|Home | About | Journals | Submit | Contact Us | Français|
Two transcription factors, Oaf1p and Pip2p (Oaf2p), are key components in the pathway by which several Saccharomyces cerevisiae genes encoding peroxisomal proteins are activated in the presence of a fatty acid such as oleate. By searching the S. cerevisiae genomic database for the consensus sequence that acts as a target for these transcription factors, we identified 40 genes that contain a putative Oaf1p-Pip2p binding site in their promoter region. Quantitative Northern analysis confirmed that the expression of 22 of the genes identified is induced by oleate and that either one or both of these transcription factors are required for the activation. In addition to known peroxisomal proteins, the regulated genes encode novel peroxisomal proteins, a mitochondrial protein, and proteins of unknown location and function. We demonstrate that Oaf1p regulates certain genes in the absence of Pip2p and that both of these transcription factors play a role in maintaining the glucose-repressed state of one gene. Furthermore, we provide evidence that the defined consensus binding site is not required for the regulation of certain oleate-responsive genes.
In the yeast Saccharomyces cerevisiae, the levels of peroxisomal enzymes and the number and size of peroxisomes are increased when the yeast is supplied with a fatty acid carbon source for growth (60). We recently characterized two proteins, Oaf1p and Oaf2p, that act as positive regulators of genes encoding peroxisomal proteins (32, 39). Oaf1p is an oleate-activated transcription factor that was purified through its function of binding to an upstream activating sequence (UAS) in POX1, the gene that encodes peroxisomal acyl coenzyme A (acyl-CoA) oxidase in this yeast (39). Binding to this specific DNA sequence results in transcriptional activation of the gene. Using a genetic approach, we identified Oaf2p, a second transcription factor that is also required for the oleate induction of genes encoding peroxisomal proteins (32). The OAF2 gene was also identified by Rottensteiner et al., who named this gene PIP2 (46). Deletion of either the OAF1 or the PIP2 gene prohibits oleate-induced proliferation of peroxisomes and prevents the yeast from being able to grow on oleate as the sole carbon source. Oaf1p and Pip2p have an overall identity of 40%, with the highest homology occurring in the amino-terminal Zn2Cys6 DNA-binding motifs (32). The proteins form a complex and bind to a UAS in the form of a heterodimer (32, 47). DNA sequences to which this heterodimer binds contain palindromic CGG triplets separated by a 15- to 18-nucleotide spacer. This sequence is present in the promoter region of several genes encoding peroxisomal proteins, and it has been termed the oleate response element (ORE) (10, 13).
The S. cerevisiae Oaf1p- and Pip2p-dependent pathway that mediates the activation of peroxisomal proteins and peroxisome proliferation resembles an analogous system in higher eukaryotes. In mammals, an increase in peroxisome number and in expression of several peroxisomal enzymes is induced by feeding a high-fat diet (28) or a wide range of compounds that have collectively been termed “peroxisome proliferators” (43). This regulation occurs at the transcriptional level and is controlled by two proteins belonging to the superfamily of nuclear hormone receptors; the peroxisome proliferator-activated receptor (PPAR) and the retinoic acid X receptor (RXR). PPAR and RXR form a heterodimer and bind to DNA elements that contain a direct repeat of the sequence AGG(A/T)CA (57). Three PPAR subtypes have been characterized; PPARδ, which is ubiquitously expressed; PPARα, which is highly expressed in the liver; and PPARγ, which is enriched in adipocytes (50). Activation of PPARγ by 15-deoxyΔ12,14-prostaglandin J2 or a synthetic analog promotes differentiation of preadipocytes into fat cells (16), whereas PPARα mediates the transcriptional effects of drugs that induce peroxisome proliferation (29). Recent studies have shown that various fatty acids and hypolipidemic drugs directly bind to each of the PPARs, but preferentially activate PPARα (9, 14, 33, 35). Furthermore, inhibitors of various steps in the mitochondrial β-oxidation pathway also lead to activation of PPARα and peroxisome proliferation (2, 19, 23). Thus, PPARs appear to act as modulators of lipid homeostasis in higher eukaryotes.
