Role of the N-terminal region of Rap1p in the transcriptional activation of glycolytic genes in Saccharomyces cerevisiae

Authors

  • Takayuki Mizuno,

    1. Institute for Biological Resources and Functions, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki 305-8566, Japan
    Search for more papers by this author
  • Tomoko Kishimoto,

    1. Institute for Biological Resources and Functions, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki 305-8566, Japan
    Search for more papers by this author
  • Tomoko Shinzato,

    1. Institute for Biological Resources and Functions, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki 305-8566, Japan
    Search for more papers by this author
  • Robin Haw,

    1. Institute for Biological Resources and Functions, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki 305-8566, Japan
    Current affiliation:
    1. Banting & Best Department of Medical Research, University of Toronto, Toronto, Ontario, M5G 1L6, Canada.
    Search for more papers by this author
  • Alistair Chambers,

    1. Institute of Genetics, University of Nottingham, Queen's Medical Centre, Nottingham NG7 2UH, UK
    Search for more papers by this author
  • Jason Wood,

    1. Cancer Center, Harvard Medical School, Boston, MA 02115, USA
    Search for more papers by this author
  • David Sinclair,

    1. Cancer Center, Harvard Medical School, Boston, MA 02115, USA
    Search for more papers by this author
  • Hiroshi Uemura

    Corresponding author
    1. Institute for Biological Resources and Functions, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki 305-8566, Japan
    • Institute for Biological Resources and Functions, National Institute of Advanced Industrial Science and Technology (AIST), AIST Tsukuba Central 6 Higashi 1-1-1, Tsukuba, Ibaraki, 305-8566, Japan.
    Search for more papers by this author

Abstract

In the yeast two-hybrid system, the N-terminal region of Rap1p was shown to interact with Gcr1p and Gcr2p. Disruption of gcr1 and/or gcr2 in the two-hybrid reporter strain demonstrated that the interaction with Gcr1p does not require Gcr2p, whereas the interaction with Gcr2p is mediated through Gcr1p. Deletion of the N-terminal region of Rap1p alone did not show a growth phenotype, but a growth defect was observed when this mutation was combined with a gcr2 deletion. The poor growth of the gcr1 null mutant was not affected further by the N-terminal deletion of Rap1p, but the growth of gcr1 strains with mutations in the DNA binding region of Gcr1p was affected by the removal of the N-terminal region of Rap1p. These results suggest that one function of the N-terminal region of Rap1p, presumably the BRCT domain, is to facilitate the binding of Gcr1p to the promoter by a protein–protein interaction. Copyright © 2004 John Wiley & Sons, Ltd.

Introduction

Glycolysis is the major route of carbohydrate metabolism in S. cerevisiae and the glycolytic enzyme genes are highly expressed (Fraenkel, 1982). Whether their expression is regulated may be strain-dependent, with reports ranging from virtual constitutivity (Clifton and Fraenkel, 1981; Baker, 1986; Uemura and Fraenkel, 1999) to apparent induction by glucose of at least some of the enzymes (Maitra and Lobo, 1971). Since many of the steps of glycolysis are shared by gluconeogenesis, most glycolytic genes, including those with glucose induction, are also highly expressed in media containing non-fermentable carbon sources. Expression of glycolytic genes is regulated by the concerted action of a number of transcription factors. The core elements of the upstream activating sequence (UAS) of glycolytic enzyme genes consist of the binding sites for Rap1p (RPG-box) and Gcr1p (CT-box) (Huie et al., 1992). Some of the UAS elements also contain binding sites for Abf1p (Brindle et al., 1990; Chambers et al., 1990; Packham et al., 1996) or Reb1p (Chasman et al., 1990; Scott and Baker, 1993; Willett et al., 1993; reviewed by Chambers et al., 1995). However, the binding sites for Abf1p (Buchman et al., 1988) and Reb1p (Morrow et al., 1989), as well as Rap1p (Shore, 1994), are not restricted to glycolytic enzyme genes, as they are known to be multifunctional transcription factors. In contrast, binding sites for Gcr1p are more or less restricted to genes encoding glycolytic enzymes and Ty elements (Turkel et al., 1997; Lopez and Baker, 2000), indicating that Gcr1p plays a major specific role in the regulation of glycolytic gene expression.

gcr1 mutants show severely reduced expression of most glycolytic enzymes at the transcriptional level (Baker, 1986; Clifton and Fraenkel, 1981; Kawasaki and Fraenkel, 1982; Uemura and Fraenkel, 1990). It has been shown that Gcr1p binds to DNA containing the CT-box in vitro (Baker, 1991) and in vivo (Huie et al., 1992). Gcr1p mediates high-level expression of glycolytic genes by interacting with Gcr2p. GCR2 also specifically regulates the expression of most glycolytic genes and the pattern of reduction of most glycolytic enzymes in gcr2 mutants is similar to that in gcr1 mutants (Uemura and Fraenkel, 1990). Uemura and Jigami (1992, 1995) presented genetic evidence for a physical interaction between Gcr2p and Gcr1p and postulated that Gcr2p probably functions as a co-activator in the Gcr1p–Gcr2p complex, presumably by changing the conformation of Gcr1p (Deminoff and Santangelo, 2001).

The observations that the UAS elements of all glycolytic genes contain binding sites for Rap1p and Gcr1p in close vicinity, and that these binding sites display a strong synergism for each other (Chambers et al., 1995), suggest that Rap1p is required for the function of the Gcr1p–Gcr2p complex in glycolytic gene expression. Since Rap1p is an essential transcription factor (Shore and Nasmyth, 1987) and is capable of carrying out many diverse cellular functions, such as activation of a large number of genes (glycolytic genes, ribosomal protein genes and the mating type genes MATα 1 and MATα 2; Huet et al., 1985; Chambers et al., 1989; Giesman et al., 1991), repression of transcription and control of telomere length, depending on the sequence context of its binding site (Shore, 1994), it has been suggested that its function may be determined by interactions with other regulatory proteins. Recent genome-wide immunoprecipitation experiments revealed that Rap1p binds to about 5% of yeast genes, and participates in the activation of 37% of RNA polymerase II initiation events in exponentially growing cells (Lieb et al., 2001).

In vivo binding studies demonstrated that Gcr1p is unable to bind to its binding site in the absence of an appropriately spaced and bound Rap1p binding site (Drazinic et al., 1996) and we demonstrated that oligonucleotides composed of tandem Gcr1p binding sites in the absence of an RPG-box can function as strong UAS elements (Uemura et al., 1997). Thus, the most likely role of Rap1p at glycolytic promoters is to facilitate the binding of Gcr1p (Drazinic et al., 1996; Uemura et al., 1997; Lopez et al., 1998). Induction of local bending and distortion of the DNA helix by the binding of Rap1p (Gilson et al., 1993) might help other regulatory factors to bind DNA (Vignais and Sentenac, 1989). Alternatively, Rap1p might mediate its function(s) by protein–protein interactions (Tornow et al., 1993). However, the interaction interface has not yet been identified.

Rap1p consists of 827 amino acids and can be divided into three regions of approximately equal size. The internal DNA binding domain (DBD) consists of two Myb-like sub-domains (Gehring et al., 1994; Ogata et al., 1994; Konig et al., 1996), essential for cell viability (Henry et al., 1990; Graham et al., 1999). The C-terminal region contains a transcriptional activation domain and a domain involved in telomere function and transcriptional silencing (Kurtz and Shore, 1991; Sussel and Shore, 1991; Hardy et al., 1992a; Hardy et al., 1992b; Moretti et al., 1994; Wotton et al., 1997). The biological function of the N-terminal region is not well understood, because its deletion produced no growth defect in medium containing glucose as a carbon source (Graham et al., 1999). It has been suggested that this region may be involved in the regulation of Rap1p activity, because it is associated with DNA bending (Vignais and Sentenac, 1989; Muller et al., 1994), and sequence analysis revealed that it contains a BRCT domain, which is thought to be important for protein–protein interactions (Callebaut and Mornon, 1997). The schematic diagrams of functional domains of Rap1p, Gcr1p, and Gcr2p are described in Figure 1.

