Purification of Protein A-tagged Yeast Ran Reveals Association with a Novel Karyopherin β Family Member, Pdr6p*

The small GTPase Ran (encoded by GSP1and GSP2 in yeast) plays a central role in nucleocytoplasmic transport. GSP1 and GSP2 were tagged with protein A and functionally expressed in a gsp1null mutant. After affinity purification of protein A-tagged Gsp1p or Gsp2p by IgG-Sepharose chromatography, known karyopherin β transport receptors (e.g. Kap121p and Kap123p) and a novel member of this protein family, Pdr6p, were found to be associated with yeast Ran. Subsequent tagging of Pdr6p with green fluorescent protein revealed association with the nuclear pore complexes in vivo. Thus, functional tagging of yeast Ran allowed the study of its in vivo distribution and interaction with known and novel Ran-binding proteins.

The small GTPase Ran (encoded by GSP1 and GSP2 in yeast) plays a central role in nucleocytoplasmic transport. GSP1 and GSP2 were tagged with protein A and functionally expressed in a gsp1 null mutant. After affinity purification of protein A-tagged Gsp1p or Gsp2p by IgG-Sepharose chromatography, known karyopherin ␤ transport receptors (e.g. Kap121p and Kap123p) and a novel member of this protein family, Pdr6p, were found to be associated with yeast Ran. Subsequent tagging of Pdr6p with green fluorescent protein revealed association with the nuclear pore complexes in vivo. Thus, functional tagging of yeast Ran allowed the study of its in vivo distribution and interaction with known and novel Ran-binding proteins.
Ran belongs to the Ras superfamily of small guanine nucleotide-binding proteins and is highly conserved within the eukaryotic kingdom (for review, see Refs. 1,2). In Saccharomyces cerevisiae, Ran is encoded by the GSP1 and GSP2 genes (3,4). Although the gene products are nearly identical, GSP1 is essential for viability, whereas a gsp2 disruption does not lead to any apparent phenotype. The lethal gsp1 null mutation can be complemented by overexpression of GSP2, suggesting that the two isogenes are functionally equivalent but differently regulated (3).
Mainly two unique features distinguish Ran from the other members of the Ras superfamily: its high abundance in the cell and its nuclear location. In agreement with these features, Ran was shown to play an essential role in nucleocytoplasmic transport, a hallmark of eukaryotic cells (for review, see Refs. [5][6][7][8]. During transport Ran is continuously shuttling between nucleus and cytoplasm and switching between a GDP-bound and a GTP-bound state. The main accessory proteins, which assist Ran during transport, are NTF2, a RanGDP-binding protein at the nuclear envelope required for nuclear import of Ran (9), RanGAP1 in (Rna1p) in S. cerevisiae, a cytoplasmically located Ran-GTPase activating protein, and RanGEF (Prp20p in S. cerevisiae), a nuclear guanine nucleotide exchange protein, as catalysts of the Ran GTPase cycle. As a consequence of the compartmentalization of RanGAP and RanGEF, Ran is thought to be predominantly GDP-bound in the cytoplasm, whereas most nuclear Ran is probably in the GTP-bound form. This asymmetric distribution of the two forms of Ran across the nuclear envelope is thought to be a prerequisite for directional transport by nuclear transport receptors. Consistent with such a model, nuclear import receptors bind their cargoes in the absence of RanGTP (in the cytoplasm) and release it after binding to RanGTP (in the nucleus). In contrast, nuclear export receptors form trimeric complexes with export cargoes and RanGTP (in the nucleus), which are then disassembled after GTP hydrolysis on Ran by RanGAP1 (in the cytoplasm). The latter step requires in addition a cytoplasmic RanGTP-binding protein called RanBP1 (Yrb1p in S. cerevisiae; Refs. 10, 11). These nuclear transport receptors (also referred to as karyopherins), which include both import (also called importins) and export receptors (also called exportins), share structural similarity in their amino-terminal RanGTP-binding domain and constitute a protein family, the founding member of which is importin ␤ (Kap95p in S. cerevisiae), the subunit of the heterodimeric import receptor complex for proteins containing a classical nuclear localization sequence (for review, see Refs. 12,13).