Given the profound phenotypic effect in yeast mutant strains lacking either OAF1 or PIP2, we hypothesized that the Oaf1p and Pip2p transcription factors may play a more global role in regulating genes encoding proteins required for peroxisome function and biogenesis. In order to determine the full extent of the role of these two regulatory proteins, we have taken advantage of the recent completion of the S. cerevisiae genome sequencing project (18). We first compared the ORE sequences of eight genes encoding peroxisomal proteins that are known to be induced by oleate and found that only two of the spacer nucleotides are conserved. Based on an ORE consensus sequence derived from these genes, we searched the yeast genome for regions in which this sequence occurs within 500 nucleotides upstream from the initiating codon of an open reading frame (ORF). The expression of genes identified by this search was then measured in wild-type and oaf1Δ and pip2Δ mutant strains grown in the presence of various carbon sources. Here we describe the role of Oaf1p and Pip2p in the oleate activation of many known peroxisomal membrane and matrix proteins. Furthermore, we demonstrate that these transcription factors are also involved in the regulation of a mitochondrial protein and additional ORFs that encode proteins whose cellular location and function have yet to be defined. Thus far, we have determined that two of these ORFs encode proteins that are localized in peroxisomes.
In this study, we have resolved two different patterns of regulation for oleate-inducible genes. One, previously described for the peroxisomal β-oxidation genes, is Oaf1p and Pip2p dependent and requires the binding of these two proteins to the consensus ORE. A second pattern of regulation does not appear to require the consensus ORE as defined here and can be carried out by either Oaf1p or Pip2p in the absence of the other protein. Whether the transcription factors bind to an ORE or ORE-like sequence in the form of a homodimer or heterodimerize with an additional, as yet unknown, protein or whether activation of the Oaf1p and Pip2p proteins leads to induction of additional transcription factors remains to be established.
The yeast strains used in this study are described in Table Table1.1. Yeast strains were grown in either YPD (1% yeast extract, 2% peptone, 2% glucose), SD (0.67% yeast nitrogen base without amino acids, 2% glucose), YPG (1% yeast extract, 2% peptone, 3% glycerol), YPGO (0.1% [wt/vol] oleic acid and 0.25% [vol/vol] Tween 40 added to YPG), YPR (1% yeast extract, 2% peptone, 2% raffinose), YPRO (0.1% [wt/vol] oleic acid and 0.25% [vol/vol] Tween 40 added to YPR), or YPE (1% yeast extract, 2% peptone, 2% [wt/vol] ethanol). Auxotrophic supplements were added to 20 μg/ml (40 μg/ml in the case of leucine) as required.
Strains were precultured in YPD to the mid-logarithmic phase. Precultures were then used to inoculate YPD, YPG, YPGO, YPR, YPRO, or YPE at an optical density at 600 nm of ≈0.1 and grown to the mid-logarithmic phase. Total yeast RNA was isolated as previously described (39). Poly(A)+ fractions were prepared with Oligotex milk as specified by the manufacturer (Quiagene). The mRNA was resolved, transferred to a nylon membrane, and hybridized overnight as described previously (32). Gene-specific probes were generated by PCR amplification with primers based on sequence from the yeast genome database (Stanford) and yeast genomic DNA. The specific primer sequences used are available on request. The PCR products were resolved in a standard 1% agarose gel, purified with a Gene Clean kit (Bio 101), and labeled with a Prime-It RmT kit (Strategene) and [α-32P]dCTP. Following hybridization, the filters were extensively washed and then exposed to a Storage Phosphor Screen. The intensity of each band was determined with Molecular Dynamics Storm 860 software (Imagequant). Values were normalized by using actin (ACT1) mRNA expression levels as an internal control for loading. Prior to reprobing, the membranes were stripped with 500 ml of 0.2% sodium dodecyl sulfate at 98°C for 30 to 60 min. Expression of each gene was examined in at least two separate mRNA preparations.
