Abstract
Far3p (factor arrest), a protein that interacts with Far7–11p, is required for the pheromone-mediated cell cycle arrest in G1 phase. We used a combination of computational and experimental strategies to identify the Far3p self-association, to map the Far3p domains that interact with Far3p itself and with other Far proteins, and to reveal the importance of the two coiled-coil motifs of Far3p in the integrity and function of the Far complex. We show that Far3p self-associates through its central region and its C-terminal coiled-coil domain, that the amino acid 61–100 region of Far3p interacts with Far7p, and that the Far3p N-terminal coiled-coil domain interacts with Far9p and Far10p. Mutation of the N-terminal coiled coil disrupts the interactions of Far3p with Far9p and Far10p, and mutation of the C-terminal domain weakens the Far3p self-interaction. Although the N- and C-terminal coiled-coil mutants reserve some of the interactions with itself and some other Far proteins, both mutants are defective in the pheromone-mediated G1 arrest, indicating that both coiled-coil motifs of Far3p are essential for the integrity and the function of the Far complex.
Introduction
In haploid Saccharomyces cerevisiae cells, mating pheromones initiate intracellular signaling pathways that lead to cell cycle arrest in G1 phase, as well as to the induction of mating-specific genes and cell morphology changes. Binding of the mating pheromone to the G-protein-coupled receptor results in the release of the Gβγ subunits from the inhibitory Gα subunit (Hirschman & Jenness, 1999). The signal then flows from Gβγ via the mitogen-activated protein kinase cascade to transcription factors and other targets (reviewed in Bardwell, 2005). Two well-studied factors that are activated by the signal transduction cascade are Far1p, which is an inhibitor of Cln-CDK and contributes to cell cycle arrest (Chang & Herskowitz, 1992), and Ste12p, a transcription factor that activates the expression of many mating genes (Roberts, 2000).
In addition to Far1p, several other Far proteins have been found to play roles in the pheromone-induced cell cycle arrest. FAR3 was first discovered to be required for the pheromone-induced arrest of the mitotic cell division cycle by a genetic screen (Ferguson, 1994), and the function of Far3p is Far1p independent (Horecka & Sprague, 1996). A yeast two-hybrid screen combined with coimmunoprecipitation (co-IP) assays identified five other Far proteins, Far7–11p, as Far3p-interacting proteins (Kemp & Sprague, 2003). The six Far proteins form a Far complex, and none of them is dispensable for maintaining pheromone-mediated G1 arrest (Kemp & Sprague, 2003). However, the mechanisms of action of these Far proteins remain elusive.
FAR3 and FAR8–10 are predicted to encode one or more coiled-coil motifs, which are generally involved in protein–protein interactions (Kemp & Sprague, 2003). Coiled coil is a ubiquitous protein folding and assembly motif for generating α-helices wrapping around each other, forming a super-coil (Yu, 2002; Parry, 2008). Approximately 20% of proteins in yeast that participate in diverse biological processes are predicted to contain one or more coiled-coil motifs (Barbara, 2007). Sequences capable of forming coiled coils are characterized by a heptad repeat pattern, (abcdefg)n. A ‘knobs-into-holes’ type of packing at the helical interface is responsible for maintaining the coiled-coil conformation, and the ‘knobs’ and ‘holes’ are the side chains at the a and d positions which are often occupied by hydrophobic residues (Yu, 2002; Barbara, 2007). Because of its simplicity and regularity and the large number of known coiled-coil sequences, coiled coils can be easily predicted by computational algorithms (Wolf, 1997). Identification of coiled-coil motifs advances the study of protein–protein interactions because the partners of coiled-coil helices are, by definition, other coiled coils (Newman, 2000).
In this study, we investigated the roles of the Far3p coiled-coil motifs and other domains in the structural and functional organization of the Far complex. We mapped the domains of Far3p involved in the interactions of Far3p with itself and with Far7–11p. Moreover, we studied the roles of the two coiled-coil domains of Far3p in the integrity of the Far complex and in the pheromone-mediated cell cycle arrest.
Materials and methods
Strains, plasmids and media
Yeast strains and plasmids used in this study are listed in Supporting Information, Tables S1 and S2, respectively. The strain with Myc-tagged FAR3 was constructed using a previously described method (Carsten, 2004). FAR9-Myc, FAR10-Myc, far3Δ and pheromone arrest strains were kindly provided by Dr George F. Sprague, Jr (Kemp & Sprague, 2003).
