A genomic integration method for the simultaneous visualization of endogenous mRNAs and their translation products in living yeast

Protein localization within cells can be achieved by the targeting and localized translation of mRNA. Yet, our understanding of the dynamics of mRNA targeting and protein localization, and how general this phenomenon is, is not clear. The authors previously developed a method (m-TAG) to localize endogenously expressed mRNAs in yeast by chromosomal insertion of the MS2 aptamer sequence between the open-reading frame (ORF) and 3′ UTR of any gene. Here they describe an advanced method (mp-TAG) that allows for the simultaneous visualization of both endogenously expressed mRNAs and their translation products in living yeast for the first time.

Abstract

Protein localization within cells can be achieved by the targeting and localized translation of mRNA. Yet, our understanding of the dynamics of mRNA targeting and protein localization, and of how general this phenomenon is, is not clear. Plasmid-based expression systems have been used to visualize exogenously expressed mRNAs and proteins; however, these methods typically produce them at levels greater than endogenous and can result in mislocalization. Hence, a method that allows for the simultaneous visualization of endogenous mRNAs and their translation products in living cells is needed. We previously developed a method (m-TAG) to localize endogenously expressed mRNAs in yeast by chromosomal insertion of the MS2 aptamer sequence between the open-reading frame (ORF) and 3′ UTR of any gene. Upon coexpression with the MS2 RNA-binding coat protein (MS2-CP) fused with GFP, the aptamer-tagged mRNAs bearing their 3′ UTRs are localized using fluorescence microscopy. Here we describe an advanced method (mp-TAG) that allows for the simultaneous visualization of both endogenously expressed mRNAs and their translation products in living yeast for the first time. Homologous recombination is used to insert the mCherry gene and MS2-CP binding sites downstream from any ORF, in order to localize protein and mRNA, respectively. As proof of the concept, we tagged ATP2 as a representative gene and demonstrated that endogenous ATP2 mRNA and protein localize to mitochondria, as shown previously. In addition, we demonstrate that tagged proteins like Hhf2, Vph1, and Yef3 localize to their expected subcellular location, while the localization of their mRNAs is revealed for the first time.

INTRODUCTION

A distinctive feature of eukaryotic cells is the targeted distribution of protein to organelles. The specific targeting of proteins can be achieved by several known mechanisms: post-translational, co-translational, and pre-translational. Post-translational localization, the best-known mechanism, is achieved via protein-targeting motifs embedded within the protein. Thus, proteins that localize to specific organelles, such as the nucleus, mitochondria, peroxisomes, as well as type II membrane proteins that translocate to the endoplasmic reticulum (ER), typically have these motifs present either in their N- or C-terminals (Dingwall and Laskey 1991; Emanuelsson and von Heijne 2001). Co-translational localization has been shown to target either secreted or type I integral membrane proteins to the ER and may also utilize targeting sequences (e.g., the signal peptide) embedded within the protein (Koch et al. 2003; Schwartz 2007). In this mechanism, the ribosome–mRNA–nascent chain complex is targeted by a signal recognition particle (SRP) to the ER through binding of the SRP to its receptor on the ER surface, and allows for subsequent protein translocation upon translation. Finally, a pre-translational mechanism involves the targeted localization of mRNAs prior to their translation. Thus, the ability to track visually both an mRNA and its translation product would be an important and critical tool for understanding the mechanism of protein localization. Until now, visualization of the dynamics between mRNA localization and local protein synthesis has been examined using plasmid-based expression systems that exogenously express mRNAs encoding gene fusions with red fluorescent protein (RFP) and binding sites for the RNA-binding MS2 coat protein from bacteriophage (MS2-CP) (Aronov et al. 2007). When coexpressed with MS2-CP fused to green fluorescent protein (GFP; MS2-CP-GFP), both mRNA and protein localization can be visualized in living cells (Aronov et al. 2007). However, exogenous gene expression can result in mRNA levels greater than normal and often leads to mRNA and protein mislocalization.

