ASH1 mRNA Anchoring Requires Reorganization of the Myo4p/She3p/She2p Transport Complex*

DNA oligonucleotide probes (11,34). Cells were subsequently stained with DAPI and mounted on slides with mounting media containing phenylenediamine. Images were captured using a Nikon Eclipse 600 epifluorescence microscope equipped with a 60X N.A. 1.4 objective, interfaced to a Micromax-Interline Transfer CCD Camera (Princeton Instruments, Inc.) and MetaMorph Imaging Software (Universal Imaging Corp.).

The establishment and maintenance of cellular polarity is a salient feature of eukaryotic cells and involves the asymmetric sorting of proteins to distinct compartments or regions of the cell. One mechanism by which proteins can be sorted in a variety of eukaryotic cell types is via mRNA localization, a process by which mRNA is specifically localized to a particular region of a cell (1)(2)(3). A primary step in the process of mRNA localization is the identification of the localization substrate by trans-acting factors, resulting in the formation of a localization mRNP.
The mRNP localization complexes are dynamic structures undergoing molecular reorganization at various stages of the localization pathway (4)(5)(6). RNA localization can result from a number of distinct mechanisms involving direct transport of the mRNA to the site of localization, generalized degradation of the mRNA with localized protection or random diffusion followed by entrapment of the mRNA at the site of localization (1,2). Ultimately, translation of the mRNA at the site of localization leads to the asymmetric distribution of the protein in the cell.
The asymmetric sorting of Ash1p in Saccharomyces cerevisiae is a paradigm for investigating the asymmetric segregation of proteins via mRNA localization. Ash1p is a transcriptional repressor that is asymmetrically sorted to daughter cells resulting in differential gene expression between mother and daughter cells (7)(8)(9). The unequal segregation of Ash1p results from the localization of ASH1 mRNA at the distal tip of budding yeast cells during anaphase of the cell cycle (10)(11)(12). Localization of ASH1 mRNA depends upon four cis-acting localization elements, named E1, E2A, E2B and E3, and each element is sufficient for localizing a heterologous reporter RNA to daughter cells (10,13,14).
Myo4p and She3p are also required for inheritance of cortical ER, but in contrast to the ASH1 mRNA localization pathway, cortical ER sorting is independent of She2p (27).
In this report we present evidence that She3p can simultaneously associate with Myo4p and She2p resulting in the Myo4p/She3p/She2p heterotrimeric complex and analyze the role of this heterotrimeric complex in the ASH1 mRNA localization pathway. In comparison to the Myo4p/She3p complex, we found that the heterotrimeric complex was present at a relatively lower abundance. Closer inspection revealed that the heterotrimeric complex which is sufficient for transport of mRNA to daughter cells is insufficient for anchoring the mRNA at the bud tip.

Yeast Strains, Media and Plasmids
The yeast strains used in this study are listed in Table 1. Yeast cells were either grown in rich media or in defined synthetic media lacking the indicated nutrients (28). Yeast cells were transformed using lithium acetate, and gene deletions were created as described previously, using PCR products generated from plasmid pUG6 (29,30). Silent deletions were subsequently generated using plasmid pSH47 (30). Endogenous yeast genes were tagged with myc 13 or 3HA epitopes using PCR products as described elsewhere (31).
Plasmids used in this study are listed in Table 2 and were constructed using standard molecular techniques (32). Details regarding plasmid construction are available upon request.

Soluble Lysate Preparation and Immunoprecipitation
Exponentially growing yeast cultures corresponding to 3 x 10 8 cells were harvested for each immunoprecipitation, and the cell pellets were resuspended in lysis buffer containing 50 mM HEPES-KOH, pH 7.3, 20 mM potassium acetate, 2 mM EDTA, 0.1% Triton X-100, 5% glycerol, 0.1 mg/ml chymostatin, 2 mg/ml aprotinin, 1 mg/ml pepstatin, 0.5 mg/ml leupeptin and 0.01 mg/ml benzamidine. Glass beads were added to the suspension, and cells were broken by vortexing. The extracts were cleared by centrifugation at 5000 x g for 2 min at 4 0 C. The soluble fraction was recovered and added to antibody bound Protein A beads (Pierce). Anti-myc immunoprecipitations were performed using 7.5 µg of monoclonal 9E10 antibody (Roche).
Anti-HA immunoprecipitations were performed using 10 µg of monoclonal HA.11 antibody (Convance). The soluble lysate was incubated with the Protein A beads for 2 hr at 4 0 C with gentle agitation. The beads were subsequently washed four times with 500 µl of wash buffer (50 mM HEPES-KOH, pH 7.3, 100 mM potassium acetate, 2 mM EDTA, 0.1% Triton X-100 and 5% glycerol). Bound proteins were eluted by boiling the Protein A-antibody complexes in Laemmli buffer. Equivalent amounts of cell extract (Total) and bound proteins (IP) were separated by SDS-PAGE and analyzed by Western-blot. For the salt titration experiment, the above wash buffer containing potassium acetate at concentrations ranging from 100 mM to 800 mM was used. Each experiment was performed at least twice and quantitative values were determined with either AlphaImager 2200 (AlphaInnotech Corp.) or FluroChem 8900 (AlphaInnotech Corp.) software.

