Nucleic Acids Research, Vol. 20, No.

k.) 1992 Oxford University Press

13 3495-3500

A negative regulating element controlling transcription of the gene encoding acyl-CoA oxidase in Saccharomyces cerevisiae T.W.Wang, A.S.Lewin' and G.M.Small2* Department of Anatomy and Cell Biology and 'Department of Immunology and Medical Microbiology, University of Florida, Gainesville FL 32610 and 2Department of Cell Biology/Anatomy, Mount Sinai School of Medicine, New York, NY 10029, USA Received January 21, 1992; Revised and Accepted May 7, 1992

ABSTRACT Peroxisomes are induced in Saccharomyces cerevisiae when this yeast is grown in the presence of oleate, and are repressed when glucose is supplied as the carbon Concomitant with this is an source. induction/repression of peroxisomal ,B-oxidation enzymes. We are investigating the transcriptional control of acyl-CoA oxidase, the first and rate-limiting enzyme in the peroxisomal (-oxidation cycle. The promoter region of POX1 from S.cerevisiae has been analyzed in POX111acZ fusions. Expression of the POX111acZ fusion protein underwent glucose repression and oleate induction. By deletion, DNA band shift and DNase I footprinting analyses we have identified a region that is involved in transcriptional repression of POX1. Elimination of this DNA sequence results in constitutive expression of POX1 when S.cerevisiae is grown on a fermentable carbon source or glycerol. INTRODUCTION Peroxisomes of eukaryotic cells are a major site of lipid metabolizing enzymes. They contain enzymes that are involved in fatty acid (3-oxidation (1), plasmalogen synthesis (2), cholesterol synthesis (3) and bile acid formation (4). The number of peroxisomes per cell varies with tissue types; in mammals hepatic and renal tissue contain greater numbers and larger sized peroxisomes than other tissues. Within these cells peroxisomes are induced by assorted stimuli which include high fat diets (5), cold environment (6) and fasting (7). Over recent years a structurally diverse range of compounds have been found to cause a massive increase in the number of peroxisomes in hepatocytes of many mammalian species (8). This increase is accompanied by an induction of several peroxisomal enzymes (9). Peroxisome proliferators include hypolipidemic drugs, pesticides and industrial plasticizers. All of these peroxisome proliferators have been found to be carcinogenic in rodents and some higher

*

To whom correspondence should be addressed

mammals (e. g. monkeys) (10, 11). Due to the widespread use and importance of these chemicals there is considerable interest in understanding the mechanisms of peroxisome proliferation and regulation of peroxisomal enzymes. Peroxisome biogenesis is unique in that all peroxisomal proteins are synthesized in the cytosol and are transported into peroxisomes posttranslationally. The organelles then grow and divide to form new peroxisomes (for reviews see (12, 13)). Yeasts provide a model system for studying peroxisome induction. In several yeasts, including Candida tropicalis and Saccharomyces cerevisiae, peroxisomes are repressed when the yeasts are grown on glucose and are induced following growth on fatty acid substrates. The level of induction is extremely high in Candida species, such that peroxisomes can become the major cell constituent. Concomitant with this induction is an increase in the activity of peroxisomal enzymes, especially those involved in peroxisomal fatty acid oxidation (14, 15). The mechanisms of peroxisome proliferation and enzyme induction are poorly understood. Experiments in rats indicate that syntheses of fatty acyl-CoA oxidase and the bifunctional enoylCoA hydratase/3-hydroxy-CoA dehydrogenase, the first two enzymes of the peroxisomal (-oxidation system, are regulated at the transcriptional level in a coordinated fashion during peroxisome proliferation (16). In order to gain a better understanding of the regulation of peroxisomal enzymes we are studying the transcriptional regulation of POXI, the gene encoding peroxisomal acyl-CoA oxidase in Saccharomyces cerevisiae (17). In this yeast fatty acid oxidation takes place solely in peroxisomes (18). Fatty acyl-CoA oxidase catalyzes the first reaction in this cycle, and it is also the rate-limiting enzyme. Because of the similarity of transcription factors in S. cerevisiae and humans (19), it is likely that the control of transcription of this gene will reflect the transcriptional control of mammalian peroxisomal proteins. In the present study we have mapped the 5' region of POX], and have used deletion analyses, DNA band shift assays and DNA footprinting to identify an upstream repression sequence (URS).

