Current Genetics
Curr Genet (1990) 17:119-123
9 Springer-Verlag 1990
When a glycolytic gene on a yeast 2[aORI-STBplasmid is made essential for growth its expression level is a major determinant of plasmid copy number Peter W. Piper and Brendan P. G. Curran*
Department of Biochemistry,UniversityCollegeLondon, London WC1E6BT, UK Received July 7/October 20, 1989
Summary. This study demonstrates how varying the promoter strength of an essential gene on a yeast 2pORISTB YEp multicopy vector can influence vector copy levels. A phosphoglycerate kinase gene (PGK) on this plasmid was made essential for fermentative growth by transformation into a p g k - yeast strain. When in these PGK + transformants the requirement for P G K expression was the sole selective criterion for plasmid maintenance, P G K promoter activity was inversely related to vector copy levels. Plasmidswith an efficiently-transcribed P G K gene were maintained at approximately ofle copy per cell, whereas those lacking the UAS that normally directs high basal P G K transcription levels were present at up to 10-15 copies. All cultures of these PGK + transformants contained only a low proportion of p g k cells. Since mitotic loss of the plasmid arrests growth through loss of a functional P G K allele, P G K confers high stability to the YEp vector in such a p g k - genetic background. In this system YEp vector levels are probably influenced by P G K transcription because high expression of P G K i s needed in rapid fermentative growth. Remarkably, low plasmid P G K promoter activity caused PGK mRNA levels slightly higher than those found in yeast with normal P G K regulation. A higher plasmid copy number is therefore not the only factor counteracting the effects of low P G K transcription, and it is possible that PGK mRNA becomes more stable in response to inefficient P G K transcription. Key words: Saccharomyces cerevisiae - Episomal plasmid - Copy number control - Plasmid maintenance - Glycolytic enzyme levels
* Current address: School of Biological Sciences, Queen Mary College, London, UK Offprint requests to: P. W. Piper
Introduction
Yeast Episomalplasmid (YEp) vectors that use replication sequences of the endogenous 2~t plasmid of Saccharomyces cerevisiae are in widespread use in yeast transformation studies [see Parent et al. (1985), Murray (1987) and Futcher (1988) for reviews]. They are multicopy plasmids which generally contain either all the sequences of 2~t DNA ("complete 2~t" plasmids) or only the 2g origin of replication (ORI) and the adjacent cis-acting STB (or REP3) region ("2gORI-STB" plasmids). The STB locus is required for efficient partitioning of plasmid DNA between mother and daughter cells in cell division (Murray and Cesareni 1986). Complete 2p plasmids can replicate in both cir + and cir~ strains, and their ability to cause loss of the endogenous 21a DNA of cir + S. cerevisiae can be used to cure strains of this plasmid (Dobson et al. 1980; Erhart and Hollenberg 1983; Harford and Peeters 1987). Complete 2g plasmids are very large, however, especially when, through the incorporation of sequences from bacterial plasmids, they are also yeast-Escherichia coli shuttle plasmids. Vectors containing only the ORI-STB region of 2g DNA are in widespread use as high copy vectors, being appreciably smaller than complete 2~t plasmids. They replicate only in cir + strains due to a requirement for functions supplied in trans by the endogenous 2~t DNA (Futcher 1988). Also they do not usually have both copies of the 21a FRT sequences, the sites of recombinations catalysed by the 2g FLP product (a site-specific invertase). These recombinations interconvert the A and B forms of 2~t DNA, and are crucial to the process whereby 2p DNA is maintained in high copy number (Murray 1987; Futcher 1988). Transformation of yeast with a YEp vector involves selecting for transformants under conditions such that a gene carried on the vector is essential for growth (e.g. medium minus leucine for a leu2- strain transformed with a LEU2-bearing plasmid). If the expression of the gene used for selection is reduced, the requirement for sufficient levels of its essential product may exert strong pressure for higher plasmid copy number. The LEU2d gene of plasmid
120 p J D B 2 1 9 (Beggs 1978) lacks most of the L E U 2 p r o m o t e r sequences. W h e n present o n YEp plasmids, L E U 2 d causes m u c h higher copy n u m b e r s [up to 300 per cell ( E r h a r d a n d H o l l e n b e r g 1983; F u t c h e r a n d Cox 1984)] in leu2- strains o n leucine-deficient m e d i u m as c o m p a r e d to YEp plasraids with the complete L E U 2 gene. This is t h o u g h t to reflect a r e q u i r e m e n t for high L E U 2 d gene dosage in order for sufficient synthesis of the L E U 2 p r o d u c t , isopropylmalate dehydrogenase (Beggs 1978; E r h a r t a n d H o l l e n berg 1983). H o w e v e r high copy n u m b e r m a y be determ i n e d by factors other t h a n the expression level of the selectable f u n c t i o n since a p l a s m i d c o n t a i n i n g L E U 2 d a n d URA3 was f o u n d to have a high copy n u m b e r even when selection was applied only for URA3 (Baldari et al. 1987). To determine how the expression level of a n o t h e r selectable gene influences p l a s m i d copy levels we constructed cells having the sole f u n c t i o n a l copy of one of their glycolytic genes ( P G I0 o n a YEp vector. We observed how variations in the level of t r a n s c r i p t i o n of this gene influenced vector copy n u m b e r d u r i n g fermentative growth, c o n d i t i o n s that necessitate P G K function.