Figure 1.

Functional domains of Rap1p, Gcr1p, and Gcr2p. In Rap1p (a total of 827 residues), the N-terminal domain including the BRCT domain (BRCT), the central DNA binding domain (DBD), and the C-terminal region containing the activation domain (AD) and the silencing domain (arrow) are indicated. The NsiI restriction sites indicate the NsiI–NsiI region which was removed to construct rap1–Δ NsiI. In Gcr1p (785 residues), a leucine zipper like motif (LZ, Deminoff et al., 1995), DNA binding domain (DBD; Huie et al., 1992), and mutation sites in the DBD (mut101 and mut102) are indicated. In Gcr2p (534 residues), the leucin zipper like region (LZ) which is required for the Gcr1p–Gcr2p interaction (Deminoff and Santangelo, 2001) and the essential central Gcr1p homologous region (hatched area; Uemura and Jigami, 1992) are indicated

Recently we cloned RAP1 from Candida glabrata. Sequence comparison showed that BRCT domains are conserved between CgRap1p and ScRap1p, regardless of the low conservation of the N-terminal region (Haw et al., 2001). Although the biological significance of the BRCT domain in Rap1p has not yet been determined, its conservation in Rap1p suggests an important function. In this report we addressed the function of the N-terminus of Rap1p and present genetic evidence for the involvement of the N-terminal region of Rap1p in the interaction with Gcr1p.

Materials and Methods

Strains

Yeast strains used in this study are listed in Table 1. 2845 (Uemura and Fraenkel, 1990) and its isogenic diploid strains C110-1 (Uemura et al., 1992) were used as standard wild-type strains. The heterozygous rap1NsiI (1–17, 342–827) deletion mutant on the 2845 background, YHU3079 [aRAP1/rap1NsiI(1–17, 342–827)], was constructed by integrating the rap1NsiI gene with a URA3 selection marker on plasmid pL1553-1 into the diploid strain C110-1. YHU3079-1D [arap1NsiI(1–17, 341–827)] is one of the segregants of YHU3079. YHU3085 (Δgcr2::KAN) was constructed by replacing GCR2 of 2845 with Δgcr2::KanMX3 module of pL1289-1. A GAL1–lacZ reporter strain SFY526 was used for two-hybrid analysis (Bartel et al., 1993b). Its Δ gcr 1 and Δ gcr 2 derivatives, YHU3086 and YHU3087, were constructed by replacing GCR1 and GCR2 with the Δgcr1::HIS3 module of pL954-33 (Sato et al., 1999) and the Δgcr2::KanMX3 module of pL1289-1, respectively. Likewise, the Δ gcr 1 Δ gcr 2 double disruptant of SFY526, YHU3088, was constructed by subsequent disruption of GCR1 and GCR2. SCR101-F (Haw et al., 2001), a spontaneous 5-fluoro-orotic acid (5-FOA) resistant derivative of SCR101 (Gonçalves et al., 1996) was used as a conditional (UASGALRAP 1) rap1 mutant strain. Its Δ gcr 1 and Δ gcr 2 derivatives, YHU3089 and YHU3090, were constructed by replacing GCR1 and GCR2 with the same Δ gcr 1::HIS 3 module and the Δ gcr 2::KanMX 3 module of pL1291-1, respectively.

Table 1. S. cerevisiae strains used
StrainGenotypeComments or source
  1. The symbols URA3::GAL1–lacZ and rap 1::UASGALL 25RAP 1 represent the insertion of the GAL1–lacZ and UASGAL 1L 25RAP 1 at the ura3 or rap1 locus, respectively. UASGALL 25RAP 1 is a rap1 conditional mutant, which can grow on SC(GAL), but not on SC(Glc) or SC(GL).

2845MATα leu2-3,112 ura3-52 his6Uemura and Fraenkel (1990)
C179-15C-F16MATα leu2-3,112 ura3-52 his6 Δgcr1::ura35-FOA resistant gcr1 disruptant of 2845
Uemura and Jigami (1992)
YHU3085MATα leu2-3,112 ura3-52 his6 Δgcr2::KANgcr2 disruptant of 2845 (this work)
C110-1MATa/MATα leu2-3,112/leu2-3,112 ura3-52/ura3-52 his6/HIS6Uemura and Jigami (1995)
YHU3079MATa/MATα leu2-3,112/leu2-3,112 ura3-52/ura3-52 his6/HIS6 RAP1/rap1NsiI (1–17, 342–827) + URA3Heterologous gcr2 disruptant of C110-1 (this work)
YHU3079-1DMATaleu2-3,112 ura3-52 rap1NsiI (1–17, 342–827) + URA3Segregant of YHU3079 (this work)
YHU3091MATa/MATα leu2-3,112/leu2-3,112 ura3-52/ura3-52 his6/HIS6 rap1NsiI (1–17, 342–827) + URA3/RAP1 Δgcr2::KAN/GCR2Diploid between YHU3079-1D and YHU3085 (this work)
YHU3091-28AMATaleu2-3,112 ura3-52Segregant of YHU3091 (this work)
YHU3091-28BMATα leu2-3,112 ura3-52 rap1NsiI (1–17, 342–827) + URA3Segregant of YHU3091 (this work)
YHU3091-28CMATα leu2-3,112 ura3-52 his6 Δgcr2::KANSegregant of YHU3091 (this work)
YHU3091-28DMATaleu2-3,112 ura3-52 Δgcr2::KAN rap1NsiI (1–17, 342–827) + URA3Segregant of YHU3091 (this work)
SFY526MATaura3-52 his3-200 ade2-101 lys2-801 trp1-901 leu2-3,112 can1 gal4-542 gal80-538 URA3::GAL1-lacZBartel et al. (1993b)
YHU3086MATaura3-52 his3-200 ade2-101 lys2-801 trp1-901 leu2-3,112 can1 gal4-542 gal80-538 URA3::GAL1-lacZ Δgcr1::HIS3gcr1 disruptant of SFY526 (this work)
YHU3087MATaura3-52 his3-200 ade2-101 lys2-801 trp1-901 leu2-3,112 can1 gal4-542 gal80-538 URA3::GAL1-lacZ Δgcr2::KANgcr2 disruptant of SFY526 (this work)
YHU3088MATaura3-52 his3-200 ade2-101 lys2-801 trp1-901 leu2-3,112 can1 gal4-542 gal80-538 URA3::GAL1-lacZ Δgcr1::HIS3 Δgcr2::KANgcr1 and gcr2 disruptant of SFY526 (this work)
SCR101-FMATα trp1-1 leu2-3, 112 ura3-1 his3-11,15 ade2-1rap 1::UASGAL-L25-RAP15-FOA resistant strain of SCR101 (Haw et al.2001)
YHU3089MATα trp1-1 leu2-3,112 ura3-1 his3-11,15 ade2-1rap 1::UASGAL-L25-RAP1 Δgcr1::HIS3gcr1 disruptant of SCR101-F (this work)
YHU3090MATα trp1-1 leu2-3,112 ura3-1 his3-11,15 ade2-1rap 1::UASGAL-L25-RAP1 Δgcr2::KANgcr2 disruptant of SCR101-F (this work)

Media

Yeast cells were grown in rich media (Uemura and Wickner, 1988) and synthetic complete (SC) media (Sherman et al., 1986) or SC drop-out media, depending on the selective pressure required to maintain plasmids. Carbon sources were added to 2% concentration (w/v). When necessary, the respiration inhibitor antimycin A was added to a final concentration of 1 µ g/ml. Geneticin (200 µ g/ml in rich medium) was used for the selection with KanMX3 gene replacement. For β-galactosidase detection, X-gal (5-bromo-4-chloro-3-indolyl-β-D-galactoside, 40 µ g/ml) was added to SC plates, which were buffered to pH 6.8 with 0.1 M potassium phosphate buffer.

