cross-pathway) control, allowing for derepression of amino acid biosynthetic enzymes under conditions of starvation for any one of a number of amino acids (18), ...
Vol. 5, No. 11
MOLECULAR AND CELLULAR BIOLOGY, Nov. 1985, p. 3139-3148 0270-7306/85/113139-10$02.00/0 Copyright C 1985, American Society for Microbiology
General Amino Acid Control and Specific Arginine Repression in Saccharomyces cerevisiae: Physical Study of the Bifunctional Regulatory Region of the ARG3 Gene MARJOLAINE CRABEEL,L* RAF HUYGEN,' KRISTIN VERSCHUEREN,1 FRANCINE MESSENGUY,2 KRISTOF TINEL,' RAYMOND CUNIN,' AND NICOLAS GLANSDORFF' 2 Laboratory of Microbiology, Vrije Universiteit Brussel,' and Research Institute, Centre d'Etude et de Recherches des Industries Alimentaires,2 B-1070 Brussels, Belgium Received 29 April 1985/Accepted 13 August 1985
To characterize further the regulatory mechanisms modulating the expression of the Saccharomyces cerevisiae ARG3 gene, i.e., the specific repression by arginine and the general amino acid control, we analyzed by deletion the region upstream of that gene, determined the nucleotide sequence of operator-constitutive-like mutations affecting the specific regulation, and examined the behavior of an ARG3-galK fusion engineered at the initiating codon of ARG3. Similarly to what was observed in previous studies on the HIS3 and HIS4 genes, our data show that the general regulation acts as a positive control and that a sequence containing the nucleotide TGACTC, between positions -364 and -282 upstream of the transcription start, functions as a regulatory target site. This sequence contains the most proximal of the two TGACTC boxes identified in front of ARG3. While the general control appears to modulate transcription efficiency, the specific repression by arginine displays a posttranscriptional component (F. Messenguy and E. Dubois, Mol. Gen. Genet. 189:148-156, 1983). Our deletion and gene fusion analyses confirm that the specific and general controls operate independently of each other and assign the site responsible for arginine-specific repression to between positions -170 and +22. In keeping with this assignment, the two operator-constitutive-like mutations were localized at positions -80 and -46, respectively, and thus in a region which is not transcribed. We discuss a hypothesis accounting for the involvement of untranscribed DNA in a posttranscriptional control.
The ARG3 gene of Saccharomyces cerevisiae codes for ornithine carbamoyltransferase (OTCase) (EC 2.1.3.3), which catalyzes step 6 of arginine biosynthesis. Expression of ARG3 is modulated in two ways: by the general (or cross-pathway) control, allowing for derepression of amino acid biosynthetic enzymes under conditions of starvation for any one of a number of amino acids (18), and by the specific repression elicited by arginine (3). Measurements of steady-state levels of HIS3 (36), HIS4 (9), HIS] (16), ARG3, ARG4 (24), and TRP5 (40) mRNA molecules, as well as estimates of TRP5 transcription rates (40), indicate that the general amino acid control acts on DNA transcription. It appears as a positive control involving in cis position at least one copy of the TGACTC consensus core sequence and in trans position, directly or indirectly, the GCN4 (formerly AAS3) gene product. The role of the TGACTC sequence was demonstrated by deletion analysis of the 5' noncoding region of HIS4 (9). A similar analysis of the HIS3 5' noncoding region and the finding of multiple TGACTC repeats in front of all the genes mentioned above corroborated the role of regulatory target assigned to this sequence. That the general control is a positive regulation is indicated by the fact that deletion of all TGACTC boxes in front of HIS4 freezes expression of that gene at a low constitutive level and by the pattern of epistasis between the two types of mutations which affect the general control in trans, i.e., the gcn mutations, which prevent derepression under conditions of amino acid deprivation, and the gcd mutations, which result in high constitutive enzyme levels. The conclusions of this analysis (17) indicate the GCN4 gene
*
Corresponding author.
product as the most likely candidate for the role of positive regulatory factor interacting with the TGACTC control site. In the case of HIS4 at least one TGACTC box and a functional GCN4 gene are required to maintain the so-called repressed level of expression observed in minimal medium. If the GCN4 product acts as a regulatory factor only, these results suggest that the repression in minimal medium is incomplete and that the amplitude of cross-pathway control is greater than previously thought; alternatively, the GCN4 product could affect promoter efficiency (20). Lucchini et al. (20) also demonstrated that the open reading frame (ORF) located 5' to HIS4 plays no role in the capacity to achieve and maintain a normal, steady-state derepressed level. The presence of such ORFs is a feature shared by several of the genes under general control (HIS4 [9, 14], HISI [16], ARG4 [2], and TRP5 [40], although not HIS3); ORFs also precede genes that are not under general control, such as LEU2 (1) and URA3 (29). Expression of ARG3 is also specifically repressed by arginine. Genetic evidence classifies this control as a negative one, involving a cis-acting, operator-like receptor site mutated in operator-constitutive (OC) strains (23; see below) and trans-acting regulatory molecules produced by at least three unlinked loci, ARGRI, ARGRII, and ARGRI!I (3, 13). Using cloned ARG3 DNA (6) as a probe for Northern blotting experiments, Messenguy and Dubois (24) examined the steady-state levels and the stability of ARG3 mRNA under various conditions and in different genetic backgrounds; they found that a large part of the variations observed at the enzyme level was not paralleled at the mRNA level and that the half-life of ARG3 mRNA was enhanced in argR and Oc mutants. They concluded, therefore, that specific repression by arginine involved a posttran3139
3140
MOL. CELL. BIOL.
CRABEEL ET AL.
scriptional component. Analysis of an ARG3-lacZ gene fusion (5) ruled out any primordial contribution to this mechanism by sequences 3' to the first 79 nucleotides of the messenger.
