Biochimica et Biophysica Acta, 1089 (1991) 47-53
47
~:~1991 Elsevier Science Publishers B.V. 0167-4781/91/$03.50 A D O N I S 016747819100127V BBAEXP 92241
Posttranscriptional regulation of the expression of MET2 gene of Saccharomyces cerevisiae Nicoletta Forlani, Fnzo Martegani and Lilia Alberghina Dipartimento di Fisiologia e Biochimica Generali, Sezione di Biochimica Comparata, Milano (Italyj
(Received 22 October 1990) (Revised manuscript received 14 January 1991)
Key words: Methionine synthesis: Posttranscriptional regulation: Homoserine-O-acetyltransferase: (S. cerevisiae)
The first step of the specific pathway for methionine biosynthesis in the yeast Sacclmromyces cerevisiae is catalyzed by the enzyme L-homoserlne-O-acetyltransferase (HSTase) (EC 23.131), encoded by the MET2 gene. in order to ascertain whether there is a posttranscriptional control on the MET2 gene expression, as suggested by previous results on the expression of the cloned gene, systems for high inducible expression of MET2 gene were constructed. In these constructs the MET2 gene was cloned in yeast expression vectors under the control of an inducible yeast GAL promoter element so that the MET2 was transcribed at very high levels under induced conditions. Measurements of the specific mRNA levels showed a strong stimulation of MET2 gene transcription in yeast traMtornumts grown on galactose as carbon source, corresponding to 50-100-fuld the repressed conditions, while only a 2-fuld increase of the enzymatic activity was observed. In addition, no evidence of a strong induced polypeptide of aplwopriate size on two dimensional gel electrophoresis was obtained. To understand the functional role of the non.coding 5 ' region of ME22 mRNA, we performed either a partial and a complete deletion of the 5 ' leader sequence, but even with these eonstngts an elevated mRNA level was not associated to a marked increase of the HSTase activity. These data support the idea of a posttranscriptional regulation of MET2 gene expression and show that the untranslated region of the specific mRNA is not inv~!:,~--Jin this regulatory mechanism.
Introduction The methionine biosynthetic pathway is rather complex; its two principal branches are connected to the one-carbon pool and sulfate metabolism of the cell and, in addition, some metabolites are intermediates also for asparagine, threonine and cysteine biosynthesis [1]. Several reports have shown that the expression of a group of enzymes for methionine production is co-regulated by repression of transcription in presence of the amino acid itself or its activated form S-adenosyimethionine (SAM) that appears to be the true effec-
Abbreviations: HSTase, homoserine-O-acetyltransferase(EC 2.3.1.31); YNB, yeast nitrogen base: SAM, S-adenosylmethionine: SDS, sodium dodecyl sulfate: UAS, upstream activation site; uORF. upstream open reading frame: [], designates plasmid-carrier state. Correspondence: L. Alberghina, Dipartimento di Fisiologia e Biochimica Generali, Sezb, ne di Biochimiea Comparata, Via Celoria 26, 20133 Milano, Italy.
tor of this negative control mechanism [2-5]. These enzymes and their genes are respectively: homoserineO-acetyltransferase, coded by the MET2 gene [2,6]; sulfate permease coded by CENI gene [7]; ATP sulfurylase coded by MET3 gene [8]; sulfite reductase putatively coded by METI, 4, 5, 8, 10, 19, 20 genes [3.6]; O-acetylhomoserine sulfhydrylase coded by MET25 gene [9]; the first of the two isoenzymes methionine-S-adenosyltransferase, coded by SAMI gene [10].
Nucleotide sequence analysis of the 5' flanking region of some of the genes mentioned above, MET2 [11], MET3 [8], MET25 [9], SAMI [10], revealed some homologous regions between 100 to 250 bases upstream of transcriptional start points, that may be relevant for the specific control [12,13]. Yet, the control elements in this pathway are far from being completely understood. We have previously cloned the MET2 gene [2] coding for the first enzyme of the specific pathway, the L-homoserine-O-acetyltransferase, that seems to be an intportam control point for the amino acid metabolism [1]. The MET2 mRNA was identified as a polyadenylated
48 species of 1700 nucleotides and its basal level in a wild type yeast strain appears to be slightly lower than 2-3 copies per cell. Determinations of HSTase activity performed on crude extracts [2] and on the partially purified enzyme [14] in the presence of methionine or SAM, demonstrated the inhibition by SAM but not by methionine, in addition, the transcription of MET2 gene was repressed by methionine and SAM [2]. These previous studies on the regulation of MET2 expression indicated that in yeast cells bearing the gene on a muhicopy plasmid, the amount of the specific mRNA was 10-fold higher than in wild type untransformed cells, but HSTase activity was only slightly higher than in wild type strain [2]. This lack of correspondence between MET2 mRNA and HSTase activity levels, in contrast with the parallel increase of both reported for MET3 gene [8], may be due to a posttranscriptional control of MET2 expression. To investigate ~bout this point we constructed inducible expression systems of the MET2 gene in order to obtain an increase in the intracellular amount of specific mRNA and to study the corresponding enzymatic activity level. We also generated a complete deletion of the 5' untranslated gene region to examine its eventual role in this type of modulation. Materials and Methods
Strains, media and growth conditions The following S. cerevisiae strains were used in this work: $288C ( MATa, SUC2, gal2, CUP1) and X40043A (MA Ta, lysS, met2, ura3, trpl ) obtained from Yeast Genetic Stock Center. Yeast cells were grown with shaking at 30°C in 0.67~ Yeast Nitrogen Base without amino acids (YNB, Difco) minimal medium supplemented with 2% glucose or 2~ raffinose or 2% galactose, and with the appropriate amino acids when necessary. Before transformation, yeast cells were grown on complete YEPD medium (glucose 2~, peptone 2%, yeast extract 1%) supplemented with tryptophan at 50 pg/ml). Growth was monitored by counting the number of cells/ml of culture with a Coulter Counter ZB! (Coulter Electronics) [15]. Escherichia coil strains HB101 and JM101 were used as hosts for recombinant DNA manipulations. Bacteria were grown in LB medium [16] at 37°C with ampicillin at 100 p g / m l . Piasmid construction and DNA manipulation Different DNA fragments containing the whole or a partially deleted MET2 gene region were cloned in the inducible expression vectors pBCL26 [17] derived from pLG2 plasmid [18], or pEMBLYex4 vector [19[. Standard DNA manipulations were performed according to general methods reported by Maniatis et al. [16]. Total
yeast DNA was extracted from transformant cells by the method of Cryer et al. [201. Enzymes were purchased from Promega or from Boehringer-Mannheim. The standard procedure of CaCI 2 was used to transform E. coil strains [18]. Yeast cells were transformed according to the LiC! method [21]. For the deletion of the 5' region of MET2 gene the 3 kb Xbal-BamHI DNA fragment bearing the whole M E T 2 0 R F and a 107 bp 5' flanking region (Fig. 1) was cloned in the corresponding restriction sites of pGEMbhi vector (Promega), generating plasmid pGMET2, which resulted in a useful system for DNA deletion and sequencing analysis. Plasmid pGMET2, digested with PstI and Xbal, was then treated with exonuclease I i i according to the method of Henikoff [22]. The extent of the deletions was verified by analysis of aliquots digested with Bgill or Clal on 1% agarose gel electrophoresis. After filling the DNA protruding ends with Klenow D N A polymerase, plasmids were then ligated with T4 DNA ligase and used to transform E. coil JM101 competent cells selecting on LB plates for ampicillin resistance. Plasmid DNA was extracted and purified on a CsC! gradient, the nucleotide sequence of MET2 deletions was determined by the dideoxy-chain termination method using the GemSeq K / R T Sequencing System (promega) in the presence of [3sS]dATP (specific activity > 1000 Ci/mmol, Amersham).