3496 Nucleic Acids Research, Vol. 20, No. 13 MATERIALS AND METHODS Yeast strains and culture conditions Saccharomyces cerevisiae strain BWGI-7a (MATa adel-100 his4-519 leu2-3 leu2-112 ura3-52) was used throughout these studies. Rich medium for yeast cultures was 2% yeast extract, 1% peptone and 2% glucose (YPD). Selective medium was 0.67% yeast nitrogen base without amino acids, 0.1% yeast extract, and either 0.5% galactose (NEGal) or 0.5% glucose (NEGlu). Glycerol medium contained 1 % yeast extract, 2% peptone and 3 % glycerol (YPG). Appropriate amino acid supplements were added as needed. Induction medium (YEOG) contained 0.3% yeast extract, 0.5% peptone, 0.5% potassium phosphate pH 6.0, 0.05% galactose, 0.1% oleic acid and 0.2% Tween-80. Preparation of yeast extracts Yeast cultures were grown in NEGal medium to stationary phase and were then transferred either to NEGal, NEGlu, YPG or YEOG. Cells were then grown for the appropriate time according to the experiment and were harvested by centrifugation. The cells were washed with sterile water and resuspended in 10 mM Tris-HCI pH 7.5 containing 0.1 M PMSF. The cells were disrupted by vortexing in the presence of glass beads (0.5 mm diameter), and cell debris was removed by centrifugation. Plasmid constructions (i) pP13570. A 2376 bp Bgll fragment containing the POX] gene was excised from pAD17 (17) and was inserted into the BamHI site of pGem3 (Promega Biotec) to create pGem3-POXl. A 1060 bp KpnI-PstI fragment, that contained 454 nucleotides 5' of the coding region of POX] was excised from pGem3-P1 and inserted into the corresponding sites in the polylinker of YIp357 to create pP13570 (Figure 1).

(ii) pP13571. To obtain a truncated 5' region of POX] in frame with lacZ, pP13570 was digested with EcoRI and PstI to release a fragment that contained 279 non-coding bases and 602 coding bases of POX1. This fragment was inserted into the EcoRI -PstI sites of YIp357 which resulted in a in-frame fusion between the POXI coding region and lacZ. (iii) pP1353. In order to introduce point mutations into the 5' region of POX] a PCR approach was taken. For this two primers were synthesized as shown: primer 1: 5' GGGTClTTTAATAAAATTAACCCTA 3',(mut2 in Figure 8); primer 2: 5' CTGCAGCTGGGCAACATTGGA 3' (coding region of POX]). Four mutations were introduced into primer 1 (underlined). The resulting 1036 bp PCR fragment was treated with DNA polymerase I to render the ends blunt, and was then digested with HindU. The resulting 783 bp fragment was gel-purified and cloned into the SnaI-Hindm sites of YIp353, which resulted in an in-frame fusion with lacZ as described above. Primer extension Total RNA was isolated using the guanidinium thiocyanate method as previously described (20). RNA from Saccharomyces cerevisiae, grown in YEOG, was annealed to a radiolabeled primer (5'CAGAACCACCGAATCGGGATTA 3') which hybridized to the POX] coding bases 21 to 43. The primer was extended with reverse transcriptase and the transcript was analyzed on an 8% sequencing gel. For molecular weight

Cot pGEM3-POXI wN KprU nd PSn P13S70

PGEM3-POXi

pGEMS~~~~~~~~~~~~~~~~~

KC

P' ,lftKP

X $b f I p~ ~~~~~~A 1ia, KA Z rg n

"""'ATO

I

pP1370 POXI hin pADI' (10kb)

Figure 1. Preparation of pP13570. See text for details. Restriction sites are as follows: B, BglII; Bm, BamHI; K, KpnI; P, PsM.

