Drosophila Y14 Shuttles to the Posterior of the Oocyte and Is Required for Oskar Mrna Transport Olivier Hachet and Anne Ephrussi

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

1666 Research Paper

Drosophila Y14 shuttles to the posterior of the oocyte and is required for oskar mRNA transport Olivier Hachet and Anne Ephrussi

Background: mRNA localization is a powerful and widely employed Address: Developmental Biology Programme, mechanism for generating cell asymmetry. In Drosophila, localization of European Molecular Biology Laboratory, Meyerhofstrasse 1, 69117 Heidelberg, Germany. mRNAs in the oocyte determines the axes of the future embryo. oskar mRNA localization at the posterior pole is essential and sufficient for the Correspondence: Anne Ephrussi specification of the germline and the abdomen. Its posterior transport along E-mail: [email protected] the microtubules is mediated by Kinesin I and several proteins, such as Mago-nashi, which, together with oskar mRNA, form a posterior localization Received: 31 August 2001 complex. It was recently shown that human Y14, a nuclear protein that Revised: 26 September 2001 associates with mRNAs upon splicing and shuttles to the cytoplasm, Accepted: 26 September 2001 interacts with MAGOH, the human homolog of Mago-nashi. Published: 30 October 2001

Results: Here, we show that Drosophila Y14 interacts with Mago-nashi in Current Biology 2001, 11:1666–1674 vivo. Immunohistochemistry reveals that Y14 is predominantly nuclear and colocalizes with oskar mRNA at the posterior pole. We show that, in y14 0960-9822/01/$ – see front matter mutant oocytes, oskar mRNA localization to the posterior pole is  2001 Elsevier Science Ltd. All rights reserved. specifically affected, while the cytoskeleton appears to be intact.

Conclusions: Our findings indicate that Y14 is part of the oskar mRNA localization complex and that the nuclear shuttling protein Y14 has a specific and direct role in oskar mRNA cytoplasmic localization.

Background Integrity of the cytoskeleton is an essential requirement Asymmetric distribution of protein activities within the for mRNA localization in many systems, as demonstrated cell cytoplasm allows for the specification of subcellular by drug inhibitor studies [10]. While ASH1 mRNA local- domains, which is of particular importance in many physi- ization appears to be mediated by active transport along ological and developmental processes. Localization of the actin cytoskeleton [11, 12], oskar mRNA transport to messenger RNAs constitutes one efficient strategy to gen- the posterior pole is more likely to involve microtubules erate local protein concentrations and is a widely used [13], as is the case for MBP mRNA [14]. Microtubules mechanism among different cell types, from unicellular are thought to be implicated in at least two steps of oskar eukaryotes to vertebrates [1]. In the budding yeast Sac- mRNA localization. Before stage 7, as the microtubules charomyces cerevisiae, during cell division, ASH1 mRNA emanate from the oocyte and extend into the adjacent localizes to the daughter cell to repress mating type nurse cells, oskar mRNA is imported from the nurse cells switching [2, 3]. In oligodendrocytes, myelin basic protein into the oocyte, presumably via a minus end-directed (MBP) mRNA is localized to distal processes, where the motor [15]. After stage 7, when the microtubule cytoskele- protein is required for myelination of the central nervous ton is reorganized and plus ends are focused toward the system [4]. During Drosophila melanogaster oogenesis, the oocyte posterior, oskar mRNA is thought to be transported oocyte accumulates maternal components that provide to its posterior destination by Kinesin I motor protein the spatial information essential for development of the [16]. Little is known, however, about how oskar mRNA future embryo. This information is encoded by three ma- might be coupled to the motor. jor determinants, gurken, bicoid, and oskar, that respectively define the dorsal, anterior, and posterior regions of the Recent results have raised the possibility that mRNAs embryo [5]. The activities of these three determinants may acquire within the nucleus their competence to be are directed to the correct subcellular sites by localization localized in the cytoplasm [17]. The nuclear RNA binding of their messenger RNAs [6–8]. gurken mRNA is thought protein Loc1p is essential for ASH1 mRNA localization to be produced locally by the oocyte nucleus [9]. In con- [18], and the shuttling protein hnRNP A2 binds the MBP trast, oskar and bicoid mRNAs, which are produced by the mRNA cytoplasmic transport sequence specifically [19]. 15 nurse cells that are interconnected to the oocyte, are During Drosophila oogenesis, the predominantly nuclear exported across cytoplasmic bridges from the nurse cells shuttling protein, Mago-nashi, has been shown to play a to the oocyte, within which they are transported to their direct role in oskar mRNA posterior transport and colocal- destination. izes with oskar mRNA at the posterior pole until stage 9 Research Paper Drosophila Y14 required for oskar RNA localization Hachet and Ephrussi 1667

[20–22]. In addition to its specific role in oskar mRNA Figure 1 transport, Mago-nashi is involved in the reorientation of the microtubule network at stage 7. Another nuclear shuttling protein, Barentsz, has also been shown to be essential for oskar mRNA posterior transport [23].