Immunoprecipitation/RT-PCR
With few modifications, immunoprecipitation/RT-PCR reactions were performed as previously described, using antibody bound Protein A beads (33). For immunoprecipitation experiments using the myc epitope 7.5 µg of anti-myc 9E10 was used. For immunoprecipitation of endogenous She2p, 20 µl of anti-She2p rabbit antiserum was used. The RT-PCR reactions were performed using the Access/RT-PCR kit (Promega) according to the directions provided by the manufacturer and were performed under conditions to ensure that amplifications were in the linear range of the assay.

RNP Purification
Yeast cells expressing the desired constructs were harvested by centrifugation, and a soluble yeast lysate was prepared as described in the above immunoprecipitation section. The soluble lysate was added to 20 µl of FLAG-Agarose beads (ANTI-FLAG ® MS2 Affinity Gel, Sigma) and incubated for 2 hr at 4 0 C with gentle agitation. The beads were recovered by centrifugation at 500 x g for 2 min at 4 0 C, and the beads were washed four times with 600 µl of wash buffer (50 mM HEPES-KOH, pH 7.3, 100 mM potassium acetate, 2 mM EDTA, 0.1% Triton X-100 and 5% glycerol). The bound proteins were eluted by heating the beads at 65 0 C for 10 min in 100 µl of elution buffer (50mM Tris-HCl, pH 8.0, 100mM NaCl, 10mM EDTA and 1% SDS). The soluble lysate (Total) and eluate (IP) fractions were subsequently separated by SDS-PAGE and analyzed by Western-blot. Myc tagged proteins were detected using the anti-myc 9E10 monoclonal antibody, while the 4XFLAG-MS2 fusion protein was detected using ANTI-FLAG M2 ® Monoclonal Antibody -Peroxidase Conjugate (Sigma).

In situ Hybridization and Fluorescent Microscopy
The in situ hybridizations were performed as described previously using ASH1-Cy3 or lacZ-Cy3 DNA oligonucleotide probes (11,34). Cells were subsequently stained with DAPI and mounted on slides with mounting media containing phenylenediamine. Images were captured using a Nikon Eclipse 600 epifluorescence microscope equipped with a 60X N.A. 1.4 objective, interfaced to a Micromax-Interline Transfer CCD Camera (Princeton Instruments, Inc.) and MetaMorph Imaging Software (Universal Imaging Corp.).