Plasmids

Standard techniques of DNA manipulation used in this study are described in Sambrook et al. (1989). Plasmids used in this study are listed in Table 2. An integration type plasmid, pL1553-1[rap1NsiI + URA3], was used to delete the N-terminal [amino acids (aa) 18–341)] region of Rap1p. It was constructed by inserting the rap1NsiI gene (SalI fragment of pAJ-T3 [rap1NsiI in pRS415], into the PvuII–PvuII gap of pAJ90, in which the C-terminal domain of RAP1 is inserted upstream of the URA3 gene (Graham et al., 1999). pL1659-2 contains RAP1 and YNL215W on pRS315 (YCp-LEU2 vector; Sikorsky and Hieter, 1989). Its deletion of YNL215W (XbaI–XhoI fragment) formed pL1720-724 [RAP1]. pL1722-9 [rap1NsiI YNL215W] and pL1721-196 [rap1NsiI] were constructed by removing an 1.0 Kb NsiI fragment [N-terminal (aa 18–341) region of Rap1p containing BRCT domain] from pL1659-2 and pL1720-724 respectively. pL1725-2 [RAP1] and pL1727-5 [rap1NsiI] are pRS316 (YCp-URA3 vector; Sikorsky and Hieter, 1989) versions of pL1659-2 and pL1722-9, respectively.

Table 2. Plasmids used
PlasmidCharacteristicsRap1p regionaType of vectorComments or source
  • a

    The symbols, N, DBD, and C indicate N-terminal (aa 1–341), DNA binding (aa 342–596), and C-terminal (aa 597–827) regions of Rap1p. The plasmids containing DBD + C and DBD were constructed by removing the NsiI–NsiI region of Rap1p from the Full and N + DBD constructs of Rap1p, respectively.

  • b

    pRS315 and pRS316 are YCp-LEU2 and YCp-URA3 vectors, respectively (Sikorsky and Hieter, 1989).

pL1720-724Rap1p(1-827)FullLEU2 cen (pRS315b )This work
pL1721-196Rap1p(1-17, 342-827)DBD + CLEU2 cen (pRS315)This work
pL1725-2Rap1p(1-827)FullURA3 cen (pRS316b )This work
pL1727-5Rap1p(1-17, 342-827)DBD + CURA3 cen (pRS316)This work
pL1729-10GBD/Rap1p(1-827)FullTRP1 2µ m (pGBT9)This work
pL1342-2GBD/Rap1p(1-596)N + DBDTRP1 2µ m (pGBT9)This work
pL1344-1GBD/Rap1p(597-827)CTRP1 2µ m (pGBT9)This work
pL1677-8GBD/Rap1p(1-17, 342-596)DBDTRP1 2µ m (pGBT9)This work
pL1810-4GBD/Rap1p(1-17, 342-827)DBD + CTRP1 2µ m (pGBT9)This work
pML77-8GAD/Gcr1p(9-785) LEU2 2µ m (pCTC13)Uemura and Jigami (1992)
pJW101GAD/Gcr2p(248-534) LEU2 2µ m (pGAD424)This work
pL1040-11Gcr1p(1-785) LEU2 2µ m (YEp351b )This work
pL1711-2Gcr1p(1-535) LEU2 2µ m (YEp351)This work
pD2005-1Gcr1p-mut101(1-779) LEU2 2µ m (YEp351)This work
pD2006-1Gcr1p-mut102(1-686) LEU2 2µ m (YEp351)This work
pGBT9GBD vector TRP1 2µ mBartel et al. (1993a)
pGAD2FGAD vector LEU2 2µ mChien et al. (1991)
pGAD424GAD vector LEU2 2µ mBartel et al. (1993a)

pL1332-8 contains the entire region of RAP1, together with an engineered EcoRI site upstream of RAP1 open reading frame (ORF). It was constructed by inserting a HindIII fragment containing RAP1-DBD (+1079 to +2073) of pAJ930 (contains the RAP1 gene with novel SphI (+1601) and SmaI (+1794) sites; Graham et al., 1999) into the HindIII gap of RAP1 in pPE711 (contains the RAP1 gene with an engineered EcoRI site upstream of ORF in pSP46) (Chambers et al., 1989). pL1343-5 contains only the C-terminal region of RAP1. It was constructed by inserting a SmaI–SalI fragment of pL1332-8 into the SmaI–SalI gap of pSP73 (Promega). pL1342-2 [GBD/Rap1p(1-596)] and pL1344-1 [GBD/Rap1p (597-827)] were used for two-hybrid analysis and were constructed by inserting an EcoRI–SmaI fragment of pL1332-8 or pL1343-5 into EcoRI–SmaI gap of pGBT9 (GBD vector; Bartel et al., 1993a), respectively. pL1729-10 [GBD/Rap1p(1-827)] was constructed by inserting a StuI–PvuII fragment containing the C-terminal region of RAP1 into the StuI–SmaI gap of pL1342-2 [GBD/Rap1p(1-596)]. pL1677-8 [GBD/Rap1p(1-17, 342-596)] and pL1810-4 [GBD/Rap1p(1-17, 342-827)] were constructed from pL1342-2 and pL1729-10 by removing an 1.0 kb NsiI–NsiI fragment (N-terminal region of Rap1p), respectively.

pML77-8 [GAD (Gal4p activation domain)/Gcr1p(9–785)] was constructed as described (Uemura and Jigami, 1992). pJW101 (GAD/Gcr2p (248–534) was constructed by fusing Gcr2p(248–534) to GAD on pGAD424 (Bartel et al., 1993a). The grc2::KanMX3 plasmids pL1289-1 and pL1291-1 were constructed by inserting the KanMX3 module from pFA6a–lacZ–KanMX3 (Wach et al., 1994) into SacI (+351)–BamHI (+1192) gap or HpaI (+139)–BglII (+1134) gap in the ORF of GCR2 on pL1277-7 [GCR2 in pT7/T3α-19 (Invitrogen)], respectively. pL1040-11 (GCR1 on YEp351) was constructed by inserting a SalI–SmaI fragment containing GCR1 into HindIII–SmaI MCS of pL997-1, a derivative of YEp351 (Hill et al., 1986) with a disruption of the SacI site. pHU57001 contains a ENO1–lacZ fusion gene in YCp-URA3 vector (Uemura et al., 1987).

PCR-based mutagenesis

The DNA binding domain (DBD) mutation library of GCR1 was constructed by PCR-based random mutagenesis of DBD by using the Diversify PCR mutagenesis kit (Clontech). Only the DBD of GCR1 was PCR-amplified by using pL1040-11 (GCR1 on YEp351) as a template and oligonucleotide primers designated GCR1–SphI-1 (5′-CCACTACTAGCCTGAAAGTCCCAC-3′, forward) and YEp351-1 primer (5′-TCGCGGTACCCGGGGATCCTCTAGAGTC-3′, reverse). The region containing DBD of Gcr1p was excised as a SphI (at codon 535 of Gcr1p)–BamHI (downstream of Gcr1p coding region) fragment and it was replaced with the wild-type DBD of pL1040-11.