standards, we used DNA sequencing reactions of POXJ that were generated using the same primer. Transformation Standard protocols were used for transformation of E.coli and yeast (21). The POXJ::lacZ fusion constructs were integrated at the URA3 locus by first digesting the YIP plasmids at a unique NcoI site. Selection for uracil prototrophs was carried out on minimal medium supplemented with the appropriate amino acids. Assays ,3-galactosidase assays were by the method of Craven et al. (22). Acyl-CoA oxidase was measured as described previously (23). Protein concentrations were determined by the method of Bradford (24). DNA band shift assays The DNA band shift assays were based on the procedures of Fried and Crothers (25) and Gamer and Revzin (26). Cellular extracts (1-5 ,ul) were mixed with 1 1l of 32P-end-labeled DNA fragment in a binding buffer composed of 12 mM Hepes pH 7.5, 60 mM KCI, 5 mM MgCl2, 4 mM Tris, 0.6 mM EDTA, 0.6 mM DTT, 10% glycerol, 0.26 tg/4l poly(dI-dC) and 0.3 tg/4l BSA. The binding reactions (total 20 Id) were incubated at room temperature for 20 min and were then separated on a 5% polyacrylamide gel at 100 V. The DNA fragments used for band shift experiments were excised from pP13570. The 174 mer was excised as an EcoRI fragment and was labeled with 32P-dATP using the Klenow fragment of DNA polymerase I. The 184 mer was amplified from pP13570 using the polymerase chain reaction using the following primers: A) 5'GGTACCGCGTTCAAACCCTGA 3'; B) 5'GCATGCAACACTCTTAAAGCC 3'. This amplified a 5' flanking region of POX] from -165 to -349 which was labeled with 32P-ATP using T4 DNA kinase. Oligonucleotides were synthesized on an Applied Biosystems DNA synthesizer model 380B. Complementary oligonucleotides were annealed together by heating them to 65°C and allowing them to slowly cool to room temperature.

DNase I footprint protection assay The 174 mer was 5' labeled, digested with KpnI to render it labeled at one end, and was then gel-purified. The assay was performed in a total volume of 25 Al containing 18 A1 of binding

Nucleic Acids Research, Vol. 20, No. 13 3497 a-ft we

x

dftft

.66

womb

1--ft

!a

I-ftft

T A

TS 1 74mer

POX1 5'

V J -454

J

7

-1--F

J

ATG

184mer

C

A ..'.

ml

-454

P 1357 1

C

52

lacZ

P 13570

C

A

lacZ -280

T G A C A

A

POX 1

coding region

C

T A A T A

Figure 3. POXJ::lacZ fusions. Regions of POX] used to prepare lacZ fusions, and nucleotides used in gel shift experiments. TS = transcriptional start site.

A

G T A T

Table 1. Endogenous acyl-CoA oxidase and ,B- galactosidase activities of strains harboring the constructs shown in Figure 3. Carbon source

Endogenous AOx mU/mg

(3-galactosidase mU/mg pP13570 pP13571 pP1353

Glucose Galactose Glycerol Oleate

U.D. .144 .815

U.D. .0009 .009 .366

A RT

G

T

A

C

Figure 2. Primer extension analysis of the POX] transcript. Lane 1: extension products obtained after an oligonucleotide complementary to the 5' end of the POX] gene was hybridized to 25 Ag of RNA from S.cerevisiae; lanes 2-5: DNA sequencing reactions of the POX] 5' non-translated region using the same primer as was used for the reverse transcriptase. Arrows indicate the position of the predominant extension products. The DNA sequence of the POX] 5' region is shown at the side of the gel. The bases that correspond to the 5' end of the transcripts are indicated by arrows and the nucleotide number relative to the bases that encode the initiating methionine.

buffer (with 5 ,tg poly(dI-dC)), 1 1l (3 ng) of end-labeled probe, 5 Al of yeast cell extract and 1 /d of the appropriate amount of DNase I depending on the experiment. The reaction was carried out at room temperature for 2 min before adding 40 ,tl of stop solution (200 mM NaCl, 20 mM EDTA, 1 % SDS and 250 jg/ml tRNA).

