Mol Gen Genet (1992) 231:296-303
MGG
© Springer-Verlag 1992
Positive regulation of the LPD1 gene of Saccharomyces cerevisiae by the HAP2[HAP3/HAP4 activation system Susan B. Bowman 1, Zaf Zaman 2, Lindsay P. Collinson 1, Alistair J.P. Brown 3, and Ian W. Dawes ~ 1 School of Biochemistry,Universityof New South Wales, Kensington,NSW 2033, Australia z Institute of Cell and MolecularBiology,Universityof Edinburgh, West Mains Road, Edinburgh EH9 3JG, UK 3 Department of Molecularand Cell Biology,Marischal CollegeUniversityof Aberdeen, AberdeenAB9 1AS, UK Received June 17, 1991
Summary. The LPD1 gene of Saccharomyces cerevisiae, encoding lipoamide dehydrogenase (LPDH), is subject to catabolite repression. The promoter of this gene contains a number of motifs for DNA-binding transcriptional activators, including three which show strong sequence homology to the core HAP2/HAP3/HAP4 binding motif. Here we report that transcription of LPD1 requires HAP2, HAP3 and HAP4 for release from glucose repression. In the wild-type strain, specific activity of LPDH was increased 12-fold by growth on lactate, 10-fold on glycerol and four- to five-fold on galactose or raffinose, compared to growth on glucose. In hap2, hap3 and hap4 null mutants, the specific activities of LPDH in cultures grown on galactose and raffinose showed only slight induction above the basal level on glucose medium. Similar results were obtained upon assaying for /%galactosidase production in wild-type, or hap2, hap3 or hap4 mutant strains carrying a single copy of the LPD1 promoter fused in frame to the lacZ gene of Escherichia coli and integrated at the URA3 locus. Transcript analysis in wild-type and hap2 mutants confirmed that the HAP2 protein regulates LPD! expression at the level of transcription in the same way as it does for the CYC1 gene. Site-directed mutagenesis of the putative HAP2/HAP3/HAP4 binding site at - 204 relative to the ATG start codon showed that this element was required for full derepression of the LPD1 gene on non-fermetable substrates. Key words: Saccharomyces cerevisiae - Lipoamide dehydrogenase - HAP activation
Introduction The nuclear LPD1 gene of Saccharomyces cerevisiae encodes lipoamide dehydrogenase (EC 1.8.1.4). This enzyme serves a common function in the multienzyme
Offprint requests to : I.W. Dawes
complexes pyruvate dehydrogenase (PDH) and 2-oxoglutarate dehydrogenase (OGDH) (Reed 1974). By an analogous series of reactions, PDH and OGDH catalyse the oxidative decarboxylation of pyruvate and 2-oxoglutarate to acetyl CoA and succinyl CoA, respectively (Reed 1974). Both complexes supply and maintain the metabolic turnover of the citric acid cycle and are therefore essential for the respiratory competence of the mitochondria. Each complex is composed of multiple copies of a dehydrogenase (El), a transferase (E2) and lipoamide dehydrogenase (E3). The E1 and E2 components are unique to each complex, while E3 is identical in both, and in yeast is encoded by the single nuclear gene LPD! (Dickinson et al. 1986). The LPD! gene of Saccharomyces cerevisiae is subject to catabolite repression, showing an increase in transcription when cells are transferred from glucose-containing medium to media containing non-fermentable carbon substrates (Roy and Dawes 1987). Catabolite repression, defined as the inhibition of the synthesis of certain enzymes by glucose or other rapidly metabolised carbon sources, is a mechanism which is not well understood in yeast (Entian 1986). Gene expression in yeast is often controlled at the level of transcription via the binding of trans-acting regulatory proteins to cis-acting DNA elements in the 5' upstream region of the gene. Some of these eis-acting elements, termed upstream activation sites (UAS), are located hundreds of base pairs upstream of the start codon in yeast DNA. Within these UASs are short motifs to which activator or regulatory proteins bind, either alone or as complexes with other proteins, to modulate gene expression. These motifs and the proteins binding to them have been reviewed by Verdier (1990). One example of this form of control is seen in a number of yeast genes involved in respiration (Forsburg and Guarente 1989). The HAP2, HAP3 and HAP4 proteins form a heteromeric complex which binds to specific UAS elements in these genes to activate transcription. This system has been most extensively characterized for the yeast CYC1 gene, encoding isocytochrome c, which is activated from two upstream elements UAS1 and
297 UAS2. Activation from UAS2 is subject to catabolite repression with CYC1 transcription increasing about 30fold when cells are shifted from a fermentable to the non-fermentable carbon source lactate (Guarente et al. 1984). The activity of UAS2 is positively regulated by the products of the HAP2, H A P 3 and H A P 4 genes (Forsburg and Guarente 1988, 1989; Hahn and Guarente 1988). These proteins form a DNA-binding complex at UAS2 to activate transcription when cells are grown on a non-fermentable carbon source (Olesen et al. 1987). Mutations in HAP2, H A P 3 or H A P 4 are pleiotropic, resulting in reduced levels of many cytochromes and in an inability of cells to grow on non-fermentable carbon sources. The HAP2/HAP3/HAP4 system has been shown to regulate other yeast genes required for respiratory competence, including H E M 1 , COX4, COX6 (Forsburg and Guarente 1988; Trawick et al. 1989), KGD1 and KGD2 which encode, respectively, the E1 and E2 subunits of OGDH (Repetto and Tzagoloff 1989, 1990). A motif, with the consensus sequence TNA/GTTGGT has been found in those genes subject to HAP2 control; linker scanning analysis of the UAS2UP1 mutation of CYC1 has shown that this motif is critical for activation (Forsburg and Guarente 1988). The LPD1 gene, including 1 kb of the 5' non-coding region, has been cloned and sequenced (Ross et al. 1988). The promoter region contains a number of known regulatory motifs, including one in the region from - 2 0 6 to - 1 9 9 that shows strong homology to the HAP2/ HAP3/HAP4 binding motif and adjacent sequences in CYC1. The role of this motif and the HAP2, HAP3 and HAP4 trans-acting factors in regulation of expression of the LPD1 gene has been determined.
Materials and methods
Yeast strains and plasmids. Table 1 lists the genotypes and sources of yeast strains used. Plasmid pGP1 contains the intact LPD1 gene in a YEp13 vector (Roy and Dawes 1987). Plasmids pSPACT9 containing the yeast actin gene (Bettany et al. 1989), and pMC1871 carrying Escherichia coli lacZ (Casadaban et al. 1983) were supplied by A. Brown, and the centromeric pYCP50-1Z Table 1. Genotypeand source of Saccharomyces cerevisiae strains
Strain
Genotype
Source
BWG1-7a
MATa, adel-lO0, his4-519, leu2-3, leu2-112, ura3-52 MATa, adel-lO0, his4-519, leu2-3, leu2-112, ura3-52, hap2::LEU2 MATa, adel-lO0, his4-519, leu2-3, leu2-112, ura3-52, hap2 MA Ta, adel-lO0, his4-519, leu2-3, [eu2-112, ura3-52, hap3: : HIS4 MATa, adel-lO0, his4-519, leu2-3, leu2-112, ura3-52, hap4::LEU2
L. Guarentea
LWGI JOl-la SHY40 SLF401
a Dept. of Biology,MIT, Cambridge, MA, USA
L. Guarente L. Guarente L. Guarente L. Guarente
with the entire CYC1 promoter and a small portion of its coding sequence fused in frame to the E. coli lacZ gene (Zitomer et al. 1987) was kindly supplied by T. Pillar (Leicester). pCS1 was constructed by C. Stirling (Edinburgh) by removal of the CYC1 insert from pLG669-Z (Guarente et al. 1982). It carries a truncated lacI-lacZ fusion with a BamHI site for the introduction of yeast promoters via in frame fusions. Growth and transfer media. Yeast strains were grown at 30°C with shaking in liquid media containing 2% bactopeptone, 1% yeast extract and, where indicated, supplemented with: 2% glucose (YEPD); 2% galactose (YEPGal); 2% glycerol (YEPG); 2% lactate (YEPLact) or 2% raffinose (YEPRaff). Cells were washed before transfer to different carbon sources in medium containing 1% yeast extract and 2% peptone. Transformants were selected on solid minimal medium containing 2% glucose, 0.16% Difco yeast nitrogen base (without amino acids and ammonium sulphate), 0.5% ammonium sulphate and 2% agar. Auxotrophic requirements were added at 20 ~tg per ml. Plates for detecting/%galactosidase activity contained minimal media with 2% raffinose as carbon source made up in potassium phosphate buffer to a final pH of 7.0; 50 gl of X-Gal (4-bromo-4-chloro-3-indolyl-fl-D-galactoside; 20 mg/ml in dimethylformamide) was spread on each plate and allowed to dry before streaking cells. Plasmid construction and integration. Plasmid pZZD-2 g was used for site-directed mutagenesis and for integration into the yeast genome as a reporter of LPD1 promoter activity. A 1.47 kb fragment carrying the LPD1 promoter, ATG start codon and putative downstream activation site (DAS) was isolated from pGP-RI (Roy and Dawes 1987). This 1.47 kb fragment with XbaI linkers containing BamHI restriction sites was ligated into the BamHI cloning site of the yeast vector pCS1 from which the 2 micron yeast DNA was removed by partial EcoRI digestion and religation. The resulting plasmid, termed pZZD-2 ~t, was linearised using a unique SmaI restriction site within the URA3 gene. The linearised plasmid was used to direct integration of the plasmid to the chromosomal ura3 locus. Plasmid pYCP50-1Z was introduced into strains BWG/-7a and LGW-1 by the lithium acetate method of yeast transformation (Ito et al. 1983). Plasmid pZZD2g was introduced into strains BWG1-7a, LGW-1, JO11A, SHY40 and SLF401 using the spheroplast method of Hinnen et al. (1978). Positive transformants were selected as Ura + and confirmed by testing for/?-galactosidase on X-Gal plates. Segregation tests were performed by comparing plasmid retention under selective and nonselective conditions. To check the integrity of transformants, plasmid DNA was re-isolated from each by the method of Sherman et al. (1982) and characterised by restriction analysis. Site-directed mutagenesis. The HAP2/HAP3 binding motif in the upstream region of the LPD1 gene was modified using the gapped duplex method of site-