RNA preparation and Northern blots hybridization Total RNA was prepared from exponentially growing cells as previously described [2]. After electrophoresis on 1% agarose formaldehyde gels [16[, RNA was blotted to Hybond-N membrane (Amersham). Hybridization was performed with in vitro labeled anti-mRNA corresponding to the 0.5 kb EcoRl-XbaI fragment of MET2 gene obtained by SP6 RNA polymerase transcription in the presence of [a-32p]UTP (specific activity 800 Ci/mmol, Amersham) as specified by Baroni et al. [2]. Enzyme assays Yeast cell free extracts were made as described by Cherest et ai. [6] with the modifications reported by Baroni et al. [2]. HSTase activity was assayed as described by Nagai and Kerr [23], using the homoserineO-acetylhomoserine exchange reaction with [14C]homoserine (Amersham) as a substrate. O-acetylhomoserine was synthesized in our laboratory according to Nagai and Flavin [24]. Protein concentration was measured with the Bio-Rad protein assay using bovine serum albumin as reference standard. Labeling of yeast protein and two dimensional gel-electrophoresis Subcultures (5-10 ml) of transformed and control yeast strains grown in minimal medium lacking both
49
methionine and uracil at a density of approx, l0 n cell/ml, were labeled with [3-~S]methionine (10/~Ci/ml, 1000 Ci/mmoi, Amersham) for 30 rain. Cells were rapidly harvested by filtration and processed as described by Popolo et al. [25]. Approx. 8- 10 ~ cpm were subjected to two dimensional electrophoresis according to O'Farrell [26]. The pH range used was 7.3-4.2. Separation on the second dimension was carried out in an LKB vertical system with 10% SDS-polyacrylamide gels. Standard proteins labeled with 14C were used as molecular weight markers (BRL). Gels were fixed for 30 min in isopropanol/water/acetic acid (25:65:10) after ~he run. The gels were then treated with Amplify (Amersham) dried on a Bio-Rad gel dryer and exposed to Kodak XR-OMAT films at - 7 0 ° C for 2-3 days. Quantitation of HSTase was clone by counting the radioactivity of the spots cut from two dimensional gels; actin was used as an internal standard [27]. Each cut spot was placed in the bottom of a scintillation vial with 0.025 ml of water, After adding 0.5 ml of Protosol (New England Nuclear), the vial was tightly capped and heated at 55-60°C for 30 min. After cooling, 5 mi of Econofluor-2 (New England Nuclear) were added and the sample was counted. In vitro transcription and translation of M E T 2 gene Plasmid p G M E T 2 was linearized before the transcription reaction by digestion with BamHl. in vitro transcription was carried out with SP6 RNA polymerase with the ribonucleoside triphosphates at 500 /tM, incubating in presence of m~G(5')ppp(5')G (Pharmacia) for capping at 40°C [28]. For a final volume of 25/~l, approx. 1/~g of linearized DNA was used. The RNA was then translated in rabbit reticulocyte lysate (Promega) with 1 ~aCi//tl of [35S]methionine (1200 C i / m M , Amersh,qm) in a 25/tl volume containing up to 5 itg RNA, at 30°C for 60-90 rain. Aliquots were taken for analysis of translation products by SDS-polyacrylamide and two dimensional gel electrophoresis. Results
Construction of a M E T 2 inducib/e expression system The M E T 2 gene coding for HSTase of S. cerevisiae was previously cloned in our laboratory from a yeast gene bank [2]. The restriction map of the structural gene and of its 5' and 3" flanking regions are shown in Fig. 1. To verify the presence of a posttranscriptional control on the regulation of M E T 2 gene expression we constructed an inducible expression system by cloning the M E T 2 gene in pBCL26 and pEMBLYea4 vectors. Both vectors contain an inducible galactose promoter that is a hybrid of the 365 bp UASoA L, and the 250 bp promoter elements of the C Y C i yeast gene. Every gene w a s inserted in the correct orientation after this hybrid promoter is transcribed under induced conditions at high levels.
ATG H
P'2 Xl
E S
'
"
C
82
E
$I
H
Bt
MIET2 ATG Xl
pBMX1
E S
"
~emme.:-
]]
C
B2
E
;
I
]
C
92
E
I
i
i
C
ATG Xl
pEMBMet
{:~z:E~-~
E S
"
i[ ATG
H
p±MET2.57
E S
-.~L~_"
82
E
)i:
I
i
I
E S
C
62
E
i
i
ATG H
p~ ML:T2-S
,m~=r
"
ll"
I
Fig. 1. Map of MET2 gene and of inducible piasmids constructed.
The black bar representsthe MET20RF (1310 bp). The positionof the two putative TATA box (-222 to -210; -124 to -118) is specified by the symbole. Transcriptionstarts (-59. -52: -40. - 3"t)are indicatedby arrows and the ATG is marked,t3 indicatesthe polyadenylation sequence [11]. The dashed bar represents the U,4SoALH0 intergenicregionof 365 bp and the dotted bar the 250 bp CYC! untranslatedleader [18l. pAMET2-57 and pAMET2-5 were obtained after MET2 leader deletion with exonucleaseIlL B1, BamHl; B2. Bglil; C, Clai; E, EcoRi; H, Hindlll; !)2. Pvull; S, Sstl; Si. Saul; X1. Xbal.
In the first construction, the pBMX1 plasmid, the 3.1 kb P v u l l - B a m H l DNA fragment containing the M E T 2 gene has been cloned in the Xhol site (blunted with Klenow DNA polymerase) and the B a m H l restriction site of the expression vector pBCL26. Since the 5" region extends for 201 nucleotides the M E T 2 gene maintains its own putative TATA box sequence [11]; its transcription is controlled by a hybrid U A S G ^ t f M E T 2 promoter. In the second plasmid, pEMBMet, the 2.65 kb Xbal-Hindlll DNA fragment containing the M E T 2 gene has been cloned in the same restriction sites of the expression vector pEMBLYex4. In this case the 5" region extends for 107 nucleotides and the two putative TATA box sequences of M E T 2 gene were deleted; thus, the expression of M E T 2 is under the control of the inducible hybrid promoter U A S ~ A L / C Y C i . These constructions are shown in Fig. 1. in each of these two constructions the transcription termination and polyadenylation sites are those of the M E T 2 gene [11]. The yeast strain X4004-3A was transformed to the Ura + phenotype either with pBMX1 or pEMBMet plasmids. Some of the indelx~dent transformants obtained with the pBMX1 plasmid and with the pEMBMet plasmid were analyzed by Southern blotting to verify the presence of plasmids in yeast cells (data not
shown). Both transformants with pBMXI and pEMBMet grow well in glucose minimal medium without