RESULTS
She2p associates with Myo4p dependent on She3p-Previous work demonstrated that She3p is capable of associating in vivo with both Myo4p and She2p (24,25), potentially resulting in two complexes: Myo4p/She3p (complex II) and She3p/She2p (complex III) (Fig. 1A). Based on the presumed ability of She3p to simultaneously interact with both Myo4p and She2p, it was hypothesized that a Myo4p/She3p/She2p heterotrimeric complex (complex IV) could also form in vivo (Fig. 1A) (24)(25)(26). Therefore, we investigated the presence of the heterotrimeric complex by immunoprecipitating Myo4p-myc 13 from a yeast lysate and analyzing the ability of She2p to co-immunoprecipitate with Myo4p-myc 13 (Fig. 1B). She2p was found to specifically coprecipitate with Myo4p dependent on She3p (Fig. 1B compare lanes 2 and 3). These results support the conclusion that a fraction of She2p is a component of a Myo4p/She3p/She2p heterotrimeric complex (complex IV) and are in agreement with results obtained from a genome-wide TAP analysis (35).
In vivo analysis of She3p-containing protein complexes-Given that we could detect an in vivo association between Myo4p and She2p, we sought further insight into the steady state abundance of She3p-containing complexes II, III and IV (Fig. 1A). Following immunoprecipitation of Myo4p-3HA and Western-blotting, we observed at least a 20-fold lower signal for She2pmyc 6 compared to the signal corresponding to She3p-myc 6 (Fig. 1C, lane 4). In this experiment, the signal for She2p-myc 6 following immunoprecipitation was more easily detected upon overexposure of the blot (data not shown). Assuming each Myo4p-3HA containing complex has an equivalent ability of being immunoprecipitated, this result suggests that the Myo4p/She3p complex (complex II) predominates over the heterotrimeric complex (complex IV).
We next determined the relative abundance of the She3p/She2p complex (complex III) in comparison to the heterotrimeric complex (complex IV). Two isogenic yeast strains were used for this analysis, one strain expressing endogenous levels of Myo4p-myc 13 and the other strain expressing endogenous levels of She3p-myc 13 . Immunoprecipitations were performed, and the abundance of co-precipitating She2p was determined by Western-blot (Fig. 1D). We observed 2-fold more She2p co-precipitating with She3p-myc 13 compared to Myo4p-myc 13 . Since She2p can associate with She3p in vivo in two potential complexes: the She2p/She3p complex (complex III) and the heterotrimeric complex (complex IV), the level of She2p that coprecipitates with She3p under these conditions represents the sum of complexes III and IV.
Consequently, our results suggest that complex III and complex IV are present at nearly equivalent levels.
It is also possible that the various complexes have different stabilities under the experimental conditions. The salt-stability of the various complexes was therefore examined.
All complexes were found to be equally stable at 100 mM KOAc (Fig. 2). However, the She3p/She2p complex and the heterotrimeric complex were not stable at salt concentrations higher than 200 mM KOAc ( Fig. 2B and C). In contrast the Myo4p/She3p complex was stable up to 800 mM KOAc ( Fig. 2A). Consequently, we conclude that the apparent lower abundance of She2p-containing complexes is not due to differences in the stabilities of the complexes under the ionic conditions used to prepare the cell lysates. associate with ASH1 mRNA via heterotrimeric complex formation (22,26). A prediction made by the current model is that She2p associates with ASH1 mRNA prior to associating with She3p and Myo4p. Thereby more ASH1 mRNA could be associated with She2p than with She3p or Myo4p. Examining the level of ASH1 mRNA associated with She2p, She3p and Myo4p tested this prediction. She2p-myc 13 , She3p-myc 13 and Myo4p-myc 13 were immunoprecipitated from yeast lysates, and co-precipitating ASH1 RNA was analyzed by RT-PCR (Fig. 3A). A similar amount of ASH1 mRNA was found to co-precipitate with She2p-myc 13 , She3p-myc 13 and Myo4p-myc 13 under these conditions, consistent with the hypothesis that very little, if any, ASH1 mRNA associated with any of these three proteins exists free of the heterotrimeric complex.
To more quantitatively assess the association of Myo4p, She3p and She2p with ASH1 cis-acting localization elements, mRNPs containing ASH1 cis-acting elements were purified, and the levels of She2p-myc 13 , She3p-myc 13 and Myo4p-myc 13 that co-purified with the mRNP were determined by Western-blot. To purify mRNPs, a lacZ mRNA construct containing MS2 stem loops as well as the ASH1 E3 cis-acting localization element was constructed (Fig. 3B).
For a negative control, lacZ-MS2 mRNA lacking the E3 cis-acting localization element was used. Expression of the fusion mRNA constructs in a strain that also expresses a 4x FLAG-MS2 fusion protein enables the purification of the MS2-containing mRNPs via the MS2 protein/MS2 RNA interaction (36). Following enrichment of the MS2-RNPs, the presence of Myo4p-myc 13 , She3p-myc 13 and She2p-myc 13 was determined by Western-blot analysis (Fig. 3C). When normalized to the amount of immunoprecipitating 4x FLAG-MS2 we observed equivalent levels of She2p and She3p co-purifying with the mRNP. In contrast, we observed a slight, 1.8-fold, enrichment for co-purifying Myo4p when compared to She2p (Fig. 3C). Using this purification scheme Myo4p, She3p and She2p are only co-precipitated from strains expressing the lacZ-MS2-E3 fusion mRNA construct (Fig. 3C). These observations indicate that the in vivo association and purification of these proteins are specific for ASH1 cis-acting localization elements. The results from the mRNP purification experiment are consistent with the IP/RT-PCR analysis, suggesting that the majority of ASH1 mRNA associated with Myo4p, She3p or She2p is present in the heterotrimeric complex. Furthermore, from our observation that approximately two fold more Myo4p is associated with ASH1 mRNA compared to She3p and She2p leads us to hypothesize that the stoichiometry of the Myo4p/She3p/She2p complex may be 2:1:1, assuming that all She2p that is associated with ASH1 mRNA is also present in the heterotrimeric complex.