Based on the pleiotropic phenotypes of alterations in the Ran-GTPase cycle, it was proposed that Ran may be involved in cellular processes other than nucleocytoplasmic transport (for review, see Ref. 1). The recent identification of a centrosomal Ran-interacting protein, RanBPM (14), and the implication of Ran in microtubule organization in the absence of nuclear trafficking (15)(16)(17) suggest a direct role of Ran at least in some of these processes.
To study the dynamic intracellular localization of Ran and its interaction with partner proteins in the living cell, Gsp1p and Gsp2p were tagged with GFP 1 and ProtA, respectively. Despite the bulky tags, Gsp1p and Gsp2p fusion proteins were functional. These tools allowed us to study the in vivo localization of Gsp1p and to identify a novel member of the karyopherin ␤ transport receptor family called Pdr6p.

EXPERIMENTAL PROCEDURES
Yeast Strains-Strains used in this study are listed in Table I. For construction and cultivation, standard procedures were applied. A GSP1 shuffle strain (Y1009) was constructed by transformation of diploid yeast strain YMO106 (gsp1::HIS3/GSP1; Ref. 4) with plasmid pRS316-GSP1 and subsequent sporulation and tetrad dissection. His ϩ , Ura ϩ spores were selected and checked for disruption at the GSP1 locus by PCR. Shuffle strains carrying additional rna1-1 or prp20-1 mutations were obtained by crossing this shuffle strain to previously described rna1-1 and prp20-1 mutant strains (18), respectively. For construction of a pdr6::HIS3 disruption strain, the entire open reading frame of PDR6 was replaced by the HIS3 gene. The diploid RS453 was transformed with the pdr6::HIS3 gene disruption construct which was * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. obtained by PCR using the HIS3 gene as a template and matching primers bearing 5Ј and 3Ј untranslated region sequences from the PDR6 gene (5Ј-GTAACAAAGAATAGCCTTCACCCTCAGCGGAGAAA-TTCTTAACCTCTTGGCCCCTCTAG-3Ј and 5Ј-CACAGATTTGAATAC-CATAGTAACTAGTCTCTTATTCTTAGTCGTTCAGAATGACACG-3Ј). His ϩ transformants were analyzed by PCR for the correct pdr6::HIS3 gene disruption. All four spores of complete tetrads from such transformants were viable, showing that PDR6 is not essential.
Affinity Purification of ProtA-TEV-Gsp1p, ProtA-TEV-Gsp2p, ProtA-TEV-Nup85, and ProtA-TEV-Pdr6p-Affinity purification of ProtAtagged fusion proteins by IgG-Sepharose chromatography was done as described earlier (22), with the following modifications. Strain gsp1::HIS3 complemented by ProtA-TEV-GSP1 or ProtA-TEV-GSP2 and strain pdr6::HIS3 complemented by ProtA-TEV-Pdr6p were grown in YPD or SDC-Leu medium, respectively, to A 600 of 1.6. Cells were harvested by centrifugation and spheroplasted. Three grams of spheroplasts were lysed in 20 ml of lysis buffer containing 150 mM KOAc, 20 mM Hepes, pH 7.0, 2 mM Mg(OAc) 2 , 0.1% Tween 20, 1 mM dithiothreitol, and a mixture of protease inhibitors (Roche Molecular Biochemicals). The whole-cell lysate was centrifuged at 15,000 rpm for 20 min at 4°C (Gsp1p/Gsp2p) or at 100,000 ϫ g for 1 h at 4°C in the ultracentrifuge (Pdr6p). The supernatant was passed over 300 l of IgG-Sepharose beads (Amersham Pharmacia Biotech). After washing the beads with 10 ml of the lysis buffer, the beads were transferred to spin columns, and 1 volume of beads was mixed with 1 volume of the lysis buffer and 1 l of recombinant TEV-protease (Life Technologies) per gram of spheroplasts. The mixture was incubated for 2 h at 16°C. The eluate was collected by centrifugation. 20 l of the eluate were used for analysis by SDS-PAGE, and 10 l were used for Western blot analysis. Affinity purification of ProtA-TEV-Gsp1p from the gsp1::HIS3/rna1-1 strain (Table I) was performed in a similar way, but cells were first grown at 23°C and shifted for 2 h to 37°C before harvesting and biochemical analysis.