50 TABLE !
Relatwe levels of MET2 mRNA and HSTase enzymatic actit'ity in transformants
Strain
Carbon source ~
X4004 $288C X400~pBMXI] X400~pBMX1] X400~pEMBMctt X4004[pEMBMetI X400~pAMET2-57] X400~pAMET2-51
glucose glucose glucose galactose glucose galactose galactose galactose
Specific mRNA levels~' 100 2 > 100 > 100 >100
HSTase specific activity" none 1 0.5 2.2 0.8 3 1.9 2.1
i All strains were tested in their exponential phase of growth (approx. (1-2). l0 s cell/ml). Growth was in minimal medium, supplemented with the required amino acids for the untransformed X4004-3A strain, but without uracil and methionine for all the others. b For mRNA levels,the densitometric analysis of autoradiograms are relative to the value detected it, the wild type strain $288C, defined as i. HSTase specific activity is expressed as nmol of acetylhomoserine produced per min per nag of extracted protein. The data reported in this table are the mean of at least three different experiments. methionine, indicating that the expression of the cloned MET2 gene was not completely repressed by glucose (see also Table l). The inducible expression systems were then tested by Northern blot analysis of total R N A extracted from several X4004-3A transformants during exponential growth in minimal selective medium (YNB) supplemented with glucose 2% (repressing conditions), or galactose 2~ (inducing conditions). As shown in Fig. 2 there is a strong induction of MET2 m R N A in galactose growing cells transformed both with pBMX1 and pEMBMet. The amount of specific m R N A was approximately determined by densitometry of autoradiographic films. Under inducing conditions the MET2
A
B
C
D
E
F
G
H
Fig. 2. Northern blot analysis of MET2 mRNA, Total RNA was extracted from X4004-3A transformants during the exponential phase
of growth ((1-2).106 cell/ml) and from the wild type strain $28gc grown in minimal medium lacking both methionine and uracil. 20/~g were loaded on denaturing gels for each lane. Alter blotting, filters were hybridized with a MET2 riboprobe. A, X4004[pBMX1] in galactose; B, X4004[pEMBMet] in galactose; C. X4004[pBMXIi in glucer,e; D, X4004[pEMBMet] in glucose; E, untransformed wild type 5288C in glucose; F, X4004[I~MET2-57] in galactose; G, X~0411~MET2-$TI in raffinose; H. X4004[p~MET2-$] in galactose.
m R N A amount appears to increase at least 50-fold in comparison with the level measured in repressing conditions and approx. 100-fold in comparison with a untransformed wild type. N o differences were observed between the two plasmids, although the promoter sequences were different. The corresponding enzymatic activity of HSTase was measured with the exchange reaction between acetyI-Lhomoserine and L-[t4C]homoserine [23]. As clearly shown in Table l, the enzymatic activity in transformed cells under induced conditions increased to a very limited extent compared to the level measured in the control, in spite of the large increment of MET2 m R N A .
Deletion of 5" MET2 untranslated region Because the HSTase activity does not parallel a large increment of its correspondent messenger, we searched for an eventual role of the leader region of m R N A in the posttranscriptional control, since 107 bp of leader sequence of the MET2 messenger were still conserved in p E M B M e t plasmid. N o upstream open reading frames (uORF) were identified in this region but some inverted repeats are present ( - 8 0 / - 44 and - 9 / + 61 region of the sequence [11]) that might contribute to the formation of a 'stem and loop' secondary structure [29]. In order to test the functional relevance of these 107 bp, we generated a partial and a complete deletion of this region. Two deletions of the MET2 5" region were selected in which respectively 57 and 5 bases before the A T G codon were preserved, and were cloned in the pEMBLYex4 expression vector. The two recombinant plasmids called p A M E T 2 - 5 7 and p A M E T 2 - 5 were used to transform X4004-3A strain. Yeast transformants bearing these two plasmids were analyzed by Southern and Northern blotting as previously described and a similar increase of M E T 2 m R N A as before was demonstrated in cells grown under inducing conditions (Fig. 2). The HSTase activity assay, in spite of the high MET2 m R N A level, again showed a very limited increase (Table I). Yeast cells transformed with pA M E T 2 - 5 7 or p A M E T 2 - 5 plasmid showed, however, a strong repression of MET2 expression in glucose medium, since they failed to grow in minimal glucose medium without methionine, while they grow well in galactose medium.
Analysis of yeast proteins b~ two dimensional gel electrophoresis The preceeding experiments demonstrated a lack of correspondence between the m R N A levels and HgTase activity, however, they did not demonstrate a translational regulation since an inactive protein could be equally induced to high levels. To investigate whether the posttranscriptionai control of MET2 gene expression is due to a lack of protein synthesis or by modulation of enzyme activity, a two dimensional electrophore-
51 sis analysis was performed. The theoretical isoelectric point ( p l ) of HSTase calculated from amino acid composition is 5.47. This p l value was confirmed by isoelectric focusing of HSTase protein synthesized in vitro. A synthetic MET2 mRNA obtained by in vitro transcription of MET2 gene cloned in pGEMblu vector was used to direct the synthesis of HSTase protein in a reticulocyte lysate. A two dimensional gel ¢lectrophoresis of translation products indicated that the MET2 mRNA is efficiently translated in vitro, showing a single polypeptide of approx. 50 kDa (Fig. 3A) and p l of 5.5 as expected on the basis of the ORF derived by sequencing the gene [11]. X4004-3A[pEMBMet] transformants, grown in selective medium where labeled with [35S]methionine under repressing or inducing conditions. Total protein was extracted and separated on two dimensional gels. The availability of HSTase polypeptide translated in vitro allowed us also to identify a yeast protein comigrating on two dimensional gels with the enzyme synthesized in vitro (Fig. 3B). This spot is located near the '130' spot identified to be a hexokinase [27]. Total extracts from $288C strain were used as untransformed controls (data not shown). Autoradiography of the gels showed the presence of several proteins induced by galactose, but we were not able to detect a strong induction of a protein with a p l = 5.47 and a size of approx. 50 kDa. Irdeed, in transformant
cells grown under inducing conditions we carl detect a slight induction (approx. 2.5-fold) of a spot that comigrates with the HSTase protein in comparison with repressing conditions (Fig.3C,D). This is comparable to the increase of enzymatic activity measured in the same growing conditions (see Fig. 3 and Table I). Discussion The MET2 gene, coding for L-homoserine-Oacetyltransferase of S. cerevisiae, belongs to the so called group I methionine biosynthetic genes. These genes are under a coordinate specific transcriptional repression by methionine and S-adenosylmethionine [5]. A comparative analysis of the 5' region of MET2, MET3, MET25 and SAMI genes [10,13] revealed some common motifs that may be important for the control of gene expression. These cis-acting elements consist in an UASmet, required for maximum expression (5'TCACGTGA), located at - 3 7 4 , and in two regions that may be involved in methionine transcriptional repression (URSmet) ( 5 ' - A A A A A T T G T G T and 5'AACTAAGTC), respectively located at - 3 0 9 and - 297. Preliminary findings of our laboratory suggested, in addition, that posttranscriptional regulation operates for MET2 expression. In fact, we found that in a strain transformed with a multicopy plasmid, bearing the
EF kOa
4.2.Z3 A
IZ3
4.2 t C
69-q) 41P
¢> e
46- 0
30-1)
•
"t-
4 6 - IR *
3 0 - ,,
oe
,r
69-p .
.qb~ll~
.~*
"
o
-
IP
....................
-~-
:.........
Fig. 3. Two dimensional gel clcctrophorcsis of yeast proteim. (A) HSTas¢ p o l ~ p l i d ¢ translated in vitro. (B) Total yeast pmleins (200000 ¢pm) from X4(]O4-3A[pEMBMet], grown in glucose, mixed with in HSTase translated in vitro. (C) Total yeast proteins (800000 cpm) from
X4004-3A[pEMBMetlgrownin glucosemediumand (D) in galactnsemedium.