In our efforts to create a genome-wide map of mRNA localization (the mRNA localizome), we previously developed m-TAG, a technique that allows for the sustained visualization of endogenous mRNAs in living yeast for the first time (Haim et al. 2007; Haim-Vilmovsky and Gerst 2009). m-TAG allows for the insertion of MS2-CP binding sites (i.e., MS2 loops; MS2L) into any gene by homologous recombination and, as a unique feature of this genome-tagging procedure, leaves the 3′ UTR intact and present in the expressed mRNA. Here we describe advancement over m-TAG that allows for the simultaneous visualization of both endogenously expressed mRNAs and their translation products in vivo for the first time. We employed mp-TAG to study the ATP2 gene and showed that its mRNA and protein localization was identical to that previously described using separate labeling and visualization methodologies. In addition, we verified that mp-TAG could correctly confirm different examples of protein localization while, at the same time, revealing the localization of the encoded mRNAs.

RESULTS

The mp-TAG technique

mp-TAG (Fig. 1) employs a genomic integration cassette that not only tags the 3′ end of an open reading frame (ORF) directly upstream of the 3′ UTR with the MS2L sequence (for RNA visualization) but also creates a gene fusion between the ORF and the mCherry gene (for protein visualization). The cassette is integrated into a gene of interest by homology-based recombination, and importantly, the auxotrophic selection marker (e.g., Schizosaccharomyces pombe his5⁺) necessary for efficient integration is flanked by loxP sites and can be subsequently eliminated upon the expression of Cre recombinase (Fig. 1). Removal of the selection marker is critical for both m-TAG and mp-TAG as it allows for the retention of the 3′ UTR within the endogenously expressed message. This differs from other genome-tagging strategies, such as the oft-used GFP-tagging system (Longtine et al. 1998; Huh et al. 2003), wherein the 3′ UTR is dissociated from the coding region due to the presence of the selection marker. Since 3′ UTRs typically bear sequences/motifs critical for mRNA localization and stability (Andreassi and Riccio 2009), use of mp-TAG should not impede the correct targeting of a given mRNA and its translation product, as might occur with other 3′-tagging strategies.

FIGURE 1.

A schematic representation of mp-TAG. (A) The template cassette contains in the following order (5′-3′): a short linker (data not shown), the mCherry gene, a selectable marker (i.e., Sphis5+) flanked by loxP sites, and 12 MS2 loop sequences (12 MS2L). As illustrated, forward and reverse primers homologous to the coding region (excluding the stop codon) and 3′ UTR of a given gene (ORF), respectively, are used to amplify the template cassette by PCR. (B) The PCR product is transformed into yeast and integrates into the genomic locus by homologous recombination. Both the linker sequence and mCherry are fused to the ORF (the stop codon of mCherry is used for translation termination), while the MS2 loops are inserted upstream of the 3′ UTR. (C) The selectable marker located between the loxP sites is removed by cre recombinase expression as shown in D. (E) After marker excision, which results in reincorporation of the 3′ UTR into the transcript, cells are transformed with a plasmid expressing MS2-CP-GFP(x3) in order to visualize the tagged mRNA via GFP fluorescence and the translated protein via mCherry fluorescence (F).