Antibodies-The generation of the anti-Yrb1p antiserum will be described elsewhere. 2 A specific antiserum against Gsp1p was obtained by immunization of rabbits with affinity-purified MBP-Gsp1p-G21V. 3 The other antisera used in this study were kindly provided by the following people Mass Spectrometric Analysis-The protein bands of interest were excised from the one-dimensional SDS-polyacrylamide gel and pro-cessed for mass spetrometric analysis as described (22,25). Tryptic peptide maps were recorded on a Bruker REFLEX matrix-assisted laser desorption ionization time-of-flight mass spectrometer equipped with delayed ion extraction. Trypsin autolysis products were used for internal calibration to give an accuracy, on average, better than 30 ppm. Proteins were uniquely identified using the Peptide Search software package and a nonredundant protein sequence data base (maintained by C. Sander, European Molecular Biology Laboratory/European Bioinformatics Institute).
Intracellular Localization of GFP-Gsp1p and GFP-Pdr6p-Intracellular location of GFP-Gsp1p and GFP-Pdr6p was analyzed in a Zeiss Axioskop fluorescence microscope. Pictures were recorded with a Xillix Microimager charge-coupled device camera. Digital pictures were sometimes processed by digital confocal imaging applying deconvolution to images, available as a module within the software program Openlab (Improvision, Coventry, UK).

RESULTS AND DISCUSSION
Yeast Ran (Gsp1p and Gsp2p) Tagged with ProtA Is Functional-The protein A (ProtA) tagging strategy was successfully applied in the past to affinity purify in a single step a protein of interest from yeast and identify its interacting partners (26). We therefore intended to construct a functional yeast Ran (Gsp1p or Gsp2p) tagged with ProtA to identify novel Ran-interacting proteins. For this purpose, it was important to control whether the large ProtA tag (15 kDa) fused to Gsp1p or Gsp2p (27 kDa) did impair the in vivo function of this small GTPase. We inserted the corresponding GSP1 and GSP2 fusion genes into a centromeric (ARS/CEN/URA3) plasmid and tested for complementation of the otherwise nonviable gsp1::HIS3 mutant. pUN100-ProtA-GSP1 and pUN100-ProtA-GSP2 complemented the otherwise nonviable gsp1::HIS3 null mutant with only a slight growth retardation compared with a strain with authentic GSP1 on centromeric pRS314 (Fig. 1A). This shows that ProtA-tagged Gsp1p/Gsp2p constructs are functional in vivo. Furthermore, we tested whether GFPtagged Gsp1p and Gsp2p are functional. Similar to ProtA-Gsp1p, GFP-Gsp1p also could complement the gsp1::HIS3 cells (Fig. 1A). As expected, this functional GFP-Gsp1p exhibits a dual location in living cells, both in the nucleus and cytoplasm, with a higher concentration inside the nucleus (Fig. 1B). The same distribution was found for GFP-Gsp2p (data not shown). This intracellular GFP-Gsp1p location is consistent with previous findings using indirect immunofluorescence microscopy (3).