DNA fragments encoding full-length wild-type or mutant Far3p, or truncation mutants of Far3p, Far7p and Far9p, were individually cloned into the EcoRI/BamHI sites of the DNA binding domain (BD) vector pGBKT7 and the activation domain (AD) vector pGADT7 (Clontech). The coding sequence of Far10p was cloned into the BamHI site in the two vectors. pL1803 was constructed by ligating a 1.8-kb SacI–SalI fragment of pSL2784, a kind gift from Dr George F. Sprague, Jr, which contains the entire FAR3 coding sequence, a C-terminal triple hemagglutinin (HA) epitope tag, plus 300 bp of upstream and 870 bp of downstream genomic sequences, into the SacI–SalI-digested pMD18-T vector. pL1804 (far3-cc1) and pL1805 (far3-cc2) were derived from pL1803 by TaKaRa MutanBEST Kit according to the manufacturer's instructions. pL1806 and pL1807 were constructed by inserting the 1.8-kb SacI–SalI fragment of pL1804 and pL1805, respectively, into YEP352. The inserts of all clones were confirmed by DNA sequencing.
Yeast transformation was performed using the lithium acetate method (Ito, 1983). Cells were grown at 30 °C in yeast extract peptone dextrose (YPD), synthetic complete medium (SCM) or dropout medium (SCM minus).
Computational prediction of coiled-coil motifs and dimeric/trimeric coiled coils
The coiled-coil motifs were predicted using the COILS server with a 28 amino acid (aa) window (Lupas, 1991). The multicoil program was used to predict the dimeric/trimeric coiled-coil regions with a 28 aa window (Wolf, 1997).
Yeast two-hybrid assay
Two-hybrid analysis was performed by cotransformation, as described previously (Yu, 2004; Kan, 2008), and/or by mating of yeast cells from two opposite mating types, AH109 and Y187 (Clontech), each containing a bait or a prey plasmid.
Co-IP assay and immunoblotting
Co-IP assay was performed as described previously (Zhang, 2002; Kemp & Sprague, 2003), and immunoblotting was carried out as described previously (Zhang, 2002).
The Far− phenotype assays
The Far− phenotype was assayed using two different methods. The first was performed with strains harboring the pGAL1∷STE4 construct as described previously (Horecka & Sprague, 1996) by streaking log-phase cells onto a control SCM-Ura with dextrose (SCM-Ura/D) plate and an SCM-Ura with raffinose and galactose (SCM-Ura/RG) plate and incubating the plates at 30 °C for 3 days. The second was the pheromone-induced arrest assay performed as described previously (Kemp & Sprague, 2003) by streaking log-phase cells onto a control YPD plate without α-factor and a YPD plate containing 0.1 μg mL−1α-factor, and incubating the plates at 30 °C for 3 days.
Results
Far3p self-associates through its central region and its C-terminal coiled-coil motif
Far3p was predicted by the coils program (Lupas, 1991) to contain two coiled-coil motifs, an N-terminal coiled-coil motif (referred to as CC1; aa 16–43) with four heptad repeats and a C-terminal coiled coil (referred to as CC2; aa 118–173) with eight heptad repeats (Figs 1a and S1a). CC2 was predicted by the multicoil program (Wolf, 1997) to form dimeric and/or trimeric coiled coils (Fig. S1b). As coiled-coil motifs, particularly multimeric coils, are frequently identified as the sites of homotypic protein–protein interactions, we employed yeast two-hybrid assays to determine if Far3p can self-associate and if the coiled-coil motifs are important for the self-interaction. As shown in Fig. 1b, Far3p could indeed interact with itself, as revealed by the yeast cell growth on the SCM-Leu-Trp-His (SCM minus leucine, tryptophan and histidine) plate, indicating expression of the reporter gene HIS3. Far3p self-association was also detected by co-IP from yeast extracts expressing two different versions of the epitope-tagged Far3p, as Far3-Myc and Far3-HA could be coprecipitated by the anti-HA antibody from the yeast extracts of the doubly tagged strain, but not of the negative control FAR3-Myc strain without expressing Far3-HA (Fig. 1c).