Yeast manipulation

Matings, diploid selection, sporulation and dissection were carried out by the usual methods (Sherman et al., 1986). Yeast cells were transformed by the method of Ito et al. (1983). Plasmid DNA was isolated from yeast by the method of Hoffman and Winston (1987). Growth on plates was scored by measurements of colony size of around 30 colonies under a microscope fitted with an ocular micrometer. The error range is approximately 20%. Growth in liquid medium was monitored at OD660nm using an automatic detector (Bio-plotter, Toyo-Sokki, Japan).

Enzyme assays

For β-galactosidase assays, transformants were grown to mid-log phase in synthetic drop-out medium containing 2% each of glycerol and lactate, and the activity of β-galactosidase was assayed as described previously (Uemura and Jigami, 1992). Units are defined as Miller units (milliunits per milliliter per unit of optical density at 600 nm; mU/ml OD600). For glycolytic enzyme assays, transformants were grown to mid-log phase in rich medium containing 2% each of glycerol and lactate. Cell extracts were prepared by vortexing cells with glass beads for 2 min (four times for 30 s each time) at 4 °C and assayed as described previously (Uemura and Fraenkel, 1990).

Preparation of polyA RNA from yeast

Yeast strains were grown in rich medium containing 2% each of glycerol and lactate to OD600 = 0.5. Total RNA was prepared as described (Mizuno and Harashima, 2003) after resuspending cells in lysis buffer (0.2 M Tris–HCl, pH 7.5, 0.5 M NaCl, 10 mM EDTA, 1% w/v SDS). PolyA RNA was further purified by using Oligotex-dT30 mRNA purification kit (Takara-Bio, Japan).

RT–PCR

RT–PCR analysis was performed by using SuperScript One-Step RT–PCR System with Platinum Taq DNA polymerase (Invitrogen). RT–PCR reactions were performed in a 15 µ l volume containing 0.2 mM deoxynucleotide triphosphate, 0.2 µM each of sense and anti-sense primers, 0.3 µ l RT/Taq mixture and 2 ng polyA RNA. Fragments were amplified by incubation at 50 °C for 30 min (reverse-transcription), 94 °C for 2 min (pre-treatment), n cycles of 94 °C for 30 s/55 °C for 30 s/72 °C for 90 s, and an additional incubation at 72 °C for 5 min (extension). Amplified fragments were run on 1% agarose gels and the intensity of each band was examined. For the amplification of the 1.2 kb ENO1/2 fragment (primers can hybridize with both ENO1 and ENO2 transcripts), primers corresponding to +48 to +75 (5′-TAACCCAACCGTCGAAGTCGAATTAACC-3′) and +1248 to +1221 (5′-GATTCTCAACAATTGGTTCAATTTAGCC-3′) of ENO1 were used. Likewise, a primer set 5′-AGGTTGCTGCTTTGGTTATTGATAACGG-3′ (+11 to +38) and 5′-AGCCAAGATAGAACCACCAATCCAGACG-3′ (+1031 to +1004) was used for the amplification of a 1.0 kb ACT1 fragment.

Western blotting

Yeast strains were grown in 8 ml SC(Glc)-Leu to OD600 = 1.0. The same amount of cells were suspended in 200 µ l sample buffer (0.06 M Tris–HCl pH 6.8, 10% v/v-glycerol, 2% w/v SDS, 5% v/v 2-mercaptoethanol, 0.0025% w/v bromophenol blue; (Horvath and Riezman, 1994) and cells were lysed by incubating at 96 °C for 3 min. After centrifugation at 15 000 rpm for 3 min, 10 µ l lysates were applied onto 10% SDS–polyacrylamide gel. Separated proteins were electrotransferred to PVDF membrane (Immobilon; Millipore Corporation, USA) and probed with the anti-Rap1 goat serum (yC-19; Santa Cruz Biotechnology Inc., USA). After washing, the membrane was probed with an HRP-conjugated bovine anti-goat antibody, and the HRP signals were detected using an ECL advanced chemiluminescence system (Amersham Biosciences Corp. USA).

Results

The N-terminal region of Rap1p interacts with Gcr1p in two-hybrid analysis

Tornow et al. (1993) showed a protein–protein interaction between Rap1p and Gcr1p by co-immunoprecipitation. To identify the interaction interface between Rap1p and the Gcr1p–Gcr2p complex, we performed two-hybrid analysis using various regions of Rap1p and Gcr1p or Gcr2p. The entire coding region of Rap1p (full or aa 1–827), N-terminus and DBD of Rap1p (N + DBD or aa 1–596) and C-terminal region of Rap1p (C or aa 597–827) were fused to the Gal4p DNA binding domain (GBD) in plasmids pL1729-10, pL1342-2 and pL1344-1, respectively. When these plasmids were introduced into the reporter strain SFY526, transformants of GBD/Rap1p (full) [pL1729-10] and GBD/Rap1p(N + DBD) [pL1342-2] formed white colonies on X-gal-containing plates, while transformants containing GBD/Rap1p(C) [pL1344-1] formed blue colonies. This result is consistent with the previous report that the C-terminus of Rap1p contains a domain involved in transcriptional activation (Hardy et al., 1992a).

We then used GBD/Rap1p (full) [pL1729-10] and GBD/Rap1p(N + DBD) [pL1342-2], neither of which activates alone, for two-hybrid analyses. To address the interaction between Rap1p and Gcr1p or Gcr2p, pML77-8 [GAD/Gcr1p(9–785)], pJW101 [GAD/Gcr2p(248–534)] and pGAD2F [GAD vector] were introduced into SFY526 harbouring GBD/Rap1p(full) or GBD/Rap1p(N + DBD). Since the N-terminal region of Gcr2p is not required for normal growth and the N-terminal deleted Gcr2p can interact with Gcr1p (Uemura and Jigami, 1992; Deminoff and Santangelo, 2001), we used pJW101 [GAD/Gcr2p(248–534)] to express the GAD–Gcr2p fusion. The β-galactosidase activity of SFY526 harbouring GBD/Rap1p (full) [pL1729-10] and GAD/Gcr1p(9–785) [pML77-8] averaged 20.7 units (Table 3, line 1). Since the growth of Δ gcr 1 mutants on glucose is very poor, we used glycerol plus lactate for the analysis, but it should be noted that glycolytic enzyme genes are highly expressed under both Glc and GL growth conditions (Clifton and Fraenkel, 1981; Baker, 1986; Uemura and Fraenkel, 1999). The Rap1p with the C-terminal deletion [GBD/Rap1p(N + DBD)] also gave a similar level of β-galactosidase activity (19.1 units; Table 3, line 3), suggesting that Rap1p (N + DBD) contains the region that interacts with Gcr1p. When the GAD–Gcr2p fusion plasmid (pJW101) was used, only a very weak interaction with the entire Rap1p was observed, but a clear two-hybrid signal was observed in combination with the GBD/Rap1p (N + DBD) plasmid (pL1342-2) (20.4 units; Table 3, line 3). The reason for this is not clear, but it is probably due to an indirect interaction between Gcr2p and Rap1p (see below).

Table 3. Interaction between Rap1p and Gcr1p or Gcr2p in two-hybrid analysis
LineHost strainGBD fusion plasmidβ-Galactosidase activity (mU/OD600/ml)b
NameRelevant GCR genotypeNameRap1p regionapML77-8 GAD/ Gcr1p (9–785)pJW101 GAD/ Gcr2p (248–534)pGAD2F GAD vector
  • a

    The symbols, N, DBD, and C indicate N-terminal (aa 1–341), DNA binding (aa 342–596), and C-terminal (aa 597–827) regions of Rap1p fused to GBD (Gal4 DNA binding domain) for two-hybrid analysis. The plasmids containing DBD + C and DBD were constructed by removing the NsiI–NsiI region of Rap1p from the Full and N + DBD constructs of Rap1p, respectively.