ATP2 mRNA and protein localize to mitochondria

In yeast, over 500 mRNAs encoding mitochondrial proteins were found to localize to mitochondria-bound polysomes (Garcia et al. 2007). ATP2 encodes the β subunit of mitochondrial F1-ATP synthase complex (Takeda et al. 1985), and its mRNA has been shown to localize to mitochondria by m-TAG, subcellular fractionation and microarray analysis, and plasmid-based expression experiments (Marc et al. 2002; Margeot et al. 2002; Gadir et al. 2011). The 3′ UTR of ATP2 mRNA contains signals that contribute to its cellular localization (Margeot et al. 2002; Gadir et al. 2011), and consequently, ATP2 mRNA can serve as a model to investigate the role of translation in mRNA localization. Thus, we used ATP2 to test the feasibility of mp-TAG. The template cassette (Fig. 1), which contains the mCherry sequence, the S. pombe histidine biosynthesis (Sphis5⁺) selectable marker, and 12 MS2 loops, was amplified using the following: (1) a forward primer complementary to the 3′ end of the ATP2 coding region (excluding the stop codon) and a linker sequence originating in the integration plasmid, and (2) a reverse primer complementary to the 5′ end of the ATP2 3′ UTR and a nonrepetitive sequence present at the 3′ end of MS2L. The PCR product containing the mCherry::loxP::his5⁺::loxP::MS2L cassette was transformed into wild-type yeast and led to the formation of colonies on solid medium lacking histidine. In order to identify cells bearing the correct integration, PCR amplification was performed on lysed samples using forward and reverse primers complementary to the ATP2 and mCherry coding regions, respectively. PCR products were sized by electrophoresis to select colonies for subsequent visual detection of mCherry signals using fluorescence microscopy. Cells passing these tests were transformed with a plasmid containing cre recombinase under the control of galactose-inducible promoter, and growth on galactose-containing medium resulted in the excision of Sphis5⁺ to yield ATP2::mCherry::loxP::MS2L::ATP2^3′UTR (ATP2_mpINT) cells.

To visualize ATP2 mRNA localization in the ATP2_mpINT strain, cells were transformed with the MS2-CP-GFP(x3) plasmid, and the fluorescence signals from both GFP (i.e., RNA granules) and mCherry (i.e., translated fusion protein) were detected (Fig. 2A). Endogenous ATP2 mRNA granules were observed to be ∼200 nm in size (i.e., their fluorescence footprint) and were not stationary but moved along mitochondrial tubules as previously described (Gadir et al. 2011). We quantified the number of mRNAs per granule using an integrated 128-mer lac operator system for RNA fluorescence quantification (Haim et al. 2007; Haim-Vilmovsky and Gerst 2009) and found an average of about nine mRNAs per granule (e.g., average ± SD = 8.9 ± 6.1, n = 12 granules measured) and a range of two to 24 mRNAs per granule. Importantly, ATP2 mRNA granules localized to mitochondria (i.e., labeled by Atp2-mCherry protein) in 84% of cells (n = 26), as was seen in earlier studies (Marc et al. 2002; Margeot et al. 2002; Gadir et al. 2011). Moreover, to confirm Atp2-mCherry protein localization to the mitochondria, ATP2_mpINT cells were transformed with a plasmid expressing GFP fused to a mitochondrial targeting sequence (mtGFP), which labels the mitochondria upon translation. A colocalization of 100% (n = 100) was seen between the translated Atp2-mCherry and mtGFP (Fig. 2B). No granules or red fluorescence signal was observed in cells expressing MS2-CP-GFP(x3) but lacking the integrated mCherry and MS2L cassette.

FIGURE 2.

Endogenous ATP2 mRNA and protein localization. (A) mp-TAG labels ATP2 mRNA and protein. Representative fluorescence microscopy and light (differential interference contrast [DIC]) images of ATP2::loxP::MS2L::ATP2^3′-UTR integrated cells (ATP2_mpINT), transformed with a plasmid expressing MS2-CP-GFP(x3), are shown. Eighty-four percent of the GFP granules (mRNA) colocalized with the Atp2-mCherry signal (Protein). Scale bar, 1 μm. (B) Atp2-mCherry labels mitochondria. Representative fluorescence microscopy and DIC images of integrated ATP2_mpINT cells expressing Atp2-mCherry (Protein) and transformed with a plasmid expressing GFP fused to a mitochondrial targeting sequence (Mitochondria) are shown. Atp2-mCherry colocalized to mitochondria in 100% of the cells. Scale bar, 1 μm. (C) mp-TAG does not affect protein function. Wild-type, atp2Δ, and ATP2_mpINT cells were grown to mid-log phase on glucose-containing medium, serially diluted, plated onto solid synthetic medium containing glycerol (Gly) as a carbon source or rich medium (YPD), and grown for 72 h and 48 h, respectively, at 26°C. Note that functional Atp2 is necessary for growth on glycerol.