298 directed mutagenesis (Kramer and Fritz 1987). The 1.47 kb B a m H I fragment of plasmid pZZD-2g containing the upstream region of L P D 1 was cloned into the BamHI site of M13mp9. Gapped duplex D N A was then constructed using BamHI-treated M 13mp9rev plasmid DNA. An oligonucleotide containing four mismatches to the HAP2/HAP3 binding element (5' C T G T T C C C G G A T C C G C G A G A 3') was hybridised to the gapped duplex DNA. Minor modifications to the method included an in vitro extension reaction using T7 D N A polymerase and the amplification step was replaced by plating aliquots of the extension reaction, after transfection into E. coli BHM71-18 m u t S cells, directly onto a lawn of E. coli MK30/3 cells (Gaata et al. 1990). The mutation was identified by plaque hybridisation, using 5'-32p-labelled mutant HAP2/HAP3 oligonucleotide. Nitrocellulose filters were washed at 20 ° C below the Tm calculated by the Wallace rule (Miyada and Wallace 1987). This allowed clear discrimination between plaques containing wild-type D N A and those containing the desired mutation. Further characterisation of the mutant was performed by dideoxy D N A sequencing using T7 D N A polymerase (Tabor and Richardson 1987). Both the wild-type and mutant L P D 1 promoters were then isolated from the M13mp9 vector using an EcoRI + HindIII digestion and cloned independently into plasmid YIp358 (Myers et al. 1986) using the E c o R I - H i n d I I I sites in the cloning region. The plasmids were termed pLC358 (wild-type promoter) and pLCH358 for the mutant promoter. These constructs, containing an in-frame lacZ fusion with the wild-type or mutant promoter, were then integrated at the ura3 locus of the wild-type strain (BWG1-7a). Enzyme assays and protein concentration. Yeast cells were
grown to a n 0 D 6 o o of 1.0 and assayed for lipoamide dehydrogenase activity using 2-acetylpyridine adenine dinucleotide (APAD) as substrate as described by Dickinson et al. (1986). Units are expressed as gmol APAD reduced/rain. Quantitative assays for fl-galactosidase acTable 2. HAP2/HAP3/HAP4 binding elements
tivity using o-nitrophenyl-fl-D-galactose (ONPG) as substrate were performed as described by Guarente (1983), and units are expressed as mmol O N P G hydrolysed/min. Protein concentration was assessed by the Bio-Rad assay method as described by the manufacturer. R N A analysis and Northern blots. For isolation of total
RNA, yeast strains were grown on YEPD or YEPGal to a n 0 D 6 o o of 1.0. Total R N A was isolated using the phenol-glass bead method of Lindquist (1981). For Northern blot analysis, 20 gg of total R N A in water was dried and resuspended in 4.5 gl of water, 2 gl MOPS buffer, 3.5 gl formaldehyde, 10 gl formamide and 1 gl ethidium bromide (400 ~tg/ml) then incubated at 65°C for 15 min. After the addition of loading buffer (50% v/v glycerol, 1 m M EDTA, 0.25% bromophenol blue, 0.25% xylene cyanol), the samples were separated electrophoretically on a 20 cm 1.0% agarose gel under nondenaturing conditions, photographed, then transferred to a nitrocellulose filter. Qualitative evaluation of transcript levels was estimated with the Ambis 2D Radioanalytic system (Lablogic Ltd.). Hybridisation probes. The following D N A fragments were radiolabelled by the random priming method (Feinberg and Vogelstein 1983) after isolation from agarose gels. A 3.6 kb XhoI fragment was isolated from pGP1 (Roy and Dawes 1987) and used to probe for L P D 1 mRNA. A 3.1 kb EcoRI fragment isolated from pMC1871 (Casadaban et al. 1983) was used as probe for lacZ mRNA. Nick-translated plasmid pSPACT9 (Bettany et al. 1989) was used as probe for actin mRNA.
Results
H A P 2 / H A P 3 binding consensus in the LPD1 gene
Analysis of the sequence 5' to the start codon of the L P D ! gene has identified an element at position - 2 0 4
Match to consensus 5'-TNA/GTTGGT-3'
Locus"
Gene product
Position of match relative to ATG
TGGTTGGT TCATTGGT TTATTGGT TGATTGGc c TTATTGGc c TAAcTGGT c TGATTGGc TCATTGGg TCATTGGc TCATTGGc TTATTGGc
CYC1 b HEM1 COX4 KGD2
Iso-l-cytochrome c 6-Aminolevulinatesynthetase Cytochrome oxidase subunit IV Dihydrolipoyl transsuccinylase
KGD1
2-Oxoglutaratedehydrogenase
LPD1
Lipoamide dehydrogenase
-=210 -=374 -605 - 357 -325 - 398 - 320 -253 - 204 - 497
-731
a The nucleotide sequences of the promoter regions of these genes are given in the following references: CYC1, HEM1, COJ(4 (Forsburg and Guarente 1988), KGDI (Repetto and Tzagoloff 1989), KGD2 (Repetto and Tzagoloff 1990), LPD1 (Ross et al. 1988) b A G ~ A transition at position -208 has been shown to increase UAS2 activity (Forsburg and Guarente 1988) ° The HAP2/HAP3/HAP4 consensus sequence appears on the noncoding strand
299 showing homology to part of the UAS2 of the yeast CYC1 gene (Guarente et al. 1984) and subsequently two other matches at - 4 9 7 and - 7 3 1 were found. The CYC1 UAS2 consists of two functionally distinct regions, each containing a binding site for a different transacting factor. Region 1 mediates the carbon source response while region 2 is required for maximal activity from region 1 (Forsburg and Guarente 1988). Region 1 has a close match to the consensus sequence T N A T T G G T which has been identified as the putative binding site for the H A P 2 / H A P 3 / H A P 4 complex. Homology to this region has also been reported for a number of other genes including COX4, H E M I (Forsburg and Guarente 1988), COX6 (Trawick et al. 1989), KGD1 (Repetto and Tzagoloff 1989) and KGD2 (Repetto and Tzagoloff J990), which are also regulated by the same H A P factors (Table 2). The LPD1 motif aligned with those of CYC1, COX4, HEM1, KGDI and KGD2 shows a number of other characteristics c o m m o n to the H A P 2 / H A P 3 / H A P 4 binding sites of these other genes. There is a GC-rich sequence of dyad symmetry, in the region immediately 3' to the consensus element showing some homology between CYC1, HEM1, COX4 and LPDI ; for the KGD genes at least one of the potential motifs has a similar GC-rich region. Other short regions of dyad symmetry are present in LPD1 as well as in CYCI and HEMI, which differ by sequence, but not in nature, from those of CYC1. These regions are important for the functional integrity of the H A P 2 / H A P 3 / H A P 4 binding site in CYCI (Forsburg and Guarente 1988) and may play a similar role in the other genes. A single base change in the CYCI UAS2 transforming it to CYC1 UAS2UP1, which conforms with the consensus, causes a 10-fold increase in UAS2 activity. The potential elements in LPDI differ by 1 bp from the consensus for the HAP2 binding site, but this involves substitution of a C in the last position, a change that is also found in the functional motifs of KGD1 and KGD2.
Effect o f hap mutations on lipoamide dehydrogenase activity under steady-state growth conditions The LPD1 gene is expressed at basal levels in cells grown on medium containing glucose. When cells are grown on carbon sources other than glucose the LPDI gene is derepressed. Partial derepression is seen in strains growing on media containing galactose or raffinose, while the LPD! gene is expressed at high levels when cells are grown on non-fermentable carbon sources such as lactate or glycerol. The level of derepression of the LPD1 gene can be observed by assaying lipoamide dehydrogenase (LPDH) activity under different growth conditions. To determine the role of the H A P 2 / H A P 3 / H A P 4 system on LPD1 expression, the specific activity of L P D H was assayed in the wild type BWG1-7a and in the strains bearing hap2, hap3 and hap4 mutations. The hap strains are unable to grow on non-fermentable carbon sources but can grow on media (YEPGal and YEPRaft) that cause partial derepression o f catabolite-repressed genes.
Table 3. Expression of the intact LPD1 gene assayed in hap2, hap3 and hap4 mutant strains Strain
BWGI-7a LGW-1 b JOl-la b SHY 40 SLF 401
Relevant genotype Wild-type
hap2 hap2 hap3 hap4
LPDH Activity a YEPD
YEPGal
YEPRaff
4.4 3.3 3.3 3.5 2.3
17.5 7.9 5.9 6.0 6.2
15.6 9.9 8.0 8.4 7.4
" LPDH activity refers to specific activity of lipoamide dehydrogenase (expressed as x 10 -3 gmol product per min per mg of protein). The values are the average of three sets of experiments. Values for duplicate assays differed from each by no more than 5% b The hap2 mutation in strain LGW-1 was a disruption, while that in JOl-la was a null mutation
Wild-type cells grown on Y E P D showed basal levels of L P D H enzyme activity (Table 3). When wild-type cells were grown on Y E P G a l or Y E P R a f f L P D H enzyme activity increased 4-fold and 3.5-fold respectively, over the basal levels in glucose. As the data presented in Table 3 shows, all four hap strains showed a reduced basal level of enzyme activity in glucose. Under partially derepressive conditions (YEPGal and YEPRaff), enzyme activity in the hap strains showed a lower level of induction, with enzyme activity increasing by only 1.7 to 2.7fold in galactose and by 2.4 to 3.2-fold in raffinose. The reduced level of L P D H enzyme activity under steady-state repressive, and partially derepressive conditions in all four hap mutants tested indicated that each of the H A P proteins exerts a measure of control on the synthesis of L P D H .