Affinity Purification of ProtA-Gsp2p Reveals Association with Pdr6p, a Novel Family Member of the Karyopherin ␤-like Transport Factors-GST-tagged Gsp1p has been recently affinity-purified from yeast, but no other co-isolating bands than Yrb1p were reported (27). It is possible that GST-Gsp1p is not fully functional in vivo. We therefore tested whether our ProtAtagged versions of Gsp1p and Gsp2p, which are functionally active, allow the co-enrichment of known and new Ran-binding proteins after affinity purification. To facilitate release of native yeast Ran from the affinity matrix, the viral TEV protease cleavage site was engineered between the ProtA and Gsp1 and 2 moieties (see "Experimental Procedures"). When ProtA-Gsp1p was isolated by IgG-Sepharose chromatography from the gsp1::HIS3-disrupted strain and released from the beads by TEV protease cleavage, Coomassie blue-stainable amounts of Gsp1p were recovered in the eluate. When the purified Gsp1p eluate was analyzed by SDS-PAGE and silver staining, several bands in the high molecular mass range became visible ( Fig.  2A). In particular the bands at ϳ100 -130 kDa appear to be specific, because since they are not seen in the purified Nup85p eluate, which served as a negative control and was isolated in a similar way by IgG-Sepharose chromatography and TEV proteolytic cleavage (Fig. 2A).
When ProtA-Gsp2p was affinity-purified and Gsp2p was released from the IgG-Sepharose beads by the TEV protease, the same prominent co-purifying bands in the range from 100 to 130 kDa were noticed (Fig. 2B). Some of these bands could be identified by mass spectrometric analysis and were shown to be known Ran-binding proteins (e.g. Kap123p and Kap121p, which are karyopherin ␤-family members involved in nuclear import of ribosomal subunits) (13). Interestingly, one of these bands in the 120-kDa range turned out to be Pdr6p, a protein previously implicated to play a possible role in pleiotropic drug resistance (Fig. 2B and Ref. 28). Although Pdr6p was reported to be distantly related to the karyopherin ␤-like transport receptors (29), binding to RanGTP could not be shown. Our data demonstrate that Pdr6p is indeed a bona fide member of the karyopherin ␤-family. According to the yeast "Kap" nomenclature, Pdr6p is called Kap122p. Additional bands in the Gsp2p eluate were Ssa1p or Eft1p (elongation factor 2). However, it is not clear whether these proteins specifically bind to Gsp2p or are contaminants. We have observed that Ssa1p tends to associate with several other ProtA-tagged fusion proteins when affinity-purified under similar conditions (data not shown). To test for the presence of other karyopherins, which were not detected by mass spectrometry, the Gsp2p eluate was analyzed by SDS-PAGE and Western blotting. This revealed that Kap95p and Kap104p, but not Mtr10p/Kap111p, are present in the eluate (data not shown). This result is consistent with the recent finding that Mtr10p has a low affinity for RanGTP (22).
Because after biochemical isolation of ProtA-tagged yeast Ran from wild-type cells, Ran-GTP may be largely hydrolyzed to Ran-GDP by Rna1p, we sought to stabilize the Gsp1p-GTP form by isolating ProtA-Gsp1p from rna1-1 mutant cells. Under these conditions, GTP hydrolysis by Ran GAP (Rna1p) should be inhibited, and a higher amount of RanGTP-binding proteins may co-purify with Gsp1p. A haploid rna1-1/gsp1::HIS3 strain was constructed, which was complemented by plasmid-borne ProtA-GSP1 (Table I). Before ProtA-Gsp1p affinity purification, this strain was shifted for 2 h to the nonpermissive temperature. Purified Gsp1p was analyzed by SDS-PAGE and silver staining (Fig. 2C). As anticipated, Gsp1p when isolated from rna1-1 cells contained a higher amount of karyopherin ␤ members. This was confirmed by Western blot analysis using antibodies against the exportins Los1p and Cse1p and the importin Kap95p (Fig. 2C). The strongest enrichment was observed for Cse1p. In addition, the RanGTP-binding protein Yrb1p was found to be significantly enriched with purified Gsp1p-GTP (Fig. 2C). In contrast, Gsp1p isolated from prp20-1 cells did not show an enhanced binding of these RanGTPbinding proteins (data not shown). Thus, biochemical purification of Gsp1p and Gsp2p from yeast allows the co-enrichment of several karyopherin ␤ members and other Ran-binding proteins.