E H
H /H
Materials and methods Strains. Strain BC3 (cir+, leu2-3,112, trpl-1, his3-15, ura3-52, pgk::TRP1) originated as a single spore colony from diploid BC2.2 (Piper et al. 1988). It possesses a TRP1 gene fragment disruption of the PGK locus on chromosome III, a disruption present at one of the two PGK alleles of the BC2.2 parent (Piper et al. 1988). In common with other pgk- strains (Ciriacy and Breitenbach 1979), BC3 grows only on glycerol plus ethanol as a carbon source. Plasmids. The plasmids used have all been described in earlier studies (Ogden et al. 1986; Piper et al. 1988), and possess the general structure shown in Fig. 1a. Transformation of BC3 to a PGK+ phenotype. The procedure used to transform this pgk- strain was essentially that of Hinnen et al. (1987), but with the following modifications: (i) cultures were grown with aeration on YEPGE (1% yeast extract, 2% bactopeptone, 3% glycerol, 2% ethanol) prior to transformation; (ii) after the incubation with DNA and polyethylene glycol according to the Hinnen et al. (1987) protocol, sphaeroplasts were sedimented by centrifugation, resuspended in 0.5 ml YEPGE plus 1.0 M sorbitol and incubated at 25 ~ for 30 min; (iii) 0.2 ml aliquots were then added to 10 ml YEPD regeneration agar (YEPD (1% yeast extract, 2% bactopeptone, 2% glucose) supplemented with 3% agar and 1.0M sorbitol) which was then plated. The efficiency of transformation of BC3 using PGK-bearing YEp vectors was low by this procedure (110 transformants i~g-1 DNA). However all 39 transformants tested for the presence ofplasmid DNA by Southern blotting were found to be maintaining the original plasmid used for the transformation in unaltered form even after prolonged culture on YEPD. Also, several modifications to the above procedure did not result in higher transformation efficiencies. Measurement of the proportion of plasmid-minus cells in cultures of PGK + BC3 transformants. When PGK + BC3 transformants become pgk-, through mitotic loss of their YEp plasmid, their growth on glucose is arrested (pgk- cells cannot maintain ATP pools in fermentative media; Ciriaey and Breitenbach 1979). However such pgk- cells do not rapidly lose viability in glucose media and can, therefore, be detected by plating on YEPGE (>97% of BC3 cells transferred from YEPGE to YEPD are still viable after 2h, as assessed by ability to subsequently form colonies on YEPGE). Even
Fig. 1. a structure of the 2gORI-STB plasmids in the BC3 transformants employed for this study. These plasmids are identical but for PGKpromoter deletions which cause substantial differences in PGK transcription level. For plasmid derivations and measurements of the levels of PGK expression caused by each deletion see Ogden et al. (1986) and Piper et al. (1988). Thin line = pBR322 sequences; solid line = double EcoR 1 fragment of yeast DNA containing 2gORI-STB and LEU2& open box = 2.95 kb HindlII fragment of S. cerevisiae DNA containing the PGK gene (coding region hatched) and a promoter deletion at the BamH 1 site (B). E = EcoR 1, H = HindIII. b determination ofplasmid copy number in BC3 transformants during growth on YEPD. Southern blots of EcoRl restriction digests of total cell DNA were probed with PGK sequences, the relative labelling of the 2.4 kb plasmid (P) and 4.25 kb chromosomal (C) DNA bands being a measure of plasmid copy number (see Piper et al. (1988) for experimental details), c Comparison of PGK mRNA levels in BC3 transformants and in a yeast strain with no chromosomal PGK disruption (MD40-4c; Ogden et al. 1986) during growth on YEPD. Samples of total cell RNA were quantitated by dot-blot hybridisation to ribosomal RNA gene sequences. Equal amounts were then gel fractionated, Northern blotted and the blot probed with PGK sequences as in Piper et al. (1988). The positions of PGK mRNA and the much less abundant transcript of the disrupted chromosomal PGK locus (arrowed; BC3 transformants only) are indicated. Unexpectedly those transformants with the weakest PGK promoter activity (Table 1) displayed the highest levels of PGK mRNA