Comparison o f L P D I and CYCI expression .following derepression The regulation of LPD1 was monitored in conjunction with that of C¥C1 in wild-type and hap2 mutant strains during a switch from growth on repressive to a partially derepressive medium. To facilitate assay o f CYC1 gene expression, a C YCl-lacZ fusion plasmid (pYCP 150-1 Z) was transformed into wild-type and hap2 strains. Plasmid pYCP150-1Z (a gift from T. Pillar) is a centromeric vector tailored to express the lacZ reporter gene product from the CYC! promoter (Zitomer et al. 1987). Strains BWG1-7a/Z (HAP2) and LGW-1/Z (hap2) were the resultant transformed derivatives of strains B W G I - 7 a and L G W - I respectively. Strains BWG1-7a/Z and LGW-1/Z were grown on Y E P D medium containing 8% glucose to ensure that expression of CYCI and LPD1 was fully repressed. The strains were then transferred to fresh media containing 2% glucose or 2% raffinose to follow derepression of both genes. L P D H and /q-galactosidase activities were monitored at mid-log phase of growth and then 3 and 8 h after transfer to 2% glucose or 2% raffinose. Following a switch from 8% glucose to 2% glucose there was
300 8-
A YEPD
galactosidase activities (Fig. 1). On a switch to raffinose m e d i u m there was an a p p r o x i m a t e eight-fold derepression o f b o t h enzyme activities in the wild-type strain but only a small change in the hap2 m u t a n t (Fig. 1). M o n i t o r i n g o f expression o f the CYCI-lacZ fusion as a control confirmed that H A P 2 protein was acting normally in the strain used to give partial derepression o f the CYC1 gene u n d e r the a p p r o p r i a t e conditions. The parallel d a t a for L P D H enzyme activity indicates that the LPDI gene is also regulated, in part, by the H A P 2 protein, when cells are shifted f r o m a repressive to a partially derepressive g r o w t h condition.
.[ 13 YEPRaf .
:>6-
• BWG-TaZ [ ] LGW-1Z
*6 -r
8
C YEPD
D YEPRaf
.I
BWG-7aZ [] LGW-1Z []
2;6. ~. o
*6 i:J
i:i!!il
:i!!!i!i! iiiiiiiiiiiii
-~2,
0
3
8
0 Time
Fig. 1A-D. Catabolite derepression of LPDH activity following a shift of HAP2 and hap2 strains from glucose to raffinose as carbon source. Cells were grown to log phase on YEPD containing 8% glucose, washed and transferred at zero time to either YEPD (2% glucose; A and C) or YEPRaff (2% raffinose; B and D). Cell extracts were prepared before the shift and at 3 and 8 h after the shift, and subsequently assayed for LPDH enzyme activity (A and B) and #-galactosidase activity (C and D). The values are averages of triplicate assays and are normalised to the activity of each enzyme in wild-type cells grown on YEPD containing 8% glucose (2 x 10 -3 U/rag protein for LPDH and 0.017 U/mg protein for #-galactosidase)
a threefold increase in L P D H activity in strains with either the wild-type or hap2 b a c k g r o u n d , bringing the basal levels near to that observed under steady state g r o w t h conditions. This increase was paralleled in the expression o f CYC1, inferred f r o m m e a s u r e m e n t o f #-
HAP regulation at the level of transcriptional control
To further assess the role of the HAP binding proteins in regulation of the LPD1 gene, LPDI-lacZ fusions were constructed. A 1.47 kb region o f the LPD1 gene was ligated to the E. coli lacZ reporter gene via an inframe translational fusion. The 1.47 kb f r a g m e n t contains the u p s t r e a m p r o m o t e r region, the A T G start c o d o n and a putative d o w n s t r e a m activation sequence (DAS) o f the LPD1 gene. This construct was integrated into wildtype, hap2, hap3 a n d hap4 strains• Levels o f #-galactosidase activity in cells t r a n s f o r m e d with this plasmid reflect the transcriptional control and activation o f the LPD1 p r o m o t e r region. #-galactosidase assays were perf o r m e d on the t r a n s f o r m e d strains g r o w n u n d e r steadystate repressive ( Y E P D ) or partially derepressive (YEPGal) conditions. All strains g r o w n on Y E P D gave basal levels o f #galactosidase activity (Table 4). W h e n wild-type cells were g r o w n on Y E P G a l the p r o m o t e r was derepressed, showing a five-fold increase in #-galactosidase activity. However, only basal levels o f #-galactosidase were seen in hap strains g r o w n u n d e r the same conditions. These results are similar to those obtained with the L P D H assays (Table 3). To confirm that the H A P 2 protein regulates activity o f L P D H at the level o f transcription, the steady-state m R N A levels o f LPD1 and lacZ were examined in the
Table 4. Effects of hap2, hap3, and hap4 on the derepression of LPD1 Strain
Relevant genotype
#-Galactosidase specific activity a Expt 2
Expt 1
BWG1-7a LGW-I ~ JOl-la c SHY 40 SLF 401
Wild-type
hap2 hap2 hap3 hap4
% Wild type b
YEPD
YEPGal
YEPD
YEPGal
YEPD
YEPGal
1.23 1.15 1.08
6.60 1.98 1.26
2.60 2.40 1.83 2.36 2.65
9.60 3.10 2.20 2.83 3.20
100 92.5 79 90.5 98
100 32 21 29 27.5
1.12
1.85
1.20
1.40
" All strains were transformed with plasmid pZZd-2lx; strains were grown on YEPD (containing 2% glucose) or YEPGal (containing 2% galactose), fi-galactosidase was assayed as described by Guarente (1983). Data from 2 different experiments are presented; activities are reported as specific #-galactosidase activity (x 10 .3 gmol product per min per mg protein)
b Average activity from the two experiments, expressed as a percentage of wild-type levels The hap2 mutation in strain LGW-1 was a disruption, while that in JOl-la was a null mutation
301 YEPD
YEPRaff
YEPGal
N
'7,
~,
6
6
rn
I
I
I
I
A
~'~
!,~i~
- LPD1
B -
C
.~
.'~
~
CYCI-lacZ
- Actin
2.2 and 3.2-fold in wild-type cells grown in raffinose, compared to cells grown on glucose. The transcripts did not show such a change in hap2 cells. LPD1 transcript levels (normalised to their respective levels in wild-type cells on glucose) were derepressed about 5-fold in wildtype cells compared with hap2 cells grown on either galactose or raffinose medium. The same filter was then reprobed with a lacZ-specific probe (Fig. 2B). The CYCl-lacZ mRNA levels were shown to be regulated in a manner identical to LPD1 mRNA on each carbon source. The origin of other bands on this filter was unclear. RNA from a yeast strain devoid of lacZ sequences treated similarly did not give any background bands with the same probe (data not shown). These bands may be degradation products of the major lacZ transcript or readthrough transcripts from plasmid sequences. Relative LPDH and fl-galactosidase activities measured in cell samples used for mRNA assays showed no change in LPDH activities in the hap2 mutant and approximately 2- and 4-fold increases in activity in the wild-type strain on raffinose and galactose respectively. This indicated that enzyme activities were a reflection of the transcriptional regulation of each gene by the HAP2 protein.
D -25S -
H A P control is mediated mainly by the HAP2/HAP3 binding motif at - 2 0 4 in the control region of the LPD1 gene
18S
Fig. 2A-D. Regulation of LPD1 gene transcription by HAP2. Northern blots of total R N A (20 gg per lane), isolated from either the wild-type strain BWG1-7A/Z (relevant genotype: LPDI HAP2 CYCI-lacZ) or the hap2 mutant LGW-1/Z (relevant genotype: LPDI hap2 CYCI-lacZ) grown to mid-log phase on YEPD, YEPRaff or YEPGal are shown. Panels A, B and C display the same filter hybridized with probes for LPD1, lacZ and actin m R N A respectively. Panel D shows the gel stained with ethidium bromide
wild-type and hap2 strains grown on glucose and less repressing carbon sources. RNA was isolated from wildtype (BWG1-7A/Z) and hap2 (LGW-lZ) cells grown to mid-log phase on glucose (YEPD), galactose (YEPGal) or raffinose (YEPRaff) medium. The strains used carried the centromeric plasmid YCp150-1Z to facilitate comparison of LPD1 and CYC1 transcription levels. The RNA species were analysed by agarose gel electrophoresis, followed by Northern blotting. The Northern filter was initially hybridised with an LPD1 probe followed by an actin probe as a loading control (Fig. 2A and C). This assay showed little change in LPD1 mRNA levels in the hap2 mutant on all carbon sources tested. Relative amounts of LPDI, lacZ and actin mRNA were estimated using the Ambis apparatus. L P D I mRNA levels, normalised to those of actin were derepressed between
The putative HAP2/HAP3 binding motif at - 2 0 4 in the LPD1 gene shows strong sequence similarity to known functional HAP2/HAP3 binding motifs (Table 2) (Ross et al. 1988). To assess the role of this region in the control of LPD1 gene expression, site-directed mutagenesis was used to introduce 4 base changes in the HAP2/HAP3 binding element in plasmid pZZD-2 ~t, altering it from 5'-TCATTGGC-3' to 5'-GGATCCGC-3'. The mutant and wild-type promoters were fused in frame to the lacZ reporter gene and these constructs were integrated at the ura3 locus of the wild-type strain (BWGI-7a) thus retaining, in addition to the fusion, an intact copy of wild-type LPD1. The presence of wildtype LPD1 and H A P genes allow the strains to be grown on non-fermentable carbon substrates and subsequently assayed for fl-galactosidase activity. When grown on the non-fermentable carbon substrates glycerol or lactate (YEPG or YEPLact), the wildtype promoter was fully derepressed, increasing specific fi-galactosidase activity 10-fold and 12-fold, respectively, compared to the basal levels of enzyme activity of YEPD grown cells (Fig. 3). The mutant promoter gave the same basal level of activity as the wild-type promoter in repressing conditions and slight derepression on the partially derepressive carbon source galactose. However, when the strain bearing the mutant promoter was grown under fully derepressive conditions, specific fi-galactosidase activity increased only two- and four-fold. These results indicate that an intact HAP2/HAP3 binding motif at - 204 is required for full derepression of the LPD1
302 5P ._>
~3 *d
g2 i
(/3
YEPD YEPGa[ YEPG YEPkact Growth media
Fig. 3. Specific fi-galactosidase activity of the wild type LPD1 promoter-lacZ fusion pLC358 and the altered LPDI promoter-lacZ fusion pLCH358 integrated into the wild-type strain BWGI-7A at the ura3 locus. The altered promoter has 4 base changes in the HAP2/HAP3 binding motif at - 2 0 4 of the LPD1 gene, altering it from 5'-TCATTGGC-3' to 5'-GGATCCGC-3'. The resulting strains, denoted as wild type and altered, were assayed for fl-galactosidase activity under repressed, partially derepressed and fully derepressed conditions; YEPD, YEPGal and YEPG or YEPLact respectively
gene on non-fermentable substrates but that there may be a weaker contribution to full derepression in YEPLact from other elements located elsewhere in the promoter region.