52 cloned M£F2 gene, the level of the specific MET2 mRNA increases approx. 10-fold, while the enzymatic activity of HSTase increased less than 2-fold, in comparison with untransformed wild type cells [2]. To obtain substantial evidence for a posttranscriptional modulation of MET2 expression, we cloned the MET2 gene in a plasmid vector, under the control of the inducible UASGAL element [181, making then progressive deletions of the 5' flanking region. in this way we obtained an inducible system for MET2 transcription, and in transformed yeast, grown in galactose medium, the steady state level of specific MET2 mRNA is increased at least 100-fold, in comparison with an untransformed wild type cell. Again the specific activity of HSTase increased only 2-fold, indicating that a strong posttranscriptional control may be operative. Furthermore, the absence of an induced polypeptide in yeast cells grown under induced conditions, analyzed by two dimensional gel electrophoresis, suggests translational control. This translational control, however, seems to be operative only in yeast cells, since the MET2 mRNA is easily translated in vitro in an heteroiogous system (i.e., reticuiocyte lysate). In yeast the mRNA untranslated 5' leader sequence may be important for a translational control. In fact, the initiation of translation in yeast follows the 'scanning model' [30] and the presence of A U G or short Open Reading Frame (uORF) upstream the initiation codon inhibits translation and this fact is at the basis of the translational control reported for GCN4 and CPA 1 genes [31,32]. The MET2 mRNA leader sequence does not contain an uORF; however, in yeast the formation of a secondary structure in the mRNA is important for translatability of the message and could provide a translational control element [29,34]. We have found some inverted repeats around the region containing the A U G codon of MET2 mRNA that could generate a secondary structure. However, the complete deletion of the leader region, up to 5 bases upstream the A U G did not increase, in induced cells, the level of HSTase, indicating that the 5' leader region of mRNA is not relevant for this posttranscriptional regulation. Since the untranslated 5' region appears not to be involved in the translational regulation of MET2 gene expression, several alternative regulatory mechanisms could be hypothesized. Sequences in the trailer part of messenger RNA or sequences within the ORF may influence the translatability of this mRNA in yeast, [34]. Internal sequences within the O R F might be recognized by a specific unknown binding protein, or MET2 protein itself could autoregulate its synthesis, as demonstrated for the synthesis of some ribosomal proteins [35]. The observed data could be also explained by a regulated fast degradation of the protein [35]. However, the half-life should be very short, in the range of few minutes, since we were not able to found a strong
increase of MET2 protein on two dimensional gels after a 30 min pulse labelling. Interestingly, evidence of posttranscriptional modulation of gene expression has also been reported for another methionine biosynthetic gene, MET25 [13], but whose regulatory mechanism is not currently known. Acknowledgements This work was partially supported by a MPI 60% grant to L.A., N.F. was supported by an A.I.R.C. (Associazione ltaliana Ricerca sul Cancro) fellowship. References I Jones, E.W. and Fink, G.R. (1982) In The molecular biology of the yeast Saccharomyces cerevisiae (Strathern, J.H., Jones, E.W. and Broach, J.R., eds.L pp. 181-300. Cold Spring Harbor, New York. 2 Baroni, M., Livian, S., Martegani, E. and Alberghina, L. (1986) Gene 46, 71-78. 3 Cherest, H., Surdin-Kerjan, J.. Antoniewski, J. and De Robichon Szulmajster, H. (1973)J. Bacteriol. 114, 928-933. 4 Cherest, H., Surdin-Kerjan, Y.. Antoniewski, J. and De Robichon-Szulmajster, H. (1973) J. Bacteriol. 115. 1084-1090. 5 Thomas, D., Rothstein, R., Rosenberg. N. and Surdin-Kerjan, Y. (1988) Mol. Cell. Biol. 8, 5132-5139. 6 Cherest, H., Surdin-Kerjan, Y. and De Robichon-Szulmajster, H. (1971) J. Bacteriol. 106, 758-772. 7 Breton, A. and Surdin-Kerjan, Y. (1977) J. Bacteriol. 132, 224-232. 8 Cherest, H., Thao, N.N. and Surdin-Kerjan, Y. (1985) Gene 34, 269-281. 9 Sangsoda, S., Cherest, FI. and Surdin-Kerjan, Y. (1985) Mol. Gen, Genet. 200. 407-414. 10 Thomas, D. and Surdin-Kerjan, Y. (1987) J, Biol. Chem. 262, 16704-16709. 11 Langin, T., Faugeron, G., Goyon, C., Nicolas, A. and Rossignol, J.L. (1986) Gene 49, 283-293. 12 Cherest, H., Kerjan, P. and Surdin-Kerjan, Y. (1987) Mol. Gen. Genet. 210, 307-313. 13 Thomas, D., Cherest, H. and Surdin-Kerjan, Y. (1989) Mol. Cell. Biol. 9, 3292-3298, 14 Yamagata, S. (1987) J. Bacteriol. 169, 3458-3463. 15 Vanoni, M., Vai, M., Popolo, L. and Alberghina, L. (1983) J. Bacteriol. 156, 1282-1291. 16 Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecular cloning. A laboratory, manual. Cold Spring Harbor Laboratory, New York. 17 Velati-Bellini, A., Pedroni, P., Martegani, E. and AIberghina, L. (1986) Appl. Microbiol. Biotechnol. 25, 124-131. 18 Guarente, L. (1983) Methods Enzymol. 101,181-191. 19 Baldari, C., Murray, J.A.H., Ghiara, P., Cesareni. (3. and Galeoni, G.L. (1987) EMBO J. 6, 229-234. 20 Cryer, D.R., Eccleshall. R. and Marmur. J. (1975) Methods Cell Biol. 12, 39-44. 21 lto, H., Fukuda, Y., Murata, K. and Kimnra, A. (1983) J. Bacteriol. 153, 163-168. 22 Henikoff, S. (1984) Gene 28, 351-359. 23 Nagai, S. and Kerr, D. (1971) Methods Enzymol. I7B, 442-445. 24 Nasal, S. and Flavin, M. (1971) Methods Enzymol. ITB, 423-424. 25 Popolo, L. and Alberghina, L. (1984) Proc. Natl. Acad. Sci. USA 81, 120-124. 26 O'Farrell, P.H. (1975) J. Biol. Chem. 250, 4007-4021. 27 Ludwig. J.R., Foy, J.J,, Elliot, S.G, and McLaughlin, C,S. (1982) Mol. Cell, Biol. 2, 117-126.
53 28 Krieg. P.A, and Melton. D,A. (1984) Nucleic Acids Res. 12, 7057-7O70. 29 Bairn. S.B. and Sherman, F. (1988) Mol. Cell. Biol. 8, 1591-1601. 30 Kozak, M. (1989) J. Cell. Biol, 108, 229-241. 31 Mueller, P.P. and Hinnesbuch, A.G. (1986) Cell 45. 201-207. 32 Werner, M., Feller, A., Messenguy, F. and Pierrard, A. (1987) Cell 49, 805-813.
33 Surdin-Kerjan, Y. and De Rohichon Szulmajster, H. (1975) J. Bacteriol. 122. 367-374. 34 Van den Heuvel, J.J., Planta, R.J. and Raut$, H.A. (1990) Yeast 6. 473-482. 35 Warner. J.R.. Mitra. O.. Schwindiger, W.F., Studer~y, M. and Fried, H.M. 0985) Mol. Cell. Biol. 5, 1512-1521.