in YEPGE, a medium which can supportpgk- cell growth, a plasmid PGK gene confers a major growth advantage with the result that plasmid loss is accompanied by at least a 3 to 4-fold increase in cell doubling time. pgk- colonies are therefore readily distinguishable from PGK + colonies by virtue of their much smaller size on YEPGE plates. In this investigation the ratios of pgk and PGK+ cells in YEPD and YEPGE cultures of BC3 transformants were determined by plating on YEPGE. PGK + cells produce large colonies which also grow on complete minimal glucose (CMD; Piper et al. 1988)medium
121
Table 1. Properties of BC3 transformants during rapid fermentative growth Transformant
3\61
3k63
3k67
3k73
3\67.73
3\77
Strain MD40-4c
YEp vector present a
pMA761 + (i) 5.3 (ii) 7.1 0.26
pMA763 + (i) 5.3 (ii) 5.5 0.30
pMA767 +\(i) 10.6 (ii) 11.8 0.30
pMA773 + + or + (i) 1.6 (ii) 2.0 0.30
p767\773 +\(i) 13.1 (ii) 14.7 0.28
pMA777 ++ (i) 0.7 (ii) 0.9 0.24
none ++ n.d.
2.69
2.74
2.69
1.75
2.63
1.22
1.01
PGK promoter activitya Plasmid copy number b gmaxat 25~ PGK activity in crude cell extract ~
a For details see Ogden et al. (1986) and Piper et al. (1988). ++: wild-type levels u Copy number was from scintillation counting of bands on Southern blots such as the blot shown in Fig. lb (see Piper et al. 1988). After the first set of copy number measurements (i) another set of measurements was made after 50 generations further growth on YEPD (ii). Taking into consideration the background count of the regions of the filter containing DNA, the maximum error for these copy number measurements was estimated as +10% for the highest and +30% for the lowest values c Expressed as I~mols NADH utilised (25~ per mg total soluble protein in a standard assay for PGK activity (Bergmeyer 1974)
with histidine and uracil supplements. Cells which have lost the YEp plasmid form tiny colonies on YEPGE, do not grow on glucose media and need leucine in addition to histidine and uracil supplements when grown on CM +3% glycerol +2% ethanol (CMGE).
Measurement of plasmid copy number and PGK mRNA levels. These measurements were on cultures in exponential growth on YEPD (25 X 10 7 cells m1-1) as in Piper et al. 1988. The promoter activity characteristic of the various deletion constructs and of wild type are as defined by Ogden et al. (1988) and by Piper et al. (1988), namely the same activities measured in transformants with high copynumbers of plasmid in the absence of selection for a functional PGK gene.
Measurement of PGK enzyme activity in crude cell extracts. Pelleted cells were broken by glass bead vortexing in three times their volume of 100 mM triethanolamine-HC1 pH 7.6. Total soluble protein was prepared by centrifugation at 30,000g for 30rain, its protein concentration measured by Bio-Rad protein assay, and its PGK activity assayed spectrophotometrically (Bergmeyer 1974).
Results and discussion
Transforming a pgk yeast with a plasmid containing the PGK gene, applying selection for the ability to support growth on fermentative media In conjunction with studies on transcription elements within the yeast PGK p r o m o t e r we constructed strains that have a PGK locus disrupted by the insertion of a TRP1 gene f r a g m e n t (Piper et al. 1988). Strains that are pgk- exhibit slow g r o w t h on only glycerol plus ethanol as c a r b o n source (Ciriacy and Breitenbach 1979) but can be t r a n s f o r m e d to a P G K + p h e n o t y p e using plasmids containing a functional PGK gene, and applying selection for the ability to g r o w on glucose (see Methods). The transformants obtained this way have an absolute requirement for the plasmid PGK gene in fermentative growth; also PGK confers more rapid gluconeogenic growth on glycerol plus ethanol, F o r this study we t r a n s f o r m e d the cir+ pgk- strain BC3 with PGK-bearing yeast-E.-coli shuttle vectors
which also contain pBR322 sequences, the 2gORI-STB region and LEU2d, These plasmids, whose general structure is shown in Fig. 1 a, are identical but for the deletion present within PGK p r o m o t e r sequences (Ogden et al. 1986; Piper et al. 1988), deletions which cause widelydisparate levels of PGK transcription (Table 1). They lack F R T sites, the p r o b a b l e reason that no r e c o m b i n a t i o n was detected between these YEp vectors and the endogenous 2g D N A of BC3 in Southern blotting experiments, such as that in Fig. 1 b.