Yeast require functional mitochondria to utilize glycerol as a carbon source. To verify that integration into the ATP2 locus did not alter protein function, we examined the ability of the ATP2_mpINT cells to grow on glycerol-containing medium (Fig. 2C). By performing drop tests on solid medium, we found that yeast expressing mCherry- and MS2L-tagged ATP2 grew essentially like wild-type cells, whereas yeast lacking the ATP2 gene was unable to grow altogether (Fig. 2C). Thus, Atp2 is fully functional after the integration of mCherry and MS2L into the ATP2 locus.

Hhf2 localizes to the nucleus, while HHF2 mRNA is polarized to the bud tip

We next determined whether other mRNAs and their translation products can be localized using mp-TAG. We tagged HHF2, which encodes a core histone required for chromatin assembly and chromosome function (Smith and Andresson 1983; Grunstein 1990), and found that HHF2 mRNA granules localized to the tip of small-budded cells HHF2::mCherry::loxP::MS2L::HHF2^3′UTR (HHF2_mpINT) cells expressing MS2-CP-GFP(x3) (e.g., 83% bud tip localization, n = 100 cells) (Fig. 3A), while Hhf2-mCherry protein localized to the nucleus (as stained by DAPI) in 100% of cells (n = 100) (Fig. 3B). Similar results for mRNA localization were obtained using m-TAG alone (e.g., HHF2::loxP::MS2L::HHF2^3′UTR (HHF2_INT) cells; 84% bud tip localization, n = 25 cells). HHF2_mpINT granules contained about 12 mRNAs (average ± SD = 12.3 ± 1.5, n = 11 granules measured), and the range was nine to 15 mRNAs per granule. In addition, we found 50% colocalization (n = 50 granules counted) of HHF2 RNA granules with membranes labeled with Sec63-RFP, an ER marker, and a low level of colocalization with mitochondria labeled by Oxa1-RFP (e.g., 20%, n = 53 granules) or nuclei labeled by DAPI (e.g., 19%, n = 53 granules). Interestingly, while HHF2 RNA localizes to the bud tip, its translation product localizes to the nucleus. This is identical to that previously described for ASH1 mRNA and protein, which functions as a transcriptional repressor and cell fate determinant (Jansen et al. 1996; Long et al. 1997; Takizawa et al. 1997; Haim et al. 2007). ASH1 mRNA is cotransported with cortical ER (cER) along with mRNAs encoding polarity and secretion factors involved in polarized cell growth (Aronov et al. 2007).

FIGURE 3.

Endogenous mRNA and protein localization. (A) HHF2 mRNA localizes to bud tip, while Hhf2 protein localizes to the nucleus. Representative light (DIC) and fluorescence microscopy images of HHF2::mCherry::loxP::MS2L::HHF2^3′UTR (HHF2_mpINT) cells transformed with a plasmid expressing MS2-CP–GFP(x3). HHF2 mRNA granules (mRNA) localize to the tip of small-budded cells, while Hhf2-mCherry (Protein) localizes to the nucleus, as indicated below in B. Scale bar, 1 μm. (B) Hhf2 protein colocalizes with DAPI staining. Representative fluorescence and light (DIC) microscopy images of HHF2_mpINT cells stained with DAPI (Nucleus). Note that Hhf2-mCherry (Protein) colocalizes with DAPI in the nucleus. Scale bar, 1 μm. (C) VPH1 mRNA and protein localization. Representative light (DIC) and fluorescence microscopy images of VPH1::mCherry::loxP::MS2L::VPH1^3′UTR (VPH1_mpINT) cells transformed with a plasmid expressing MS2-CP–GFP(x3). VPH1 mRNA granules (mRNA) do not colocalize with Vph1-mCherry (Protein), which labels the vacuole membrane, as indicated below in D. Scale bar, 1 μm. (D) Vph1-mCherry labels the vacuole membrane. Representative light (DIC) and fluorescence microscopy images of VPH1_mpINT cells expressing Vph1-mCherry (Protein) and transformed with a plasmid expressing CPS1-GFP, a vacuolar membrane marker (Vacuole). Scale bar, 1 μm. (E) VPH1 mRNA localizes to mitochondria. Representative light (DIC) and fluorescence microscopy images of VPH1::loxP::MS2L::VPH1^3′UTR (VPH1_INT) cells transformed with plasmids expressing MS2-CP–GFP(x3) to localize VPH1 mRNA (mRNA) and Oxa1-RFP, a marker that labels the mitochondria (Mitochondria). Scale bar, 1 μm. (F) YEF3 mRNA and protein localization. Representative light (DIC) and fluorescence microscopy images of YEF3::mCherry::loxP::MS2L::YEF3^3′UTR (YEF3_mpINT) cells transformed with a plasmid expressing MS2-CP–GFP(x3) to localize YEF3 mRNA (mRNA) and Yef3-mCherry protein (Protein). Scale bar, 1 μm. (G) Yef3-mCherry labels the cytoplasm. Representative light (DIC) and fluorescence microscopy images of YEF3_mpINT cells expressing Yef3-mCherry (Protein) and labeled with DAPI to stain the nucleus (Nucleus). (H) YEF3 mRNA localizes to ER. Representative light (DIC) and fluorescence microscopy images of YEF3::loxP::MS2L::YEF3^3′UTR (YEF3_INT) cells transformed with plasmids expressing MS2-CP–GFP(x3) to visualize YEF3 mRNA (mRNA) and Sec63-RFP to label the ER (ER). Seventy-three percent of YEF3 mRNA granules localized to the ER. Scale bar, 1 μm.