47
~:~1991 Elsevier Science Publishers B.V. 0167-4781/91/$03.50 A D O N I S 016747819100127V BBAEXP 92241
Posttranscriptional regulation of the expression of MET2 gene of Saccharomyces cerevisiae Nicoletta Forlani, Fnzo Martegani and Lilia Alberghina Dipartimento di Fisiologia e Biochimica Generali, Sezione di Biochimica Comparata, Milano (Italyj
(Received 22 October 1990) (Revised manuscript received 14 January 1991)
Key words: Methionine synthesis: Posttranscriptional regulation: Homoserine-O-acetyltransferase: (S. cerevisiae)
The first step of the specific pathway for methionine biosynthesis in the yeast Sacclmromyces cerevisiae is catalyzed by the enzyme L-homoserlne-O-acetyltransferase (HSTase) (EC 23.131), encoded by the MET2 gene. in order to ascertain whether there is a posttranscriptional control on the MET2 gene expression, as suggested by previous results on the expression of the cloned gene, systems for high inducible expression of MET2 gene were constructed. In these constructs the MET2 gene was cloned in yeast expression vectors under the control of an inducible yeast GAL promoter element so that the MET2 was transcribed at very high levels under induced conditions. Measurements of the specific mRNA levels showed a strong stimulation of MET2 gene transcription in yeast traMtornumts grown on galactose as carbon source, corresponding to 50-100-fuld the repressed conditions, while only a 2-fuld increase of the enzymatic activity was observed. In addition, no evidence of a strong induced polypeptide of aplwopriate size on two dimensional gel electrophoresis was obtained. To understand the functional role of the non.coding 5 ' region of ME22 mRNA, we performed either a partial and a complete deletion of the 5 ' leader sequence, but even with these eonstngts an elevated mRNA level was not associated to a marked increase of the HSTase activity. These data support the idea of a posttranscriptional regulation of MET2 gene expression and show that the untranslated region of the specific mRNA is not inv~!:,~--Jin this regulatory mechanism.
Introduction The methionine biosynthetic pathway is rather complex; its two principal branches are connected to the one-carbon pool and sulfate metabolism of the cell and, in addition, some metabolites are intermediates also for asparagine, threonine and cysteine biosynthesis [1]. Several reports have shown that the expression of a group of enzymes for methionine production is co-regulated by repression of transcription in presence of the amino acid itself or its activated form S-adenosyimethionine (SAM) that appears to be the true effec-
Abbreviations: HSTase, homoserine-O-acetyltransferase(EC 2.3.1.31); YNB, yeast nitrogen base: SAM, S-adenosylmethionine: SDS, sodium dodecyl sulfate: UAS, upstream activation site; uORF. upstream open reading frame: [], designates plasmid-carrier state. Correspondence: L. Alberghina, Dipartimento di Fisiologia e Biochimica Generali, Sezb, ne di Biochimiea Comparata, Via Celoria 26, 20133 Milano, Italy.
tor of this negative control mechanism [2-5]. These enzymes and their genes are respectively: homoserineO-acetyltransferase, coded by the MET2 gene [2,6]; sulfate permease coded by CENI gene [7]; ATP sulfurylase coded by MET3 gene [8]; sulfite reductase putatively coded by METI, 4, 5, 8, 10, 19, 20 genes [3.6]; O-acetylhomoserine sulfhydrylase coded by MET25 gene [9]; the first of the two isoenzymes methionine-S-adenosyltransferase, coded by SAMI gene [10].
Nucleotide sequence analysis of the 5' flanking region of some of the genes mentioned above, MET2 [11], MET3 [8], MET25 [9], SAMI [10], revealed some homologous regions between 100 to 250 bases upstream of transcriptional start points, that may be relevant for the specific control [12,13]. Yet, the control elements in this pathway are far from being completely understood. We have previously cloned the MET2 gene [2] coding for the first enzyme of the specific pathway, the L-homoserine-O-acetyltransferase, that seems to be an intportam control point for the amino acid metabolism [1]. The MET2 mRNA was identified as a polyadenylated
48 species of 1700 nucleotides and its basal level in a wild type yeast strain appears to be slightly lower than 2-3 copies per cell. Determinations of HSTase activity performed on crude extracts [2] and on the partially purified enzyme [14] in the presence of methionine or SAM, demonstrated the inhibition by SAM but not by methionine, in addition, the transcription of MET2 gene was repressed by methionine and SAM [2]. These previous studies on the regulation of MET2 expression indicated that in yeast cells bearing the gene on a muhicopy plasmid, the amount of the specific mRNA was 10-fold higher than in wild type untransformed cells, but HSTase activity was only slightly higher than in wild type strain [2]. This lack of correspondence between MET2 mRNA and HSTase activity levels, in contrast with the parallel increase of both reported for MET3 gene [8], may be due to a posttranscriptional control of MET2 expression. To investigate ~bout this point we constructed inducible expression systems of the MET2 gene in order to obtain an increase in the intracellular amount of specific mRNA and to study the corresponding enzymatic activity level. We also generated a complete deletion of the 5' untranslated gene region to examine its eventual role in this type of modulation. Materials and Methods
Strains, media and growth conditions The following S. cerevisiae strains were used in this work: $288C ( MATa, SUC2, gal2, CUP1) and X40043A (MA Ta, lysS, met2, ura3, trpl ) obtained from Yeast Genetic Stock Center. Yeast cells were grown with shaking at 30°C in 0.67~ Yeast Nitrogen Base without amino acids (YNB, Difco) minimal medium supplemented with 2% glucose or 2~ raffinose or 2% galactose, and with the appropriate amino acids when necessary. Before transformation, yeast cells were grown on complete YEPD medium (glucose 2~, peptone 2%, yeast extract 1%) supplemented with tryptophan at 50 pg/ml). Growth was monitored by counting the number of cells/ml of culture with a Coulter Counter ZB! (Coulter Electronics) [15]. Escherichia coil strains HB101 and JM101 were used as hosts for recombinant DNA manipulations. Bacteria were grown in LB medium [16] at 37°C with ampicillin at 100 p g / m l . Piasmid construction and DNA manipulation Different DNA fragments containing the whole or a partially deleted MET2 gene region were cloned in the inducible expression vectors pBCL26 [17] derived from pLG2 plasmid [18], or pEMBLYex4 vector [19[. Standard DNA manipulations were performed according to general methods reported by Maniatis et al. [16]. Total
yeast DNA was extracted from transformant cells by the method of Cryer et al. [201. Enzymes were purchased from Promega or from Boehringer-Mannheim. The standard procedure of CaCI 2 was used to transform E. coil strains [18]. Yeast cells were transformed according to the LiC! method [21]. For the deletion of the 5' region of MET2 gene the 3 kb Xbal-BamHI DNA fragment bearing the whole M E T 2 0 R F and a 107 bp 5' flanking region (Fig. 1) was cloned in the corresponding restriction sites of pGEMbhi vector (Promega), generating plasmid pGMET2, which resulted in a useful system for DNA deletion and sequencing analysis. Plasmid pGMET2, digested with PstI and Xbal, was then treated with exonuclease I i i according to the method of Henikoff [22]. The extent of the deletions was verified by analysis of aliquots digested with Bgill or Clal on 1% agarose gel electrophoresis. After filling the DNA protruding ends with Klenow D N A polymerase, plasmids were then ligated with T4 DNA ligase and used to transform E. coil JM101 competent cells selecting on LB plates for ampicillin resistance. Plasmid DNA was extracted and purified on a CsC! gradient, the nucleotide sequence of MET2 deletions was determined by the dideoxy-chain termination method using the GemSeq K / R T Sequencing System (promega) in the presence of [3sS]dATP (specific activity > 1000 Ci/mmol, Amersham).