Association of mRNA with the heterotrimeric complex is not sufficient for mRNA localization-
The observations that the heterotrimeric complex is present in yeast cells and that Myo4p, She3p  Consequently, these results imply that when ASH1-MS2 mRNA is associated with the She2p portion of the She2p-MS2 fusion protein the mRNA is capable of being correctly localized, but when this mRNA is associated with the MS2 portion of the fusion protein, the localization substrate is not able to be localized. to localize to daughter cells could be due to the inability of the She2p-MS2 fusion protein to transport mRNA to daughter cells, and/or the inability of the fusion protein to anchor mRNA at the bud tip. We reasoned that if we could pinpoint where in the localization pathway that the MS2 mRNA-binding activity could not substitute for She2p mRNA-binding activity, we might gain further insight into the mechanism of mRNA localization to daughter cells.
In an effort to distinguish between an mRNA localization transport defect or an anchoring defect related to MS2 mRNA-binding activity, we sought to separate transport from anchoring.
By investigating the distribution of mRNA localization substrates throughout the entire cell cycle, the transport and anchoring components could be individually investigated. Transport and anchoring of mRNA dependent on MS2 mRNA binding activity throughout the cell cycle was studied by fusing wild-type She2p to MS2 and individually monitoring the intracellular distribution of galactose-inducible lacZ mRNA, lacZ-ASH1 mRNA or lacZ-MS2 mRNA (Fig.   5). Since She2p-MS2 is unable to associate with lacZ mRNA, as expected this mRNA is uniformly distributed between mother and daughter cells at all stages of the cell cycle ( examining the distribution of lacZ-MS2 mRNA, we observed that this mRNA was correctly sorted to daughter cells prior to anaphase (Fig. 5/top panel). However, at anaphase when ASH1 mRNA is normally localized, lacZ-MS2 mRNA is uniformly distributed between mother and daughter cells (Fig. 5/bottom panel). Consequently, these results suggest that prior to anaphase She2p-MS2 is able to participate in the transport of lacZ-MS2 mRNA to daughter cells, but is not able to properly anchor lacZ-MS2 mRNA in anaphase cells. In support of the heterotrimeric complex model, we observed that Myo4p and She2p associate dependent on She3p (Fig. 1). Further examination of She3p-containing complexes revealed that the Myo4p/She3p complex is more abundant than either the She3p/She2p complex or the heterotrimeric complex (Fig. 1). One possible explanation for the greater abundance of the Myo4p/She3p complex in comparison to the She2p-containing complexes is that She2p may be expressed at a much lower level than either She3p or Myo4p. However, She2p is approximately 2-fold more abundant than Myo4p and 4-fold more abundant than She3p (38). Therefore, the predominance of the Myo4p/She3p complex over the other two complexes is apparently independent of concentration differences of these three proteins and is apparently related to intrinsic differences in the ability of She3p to associate with Myo4p versus She2p. Consistent with this assertion is our finding that the She3p/She2p association is more sensitive to ionic strength compared to the Myo4p/She3p interaction (Fig. 2). Based on these results we hypothesize that She3p may have differential affinities for Myo4p and She2p, thus resulting in the preferential formation of the Myo4p/She3p complex. Alternatively, formation of the three protein complexes could be subject to post-translational regulation. If formation of the heterotrimeric complex or the She3p/She2p complex is subject to regulation, it is apparently unrelated to the ability of mRNA to associate with She2p. We found no difference in the amount of She2p associated with Myo4p or She3p when She2p is unable to associate with mRNA localization substrates or when mRNA localization substrates are over-expressed (data not shown).
An implicit prediction of the heterotrimeric model is that RNA artificially tethered to She2p should localize to daughter cells in an analogous fashion to ASH1 mRNA. Even though She2p-R63K-MS2 is capable of associating with Myo4p and ASH1-MS2 mRNA, ASH1-MS2 mRNA is delocalized in anaphase cells (Fig. 4). A trivial explanation for the inability of ASH1-MS2 mRNA to localize to daughter cells dependent on She2p-R63K-MS2 is that fusing MS2 to She2p could interfere with a She2p function in ASH1 mRNA localization. However, wild-type She2p-MS2 is fully competent for localizing ASH1 mRNA (data not shown), indicating that the MS2 RNA-binding domain does not impede the ability of wild-type She2p to localize ASH1 mRNA.