Vph1 protein is found in the vacuole membrane, while its transcript is associated with mitochondria

We then tagged VPH1, which encodes a vacuolar membrane protein (Manolson et al. 1992) that serves as a subunit of the yeast vacuolar ATPase (V-ATPase) and is required for its assembly and enzymatic activity. The V-ATPase consists of two separable domains: the V₁ domain is peripherally associated with the vacuolar membrane and catalyzes ATP hydrolysis, while the V₀ domain is an integral membrane structure of five subunits that transports protons across the membrane. Vph1 is part of the V-ATPase V₀ domain (Manolson et al. 1992). Upon tagging, we found that VPH1::mCherry::loxP::MS2L::VPH1^3′UTR (VPH1_mpINT) RNA granules were not localized to the vacuole, although Vph1-mCherry localized to the vacuole membrane (Fig. 3C), which could be verified by vacuolar labeling with Cps1-GFP (Fig. 3D; Manolson et al. 1992). VPH1_mpINT RNA granules contained about 14 mRNAs (14.0 ± 7.8, n = 13 granules measured) and ranged from two to 28 mRNAs per granule, as determined using quantitative fluorescence analysis. The intracellular localization of VPH1 mRNA localization was examined further using a VPH1::loxP::MS2L::VPH1^3′-UTR (VPH1_INT) strain. We found that 84% of VPH1_INT mRNA granules (n = 100) (Fig. 3E) localized with mitochondria (i.e., organelles labeled by Oxa1-RFP), while in separate experiments using other organellar markers, we saw a low level of localization with ER labeled by Sec63-RFP (e.g., 32% colocalization, n = 50 granules), and little to no localization with peroxisomes labeled by RFP-PTS1 (e.g., 8%, n = 50 granules).

To determine whether Vph1 is functional after mCherry and MS2L integration, we examined the growth of VPH1_INT cells in comparison with wild-type and vph1Δ cells. We found that both wild-type and VPH1_INT cells grow normally (at 30°C), while control vph1Δ cells grow slowly (Supplemental Fig. S1). Thus, Vph1 is functional after the integration of mCherry and MS2L into the VPH1 locus. Overall, the results indicate that VPH1 mRNA is targeted primarily to mitochondria, although Vph1 protein localized correctly to the vacuolar membrane.