The human homolog of Mago-nashi, MAGOH, was recently reported to interact with a novel RNA binding protein called Y14, encoded by the RBM8 gene [24, 25]. Y14 is a nuclear protein that is able to shuttle to the cytoplasm and is recruited to messenger RNAs upstream of exon-exon junctions upon splicing [26–28]. Y14 has been proposed to serve as a landmark that is deposited on mRNAs, influencing their cytoplasmic fate: nuclear export [26], nonsense-mediated decay [29–31], cytoplasmic localization, or translational control.

Here, we show that Drosophila Y14 interacts with Mago- nashi in vivo and shuttles out of the nucleus to the cytoplasm where it colocalizes with the oskar mRNA localization complex. Moreover, our data reveal that Y14 has a specific and direct role in oskar mRNA posterior transport.

Results Drosophila Y14 interacts with Mago-nashi in a yeast two-hybrid assay and in vivo Y14 and Mago-nashi are highly conserved proteins, from unicellular eukaryotes to vertebrates. MAGOH and Mago-nashi share 89% identity, and human and Drosophila Y14 share 59% identity. However, the Y14 homologs differ noticeably in one respect: Drosophila Y14 lacks most of the SR domain present in the human form. To test whether the Y14:MAGOH interaction is conserved in Dro- sophila melanogaster, we cloned the Drosophila y14 gene Y14 interacts with Mago-nashi in the yeast two-hybrid assay and in and designed an interaction test, making use of the LexA ovaries. (a) In the two-hybrid assay, the Y14:Mago-nashi interaction two-hybrid system [32]. We detected an interaction be- is detected both with LexA-NLS-Mago and LexA-Mago as baits (lanes 1 and 3) and with AD-Mago as prey (lane 5). In contrast, no self- tween Mago-nashi and Drosophila Y14 in both assay orien- interaction was detected for Y14 and Mago-nashi (lanes 2, 4, and 6). tations, using two independent Mago-nashi baits that dif- The upper panel shows yeast grown on glucose medium, conditions fer with regard to inclusion of a nuclear localization signal under which the prey fusion is not expressed. The lower panel shows (NLS) (Figure 1a, lanes 1, 3, and 5). In contrast, no self- growth on galactose-raffinose medium, which induces expression of the prey fusion protein. LexA: LexA DNA binding domain; AD: interaction was detected for either Mago-nashi or Y14 activating domain. (b) Y14 is detected in both bound and unbound (Figure 1a, lanes 2, 4, and 6). fractions in a coimmunoprecipitation assay using anti-Myc antibody and ovarian extract of myc-mago flies (lanes 2 and 3). Y14 is absent To address the physiological relevance of the Y14:Mago- from the bound fraction when preimmune serum and anti-Tubulin nashi interaction, we raised an antibody against Y14 and antibody are used as negative controls (lanes 4–7). No Y14 is pulled down using the anti-Myc antibody when wild-type ovarian extract is performed coimmunoprecipitations from ovarian extracts used (lane 9 and 10). In addition, Mago-nashi is of myc-mago transgenic flies, which produce a Myc-mago coimmunoprecipitated with Y14 when wild-type extract is treated with protein in a wild-type background [22]. As shown in Fig- anti-Y14 antibody (lanes 12 and 13), whereas Mago-nashi is absent ure 1b, we were able to coimmunoprecipitate Y14 with from the bound fraction when preimmune serum is used (lanes 14 and 15). In lanes 1, 8, and 11, untreated extract corresponding to the Myc-mago protein when using the anti-Myc antibody the amount used for each immunoprecipitation was loaded. In lanes (lanes 2 and 3), but not when using a preimmune serum 2, 4, 6, and 9, half of the unbound fraction was loaded. U: unbound (lanes 4 and 5) or an unrelated antibody (lanes 6 and 7). fraction; B: bound fraction. The entire fractions were loaded in all other In addition, no Y14 was precipitated from ovarian extract lanes. All lanes are taken from an image of the same gel. of wild-type flies, which lack the myc-mago transgene, showing that the precipitation of Y14 is specifically mediated by Myc-mago protein (lanes 9 and 10). Conversely, Mago-nashi coimmunoprecipitated with Y14 when wild- 1668 Current Biology Vol 11 No 21

Figure 2

Y14 is predominantly nuclear and colocalizes with Mago-nashi at the posterior pole. (b,e) Y14 accumulates to high levels in the nucleus of both the germline and the follicle cells and shows a cytoplasmic localization in the oocyte. (a,b,c) At stage 6, Y14 is enriched in the posterior half of the oocyte to the same extent as Mago-nashi. (d,e,f) At stage 9, Y14 accumulates at the posterior pole of the oocyte, where it colocalizes with Mago-nashi. ,(d؅,e؅,f؅) A 2.5-fold magnification of (d), (e) and (f). The upper panels show GFP-Mago distribution as revealed by GFP, the intermediate panels show Y14 distribution revealed by antibody staining, and the lower panels show the merge of GFP-Mago and Y14 signals. Fc: follicle cells; oo: oocyte; nc: nurse cells. Anterior is toward the left, posterior is toward the right, and dorsal is toward the top.