  • b

    Cells were grown in SC(GL) (2% lactate and 2% glycerol) medium at 30 °C to a mid-log phase. Values of β-galactosidase activity (mU/ml/OD600) are an average of at least three independent measurements.

1SFY526WTpL1729-10Full20.7 ± 5.70.98 ± 0.020.15 ± 0.17
2SFY526WTpL1810-4DBD + C<0.10<0.10<0.10
3SFY526WTpL1342-2N + DBD19.1 ± 0.320.4 ± 3.90.15 ± 0.20
4SFY526WTpL1677-8DBD0.15 ± 0.010.28 ± 0.040.28 ± 0.04
5YHU3086Δgcr1pL1342-2N + DBD6.3 ± 1.70.14 ± 0.080.19 ± 0.03
6YHU3087Δgcr2pL1342-2N + DBD39.6 ± 6.831.3 ± 8.6<0.10
7YHU3088Δgcr1 Δ gcr2pL1342-2N + DBD4.0 ± 0.20.14 ± 0.040.60 ± 0.02

To delimit the interaction region further, the N-terminal NsiI–NsiI fragment (aa 18–341) of Rap1p was removed from pL1342-2 [GBD/Rap1p (N + DBD)] to construct pL1677-8 [GBD/Rap1p (DBD or 1–17, 342–596)]. The interaction between Rap1p and Gcr1p or Gcr2p was completely abolished (Table 3, cf. lines 3 and 4), suggesting that the N-terminal region of Rap1p is required for the interaction between Rap1p and Gcr1p. Since the removal of the same NsiI–NsiI region from the full-length Rap1p, pL1810-4 [GBD/Rap1p (DBD + C)], also abolished the interaction (Table 3, cf. lines 1 and 2), the loss of signal was probably not due to the instability of the truncated Rap1p product (see the next section).

To investigate if both Gcr1p and Gcr2p are required for the observed interactions involving Rap1p, we performed similar experiments in Δ gcr 1 (YHU3086), Δ gcr 2 (YHU3087) and Δ gcr 1 Δ gcr 2 double (YHU3088) disruptant strains. For the interaction between GBD/Rap1p (N + DBD) and GAD/Gcr1p, similar levels of β-galactosidase activities were observed in YHU3086 (Δ gcr 1), YHU3087 (Δ gcr 2) and YHU3088 (Δ gcr 1 Δ gcr 2) (Table 3, lines 3, 5, 6, and 7 in the columns of GAD/Gcr1p). However, the β-galactosidase signals generated by the interaction between GBD/Rap1p (N + DBD) and GAD/Gcr2p were abolished specifically in Δ gcr 1 background strains YHU3086 (Δ gcr 1) and YHU3088 (Δ gcr 1 Δ gcr 2) (Table 3, cf. line 3 with lines 5, 6 and 7 in the columns of GAD/Gcr2p). These results indicated that the interaction between Rap1p and Gcr1p did not require Gcr2p, but the interaction between Rap1p and Gcr2p was mediated via endogenous Gcr1p.

Deletion of the N-terminal region of Rap1p affected the growth of the gcr2 mutant

Tornow et al. (1993) showed biochemical evidence for the interaction between Rap1p and Gcr1p by co-immunoprecipitation experiment. Our two-hybrid experiments suggested that the N-terminal region of Rap1p is important for this interaction. To verify the biological significance of this region, we examined the effect of deletion of the N-terminal region of Rap1p on cell growth. An NsiI–NsiI fragment in the N-terminus of RAP1 was eliminated from wild-type RAP1 to construct the N-terminus deleted rap1 gene [rap1NsiI (aa 1–17, 342–827)] (see Materials and methods). The rap1NsiI gene with a URA3 selectable marker was introduced into the diploid strain C110-1, and the resultant diploid strain YHU3079 (RAP1/rap1NsiI with a URA3 marker) was dissected after sporulation. The Ura+ phenotype segregated 2+:2 and all segregants (both wild-type and rap1NsiI strains) grew normally on SC(Glc), SC(GL), and SC(Glc) containing antimycin A plates at 30 °C (data not shown).

We previously reported the interaction between Gcr1p and Gcr2p, and speculated that the function of Gcr2p is to provide an activation domain to the Gcr1p–Gcr2p complex (Uemura and Jigami, 1992). In contrast to the very poor growth of Δ gcr 1 mutants on glucose, the growth of Δ gcr 2 mutants is almost normal on glucose plates at 30 °C (semi-permissive condition), and more severe conditions [glucose containing antimycin A (1 µ g/ml) at 37 °C (non-permissive condition)] are required to inhibit the growth of the Δ gcr 2 mutant (see below; Uemura and Fraenkel, 1990). Based on the growth phenotypes of Δ gcr 1 and Δ gcr 2 strains described above, we postulated that we might be able to observe a synthetic growth defect in a rap1NsiI and Δ gcr 2 double mutant. One of the rap1NsiI strains, YHU3079-1D, was crossed with YHU3085 (Δ gcr 2::KAN) and the resultant diploid strain was dissected after sporulation to construct the double mutant of rap1NsiI and Δ gcr 2. Tetrad analysis showed a 2+:2 segregation of both Ura+ and geneticin resistance (KAN) phenotypes. A set of tetrads with a tetra-type segregation, YHU3091-28A (WT), YHU3091-28B (rap1NsiI), YHU3091-28C (Δ gcr 2) and YHU3091-28D (Δ gcr 2rap 1 − Δ NsiI), were examined for their growth phenotypes. The growth of the Δ gcr 2 single mutant (YHU3091-28C) was normal on SC(GL) and slightly affected on SC(G1c) (Table 4, line 3). In contrast, the rap1NsiI single mutant strain (YHU3091-28B) grew normally on both SC(GL) and SC(Glc) plates (Table 4, line 2), consistent with a previous report (Graham et al.1999). However, the growth of the Δ gcr 2rap 1 − Δ NsiI double mutant (YHU3091-28D) was drastically impaired on SC(Glc) and a growth defect was also observed on SC(GL) plates (Table 4, line 4). The slow growth phenotype of the Δ gcr 2rap 1 − Δ NsiI double mutant (YHU3091-28D) was not due to the instability of the truncated Rap1p protein, because almost the same amount of intact Rap1p (expected size of 92 kDa) and the Rap1 lacking the N-terminal region (Rap1p-Δ NsiI, expected size of 58 kDa) were detected in both GCR2+ and Δ gcr 2 strain backgrounds (Figure 2, lanes 1–4).

Figure 2.

Western blot detection of the full length Rap1p and the N-terminal deleted Rap1p. Yeast strains were grown in SC(Glc) or SC(Glc)-Leu to OD600 = 1.0. Same amount of cells were suspended in lysis buffer and 10 µ l lysates were applied onto 10% SDS–polyacrylamide gel. Separated proteins were electrotransferred to PVDF membrane and the Rap1p proteins were detected by using ECL advance chemiluminescence system. Lane 1, YHU3091-28A (WT); lane 2, YHU3091-28B (rap1NsiI); lane 3, YHU3091-28C (Δ gcr2); lane 4, YHU3091-28D (Δ gcr2 rap1NsiI); lane 5, SCR101-F [UASGALRAP 1] with pL1720-724 (RAP1); lane 6, SCR101-F with pL1721-196 (rap1NsiI); lane 7, YHU3090 [Δ gcr2 UASGALRAP 1] with pL1720-724 (RAP1); lane 8, YHU3090 with pL1721-196 (rap1NsiI)

Table 4. Growth and glycolytic enzyme activities of rap1 mutant strains in which the N-terminal region of Rap1p is deleted
LineStrainGenotypeaPlasmidRAP1geneaGrowth on plateb[colony size (mm)]Generation timec(min)Relative enzyme activityd
SC(Glc)SC(GL)GPMENOPYKZWF
  • a

    rap1NsiI indicate the RAP1 gene in which the NsiI region (aa 18–341) of Rap1p is deleted.