RNA preparation and Northern blots hybridization Total RNA was prepared from exponentially growing cells as previously described [2]. After electrophoresis on 1% agarose formaldehyde gels [16[, RNA was blotted to Hybond-N membrane (Amersham). Hybridization was performed with in vitro labeled anti-mRNA corresponding to the 0.5 kb EcoRl-XbaI fragment of MET2 gene obtained by SP6 RNA polymerase transcription in the presence of [a-32p]UTP (specific activity 800 Ci/mmol, Amersham) as specified by Baroni et al. [2]. Enzyme assays Yeast cell free extracts were made as described by Cherest et ai. [6] with the modifications reported by Baroni et al. [2]. HSTase activity was assayed as described by Nagai and Kerr [23], using the homoserineO-acetylhomoserine exchange reaction with [14C]homoserine (Amersham) as a substrate. O-acetylhomoserine was synthesized in our laboratory according to Nagai and Flavin [24]. Protein concentration was measured with the Bio-Rad protein assay using bovine serum albumin as reference standard. Labeling of yeast protein and two dimensional gel-electrophoresis Subcultures (5-10 ml) of transformed and control yeast strains grown in minimal medium lacking both
49
methionine and uracil at a density of approx, l0 n cell/ml, were labeled with [3-~S]methionine (10/~Ci/ml, 1000 Ci/mmoi, Amersham) for 30 rain. Cells were rapidly harvested by filtration and processed as described by Popolo et al. [25]. Approx. 8- 10 ~ cpm were subjected to two dimensional electrophoresis according to O'Farrell [26]. The pH range used was 7.3-4.2. Separation on the second dimension was carried out in an LKB vertical system with 10% SDS-polyacrylamide gels. Standard proteins labeled with 14C were used as molecular weight markers (BRL). Gels were fixed for 30 min in isopropanol/water/acetic acid (25:65:10) after ~he run. The gels were then treated with Amplify (Amersham) dried on a Bio-Rad gel dryer and exposed to Kodak XR-OMAT films at - 7 0 ° C for 2-3 days. Quantitation of HSTase was clone by counting the radioactivity of the spots cut from two dimensional gels; actin was used as an internal standard [27]. Each cut spot was placed in the bottom of a scintillation vial with 0.025 ml of water, After adding 0.5 ml of Protosol (New England Nuclear), the vial was tightly capped and heated at 55-60°C for 30 min. After cooling, 5 mi of Econofluor-2 (New England Nuclear) were added and the sample was counted. In vitro transcription and translation of M E T 2 gene Plasmid p G M E T 2 was linearized before the transcription reaction by digestion with BamHl. in vitro transcription was carried out with SP6 RNA polymerase with the ribonucleoside triphosphates at 500 /tM, incubating in presence of m~G(5')ppp(5')G (Pharmacia) for capping at 40°C [28]. For a final volume of 25/~l, approx. 1/~g of linearized DNA was used. The RNA was then translated in rabbit reticulocyte lysate (Promega) with 1 ~aCi//tl of [35S]methionine (1200 C i / m M , Amersh,qm) in a 25/tl volume containing up to 5 itg RNA, at 30°C for 60-90 rain. Aliquots were taken for analysis of translation products by SDS-polyacrylamide and two dimensional gel electrophoresis. Results
Construction of a M E T 2 inducib/e expression system The M E T 2 gene coding for HSTase of S. cerevisiae was previously cloned in our laboratory from a yeast gene bank [2]. The restriction map of the structural gene and of its 5' and 3" flanking regions are shown in Fig. 1. To verify the presence of a posttranscriptional control on the regulation of M E T 2 gene expression we constructed an inducible expression system by cloning the M E T 2 gene in pBCL26 and pEMBLYea4 vectors. Both vectors contain an inducible galactose promoter that is a hybrid of the 365 bp UASoA L, and the 250 bp promoter elements of the C Y C i yeast gene. Every gene w a s inserted in the correct orientation after this hybrid promoter is transcribed under induced conditions at high levels.
ATG H
P'2 Xl
E S
'
"
C
82
E
$I
H
Bt
MIET2 ATG Xl
pBMX1
E S
"
~emme.:-
]]
C
B2
E
;
I
]
C
92
E
I
i
i
C
ATG Xl
pEMBMet
{:~z:E~-~
E S
"
i[ ATG
H
p±MET2.57
E S
-.~L~_"
82
E
)i:
I
i
I
E S
C
62
E
i
i
ATG H
p~ ML:T2-S
,m~=r
"
ll"
I
Fig. 1. Map of MET2 gene and of inducible piasmids constructed.
The black bar representsthe MET20RF (1310 bp). The positionof the two putative TATA box (-222 to -210; -124 to -118) is specified by the symbole. Transcriptionstarts (-59. -52: -40. - 3"t)are indicatedby arrows and the ATG is marked,t3 indicatesthe polyadenylation sequence [11]. The dashed bar represents the U,4SoALH0 intergenicregionof 365 bp and the dotted bar the 250 bp CYC! untranslatedleader [18l. pAMET2-57 and pAMET2-5 were obtained after MET2 leader deletion with exonucleaseIlL B1, BamHl; B2. Bglil; C, Clai; E, EcoRi; H, Hindlll; !)2. Pvull; S, Sstl; Si. Saul; X1. Xbal.
In the first construction, the pBMX1 plasmid, the 3.1 kb P v u l l - B a m H l DNA fragment containing the M E T 2 gene has been cloned in the Xhol site (blunted with Klenow DNA polymerase) and the B a m H l restriction site of the expression vector pBCL26. Since the 5" region extends for 201 nucleotides the M E T 2 gene maintains its own putative TATA box sequence [11]; its transcription is controlled by a hybrid U A S G ^ t f M E T 2 promoter. In the second plasmid, pEMBMet, the 2.65 kb Xbal-Hindlll DNA fragment containing the M E T 2 gene has been cloned in the same restriction sites of the expression vector pEMBLYex4. In this case the 5" region extends for 107 nucleotides and the two putative TATA box sequences of M E T 2 gene were deleted; thus, the expression of M E T 2 is under the control of the inducible hybrid promoter U A S ~ A L / C Y C i . These constructions are shown in Fig. 1. in each of these two constructions the transcription termination and polyadenylation sites are those of the M E T 2 gene [11]. The yeast strain X4004-3A was transformed to the Ura + phenotype either with pBMX1 or pEMBMet plasmids. Some of the indelx~dent transformants obtained with the pBMX1 plasmid and with the pEMBMet plasmid were analyzed by Southern blotting to verify the presence of plasmids in yeast cells (data not
shown). Both transformants with pBMXI and pEMBMet grow well in glucose minimal medium without