type ovarian extracts were treated with anti-Y14 antibody ing using affinity-purified anti-Y14 antibody. This anti- (lanes 12 and 13), but not when a preimmune serum was body recognizes Y14 protein specifically, both on Western used (lanes 14 and 15), showing that the in vivo interaction blots (data not shown) and in situ (see below). The in situ is detected in both orientations. This in vivo pull-down localization study highlights the predominantly nuclear assay demonstrates that the two-hybrid interaction corre- localization of Drosophila Y14 (Figure 2b,e), as is the case sponds to a physiologically relevant complex. Moreover, for the human protein [26]. This distribution correlates since all the unbound fraction was loaded in lanes 12 and with that of Mago-nashi (Figure 2a,c,d,f) and is consistent 14 and the amount of Mago-nashi protein, as detected on with the detected protein interaction. Interestingly, Y14 the blot, is slightly greater in lane 13 than in lane 12, we also shows a cytoplasmic distribution. During the early can infer that more than 50% of Mago-nashi coimmuno- stages of oogenesis, Y14 is enriched in the posterior half precipitates with Y14. This indicates that the Y14:Mago- of the oocyte, like Mago-nashi (Figure 2a–c). After stage nashi interaction is remarkably robust and that a large 7, Y14 localizes to the posterior of the oocyte (Figure proportion of Y14 is associated with Mago-nashi in the 2e,eЈ). Thus, Y14 is asymmetrically distributed in the ovary. Similar results were obtained when using ovarian oocyte, where it colocalizes with Mago-nashi at the poste- extracts from myc-mago1 transgenic flies (data not shown). rior pole (Figure 2d,dЈ,fЈ). This suggests that Y14 might This line expresses the Myc-mago1 protein, in which gly- be part of the oskar mRNA localization complex. cine 19 is replaced by an arginine (G19R), a mutation encoded by the mago1 allele [22]. This reveals that the G19R mutation does not affect the ability of Mago-nashi To test whether Y14 shuttles with the oskar mRNA poste- to bind Y14, indicating that the phenotypes observed in rior transport machinery, we examined the localization of mago1/Df are probably not the result of a failure of Mago- Y14 in a genetic context in which oskar mRNA and its nashi to interact with Y14. partner, the double-stranded RNA binding protein, Staufen [33], are mislocalized. To do so, we employed a par-1 Y14 colocalizes with Mago-nashi and Staufen mutant combination (9A/w3), in which oskar mRNA is at the posterior pole of the oocyte targeted to an ectopic site, due to misorientation of the To investigate the subcellular localization of Y14 in Dro- microtubule network [34, 35]. In this genetic background, sophila egg chambers, we performed in situ immunostain- as in the wild-type, Y14 colocalizes with Staufen (Figure Research Paper Drosophila Y14 required for oskar RNA localization Hachet and Ephrussi 1669

Figure 3 lethal, to test whether y14 expression is affected by the EP insertion, we generated homozygous clones in mosaic tissue, making use of the FLP/FRT system [37]. Homozy- gous mutant clones were marked by the absence of a GFP signal, whereas heterozygous and wild-type cells were marked by a GFP signal of proportional intensity. This clonal analysis revealed that the level of Y14 expression is reduced in heterozygous compared to wild-type cells (Figure 4b). Importantly, no Y14 protein was detected in EP(2)0567 homozygous cells, either in somatic or germline clones (Figure 4c,d). This shows that EP(2)0567 strongly reduces or even completely abolishes the expression of y14 and thus constitutes a strong y14 allele. This is confirmed by the fact that the excision of the EP element restores Y14 expression (Figure 4e).