Discussion The LPDI gene can be derepressed 50-fold in cells grown on lactate, 30-fold on glycerol, and 4- to 5-fold on galactose and raffinose relative to the level seen in cells grown on glucose. This derepression has been found to be mediated by the action of the HAP2, HAP3 and HAP4 proteins, since the levels of expression of the gene seen in hap2, hap3 and hap4 mutants are comparable with those seen on glucose medium. The majority of this effect is due to the HAP2 binding motif at - 2 0 4 in the upstream region of the gene since site-directed mutagenesis of this element leads to a very marked reduction in LPDI expression under strongly derepressing conditions. There is, however, some residual level of expression above the basal level in cells grown on lactate medium and carrying this mutant construct. This may be due to other elements within the control region of the gene. One candidate for such an element is another HAP2 motif at - 7 3 1 ; this is supported to some extent by the finding that a deletion from - 7 9 4 to - 6 9 7 does lead to a decrease in expression from an integrated lacZ fusion construct (J. Waterkeyn and I. Dawes, unpublished). The above results are similar to those seen for expression of the CYC1 gene on glucose and non-fermentable carbon sources. On glucose, CYC1 is regulated by the HAP1 activator protein binding to a site (UAS1) distinct from that (UAS2) of the HAP2, HAP3 and HAP4 complex. LPD1 does not appear to have a region showing
homology to the CYC1 UAS1. At present it is not known whether there is a sequence analogous to the CYC1 UAS1 involved in mediating the expression of LPD1 in glucose-grown cells. It would be interesting to determine whether HAP1 does regulate expression of LPD1 despite the absence of an obvious binding site, since Pfeifer etal. (1987) clearly demonstrated that HAP1 binds to a dissimilar sequence at the CYC7 promoter. The HAP2, HAP3 and HAP4 system has been found to regulate a number of other mitochondrial enzymes and may represent one of the main mechanisms for coordinating the derepression of these enzymes in response to changes in carbon status of the medium. The LPD1 gene encodes an enzyme subunit c o m m o n to the pyruvate dehydrogenase (PDH) and 2-oxoglutarate dehydrogenase ( O G D H ) multienzyme complexes. The other two genes encoding subunits of O G D H (KGD1 and KGD2) have been shown to contain HAP2-binding motifs, and these genes are regulated in a manner analogous to LPD1 (Repetto and Tzagoloff 1989, 1990). It is not surprising that common transcriptional regulators exist for the subunits of O G D H since it reflects at least one mechanism to ensure their coordinated response to growth on different carbon sources. It will be interesting to see whether the genes unique to P D H are controlled in the same way. There is, however, an additional dimension in the regulation of LPD1 since its product serves the two different complexes which, in E. coli at least, can be subject to some degree of differential control (Spencer and Guest 1985). This is reflected in the fact that the LPD1 gene in yeast has a number of other potential control motifs in its upstream region that are not found in the KGD1 and KGD2 genes (Dawes 1989). Acknowledgments. We wish to thank L. Guarente and his colleagues for kindly providing us with the hap mutants used. This work was supported by grants from the Science and Engineering Research Council and the Australian Research Council. Z.Z. was supported by an SERC Postgraduate Studentship and S.B. is the recipient of an Australian Postgraduate Research Award.
References Bettany AJE, Moore PA, Cafferkey R, Bell LD, Goodey AR, Carter BL, Brown AIP (1989) 5'-secondary structure formation, in contrast to a short string of non-preferred codons, inhibits the translation of the pyruvate kinase mRNA in yeast. Yeast 5:187-198 Casadaban MJ, Martinez-Arias A, Shapira SK, Chou J (1983) figalactosidase gene fusions for analyzing gene expression in Escherichia coli and yeast. Meth Enzymol 100:293-308 Dawes IW (1989) Complex regulation of gene expression in Saccharomyces cerevisiae and implications for biotechnology. Aust J Biotechnol 3 : 117-124 Dickinson JR, Roy DJ, Dawes IW (1986) A mutation affecting lipoamide dehydrogeanse, pyruvate dehydrogenase and 2-oxoglutarate dehydrogenase activities in Saccharomyces cerevisiae. J. Gen Microbiol 133:925-933 Entian KD (1986) Glucose repression: a complex regulatory system in yeast. Microbiol Sci 3 : 366-371 Feinberg AP, Vogelstein B (1983) A technique for radiolabelling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 132: 6-13
303 Forsburg S, Guarente L (1988) Mutational analysis of upstream activation sequence 2 of the CYC1 gene of Saccharomyces cerevisiae: a HAP2-HAP3-responsive site. Mol Cell Biol 8 : 647654 Forsburg SL, Guarente L (1989) Identification and characterization of HAP4: a third component of the CCAAT-bound HAP2/ HAP3 heteromer. Genes Dev 3 : 1166-1178 Gafita BA, Sharp SJ, Stewart TS (1990) Saturation mutagenesis of the Drosophila tRNA A~ggene B-box intragenic promoter element: requirements for transcription activation and stable complex formation. Nucleic Acid Res 18:1541-1548 Guarente L, Yocum RR, Gifford P (1982) A GALIO-CYC1 hybrid yeast promoter identifies the GAL4 regulatory region as an upstream site. Proc Natl Acad Sci USA 79:7410-7414 Guarente L (1983) Yeast promoters and lacZ fusions designed to study expression of cloned genes in yeast. Meth Enzymol 101:181-191 Guarente L, Lalonde B, Gifford P, Alani E (1984) Distinctly regulated tandem upstream activation sites mediate eatabolite repression of the CYC1 gene of Saceharornyces cerevisiae. Cell 36: 503-511 Hahn S, Guarente L (1988) Yeast HAP2 and HAP3: Transcriptional activators in a heteromeric complex. Science 240:317-321 Hinnen A, Hicks JB, Fink GR (1978) Transformation of yeast. Proc Natl Acad Sci USA 75:1929-1933 Ito H, Fukuda Y, Murata K, Kimura A (1983) Transformation of intact yeast treated with alkali cations. J Bacteriol 153:163168 Kramer W, Fritz HJ (1987) Oligonucleotide-directed construction of mutations via gapped duplex DNA. Meth Enzymol 154: 350 367 Lindquist S (1981) Regulation of protein synthesis during heat shock. Nature 293:311-314 Miyada CG, Wallace RB (1987) Oligonucleotide hybridization techniques. Meth Enzymol 154:94-107 Myers AM, Tzagotoff A, Kinney DM, Lusty CJ (1986) Yeast shuttle and integrative vectors with multiple cloning sites suitable for construction of lacZ fusions. Gene 45: 299-310 Olesen J, Hahn S, Guarente L (1987) Yeast HAP2 and HAP3 activators both bind to the CYC1 upstream activation site, UAS2, in an interdependent manner. Cell 51:953-961
Pfeifer K, Arcangioli B, Guarente L (1987) Yeast HAP1 activator competes with the factor RC2 for binding to the upstream activation site UASI of the CYC1 gene. Cell 49:9-18 Reed LJ (1974) Multienzyme complexes. Accounts Chem Res 7:4046 Repetto B, Tzagoloff A (1989) Structure and regulation of KGD1, the structural gene for the yeast c~-ketoglutarate dehydrogenase. Mol Cell Biol 9:2695-2705 Repetto B, Tzagoloff A (1990) Structure and regulation of KGD2, the structural gene for yeast dihydrolipoyl transsuccinylase, Mol Celt Biol 10:4221-4232 Ross J, Reid GA, Dawes IW (1988) The nucleotide sequence of the LPD1 gene encoding lipoamide dehydrogenase in Saecharomyces cerevisiae: comparison between eukaryotic and prokaryotic sequences for related enzymes and identification of potential upstream control sites. J Gen MicrobioI 134: l 131-1139 Roy DJ, Dawes 1W (1987) Cloning and characterization of the gene encoding lipoamide dehydrogenase in Saccharomyces cerevisiae. J Gen Microbiol 133:925-933 Sherman F, Fink GR, Hicks JB (1982) Laboratory course manual for methods in yeast genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York Spencer ME, Guest JR (1985) Transcriptional analysis of the sucAB, aceEF and lpd genes of Eseherichia coli. Mol Gen Genet 200:145-154 Tabor S, Richardson C (1987) DNA sequence analysis with a modified bacteriophage T7 DNA polymerase. Proc Nat Acad Sci USA 84:4767-4771 Trawick JD, Rugness C, Poyton RO (1989) Identification of an upstream activation sequence and other cis-acting elements required for transcription of COX6 from Saccharomyces cerevisiae. Mol Cell Biol 9:5350-5358 Verdier JM (1990) Regulatory DNA-binding proteins in yeast: an overview. Yeast 6: 271-297 Zitomer RS, Sellers JW, McCarter DW, Hastings GA, Wick P, Lowry CV (1987) Elements involved in oxygen regulation of the Saccharomyces cerevisiae CYC7 gene. Mol Cell Biol 7: 2212-2220 C o m m u n i c a t e d b y B.J. K i l b e y
MGG
© Springer-Verlag 1992
Positive regulation of the LPD1 gene of Saccharomyces cerevisiae by the HAP2[HAP3/HAP4 activation system Susan B. Bowman 1, Zaf Zaman 2, Lindsay P. Collinson 1, Alistair J.P. Brown 3, and Ian W. Dawes ~ 1 School of Biochemistry,Universityof New South Wales, Kensington,NSW 2033, Australia z Institute of Cell and MolecularBiology,Universityof Edinburgh, West Mains Road, Edinburgh EH9 3JG, UK 3 Department of Molecularand Cell Biology,Marischal CollegeUniversityof Aberdeen, AberdeenAB9 1AS, UK Received June 17, 1991