A PGK-bearing YEp vector can ensureplasmid maintenance on complex media, in a pgk- genetic background even cultures with low vector levels accumulating only small proportions of plasmid-deficient cells Y E P D cultures of P G K + BC3 transformants were plated on Y E P G E so as to determine the p r o p o r t i o n o f the cells that werepgk- due to mitotic loss o f the YEp plasmid (see Methods). These Y E P D cultures were maintaining their YEp vector at relatively low copy levels (Table 1) and would therefore be expected to show appreciable segregation of plasmid-minus cells. H o w e v e r they contained only a very small percentage of pgk- cells, invariably no more than 2% of the total viable cell count. This small pgk- p r o p o r t i o n must be due substantially to the g r o w t h arrest that accompanies plasmid loss. Even on Y E P G E the YEp plasmid confers P G K + cells with a m a r k e d growth advantage over pgk- ceils (see Methods). We f o u n d the p r o p o r t i o n ofpgk- cells in Y E P G E cultures of these transformants to remain stable at between 2 - 7 % . The use of a PGK-bearing plasmid in a pgk- strain therefore provides a self-regulating system ensuring maintenance o f the plasmid, plasmid-minus segregants arresting growth on fermentative media and growing at very m u c h reduced rate on glycerol plus ethanol. It is improbable that the results that follow were appreciably influenced by the presence of the small p r o p o r t i o n ofpgk- cells in the Y E P D cultures analysed.
122 When PGK is the sole plasmid function needed in fermentative growth its expression level influences vector copy levels On CMD medium supplemented with histidine and uracil, but lacking leucine, growth of the transformants in Table 1 requires two genes of the plasmid (LEU2d and PGK). Under these conditions the YEp plasmid was present at very high copy number, as is normal when selection is applied for LEU2d; also PGK was overexpressed relative to normal haploid cells with a single wild type PGK gene on chromosome III except in those BC3 transformants with the weakest PGK promoters (data not shown). This was consistent with LEU2d, not PGK expression, determining plasmid copy number on leucine-deficient medium. On rich glucose medium (YEPD), LEU2d is not required and the sole selective factor for plasmid maintenance is the need for PGK in fermentative growth. Under these conditions YEp vector sequences were found at relatively low copy numbers, yet not integrated into chromosomal sequences (Fig. l b). Also these copy numbers remained stable during more than 50 generations growth on YEPD (Table 1). Any effect of the level of expression of the plasmid PGK gene on growth rate was relatively small (Table 1), and the endogenous 2bt DNA of the cells was maintained at high copy levels (data not shown). The PGK promoter deletions of the transformants in Table 1 give rise to levels of PGK transcription varying from as little as 0.5-1% of that of the wild-type PGK gene, to the level displayed by wild-type PGK. These measurements of transcription level ("promoter activity" in Table 1) are taken from PGK mRNA levels when the same plasmids are present at higher copy in other yeast transformants (Ogden et al. 1986; Piper et al. 1988). Despite considerable variation in their PGK promoter activities the transformants were all able to maintain high PGK mRNA (Fig. 1c) and PGK enzyme (Table 1) levels during rapid fermentative growth. These levels did not directly reflecting the plasmid PGK promoter activity. Fig. 1b and Table 1 show that this was achieved, at least in part, by increasing plasmid copy number in response to weak PGK expression. Plasmid pMA777, whose PGK gene is transcribed at approximately the same levels as wild-type PGK, is maintained at about one copy per cell. Under identical culture conditions, plasmids lacking the promoter element that normally directs efficient PGK transcription (UASpGK; Ogden et al. 1986) are maintained at higher copy levels, the weakest PGK promoters (pMA767; p767\773) causing copy numbers in the region of 10-15. Unexpectedly, low PGK promoter activity in BC3 transformants is associated with PGK mRNA levels 2-3fold higher than normal To support rapid fermentative growth S. cerevisiae must maintain high glycolytic flux rates and glycolytic enzyme levels. However phosphoglycerate kinase levels in yeast