RESULTS Mapping the 5' end of POX] We used primer extension analysis in order to map the 5' end of the mature transcript of the Saccharomyces cerevisiae POX] gene. Figure 2 shows that the main 5' ends are 52 and 56 nucleotides upstream from the translational start site of POX]. Multiple transcriptional start sites is a common feature for yeast genes (27). This may explain why our result is somewhat different to that of Dmochowska et al. who reported two major 5' ends of POX] at 55 and 69 nucleotides upstream from the start of

translation (17).

Regulation of POXI::lacZ constructs The endogenous acyl-CoA oxidase activity of S. cerevisiae undergoes glucose repression and oleate induction (28). Therefore we prepared fusion constructs between the 5' region of the AOx gene from this yeast (POXJ) and the lacZ gene from E. coli. Two POXI::lacZ fusions were prepared, pP13570 and pP13571

.09 .089 .166 .86

.033 .110 1.18

Figures are the mean of three experiments. Stris were grown in minimal medium supplemented with glucose, galactose, glycerol or oleic acid. U.D. no detectable activity, - = not done.

(Figure 3). Both constructs were transformed into S. cerevisiae and 3-galactosidase activities were measured in extracts from cells grown in different media. S.cerevisiae transformed with pP13570, which contains 454 nucleotides of 5' sequence, was grown on glucose, glycerol or oleate. The 13-galactosidase activity was undetectable in extracts from glucose-grown cells, whereas measurable amounts of activity were present in extracts from cells grown on glycerol. The highest activity was measured in extracts from cells grown in media containing oleate as the main carbon source (Table 1). The glucose repression and oleate induction observed with this construct is similar to that of endogenous acylCoA oxidase activity in S. cerevisiae (Table 1). This suggests that the 5' flanking region of POX] present in this construct (454 bp) contains the necessary sequences for regulation by glucose and oleate. However, the ,B-galactosidase activity of cells grown in glycerol is lower than the acyl-CoA oxidase activity. This could indicate that there are further upstream elements in POX], that are not present in this construct, that increase expression of acylCoA oxidase in uninduced conditions. Cells transformed with the deletion construct pP13571, which contains only 279 5' nucleotides had higher j3-galactosidase activities than those transformed with pP13570 in each of the growth media (Table 1). Glucose repression was abolished in cells transformed with the truncated construct, suggesting that there is at least one negative regulating element located within the deleted 174 nucleotides. We believe that there is an oleateresponsive activating element located within the 279 5' nucleotides present in pP13571 because cells transformed with this construct had elevated ,B-galactosidase activity when grown in oleate media.

3498 Nucleic Acids Research, Vol. 20, No. 13

r

i

_...

tj

ww

wWWw (.w.

~ ~ ~ ~ ~ ~ ~ ~ ~ ~W~~~~~~~~~

j j V .

i,.A r

ILA1

1111-

Figure 4. DNA band shift analysis of the 5' flanking region of POX]. (a) A 174 bp radiolabeled DNA fragment (-280 to -454) was incubated with yeast extracts containing 30 itg of protein from glucose-grown (lanes 1 and 2) or galactose-grown (lanes 3 and 4) cells. The assay was carried out in the presence (lanes 2 and 4), or absence (lanes 1,3 and 5) of 150 ng of non-labeled 174 mer. Nucleoprotein complexes (c) were resolved from free DNA (f) by nondenaturing polyacrylamide gel electrophoresis and were revealed by autoradiography. (b) lanes 1-4: labeled 174 mer in the presence of 10 ,sg of glucose-grown (lanes 1 and 3) or galactose-grown (lanes 2 and 4) cell extract. Samples in lanes 3 and 4 were prepared in the presence of 150 ng of unlabeled 184 mer (-165 to -249). Lanes 5 in both (a) and (b) contained labeled DNA in the absence of protein.