Summary. The LPD1 gene of Saccharomyces cerevisiae, encoding lipoamide dehydrogenase (LPDH), is subject to catabolite repression. The promoter of this gene contains a number of motifs for DNA-binding transcriptional activators, including three which show strong sequence homology to the core HAP2/HAP3/HAP4 binding motif. Here we report that transcription of LPD1 requires HAP2, HAP3 and HAP4 for release from glucose repression. In the wild-type strain, specific activity of LPDH was increased 12-fold by growth on lactate, 10-fold on glycerol and four- to five-fold on galactose or raffinose, compared to growth on glucose. In hap2, hap3 and hap4 null mutants, the specific activities of LPDH in cultures grown on galactose and raffinose showed only slight induction above the basal level on glucose medium. Similar results were obtained upon assaying for /%galactosidase production in wild-type, or hap2, hap3 or hap4 mutant strains carrying a single copy of the LPD1 promoter fused in frame to the lacZ gene of Escherichia coli and integrated at the URA3 locus. Transcript analysis in wild-type and hap2 mutants confirmed that the HAP2 protein regulates LPD! expression at the level of transcription in the same way as it does for the CYC1 gene. Site-directed mutagenesis of the putative HAP2/HAP3/HAP4 binding site at - 204 relative to the ATG start codon showed that this element was required for full derepression of the LPD1 gene on non-fermetable substrates. Key words: Saccharomyces cerevisiae - Lipoamide dehydrogenase - HAP activation
Introduction The nuclear LPD1 gene of Saccharomyces cerevisiae encodes lipoamide dehydrogenase (EC 1.8.1.4). This enzyme serves a common function in the multienzyme
Offprint requests to : I.W. Dawes
complexes pyruvate dehydrogenase (PDH) and 2-oxoglutarate dehydrogenase (OGDH) (Reed 1974). By an analogous series of reactions, PDH and OGDH catalyse the oxidative decarboxylation of pyruvate and 2-oxoglutarate to acetyl CoA and succinyl CoA, respectively (Reed 1974). Both complexes supply and maintain the metabolic turnover of the citric acid cycle and are therefore essential for the respiratory competence of the mitochondria. Each complex is composed of multiple copies of a dehydrogenase (El), a transferase (E2) and lipoamide dehydrogenase (E3). The E1 and E2 components are unique to each complex, while E3 is identical in both, and in yeast is encoded by the single nuclear gene LPD! (Dickinson et al. 1986). The LPD! gene of Saccharomyces cerevisiae is subject to catabolite repression, showing an increase in transcription when cells are transferred from glucose-containing medium to media containing non-fermentable carbon substrates (Roy and Dawes 1987). Catabolite repression, defined as the inhibition of the synthesis of certain enzymes by glucose or other rapidly metabolised carbon sources, is a mechanism which is not well understood in yeast (Entian 1986). Gene expression in yeast is often controlled at the level of transcription via the binding of trans-acting regulatory proteins to cis-acting DNA elements in the 5' upstream region of the gene. Some of these eis-acting elements, termed upstream activation sites (UAS), are located hundreds of base pairs upstream of the start codon in yeast DNA. Within these UASs are short motifs to which activator or regulatory proteins bind, either alone or as complexes with other proteins, to modulate gene expression. These motifs and the proteins binding to them have been reviewed by Verdier (1990). One example of this form of control is seen in a number of yeast genes involved in respiration (Forsburg and Guarente 1989). The HAP2, HAP3 and HAP4 proteins form a heteromeric complex which binds to specific UAS elements in these genes to activate transcription. This system has been most extensively characterized for the yeast CYC1 gene, encoding isocytochrome c, which is activated from two upstream elements UAS1 and
297 UAS2. Activation from UAS2 is subject to catabolite repression with CYC1 transcription increasing about 30fold when cells are shifted from a fermentable to the non-fermentable carbon source lactate (Guarente et al. 1984). The activity of UAS2 is positively regulated by the products of the HAP2, H A P 3 and H A P 4 genes (Forsburg and Guarente 1988, 1989; Hahn and Guarente 1988). These proteins form a DNA-binding complex at UAS2 to activate transcription when cells are grown on a non-fermentable carbon source (Olesen et al. 1987). Mutations in HAP2, H A P 3 or H A P 4 are pleiotropic, resulting in reduced levels of many cytochromes and in an inability of cells to grow on non-fermentable carbon sources. The HAP2/HAP3/HAP4 system has been shown to regulate other yeast genes required for respiratory competence, including H E M 1 , COX4, COX6 (Forsburg and Guarente 1988; Trawick et al. 1989), KGD1 and KGD2 which encode, respectively, the E1 and E2 subunits of OGDH (Repetto and Tzagoloff 1989, 1990). A motif, with the consensus sequence TNA/GTTGGT has been found in those genes subject to HAP2 control; linker scanning analysis of the UAS2UP1 mutation of CYC1 has shown that this motif is critical for activation (Forsburg and Guarente 1988). The LPD1 gene, including 1 kb of the 5' non-coding region, has been cloned and sequenced (Ross et al. 1988). The promoter region contains a number of known regulatory motifs, including one in the region from - 2 0 6 to - 1 9 9 that shows strong homology to the HAP2/ HAP3/HAP4 binding motif and adjacent sequences in CYC1. The role of this motif and the HAP2, HAP3 and HAP4 trans-acting factors in regulation of expression of the LPD1 gene has been determined.
Materials and methods
Yeast strains and plasmids. Table 1 lists the genotypes and sources of yeast strains used. Plasmid pGP1 contains the intact LPD1 gene in a YEp13 vector (Roy and Dawes 1987). Plasmids pSPACT9 containing the yeast actin gene (Bettany et al. 1989), and pMC1871 carrying Escherichia coli lacZ (Casadaban et al. 1983) were supplied by A. Brown, and the centromeric pYCP50-1Z Table 1. Genotypeand source of Saccharomyces cerevisiae strains
Strain
Genotype
Source
BWG1-7a
MATa, adel-lO0, his4-519, leu2-3, leu2-112, ura3-52 MATa, adel-lO0, his4-519, leu2-3, leu2-112, ura3-52, hap2::LEU2 MATa, adel-lO0, his4-519, leu2-3, leu2-112, ura3-52, hap2 MA Ta, adel-lO0, his4-519, leu2-3, [eu2-112, ura3-52, hap3: : HIS4 MATa, adel-lO0, his4-519, leu2-3, leu2-112, ura3-52, hap4::LEU2
L. Guarentea
LWGI JOl-la SHY40 SLF401
a Dept. of Biology,MIT, Cambridge, MA, USA
L. Guarente L. Guarente L. Guarente L. Guarente
with the entire CYC1 promoter and a small portion of its coding sequence fused in frame to the E. coli lacZ gene (Zitomer et al. 1987) was kindly supplied by T. Pillar (Leicester). pCS1 was constructed by C. Stirling (Edinburgh) by removal of the CYC1 insert from pLG669-Z (Guarente et al. 1982). It carries a truncated lacI-lacZ fusion with a BamHI site for the introduction of yeast promoters via in frame fusions. Growth and transfer media. Yeast strains were grown at 30°C with shaking in liquid media containing 2% bactopeptone, 1% yeast extract and, where indicated, supplemented with: 2% glucose (YEPD); 2% galactose (YEPGal); 2% glycerol (YEPG); 2% lactate (YEPLact) or 2% raffinose (YEPRaff). Cells were washed before transfer to different carbon sources in medium containing 1% yeast extract and 2% peptone. Transformants were selected on solid minimal medium containing 2% glucose, 0.16% Difco yeast nitrogen base (without amino acids and ammonium sulphate), 0.5% ammonium sulphate and 2% agar. Auxotrophic requirements were added at 20 ~tg per ml. Plates for detecting/%galactosidase activity contained minimal media with 2% raffinose as carbon source made up in potassium phosphate buffer to a final pH of 7.0; 50 gl of X-Gal (4-bromo-4-chloro-3-indolyl-fl-D-galactoside; 20 mg/ml in dimethylformamide) was spread on each plate and allowed to dry before streaking cells. Plasmid construction and integration. Plasmid pZZD-2 g was used for site-directed mutagenesis and for integration into the yeast genome as a reporter of LPD1 promoter activity. A 1.47 kb fragment carrying the LPD1 promoter, ATG start codon and putative downstream activation site (DAS) was isolated from pGP-RI (Roy and Dawes 1987). This 1.47 kb fragment with XbaI linkers containing BamHI restriction sites was ligated into the BamHI cloning site of the yeast vector pCS1 from which the 2 micron yeast DNA was removed by partial EcoRI digestion and religation. The resulting plasmid, termed pZZD-2 ~t, was linearised using a unique SmaI restriction site within the URA3 gene. The linearised plasmid was used to direct integration of the plasmid to the chromosomal ura3 locus. Plasmid pYCP50-1Z was introduced into strains BWG/-7a and LGW-1 by the lithium acetate method of yeast transformation (Ito et al. 1983). Plasmid pZZD2g was introduced into strains BWG1-7a, LGW-1, JO11A, SHY40 and SLF401 using the spheroplast method of Hinnen et al. (1978). Positive transformants were selected as Ura + and confirmed by testing for/?-galactosidase on X-Gal plates. Segregation tests were performed by comparing plasmid retention under selective and nonselective conditions. To check the integrity of transformants, plasmid DNA was re-isolated from each by the method of Sherman et al. (1982) and characterised by restriction analysis. Site-directed mutagenesis. The HAP2/HAP3 binding motif in the upstream region of the LPD1 gene was modified using the gapped duplex method of site-