50 TABLE !
Relatwe levels of MET2 mRNA and HSTase enzymatic actit'ity in transformants
Strain
Carbon source ~
X4004 $288C X400~pBMXI] X400~pBMX1] X400~pEMBMctt X4004[pEMBMetI X400~pAMET2-57] X400~pAMET2-51
glucose glucose glucose galactose glucose galactose galactose galactose
Specific mRNA levels~' 100 2 > 100 > 100 >100
HSTase specific activity" none 1 0.5 2.2 0.8 3 1.9 2.1
i All strains were tested in their exponential phase of growth (approx. (1-2). l0 s cell/ml). Growth was in minimal medium, supplemented with the required amino acids for the untransformed X4004-3A strain, but without uracil and methionine for all the others. b For mRNA levels,the densitometric analysis of autoradiograms are relative to the value detected it, the wild type strain $288C, defined as i. HSTase specific activity is expressed as nmol of acetylhomoserine produced per min per nag of extracted protein. The data reported in this table are the mean of at least three different experiments. methionine, indicating that the expression of the cloned MET2 gene was not completely repressed by glucose (see also Table l). The inducible expression systems were then tested by Northern blot analysis of total R N A extracted from several X4004-3A transformants during exponential growth in minimal selective medium (YNB) supplemented with glucose 2% (repressing conditions), or galactose 2~ (inducing conditions). As shown in Fig. 2 there is a strong induction of MET2 m R N A in galactose growing cells transformed both with pBMX1 and pEMBMet. The amount of specific m R N A was approximately determined by densitometry of autoradiographic films. Under inducing conditions the MET2
A
B
C
D
E
F
G
H
Fig. 2. Northern blot analysis of MET2 mRNA, Total RNA was extracted from X4004-3A transformants during the exponential phase
of growth ((1-2).106 cell/ml) and from the wild type strain $28gc grown in minimal medium lacking both methionine and uracil. 20/~g were loaded on denaturing gels for each lane. Alter blotting, filters were hybridized with a MET2 riboprobe. A, X4004[pBMX1] in galactose; B, X4004[pEMBMet] in galactose; C. X4004[pBMXIi in glucer,e; D, X4004[pEMBMet] in glucose; E, untransformed wild type 5288C in glucose; F, X4004[I~MET2-57] in galactose; G, X~0411~MET2-$TI in raffinose; H. X4004[p~MET2-$] in galactose.
m R N A amount appears to increase at least 50-fold in comparison with the level measured in repressing conditions and approx. 100-fold in comparison with a untransformed wild type. N o differences were observed between the two plasmids, although the promoter sequences were different. The corresponding enzymatic activity of HSTase was measured with the exchange reaction between acetyI-Lhomoserine and L-[t4C]homoserine [23]. As clearly shown in Table l, the enzymatic activity in transformed cells under induced conditions increased to a very limited extent compared to the level measured in the control, in spite of the large increment of MET2 m R N A .
Deletion of 5" MET2 untranslated region Because the HSTase activity does not parallel a large increment of its correspondent messenger, we searched for an eventual role of the leader region of m R N A in the posttranscriptional control, since 107 bp of leader sequence of the MET2 messenger were still conserved in p E M B M e t plasmid. N o upstream open reading frames (uORF) were identified in this region but some inverted repeats are present ( - 8 0 / - 44 and - 9 / + 61 region of the sequence [11]) that might contribute to the formation of a 'stem and loop' secondary structure [29]. In order to test the functional relevance of these 107 bp, we generated a partial and a complete deletion of this region. Two deletions of the MET2 5" region were selected in which respectively 57 and 5 bases before the A T G codon were preserved, and were cloned in the pEMBLYex4 expression vector. The two recombinant plasmids called p A M E T 2 - 5 7 and p A M E T 2 - 5 were used to transform X4004-3A strain. Yeast transformants bearing these two plasmids were analyzed by Southern and Northern blotting as previously described and a similar increase of M E T 2 m R N A as before was demonstrated in cells grown under inducing conditions (Fig. 2). The HSTase activity assay, in spite of the high MET2 m R N A level, again showed a very limited increase (Table I). Yeast cells transformed with pA M E T 2 - 5 7 or p A M E T 2 - 5 plasmid showed, however, a strong repression of MET2 expression in glucose medium, since they failed to grow in minimal glucose medium without methionine, while they grow well in galactose medium.
Analysis of yeast proteins b~ two dimensional gel electrophoresis The preceeding experiments demonstrated a lack of correspondence between the m R N A levels and HgTase activity, however, they did not demonstrate a translational regulation since an inactive protein could be equally induced to high levels. To investigate whether the posttranscriptionai control of MET2 gene expression is due to a lack of protein synthesis or by modulation of enzyme activity, a two dimensional electrophore-
51 sis analysis was performed. The theoretical isoelectric point ( p l ) of HSTase calculated from amino acid composition is 5.47. This p l value was confirmed by isoelectric focusing of HSTase protein synthesized in vitro. A synthetic MET2 mRNA obtained by in vitro transcription of MET2 gene cloned in pGEMblu vector was used to direct the synthesis of HSTase protein in a reticulocyte lysate. A two dimensional gel ¢lectrophoresis of translation products indicated that the MET2 mRNA is efficiently translated in vitro, showing a single polypeptide of approx. 50 kDa (Fig. 3A) and p l of 5.5 as expected on the basis of the ORF derived by sequencing the gene [11]. X4004-3A[pEMBMet] transformants, grown in selective medium where labeled with [35S]methionine under repressing or inducing conditions. Total protein was extracted and separated on two dimensional gels. The availability of HSTase polypeptide translated in vitro allowed us also to identify a yeast protein comigrating on two dimensional gels with the enzyme synthesized in vitro (Fig. 3B). This spot is located near the '130' spot identified to be a hexokinase [27]. Total extracts from $288C strain were used as untransformed controls (data not shown). Autoradiography of the gels showed the presence of several proteins induced by galactose, but we were not able to detect a strong induction of a protein with a p l = 5.47 and a size of approx. 50 kDa. Irdeed, in transformant
cells grown under inducing conditions we carl detect a slight induction (approx. 2.5-fold) of a spot that comigrates with the HSTase protein in comparison with repressing conditions (Fig.3C,D). This is comparable to the increase of enzymatic activity measured in the same growing conditions (see Fig. 3 and Table I). Discussion The MET2 gene, coding for L-homoserine-Oacetyltransferase of S. cerevisiae, belongs to the so called group I methionine biosynthetic genes. These genes are under a coordinate specific transcriptional repression by methionine and S-adenosylmethionine [5]. A comparative analysis of the 5' region of MET2, MET3, MET25 and SAMI genes [10,13] revealed some common motifs that may be important for the control of gene expression. These cis-acting elements consist in an UASmet, required for maximum expression (5'TCACGTGA), located at - 3 7 4 , and in two regions that may be involved in methionine transcriptional repression (URSmet) ( 5 ' - A A A A A T T G T G T and 5'AACTAAGTC), respectively located at - 3 0 9 and - 297. Preliminary findings of our laboratory suggested, in addition, that posttranscriptional regulation operates for MET2 expression. In fact, we found that in a strain transformed with a multicopy plasmid, bearing the
EF kOa
4.2.Z3 A
IZ3
4.2 t C
69-q) 41P
¢> e
46- 0
30-1)
•
"t-
4 6 - IR *
3 0 - ,,
oe
,r
69-p .
.qb~ll~
.~*
"
o
-
IP
....................
-~-
:.........
Fig. 3. Two dimensional gel clcctrophorcsis of yeast proteim. (A) HSTas¢ p o l ~ p l i d ¢ translated in vitro. (B) Total yeast pmleins (200000 ¢pm) from X4(]O4-3A[pEMBMet], grown in glucose, mixed with in HSTase translated in vitro. (C) Total yeast proteins (800000 cpm) from
X4004-3A[pEMBMetlgrownin glucosemediumand (D) in galactnsemedium.