are normally sufficiently high as not to be a factor limiting glycolytic flux (Brindle 1988). When the only selection for YEp vector maintenance was the requirement for PGK expression, both PGK mRNA (Fig. lc) and PGK enzyme levels (Table 1) remained high irrespective of PGK promoter activity. Thus levels of PGK enzyme are sustained primarily through controls that ensure that the levels of its mRNA are kept high. In contrast to what was expected, these mRNA levels were > 2 - 3 times higher in those BC3 transformants with low activity PGK promoter plasmids as compared to either BC3 transformants with high activity PGK promoter plasmids or a strain (MD404c) with a single wild-type chromosomal PGK locus (Fig. 1 c and Table 1). Indeed, mRNA levels were approximately proportional to the copy-number of the PGK gene present in these growth conditions. If the lower mRNA levels characteristic of wild-type and high-activity promoters (Fig. 1 c) are sufficient for active growth on fermentable substrates, it is difficult to understand either what the selection pressures are that generate a higher copynumber of the PGK gene in the low-activity constructs or, on the other hand, what are the controls that apparently raise the specific activity of these promoters to that of the wild-type. One obvious possibility is that low PGK transcription may cause a marked increase in the stability of PGK mRNA. The unexpectedly high levels of PGK mRNA in transformants 3\67.73, 3\67 and 3\61 (Fig. 1c) would then be the result of weak PGK transcription causing a dramatic increase in the stability of the generated transcript. In strain MD40-4c, PGK mRNA has a halflife of about 30rain during conditions of rapid fermentative growth (Piper et al. 1986), yet this mRNA may be more stable under certain circumstances since other yeast mRNAs have halflives considerably in excess of this value (Santiago et al. 1986). We have not investigated if low levels of PGK transcription act to considerably increase PGK mRNA halflife, but suggest that this, and increased plasmid copy levels, may act together in ensuring that the transformants in Table 1 maintain PGK mRNA levels. The levels of PGK mRNA are therefore probably the subject of complex regulation, as also are the transcripts of at least one other yeast glycolytic gene (PYK1; Moore et al. 1989). In the latter case translation of PYK1 mRNA is inhibited as its abundance increases. Also the abundance ofPYK1 mRNA per gene copy decreases as copy number increases, an effect that also indicates controls acting either at transcription or at PYK mRNA turnover (Moore et al. 1989). Acknowledgements. This work was supported by the U.K. Science and Engineering Research Council. We thank Kevin Seward for technical assistanceand RajeeshPatel for the growth rate determinations.
References
Baldari C, Murray JAH, Ghiara P, Cesareni G, Galeotti CL (1987) EMBO J 6:229-234 Beggs JD (1978) Nature 275:104-109 Bergmeyer HU (ed) (1974) Methods of enzymatic analysis, vol 1. Academic Press, New York, pp 466-502
123 Brindle KM (1988) Biochemistry 27:6187-6196 Ciriacy M, Breitenbach I (1979) J Bacteriol 139:152-160 Dobson MJ, Futcher AB, Cox BS (1980) Curr Genet 2:201-205 Erhart E, Hollenberg CP (1983) J Bacteriol 156:625-635 Futcher AB (1988) Yeast 4:27-40 Futcher AB, Cox BS (1984) J Bacteriol 157:283-290 Harford MN, Peeters M (1987) Curr Genet 11:315-319 Hinnen A, Hicks JB, Fink GR (1978) Proc Natl Acad Sci USA 75:1929-1933 Moore PA, Bettany AJE, Brown AJP (1990) (submitted for publication) Murray JAH (1987) Mol Microbiol 1 : 1-4 Murray JAH, Cesareni G (1986) EMBO J 5:3391-3399
Ogden JE, Stanway C, Kim S, Mellor J, Kingsman A J, Kingsman SM (1986) Mol Cell Biol 6:4335-4343 Parent SA, Fenimore CM, Bostian KA (1985) Yeast 1:83-138 Piper PW, Curran B, Davies MW, Hirst K, Lockheart A, Ogden JE, Stanway CA, Kingsman A J, Kingsman SM (1988) Nucleic Acids Res 16:1333-1348 Piper PW, Curran B, Davies MW, Lockheart A, Reid G (1986) Eur J Biochem 161:525-531 Santiago TC, Purvis IJ, Bettany AJE, Brown AJP (1986) Nucleic Acids Res 14: 8347-8359 C o m m u n i c a t e d b y B. S. C o x
Curr Genet (1990) 17:119-123
9 Springer-Verlag 1990
When a glycolytic gene on a yeast 2[aORI-STBplasmid is made essential for growth its expression level is a major determinant of plasmid copy number Peter W. Piper and Brendan P. G. Curran*
Department of Biochemistry,UniversityCollegeLondon, London WC1E6BT, UK Received July 7/October 20, 1989