298 directed mutagenesis (Kramer and Fritz 1987). The 1.47 kb B a m H I fragment of plasmid pZZD-2g containing the upstream region of L P D 1 was cloned into the BamHI site of M13mp9. Gapped duplex D N A was then constructed using BamHI-treated M 13mp9rev plasmid DNA. An oligonucleotide containing four mismatches to the HAP2/HAP3 binding element (5' C T G T T C C C G G A T C C G C G A G A 3') was hybridised to the gapped duplex DNA. Minor modifications to the method included an in vitro extension reaction using T7 D N A polymerase and the amplification step was replaced by plating aliquots of the extension reaction, after transfection into E. coli BHM71-18 m u t S cells, directly onto a lawn of E. coli MK30/3 cells (Gaata et al. 1990). The mutation was identified by plaque hybridisation, using 5'-32p-labelled mutant HAP2/HAP3 oligonucleotide. Nitrocellulose filters were washed at 20 ° C below the Tm calculated by the Wallace rule (Miyada and Wallace 1987). This allowed clear discrimination between plaques containing wild-type D N A and those containing the desired mutation. Further characterisation of the mutant was performed by dideoxy D N A sequencing using T7 D N A polymerase (Tabor and Richardson 1987). Both the wild-type and mutant L P D 1 promoters were then isolated from the M13mp9 vector using an EcoRI + HindIII digestion and cloned independently into plasmid YIp358 (Myers et al. 1986) using the E c o R I - H i n d I I I sites in the cloning region. The plasmids were termed pLC358 (wild-type promoter) and pLCH358 for the mutant promoter. These constructs, containing an in-frame lacZ fusion with the wild-type or mutant promoter, were then integrated at the ura3 locus of the wild-type strain (BWG1-7a). Enzyme assays and protein concentration. Yeast cells were
grown to a n 0 D 6 o o of 1.0 and assayed for lipoamide dehydrogenase activity using 2-acetylpyridine adenine dinucleotide (APAD) as substrate as described by Dickinson et al. (1986). Units are expressed as gmol APAD reduced/rain. Quantitative assays for fl-galactosidase acTable 2. HAP2/HAP3/HAP4 binding elements
tivity using o-nitrophenyl-fl-D-galactose (ONPG) as substrate were performed as described by Guarente (1983), and units are expressed as mmol O N P G hydrolysed/min. Protein concentration was assessed by the Bio-Rad assay method as described by the manufacturer. R N A analysis and Northern blots. For isolation of total
RNA, yeast strains were grown on YEPD or YEPGal to a n 0 D 6 o o of 1.0. Total R N A was isolated using the phenol-glass bead method of Lindquist (1981). For Northern blot analysis, 20 gg of total R N A in water was dried and resuspended in 4.5 gl of water, 2 gl MOPS buffer, 3.5 gl formaldehyde, 10 gl formamide and 1 gl ethidium bromide (400 ~tg/ml) then incubated at 65°C for 15 min. After the addition of loading buffer (50% v/v glycerol, 1 m M EDTA, 0.25% bromophenol blue, 0.25% xylene cyanol), the samples were separated electrophoretically on a 20 cm 1.0% agarose gel under nondenaturing conditions, photographed, then transferred to a nitrocellulose filter. Qualitative evaluation of transcript levels was estimated with the Ambis 2D Radioanalytic system (Lablogic Ltd.). Hybridisation probes. The following D N A fragments were radiolabelled by the random priming method (Feinberg and Vogelstein 1983) after isolation from agarose gels. A 3.6 kb XhoI fragment was isolated from pGP1 (Roy and Dawes 1987) and used to probe for L P D 1 mRNA. A 3.1 kb EcoRI fragment isolated from pMC1871 (Casadaban et al. 1983) was used as probe for lacZ mRNA. Nick-translated plasmid pSPACT9 (Bettany et al. 1989) was used as probe for actin mRNA.
Results
H A P 2 / H A P 3 binding consensus in the LPD1 gene
Analysis of the sequence 5' to the start codon of the L P D ! gene has identified an element at position - 2 0 4
Match to consensus 5'-TNA/GTTGGT-3'
Locus"
Gene product
Position of match relative to ATG
TGGTTGGT TCATTGGT TTATTGGT TGATTGGc c TTATTGGc c TAAcTGGT c TGATTGGc TCATTGGg TCATTGGc TCATTGGc TTATTGGc
CYC1 b HEM1 COX4 KGD2
Iso-l-cytochrome c 6-Aminolevulinatesynthetase Cytochrome oxidase subunit IV Dihydrolipoyl transsuccinylase
KGD1
2-Oxoglutaratedehydrogenase
LPD1
Lipoamide dehydrogenase
-=210 -=374 -605 - 357 -325 - 398 - 320 -253 - 204 - 497
-731
a The nucleotide sequences of the promoter regions of these genes are given in the following references: CYC1, HEM1, COJ(4 (Forsburg and Guarente 1988), KGDI (Repetto and Tzagoloff 1989), KGD2 (Repetto and Tzagoloff 1990), LPD1 (Ross et al. 1988) b A G ~ A transition at position -208 has been shown to increase UAS2 activity (Forsburg and Guarente 1988) ° The HAP2/HAP3/HAP4 consensus sequence appears on the noncoding strand
299 showing homology to part of the UAS2 of the yeast CYC1 gene (Guarente et al. 1984) and subsequently two other matches at - 4 9 7 and - 7 3 1 were found. The CYC1 UAS2 consists of two functionally distinct regions, each containing a binding site for a different transacting factor. Region 1 mediates the carbon source response while region 2 is required for maximal activity from region 1 (Forsburg and Guarente 1988). Region 1 has a close match to the consensus sequence T N A T T G G T which has been identified as the putative binding site for the H A P 2 / H A P 3 / H A P 4 complex. Homology to this region has also been reported for a number of other genes including COX4, H E M I (Forsburg and Guarente 1988), COX6 (Trawick et al. 1989), KGD1 (Repetto and Tzagoloff 1989) and KGD2 (Repetto and Tzagoloff J990), which are also regulated by the same H A P factors (Table 2). The LPD1 motif aligned with those of CYC1, COX4, HEM1, KGDI and KGD2 shows a number of other characteristics c o m m o n to the H A P 2 / H A P 3 / H A P 4 binding sites of these other genes. There is a GC-rich sequence of dyad symmetry, in the region immediately 3' to the consensus element showing some homology between CYC1, HEM1, COX4 and LPDI ; for the KGD genes at least one of the potential motifs has a similar GC-rich region. Other short regions of dyad symmetry are present in LPD1 as well as in CYCI and HEMI, which differ by sequence, but not in nature, from those of CYC1. These regions are important for the functional integrity of the H A P 2 / H A P 3 / H A P 4 binding site in CYCI (Forsburg and Guarente 1988) and may play a similar role in the other genes. A single base change in the CYCI UAS2 transforming it to CYC1 UAS2UP1, which conforms with the consensus, causes a 10-fold increase in UAS2 activity. The potential elements in LPDI differ by 1 bp from the consensus for the HAP2 binding site, but this involves substitution of a C in the last position, a change that is also found in the functional motifs of KGD1 and KGD2.
Effect o f hap mutations on lipoamide dehydrogenase activity under steady-state growth conditions The LPD1 gene is expressed at basal levels in cells grown on medium containing glucose. When cells are grown on carbon sources other than glucose the LPDI gene is derepressed. Partial derepression is seen in strains growing on media containing galactose or raffinose, while the LPD! gene is expressed at high levels when cells are grown on non-fermentable carbon sources such as lactate or glycerol. The level of derepression of the LPD1 gene can be observed by assaying lipoamide dehydrogenase (LPDH) activity under different growth conditions. To determine the role of the H A P 2 / H A P 3 / H A P 4 system on LPD1 expression, the specific activity of L P D H was assayed in the wild type BWG1-7a and in the strains bearing hap2, hap3 and hap4 mutations. The hap strains are unable to grow on non-fermentable carbon sources but can grow on media (YEPGal and YEPRaft) that cause partial derepression o f catabolite-repressed genes.
Table 3. Expression of the intact LPD1 gene assayed in hap2, hap3 and hap4 mutant strains Strain
BWGI-7a LGW-1 b JOl-la b SHY 40 SLF 401
Relevant genotype Wild-type
hap2 hap2 hap3 hap4
LPDH Activity a YEPD
YEPGal
YEPRaff
4.4 3.3 3.3 3.5 2.3
17.5 7.9 5.9 6.0 6.2
15.6 9.9 8.0 8.4 7.4
" LPDH activity refers to specific activity of lipoamide dehydrogenase (expressed as x 10 -3 gmol product per min per mg of protein). The values are the average of three sets of experiments. Values for duplicate assays differed from each by no more than 5% b The hap2 mutation in strain LGW-1 was a disruption, while that in JOl-la was a null mutation
Wild-type cells grown on Y E P D showed basal levels of L P D H enzyme activity (Table 3). When wild-type cells were grown on Y E P G a l or Y E P R a f f L P D H enzyme activity increased 4-fold and 3.5-fold respectively, over the basal levels in glucose. As the data presented in Table 3 shows, all four hap strains showed a reduced basal level of enzyme activity in glucose. Under partially derepressive conditions (YEPGal and YEPRaff), enzyme activity in the hap strains showed a lower level of induction, with enzyme activity increasing by only 1.7 to 2.7fold in galactose and by 2.4 to 3.2-fold in raffinose. The reduced level of L P D H enzyme activity under steady-state repressive, and partially derepressive conditions in all four hap mutants tested indicated that each of the H A P proteins exerts a measure of control on the synthesis of L P D H .