Yef3 protein localizes with the cytoplasm, while its mRNA localizes to the ER

Finally, we tagged YEF3, which encodes the γ subunit of translation elongation factor eEF1B (Qin et al. 1987). We created a YEF3::mCherry::loxP::MS2L::YEF3^3′UTR (YEF3_mpINT) strain and visualized both YEF3 mRNA granules (Fig. 3F) and Yef3-mCherry, the latter being spread throughout the cytoplasm (Fig. 3F,G). To better examine YEF3 mRNA localization, we created a YEF3::loxP::MS2L::YEF3^3′-UTR (YEF3_INT) strain and transformed it with plasmids expressing MS2-CP-GFP(x3) and Sec63-RFP. We found that 73% (n = 26 granules) of YEF3_INT granules localized to the ER (Fig. 3H), both ER peripheral to the nucleus and cER. We noted that YEF3_mpINT RNA granules were not highly polarized and showed only 36.6% (n = 30 granules) localization to the bud tip. In addition, only a low level of colocalization (20%, n = 25 granules) with mitochondria (labeled by Oxa1-RFP) was observed. Finally, we determined using quantitative fluorescence analysis that YEF3_mpINT granules contained about six mRNAs (5.8 ± 2.2, n = 10 granules measured) and had a range of two to eight mRNAs per individual granule.

DISCUSSION

Understanding the connection between mRNA targeting and subsequent protein localization is of great interest. Although plasmid-based expression systems have been used to visualize both mRNA and protein, these methods typically result in exogenous mRNA levels that are greater than the endogenous and could lead to both mRNA and protein mislocalization. Consequently, the localization of most mRNAs in cells is still undetermined, and there is a pressing need for a simple and fast method that allows for the visualization of a wide range of endogenous mRNAs and their translated proteins in living cells. Moreover, the power to follow mRNA and protein simultaneously in vivo can facilitate the investigation of whether protein import to a given organelle occurs in a post- or co-translational manner, which is a matter of current debate for those studying protein translocation into mitochondria or peroxisomes, for example.

We previously devised a novel method for visualizing endogenously expressed mRNAs in living yeast for the first time. m-TAG has been employed successfully by us to determine the localization of representative mRNAs that encode proteins of the mitochondria, ER, and peroxisome (Zipor et al. 2009; Gadir et al. 2011). We have now formulated a superior method, mp-TAG, that allows for the simultaneous visualization of endogenously expressed mRNA and protein. We employed mp-TAG to test the localization of several mRNAs and their translation products and to verify that tagging did not alter localization. We first demonstrated that both ATP2 mRNA and Atp2 protein colocalize with mitochondria (Fig. 2A,B), as seen in earlier studies (Marc et al. 2002; Margeot et al. 2002; Gadir et al. 2011). This result served as a proof of the concept for the validity of the mp-TAG method. Interestingly, when we tagged HHF2, we observed that while HHF2 mRNA localizes to the bud tip, its translation product localizes to the nucleus (Fig. 3A and B). This result parallels that of ASH1 mRNA (the archetypal polarized mRNA in yeast) and protein (Jansen et al. 1996; Long et al. 1997; Takizawa et al. 1997; Haim et al. 2007) and suggests that both HHF2 and ASH1 mRNA and protein undergo co-trafficking, respectively. Additional investigation shows that VPH1 mRNA is targeted primarily to mitochondria, although the Vph1 protein localized correctly to the vacuolar membrane (Fig. 3C–E). This surprising result might indicate that Vph1 protein is translated at mitochondrial—ER junctions and can access the secretory pathway (i.e., ER) via such sites. This will be the subject of future studies. Finally, we found that YEF3 mRNA localized with the ER, while its protein was found present in the cytoplasm (Fig. 3F–H). YEF3 mRNA was previously shown to be associated with the She2 RNA-binding protein, which is involved in mRNA localization to the bud tip (Oeffinger et al. 2007). However, we observed only partial YEF3 mRNA localization to the bud, in contrast to HHF2 mRNA. Thus, YEF3 mRNA is not highly polarized and is probably less likely to undergo cotransport with cER to the bud tip, as seen with ASH1 mRNA and mRNAs encoding polarity and secretion factors (Aronov et al. 2007) and predicted here for HHF2 mRNA.