To identify the domains involved in promoter activity and in each of the two mechanisms controlling the expression of ARG3, we analyzed the 5' noncoding region of that gene by three converging approaches: deletion analysis, sequence determinations of Oc mutations, and analysis of an ARG3galK gene fusion engineered at the initiating codon. MATERIALS AND METHODS
Strains. Strains 10S46b (gcn4-101), 10S48c (gcn4-101 arglJl), 10S48b (pA223 argIlI), and 10S48d (pA223, gcn4-101) are segregants issued from a cross between MG471::pA223 and 1C1784b (gcn4-101 argllJ). The mutation argll+ (previously called argJ+) is derived from the MG409 strain, and gcn4-101 is derived from the L870 strain isolated by G. Fink. "p" refers to ARG3 promoter deletions (see below). Strains 10S58b (pA282 argllI ) and 10S86b (pA363 argJ I ') are segregants from crosses of strain 1C1784b (gcn4-101 argJlJ) with, respectively, MG471::pA282 and MG471::pA363 [with the 1C1784b strain gcn4-101, argll+]. Strain lC1315a (ARG3+ 02C argRII-) is issued from a cross between strains 1C1288a (MATa ARG3+ 02c) and BJ210 (MATot argRII-); strain 1C1932c (ARG3+ 02C argll+) was obtained after crossing strains 8850b (MATa argll+) and 1C1304c (MATot ARG3+ 02c). Media. The basic minimal medium (M medium) containing 3% glucose, vitamins, and mineral traces has been previously described (23). Mam medium is M medium supplemented with 0.02 M (NH4)2SO4 as the nitrogen source. Yeast RNA preparations. Total RNAs were prepared from log-phase yeast cells (5 x 106 cells per ml) by the method of Waldron and Lacroute as modified by Dubois et al. (13). Yeast DNA preparations. Yeast DNA was prepared by the method of Cryer et al. (7) with minor modifications. Southern analysis. The DNA restriction fragments resulting from the digestion of 10 p.g of genomic DNA were fractionated by electrophoresis on a 0.7% agarose slab gel. The transfer of the in situ-denatured DNA was performed by the Southern blotting procedure (33). Prehybridization and hybridization were performed in a mixture of 4x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate), 0.2% Ficoll, 0.2% polyvinylpyrrolidone, 0.2% bovine serum albumin, 0.1% sodium dodecyl sulfate, 10% dextran sulfate, 1 mM EDTA, and 1 mg of sonicated calf thymus DNA per ml. The probe, radioactively labeled by nick translation (28) was heat denatured by boiling for 5 min and allowed to hybridize overnight. Transformation procedures. The methods used for transformation have been described previously (5). DNA fragment purification. The desired DNA fragments separated by agarose gel electrophoresis were cut out and electroeluted out of the agarose as follows. The gel pieces were introduced into a Pasteur pipette on top of 1 cm of hydroxyapatite layered on 2 cm of Sephadex G50, both preequilibrated in electrophoresis buffer (TEB) (Tris, 54 mg/ml; EDTA, 4.65 mg/ml; boric acid, 27.5 mg/ml; ethidium bromide, 0.5 p.g/ml). The column was then immersed without air bubbles in TEB in a horizontal electrophoresis apparatus, and electrophoresis was applied overnight at low voltage (30 V). Visualized under UV light, the DNA trapped in the hydroxyapatite was eluted with 1 M phosphate buffer (pH 7.5), it was purified partially by its passage through the
Sephadex and further by dialysis against TE buffer (Tris, 10 mM, pH 7.5; EDTA, 1 mM). Construction of deletions by using BAI 31 nuclease. The conditions for BAl 31 digestion were as follows. Restricted and ethanol-precipitated plasmid (50 ,ug) was suspended in 50 p.1 of bovine serum albumin (500 ,ug/ml) plus 50 ,ul of 2x BAl 31 buffer (24 mM CaC12, 24 mM MgCl2, 0.4 M NaCl, 40 mM Tris hydrochloride [pH 8], 2 mM EDTA). BAl 31 (0.35 U; Boehringer) was added to this mixture, which was first preincubated for 3 min at 30°C. Samples of 20 p.l were removed successively after 1, 2, 5, 10, and 20 min, and the action of BAl 31 was immediately blocked by addition of 20 mM EDTA. The nuclease was then extracted from the pooled samples with phenol, phenol-chloroform (1:1), and chloroform (supplemented in a 24:1 proportion with isoamyl alcohol). The deleted DNA was precipitated with ethanol and suspended in SmaI restriction buffer for secondary restriction with this enzyme. These conditions were chosen on the basis of calculations taking into account the size of the plasmid, the number of free ends, and the length of the deletions desired. Si and ExoVII nuclease manipulations. The suitable singlestrand 5'-end-labeled DNA fragment and 20 or 40 p.g of total RNA were solubilized in 10 p.l of 1x HB buffer (40 mM PIPES [piperazine-N,N'-bis(2-ethanesulfonic acid)] 0.4 M NaCl, 1 mM EDTA). The nucleic acids were first denatured for 10 min at 80°C and then allowed to hybridize during an overnight incubation at 45°C. The enzymatic digestion occurred afterwards during a 45-min incubation at 20°C after addition of 300 p.l of either S1 or ExoVII solution, consisting of 1,000 U of S1 or ExoVII per ml in, respectively, 280 mM NaCl-25 mM sodium acetate (pH 4.4)-4.5 mM zinc acetate or 30 mM KCl-10 mM Tris (pH 7.4)-10 mM EDTA. Controls were incubated with the same volume of buffer without enzyme. In other controls, no RNA was added. The nucleic acids were precipitated with ethanol, dissolved in 6 p.l of a mixture of formamide dyes (2 ml of formamide, 0.1 ml of 2% xylene cyanol FF in formamide, and 0.1 ml of 2% bromophenol blue in formamide), and resolved by electrophoresis on denaturing 6% polyacrylamide-7 M urea gels together with a chemical sequencing ladder of the same 5'-endlabeled DNA single-strand fragment. Enzyme assays. OTCase activity was measured as described previously (25). P-Galactosidase was assayed by the method of Miller (26); the assays were run on yeast extracts obtained with a French press. Galactokinase activity was measured by the procedure described by Rymond et al. (30), except that the yeast extracts were obtained with the French press. Proteins were measured by the Folin method (19). RESULTS Nucleotide sequence of the control region of ARG3 and determination of transcription start points. The sequence of the 5' noncoding region of ARG3, previously established up to position -182 (5; R. Huygen, M. Crabeel, R. Cunin, N. Glansdorff, R. Contreras, W. Degrave, and W. Fiers, Arch. Intern. Physiol. Biochim. 89:B172, 1981), was extended upstream to position -500 by the strategy depicted in Fig. 1. Nucleotide positions are numbered relative to the start of transcription (see below). A sequence matching the consensus for a GoldbergHogness box is present around position -100. Two TGACTC sequences, the proposed target sites for the general amino acid control, and a third related sequence (TGATTC) are found at positions -292, -416, and -497,
respectively.