Y14 has a specific function in oskar mRNA posterior localization Because Y14 is likely to belong to the oskar mRNA localization complex, we investigated whether Y14 has a role in oskar mRNA localization. We analyzed by in situ hybridization the effect of the y14EP(2)0567 allele on the distribution of three localized mRNAs, oskar, gurken, and bicoid. For this purpose, we generated homozygous y14EP(2)0567 germline clones by using the FLP/FRT ovoD system, which allows only homozygous mutant germline clones to proceed through oogenesis [37]. Interestingly, the anal- Y14 colocalizes with Staufen at the plus ends of the microtubules. ysis of oskar mRNA distribution revealed that the early (a,b) In wild-type oocytes (stage 9), Y14 colocalizes with Staufen transport of oskar mRNA is not affected by the y14EP(2)0567 9A/W3 at the posterior pole. (c,d) In par-1 mutant oocytes (stage 9), Y14 mutation. From stages 3 to 5, oskar mRNA accumulates colocalizes with Staufen at an ectopic site, where the microtubule plus ends are focused [34]. The left panels show Y14 localization in the oocyte (Figure 5a,b), showing that its transport from revealed by the Y14 antibody, and the right panels show Staufen the nurse cells to the oocyte does not depend on Y14. At antibody staining. An arrowhead indicates Y14 localization in the stage 7, when the microtubule network is reoriented, oocyte. while oskar mRNA is transported toward the posterior in the wild-type, in y14EP(2)0567 egg chambers, oskar mRNA fails to localize at the posterior (Figure 5c,d). This defect 3), indicating that Y14 travels with the oskar mRNA local- in oskar mRNA posterior localization is persistent, as oskar ization complex. This supports the idea that Y14 may be mRNA never accumulates at the posterior pole and even- part of this complex. tually diffuses throughout the oocyte during the ooplasmic streaming that occurs at stage 10b (Figure 5e–h). In con- EP(2)0567 is an allele of y14 trast, both gurken and bicoid mRNAs reach their final desti- To analyze the function of Y14 in vivo, we decided to nation in y14 mutant egg chambers, even if the amount perform a genetic analysis of the y14 gene. Through data- of localized mRNA seems reduced (Figure 5i–l). This base searches, we identified a transposable element (EP shows that their localization competence per se is not element) inserted 70base pairs upstream of the y14 start altered in the absence of Y14. As Staufen colocalizes with codon in the EP(2)0567 line (Figure 4a). This line is oskar mRNA at the oocyte posterior [33], we next analyzed homozygous lethal, and lethality is observed when the EP the effect of y14EP(2)0567 on Staufen protein distribution. As element is placed over three independent chromosomal expected, Staufen accumulates properly in the oocyte deletions covering the y14 locus [Df(2R)Np3, Df(2R)Np4 during the early stages, but is not observed at the posterior and Df(2R)Np5]. Hence, the lethality is closely associated pole after stage 9, in contrast to the wild-type (Figure with the EP element insertion. In addition, the lethality 6a,b). Instead, Staufen is detected throughout the oocyte can be rescued by mobilizing the transposable element. and at the anterior pole, confirming the oskar mRNA This indicates that the EP(2)0567 insertion affects the mislocalization phenotype. Staufen localization is rescued expression of an essential gene. Transposable element by excision of the P element (Figure 6c). In addition, no insertions are known to interfere with the expression of Oskar protein was detected at the posterior of y14EP(2)0567 downstream coding sequences [36]. As this mutation is egg chambers (Figure 6d,e), consistent with the defect 1670 Current Biology Vol 11 No 21

Figure 4

The EP(2)0567 insertion severely reduces or abolishes y14 expression. (a) EP(2)0567 is inserted 70 bp upstream of the start codon of the predicted y14 gene (CG8781). The y14 coding sequence is 498 bp long and is organized in two exons. (b) y14 expression levels correlate with wild-type gene dosage. In wild-type clones (high GFP signal and white arrowhead), the Y14 signal is stronger than in heterozygous cells (intermediate GFP signal and red arrowhead). (c) No Y14 protein is detected in homozygous follicle cell clones (no GFP signal and black arrowhead) as compared to heterozygous follicle cells (GFP signal and red arrowhead). (d) No Y14 protein is detected in homozygous germline clones (no GFP signal and black arrowhead) as compared to heterozygous germline (GFP signal and red arrowhead). (b,c,d) The left panels show the GFP signal that marks the wild-type chromosomes. The intermediate panel shows Y14 antibody staining. The right panel shows the merge. This analysis also demonstrates the specificity of the affinity- purified anti-Y14 antibody. (e) When the EP(2)0567 element is excised, Y14 staining is fully recovered.

in oskar mRNA localization, as oskar mRNA translation The effect of y14EP(2)0567 on oskar mRNA localization is only activated at the posterior pole [38]. Our in situ is independent of the cytoskeleton localization study reveals that Y14 is specifically required To show that the oskar mRNA localization defect ob- for oskar mRNA posterior localization, suggesting that served in the y14 mutant is direct and not due to cytoskele- Y14, like Mago-nashi, may have a direct function in the tal defects, we analyzed the organization of the cytoskele- posterior transport of oskar mRNA. ton in y14 mutant egg chambers. As shown in Figure 7a, the actin cytoskeleton appears unaffected by the y14EP(2)0567 mutation; normal ring canals and the typical cortical actin During later stages of oogenesis, the nurse cells degener- layer are observed. To assess the organization of the mi- ate, expelling their entire cytoplasm into the oocyte in a crotubules, we performed an anti-Tubulin, anti-Staufen process called “dumping”. The y14 mutation leads to a double staining. As shown in Figure 7b–d, both wild-type dumpless phenotype; laid eggs are smaller than wild-type and y14 mutant oocytes show the normal anterolateral and are unfertilized (data not shown). This prevented us accumulation of microtubules; whereas, in par-1 mutants, from analyzing the effect of the mutation on embryonic microtubules emanate from the entire cortex. This shows patterning. that the orientation of the microtubule network is correct Research Paper Drosophila Y14 required for oskar RNA localization Hachet and Ephrussi 1671