  • b

    Cells were streaked on SC(Glc) or SC(GL) plates, and average colony size (mm) was measured after incubation at 30 °C for 2 or 4 days, respectively.

  • c

    Generation times were determined by growing cells in SC(Glc) at 30 °C.

  • d

    The strains were grown to mid-log phase in SC(GL). Specific activities (in µM NAD or NADH formed (product)/min/mg protein) for YHU3091-28A are 1.9 for GPM, 1.4 for ENO, 1.3 for PYK and 0.9 for ZWF.

 1YHU3091-28AGCR2 RAP1None 0.60.6 (1.00)(1.00)(1.00)(1.00)
 2YHU3091-28BGCR2 rap1–Δ NsiINone 0.60.6 1.43 ± 0.301.22 ± 0.161.51 ± 0.081.12 ± 0.26
 3YHU3091-28CΔ gcr 2RAP 1None 0.40.6 0.16 ± 0.040.19 ± 0.090.36 ± 0.341.60 ± 0.24
 4YHU3091-28DΔ gcr 2rap 1–Δ NsiINone 0.20.3 0.12 ± 0.020.10 ± 0.040.17 ± 0.101.56 ± 0.29
 5YHU3091-28BGCR 2rap 1–Δ NsiIpL1720-724RAP10.50.6142 ± 4    
 6YHU3091-28BGCR 2rap 1–Δ NsiIpL1721-196rap1–Δ NsiI0.50.6147 ± 3    
 7YHU3091-28BGCR 2rap 1–Δ NsiIpRS315Vector0.40.6166 ± 1    
 8YHU3091-28DΔ gcr 2rap 1–Δ NsiIpL1720-724RAP10.40.7200 ± 13    
 9YHU3091-28DΔ gcr 2rap 1–Δ NsiIpL1721-196rap1–Δ NsiI0.20.4257 ± 20    
10YHU3091-28DΔ gcr 2rap 1–Δ NsiIpRS315Vector0.20.4267 ± 15    

To confirm the importance of the N-terminal region of Rap1p for growth of the gcr2 mutant, pL1720-724 (RAP1), pL1721-196 (rap1NsiI) and pRS315 (empty vector) plasmids were introduced into YHU3091-28B (rap1NsiI) and YHU3091-28D (Δ gcr 2rap 1 − Δ NsiI) strains. Judging from the colony size on the plates, the growth of rap1NsiI strain YHU3091-28B was not affected by the transformation of RAP1 regardless of the presence of the N-terminal region of Rap1p (Table 4, lines 5 and 6). On the other hand, for the Δ gcr 2 and rap1NsiI double mutant strain YHU3091-28D, transformation of rap1NsiI (pL1721-196) or an empty vector (pRS315) did not rescue growth (Table 4, lines 9 and 10), while transformation of wild-type RAP1 (pL1720-724) restored its growth to the same level as the Δ gcr 2 mutant strain (YHU3091-28C) (Table 4, lines 3 and 10). Monitoring of growth in liquid media confirmed these results (Figure 3A. cf. open and closed diamonds). The generation times of the Δ gcr 2 and rap1NsiI double mutant strain harbouring pL1720-724 (RAP1), pL1721-196 (rap1NsiI), or pRS315 (empty vector) are 200, 257 and 267 min, respectively (Table 4), indicating the biological significance of the N-terminal region of Rap1p.

Figure 3.

Growth of transformants. Cells were grown in SC(Glc)-Leu medium at 30 °C and the turbidity of the cells was monitored at 660 nm by an automatic detector (Bio-Plotter, Toyo-Sokki, Japan). (A) Slow growth phenotype of YHU3091-28D (Δ gcr 2rap 1 − Δ NsiI) was restored by the wild-type RAP1 but not by the N-terminal truncated rap1. Symbols: open diamond, pL1720-724 (RAP1); closed diamond, pL1721-196 (rap1NsiI); open circle, pRS315 (empty vector). (B) Effect of N-terminal deletion of Rap1p in the rap1 conditional mutants. Symbols: open triangle, SCR101-F (UASGALRAP 1) with pL1720-724 (RAP1); closed triangle, SCR101-F with pL1721-196 (rap1NsiI); open square, YHU3090 (Δ gcr 2UASGALRAP 1) with pL1720-724 (RAP1); closed square, YHU3090 with pL1721-196 (rap1NsiI)

Activities of several glycolytic enzymes (enolase [ENO], phosphoglycerate mutase [GPM], pyruvate kinase[PYK]), and glucose 6-phosphate dehydrogenase [ZWF] (non-glycolytic enzyme used as a control) were measured. The rap1NsiI single mutation hardly affected the enzyme activities (Table 4, line 2). However, the combination of rap1NsiI with gcr2 mutation further decreased the glycolytic enzyme activities compared to those in the gcr2 single mutant (Table 4, cf. line 3 with line 4). The transcriptional effect of rap1NsiI in the gcr2 mutant was further confirmed by measuring the amount of ENO1 transcript in these mutants by RT–PCR. Consistent with the enzyme activities, ENO1/2 expression was more drastically affected in the double mutant of rap1NsiI and gcr2 than that in the single mutant of gcr2. On the other hand, the single mutation of rap1NsiI slightly increased the ENO1/2 expression (Figure 4).

Figure 4.

Expression of ENO1/2 in rap1NsiI, Δ gcr 2, and Δ gcr 2rap 1 − Δ NsiI mutants. Levels of ENO1/2 and ACT1 mRNA in YHU3091-28A (WT), YHU3091-28B (rapl1NsiI), YHU3091-28C (Δ gcr 2) and YHU3091-28D (Δ gcr 2rap 1 − Δ NsiI) strains were examined by RT–PCR. mRNA was prepared from the cells grown in YPAGL to OD600 = 0.5. ENO1/2 fragment of 1.2 kb and ACT1 fragment of 1.0 kb were amplified by RT–PCR and analysed by 1% agarose gels as described in Materials and methods. Cycles of PCR were fixed to 18, 21 and 24 for ENO and ACT1 amplification with 2 ng poly A RNA

To exclude the possibility that the apparent influence of the N-terminal region of Rap1p was an artifact due to the genetic background of the strain, the rap1 conditional mutant strain SCR101-F, in which RAP1 is expressed under the control of the GAL1 promoter (Haw et al.2001; Gonçalves et al. 1996), was used. Since RAP1 is an essential gene, SCR101-F is able to grow on galactose plates (permissive condition) but not on glucose or glycerol + lactate plates (non-permissive conditions). To investigate whether the N-terminal deletion of Rap1p affects growth in the Δ gcr 2 background, pL1720-724 (RAP1), pL1721-196 (rap1NsiI) and pRS315 (empty vector) plasmids were introduced into the conditional rap1 mutant SCR101-F and its Δ gcr 2 derivative YHU3090. The rap1NsiI gene (pL1721-196) worked, as well as the wild-type RAP1 (pL1720-724), and no difference in growth was observed in the GCR 2+rap1 conditional mutant strain SCR101-F, either on plates (Table 5, lines 1 and 2) or in liquid media (Figure 3B, see triangles). On the other hand, in the Δ gcr 2 background (YHU3090) the strain harbouring pL1720-724 (RAP1) grew faster than that harbouring pL1721-196 (rap1NsiI) on glucose plates (Table 5, lines 7 and 8). Their generation times were 198 and 242 minutes, respectively (Table 5, lines 7 and 8; and Figure 3B, see squares). The growth effects were not due to the instability of the truncated protein because almost the same amount of intact and the N-terminal region truncated Rap1 proteins were detected in both GCR 2+ and Δ gcr 2 background strains (Figure 2, lanes 5–8). Thus, the importance of the N-terminal region of Rap1p was also demonstrated in the RAP1 conditional mutants.

