-GFP antibodies on Western blots (not shown). Interestingly, a 55-kDa protein band corresponding to the size of the endogenous Staufen protein also coprecipitated with mStau-HA and mStauⁱ-HA. That the mStau fusion proteins are localized with the endogenous protein is particularly significant. These results demonstrate that the two isoforms are components of the same complexes in vivo.

View larger version (38K): [in this window] [in a new window]   FIG. 7.   mStau isoforms are present in the same complexes. Coimmunoprecipitation assay. Different combinations of tagged-Staufen isoforms were expressed in COS-1 cells as indicated above the gels. Cell extracts were immunoprecipitated with anti-HA antibodies, and the proteins were revealed by Western blotting using anti-hStau (A) and anti-GFP (B) antibodies. The positions of mStau/mStaui-GFP (Stau-GFP), mStau/mStaui-HA (Stau-HA), IgG, and GFP are indicated on the right. Numbers on the left are protein molecular weight markers in kilodaltons. The large arrowhead represents the position of the endogenous 55-kDa Staufen isoform.

mStauⁱ modulates the amount of RNA in mStau-containing complexes. The reduced capacity of mStauⁱ to bind RNAs and its ability to associate with mStau in vivo suggest that mStauⁱ could modulate the RNA-binding activity of the complexes and thus regulate the amount of RNA associated with it. To test this hypothesis, we transfected COS-1 cells with different amounts of the two isoforms, immunoprecipitated HA-tagged complexes, and analyzed the RNA content of the resulting precipitates (Fig. 8A). Multiple RNA bands were easily visible when immunoprecipitation was done from mStau-HA/mStau-GFP-transfected cells (Fig. 8A, lane 3). In contrast, a significant reduction in the amount of RNAs was attained when immunoprecipitated RNAs from the mStauⁱ-HA/mStau-GFP-transfected cells were analyzed (lane 4). This reduction was even more striking in the mStauⁱ-HA/mStauⁱ-GFP immunoprecipitates (lane 2). Western analysis of the amount of the immunoprecipitated proteins showed that the difference in the quantity of RNA cannot be due to less efficient immunoprecipitation of mStauⁱ-containing complexes (Fig. 8B). This experiment was repeated three more times with the same results. We conclude that mStau is a component of complexes containing RNAs and that mStauⁱ incorporation in these complexes modulates the amounts of associated RNA.

View larger version (46K): [in this window] [in a new window]   FIG. 8.   Expression of mStaui modulates the amount of RNAs in Staufen-containing particles. COS-1 cells were transfected with different combinations of the two Staufen isoforms as indicated above the gel, and the proteins were immunoprecipitated with anti-HA antibodies. RNAs were purified and separated on formaldehyde-agarose gels and analyzed by hybridization using 32P-labeled random oligonucleotides (A). (B) Proteins from the same immunoprecipitates were analyzed by SDS-PAGE and detected by Western blotting using anti-Staufen antibodies.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Despite an increasing number of reports describing mRNA transport, only recently have the first trans-acting proteins in mammals been identified. Their functions, molecular characteristics, interacting partners, and regulation remain largely unknown. Staufen is known to bind dsRNA, to associate with polysomes, the RER, and elements of the cytoskeleton, and to be expressed as multiple isoforms in mammalian cells. However, it was unclear how these molecular and cellular characteristics are integrated to fulfill and regulate all Staufen functions. In this paper, we report the following: (i) the characterization of a novel Staufen isoform with impaired RNA-binding properties; (ii) that dsRBD3 is involved not only in RNA binding but also in the proper localization on the RER; (iii) that Staufen isoforms are components of a common RNA-protein complex in vivo; and (iv) that the ratio of mStau and mStauⁱ modulates the amount of RNA present in the complex.

mStauⁱ modulates the RNA content of the Stau complexes. This study provides the first biochemical evidence that Staufen is a component of RNA-protein complexes. The immunoprecipitation experiment demonstrates that mStau complexes contain a limited number of RNAs. This corroborates previous in vivo results that showed that Staufen-containing particles colocalized with RNA-containing granules (19, 25). This is also consistent with a putative role for mammalian Staufen in mRNA transport and localization and suggests that if Staufen plays additional roles in mammals, this role is likely to be related at least to some aspect of RNA processing. In Drosophila, genetic evidence suggests that Staufen may also be involved in the regulation of translation of localized RNAs (5).

View larger version (61K): [in this window] [in a new window]   FIG. 9.   Effect of the SFLLTQ insertion on the structure of mStaui. (Top) The -helical surface of the mStaui region that includes the insertion. The six inserted residues are shown in fancy letters. Enclosed in large circles, the three inserted residues FLL plus residues F20, V22, and A23 form a wide hydrophobic cluster that can effect the secondary structure of the whole region. (Bottom) Prediction of the secondary structure for mStau (upper part) and mStaui (lower part) with use of the PHD software package. AA, amino acid sequence. PHD_sec (predicted secondary structure): H, helix; E, extended (sheet); blank, other (loop). Rel_sec, reliability index for the PHD_sec prediction (0 = low to 9 = high); Sub_sec, subset of the PHD_sec predictions, for all residues with an expected average accuracy >82%. The insertion of the six amino acids between residues 21 and 22 changes the secondary structure of the whole region from  structure to  helix.

mStauⁱ and the RER. The localization of mStauⁱ to cytoplasmic granules is likely a consequence of the overexpression of mStauⁱ: these structures are not observed endogenously and are visible in a low percentage of mStau-transfected cells. However, the rescue experiment shows that the formation of granules is due not to the overexpression of mStauⁱ per se but rather to an imbalance in the ratio of the two isoforms and that under normal conditions, mStauⁱ is associated with the reticular RER. One plausible explanation is that mStauⁱ cannot be properly localized independent of mStau, with which it associates in vivo. Can its normal distribution to reticular RER occur only in this form? This would suggest that mStauⁱ is likely a regulator of mStau function, as shown here for RNA incorporation in the mStau complexes. Interestingly, tendon cell differentiation in Drosophila was shown to be modulated by the balance between two isoforms of the RBP How (31). It is likely that the granular mStauⁱ phenotype is due to the structural modification of dsRBD3 or to its reduced RNA-binding activity. The formation of similar granules following overexpression of a mutant with a point mutation that completely abolishes the dsRBD3 RNA-binding activity, but not the structure of the domain, is consistent with the second hypothesis (M. Luo and L. DesGroseillers, unpublished data).