Figure 5 y14 has a specific function in oskar mRNA posterior localization. (a,b) During stages 3–5 of oogenesis, oskar mRNA (in green) is correctly enriched in oocytes of y14EP(2)0567 egg chambers. (c,d) At stage 6, oskar mRNA is enriched in the posterior half of the oocyte in wild-type and y14EP(2)0567 egg chambers, but fails to be actively transported to the posterior pole in y14EP(2)0567 stage 7 egg chambers as compared to wild-type. (e,f) At stage 9, oskar mRNA is dispersed throughout y14EP(2)0567 oocytes, whereas it accumulates at the posterior in the wild-type. (g,h) At stage 10b, oskar mRNA is randomly distributed in y14EP(2)0567 oocytes, in contrast to its posterior localization in the wild-type. (i,j) gurken mRNA localizes in the dorsal-anterior corner of y14EP(2)0567 oocytes at stage 9, as in the wild-type. (k,l) A surface view of bicoid mRNA that correctly accumulates at the anterior cortex of y14EP(2)0567 oocytes at stage 10a, forming an anterior ring as in the wild-type. The RNA signal in the oocyte is shown in green, and autofluorescence of the samples allows visualization of egg chamber morphology (red). The left panels show wild- type egg chambers, and the right panels show y14EP(2)0567 egg chambers. The scale bar represents 50 ␮m.

in the y14 mutant and that the effect of the y14EP(2)0567 toplasmic shuttling competence of Y14 demonstrated in mutation on Staufen and oskar mRNA localization cannot heterokaryons [26] and constitutes the first description of be attributed to a defect in cytoskeletal organization. This Y14 cytoplasmic localization under physiological condi- indicates that Y14 has a direct role in oskar mRNA poste- tions. Y14 posterior localization was not detected after rior transport. stage 9 by our methods, suggesting that its posterior localization may be transient, as described for Mago-nashi and Barentsz [22, 23]. The colocalization of the Y14:Mago- Discussion nashi complex with Staufen and oskar mRNA, both at the In this study, we have shown that the Y14:MAGOH inter- posterior pole and at an ectopic site in the par-1 mutant, action is conserved in Drosophila in vivo, indicating that is entirely consistent with Y14 being part of the oskar the Y14:Mago complex is likely to be conserved through- mRNA posterior transport machinery. out evolution. Immunochemistry reveals that Y14 is predominantly nuclear and shuttles out of the nucleus to localize in the posterior cytoplasm of the oocyte. This Examination of the effect of a strong allele of y14 on observation is consistent with the previously reported cy- mRNA localization revealed that Y14 protein is specifi- 1672 Current Biology Vol 11 No 21

Figure 6 upon splicing of oskar mRNA constitutes a critical step in the assembly of the posterior transport machinery. If this were the case, this mechanism would constitute a new function for the splicing event, the deposit of a landmark addressing mRNAs to their cytoplasmic destinations.

Materials and methods Mutant alleles and transgenic lines The y14 mutant EP(2)0567 (obtained from Exelixis), mago1 and mago2 mutant alleles, and par-19A and par-1w3 mutant alleles were used in this analysis. Df(2R)Np3, Df(2R)Np4, and Df(2R)Np5 (obtained from the Bloomington stock center) were used as deficiencies for y14. For the clonal analysis, we used the following stocks: FRT42BG13 EP(2)0567, hsFLP; FRT42BG13 NLS-GFP, and hsFLP; FRT42BG13 ovoD. myc-mago, myc-mago1, and GFP-mago transgenic lines were used in this study. For EP element excision, flies bearing a transgene encoding the ⌬2–3 transposase were crossed to the EP(2)0567 line.