The aa 61–100 region and the C-terminal coiled coil mediate the self-association of Far3p. (a) Diagram of the Far3p fragments and coiled-coil domain mutants. Numbers refer to amino acid residues. The filled boxes indicate the predicted coiled coils, CC1 and CC2. The residues mutated to glycine to disrupt the coiled-coil domains in Far3-cc1 and Far3-cc2 are indicated. (b) Yeast two-hybrid analysis of the self-association of Far3p and its fragments and coiled-coil domain mutants. Potential interactions of the proteins were tested in pairwise combinations consisting of a bait plasmid (the AD vector or AD fused to the full-length FAR3, FAR3 fragments or mutants) and a prey plasmid (the BD vector or BD fused to the full-length FAR3, FAR3 fragments or mutants). Tenfold dilutions of yeast cells containing the plasmids were spotted onto SCM-Leu-Trp plates (nonselective for the report gene) and SCM-Leu-Trp-His plates (selective for the report gene) and incubated for 3 days. Results not shown here are summarized in Table S3. (c) Whole cell extracts (WCE) prepared from YL1313 cells (FAR3-Myc FAR3-HA) were precipitated with the control mouse immunoglobulin G (IgG) or anti-HA antibody. WCE from the control strain YL1314 (FAR3-Myc) were precipitated with the anti-HA antibody. Immunoprecipitants were immunoblotted with the anti-HA and anti-Myc antibodies, separately. The asterisk indicates a nonspecific signal from the anti-HA antibody.
To map the domains for the Far3p self-interaction and to determine if the two coiled-coil motifs of Far3p are involved in the Far3p self-interaction, we constructed yeast two-hybrid plasmids to test pairwise interactions of six different Far3p fragments (Far3-F1 to -F6; Fig. 1a) and the full-length Far3p. In addition, we created an N-terminal coiled-coil motif mutant (Far3-cc1) by substituting two leucines at residues 19 and 40 and one isoleucine at residue 33 with glycines, and a C-terminal coiled-coil motif mutant (Far3-cc2) by replacing three leucines at residues 135, 142 and 149 and one isoleucine residue located at residue 163 with glycines (Fig. 1a). All mutations are located at the d positions of the heptad repeats of the coiled-coil motifs. Using the coils and multicoil programs, we predicted that the mutations in Far3-cc1 would disrupt the CC1 structure without affecting CC2 (Fig. S1c and d), and that the mutations in Far3-cc2 would abolish the coiled-coil structure of CC2 without affecting CC1 (Fig. S1e and f).
The results from yeast two-hybrid assays showed that Far3-F1 (aa 1–120), Far3-F5 (aa 101–204), Far3-F6 (aa 61–204), Far3-cc1 and Far3-cc2 could self-interact and interact with the full-length Far3p (Fig. 1b; some data not shown; see summary in Table S3). Among these interactions, Far3-F5 and Far3-cc1 could self-associate almost as strongly as the wild-type Far3p, as judged by the extent of cell growth (Fig. 1b), suggesting that the N-terminal region (aa 1–100) containing CC1 is not involved in the Far3p self-interaction. This is consistent with the results that Far3-F2 (aa 1–100), Far3-F3 (aa 1–60) and Far3-F4 (aa 61–100) did not self-interact and did not interact with other Far3p fragments or the full-length Far3p (data not shown; Table S3). Therefore, the self-interaction of Far3-F1, which was weaker than that of wild-type Far3p (Fig. 1b), was most likely mediated by the central region of Far3p (aa 101–120). This is further supported by the interaction between Far3-F1 and Far3-F5 (data not shown; Table S3), which overlap with each other only in the central region (aa 101–120) of Far3p. The strength of self-interaction of Far3-cc2 was similar to that of Far3-F1 and was weaker than that of wild-type Far3p (Fig. 1b), suggesting that CC2 also mediates the self-interaction of Far3p. Together, these data suggest that both the central region (aa 101–120) and the C-terminal coiled-coil domain of Far3p are involved in the self-association of Far3p.