To disrupt the YPL095c gene, a DNA fragment containing the entire gene was first amplified from yeast genomic DNA. The purified fragment was then subcloned into the PCR 2.1 TA vector, resulting in p95c. This plasmid was digested with ClaI, blunt ended, and dephosphorylated, and a 1.7-kb blunt-ended BamHI fragment containing the S. cerevisiae HIS3 gene was inserted, resulting in p95c::HIS3. The plasmid was digested with EcoRI, and the reaction mixture was used for transformation into S. cerevisiae W3031A. Selected clones were screened for correct integration by PCR analysis of total DNA isolated from the transformants. A strain carrying a deletion in YPL095c was named 95Δ and was used for further studies.
Standard procedures were used for cloning and for transformation of Escherichia coli and S. cerevisiae cells (48).
The Oaf1-Pip2 heterodimer mediates the oleate-dependent induction of POX1 and FOX3 by binding to a response element (ORE) present in the promoter regions of these genes (32, 47). Similar elements are found in the promoter regions of other genes encoding peroxisomal proteins in S. cerevisiae. By comparing this sequence in eight of these genes, we defined a consensus ORE to consist of the following nucleotides: CGGNNNTNAN9–12CCG (Table (Table2).2). With the help of the yeast curator of Stanford University, we carried out a search for the occurrence of this sequence within the yeast genome. In total, the consensus sequence occurs approximately 200 times. By selecting those genes or ORFs in which the ORE is located within 500 nucleotides upstream from the initiation codon, we identified the 40 genes or ORFs that are shown in Table Table3.3.
Thirteen of the genes shown in Table Table33 encode known peroxisomal proteins (shown in boldface). As predicted, these include the eight genes shown in Table Table2.2. Several of the genes identified have previously been shown to be induced by oleate, and in some cases, Oaf1p and/or Pip2p has been implicated in their regulation (10, 13, 24, 31, 46, 59, 61). We examined the expression of each of these genes in a wild-type yeast strain and in oaf1Δ and pip2Δ null mutants, as well as in a strain carrying deletions in both of these genes. The strains were each grown in the presence of glucose, glycerol, or glycerol and oleate.
Our results demonstrated that the three genes encoding peroxisomal β-oxidation enzymes (POX1, FOX2, and FOX3) are expressed and regulated in an identical fashion. Each transcript is repressed by growth in glucose medium, derepressed when the cells are grown in the presence of glycerol, and activated approximately 10-fold by growth in glycerol-oleate medium. The induction by oleate, but not the derepression in glycerol, requires Oaf1p and Pip2p (Fig. (Fig.1a1a to c). The peroxisomal proteins encoded by TES1 (31), SPS19 (24), and PEX11 (12, 40) are regulated in the same manner as the β-oxidation genes (Fig. (Fig.1d1d to f).
A different pattern of regulation was seen for FAA2, PEX5, MDH3, PXA2, YCAT, and IDP3, genes that encode peroxisomal matrix and membrane proteins. Each of these genes is repressed by growth in glucose to variable extents and is only moderately induced by oleate (two- to threefold) in our wild-type strain (Fig. (Fig.2).2). In the oaf1Δ and pip2Δ strains grown in the presence of oleate, expression of each of these transcripts is marginally higher than that in glycerol-grown cells (with the exception of MDH3 in oaf1Δ). However, in a strain carrying disruptions of both OAF1 and PIP2 (ΔO1P2), the expression of these genes in glycerol-oleate-grown cells is equal to or lower than the expression in cells grown in glycerol (Fig. (Fig.2).2).
Regulation of CTA1, the gene encoding peroxisomal catalase, appears to be unique among the genes described here. This gene is induced approximately fivefold in glycerol-oleate-grown cells compared to the expression in cells grown in glycerol alone, and the induction is dependent on Oaf1p but not Pip2p (Fig. (Fig.3).3).