Summary. This study demonstrates how varying the promoter strength of an essential gene on a yeast 2pORISTB YEp multicopy vector can influence vector copy levels. A phosphoglycerate kinase gene (PGK) on this plasmid was made essential for fermentative growth by transformation into a p g k - yeast strain. When in these PGK + transformants the requirement for P G K expression was the sole selective criterion for plasmid maintenance, P G K promoter activity was inversely related to vector copy levels. Plasmidswith an efficiently-transcribed P G K gene were maintained at approximately ofle copy per cell, whereas those lacking the UAS that normally directs high basal P G K transcription levels were present at up to 10-15 copies. All cultures of these PGK + transformants contained only a low proportion of p g k cells. Since mitotic loss of the plasmid arrests growth through loss of a functional P G K allele, P G K confers high stability to the YEp vector in such a p g k - genetic background. In this system YEp vector levels are probably influenced by P G K transcription because high expression of P G K i s needed in rapid fermentative growth. Remarkably, low plasmid P G K promoter activity caused PGK mRNA levels slightly higher than those found in yeast with normal P G K regulation. A higher plasmid copy number is therefore not the only factor counteracting the effects of low P G K transcription, and it is possible that PGK mRNA becomes more stable in response to inefficient P G K transcription. Key words: Saccharomyces cerevisiae - Episomal plasmid - Copy number control - Plasmid maintenance - Glycolytic enzyme levels
* Current address: School of Biological Sciences, Queen Mary College, London, UK Offprint requests to: P. W. Piper
Introduction
Yeast Episomalplasmid (YEp) vectors that use replication sequences of the endogenous 2~t plasmid of Saccharomyces cerevisiae are in widespread use in yeast transformation studies [see Parent et al. (1985), Murray (1987) and Futcher (1988) for reviews]. They are multicopy plasmids which generally contain either all the sequences of 2~t DNA ("complete 2~t" plasmids) or only the 2g origin of replication (ORI) and the adjacent cis-acting STB (or REP3) region ("2gORI-STB" plasmids). The STB locus is required for efficient partitioning of plasmid DNA between mother and daughter cells in cell division (Murray and Cesareni 1986). Complete 2p plasmids can replicate in both cir + and cir~ strains, and their ability to cause loss of the endogenous 21a DNA of cir + S. cerevisiae can be used to cure strains of this plasmid (Dobson et al. 1980; Erhart and Hollenberg 1983; Harford and Peeters 1987). Complete 2g plasmids are very large, however, especially when, through the incorporation of sequences from bacterial plasmids, they are also yeast-Escherichia coli shuttle plasmids. Vectors containing only the ORI-STB region of 2g DNA are in widespread use as high copy vectors, being appreciably smaller than complete 2~t plasmids. They replicate only in cir + strains due to a requirement for functions supplied in trans by the endogenous 2~t DNA (Futcher 1988). Also they do not usually have both copies of the 21a FRT sequences, the sites of recombinations catalysed by the 2g FLP product (a site-specific invertase). These recombinations interconvert the A and B forms of 2~t DNA, and are crucial to the process whereby 2p DNA is maintained in high copy number (Murray 1987; Futcher 1988). Transformation of yeast with a YEp vector involves selecting for transformants under conditions such that a gene carried on the vector is essential for growth (e.g. medium minus leucine for a leu2- strain transformed with a LEU2-bearing plasmid). If the expression of the gene used for selection is reduced, the requirement for sufficient levels of its essential product may exert strong pressure for higher plasmid copy number. The LEU2d gene of plasmid
120 p J D B 2 1 9 (Beggs 1978) lacks most of the L E U 2 p r o m o t e r sequences. W h e n present o n YEp plasmids, L E U 2 d causes m u c h higher copy n u m b e r s [up to 300 per cell ( E r h a r d a n d H o l l e n b e r g 1983; F u t c h e r a n d Cox 1984)] in leu2- strains o n leucine-deficient m e d i u m as c o m p a r e d to YEp plasraids with the complete L E U 2 gene. This is t h o u g h t to reflect a r e q u i r e m e n t for high L E U 2 d gene dosage in order for sufficient synthesis of the L E U 2 p r o d u c t , isopropylmalate dehydrogenase (Beggs 1978; E r h a r t a n d H o l l e n berg 1983). H o w e v e r high copy n u m b e r m a y be determ i n e d by factors other t h a n the expression level of the selectable f u n c t i o n since a p l a s m i d c o n t a i n i n g L E U 2 d a n d URA3 was f o u n d to have a high copy n u m b e r even when selection was applied only for URA3 (Baldari et al. 1987). To determine how the expression level of a n o t h e r selectable gene influences p l a s m i d copy levels we constructed cells having the sole f u n c t i o n a l copy of one of their glycolytic genes ( P G I0 o n a YEp vector. We observed how variations in the level of t r a n s c r i p t i o n of this gene influenced vector copy n u m b e r d u r i n g fermentative growth, c o n d i t i o n s that necessitate P G K function.