Comparison o f L P D I and CYCI expression .following derepression The regulation of LPD1 was monitored in conjunction with that of C¥C1 in wild-type and hap2 mutant strains during a switch from growth on repressive to a partially derepressive medium. To facilitate assay o f CYC1 gene expression, a C YCl-lacZ fusion plasmid (pYCP 150-1 Z) was transformed into wild-type and hap2 strains. Plasmid pYCP150-1Z (a gift from T. Pillar) is a centromeric vector tailored to express the lacZ reporter gene product from the CYC! promoter (Zitomer et al. 1987). Strains BWG1-7a/Z (HAP2) and LGW-1/Z (hap2) were the resultant transformed derivatives of strains B W G I - 7 a and L G W - I respectively. Strains BWG1-7a/Z and LGW-1/Z were grown on Y E P D medium containing 8% glucose to ensure that expression of CYCI and LPD1 was fully repressed. The strains were then transferred to fresh media containing 2% glucose or 2% raffinose to follow derepression of both genes. L P D H and /q-galactosidase activities were monitored at mid-log phase of growth and then 3 and 8 h after transfer to 2% glucose or 2% raffinose. Following a switch from 8% glucose to 2% glucose there was
300 8-
A YEPD
galactosidase activities (Fig. 1). On a switch to raffinose m e d i u m there was an a p p r o x i m a t e eight-fold derepression o f b o t h enzyme activities in the wild-type strain but only a small change in the hap2 m u t a n t (Fig. 1). M o n i t o r i n g o f expression o f the CYCI-lacZ fusion as a control confirmed that H A P 2 protein was acting normally in the strain used to give partial derepression o f the CYC1 gene u n d e r the a p p r o p r i a t e conditions. The parallel d a t a for L P D H enzyme activity indicates that the LPDI gene is also regulated, in part, by the H A P 2 protein, when cells are shifted f r o m a repressive to a partially derepressive g r o w t h condition.
.[ 13 YEPRaf .
:>6-
• BWG-TaZ [ ] LGW-1Z
*6 -r
8
C YEPD
D YEPRaf
.I
BWG-7aZ [] LGW-1Z []
2;6. ~. o
*6 i:J
i:i!!il
:i!!!i!i! iiiiiiiiiiiii
-~2,
0
3
8
0 Time
Fig. 1A-D. Catabolite derepression of LPDH activity following a shift of HAP2 and hap2 strains from glucose to raffinose as carbon source. Cells were grown to log phase on YEPD containing 8% glucose, washed and transferred at zero time to either YEPD (2% glucose; A and C) or YEPRaff (2% raffinose; B and D). Cell extracts were prepared before the shift and at 3 and 8 h after the shift, and subsequently assayed for LPDH enzyme activity (A and B) and #-galactosidase activity (C and D). The values are averages of triplicate assays and are normalised to the activity of each enzyme in wild-type cells grown on YEPD containing 8% glucose (2 x 10 -3 U/rag protein for LPDH and 0.017 U/mg protein for #-galactosidase)
a threefold increase in L P D H activity in strains with either the wild-type or hap2 b a c k g r o u n d , bringing the basal levels near to that observed under steady state g r o w t h conditions. This increase was paralleled in the expression o f CYC1, inferred f r o m m e a s u r e m e n t o f #-
HAP regulation at the level of transcriptional control
To further assess the role of the HAP binding proteins in regulation of the LPD1 gene, LPDI-lacZ fusions were constructed. A 1.47 kb region o f the LPD1 gene was ligated to the E. coli lacZ reporter gene via an inframe translational fusion. The 1.47 kb f r a g m e n t contains the u p s t r e a m p r o m o t e r region, the A T G start c o d o n and a putative d o w n s t r e a m activation sequence (DAS) o f the LPD1 gene. This construct was integrated into wildtype, hap2, hap3 a n d hap4 strains• Levels o f #-galactosidase activity in cells t r a n s f o r m e d with this plasmid reflect the transcriptional control and activation o f the LPD1 p r o m o t e r region. #-galactosidase assays were perf o r m e d on the t r a n s f o r m e d strains g r o w n u n d e r steadystate repressive ( Y E P D ) or partially derepressive (YEPGal) conditions. All strains g r o w n on Y E P D gave basal levels o f #galactosidase activity (Table 4). W h e n wild-type cells were g r o w n on Y E P G a l the p r o m o t e r was derepressed, showing a five-fold increase in #-galactosidase activity. However, only basal levels o f #-galactosidase were seen in hap strains g r o w n u n d e r the same conditions. These results are similar to those obtained with the L P D H assays (Table 3). To confirm that the H A P 2 protein regulates activity o f L P D H at the level o f transcription, the steady-state m R N A levels o f LPD1 and lacZ were examined in the
Table 4. Effects of hap2, hap3, and hap4 on the derepression of LPD1 Strain
Relevant genotype
#-Galactosidase specific activity a Expt 2
Expt 1
BWG1-7a LGW-I ~ JOl-la c SHY 40 SLF 401
Wild-type
hap2 hap2 hap3 hap4
% Wild type b
YEPD
YEPGal
YEPD
YEPGal
YEPD
YEPGal
1.23 1.15 1.08
6.60 1.98 1.26
2.60 2.40 1.83 2.36 2.65
9.60 3.10 2.20 2.83 3.20
100 92.5 79 90.5 98
100 32 21 29 27.5
1.12
1.85
1.20
1.40
" All strains were transformed with plasmid pZZd-2lx; strains were grown on YEPD (containing 2% glucose) or YEPGal (containing 2% galactose), fi-galactosidase was assayed as described by Guarente (1983). Data from 2 different experiments are presented; activities are reported as specific #-galactosidase activity (x 10 .3 gmol product per min per mg protein)
b Average activity from the two experiments, expressed as a percentage of wild-type levels The hap2 mutation in strain LGW-1 was a disruption, while that in JOl-la was a null mutation
301 YEPD
YEPRaff
YEPGal
N
'7,
~,
6
6
rn
I
I
I
I
A
~'~
!,~i~
- LPD1
B -
C
.~
.'~
~
CYCI-lacZ
- Actin
2.2 and 3.2-fold in wild-type cells grown in raffinose, compared to cells grown on glucose. The transcripts did not show such a change in hap2 cells. LPD1 transcript levels (normalised to their respective levels in wild-type cells on glucose) were derepressed about 5-fold in wildtype cells compared with hap2 cells grown on either galactose or raffinose medium. The same filter was then reprobed with a lacZ-specific probe (Fig. 2B). The CYCl-lacZ mRNA levels were shown to be regulated in a manner identical to LPD1 mRNA on each carbon source. The origin of other bands on this filter was unclear. RNA from a yeast strain devoid of lacZ sequences treated similarly did not give any background bands with the same probe (data not shown). These bands may be degradation products of the major lacZ transcript or readthrough transcripts from plasmid sequences. Relative LPDH and fl-galactosidase activities measured in cell samples used for mRNA assays showed no change in LPDH activities in the hap2 mutant and approximately 2- and 4-fold increases in activity in the wild-type strain on raffinose and galactose respectively. This indicated that enzyme activities were a reflection of the transcriptional regulation of each gene by the HAP2 protein.
D -25S -
H A P control is mediated mainly by the HAP2/HAP3 binding motif at - 2 0 4 in the control region of the LPD1 gene
18S
Fig. 2A-D. Regulation of LPD1 gene transcription by HAP2. Northern blots of total R N A (20 gg per lane), isolated from either the wild-type strain BWG1-7A/Z (relevant genotype: LPDI HAP2 CYCI-lacZ) or the hap2 mutant LGW-1/Z (relevant genotype: LPDI hap2 CYCI-lacZ) grown to mid-log phase on YEPD, YEPRaff or YEPGal are shown. Panels A, B and C display the same filter hybridized with probes for LPD1, lacZ and actin m R N A respectively. Panel D shows the gel stained with ethidium bromide
wild-type and hap2 strains grown on glucose and less repressing carbon sources. RNA was isolated from wildtype (BWG1-7A/Z) and hap2 (LGW-lZ) cells grown to mid-log phase on glucose (YEPD), galactose (YEPGal) or raffinose (YEPRaff) medium. The strains used carried the centromeric plasmid YCp150-1Z to facilitate comparison of LPD1 and CYC1 transcription levels. The RNA species were analysed by agarose gel electrophoresis, followed by Northern blotting. The Northern filter was initially hybridised with an LPD1 probe followed by an actin probe as a loading control (Fig. 2A and C). This assay showed little change in LPD1 mRNA levels in the hap2 mutant on all carbon sources tested. Relative amounts of LPDI, lacZ and actin mRNA were estimated using the Ambis apparatus. L P D I mRNA levels, normalised to those of actin were derepressed between
The putative HAP2/HAP3 binding motif at - 2 0 4 in the LPD1 gene shows strong sequence similarity to known functional HAP2/HAP3 binding motifs (Table 2) (Ross et al. 1988). To assess the role of this region in the control of LPD1 gene expression, site-directed mutagenesis was used to introduce 4 base changes in the HAP2/HAP3 binding element in plasmid pZZD-2 ~t, altering it from 5'-TCATTGGC-3' to 5'-GGATCCGC-3'. The mutant and wild-type promoters were fused in frame to the lacZ reporter gene and these constructs were integrated at the ura3 locus of the wild-type strain (BWGI-7a) thus retaining, in addition to the fusion, an intact copy of wild-type LPD1. The presence of wildtype LPD1 and H A P genes allow the strains to be grown on non-fermentable carbon substrates and subsequently assayed for fl-galactosidase activity. When grown on the non-fermentable carbon substrates glycerol or lactate (YEPG or YEPLact), the wildtype promoter was fully derepressed, increasing specific fi-galactosidase activity 10-fold and 12-fold, respectively, compared to the basal levels of enzyme activity of YEPD grown cells (Fig. 3). The mutant promoter gave the same basal level of activity as the wild-type promoter in repressing conditions and slight derepression on the partially derepressive carbon source galactose. However, when the strain bearing the mutant promoter was grown under fully derepressive conditions, specific fi-galactosidase activity increased only two- and four-fold. These results indicate that an intact HAP2/HAP3 binding motif at - 204 is required for full derepression of the LPD1
302 5P ._>
~3 *d
g2 i
(/3
YEPD YEPGa[ YEPG YEPkact Growth media
Fig. 3. Specific fi-galactosidase activity of the wild type LPD1 promoter-lacZ fusion pLC358 and the altered LPDI promoter-lacZ fusion pLCH358 integrated into the wild-type strain BWGI-7A at the ura3 locus. The altered promoter has 4 base changes in the HAP2/HAP3 binding motif at - 2 0 4 of the LPD1 gene, altering it from 5'-TCATTGGC-3' to 5'-GGATCCGC-3'. The resulting strains, denoted as wild type and altered, were assayed for fl-galactosidase activity under repressed, partially derepressed and fully derepressed conditions; YEPD, YEPGal and YEPG or YEPLact respectively
gene on non-fermentable substrates but that there may be a weaker contribution to full derepression in YEPLact from other elements located elsewhere in the promoter region.