Five of the remaining genes that are listed in Table Table33 (MAL33, UGA3, CUP2, RIM1, and PIP2) encode regulatory proteins. We and others have previously demonstrated that Pip2p (Oaf2p) is induced by oleate (32, 47). To determine whether induction of the PIP2 transcript requires the presence of OAF1, we performed a Northern analysis of total RNA isolated from our wild-type strain and from the oaf1Δ and pip2Δ strains. Our results clearly show increased expression of PIP2 mRNA in cells grown in the presence of oleate, and this induction is abolished in the absence of OAF1 (Fig. (Fig.4).4). These findings are in agreement with those from a previous report demonstrating that β-galactosidase activity expressed from a PIP2-lacZ reporter construct was induced by oleate, and this induction was absent in an oaf1Δ strain (47).
Expression of the remaining genes in Table Table33 that encode regulatory proteins was not induced by growth in the presence of oleate and was no different in the oaf1Δ and pip2Δ strains compared to that in our wild-type strain (data not shown).
Seven of the genes in Table Table33 encode nonperoxisomal proteins of known function. These include YAT1, encoding a mitochondrial acetyltransferase (49); TPS2, encoding trehalose phosphatase (7); HXK1, encoding hexokinase 1 (52); SER1, encoding serine aminotransferase (42); ATF1, encoding alcohol acetyltransferase (17); CIN1, encoding a protein required for microtubule stability (53); and CIT1, encoding citrate synthase (55). Of these, only the CIT1 transcript was induced by oleate and regulated by Oaf1p and Pip2p (Fig. (Fig.5a).5a). This result was somewhat surprising in light of the fact that CIT1 encodes a mitochondrial isoform of citrate synthase, whereas the CIT2 gene encodes a peroxisomal isoform (36, 45, 55). Increased transcription of CIT2, via a retrograde mechanism of regulation involving the RTG genes, occurs when mitochondrial function is impaired (5, 37, 38). Interestingly, these studies also demonstrated that the RTG genes are required for oleate-dependent peroxisome proliferation and that transcription of CIT2 is increased in an RTG1- and RTG2-dependent manner in cells grown in raffinose-oleate media (5). In order to determine whether Oaf1p and Pip2p are also involved in the regulation of CIT2, we compared the expression of this gene in wild-type and oaf1Δ pip2Δ strains grown in several different media. CIT2 is expressed at low levels in YPD-, YPG-, and YPGO-grown wild-type cells and at higher levels in cells grown in YPR. The expression is elevated when oleate is added to the raffinose medium (Fig. (Fig.5b).5b). In the oaf1Δ pip2Δ double disruption strain, CIT2 is expressed at similar levels to the wild-type strain in all media tested. Thus, in contrast to the mode of CIT1 regulation, Oaf1p and Pip2p appear to play no role in the oleate-dependent induction of CIT2.
Our finding that CIT1 is induced by oleate in an Oaf1p- and Pip2p-dependent manner demonstrates that regulation by these two transcription factors is not restricted to genes required for peroxisome biogenesis or function.
Our search for ORE-containing genes retrieved 15 ORFs that contain this sequence in their promoter regions. Seven of these genes are induced in cells grown in the presence of oleate via pathways that require Oaf1p or the Oaf1p-Pip2p heterodimer (Table (Table4).4).
We noted that two of the ORFs shown in Table Table44 (YLR284c and YOR180c) encode putative proteins that contain amino acid sequence motifs similar to the peroxisomal targeting signals 1 and 2 (PTS1 and PTS2, respectively) (20, 21). The transcripts for both of these ORFs are induced in the presence of oleate in a fashion similar to that of the genes encoding the peroxisomal β-oxidation enzymes, and this induction requires the presence of Oaf1p and Pip2p (data not shown). We have subsequently demonstrated that these ORFs encode novel peroxisomal proteins belonging to the enoyl-CoA hydratase family (unpublished data).
The transcripts expressed by YJL218w and YPL095c are also induced approximately 10-fold when cells are grown in the presence of oleate and are expressed at low levels in both glucose- and glycerol-grown cells (Fig. (Fig.6a6a and and6b).6b).