E H
H /H
Materials and methods Strains. Strain BC3 (cir+, leu2-3,112, trpl-1, his3-15, ura3-52, pgk::TRP1) originated as a single spore colony from diploid BC2.2 (Piper et al. 1988). It possesses a TRP1 gene fragment disruption of the PGK locus on chromosome III, a disruption present at one of the two PGK alleles of the BC2.2 parent (Piper et al. 1988). In common with other pgk- strains (Ciriacy and Breitenbach 1979), BC3 grows only on glycerol plus ethanol as a carbon source. Plasmids. The plasmids used have all been described in earlier studies (Ogden et al. 1986; Piper et al. 1988), and possess the general structure shown in Fig. 1a. Transformation of BC3 to a PGK+ phenotype. The procedure used to transform this pgk- strain was essentially that of Hinnen et al. (1987), but with the following modifications: (i) cultures were grown with aeration on YEPGE (1% yeast extract, 2% bactopeptone, 3% glycerol, 2% ethanol) prior to transformation; (ii) after the incubation with DNA and polyethylene glycol according to the Hinnen et al. (1987) protocol, sphaeroplasts were sedimented by centrifugation, resuspended in 0.5 ml YEPGE plus 1.0 M sorbitol and incubated at 25 ~ for 30 min; (iii) 0.2 ml aliquots were then added to 10 ml YEPD regeneration agar (YEPD (1% yeast extract, 2% bactopeptone, 2% glucose) supplemented with 3% agar and 1.0M sorbitol) which was then plated. The efficiency of transformation of BC3 using PGK-bearing YEp vectors was low by this procedure (110 transformants i~g-1 DNA). However all 39 transformants tested for the presence ofplasmid DNA by Southern blotting were found to be maintaining the original plasmid used for the transformation in unaltered form even after prolonged culture on YEPD. Also, several modifications to the above procedure did not result in higher transformation efficiencies. Measurement of the proportion of plasmid-minus cells in cultures of PGK + BC3 transformants. When PGK + BC3 transformants become pgk-, through mitotic loss of their YEp plasmid, their growth on glucose is arrested (pgk- cells cannot maintain ATP pools in fermentative media; Ciriaey and Breitenbach 1979). However such pgk- cells do not rapidly lose viability in glucose media and can, therefore, be detected by plating on YEPGE (>97% of BC3 cells transferred from YEPGE to YEPD are still viable after 2h, as assessed by ability to subsequently form colonies on YEPGE). Even
Fig. 1. a structure of the 2gORI-STB plasmids in the BC3 transformants employed for this study. These plasmids are identical but for PGKpromoter deletions which cause substantial differences in PGK transcription level. For plasmid derivations and measurements of the levels of PGK expression caused by each deletion see Ogden et al. (1986) and Piper et al. (1988). Thin line = pBR322 sequences; solid line = double EcoR 1 fragment of yeast DNA containing 2gORI-STB and LEU2& open box = 2.95 kb HindlII fragment of S. cerevisiae DNA containing the PGK gene (coding region hatched) and a promoter deletion at the BamH 1 site (B). E = EcoR 1, H = HindIII. b determination ofplasmid copy number in BC3 transformants during growth on YEPD. Southern blots of EcoRl restriction digests of total cell DNA were probed with PGK sequences, the relative labelling of the 2.4 kb plasmid (P) and 4.25 kb chromosomal (C) DNA bands being a measure of plasmid copy number (see Piper et al. (1988) for experimental details), c Comparison of PGK mRNA levels in BC3 transformants and in a yeast strain with no chromosomal PGK disruption (MD40-4c; Ogden et al. 1986) during growth on YEPD. Samples of total cell RNA were quantitated by dot-blot hybridisation to ribosomal RNA gene sequences. Equal amounts were then gel fractionated, Northern blotted and the blot probed with PGK sequences as in Piper et al. (1988). The positions of PGK mRNA and the much less abundant transcript of the disrupted chromosomal PGK locus (arrowed; BC3 transformants only) are indicated. Unexpectedly those transformants with the weakest PGK promoter activity (Table 1) displayed the highest levels of PGK mRNA