To verify that mp-TAG did not distort or impair protein function, we examined, where possible, the ability of the tagged strains to grow under conditions whereby a loss in function would compromise cell growth. As we have shown previously for other MS2L-tagged mRNAs (Haim et al. 2007; Zipor et al. 2009; Gadir et al. 2011), tagging of the mRNA with either MS2L and the protein with mCherry did not affect the functionality of either Atp2 or Vph1 (Fig. 2C; Supplemental Fig. S1). Taken together, these results demonstrate that mp-TAG is a simple and effective means for visualizing both endogenous mRNAs and their translated proteins in vivo for the first time. Although this technique works best for highly expressed genes and their proteins (probably due to the weaker signals and/or slower folding rate of mCherry), it does not remove the 3′ UTR from the coding region, which can result in hypomorphic phenotypes. Thus, the protein-labeling component of the mp-TAG technique represents a significant advance over existing genomic tagging techniques that produce C-terminal-fluorescent protein fusions in yeast.

In addition to the ability of mp-TAG to localize mRNA and protein simultaneously, it can also be used to quantify the relative amounts of a given transcript and its translation product in cells. For example, by employing qRT-PCR to quantify RNA and Western analysis to quantify protein with antibodies against mCherry (or the native component of the fusion protein), their relative abundance can be accurately measured. Protein quantification necessitates generation of a standard curve using increasing amounts of recombinant mCherry, e.g., in Westerns performed in parallel and under the same conditions as the query. By examining RNA and protein levels with mp-TAG under changing environmental conditions, it is then possible to quantify variations in both transcript and protein expression for any given gene in a simple and efficient manner. In addition, this dual mRNA-protein tagging strategy also allows for the use of yeast genetics to identify the genes that ultimately control the mRNA and protein targeting processes.

MATERIALS AND METHODS

Media, DNA, genetic manipulations, and yeast strains

Yeast was grown in standard growth media containing either 2% glucose or 3.5% galactose. Synthetic complete (SC) and drop-out media containing glucose were prepared essentially as described (Haim-Vilmovsky and Gerst 2009). Standard rich medium (YPD) was prepared as described (Rose et al. 1990). For growth on synthetic medium containing glycerol (see below), 3% glycerol was substituted for glucose. Standard methods were used for the introduction of DNA into yeast and the preparation of genomic DNA (Rose et al. 1990). The primers used in this study are listed in Supplemental Table 1. The yeast strains used in this study are listed in Supplemental Table 2.

Plasmids

Plasmid pLOXHIS5MS2L (Haim et al. 2007), which contains the loxP-Sphis5+-loxP-MS2L cassette, was used as the vector backbone to create the template plasmid for generating constructs for genomic integration by PCR. Plasmid pNat-TEF2p-mCherry, which contains the mCherry coding sequence, was provided by M. Schuldiner (Weizmann Institute of Science, Rehovot, Israel). mCherry without its ATG was amplified from the pNat-TEF2p-mCherry vector using the mCherryF and mCherryR primers (Supplemental Table 1) containing HindIII sites and cloned into pGEM-Teasy (Promega) to yield plasmid pmCherry-HIII. Next, a 711-bp HindIII fragment digested from pmCherry-HIII was cloned (in the correct orientation) into the HindIII site located upstream of the 5′ LoxP site in pLOXHIS5MS2L to yield plasmid pmChLOXHIS5MS2L (see Supplemental Information Sequence). Plasmid pMS2-CP-GFP(x3), which expresses MS2-CP fused to three GFPs under the control of the MET25 promoter; plasmid RFP-PTS1, which labels peroxisomes; and plasmid pAD4Δ-OXA1-RFP-3′ UTR, which expresses OXA1-RFP, were previously described (Haim et al. 2007; Zipor et al. 2009). A multi-copy plasmid expressing Sec63-RFP (pSM1960) was provided by S. Michaelis (John Hopkins University, Baltimore, MD). Plasmid pSH47, which expresses cre recombinase from under the control of a galactose-inducible promoter, was obtained from Euroscarf. Plasmid pGO426-CPS1-GFP, which expresses CPS1-GFP, was provided by S. Emr (Cornell University, Ithaca, NY). Plasmid pA521–PYX122, which expresses GFP fused to a mitochondrial targeting sequence under the control of a TPI promotor, was provided by B. Westermann (Universität Bayreuth, Bayreuth, Germany).