GFP-Pdr6p Localizes Predominantly to the Nuclear Envelope-To study the in vivo role of Pdr6p, which we found to be physically associated with yeast Ran, we disrupted the PDR6 gene. Haploid pdr6::HIS3 progeny are viable, showing that PDR6 is not essential for cell growth. No apparent growth defect was observed in pdr6 Ϫ cells when grown at various temperatures (e.g. 23, 30, and 37°C; data not shown). Because Pdr6p was implicated to be involved in pleiotropic drug resistance (for review, see Ref. 28), we analyzed whether drugs such as cycloheximide cause different growth properties of pdr6 Ϫ cells compared with PDR6 ϩ cells. However, this was not the case (data not shown). When other stress pathways were tested in pdr6 Ϫ cells (e.g. thermotolerance, oxidative stress such as H 2 O 2 , and amino acid analogues), the cells also appeared to be normal compared with PDR6 ϩ cells (data not shown). Therefore, it remains to be shown whether Pdr6p is at all involved in pleiotropic multidrug resistance or other stress pathways. To find out whether Pdr6p not only binds to Gsp1p but also associates, like other karyopherin ␤-like transport receptors, with nuclear pore proteins, we examined its subcellular location. Pdr6p was tagged with GFP and analyzed by fluorescence microscopy. GFP-Pdr6p exhibits a punctate nuclear envelope staining typical for a nuclear pore complex association (Fig. 3A,  upper panel). This is in agreement with the finding that GFP-Pdr6p co-clusters with nuclear pores in nup133 Ϫ cells (Fig. 3A,  lower panel). In addition, some GFP-Pdr6p can also be detected in the cytoplasm and nucleoplasm. To find out whether Pdr6p is an import receptor for a specific cargo(es), ProtA-tagged Pdr6p was affinity-purified from a whole-cell yeast lysate. Although ProtA-Pdr6p purified very well and was easily detected by SDS-PAGE and Coomassie blue staining, no major transport substrate appeared to co-purify (Fig. 3B). Only Hsp70 proteins and some minor bands were noticed (Fig. 3B). Thus, it remains to be shown whether Pdr6p/Kap122p is an importin or exportin and to which transport cargo it binds. However, the fact that Pdr6p was found to be associated with ProtA-Gsp2p under standard isolation conditions, in which mainly importins were recovered, would favor a model of Pdr6p/Kap122p being an import receptor. Attempts to find genetically interacting partners of PDR6 failed so far. A synthetic lethal screen with the pdr6::HIS3 null allele as described previously (24) did not yield synthetic lethal mutants. 4 Furthermore, double haploid mutants were constructed between the pdr6::HIS3 strain and several mutant or knockout karyopherin alleles (e.g. pdr6::HIS3/los1::HIS3, pdr6::HIS3/mtr10-7, pdr6::HIS3/ kap104::HIS3, pdr6::HIS/kap121-1, pdr6::HIS3/kap123:: HIS3, pdr6::HIS3/sxm1::HIS3, pdr6::HIS3/nmd5::HIS3, pdr6::HIS3/cse1-1, pdr6::HIS3/xpo1-1, and pdr6::HIS3/ msn5::TRP1) and found to be viable. 4 This suggests that Pdr6p either fulfills a completely nonessential function in yeast (at least under laboratory growth conditions), or that Pdr6p is highly redundant with other karyopherin ␤-family members. It could also be possible that Pdr6p is involved in a regulated nuclear import or export pathway and thus only required under special growth or stress conditions. This was recently found for some of the uncharacterized yeast ␤-like karyopherins, which were shown to be involved in regulated nuclear import and export mechanisms under conditions of stress or under distinct metabolic stages (30 -33).
In conclusion, yeast Ran was functionally modified with both the GFP and ProtA tag, which allowed the study of the steady state in vivo location and its interaction with partner proteins. This enabled us to identify a novel karyopherin ␤-like transport receptor, and this system will serve in the future as a useful tool to further study the cell biological and biochemical aspects of this essential regulator of eukaryotic nucleocytoplasmic transport.