The aa 61–100 region of Far3p interacts with Far7p, and the Far3p N-terminal coiled-coil domain interacts with Far9p and Far10p
It was reported that Far3p interacts with Far7–11p as determined by yeast two-hybrid and/or co-IP assays (Kemp & Sprague, 2003). We also observed most of these interactions in one or both directions by yeast two-hybrid assays (Figs 1–3; some data not shown; see summary in Table S4). The only reported interactions that were not observed in this study were the Far3p–Far8p and Far9p–Far11p interactions. On the other hand, we detected the interactions between Far3p and Far10p and between Far8p and Far9p which were previously detected only by co-IP (Kemp & Sprague, 2003), and we also identified the self-association of Far9p as well as Far3p which has not been reported previously. Note that a truncated form of Far7p (Far7A; aa 65–222) in the AD vector and a truncated form of Far9p (Far9B; aa 138–572) in both the AD and BD vectors, instead of the full-length proteins, were used for the yeast two-hybrid assays in this and the previous studies (Kemp & Sprague, 2003). This is because the full-length Far7p, but not the truncated Far7A, fused to the AD vector could self-activate the reporter gene expression without an interacting partner (data not shown), and the N-terminus of Far9p would inhibit the interaction between Far9p and Far3p (Kemp & Sprague, 2003).
The aa 61–100 region of Far3p interacts with Far7p, and the Far3p N-terminal coiled-coil domain interacts with Far9p and Far10p. (a) Yeast two-hybrid interactions of Far3p with Far7–10p. Potential interactions of the proteins were tested in pairwise combinations consisting of a bait plasmid (the AD vector or AD fused to FAR genes) and a prey plasmid (the BD vector or BD fused to the FAR3 fragments or mutants). Yeast cell growth was measured as in Fig. 1b. Results not shown here are summarized in Tables S5 and S6. (b, c) Whole cell extracts (WCE) prepared from strains SY4070 (FAR9-Myc FAR3-HA), YL1316 (FAR9-Myc far3-cc1-HA), SY4072 (FAR10-Myc FAR3-HA), YL1318 (FAR10-Myc far3-cc1-HA) were precipitated with the control mouse immunoglobulin G (IgG) or anti-HA antibody. WCE from the control strains SY4069 (FAR9-Myc) and SY4071 (FAR10-Myc) were precipitated with the anti-HA antibody. Immunoprecipitants were immunoblotted with the anti-HA and anti-Myc antibodies, separately. The asterisk indicates a nonspecific signal from the anti-HA antibody. Data for Far9p are in (b), and those for Far10p in (c).
Both Far3p-cc1 and Far3p-cc2 mutants are defective in pheromone-mediated G1 arrest, although they reserve some interactions of Far3p. (a) Yeast two-hybrid interactions of Far3p mutants with Far7–10p, tested using the same method as in Figs 1b and 2a. (b) Cell growth in response to the induction of pGAL1-STE4. Log-phase cells from strains with wild-type (WT) FAR3, far3Δ, far3-cc1 and far3-cc2 in the pGAL1-STE4 background were streaked onto a SCM-Ura with dextrose plate (SCM-Ura/D) and a SCM-Ura with raffinose and galactose plate (SCM-Ura/RG) and incubated at 30°C for 3 days. (c) Cell growth in response to α-factor. Log-phase cells from strains with WT FAR3, far3, far3-cc1 and far3-cc2 were streaked onto a YPD plate (YPD without α-factor) and a YPD plate containing 0.1 μg mL−1α-factor and incubated at 30°C for 3 days.
To map the Far3p domains that interact with Far7–11p, we tested the interactions of different Far3p fragments with Far7–11p using yeast two-hybrid assays. The results show that Far3-F1 (aa 1–120), Far3-F2 (aa 1–100), Far3-F4 (aa 61–100), Far3-F6 (aa 61–204) and full-length Far3p interacted with Far7p, and that Far3-F1, -F2 and -F3 (aa 1–60) interacted with Far9p and Far10p (Fig. 2a; some data not shown; see summary in Tables S5 and S6). These data suggest that the aa 61–100 region of Far3p interacts with Far7p, and an N-terminal region within the first 60 residues of Far3p interacts with Far9p and Far10p.
As CC1 is within Far3-F3 (aa 1–60), which interacts with Far9p and Far10p, we tested if the Far3-cc1 mutant could still interact with Far9p and Far10p. As shown in Fig. 2a, the mutations in CC1 completely abolished the interactions of Far3p with Far9p and Far10p, suggesting that CC1 is essential for the interactions of Far3p with Far9p and Far10p.