oi

Figure 6. Competition of band shift with a 25 base oligonucleotide. The DNA band shift assay was carried out in the presence of extract from glucose-grown cells as described in Figure 4. The assay was performed in the absence (lane 2) or presence (lanes 3-7) of increasing amounts (1-25 ng) of double stranded oligonucleotide (for details see text). Assays were also carried out in the presence of 1 ug of single stranded oligonucleotide top strand (lane 8) or bottom strand (lane 9). Lane 1 is a control in the absence of cell extract.

w

sit iI '

Ii,.IW W.-

Figure 7. A 25 nucleotide fragment contains the cis-element. Lane 1 serves as a control and contains the radiolabeled 25 mer in the absence of cell extract. Lane 2 contains the labeled DNA in the presence of 25 isg of glucose-grown cell extract. Lanes 3 -8 are as for lane 2 in the presence of increasing amounts (10 ng to 1 4g) of unlabeled 25 mer. Lanes 9 and 10 are the same as lanes 1 and 2 respectively in Figure 6.

.;

A;

i,

R ii

Figure 5. DNase I footprint protection assay. The 174 mer was end labeled as described (materials and methods). Lanes 1-3: digestion of the free DNA with increasing amounts of DNase I, lanes 4 and 5: 2 concentrations of DNase I used to digest radiolabeled DNA in the presence of extacts from glucose grown cells. The region of protected nucleotides are indicated by a bracket. Lanes 6-9: DNA sequencing reactions of POX], used as a size marker.

Specific protein-DNA interactions correlate with the regulation of POXI expression The 5' region of POXI in pP13570 is 174 nucleotides longer than that in pP13571. Presence of the extra sequence correlates with regulation by fermentable carbon sources and glycerol. To test for a protein binding site in this region, the 174 bp fragment

was isolated, labeled and used in DNA mobility shift assays. As shown in Figure 4a, the 174 bp fragment gave rise to a positive band shift with extracts derived from glucose- and galactosegrown yeast. The unlabeled 174 mer was an effective competitor in the band shift assays (Figure 4a, lanes 2 and 4). A 184 mer (bases -165 to -349, Figure 3) did not compete in similar assays (Figure 4b). These results suggest that a protein or proteins present in these cells binds to the 174 bp fragment and represses transcription from the POXI promoter.

Mapping the binding site of the repressive element by DNase I footprinting The binding site of the repressive element of POXJ was pinpointed by DNase I footprinting. We detected protection by a crude extract of S. cerevisiae cells that were grown in NEGlu medium. The amount of protein used (50 rig) was that found sufficient to bind all of the probe in band shift experiments. The protected sequences are bases -434 to -413 of the POX] 5' region, indicated by a bracket in Figure 5. In order to confirm that this sequence is a binding site a double stranded 25 bp

Nucleic Acids Research, Vol. 20, No. 13 3499 a)

25 mer

5' GGTAGGGTA.AAAAATTAACCCTA 3'

mut 1:

5' TTTG --- 3

mut2:

5 -

mut3:

5 ---------GCCG-----

mut4:

5-------------CCCC----------- 3'

b

CTTT-------------------- 3' ----

3

NW

a

b)

1

2 3 4 5 6

7 8 9 1011 12

Figure 8. Oligonucleotide mutagenesis and DNA bandshift competition assay to precisely map the cis element. (a) Top strands of double-stranded oligonucleotides are shown. The 25 nucleotides represent the region of POX] identified from the DNase I footprint assay. Mutl-mut4 represent mutated versions of the 25 mer that were used in a competition assay (the mutated bases are shown). Nucleotides that are underlined indicate the region of DNA-protein interaction. (b) Competition of DNA band shift using mutl-mut4. The DNA band shift assay was carried out in the presence of extracts from glucose-grown cells and labeled 174 mer as described in Figure 4. The assay was performed in the absence (lane 2) or presence of 3 ng or 6 ng of 25 mer (lanes 3 and 4), 3 ng or 6 ng of mutl (lanes 5 and 6), 3 ng or 6 ng of mut2 (lanes 7 and 8), 3 ng or 6 ng of mut3 (lanes 9 and 10), 3 ng or 6 ng of mut4 (lanes 11 and 12). Lane 1 is a control in the absence of cell extract.