y14 cloning and two-hybrid assay The y14 coding sequence was cloned into the pCRII-Topo vector by the polymerase chain reaction (PCR) from a Drosophila cDNA library, using ttatctgcgacgcttttcgg and atggccgatgtgttggac as primers and a TOPO cloning kit (Stratagene). The clone was verified by sequencing, Staufen protein is mislocalized, and Oskar protein is not produced in and its orientation was found to be T7 y14 CDS SP6. The EcoRI TOPO the y14 mutant. (a,b) Staufen protein is mislocalized throughout the y14 fragment was subcloned into the EcoRI site of pLexA and pB42AD oocyte cytoplasm in (a) y14 mutant egg chambers compared to (b) vectors (Clontech). mago-nashi was subcloned by PCR into the pCRII- wild-type. (c) Staufen posterior localization is rescued in EP(2)0567 Topo vector as described above, using tcgcaattgatgtccacggaggac and excision lines. (d,e) Oskar protein is not detected at the posterior of tttctcgaggaggtggtggctacggc primers, inserting MfeI and XhoI restriction the oocyte in (e) y14 mutant egg chambers compared to (d) wild- sites at the 5Ј and 3Ј ends, respectively. The MfeI XhoI TOPO mago- type. Oskar protein is shown in red. Autofluorescence of the samples nashi fragment was then cloned between EcoRI and SalI of pLexA-NLS allows visualization of egg chamber morphology (green). (adapted from pLexA clontech) and between EcoRI and XhoI of pB42AD and pLexA. Bait constructs were transformed into yeast strain EGY48, and prey vectors were transformed into RSY47-3, using the lithium acetate transformation procedure. Interaction tests were performed by cally involved in oskar mRNA posterior transport, while mating the strains and growing them on the appropriate media [32]. gurken and bicoid mRNA localization is not affected. oskar Antibody generation mRNA was found distributed throughout the oocyte, in- The full-length coding sequence of y14 and mago-nashi were cloned cluding at the anterior, as we also observe by fluorescent into pGEX-4T-1 and expressed as COOH-terminal fusions to glutathione RNA detection and confocal microscopy of other mutants S-transferase (GST) in E. coli strain BL21. GST-Y14 and GST-Mago affecting oskar mRNA localization (A. M. Michon and protein were purified by affinity chromatography using glutathione sepha- gs1 D3 rose 4B (Amersham Pharmacia Biotech); GST-Y14 was injected into A.E., unpublished data), such as Tm2 and stau [33, 39] rats, and GST-Mago was injected into rabbits with RIBI adjuvant (serum and mago1/mago2 (data not shown). Consistent with this, production by LAR EMBL service). Y14 antiserum was affinity purified Staufen protein is to some extent also dispersed in from nitrocellulose strips bearing the GST-Y14 protein as previously y14EP(2)0567 oocytes. However, the apparent accumulation described [40]. of Staufen protein at the anterior of these oocytes may Coimmunoprecipitation assay suggest that the mutation partially uncouples Staufen Ovarian extracts were obtained by hand dissecting 2- to 3-day-old well- from oskar mRNA. The absence of Oskar protein at the fed females, using 20 ovaries per sample. Ovaries were washed in 50% posterior of y14 mutant oocytes confirms the oskar mRNA extraction buffer (20 mM HEPES [pH 7.5], 400 mM NaCl, 0.1 mM localization defect. EGTA, 1 mM PMSF, 10 ␮g/ml leupeptin and pepstatin) in PBS. Typically, 200 ovaries were homogenized with 50 ␮l extraction buffer. The homoge- nate was centrifuged at 10,000 ϫ g for 10 min at 4ЊC, and the intermedi- We have shown that the cytoskeletal organization of the ate phase was saved. A total of 5 ␮l fresh extract was incubated for 2 egg chamber is not affected in y14EP(2)0567 mutants under hr with 9 ␮l2ϫ immunoprecipitation buffer (40 mM HEPES [pH 7.5], ␮ our experimental conditions. This indicates that, like 300 mM NaCl, 5 mM MgCl2, 500 mM sucrose, 2 mM PMSF, 20 g/ ␮ Staufen, Mago-nashi, and Barentsz, Y14 has a direct role ml leupeptin and pepstatin) and either 4 l anti-Myc 9E10 antibody, 4 ␮l anti-Tubulin antibody (Sigma), 2 ␮l mouse preimmune crude serum in oskar mRNA localization. Moreover, the fact that oskar with 2 ␮l PBS, 4 ␮l anti-Y14 crude serum, or 4 ␮l rat preimmune crude EP(2)0567 mRNA is dispersed in y14 oocytes, rather than ad- serum. A total of 50 ␮l 50% washing buffer (20 mM HEPES [pH 7.5], dressed to an ectopic site, as occurs in par-1 and gurken 400 mM NaCl, 0.01% BSA, 0.5% Triton X-100, 0.5% NP-40, 1 mM mutants, also argues for a direct role of y14 in oskar mRNA PMSF, 10 ␮g/ml leupeptin and pepstatin) protein A agarose slurry (Amer- posterior transport. As human Y14 has been described as sham Pharmacia Biotech) was added and incubated for 1 hr. Beads were pelleted to collect the supernatant (unbound fraction) and were imprinting mRNAs upon splicing [26], it is tempting to washed five times in washing buffer. The bound fraction was recovered imagine that recruitment of the Y14:Mago-nashi complex by boiling the beads for 5 min at 95ЊCin2ϫ loading buffer. Samples Research Paper Drosophila Y14 required for oskar RNA localization Hachet and Ephrussi 1673

Figure 7

Cytoskeletal organization is not affected by the y14EP(2)0567 mutation. (a,a؅) The actin cytoskeleton of a y14 mutant egg chamber (late stage 9) appears to be unaffected, with normal ring canals (black arrowheads) and cortical actin (white arrowhead). A 4-fold magnification of cortical actin is shown in (aЈ). fc: follicle cells; oo: oocyte. (b,c) At early stage 9, although microtubule orientation is normal in the (b) y14 mutant oocyte, Staufen protein does not localize at the posterior pole, in contrast to (c) wild-type. (d) The misorganization of the microtubule network in par-19A/W3 oocytes and the consequent mislocalization of Staufen are shown for comparison (early stage 9). The microtubule cytoskeleton is revealed by anti-Tubulin staining (in red), and anti-Staufen antibody staining is shown in green.

were loaded on a 12% SDS-PAGE gel. Y14 and Mago-nashi proteins critical comments on the manuscript. We thank Robert Boswell for the mago- were detected by Western blot (1/2500 dilution of the anti-Y14 and nashi lines. O.H. is supported by a predoctoral “Allocation de recherche” anti-Mago crude serum were used). from the French government.