The precise defect in ASH1 mRNA localization associated with tethering mRNA to She2p through MS2 appears to be related to anchoring. A completely artificial mRNA localization substrate, lacZ-MS2 mRNA, localizes to small and medium sized buds dependent on She2p-MS2 (Fig. 5). Consequently, She2p-MS2 is able to participate in the delivery of RNA localization substrates during this period of the cell cycle when the actin cytoskeleton is highly polarized for efficient delivery of various cargos to the bud. In contrast, She2p-MS2 is unable to localize lacZ-MS2 mRNA during anaphase when ASH1 mRNA is normally localized. We hypothesize that upon reaching the bud tip She2p-MS2/lacZ-MS2 mRNA is unable to mimic a She2p/ASH1 mRNA molecular remodeling event that results in the anchoring of the mRNA localization substrate. Since in contrast to She2p, mRNA tethered to She3p is localized in anaphase cells, anchoring may require transfer of mRNA from She2p onto an anchoring factor that is part of the Myo4p/She3p complex (Fig. 6) (24). By artificially tethering mRNA to She2p through high affinity binding, the transfer of mRNA required for anchoring may be abrogated.
Once the mRNA has been transferred from She2p onto the anchoring factor, She2p may dissociate from the anchoring complex, thus resulting in the observed uniform cellular distribution of She2p.
If She2p disassociates from the anchoring complex, it is possible that in addition to the heterotrimeric complex, Myo4p and She3p also associates with ASH1 mRNA in complexes that are devoid of She2p. Thus in the IP/RT-PCR experiments there is the potential for purifying transport complexes containing She2p as well as anchoring complexes that are devoid of She2p (Fig. 3). Similarly, in the mRNP purification assay all ASH1 containing mRNPs should be purified. Thus if Myo4p and She3p also associate with ASH1 mRNA in complexes devoid of Previous evidence for molecular remodeling of mRNA localization complexes has been observed in oocytes. In Drosophila oocytes a protein complex containing Swa and PABP have been shown to associate with bicoid mRNA (5). Like bicoid mRNA, both proteins accumulate in the anterior region of the oocyte, suggesting a potential role for these proteins in anchoring bicoid mRNA at the anterior cortex of the oocytes (5,39,40). In addition, the Mod and Nod proteins have also been shown to associate with bicoid mRNA and to form a complex with Swa and PABP in vivo (5). However unlike Swa and PABP, Mod and Nod do not accumulate at the anterior cortex of the oocyte, thus implicating these proteins in the transport phase of bicoid mRNA localization (5,41,42). Therefore the proper localization of an mRNA might be mediated by distinct transport and anchoring steps involving the remodeling of the RNP at various stages during the course of localization. In yeast, certain proteins like She2p may be a unique component of the transport complex while other proteins like She3p and Myo4p may be present in both transport and anchoring complexes.
In conclusion, this work provides new evidence that the Myo4p/She3p/She2p heterotrimeic complex participates in the transport of ASH1 mRNA, but is not actively involved in anchoring ASH1 mRNA at the bud tip. Furthermore since She3p and Myo4p appear to have roles in both transport and anchoring, this data suggests that ASH1 mRNA may be transferred from She2p onto a factor that is part of the Myo4p/She3p complex. The identity of the anchoring factor and the precise mechanism by which ASH1 mRNA is anchored at the bud tip awaits further characterization.

43.
Gietz, R. D., and Sugino, A. (1988) Gene 74, 527-534. Transformants were grown to mid-log phase in synthetic media containing 2% raffinsoe but lacking leucine and uracil. Expression of the reporter mRNA and fusion protein constructs was subsequently induced by the addition of galactose to a final concentration of 2% and incubation for an additional three hours. Subsequently, the cells were harvested and lysates were prepared for immunoprecipitation. The RNP complexes were purified using anti-FLAG antibody beads, and the co-precipitating proteins were separated by SDS-PAGE and analyzed by Western-blot using anti-myc and anti-FLAG antibodies.