oligonucleotide corresponding to bases -434 to -410 of POX] (top strand: 5' GGGTAGGGTAATAAAATTAACCCTA 3') was synthesized and used as a competitor in the gel shift assay. This oligonucleotide competed out binding to the 174 mer, previously shown to contain the binding site (Figure 6). The labeled oligonucleotide itself gave rise to a positive band shift with extracts from glucose grown cells (Figure 7). Increasing amounts of unlabeled oligonucleotide progressively competed binding of the labeled DNA (Figure 7). Point mutations in the footprint region affect both glucose repression and DNA band retardation To further identify the exact nucleotides that comprise the cis element, we systematically mutated the 25 bp region. We introduced four mutations a time by synthesizing mutated versions of the 25 bp oligonucleotide (Figure 8a). Two of these oligonucleotides (mut2 and mut3) failed to compete for the band shift obtained with labeled 174 mer (Figure 8b, lanes 7-10). Thus we identified an 8 bp region (5' AGGGTAAT 3') that is involved in the DNA-protein interaction. To functionally demonstrate the importance of this region we introduced point mutations into this sequence in the 5' region of POX] (material and methods). The resulting construct, pP1353 was then transformed into S.cerevisiae. The f3-galactosidase activities from cells transformed with this construct were similar to those obtained from the deletion construct pP13571 (Table 1). This confirms that these mutated nucleotides are an integral part of the URS we have identified.

Oleate-grown cells give rise to a different mobility shift To test the glucose-dependence of the shift we observed with the 174 nucleotide fragment we also carried out DNA bandshift assays with extracts from cells grown on glycerol or oleate. Using

3

4

5

6

1

2

3

4

5

Figure 9. Extracts from oleate-grown cells cause a different mobility shift than extracts from glycerol- and glucose-grown

cells.

(a) Lane 1

serves as a

control

and contains the 174 bp radiolabeled DNA fragment in the absence of protein. Lanes 2-4 contain the same amount of labeled DNA in the presence of 10 ag of extract from cells grown on glucose (lane 2), glycerol (lane 3), or oleate (lane 4). Lanes 5 and 6 are the same as lanes 3 and 4 but in the presence of excess (1 tg) unlabeled 25 mer. (b) DNA bandshift with radiolabeled 25 mer. Lane 1 is a control in the absence of protein. Lanes 2 and 3 contain the same amount of labeled 25 mer in the presence of 30 jig protein from cells grown on glycerol (lane 2) or oleate (lane 3). Lanes 4 and 5 are the same as lanes 2 and 3, but in the presence of excess (1 ,ug) unlabeled 25 mer.

the radiolabeled 174 mer and extracts from glycerol-grown cells we observed DNA shifts that appear to be the same size as those seen with glucose or galactose extracts. However, with extracts from oleate-grown cells we obtained a DNA shift at a lower molecular weight (Figure 9a). In each case excess amounts of 25 mer abolished the DNA shifts. In assays using radiolabeled 25 mer we also observed DNA shifts at different molecular weights with extracts from glycerol and oleate-grown cells (Figure 9b).

DISCUSSION In order to understand the elements controlling the regulation of peroxisomal (3-oxidation enzymes we have initiated studies on the promoter region of POX], the gene encoding acyl-CoA oxidase in the yeasts S.cerevisiae. The 454 bp 5' flanking region of POX] used in these experiments appears to contain the necessary sequences for regulation of POX] by glucose and oleate. Deletion of a 174 bp fragment (-454 to -280) from the 5' flanking region of POX] results in increased expression of this gene in all growth media, and abolishes the glucose repression observed with the pP13570 construct. This suggests that the 174 bp sequence contains a ciselement that is responsible for glucose repression. In accordance with this, we found that glucose-grown cells transformed with the deletion construct expressed approximately the same level of,-galactosidase activity as those grown on galactose or glycerol, whereas there was no detectable activity in glucosegrown cells transformed with the undeleted construct. The DNA band shift and DNA footprinting experiments mapped one responsive site to a region of 24 nucleotides (-434 to -410).