To further validate the conclusions from the yeast two-hybrid assays, the interactions of the wild-type Far3p and mutant Far3-cc1 with Far9p and Far10p were examined by co-IP analysis. Far9-Myc (Fig. 2b) and Far10-Myc (Fig. 2c) could be coimmunoprecipitated by the anti-HA antibody that recognizes Far3-HA, as reported previously (Kemp & Sprague, 2003). As negative controls, Far9-Myc and Far10-Myc could not be coimmunoprecipitated by the control immunoglobulin G, or by the anti-HA antibody from extracts prepared from the control strains with untagged FAR3. In contrast to the results from the wild-type Far3-HA, neither Far9-Myc nor Far10-Myc was coimmunoprecipitated with Far3p-cc1-HA by the anti-HA antibody, while the levels of the Far3-cc1-HA mutant protein in the extracts and in the anti-HA immunoprecipitants were comparable to the wild-type Far3-HA (Fig. 2b and c). These results confirmed that the mutations in Far3-cc1 abolished the interactions of Far3p with Far9p and Far10p. Taken together, the yeast two-hybrid and co-IP data indicate that Far3p interacts with Far9p and Far10p through its N-terminal coiled coil.
Both coiled-coil motifs of Far3p are required for the pheromone-mediated G1 arrest
The Far3-cc1 mutant, while it could no longer interact with Far9p or Far10p (Fig. 2a), retained its ability to interact with itself (Fig. 1b) and with Far7A (Fig. 3a). Likewise, Far3-cc2 could still self-associate through its central region, though not as efficiently as the wild-type Far3p (Fig. 1b), and could also interact with Far7p, Far9p and Far10p (Fig. 3a). These results indicate that both mutants at least partially reserved the folding of Far3p and retained some of the Far3p interactions.
As Far3p is required for the pheromone-mediated G1 arrest together with Far7–11p, we examined the biological activities of the Far3-cc1 and Far3-cc2 mutants using two different methods. The first was the Far− phenotype assay in which overexpression of Ste4p causes G1 arrest in wild-type cells, mimicking pheromone treatment (Whiteway, 1990; Horecka & Sprague, 1996). The second method was the pheromone-induced arrest assay, wherein wild-type cells arrest in G1 phase in response to α-factor (Kemp & Sprague, 2003). As expected, cells harboring wild-type FAR3 did not grow in the presence of galactose due to the overexpression of Ste4p in the GAL1-STE4 strain (Fig. 3b) or in the presence of α-factor (Fig. 3c). In contrast, far3Δ, far3-cc1 and far3-cc2 mutant cells did not respond to, and were not arrested by the overexpression of Ste4p in the GAL1-STE4 strain (Fig. 3b) or by the presence of α-factor (Fig. 3c). These results demonstrate that both coiled-coil motifs of Far3p are essential for the function of Far3p in the pheromone-mediated G1 arrest.
Discussion
Initially, we identified Far3p as a potential yeast homolog of geminin, an inhibitor of metazoan DNA replication through binding to the replication-initiation protein Cdt1p (McGarry & Kirschner, 1998; Wohlschlegel, 2000). We found that the C-terminal region of Far3p (aa 122–196) is highly similar in sequence to the Cdt1p-interacting region of geminin (aa 70–152), and Far3p and the human geminin share 29% overall sequence identity and 56% similarity. Furthermore, both proteins are predicted to have a dimeric/trimeric coiled-coil motif at the C-terminal region. We also found that Far3p could interact with Cdt1p and with itself by yeast two-hybrid assays (data not shown), as does geminin in higher eukaryotes (Saxena, 2004). However, overexpression of Far3p did not affect cell cycle progression, and deletion of FAR3 did not result in DNA re-replication (data not shown). It is possible that our experimental conditions did not allow detection of the possible inhibitory effects of Far3p on yeast DNA replication. Therefore, at present it is not certain whether Far3p is a functional homolog of geminin. In this study, we focused on interactions of Far3p with itself and other Far proteins and on the functions of the Far3p coiled coils in the pheromone-induced G1 arrest.
We used yeast two-hybrid assays combined with co-IP to identify and characterize the interactions of the full-length Far3p and its domains with Far3p itself and with other Far proteins. In addition to most of the interactions reported, we detected the interactions between Far3p and Far10p and between Far8p and Far9p that were previously detected only by co-IP, and we also identified the self-association of Far3p and Far9p for the first time. Furthermore, through interaction domain mapping, we demonstrate that the central region and C-terminal coiled coil of Far3p mediate the Far3p self-association, that the N-terminal coiled-coil domain of Far3p interacts with Far9p and Far10p, and that the aa 61–100 region of Far3p interacts with Far7p. Mutation of either coiled-coil motif of Far3p disrupts some but not all of the interactions in the Far complex, and functional analyses show that both coiled coils are essential for the pheromone-mediated G1 arrest. Disruption of the Far3p CC1 abolishes the Far3p interactions with Far9p and Far10p. Mutation of the Far3p CC2 only weakens the Far3p self-interaction without affecting the Far3p interaction with Far7, -9p or -10p. It is possible that the Far3-cc2 mutant cannot maintain a stable Far complex with other Far proteins. Alternatively, the Far3p CC2 may be required for interactions with some proteins that relay the pheromone response signals to and from the Far complex.