Immunostaining and in situ hybridization References The destinies and destinations of RNAs. Ovaries from 2- to 3-day-old females were dissected in PBS, fixed for 1. Hazelrigg T: Cell 1998, 95:451-460. 20 min in 4% paraformaldehyde in PBS, washed twice for 20 min in 2. Long RM, Singer RH, Meng X, Gonzalez I, Nasmyth K, Jansen RP: PBT (0.1% Triton X-100), incubated for 1.5 hr in PBS 1% Triton X-100, Mating type switching in yeast controlled by asymmetric and blocked for 1 hr in PBS 0.3% Triton X-100, 0.5% BSA (BSA localization of ASH1 mRNA. Science 1997, 277:383-387. blocking buffer). Ovaries were then incubated overnight with the primary 3. Takizawa PA, Sil A, Swedlow JR, Herskowitz I, Vale RD: Actin- antibodies in BSA blocking buffer. Preadsorbed anti-Staufen, pread- dependent localization of an RNA encoding a cell-fate sorbed anti-Oskar, and affinity-purified anti-Y14 were used at 1/250, 1/ determinant in yeast. Nature 1997, 389:90-93. 3000, and 1/50 dilutions, respectively. Samples were washed twice for 4. Ainger K, Avossa D, Morgan F, Hill SJ, Barry C, Barbarese E, et al.: 20 min in BSA blocking buffer, incubated for 1 hr in PBT 10% normal Transport and localization of exogenous myelin basic goat serum, incubated for 1.5 hr with fluorescently coupled secondary protein mRNA microinjected into oligodendrocytes. J Cell Biol 1993, 123:431-441. antibodies (1/500 in PBT), washed twice for 20 min in PBT, and mounted 5. St Johnston D, Nusslein-Volhard C: The origin of pattern and in 80% glycerol, 2% N-Propyl-Gallate. The actin cytoskeleton was decor- polarity in the Drosophila embryo. Cell 1992, 68:201-219. ated by incubating ovaries in 1/250 rhodamine-phalloidin, in parallel with 6. Neuman-Silberberg FS, Schu¨pbach T: The Drosophila the secondary antibody incubation, when required. For Tubulin staining, dorsoventral patterning gene gurken produces a dorsally ovaries were fixed in methanol at Ϫ20ЊC for 10 min, rehydrated progres- localized RNA and encodes a TGF␣-like protein. Cell 1993, sively in PBT, and then processed as described above. Anti-Tubulin 75:165-174. (Sigma) was used at 1/400 dilution. For in situ hybridization, ovaries 7. Berleth T, Burri M, Thoma G, Bopp D, Richstein S, Frigerio G, et al.: were fixed in 4% paraformaldehyde and processed according to [41] The role of localization of bicoid RNA in organizing the using oskar, gurken,orbicoid digoxigenin-labeled probes, and the signal anterior pattern of the Drosophila embryo. EMBO J 1988, 7: was amplified using FITC-tyramide (NEN Life Science). Microscopy was 1749-1756. 8. Ephrussi A, Dickinson LK, Lehmann R: Oskar organizes the germ carried out using a Leica TCS SP confocal microscope. plasm and directs localization of the posterior determinant nanos. Cell 1991, 66:37-50. Acknowledgements 9. Saunders C, Cohen RS: The role of oocyte transcription, the 5؅ We are grateful to Juan Mata for exciting discussions and suggestions during UTR, translation repression and derepression in Drosophila the early stages of this work, to Thomas Vaccari and all the other members gurken mRNA and protein localization. Mol Cell 1999, 3:43- of the lab for technical advice and discussions, and to Virginie Hachet for 54. 1674 Current Biology Vol 11 No 21