3500 Nucleic Acids Research, Vol. 20, No. 13 By mutating this region we identified a sequence of 8 nucleotides (5' AGGGTAAT 3') that appear to be essential for the DNAprotein interaction. 3-galactosidase activities expressed from a construct carrying mutations within these 8 bp (pP1353) were higher in all growth media compared to activities expressed from cells transformed with pP13570, confirming that this is a repressive sequence (URS). However, the activity in extracts from glucose-grown cells was not as high as that obtained from the deletion construct pP13571. Thus it is possible that there is another glucose-repressive element located within the 174 bp region. The bandshift obtained with extracts from cells grown on glucose, galactose or glycerol appeared to be identical to each other, suggesting that the trans-acting factor that is binding to the URS we have identified is the same. Retardation of this same DNA fragment with extracts from oleate-grown cells was less than that observed with other media. In this case it is possible that the same factor is binding, but is modified in some manner and thus causes a suppression of the negative regulation. An alternative possibility is that a different protein is associating with the same region of DNA, thus preventing the negative-regulating factor from binding. In either case, the outcome contributes to the increased acyl-CoA oxidase activity observed when cells are grown in oleate medium. We note that the URS we have identified in POX] is also present in the 5' non-coding region of the gene encoding peroxisomal thiolase in S. cerevisiae (29). This enzyme catalyzes the cleavage reaction in the peroxisomal (-oxidation cycle, and is repressed by glucose. CTAI encodes peroxisomal catalase in S. cerevisiae, this gene is also repressed by glucose. There does not appear to be a similar sequence to the URS we have identified in POX] in the 5' region of CTA]. It was recently demonstrated that ADRJ, which functions in the activation of ADH2 encoding glucose-repressible alcohol dehydrogenase, binds to a CTAI upstream element (30). ADRI acts as a positive regulator of CTAI, thus causing suppression of the normal glucose repression. It was also shown to be involved in the regulation of two genes encoding fatty acid ,3-oxidation enzymes (trifunctional hydratasedehydrogenase-epimerase and thiolase), and PAS], a gene encoding a protein involved in peroxisome assembly (30, 31). We are currently testing whether ADRI can also act as a positive regulator of POX]. Einerhand et al. have recently proposed a sequence, termed the ,B-oxidation box, that acts as a UAS in genes encoding peroxisomal (-oxidation enzymes in S. cerevisiae during growth on oleate (29). The authors described this sequence in the thiolase gene, in the gene encoding the trifunctional protein, and a version of the sequence in POX] (-242 to -261). The sequence they described in POX] is within the 5' region of pPl3571 used in our experiments. Activity expressed form this construct is induced by oleate. However, we were unable to obtain a DNA shift with crude extracts from oleate-grown cells using radiolabeled DNA which contained this sequence. Experiments to determine whether the 20 nucleotides of POX], identified by Einerhand et al. are able to give rise to a band shift with purified proteins from oleate grown cells are currently in progress. To our knowledge this is the first negative regulatory element found to be associated with a gene encoding a peroxisomal enzyme. This finding should allow us to purify and characterize trans acting factors that bind to this region and that play a role in the regulation of peroxisomal enzymes. We have recently found that some yeast peroxisomal enzymes are regulated in response

to certain peroxisome proliferators which cause liver tumors in mammals (unpublished results). The studies on the transcriptional regulation of these enzymes should allow us to elucidate the general control mechanisms involved in repression and induction of peroxisomal enzymes in response to physiological or pharmacological stimuli.

ACKNOWLEDGEMENTS We thank Dr Henry Baker and Mr Edward Scott for assisting us with bandshift experiments, and Dr Henry Baker for useful suggestions and for critically reading this manuscript. We also thank Dr David Thomas for providing us with the plasmid pAD17. This work was supported by grant GIA643 from the American Heart Association-Florida Affiliate.