Genomic integration of mCherry and the MS2-CP binding sites

For each gene to be tagged, a forward primer complementary to the 3′ end of the coding region (i.e., overlapping by ∼40 bp, excluding stop codon) and 5′ end of the mCherry::loxP::Sphis5+::loxP::MS2L cassette was used, along with a reverse primer complementary to the 5′ end of the 3′ UTR (i.e., overlapping by ∼40 bp) and 3′ end of the mCherry::loxP::Sphis5+::loxP::MS2L cassette, in the PCR reaction with plasmid pmChLOXHIS5MS2L as the template (for a schematic representation, see Fig. 1). See previous works for a complete description of the m-TAG genome tagging procedure (Haim et al. 2007; Haim-Vilmovsky and Gerst 2009), which is essentially the same as that used for mp-TAG. PCR products of the correct size (e.g., 2890 bp) were transformed into yeast and grown on plates containing SC medium lacking histidine for 3–5 d in 26°C. To confirm the proper integration into the genome, samples of individual yeast colonies were treated with zymolase and subjected to PCR amplification using a forward primer complementary to the coding region and a reverse primer complementary to the mCherry sequence. PCR products were sized on agarose gels to identify colonies bearing the correct integration. Samples of cells were then taken for the detection of red fluorescence signals using fluorescence microscopy (i.e., as a phenotypic test for integration and mCherry protein expression).

Next, yeast bearing the correct-sized mCherry::loxP::Sphis5+::loxP::MS2L integrations were transformed with plasmid pSH47 and grown on SC medium lacking histidine and uracil. Cre recombinase expression was induced by growing the transformed cells in SC medium containing galactose, but lacking uracil, for 16 h in 26°C. Cells were then diluted, plated and grown on SC medium lacking uracil, and replica-plated to determine the presence or absence of the Sphis5+ selection marker. Yeast bearing the mCherry::loxP::MS2L integration (∼1400 bp) was verified by PCR amplification (i.e., using primers complementary to the coding region and 3′ UTR, respectively).

MS2-CP-GFP(x3) expression and mRNA visualization

MS2L integrated strains were transformed with a plasmid expressing MS2-CP-GFP(x3) and fusion protein expression was induced by growth in synthetic medium lacking methionine for 1hr at 26°C. Cells were visualized by confocal fluorescence microscopy using a Zeiss LSM510 Meta confocal microscope and a PLANApo 100x/oil objective. The following wavelengths were used: for GFP, excitation at 480 nm and emission at 530 nm; for mCherry, excitation at 545 nm and emission at 560 to 580 nm. For scoring mRNA granule localization, GFP-labeled granules were counted relative to a given secondary label (either the tagged mCherry fusion protein or an RFP-labeled organellar marker. For every aptamer-tagged RNA examined, the number of granules scored for localization and the corresponding statistics regarding localization are given in the text. The average number of granules observed per cell (±SD) was 1.17 ± 0.41 (n = 71 cells), 1.13 ± 0.36 (n = 102 cells), 1.06 ± 0.24 (n = 69 cells), and 1.00 ± 0.00 (n = 46 cells) for the aptamer-tagged ATP2, HHF2, YEF3, and VPH1 mRNAs, respectively. The range of granules per cell observed was one to three for ATP2 and HHF2, one to two for YEF3, and one for VPH1.

Growth tests

Yeast was grown on standard synthetic and rich growth media. For growth tests on solid medium, yeast strains were grown to mid-log phase, normalized for optical density (absorbance at 600 nm), diluted serially, and plated by drops onto the appropriate solid medium preincubated at the different temperatures used in the experiments. To assess the functionality of the Atp2-mCherry fusion, yeast was plated onto rich growth medium containing glucose (YPD) and synthetic growth medium containing 3% glycerol. YPD plates were incubated at for 48 h 26°C, while the synthetic plates were incubated for 72 h at 26°C . To assess functionality of the Vph1-mCherry fusion, yeast strains were plated onto YPD and incubated for 2 d at 30°C.

SUPPLEMENTAL MATERIAL

Supplemental material is available for this article.