10. Bassell G, Singer RH: mRNA and cytoskeletal filaments. Curr 34. Shulman JM, Benton R, St Johnston D: The Drosophila homolog Opin Cell Biol 1997, 9:109-115. of C. elegans PAR-1 organizes the oocyte cytoskeleton and 11. Bertrand E, Chartrand P, Schaefer M, Shenoy SM, Singer RH, Long directs oskar mRNA localization to the posterior pole. Cell RM: Localization of ASH1 mRNA particles in living yeast. 2000, 101:377-388. Mol Cell 1998, 2:437-445. 35. Tomancak P, Piano F, Riechmann V, Gunsalus KC, Kemphues KJ, 12. Beach DL, Salmon ED, Bloom K: Localization and anchoring of Ephrussi A: A Drosophila melanogaster homologue of mRNA in budding yeast. Curr Biol 1999, 9:569-578. Caenorhabditis elegans par-1 acts at an early step in 13. Clark I, Giniger E, Ruohola-Baker H, Jan LY, Jan YN: Transient embryonic-axis formation. Nat Cell Biol 2000, 2:458-460. posterior localization of a kinesin fusion protein reflects 36. Cooley L, Kelley R, Spradling A: Insertional mutagenesis of the anteroposterior polarity of the Drosophila oocyte. Curr Biol Drosophila genome with single P elements. Science 1988, 1994, 4:289-300. 239:1121-1128. 14. Carson JH, Worboys K, Ainger K, Barbarese E: Translocation of 37. Chou TB, Perrimon N: Use of a yeast site-specific recombinase myelin basic protein mRNA in oligodendrocytes requires to produce female germline chimeras in Drosophila. microtubules and kinesin. Cell Motil Cytoskeleton 1997, 38:318- Genetics 1992, 131:643-653. 328. 38. Markussen FH, Michon AM, Breitwieser W, Ephrussi A: 15. Theurkauf WE, Smiley S, Wong ML, Alberts BM: Reorganization Translational control of oskar generates short OSK, the of the cytoskeleton during Drosophila oogenesis: isoform that induces pole plasma assembly. Development implications for axis specification and intercellular transport. 1995, 121:3723-3732. Development 1992, 115:923-936. 39. Erdelyi M, Michon AM, Guichet A, Glotzer JB, Ephrussi A: 16. Brendza RP, Serbus LR, Duffy JB, Saxton WM: A function for Requirement for Drosophila cytoplasmic tropomyosin in kinesin I in the posterior transport of oskar mRNA and oskar mRNA localization. Nature 1995, 377:524-527. Staufen protein. Science 2000, 289:2120-2122. 40. Pringle JR, Adams AE, Drubin DG, Haarer BK: 17. Cote CA, Gautreau D, Denegre JM, Kress TL, Terry NA, Mowry KL: Immunofluorescence methods for yeast. Methods Enzymol A Xenopus protein related to hnRNP I has a role in 1991, 194:565-602. cytoplasmic RNA localization. Mol Cell 1999, 4:431-437. 41. Wilkie G, Davis I: Visualizing mRNA by in situ hybridization 18. Long RM, Gu W, Meng X, Gonsalvez G, Singer RH, Chartrand P: using “high resolution” and sensitive tyramide signal An exclusively nuclear RNA-binding protein affects amplification [Technical Tips]. TTO 1998, 1:T01458. asymmetric localization of ASH1 mRNA and Ash1p in yeast. J Cell Biol 2001, 153:307-318. 19. Hoek KS, Kidd GJ, Carson JH, Smith R: hnRNP A2 selectively binds the cytoplasmic transport sequence of myelin basic protein mRNA. Biochemistry 1998, 37:7021-7029. 20. Newmark PA, Boswell RE: The mago nashi locus encodes an essential product required for germ plasm assembly in Drosophila. Development 1994, 120:1303-1313. 21. Micklem DR, Dasgupta R, Elliott H, Gergely F, Davidson C, Brand A, et al.: The mago nashi gene is required for the polarization of the oocyte and the formation of perpendicular axes in Drosophila. Curr Biol 1997, 7:468-478. 22. Newmark PA, Mohr SE, Gong L, Boswell RE: mago nashi mediates the posterior follicle cell-to-oocyte signal to organize axis formation in Drosophila. Development 1997, 124:3197-3207. 23. van Eeden FJ, Palacios IM, Petronczki M, Weston MJ, St Johnston D: Barentsz is essential for the posterior localization of oskar mRNA and colocalizes with it to the posterior pole. J Cell Biol 2001, 154:511-524. 24. Zhao XF, Nowak NJ, Shows TB, Aplan PD: MAGOH interacts with a novel RNA-binding protein. Genomics 2000, 63:145-148. 25. Mingot JM, Kostka S, Kraft R, Hartmann E, Gorlich D: Importin 13: a novel mediator of nuclear import and export. EMBO J 2001, 20:3685-3694. 26. Kataoka N, Yong J, Kim VN, Velazquez F, Perkinson RA, Wang F, et al.: Pre-mRNA splicing imprints mRNA in the nucleus with a novel RNA-binding protein that persists in the cytoplasm. Mol Cell 2000, 6:673-682. 27. Le Hir H, Izaurralde E, Maquat LE, Moore MJ: The spliceosome deposits multiple proteins 20-24 nucleotides upstream of mRNA exon-exon junctions. EMBO J 2000, 19:6860-6869. 28. Kim VN, Yong J, Kataoka N, Abel L, Diem MD, Dreyfuss G: The Y14 protein communicates to the cytoplasm the position of exon-exon junctions. EMBO J 2001, 20:2062-2068. 29. Le Hir H, Gatfield D, Izaurralde E, Moore MJ: The exon-exon junction complex provides a binding platform for factors involved in mRNA export and nonsense-mediated mRNA decay. EMBO J 2001, 20:4987-4997. 30. Kim VN, Kataoka N, Dreyfuss G: Role of the nonsense-mediated decay factor hUpf3 in the splicing-dependent exon-exon junction complex. Science 2001, 293:1832-1836. 31. Lykke-Andersen J, Shu MD, Steitz JA: Communication of the position of exon-exon junctions to the mRNA surveillance machinery by the protein RNPS1. Science 2001, 293:1836- 1839. 32. Gyuris J, Golemis E, Chertkov H, Brent R: Cdi1, a human G1 and S phase protein phosphatase that associates with Cdk2. Cell 1993, 75:791-803. 33. St Johnston D, Beuchle D, Nusslein-Volhard C: Staufen, a gene required to localize maternal RNAs in the Drosophila egg. Cell 1991, 66:51-63.