Discussion The LPDI gene can be derepressed 50-fold in cells grown on lactate, 30-fold on glycerol, and 4- to 5-fold on galactose and raffinose relative to the level seen in cells grown on glucose. This derepression has been found to be mediated by the action of the HAP2, HAP3 and HAP4 proteins, since the levels of expression of the gene seen in hap2, hap3 and hap4 mutants are comparable with those seen on glucose medium. The majority of this effect is due to the HAP2 binding motif at - 2 0 4 in the upstream region of the gene since site-directed mutagenesis of this element leads to a very marked reduction in LPDI expression under strongly derepressing conditions. There is, however, some residual level of expression above the basal level in cells grown on lactate medium and carrying this mutant construct. This may be due to other elements within the control region of the gene. One candidate for such an element is another HAP2 motif at - 7 3 1 ; this is supported to some extent by the finding that a deletion from - 7 9 4 to - 6 9 7 does lead to a decrease in expression from an integrated lacZ fusion construct (J. Waterkeyn and I. Dawes, unpublished). The above results are similar to those seen for expression of the CYC1 gene on glucose and non-fermentable carbon sources. On glucose, CYC1 is regulated by the HAP1 activator protein binding to a site (UAS1) distinct from that (UAS2) of the HAP2, HAP3 and HAP4 complex. LPD1 does not appear to have a region showing
homology to the CYC1 UAS1. At present it is not known whether there is a sequence analogous to the CYC1 UAS1 involved in mediating the expression of LPD1 in glucose-grown cells. It would be interesting to determine whether HAP1 does regulate expression of LPD1 despite the absence of an obvious binding site, since Pfeifer etal. (1987) clearly demonstrated that HAP1 binds to a dissimilar sequence at the CYC7 promoter. The HAP2, HAP3 and HAP4 system has been found to regulate a number of other mitochondrial enzymes and may represent one of the main mechanisms for coordinating the derepression of these enzymes in response to changes in carbon status of the medium. The LPD1 gene encodes an enzyme subunit c o m m o n to the pyruvate dehydrogenase (PDH) and 2-oxoglutarate dehydrogenase ( O G D H ) multienzyme complexes. The other two genes encoding subunits of O G D H (KGD1 and KGD2) have been shown to contain HAP2-binding motifs, and these genes are regulated in a manner analogous to LPD1 (Repetto and Tzagoloff 1989, 1990). It is not surprising that common transcriptional regulators exist for the subunits of O G D H since it reflects at least one mechanism to ensure their coordinated response to growth on different carbon sources. It will be interesting to see whether the genes unique to P D H are controlled in the same way. There is, however, an additional dimension in the regulation of LPD1 since its product serves the two different complexes which, in E. coli at least, can be subject to some degree of differential control (Spencer and Guest 1985). This is reflected in the fact that the LPD1 gene in yeast has a number of other potential control motifs in its upstream region that are not found in the KGD1 and KGD2 genes (Dawes 1989). Acknowledgments. We wish to thank L. Guarente and his colleagues for kindly providing us with the hap mutants used. This work was supported by grants from the Science and Engineering Research Council and the Australian Research Council. Z.Z. was supported by an SERC Postgraduate Studentship and S.B. is the recipient of an Australian Postgraduate Research Award.
References Bettany AJE, Moore PA, Cafferkey R, Bell LD, Goodey AR, Carter BL, Brown AIP (1989) 5'-secondary structure formation, in contrast to a short string of non-preferred codons, inhibits the translation of the pyruvate kinase mRNA in yeast. Yeast 5:187-198 Casadaban MJ, Martinez-Arias A, Shapira SK, Chou J (1983) figalactosidase gene fusions for analyzing gene expression in Escherichia coli and yeast. Meth Enzymol 100:293-308 Dawes IW (1989) Complex regulation of gene expression in Saccharomyces cerevisiae and implications for biotechnology. Aust J Biotechnol 3 : 117-124 Dickinson JR, Roy DJ, Dawes IW (1986) A mutation affecting lipoamide dehydrogeanse, pyruvate dehydrogenase and 2-oxoglutarate dehydrogenase activities in Saccharomyces cerevisiae. J. Gen Microbiol 133:925-933 Entian KD (1986) Glucose repression: a complex regulatory system in yeast. Microbiol Sci 3 : 366-371 Feinberg AP, Vogelstein B (1983) A technique for radiolabelling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 132: 6-13
303 Forsburg S, Guarente L (1988) Mutational analysis of upstream activation sequence 2 of the CYC1 gene of Saccharomyces cerevisiae: a HAP2-HAP3-responsive site. Mol Cell Biol 8 : 647654 Forsburg SL, Guarente L (1989) Identification and characterization of HAP4: a third component of the CCAAT-bound HAP2/ HAP3 heteromer. Genes Dev 3 : 1166-1178 Gafita BA, Sharp SJ, Stewart TS (1990) Saturation mutagenesis of the Drosophila tRNA A~ggene B-box intragenic promoter element: requirements for transcription activation and stable complex formation. Nucleic Acid Res 18:1541-1548 Guarente L, Yocum RR, Gifford P (1982) A GALIO-CYC1 hybrid yeast promoter identifies the GAL4 regulatory region as an upstream site. Proc Natl Acad Sci USA 79:7410-7414 Guarente L (1983) Yeast promoters and lacZ fusions designed to study expression of cloned genes in yeast. Meth Enzymol 101:181-191 Guarente L, Lalonde B, Gifford P, Alani E (1984) Distinctly regulated tandem upstream activation sites mediate eatabolite repression of the CYC1 gene of Saceharornyces cerevisiae. Cell 36: 503-511 Hahn S, Guarente L (1988) Yeast HAP2 and HAP3: Transcriptional activators in a heteromeric complex. Science 240:317-321 Hinnen A, Hicks JB, Fink GR (1978) Transformation of yeast. Proc Natl Acad Sci USA 75:1929-1933 Ito H, Fukuda Y, Murata K, Kimura A (1983) Transformation of intact yeast treated with alkali cations. J Bacteriol 153:163168 Kramer W, Fritz HJ (1987) Oligonucleotide-directed construction of mutations via gapped duplex DNA. Meth Enzymol 154: 350 367 Lindquist S (1981) Regulation of protein synthesis during heat shock. Nature 293:311-314 Miyada CG, Wallace RB (1987) Oligonucleotide hybridization techniques. Meth Enzymol 154:94-107 Myers AM, Tzagotoff A, Kinney DM, Lusty CJ (1986) Yeast shuttle and integrative vectors with multiple cloning sites suitable for construction of lacZ fusions. Gene 45: 299-310 Olesen J, Hahn S, Guarente L (1987) Yeast HAP2 and HAP3 activators both bind to the CYC1 upstream activation site, UAS2, in an interdependent manner. Cell 51:953-961
Pfeifer K, Arcangioli B, Guarente L (1987) Yeast HAP1 activator competes with the factor RC2 for binding to the upstream activation site UASI of the CYC1 gene. Cell 49:9-18 Reed LJ (1974) Multienzyme complexes. Accounts Chem Res 7:4046 Repetto B, Tzagoloff A (1989) Structure and regulation of KGD1, the structural gene for the yeast c~-ketoglutarate dehydrogenase. Mol Cell Biol 9:2695-2705 Repetto B, Tzagoloff A (1990) Structure and regulation of KGD2, the structural gene for yeast dihydrolipoyl transsuccinylase, Mol Celt Biol 10:4221-4232 Ross J, Reid GA, Dawes IW (1988) The nucleotide sequence of the LPD1 gene encoding lipoamide dehydrogenase in Saecharomyces cerevisiae: comparison between eukaryotic and prokaryotic sequences for related enzymes and identification of potential upstream control sites. J Gen MicrobioI 134: l 131-1139 Roy DJ, Dawes 1W (1987) Cloning and characterization of the gene encoding lipoamide dehydrogenase in Saccharomyces cerevisiae. J Gen Microbiol 133:925-933 Sherman F, Fink GR, Hicks JB (1982) Laboratory course manual for methods in yeast genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York Spencer ME, Guest JR (1985) Transcriptional analysis of the sucAB, aceEF and lpd genes of Eseherichia coli. Mol Gen Genet 200:145-154 Tabor S, Richardson C (1987) DNA sequence analysis with a modified bacteriophage T7 DNA polymerase. Proc Nat Acad Sci USA 84:4767-4771 Trawick JD, Rugness C, Poyton RO (1989) Identification of an upstream activation sequence and other cis-acting elements required for transcription of COX6 from Saccharomyces cerevisiae. Mol Cell Biol 9:5350-5358 Verdier JM (1990) Regulatory DNA-binding proteins in yeast: an overview. Yeast 6: 271-297 Zitomer RS, Sellers JW, McCarter DW, Hastings GA, Wick P, Lowry CV (1987) Elements involved in oxygen regulation of the Saccharomyces cerevisiae CYC7 gene. Mol Cell Biol 7: 2212-2220 C o m m u n i c a t e d b y B.J. K i l b e y