The YJL218w ORF encodes a putative 21.5-kDa protein predicted to contain at least one transmembrane domain. It has no close homolog in S. cerevisiae, but has similarity to galactoside O-acetyltransferase from E. coli. Experiments are in progress to determine the localization and function of the YJL218w protein.
The YPL095c ORF encodes a protein of 456 amino acids that has a predicted molecular mass of 51.7 kDa and an isoelectric point of 7.61. This protein has no close homology to any known protein family; however, a database search revealed a close yeast homolog encoded by the YBR177c ORF. The YPL095c protein contains no obvious localization signal. As a first step toward determining the function of this protein, the genomic copy of YPLO95c was replaced with HIS3, yielding the null mutant strain 95Δ (see Materials and Methods). 95Δ cells fail to grow on a nonfermentable carbon source (glycerol or ethanol), suggesting that mitochondrial respiration may be impaired in these cells. Experiments are in progress to further define the role of the YPL095c protein.
The remaining ORFs that are regulated by the Oaf1 and Pip2 proteins are each predicted to encode membrane proteins. The expression and regulation of each of these transcripts are somewhat different. YBR159w is expressed at approximately equal levels in glucose- and glycerol-grown cells and is induced two- to threefold in cells grown in the presence of oleate. This induction requires both Oaf1p and Pip2p (Fig. (Fig.66c).
The YIL120w ORF encodes a protein that contains 12 regions predicted to be transmembrane domains. The gene is expressed at very low levels in the presence of glucose and is induced approximately twofold by growth in glycerol-oleate compared to that in glycerol alone. This induction is abolished in an oaf1Δ strain, but it is still evident, although reduced, in a pip2Δ strain. Thus, regulation of this gene resembles the mode of CTA1 regulation (compare Fig. Fig.6d6d with Fig. Fig.33).
The YOL002c ORF is predicted to encode a protein that contains seven transmembrane domains. In our wild-type yeast strain, the YOL002c transcript is highly expressed in the presence of glucose and is expressed at lower levels in cells grown in glycerol or glycerol and oleate. In the absence of Oaf1p or Pip2p, the expression of this transcript is reduced in each growth medium tested (Fig. (Fig.6e).6e). These data suggest that the role of Oaf1p and Pip2p is not restricted to activating oleate-responsive genes, but appears to be more complex.
The YOR100c ORF is predicted to encode a protein of 327 amino acids that has similarity to mitochondrial carnitine acyltransferase proteins. While expression of this transcript is repressed by glucose in wild-type and oaf1pΔ and pip2pΔ strains, it is not repressed in a strain carrying disruptions in both of these genes (Fig. (Fig.6f).6f). This result further supports the idea that the function of Oaf1p and Pip2p is not restricted to mediating oleate-dependent induction, because they appear to also play a role in maintaining the glucose-repressed state of this gene.
Pat1p and Pat2p are peroxisomal membrane proteins that comprise the two halves of an ABC transporter required for the import of long-chain fatty acids into peroxisomes (25). The proteins are encoded by the PXA2 and PXA1 (PAL1) genes, respectively (51, 56). Both proteins are reported to be induced by oleate; however, while PXA2 contains a consensus ORE, PXA1 lacks this binding sequence in its promoter region. To determine whether the regulation of the Pat2 protein requires Oaf1p or Pip2p, we examined the expression of the PXA1 transcript in our wild-type and oaf1 pip2 null strains. We found that PXA1 is induced approximately twofold in the presence of oleate, and this induction is abolished in the oaf1Δ pip2Δ strain (Fig. (Fig.7).7). Thus, moderate increases in PXA1 transcription in cells grown in the presence of oleate resemble the increased expression of the genes described in Fig. Fig.2.2. These data suggest that while such regulation requires either Oaf1p or Pip2p, it differs from the mechanism by which highly induced peroxisomal proteins (e.g., β-oxidation enzymes) are activated, since it does not necessarily require a consensus ORE.