VOL. 5, 1985
BIFUNCTIONAL REGULATORY REGION OF S. CEREVISIAE ARG3 _
=
=
.u
X
s
H
I
H T
T
T
3141
I
49 2
E
II-
I
I
-
I
-600
700
- SOO
-300 -.300
-400
I
I1 +1
I
-20(()
-10 a
4-
1 01)I
aI__
F
I
de
we
..
I
I
_-e
.
.
HinfI -590 S'-AA CCCTGATTCG CCATCAGTTA CGTCTTCTAA GAAAAAATAT TTAGATCATA
TTATTTTAGA TAACCGAGAC
6- 363 HinfI
-400
Taq I
ATCGTTAGCA ACCATGACTC CAGTAAACAA AAATTCAAGA TCCGAAATAT TTTGAACTCG ACCTTCTAAC ATTAC A-282 TaqI -390 Hinf I GCTCC TTCGTATTAC TCATTCAGCT -CTTCTTCTGA TAGCAGTGAA TTTTCGAGGG TCACGTCGTG ACTCATATGC
I
A-223 Taq I I TTTCTTGTTC CGTTTCGTTT CGAGATGACA AAAAACTGGT CATTTTTTCC GTTAAGTGCA ACTCACAGCA GTATC
HaeIII -200
Est E II GGCCG CTGAGAAATG CCCGGACAAA TTTTTTTGAG CCGGATTGGT CACCGTTTCT TTCTTCGGCG CGGCTTCCCA
Cf2 -190 TaqI TTCCCGTCCA TCCAAAAAAA TCTACCTATA TAAATCGACT TTTCACCTCT AAAGGCAGTT TATTCCTTGT ATGTC C 0 1
5, 'mRNA
A CTTTA AGTACAGTTA ATAACGAGCA ATTTTTTTTT TTTTTTTTAG CCATCTACCC ATCAACTTGT ACACTCGTTA
Bgl CC
ATG
TCA
ACC
ACA
GCA
TCC
ACG
CCT
TCA
TCT
TTA
CGT
CAT
TTG
ATT
TCT
ATA
AAA
GAT
II
CTT
met ser thr thr ala ser thr pro ser ser leu arg his lou ile ser ile lys asp leu
Eco RI TCT GAT GA GM TTC AGA-3' ser asp glu glu phe arg
FIG. 1. Nucleotide sequence of the 5' region of ARG3. The strategy followed is indicated by arrows below the restriction map (H, Hinfl; T, TaqI). Both the Maxam and Gilbert (22) and the dideoxynucleotide termination (H; 31) procedures were used. The position of the 3' termini of the sequenced deletions (A - 363, A - 282, A - 223) and the nucleotide substitutions found in the Olc and 02C mutants are indicated. The two TGACTC boxes (at positions -416 and -292) and a Goldberg-Hogness consensus sequence are underlined. The heavy bars under the restriction map indicate ORFs, i.e., the ARG3 structural gene beginning at +22 and two other ORFs (-193 to -97 and -284 to ARG3).
3142
MOL. CELL. BIOL.
CRABEEL ET AL.
SmaI
FIG. 2. Plasmid used for the construction of a family of deletions with identical 5' ends and variable 3' ends. Plasmid pMC304 was linearized at the unique XhoI site near position -750. Nuclease BAl 31 was allowed to degrade the free ends under controlled conditions. Restriction at the unique SmaI site near position -1500 was performed before ligation. Since the yeast URA3 gene of pMC304 complements pyrF E. coli mutants (29), the resulting plasmids were transformed into E. coli C600 pyrF, with selection for ampicillin resistance and thereafter screening out of Ura+ transformants producing blue colonies on X-Gal-supplemented minimal medium (about 10% of the total). Plasmids from these strains, in which both URA3 and lacZ should be intact, were used for integration into the yeast genome (see Fig. 3). p, Promoter.
Besides the ORF of the ARG3 structural gene starting at nucleotide +23, there is a small ORF between positions -193 and -97; another ORF extends from position -282 to the AUG codon of the structural gene, in the same reading frame. The 5' ends of ARG3 mRNA molecules have already been mapped with nuclease S1 or ExoVII and the separated strands of a DNA fragment between a HaeIII (-205) and an EcoRI (+94) site (Huygen et al., Arch. Intern. Physiol. Biochem. 89:B172). With ExoVII the main 5' end fell in the CAT sequence centered 22 nucleotides upstream of the initiating codon. S1 gave a narrow cluster of bands, the major one corresponding to a leader beginning 26 nucleotides before the initiating codon. In further discussion, position + 1 will be assigned to the adenine situated 22 nucleotides before the ATG codon. A comparison of the results obtained with exonuclease ExoVII and exoendonuclease S1 shows that there is no intron in ARG3 5' to the EcoRI site (Fig. 1). No variation in the pattern of 5' termini could be detected whether the RNA was extracted from repressed or derepressed cells or from argR or Oc regulatory mutants. In more recent S1 mapping experiments we used separate strands of both the XhoI (-'750)-BstEII (-163) and XhoI (-750)-HaeIII (-205) DNA fragments, both located 5' to the fragment used previously. 'No protected DNA sequence could be detected whether RNA was extracted from steady state derepressed or repressed cells; therefore, the two ORFs mentioned above do not appear to be transcribed.