Table 5. Effect of N-terminal deletion of Rap1p in the rap1 conditional (UASGALRAP 1) mutant strains
LineHost strainPlasmidGrowth on plate [colony size (mm)]bGeneration timec(min) in SC(Glc)
    30 °C37 °C
NameGCRgenotypeNameRAP1geneaSC(Glc) (2 days)SC(GL) (4 days)SC(Glc) +Anti (2 days)SC(Glc) (2 days)SC(Glc) +anti (2 days)
  • a

    As described in Table 4.

  • b

    Cells were streaked on SC(Glc), SC(GL) or SC(Glc) + anti (antimycin A) plates, and average colony size (mm) were measured after incubation at 30 °C or 37 °C. Colony size in the parentheses are measured after 7 day incubation. Symbols (−) indicate no growth was observed.

  • c

    As described in Table 4.

1SCR101FWTpL1720-724RAP10.651.20.550.60.6124 ± 5
2SCR101FWTpL1721-196rap1–Δ NsiI0.651.20.550.60.6126 ± 9
3SCR101FWTpRS315Vector 
4YHU3089Δ gcr 1pL1720-724RAP1<0.05 (0.4)0.4<0.05<0.05<0.051200 ± 40
5YHU3089Δ gcr 1pL1721-196rap1–Δ NsiI<0.05 (0.4)0.4<0.05<0.05<0.051170 ± 40
6YHU3089Δ gcr 1pRS315Vector 
7YHU3090Δ gcr 2pL1720-724RAP10.450.60.30.5<0.05198 ± 14
8YHU3090Δ gcr 2pL1721-196rap1–Δ NsiI0.250.60.150.2<0.05242 ± 10
9YHU3090Δ gcr 2pRS315Vector 

To examine whether the N-terminal deletion of Rap1 also affects the growth of the Δ gcr 1 mutant, we introduced pL1720-724 (RAP1) and pL1721-196 (rap1NsiI) into the Δ gcr 1 version of the rap1 conditional mutant YHU3089. Since the growth of the gcr1 single mutant is already very poor on glucose, substantial growth was not observed in 2 days. No difference was observed between pL1720-724 (RAP1) and pL1721-196 (rap1NsiI) even after a prolonged incubation (Table 5, lines 4 and 5; see growth after 7 days incubation on glucose). No difference was observed in their doubling times in liquid culture, either (data not shown). This result agreed well with our two-hybrid result; interaction with Gcr1p is direct, but the interaction with Gcr2p is mediated by Gcr1p. We suggest that an additional effect was not observed in the Δ gcr 1 mutant, because Gcr1p, the direct target of Rap1p, was absent, but that an additional effect was observed in the Δ gcr 2 mutant, because intact Gcr1p exists in the Δ gcr 2 mutant.

Growth of the strains with mutations in the DNA binding region of GCR1 requires the N-terminal region of Rap1p

Based on the observations described above, we suggest that not only the activation domain of Rap1p, but also its N-terminal region, are involved in the recruitment of Gcr1p. But how does the N-terminus of Rap1p contribute to this process? If the physical interaction between Rap1p and Gcr1p is important for stabilizing Gcr1p binding at its target sites, mutated derivatives of Gcr1p that have mutations in the DNA binding domain might show more drastic phenotype in the strain carrying the N-terminal deleted Rap1p. To address this point, we randomly mutagenized the DNA binding domain of GCR1 (see Materials and methods). The mutant GCR1 library was transformed into the gcr1 mutant (C179-15C-F16) to isolate mutants with a partially impaired growth phenotype on SC(Glc). We isolated two clones (pD2005-1 and pD2006-1) that showed decreased growth compared to a transformant with the wild-type GCR1 (pL1040-11) on SC-Leu(Glc) plates at 30 °C, but faster growth rate than the gcr1 null mutant. Their plasmid dependency was confirmed by the isolation and retransformation of the plasmids into the original gcr1 mutant (C179-15C-F16) (Table 6, cf lines 1 and 10 with lines 4 and 7). DNA sequencing identified that pD2005-1 and pD2006-1 had single frame-shift mutations in codon 779 (deletion of A at nucleotide +2335 counting from the initiation codon, GCR1mut101) and 686 (deletion of T at nucleotide +2058, GCR1mut102) within the DNA binding domain of Gcr1p (aa 632–785; see Figure 1, Huie et al.1992). The DNA sequence predicted the truncation of the last seven amino acids and the addition of 17 unrelated amino acids in pD2005-1 and the truncation of the last 99 amino acids and the addition of 26 unrelated amino acids in pD2006-1. When pD2005-1, pD2006-1 and pL1711-2 (gcr1-Δ DBD) were transformed into the gcr1 mutant (C179-15C-F16) harbouring ENO1–lacZ reporter plasmids (pHU57 001), they showed 26%, 34% and 10% β-galactosidase activity compared to the strain with the wild-type GCR1 plasmid pL1040-11, respectively, suggesting that the mutated Gcr1p proteins have weaker transcriptional activation function, presumably due to the defect in DNA binding activity. When the rap1 conditional Δ gcr 1 mutant (YHU3089) carries the wild-type GCR1 plasmid (pL1040-11), additional transformation of pL1725-2 (RAP1) and pL1727-5 (rap1NsiI) showed no difference (Table 6, lines 1 and 2). However, when YHU3089 carries DBD mutants of gcr1 [GCR1mut101 (pD2005-1) or GCR1mut102 (pD2006-1)], transformation with the full-length RAP1 (pL1725-2) produced cells that grew faster than those with the N-terminus truncated rap1 (pL1727-5). This was reflected in their doubling times in liquid cultures (Table 6, lines 4, 5, 7 and 8; Figure 5). The fact that growth defect of the rap1 conditional Δ gcr 1 mutant strain harbouring the DBD mutated gcr1 was less well restored by the N-terminal truncated rap1 supported our idea that the N-terminus of Rap1p was important for the binding of Gcr1p to DNA and the subsequent activation of glycolytic enzyme genes.

Figure 5.

Growth of the strains harbouring GCR1 with a mutation in the DNA binding domain. The rap1 conditional Δ gcr 1 mutant strain (YHU3089) carrying various combinations of RAP1 and GCR1 plasmids were grown in SC(Glc)-Leu, Ura medium at 30 °C and the turbidity of the cells was monitored at 660 nm by an automatic detector (Bio-Plotter, Toyo-Sokki, Japan). Symbols: open triangle, YHU3089 with pL1040-11 (GCR1) and pL1725-2 (RAP1); closed triangle, pL1040-11 (GCR1) and pL1727-5 (rap1NsiI); open square, pD2005-1 (GCR1mut101) and pL1725-2 (RAP1); closed square, pD2005-1 (GCR1mut101) and pL1727-5 (rap1NsiI); open circle, pL1711-2 (gcr1-Δ DBD) and pL1725-2 (RAP1); closed circle, pL1711-2 (gcr1-Δ DBD) and pL1727-5 (rap1NsiI)

Table 6. Complementation of the Δgcr1 mutation by the DBD-mutant gcr1 gene in the rap1 conditional (UASGALRAP1) and Δgcr1 double mutant strain YHU3089
LineRAP1plasmidGCR1plasmidGrowth on plate (30 °C) [colony size (mm)]Generation time (min) in SC(Glc)
PlasmidGenePlasmidGeneSC(Glc) (2 days)SC(Glc) + anti (2 days)SC(GL) (4 days)
  1. YHU3089 has GAL promoter driven RAP1 (UASGALRAP1) instead of the native RAP1 gene, and it cannot grow on SC(Glc) and SC(GL) without the RAP1 gene on the plasmid.