We have taken advantage of the completed yeast genome database to search for genes containing a consensus binding site for the transcription factors that are responsible for mediating induced expression of genes encoding peroxisomal proteins. This search identified 40 genes that have oleate response-like elements in their promoter regions (Table (Table3).3). Twenty-two of the transcripts expressed from these genes are induced in cells grown in the presence of oleate, under the control of Oaf1p and Pip2p. These include genes encoding known, as well as newly identified, peroxisomal proteins in addition to genes encoding a mitochondrial protein and proteins of unknown location and function. mRNA expression of 11 of these genes was approximately 10-fold higher in glycerol- oleate-grown cells than that in cells grown in glycerol alone, whereas induction of the remaining genes was only 2- to 3-fold.
The OREs in the regulated genes are localized on either the Watson or the Crick DNA strand, and they are not identical (Table (Table5).5). However, nucleotides at certain positions appear to be conserved. Following the first CGG, positions 1 and 2 are variable, position 3 is A/T, position 4 is T, position 5 is A/T, and position 6 is A. Nucleotides at all other positions are totally random, except for those at positions 15 and 17, which are A, T, or C. There are four regulated genes with OREs that do not conform to this consensus (footnote b in Table Table5).5). It is possible that there are additional ORE-like elements that act as targets for the Oaf1-Pip2 heterodimer that would not be retrieved in our search. For example, there may be some variability at positions 4 and 6, or the number of nucleotides in the spacer may be less than 9 or greater than 12. Nevertheless, the defined consensus ORE sequence that we used in this search enabled us to identify 13 genes encoding known peroxisomal proteins, as well as at least two ORFs that encode novel peroxisomal proteins, all of which are induced by oleate and regulated either by Oaf1p or by the Oaf1p-Pip2p heterodimer.
Four of the 18 genes that are not regulated by Oaf1p and Pip2p contain putative OREs that conform to the consensus sequence described above (Table (Table5).5). This finding, together with the fact that certain regulated genes do not contain the consensus ORE as defined here (e.g., PXA1), implies that there must be additional factors that determine whether this sequence acts as a functional regulatory element. It is likely that the flanking sequences around the putative ORE influences the selectivity or strength of Oaf1p and Pip2p binding. The fact that the nucleotides in the flanking region upstream of the PPRE (the binding site for the PPAR-RXR heterodimer) play an important role in the strength of binding, and thus in the level of transcriptional activation of PPAR-regulated genes (30), supports this hypothesis. In our own experiments we have found that the flanking regions of the POX1 ORE are crucial for its activity, since the introduction of XhoI sites on either side of the ORE abolished its function (unpublished results).
Several peroxisomal proteins that are reported to be induced by oleate are encoded by genes that do not contain an obvious ORE consensus sequence. One such example is the PXA1 gene (also called PAL1), which encodes Pat2p, a peroxisomal ATP-binding cassette transporter protein (51, 56). β-Galactosidase activity expressed from a PXA1-LacZ fusion is induced 10-fold in cells grown in the presence of oleate compared to that in cells grown in glycerol (56). Our results show that expression of PXA1 in cells grown in the presence of oleate is approximately twofold greater than that in cells grown in glycerol and appears to be regulated in a similar manner to those of several other genes encoding peroxisomal proteins that do contain the consensus ORE (compare Fig. Fig.77 with Fig. Fig.2).2). This finding raises the possibility that the consensus ORE defined here is only critical in those genes that are highly induced by oleate (10-fold) and that require both Oaf1p and Pip2p for their induction. The manner in which the genes that are marginally induced by oleate are regulated remains unclear. For these genes, the presence of Oaf1p or Pip2p alone appears to be sufficient for induced expression by oleate, whereas in the absence of both of these proteins, the induction is abolished. Whether the transcription factors bind to the ORE or an ORE-like sequence or whether they bind to an alternative, as yet unidentified, DNA sequence remains to be determined. A further possibility to consider is that activation of the Oaf1 and Pip2 proteins may lead to induction of an additional factor(s) that may bind to a different cis-acting element in the PXA1 promoter and thus mediate transcriptional activation.