Construction, chromosomal integration, and physical characterization of deletions with identical 5' ends and variable 3' ends in the 5' upstream region of ARG3. We used plasmid pMC304, which carries URA3 and an ARG3-lacZ fusion engineered at the codon 19 of ARG3 (5), to construct a set of deletions with their 5' ends at the unique SmaI site of this plasmid and their 3' ends at different points in front of the transcription start point of ARG3 (Fig. 2). Enzyme assays (data not shown) indicate that cutting ;at the SmaI site located at position 1104 in the 1,166-nucle6tide-long HindlIl insert bearing URA3 (29) does not interefere with expression of that gene. In keeping with this observation, an exemplar of the consensus sequence for DNA transcription termination in S. cerevisiae (41) is present at a short distance upstream from the SmaI site. Besides, since the shortest deletion examined in the ARG3 promoter region (A363; see Fig. 4) expresses almost wild-type OTCase activity, no readthrough transcription from URA3 appears to interfere with ARG3 transcription. Our aim was to insert the deleted plasmids, which are unable to replicate in S. cerevisiae, in front of the unique genomic copy of ARG3 in order to assay the influence of the deletions directly in terms of OTCase activity and thus to avoid complications arising from multiple integrations of the hybrid gene or from the nature of the fusion itself. Since no restriction site was available to direct integration at ARG3 by double-strand break-repair recombination, we selected for integration at either ARG3 or URA3 by transforming to Ura3+ mutant MG471, which is isogenic to our standard strain Y. 1278b (Fig. 3). Screening of integrated plasmids was performed by Southern analysis (Fig. 3a), characterization of plasmids extracted from integrant DNA after restriction by HindIII (Fig. 3b), and genetic analysis; a cross between a Ura3+ strain and a Ura3+ integrant should regularly produce Ura3- spores only if no integration occurred at URA3. The results established that in the case of deletions A -170, A -223, A -282, and A -363, at least one exemplar had been integrated at ARG3; in the case of A70, integration had occurred at URA3 only. To leave no ambiguity in the link established between phenotype and genotype, we reextracted from the cultures used for enzyme assays (see below) the plasmids to be characterized. DNA was treated with the BstEII enzyme (cutting at -162) and ligated, and the plasmids with the configuration shown in Fig. 3c were obtained in Escherichia coli after transformation to ampicillin resistance. The HindIII-SalI fragments bearing the deletions were subcloned in phage M13 mpll, and the exact length of the remaining portion of the ARG3 control region was established by the dideoxynucleotide termination sequencing method for deletions A -363, A -282, and A -223. In all cases the 5' end of the deletion was at the SmaI site of the URA3 fragment, as
expected. The remaining segments of promoter region for the longer deletions A "170 and A =-'70 were established in the original plasmids by restriction patterns obtained with BstEII, ClaI,
FIG. 3. Chromosomal integration and characterization of deleted plasmids. (a) Genomic structures resulting from single or multiple integrations at ARG3 or URA3 in strain MG471. Southern blots of HindlIl digests of integrant DNA were probed with a URA3 HindlIl fragment to distinguish integration at ARG3 (a 3.0- to 3.8-kilobase band is expected) from integration at URA3 (a plasmid-length, ca. 11-kilobase band is expected) or from multiple integrations either at ARG3 or at ARG3 and URA3 (both bands are expected). (b) Plasmids that can be recovered after treatment of integrant DNA with HindlIl and transformation into E. coli pyrF, with selection for ampicillin resistance. The two types of plasmids can be distinguished by their Xhol-BstEIl and Pstl restriction patterns and by the Ura+ phenotype of the E. coli transformants. (c) Plasmids recovered after treatment of integrant DNA with BstEII in the case of integration at ARG3 or tandem integration at URA3. p, Promoter; B, BstEII; H, HindlIl; P, PstI; S, SalIl.
CEREVISIAE ARG3
BIFUNCTIONAL REGULATORY REGION OF S.
VOL. 5, 1985
chromosome
a) Integration in the
single
single copy integrated at ARG 3
lacZ
pBR32.2
W. T. pARG 3 B
;
BB
lacZ
WTp
pBR322
,
ARG 3 H
pBR
EB
BB f S
B
H
H
ura3+/H4
URA3+ ARG3
H
pBR
lac Z
H
ura3-/+
AP
B B_+Z_>wf
B B
p ARG3
ura3-
6p
B
lacZ
ARG3R3
URA3
H
copy integrated at URA 3
H
H
tandem integration
ARG 3 ~~~~~~mat ~~~~~~~~
AP tandem integration at URA 3
Ap B
H
7B
B
H
H
Lrr AP
b) Plasmids that
BB
H
is
AP
can be recovered after treatment of B B with Hind III
the
of the integrants B B
genome
p BR
p BR
lac 7
lac Z
P
P H
H.
ura3 P
W.T pARG 3 X recoveredfrom single integration at ARG3
c) Plasmid that
recovered from single integratlon at URA3 and from tandem integrations
can be recovered after treatment of the genome of H
with
u ra 3
ApARG3 lac Z BS
pBR322
integrants Bst E
II
3143
3144
MOL. CELL. Bioi-
CRABEEL ET AL.
OC2
TGACTC -292
TGACTC
-41 6
0C1
+1 ST
B
T
A--70
A--170 1' A-223
A-282 A-363r
-I-
General Control
Specific Repression by Arginine FIG. 4. Portions of the 5' region of ARG3 left intact in various deletions (straight lines) and inferred allocation of regulatory domains. B, BstEII; T, TaqI.