 1pL1725-2RAP1pL1040-11GCR10.80.40.7143 ± 3
 2pL1727-5rap1–Δ NsiIpL1040-11GCR10.80.40.7144 ± 5
 3pRS316VectorpL1040-11GCR1 
 4pL1725-2RAP1pD2005-1GCR1mut1010.50.250.4249 ± 21
 5pL1727-5rap1–Δ NsiIpD2005-1GCR1mut1010.30.150.4320 ± 21
 6pRS316VectorpD2005-1GCR1mut101 
 7pL1725-2RAP1pD2006-1GCR1mut1020.40.20.6250 ± 25
 8pL1727-5rap1–Δ NsiIpD2006-1GCR1mut1020.2<0.050.6289 ± 20
 9pRS316VectorpD2006-1GCR1mut102 
10pL1725-2RAP1pL1711-2gcr1–Δ DBD<0.05<0.050.6932 ± 30
11pL1727-5rap1–Δ NsiIpL1711-2gcr1–Δ DBD<0.05<0.050.6936 ± 40
12pRS316VectorpL1711-2gcr1–Δ DBD 

Discussion

It was suggested that the N-terminal region of Rap1p may be involved in the regulation of Rap1p activity (Vignais and Sentenac, 1989; Muller et al., 1994), but its biological function was not well understood because its deletion did not result in any growth defect under standard growth conditions (Graham et al., 1999). In this study we provide the first genetic evidence for the importance of the N-terminal region of Rap1p for the function of Gcr1p. In two-hybrid experiments, deletion of a large part of the N-terminus of Rap1p abolished the interaction with Gcr1p. We also demonstrated that the interaction between Rap1p and Gcr2p was mediated by Gcr1p, because their interaction was not observed in Δ gcr 1 strains [YHU3086 (Δ gcr 1) and YHU3088 (Δ gcr 1 Δ gcr 2); Table 3, lines 6 and 7]. The weaker signal generated by the interaction between the full-length Rap1p and Gcr2p compared to that generated by Rap1p and Gcr1p is probably due to the difference between indirect and direct interactions. As has been reported by Graham et al. (1999), deletion of the N-terminal region of Rap1p did not show any growth phenotype by itself (Table 4, line 2, and Table 5, line 2), but an effect on growth was observed in the gcr2 mutation background (Tables 4 and 5; Figure 3).

The sequence analysis of Rap1p revealed that its N-terminal region contains a BRCT domain. Our phylogenetical approach suggested the biological significance of the BRCT domain in Rap1p function. We isolated a RAP1 gene from C. glabrata based on complementation activity (Haw et al., 2001) and compared its sequence to that of ScRap1p. In contrast to the high degree of identities in the DNA binding domain and the C-terminal region, the N-terminal region was less well conserved in CgRap1p [28% in ScRap1p (aa 1–361) vs. CgRap1p (aa 1–232)]. In spite of the low homology in the entire N-terminal region, the BRCT domain (aa 121–208 in ScRap1) was highly conserved [51% in CgRap1p (aa 82–169)]. Recently, Wahlin and Cohn isolated another RAP1 homologue from Saccharomyces castellii (ScasRap1p), based on complementation activity (2001). Like CgRap1p, the N-terminal region of Rap1p was less well conserved in ScasRap1p [27% in ScRap1p (aa 1–361) vs. ScasRap1p (aa 1–343)], but the BRCT domain was highly conserved [56% in ScasRap1p (aa 120–207)]. Conservation of the BRCT domain in Rap1p homologues suggests its importance in the function of Rap1p. Since the BRCT domain is a protein–protein interaction region found in a large number of proteins involved in DNA repair, recombination and cell cycle control (Koonin et al., 1996; Bork et al., 1997; Callebaut and Mornon, 1997) and the BRCT domain is the only motif found in this region, we speculated that this region may be the interaction interface between Rap1p and Gcr1p. Recent work by Miyake et al. (2000) demonstrated that the BRCT domain of Rap1p is able to activate transcription. Since Rap1p cannot activate glycolytic genes in the absence of a functional Gcr1p or CT-box, while it can bind to the RPG-box (Drazinic et al., 1996), transcriptional activation function of the BRCT domain might be mediated by the interaction with Gcr1p.

The UAS elements of all glycolytic genes contain binding sites for Rap1p and Gcr1p in close vicinity and these binding sites display a strong synergism for each other (Chambers et al., 1995). In vivo binding studies demonstrated that Gcr1p is unable to bind at its binding site in the absence of a bound Rap1p-binding site (Drazinic et al., 1996). Thus, the most likely role of Rap1p at glycolytic promoters is to facilitate the binding of Gcr1p (Drazinic et al., 1996; Uemura et al., 1997; Lopez et al., 1998), either by changing the DNA conformation or through a protein–protein interaction between Rap1p and Gcr1p. The first model proposed that Rap1p plays a role in maintaining accessibility of transcription factor binding sites within chromatin, based on the observation that the binding of Rap1p induces local bending and distortion of the DNA helix, which might help other regulatory factors to bind DNA (Vignais and Sentenac, 1989; Gilson et al., 1993). The second model was based on the co-immunoprecipitation of Rap1p and Gcr1p (Tornow et al., 1993). Lopez et al. (1998) demonstrated that both N-terminal and C-terminal regions of Rap1p independently facilitate the binding of Gcr1p to DNA. However, how this is achieved is not well understood. The contribution of the N-terminus of Rap1p in this process may be mediated by a direct interaction with Gcr1p, which we observed in two-hybrid analysis. The fact that the N-terminus deleted Rap1p rescued the DBD mutation of gcr1 less well indicates the importance of their interaction in stabilizing the binding of Gcr1p to DNA.

Our working hypothesis is that Rap1p facilitates the binding of Gcr1p to DNA in two ways, one of which involves a physical interaction between the Rap1p N-terminus and Gcr1p, and the second of which involves another region of Rap1p and could be indirect, perhaps via chromatin structure. We speculate that the functions of Gcr2p are to interact with Gcr1p in order to change the conformation of Gcr1p, and to stabilize Gcr1p within the activation complex; therefore, the function of the N-terminal region of Rap1p and Gcr2p might be partly redundant. When the N-terminus of Rap1p is deleted, one stabilizing effect on the complex is lost. Under this condition, binding of Gcr1p to DNA is still facilitated by the other region of Rap1p, but the stabilizing effect of Gcr2p becomes more important. Thus, when both Gcr2p and the N-terminus of Rap1p are deleted, the Gcr1p/Rap1p activation complex is destabilized, leading to inefficient gene activation. Recently we have started DNA microarray analysis of the strain with the N-terminus deleted rap1 gene [rap1NsiI]. Further analyses of the microarray results will shed light on the function of the N-terminal region of Rap1p in the wild-type strain.

Acknowledgements

We thank Tsuyoshi Miyake and Rong Li for communications. T.M. is a recipient of a postdoctoral fellowship of New Energy and Industrial Technology Development Organization (NEDO) of Japan. R.H. was a recipient of a post-doctoral fellowship (STA fellowship) of the Science and Technology Agency of Japan. Part of this work was supported by a grant from the Ministry of Education, Culture, Sports, Science and Technology of Japan and from the British Council to H.U. Early work on this project was supported by a UK/Japan joint grant (Royal Society, UK) to A.C. and H.U.

Ancillary