Far3p and Far7–11p are predicted to contain one or more coiled-coil motifs, which may be responsible for homo- or heterodimerization. Far3p is present in a complex with >900 kDa in molecular weight (Kemp & Sprague, 2003), yet the sum of the molecular weight of the six Far proteins is ∼600 kDa, suggesting that more than one molecule of some of the Far proteins are present in the Far complex. Our findings of the self-association of Far3p and Far9p support this notion, and our functional studies of the Far3-cc2 mutant demonstrate the importance of the Far3p self-interaction for the function of the Far complex. Although the multicoil program predicted that Far8p and Far10p could also form dimeric coiled coils (Fig. S2), we did not find self-interaction of Far8p or Far10p.
Combining the data from this study and a previous report (Kemp & Sprague, 2003), we depict the interactions of the Far complex in Fig. 4. For simplicity, we assume that Far3p forms a dimer, although Far3p is predicted to form homodimers and/or -trimers. We still do not know the number of molecule(s) of each Far protein in the Far complex, nor do we know the manner in which Far proteins interact. For example, does one molecule of a Far7–11 protein interact with a structure formed by two Far3p molecules in a dimer, or do perhaps two Far3p molecules in a dimer bind two molecules of the same Far7–11 protein?
Model of the interactions among Far proteins based on yeast two-hybrid assays. Thick arrows indicate new interactions identified in this study. Thin arrows indicate interactions of Far7p and Far9p with the Far3p domains identified in this study, but the protein–protein interactions have been reported. Thick dashed arrows indicate the reported interactions that were not observed in this study. Thin dashed arrows indicate the reported interactions that were also observed in this study.
The Far complex is involved in the pheromone-induced cell cycle arrest in a Far1p-independent manner (Horecka & Sprague, 1996). The molecular mechanism of how the Far complex carries out its function is still an open question. Far3p and Far8-10p localize to the endoplasmic reticulum (ER) membrane (Huh, 2003), while Far11p is a member of the highly conserved family of eukaryotic transmembrane proteins (Xiang, 2002). Based on their restricted intracellular distributions and their coiled-coil structures, the Far complex has been proposed to play a role in membrane trafficking (Bonangelino, 2002). We consider it possible that the Far complex is involved in the retention of some important regulators of the START checkpoint, such as the G1 cyclin Cln3p, in the ER, as the HSP70 chaperone Ssa1p does (Verge, 2007). The highly regulated localization of Cln3p assures the precise cell cycle entry. Overexpression of the chaperone increases ER release and nuclear accumulation of Cln3p and promotes premature cell cycle entry from START. This phenomenon is to some extent similar to the Far− phenotypes.
Recent studies on the genome-wise physical and genetic interactions also shed some light on the role of the Far complex. According to a comprehensive study with a synthetic genetic array (Costanzo, 2010), FAR3, FAR7, FAR10 and FAR11 interact genetically with CDC48, which is a chaperone that mediates protein degradation, and with some components of the proteasome, suggesting that the Far complex may be involved in protein degradation at the ER lumen. It is possible that together with Cdc48p and perhaps other proteins, the Far complex controls the degradation of some regulators of the G1 CDK (Fu, 2003). This and other possibilities warrant further investigation.
Authors' contribution
F.L. and R.W. contributed equally to this study.
Acknowledgements
We are very grateful to Dr George F. Sprague, Jr for providing yeast strains and plasmids. Supported by the Hong Kong Research Grants Council (HKUST6430/06M).
References
Supporting Information
Fig. S2. The coiled-coil motif and dimeric/trimeric motif prediction of Far7–11p.
Table S1. Yeast strains used in this study.
Table S2. Plasmids used in this study.
Table S4. Yeast two-hybrid interactions among the Far proteins.
Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
Author notes
Editor: Ian Dawes