AccI, HincII, TaqI, Hinfl, HpaI, HaeIII, and combinations of these enzymes. The presence of the BstEII site at -162 in mutant A '170 and the absence of the TaqI site at -90 in mutant A =70 were clearly demonstrated. The different deletions are drawn to scale in Fig. 4. OTCase assays of deletion mutants under different growth conditions and genetic backgrounds. None of the deletions integrated in front of ARG3 affect growth on minimal medium; thus, even if reduced, OTCase specific activity does not become growth limiting in those strains. To test whether the deletions affected promoter efficiency, the specific repression by arginine, and the general control response, we assayed OTCase activity both in the mutants themselves after growth in different media and in recombinants (see Materials and Methods) mutated in trans-acting regulatory elements affecting the specific and the general controls. During the construction of these recombinants, the gcn4mutation was recognized by the sensitivity it confers to 20 mM 3-aminotriazole and by moderate arginine bradytrophy; argll+ was recognized by strong arginine bradytrophy. Strains bearing the ARG3 gene with a deleted promoter was identified by the blue color of their colonies in X-Gal (5 bromo 4 chloro 3 indolyl , D galactopyranoside)supplemented medium due to the presence of the lacZ gene linked to the ARG3 locus (Fig. 3a). This phenotype segregated 2:2, as expected in those cases in which integration had occurred exclusively at ARG3 (A -363 and A -223). In the case of A -282, where integration had occurred also at URA3, all the segregants of the relevant tedrads were assayed. To interpret the results (Table 1), let us first recall the response of the wild-type ARG3 gene to the physiological treatments and genetic backgrounds considered. Normal arginine-specific repression obtained by adding 1 mg of L-arginine per ml to minimal medium reduces the wild-type level (about 30 ,umol of citrulline per h per mg of protein) by a factor of at least 10; in the argR- background this level is raised 1.5- to 2-fold. The maximal amplitude of this specific control is thus about 30. OTCase activity reaches its maxi-
-
-
-
-
-
-
-
mal level (up to 200 ,umol of citrulline per h per mg of protein) when derepression due to the general amino acid control is superimposed to arginine-specific derepression in argll+ bradytroph grown in minimal medium. gcn4- mutants are affected in their ability to derepress via the general control, but since the gcn4- mutation confers arginine bradytrophy, such strains experience arginine-specific derepression in minimal mnedium; indeed, no further increase in OTCase activity takes place when a bradytrophic argl l+ mutation is introduced in a gcn4- strain. In gcn4- strains grown in minimal medium the level of OTCase is about 20; thus, if specific derepression contributes a factor of 2 to this value (see above), it is a theoretical activity of about 10, which should be compared to the 30 U obtained in the wild type under the same conditions. We therefore estimate that the positive GCN4 factor, when present, enhances the enzyme level about threefold in minimal medium by increasing promotion or by a regulatory effect. Let us now examine the results obtained with the deletions. In minimal medium three levels of expression can be distinguished: quasi-normal (A -363), two- to fourfold reduced (A -282, A -223, and A -170), and not detectable (A '70). In all the mutants in which OTCase activity remains measurable, the addition of 1 mg of L-arginine per ml reduces expression by a factor of 10, as in the wild type. Therefore, specific repression by arginine remains untouched in deletions retaining promoter activity. Normal ability to derepress the general control system (seven- to eightfold) is observed in the argl l+ derivative of the A -363 mutant. Moreover, OTCase activity in the A - 363 mutant is close to normal in minimal mediumn. The slight difference with respect to the wild type could be due to the loss of a weak promoter element, to a secondary site of action for the general control, or to the presence of bacterial sequences 5' to the deletion novel joint. In marked contrast with the A -363 mutant, the other mutants display only a 1.5- to 2-fold increase in activity, whether in the argll+ or the gcn4- context; considering the bradytrophic character of
BIFUNCTIONAL REGULATORY REGION OF S. CEREVISIAE ARG3
VOL. 5, 1985
3145
TABLE 1. OCTase assays in the deletion mutants Strain
Status of ARG3 promotor
Genotype
Y1278b E1278b
Growth medium (29°C)
Mam Mam + L-arginine (1 mg/ml) Mam Mam
OTCase sp act (i±mol of product per h per mg of protein at 30°C)
25-28 2-3 19 170-200 20 53
10S46b MG409 10S48c BJ210
Wild type Wild type gcn4argI+argll+ gcn4argRII-
Wild Wild Wild Wild
MG471::A363 MG471::A363 10S86a
A -363 A -363 A/-363 argl I
A -363 A -363 A -363
Mam Mam + L-arginine (1/mg/ml) Mam
MG471::A282 MG471::A282 10S58b
A -282 A -282 A -282 argll+
A -282 A -282 A -282
Mam Mam + L-arginine (1 mg/ml) Mam
14 1.8 23
MG471::A223 MG471::A223 10S48d 10S48b
A -223 A -223 A -223 gcn4A -223 argll+
A A A A
Mam Mam + L-arginine (1 mg/ml) Mam
6-10
MG471::A =170 MG471::A =4170
A =170 A ~170
A -170 A -170
type type type type
-223 -223 -223 -223
Mam Mam
Mam
Mam Mam + L-arginine (1 mg/ml)
Mami a For this deletion strain, in which integration occurred at URA3 (see Fig. 3a) fB-galactosidase activity was measured. A ~=70
MG471::A =70
both types of mutations, this increase should be ascribed to arginine-specific derepression. Thus, in the A -282, A -223, and A -170 mutants, the ability to derepress via the general control appears to be lost. It is noteworthy that a TGACTC sequence is present between positions -363 and -282. In summary, we can draw the following conclusions. (i) The domains involved in the specific repression and in the general control are totally distinct, the sites involved in the specific control being located downstream from position -170. (ii) An 80-nucleotide-long domain (from position -363 to -282) containing the most proximal of the two TGACTC sequences identified in front of ARG3 is needed, as well as the GCN4 product, to achieve the general derepressed level and to maintain normal expression in minimal medium. (iii)
LEU 2
ARG3
ARG3
p ARG 3
gal K
oARC3
v
I
ter ARG3
FIG. 5. Directed integration of plasmid HP1 at ARG3 by using the unique XhoI site present in the control region of that gene (see text).