The manner in which CTA1, the gene encoding peroxisomal catalase, is regulated differs from that of the other genes encoding peroxisomal proteins described thus far. Oleate-dependent induction of CTA1 requires Oaf1p, but not Pip2p (Fig. (Fig.3).3). This finding is in agreement with our previous data obtained with a strain carrying a mutation in the PIP2 (OAF2) gene (32). It has also been demonstrated that catalase A activity in oleate-grown cells is higher in a pip2Δ strain than in an oaf1Δ or oaf1Δ pip2Δ strain (47). Together these results suggest that Oaf1p alone is able to mediate the transcriptional activation of CTA1, perhaps in the form of a homodimer or complexed with additional as yet unidentified protein(s). The presence of Pip2p appears to enhance this induction. YIL120w, one of the ORFs identified by our search, is regulated in a fashion similar to that of the CTA1 transcript (see below).
In this study, we demonstrated that expression of the gene encoding mitochondrial citrate synthase is stimulated by the Oaf1-Pip2 heterodimer in response to oleate. To our knowledge, this is the first example in yeast in which the transcription factors required for peroxisome proliferation also regulate expression of a mitochondrial protein. However, such examples have been reported in higher eukaryotes. The PPAR-RXR heterodimer mediates induction of mitochondrial acyl-CoA dehydrogenase in rat liver and in cultured cells when the enzyme carnitine palmitoyl transferase I (CPT-I) is inhibited (23). In addition, PPAR and RXR activate the mitochondrial 3-hydroxy-3-methylglutaryl (HMG)-CoA synthase gene in response to the hypolipidemic drug clofibrate or to fatty acids, such as linoleic acid and oleic acid (44). Thus, these mammalian transcription factors play a more global role in regulating lipid metabolism than was initially recognized.
By implementing the approach described here to identify Oaf1p- and Pip2p-regulated genes, we retrieved seven novel ORFs that are induced by oleate. It is now of great interest for us to determine the subcellular location and function of these ORFs. Thus far, we have demonstrated that YLR284c and YOL180c encode novel peroxisomal proteins (unpublished data). Furthermore, our initial results suggest that the YPL095c protein is required for normal mitochondrial function.
In addition to their role in activation of oleate-responsive genes, we obtained evidence that Oaf1p and Pip2p are also involved in maintaining the inactive state of the glucose-repressed YOR100c gene. In contrast to the requirement of both of these proteins for oleate-dependent transcriptional activation of this gene, the presence of either Oaf1p or Pip2p is sufficient for maintaining its glucose-repressed state. One possible explanation for this mode of regulation would postulate that either one of these proteins is capable of binding, either as a homodimer or complexed with an additional factor(s), to a repression element in the YOR100c promoter when cells are grown in the presence of glucose. When the cells are shifted to an oleate medium, Pip2p is induced, which may in turn cause the Oaf1p-Pip2p heterodimer to predominate and bind to the ORE, thus activating transcription of the gene. Alternatively, regulation of YOR100c may represent an example in which transcription factors can act as repressors or activators while bound to the same regulatory element, a phenomenon that has been described for some members of the steroid hormone superfamily of nuclear receptors. For example, heterodimers of the thyroid hormone receptor and the retinoic acid receptor can activate or repress transcription of target genes, depending on the presence or absence of ligand (15, 22). It was postulated that negative control in the absence of ligand would greatly increase the specificity of gene regulation for response elements with overlapping receptor specificity (22).
In summary, our search for genes containing a consensus DNA-binding site for the Oaf1p-Pip2p heterodimer has led to the identification of 22 genes that are regulated by one or both of these transcription factors. In addition, subsequent experiments have revealed that the mechanism by which Oaf1p and Pip2p mediate transcriptional regulation depends on the nature of the target gene.
We acknowledge and thank Murl Casey for excellent technical assistance. We also extend thanks to the yeast curator at Stanford University for assistance with the yeast genome database search and to Joel Lopez for assistance with the Northern analysis of YOL002c.
This research was supported by American Heart Association grants 95008910 and 92001690 and by NIH grant R55DKOD51992.