REFERENCES Natl. Acad. Sci. USA 73, 2043 -2046. Hajra,A.K. and Bishop,J.E. (1982) Ann. N. Y. Acad. Sci. 386, 170-181. Keller,G.-A., Barton,M.C., Shapiro,D.J. and Singer,S.J. (1985) Proc. Natl. Acad. Sci. USA 82, 770-774. Krisans,S.K., Thompson,S.L., Pena,L.A., Kok,E. and Javitt,N.B. (1985) J. Lipid Res. 26, 1324-1332. Ishii,H., Fuicumori,N., Morie,S. and SugaT. (1980) Biochem. Biophys. Acta 617, 1 -1. Needergaard,J., Alexson,S. and Cannon,B. (1980) Am. J. Physiol. 239, 208-216. Ishii,H., Horie,S. and SugaT. (1980) J. Biochem. 87, 1855-1858. Reddy,J.K. and Krishnakantha,T.P. (1975) Science 190, 787-789. Nemali,M.R., Reddy,M.K., Usuda,N., Reddy,P.G., Comeau,L.D., Rao,M.S. and Reddy,J.K. (1989) Toxicol. Appl. Pharnacol. 97, 72-87. Reddy,J.K., Azarnoff,D.L. and Hignite,C.E. (1980) Nature 283, 397-398. Rao,M.S. and Reddy,J.K. (1987) Carcinogenesis 8, 631-636. Lazarow,P.B. and Fujiki,Y. (1985) Annu. Rev. Cell Biol. 1, 489-530. Borst,P. (1989) Biochim. Biophys. Acta 1008, 1 -13. Kawamoto,S., Nozaki,c., Tanaka,A. and Fukui,S. (1978) Eur. J. Biochem. 83, 609-613. Ueda,M., Yamanoi,K., Morikawa,T., Okada,H. and Tanaka,A. (1985) Agric. Biol. Chem. 49, 1821-1828. Reddy,J.K., Goel,S.K., Nemali,M.R., Carrino,J.J., Laffler,T.G.,

1. Lazarow,P.B. and de Duve,C. (1976) Proc. 2.

3. 4.

5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

15. 16.

17. 18. 19. 20.

21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

Reddy,M.K., Sperbeck,S.J., Osumi,T., Hashimoto,T., Lalwani,N.D. and Rao,M.S. ('1986) Proc. Natl. Acad. Sci. USA 83, 1747-1751. Dmochowska,A., Dignard,D., Maleszka,R. and Thomas,D.Y. (1990) Gene 88, 247-252. McCammon,M.T., Veenhuis,M., Trapp,S.B. and Goodman,J.M. (1990) J. Bacteriol. 172, 5816-5827. Harshman,K.D., Moye-Rowley,W.S. and Parker,C.S. (1988) Cell 53, 321 -330. Small,G.M., Imanaka,T., Shio,H. and Lazarow,P.B. (1987) Mol. Cell. Biol. 7, 1848- 1855. Sambrook,J., Fritsch,E.F. and Maniatis,T. (1989) Molecular Cloning: a Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Craven,G.R., Steers,E. and Anfinsen,C.B. (1965) J. Biol. Chem. 240, 2468-2477. Small,G.M., Burdett,K. and Connock,M.J. (1985) Biochem. J. 227, 205-210. Bradford,M. (1976) Anal. Biochem. 72, 248-254. Fried,M. and Crothers,D.M. (1981) Nucleic Acids Res. 9, 6505-6525. Garner,M.M. and Revzin,A. (1981) Nucleic Acids Res. 9, 3047-3060. Hahn,S., Hoar,E.T. and Guarente,L. (1985) Proc. Natl. Acad. Sci. USA 82, 8562-8566. Veenhuis,M., Mateblowski,M., Kunau,W.H. and Harder,W. (1987) Yeast 3, 77-84. Einerhand,A.W.C., Voorn-Brouwer,T.M., Erdmann,R., Kunau,W.-H. and Tabak,H.F. (1991) Eur. J. Biochem. 200, 113-122. Simon,M., Adam,G., Rapatz,W., Spevak,W. and Ruis,H. (1991) Mol. Cell.

Biol. 11, 699-704. 31. Erdmann,R., Wiebel,F.F., Flessau,A., Rytka,J., Beyer,A., Frohlich,K.-U. and Kunau,W.-H. (1991) Cell 64, 499-510.