20 2 150
0.6 13 10-20
7-11 1 0"
A strong, downstream promoter element located between positions -170 and -70 is necessary to achieve a basal expression level of 6 to 10 U; this element probably involves the TATATAAT sequence centered around position -100. The next experiments were designed to narrow the limits of the domain involved in arginine-specific repression. Galactokinase assays in an ARG3-gaUK fusion. Plasmid HP1 (Fig. 5) a gift from T. Cabezon, N. Harford, and M. De Wilde at Smith Kline-RIT, bears the upstream region of ARG3 from the HindIII site located at position -1800 to the translation-initiating AUG codon at position +22. A BamHI restriction sequence was engineered at that site, a polylinker was built in, and the E. coli structural gene for galactokinase (galK) was inserted in frame with the AUG codon. We directed HP1 integration into the chromosomal ARG3 gene of strain 10S44d (leu2-) by linearizing the plasmid with enzyme XhoI at position -750 and selecting for LEU2+ transformants (Fig. 5); those having acquired intact plasmids were eliminated on the basis of the high mitotic instability of their LEU2+ marker. Galactokinase was assayed after growth of the cells in minimal medium containing 3% glucose and in the same medium supplemented with 1 mg of L-arginine per ml; values of, respectively, 18.5 and 1 nmol of galactose phosphate per min per mg of protein were reproducibly found. No contribution of the yeast gall gene could be detected under those conditions. OTCase specific activity was respectively, 21.5 and 1.4 pM/h per mg of protein. Thus, in strain 10S44d the native ARG3 gene and the galK gene were both normally repressed by arginine, implying that repression involves sequences located 5' with respect to the AUG codon. Cloning and sequencing of two ARG3+-O'-1ike mutations. The first Oc0like mutation relieving ARG3 from argininespecific repression was isolated in our laboratory by Messenguy (23). By the same strategy, a second, more
3146
MOL . CELL . B IOL .
CRABEEL ET AL.
TABLE 2. OTCase in the OC mutants Strain
Y1278b
Growth medium
Genotype
Wild type
BJ210
argRII-10
MG409 7480a
argl I ARG3+ Olc
7546b
ARG3+ O1" argRII-10
1C1634a 1C1297c
ARG3+ O1 arglI+ ARG3+02C
1C1315a 1C1932c
ARG3+02cargRII-1O ARG3+ 02c argJl+
Mam Mam Mam Mam Mam Mam Mam Mam Mam Mam Mam Mam Mam Mam
+ L-arginine (1 mg/ml) + L-arginine (1 mg/ml) + L-arginine (1 mg/ml)
+ L-arginine (1 mg/ml) + L-arginine (1 mg/ml)
OTCase sp act (,umol of product per h per mg of protein at 30°C)
29 4 44 45 170-200 46"l 31" 51" 41" 111 48 20 62 173
a These values have been published previously (23).
partially constitutive mutation was obtained (Table 2). Both mutations were cloned by using the gene transplacement technique (27, 32), following the strategy outlined in Fig. 6. Unique but different substitutions were identified for each of the Oc mutants, i.e., a CG-to-AT transversion at position -46 in mutant O1c and an AT-to-CG transversion at position -80 in mutant 02C. The sequences were examined up to position -180, upstream of which the deletion analysis indicates that no sequences playing a role in arginine repression should be present. The O1c mutation appears to interfere with promoter activity since, in minimal medium, the OTCase level observed in an Olc-argllJ double mutant (111 ,umol of citrulline per h per mg of protein) stays well below the value found in argil ± itself (170 to 200 ,umol of citrulline per h per mg of protein); the mRNA level as estimated by Northern blotting (data not shown) is affected to the same extent. DISCUSSION
The aim of this study was to identify the domains of DNA involved in promoter function and in each of the two controls modulating the expression of ARG3, i.e., the general amino acid control and arginine-specific repression. Concerning the general amino acid control, it is clear that the system works positively by stimulating expression of ARG3, in keeping with the results obtained for other genes coregulated by the cross-pathway control (9, 34). Indeed, removing the DNA between positions -363 and -282 suppresses the capacity to derepress ARG3 under conditions of amino acid deprivation. This 80-nucleotide-long sequence, essential for derepression, contains the most proximal of the two TGACTC sequences present at positions -290 and -415 in the 500 nucleotides proximal to ARG3. It therefore appears likely that in ARG3 also this sequence is the target site for the general control. The sequence at -415 appears to be unnecessary or to exert only a weak effect. The TGAGCC sequence present at position -170 might either be inactive or act only in concert with a functional copy. It would be interesting to mutate this sequence into TGAGTC (demonstrated to be functional in the R2A136 HIS4 mutant [9]) in a strain devoid of the two TGACTC sequences to see whether it becomes functional at its particular location. It is not
known whether the TGATTC sequence located at -499 may be considered as a putative functional target site. Several questions concerning the way the TGACTC boxes act in derepression remain unanswered. Are all the repeated elements equivalent? Are there context or distance effects? Do several TGACTC elements act concertedly, and do they play a role in stimulating transcription only under derepressing conditions or do they also contribute to promoter efficiency in the absence of the physiological signal elicited by amino acid starvation? As in the case of HIS4 our results show that the combination of only one TGACTC copy and a functional GCN4 product is required not only for derepression under conditions of starvation but also to maintain a normal "repressed" level in minimal medium; it is not known whether this combination of elements acts on promoter efficiency or whether the physiological signal elicited by amino acid starvation is not entirely silent on minimal medium. In that case, the amplitude of regulation by the general control would be greater than previously assumed. Besides the TGACTC sequences, another structural analogy shared by most genes coregulated by the general control is the presence of upstream ORFs (see above). A short ORF of 30 codons starts at position -284 but does not appear to be transcribed, at least under steady-state conditions. Also, no transcript could be correlated with the ORF extending the ARG3 gene for 102 codons in the 5' direction. From the point of view of promoter activity, our analysis distinguishes two important domains. The DNA sequence between -363 and -282 contains a promoter element that may or may not be distinct from the TGACTC sequence centered at position -290. Another critical sequence is localized between positions -170 and -70; the TATATAAAT oligonucleotide centered at position -100 is therefore probably functional, while the TAATAA sequence at position -40 appears to be irrelevant. The physical data confirm the genetic evidence for the existence of two independent regulations acting on the expression of ARG3. Indeed, specific repression by arginine is clearly operating in the three deletion mutants that have lost the general control. S1 mapping experiments have eliminated the possibility that the mRNA molecules obtained at different levels of activity of the specific control would differ at their 5' termini, while enzyme assays of the five
VOL. 5, 1985
BIFUNCTIONAL REGULATORY REGION OF S. CEREVISIAE ARG3
deletion strains and of the ARG3-galK gene fusion engineered at the first AUG codon of ARG3 mRNA bracket the target sites involved in the specific repression between positions -170 and +23. The possibility that a messenger sequence between positions +1 and +23 is involved in repression is being investigated presently. Here, however, the localization of Oc mutations 1 and 2 outside the transcribed region provides a most important indication regarding the mechanism of the specific control. This localization to the right side of the Goldberg-Hogness
a) Directed
3147
start of transcription
v
*1
Integration
p8R322
5
URA 3 +1
RNA pol II
FIG. 7. Hypothetical interaction between the arginine repressor and the 5' region of ARG3 mRNA. Repressor bound at the operator is translocated by the transcription machinery to the nascent mRNA molecules (see text).