52 cloned M£F2 gene, the level of the specific MET2 mRNA increases approx. 10-fold, while the enzymatic activity of HSTase increased less than 2-fold, in comparison with untransformed wild type cells [2]. To obtain substantial evidence for a posttranscriptional modulation of MET2 expression, we cloned the MET2 gene in a plasmid vector, under the control of the inducible UASGAL element [181, making then progressive deletions of the 5' flanking region. in this way we obtained an inducible system for MET2 transcription, and in transformed yeast, grown in galactose medium, the steady state level of specific MET2 mRNA is increased at least 100-fold, in comparison with an untransformed wild type cell. Again the specific activity of HSTase increased only 2-fold, indicating that a strong posttranscriptional control may be operative. Furthermore, the absence of an induced polypeptide in yeast cells grown under induced conditions, analyzed by two dimensional gel electrophoresis, suggests translational control. This translational control, however, seems to be operative only in yeast cells, since the MET2 mRNA is easily translated in vitro in an heteroiogous system (i.e., reticuiocyte lysate). In yeast the mRNA untranslated 5' leader sequence may be important for a translational control. In fact, the initiation of translation in yeast follows the 'scanning model' [30] and the presence of A U G or short Open Reading Frame (uORF) upstream the initiation codon inhibits translation and this fact is at the basis of the translational control reported for GCN4 and CPA 1 genes [31,32]. The MET2 mRNA leader sequence does not contain an uORF; however, in yeast the formation of a secondary structure in the mRNA is important for translatability of the message and could provide a translational control element [29,34]. We have found some inverted repeats around the region containing the A U G codon of MET2 mRNA that could generate a secondary structure. However, the complete deletion of the leader region, up to 5 bases upstream the A U G did not increase, in induced cells, the level of HSTase, indicating that the 5' leader region of mRNA is not relevant for this posttranscriptional regulation. Since the untranslated 5' region appears not to be involved in the translational regulation of MET2 gene expression, several alternative regulatory mechanisms could be hypothesized. Sequences in the trailer part of messenger RNA or sequences within the ORF may influence the translatability of this mRNA in yeast, [34]. Internal sequences within the O R F might be recognized by a specific unknown binding protein, or MET2 protein itself could autoregulate its synthesis, as demonstrated for the synthesis of some ribosomal proteins [35]. The observed data could be also explained by a regulated fast degradation of the protein [35]. However, the half-life should be very short, in the range of few minutes, since we were not able to found a strong
increase of MET2 protein on two dimensional gels after a 30 min pulse labelling. Interestingly, evidence of posttranscriptional modulation of gene expression has also been reported for another methionine biosynthetic gene, MET25 [13], but whose regulatory mechanism is not currently known. Acknowledgements This work was partially supported by a MPI 60% grant to L.A., N.F. was supported by an A.I.R.C. (Associazione ltaliana Ricerca sul Cancro) fellowship. References I Jones, E.W. and Fink, G.R. (1982) In The molecular biology of the yeast Saccharomyces cerevisiae (Strathern, J.H., Jones, E.W. and Broach, J.R., eds.L pp. 181-300. Cold Spring Harbor, New York. 2 Baroni, M., Livian, S., Martegani, E. and Alberghina, L. (1986) Gene 46, 71-78. 3 Cherest, H., Surdin-Kerjan, J.. Antoniewski, J. and De Robichon Szulmajster, H. (1973)J. Bacteriol. 114, 928-933. 4 Cherest, H., Surdin-Kerjan, Y.. Antoniewski, J. and De Robichon-Szulmajster, H. (1973) J. Bacteriol. 115. 1084-1090. 5 Thomas, D., Rothstein, R., Rosenberg. N. and Surdin-Kerjan, Y. (1988) Mol. Cell. Biol. 8, 5132-5139. 6 Cherest, H., Surdin-Kerjan, Y. and De Robichon-Szulmajster, H. (1971) J. Bacteriol. 106, 758-772. 7 Breton, A. and Surdin-Kerjan, Y. (1977) J. Bacteriol. 132, 224-232. 8 Cherest, H., Thao, N.N. and Surdin-Kerjan, Y. (1985) Gene 34, 269-281. 9 Sangsoda, S., Cherest, FI. and Surdin-Kerjan, Y. (1985) Mol. Gen, Genet. 200. 407-414. 10 Thomas, D. and Surdin-Kerjan, Y. (1987) J, Biol. Chem. 262, 16704-16709. 11 Langin, T., Faugeron, G., Goyon, C., Nicolas, A. and Rossignol, J.L. (1986) Gene 49, 283-293. 12 Cherest, H., Kerjan, P. and Surdin-Kerjan, Y. (1987) Mol. Gen. Genet. 210, 307-313. 13 Thomas, D., Cherest, H. and Surdin-Kerjan, Y. (1989) Mol. Cell. Biol. 9, 3292-3298, 14 Yamagata, S. (1987) J. Bacteriol. 169, 3458-3463. 15 Vanoni, M., Vai, M., Popolo, L. and Alberghina, L. (1983) J. Bacteriol. 156, 1282-1291. 16 Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecular cloning. A laboratory, manual. Cold Spring Harbor Laboratory, New York. 17 Velati-Bellini, A., Pedroni, P., Martegani, E. and AIberghina, L. (1986) Appl. Microbiol. Biotechnol. 25, 124-131. 18 Guarente, L. (1983) Methods Enzymol. 101,181-191. 19 Baldari, C., Murray, J.A.H., Ghiara, P., Cesareni. (3. and Galeoni, G.L. (1987) EMBO J. 6, 229-234. 20 Cryer, D.R., Eccleshall. R. and Marmur. J. (1975) Methods Cell Biol. 12, 39-44. 21 lto, H., Fukuda, Y., Murata, K. and Kimnra, A. (1983) J. Bacteriol. 153, 163-168. 22 Henikoff, S. (1984) Gene 28, 351-359. 23 Nagai, S. and Kerr, D. (1971) Methods Enzymol. I7B, 442-445. 24 Nasal, S. and Flavin, M. (1971) Methods Enzymol. ITB, 423-424. 25 Popolo, L. and Alberghina, L. (1984) Proc. Natl. Acad. Sci. USA 81, 120-124. 26 O'Farrell, P.H. (1975) J. Biol. Chem. 250, 4007-4021. 27 Ludwig. J.R., Foy, J.J,, Elliot, S.G, and McLaughlin, C,S. (1982) Mol. Cell, Biol. 2, 117-126.
53 28 Krieg. P.A, and Melton. D,A. (1984) Nucleic Acids Res. 12, 7057-7O70. 29 Bairn. S.B. and Sherman, F. (1988) Mol. Cell. Biol. 8, 1591-1601. 30 Kozak, M. (1989) J. Cell. Biol, 108, 229-241. 31 Mueller, P.P. and Hinnesbuch, A.G. (1986) Cell 45. 201-207. 32 Werner, M., Feller, A., Messenguy, F. and Pierrard, A. (1987) Cell 49, 805-813.
33 Surdin-Kerjan, Y. and De Rohichon Szulmajster, H. (1975) J. Bacteriol. 122. 367-374. 34 Van den Heuvel, J.J., Planta, R.J. and Raut$, H.A. (1990) Yeast 6. 473-482. 35 Warner. J.R.. Mitra. O.. Schwindiger, W.F., Studer~y, M. and Fried, H.M. 0985) Mol. Cell. Biol. 5, 1512-1521.