0c7. : 7 7 Xhol
Xho I
b)Plasmid With Oc is recovered by XhoI restriction of the genome of integrants. HindIM-BgtI1 fragment is subcloned in M13mpll
HindIII
C
0
Bglfl FIG. 6. Cloning of Oc mutations by
gene transplacement. (a) Plasmid pMC304 was linearized at a unique BglII site in ARG3 to direct integration at the ARG3 locus in each ARG3+ Oc ura3recipient strain. Ura+ transformants harboring only one copy of the plasmid were screened out by OTCase assay. (b) DNA from such strains was restricted by XhoI, ligated, and used to transform E. coli to ampicillin resistance. Plasmids with the same structure as pMC304 were recovered. Fragments delineated by HindIII (-1800) and BglII (+79) were subcloned in M13 mpll opened by HindIIl and BamHI, and the nucleotide sequence was determined from the BglII extremity by the dideoxy termination procedure using a universal
sequencing primer.
box is in itself unique. It is at first sight paradoxical, considering the posttranscriptional component involved in arginine-specific repression (see above and reference 24). The model outlined in Fig. 7 proposes a solution to this paradox. The present data point to specific contacts occurring between argR regulatory molecules and the -80 to -40 region; this interaction may affect transcription to a certain extent, but to explain the posttranscriptional component, we assume that the repressor or part of it is transported by RNA polymerase molecules to the nascent transcript and, at this site, modulates translation efficiency or mRNA stability (or both). This could occur by recognition of a specific site on the short (22-nucleotide) leader or by interaction with the transcript in a less specific way; e.g., one could envisage an interference with messenger capping. Such a mechanism of posttranscriptional control would be different from those inferred for the GCN4 (15, 38) or CPA] (39) gene: short ORFs are present in the leader sequence of the messengers of both genes and, at least in the case of GCN4, are instrumental in translational control. Mutation 02C is located at the 3' end of the octanucleotide 5'-CACCTCTA-3', the only region of significant homology pinpointed by Sumrada and Cooper (37) between the 5' noncoding regions of ARG3 and -CAR], the gene for the enzyme arginase. Results from our laboratory indicate that arginine-activated ARGR products act negatively at ARG3 but also trigger CARI induction; it has been assumed that active ARGR repressor prevents the repressor specific for the CAR gene (CARGR) from functioning by forming an ARGR-CARGR inactive complex (12). However, it is not excluded that the ARG R products instead recognize a DNA target (i.e., the above-mentioned octanucleotide) in the 5' noncoding region of CAR] and in this way, directly or indirectly, counteract the negative effect of CARGR. Further mutations in the CACCTCTA sequence, in both ARG3 and CAR1, will be most interesting to examine.
3148
CRABEEL ET AL.
ACKNOWLEDGMENTS We thank P. Hilven for excellent and efficient assistance with these experiments. R.H. and K.V. were holders of a fellowship from the Instituut voor Aanmoediging van Wetenschappelijk Onderzoek in Nijverheid en Landbouw. This work was partly supported by a grant from the National Fonds voor Wetenschappelijk Onderzoek to M.C. and by grant no. 2.9012.83 from the Fonds voor Kollectief en Fundamenteel Onderzoek to N.G. 1. 2. 3. 4. 5.
6.
7.
8.
9.
10.
11.
12.
13.
14. 15.
16. 17.
18.
LITERATURE CITED Andreadis, A., Y. P. Hsu, G. B. Kohlaw, and P. Schimmel. 1982. Nucleotide sequence of yeast LEU2 shows 5' non-coding region has sequences cognate to leucine. Cell 31:319-325. Beacham, I. R., B. W. Schweitzer, H. M. Warrick, and J. Carbon. 1984. The nucleotide sequence of the yeast ARG4 gene. Gene 29:271-279. Bechet, J., M. Grenson, and J. M. Wiame. 1970. Mutations affecting the repressibility of arginine biosynthetic enzymes in S. cerevisiae. Eur. J. Biochem. 12:31-39. Berk, A., and P. Sharp. 1977. Sizing and mapping of early adenovirus mRNA's by gel electrophoresis of endonucleasedigested hybrids. Cell 12:721-732. Crabeel, M., R. Huygen, R. Cunin, and N. Glansdorff. 1983. The promoter region of the ARG3 gene in S. cerevisiae: nucleotide sequence and regulation in an ARG3-lacZ gene fusion. EMBO J. 2:205-212. Crabeel, M., F. Messenguy, F. Lacroute, and N. Glansdorff. 1981. Cloning ARG3, the gene for ornithine carbamoyltransferase from S. cerevisiae: expression in E. coli requires secondary mutations; production of plasmid 1-lactamase in yeast. Proc. Natl. Acad. Sci. USA 78:5026-5030. Cryer, D. R., R. Eccleshall, and J. Marmur. 1975. Isolation of yeast DNA. Methods Cell. Biol. 12:39-44. Delforge, J., F. Messenguy, and J. M. Wiame. 1975. The regulation of arginine biosynthesis in S. cerevisiae: the specificity of argR- mutations and the general control of amino acid biosynthesis. Eur. J. Biochem. 57:231-239. Donahue, T. E., R. S. Daves, G. Lucchini, and G. R. Fink. 1983. A short nucleotide sequence required for regulation of HIS4 by the general control system of yeast. Cell 32:89-98. Driscoll Penn, M., B. Galgoci, and H. Greer. 1983. Identification of AAS genes and their regulatory role in general control of amino acid biosynthesis in yeast. Proc. Natl. Acad. Sci. USA 80:2704-2708. Driscoll Penn, M., G. Thiraos, and H. Greer. 1984. Temporal analysis of general control of amino acid biosynthesis in Saccharomyces cerevisiae: role of positive regulatory genes in initiation and maintenance of mRNA derepression. Mol. Cell. Biol. 4:520-528. Dubois, E., D. Hiernaux, M. Grenson, and J. M. Wiame. 1978. Specific induction of catabolism and its relation to repression of biosynthesis in arginine metabolism of S. cerevisiae. J. Mol. Biol. 122:383-406. Dubois, E., and F. Messenguy. 1985. Isolation and characterization of the yeast ARG RII gene involved in regulating both anabolism and catabolism of arginine. Mol. Gen. Genet. 198:283-289. Greer, H., M. Penn, B. Hauge, and B. Galgoci. 1982. Control of amino acid biosynthesis in yeast. Recent Adv. Yeast Mol. Biol. 1:122-134. Hinnebusch, A. G. 1984. Evidence for translational regulation of the activator of general amino acid control in yeast. Proc. Natl. Acad. Sci. USA 81:6442-6446. Hinnebusch, A. G., and G. R. Fink. 1983. Repeated DNA sequences upstream of HIS1 also occur at several other coregulated genes in S. cerevisiae. J. Biol. Chem. 258:5238-5247. Hinnebusch, A. G., and G. R. Fink. 1983. Positive regulation in the general amino acid control of S. cerevisiae. Proc. Natl. Acad. Sci. USA 80:5374-5378. Jones, E. W., and G. R. Fink. 1982. Regulation of amino acid and nucleotide biosynthesis in yeast, p. 181-299. In J. N.
MOL. CELL. BIOL.
19. 20. 21.
22.
23.
24.
25. 26. 27. 28. 29. 30. 31. 32. 33. 34.
35. 36. 37.
38.
39.
40. 41.
Strathern, E. W. Jones, and J. R. Broach (ed.), The molecular biology of the yeast Saccharomyces cerevisiae: metabolism and gene expression. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1961. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. Lucchini, G., A. G. Hinnebusch, C. Chen, and G. R. Fink. 1984. Positive regulatory interactions of the HIS4 gene of Saccharomyces cerevisiae. Mol. Cell. Biol. 4:1326-1333. Maniatis, T., E. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Maxam, A., and W. Gilbert. 1980. Sequencing end-labelled DNA with base-specific chemical cleavages. Methods Enzymol. 65:499-560. Messenguy, F. 1976. Regulation of arginine biosynthesis in Saccharomyces cerevisiae: isolation of a cis-dominant, constitutive mutant for ornithine carbamoyltransferase synthesis. J. Bacteriol. 128:49-55. Messenguy, F., and E. Dubois. 1983. Participation of transcriptional and post-transcriptional regulatory mechanisms in the control of arginine metabolism in yeast. Mol. Gen. Genet. 189:148-156. Messenguy, F., M. Penninckx, and J. M. Wiame. 1971. Interaction between arginase and ornithine carbamoyltransferase in Saccharomyces cerevisiae. Eur. J. Biochem. 22:277-286. Miller, J. H. 1972. Assay of 0-galactosidase, p. 352-355. In J. H. Miller (ed.), Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Orr-Weaver, T. L., J. W. Szostak, and R. J. Rothstein. 1983. Genetic applications of yeast transformation with linear and gapped plasmids. Methods Enzymol. 100:228-245. Rigby, W., M. Dieckerman, C. Rhodes, and P. Berg. 1977. Labelling DNA to high specific activity in vitro by nick translation with DNA polymerase I. J. Mol. Biol. 113:237-251. Rose, M., P. Grifasi, and D. Botstein. 1984. Structure and function of the yeast URA3 gene: expression in E. coli. Gene 29:113-124. Rymond, B. C., R. S. Zitomer, D. Schumperli, and M. Rosenberg. 1983. The expression in yeast of the E. coli galK gene on a CYC1IgalK fusion plasmid. Gene 5:249-262. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 75:5463-5469. Scherer, S., and R. Davis. 1979. Replacement of chromosome segments with altered DNA sequences constructed in vitro. Proc. Natl. Acad. Sci. USA 76:4951-4955. Southern, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517. Struhl, K. 1981. Deletion mapping a eukaryotic promoter. Proc. Natl. Acad. Sci. USA 78:4461-4465. Struhl, K. 1982. Regulatory sites for HIS3 gene expression in yeast. Nature (London) 300:284-286. Struhl, K., and R. W. Davis. 1981. Transcription of HIS3 gene region in S. cerevisiae. J. Mol. Biol. 152:535-552. Sumrada, R. A., and T. G. Cooper. 1984. Nucleotide sequence of Saccharomyces cerevisiae arginase gene (CAR1) and its transcription under various physiological conditions. J. Bacteriol. 160:1078-1087. Thireos, G., M. Driscoll-Penn, and H. Greer. 1984. 5' Untranslated sequences are required for the translational control of a yeast regulatory gene. Proc. Natl. Acad. Sci. USA 81: 5096-5100. Werner, M., A. Feller, and A. Pierard. 1985. Nucleotide sequence of yeast gene CPA1 encoding the small subunit of arginine-pathway carbamoylphosphate synthetase. Eur. J. Biochem. 146:371-381. Zalkin, H., and C. Yanofsky. 1982. Yeast gene TRP5: structure, function, regulation. J. Biol. Chem. 157:1491-1500. Zaret, K. S., and F. Sherman. 1982. DNA sequence required for efficient transcription termination in yeast. Cell 28:563-573.