in YEPGE, a medium which can supportpgk- cell growth, a plasmid PGK gene confers a major growth advantage with the result that plasmid loss is accompanied by at least a 3 to 4-fold increase in cell doubling time. pgk- colonies are therefore readily distinguishable from PGK + colonies by virtue of their much smaller size on YEPGE plates. In this investigation the ratios of pgk and PGK+ cells in YEPD and YEPGE cultures of BC3 transformants were determined by plating on YEPGE. PGK + cells produce large colonies which also grow on complete minimal glucose (CMD; Piper et al. 1988)medium
121
Table 1. Properties of BC3 transformants during rapid fermentative growth Transformant
3\61
3k63
3k67
3k73
3\67.73
3\77
Strain MD40-4c
YEp vector present a
pMA761 + (i) 5.3 (ii) 7.1 0.26
pMA763 + (i) 5.3 (ii) 5.5 0.30
pMA767 +\(i) 10.6 (ii) 11.8 0.30
pMA773 + + or + (i) 1.6 (ii) 2.0 0.30
p767\773 +\(i) 13.1 (ii) 14.7 0.28
pMA777 ++ (i) 0.7 (ii) 0.9 0.24
none ++ n.d.
2.69
2.74
2.69
1.75
2.63
1.22
1.01
PGK promoter activitya Plasmid copy number b gmaxat 25~ PGK activity in crude cell extract ~
a For details see Ogden et al. (1986) and Piper et al. (1988). ++: wild-type levels u Copy number was from scintillation counting of bands on Southern blots such as the blot shown in Fig. lb (see Piper et al. 1988). After the first set of copy number measurements (i) another set of measurements was made after 50 generations further growth on YEPD (ii). Taking into consideration the background count of the regions of the filter containing DNA, the maximum error for these copy number measurements was estimated as +10% for the highest and +30% for the lowest values c Expressed as I~mols NADH utilised (25~ per mg total soluble protein in a standard assay for PGK activity (Bergmeyer 1974)
with histidine and uracil supplements. Cells which have lost the YEp plasmid form tiny colonies on YEPGE, do not grow on glucose media and need leucine in addition to histidine and uracil supplements when grown on CM +3% glycerol +2% ethanol (CMGE).
Measurement of plasmid copy number and PGK mRNA levels. These measurements were on cultures in exponential growth on YEPD (25 X 10 7 cells m1-1) as in Piper et al. 1988. The promoter activity characteristic of the various deletion constructs and of wild type are as defined by Ogden et al. (1988) and by Piper et al. (1988), namely the same activities measured in transformants with high copynumbers of plasmid in the absence of selection for a functional PGK gene.
Measurement of PGK enzyme activity in crude cell extracts. Pelleted cells were broken by glass bead vortexing in three times their volume of 100 mM triethanolamine-HC1 pH 7.6. Total soluble protein was prepared by centrifugation at 30,000g for 30rain, its protein concentration measured by Bio-Rad protein assay, and its PGK activity assayed spectrophotometrically (Bergmeyer 1974).
Results and discussion
Transforming a pgk yeast with a plasmid containing the PGK gene, applying selection for the ability to support growth on fermentative media In conjunction with studies on transcription elements within the yeast PGK p r o m o t e r we constructed strains that have a PGK locus disrupted by the insertion of a TRP1 gene f r a g m e n t (Piper et al. 1988). Strains that are pgk- exhibit slow g r o w t h on only glycerol plus ethanol as c a r b o n source (Ciriacy and Breitenbach 1979) but can be t r a n s f o r m e d to a P G K + p h e n o t y p e using plasmids containing a functional PGK gene, and applying selection for the ability to g r o w on glucose (see Methods). The transformants obtained this way have an absolute requirement for the plasmid PGK gene in fermentative growth; also PGK confers more rapid gluconeogenic growth on glycerol plus ethanol, F o r this study we t r a n s f o r m e d the cir+ pgk- strain BC3 with PGK-bearing yeast-E.-coli shuttle vectors
which also contain pBR322 sequences, the 2gORI-STB region and LEU2d, These plasmids, whose general structure is shown in Fig. 1 a, are identical but for the deletion present within PGK p r o m o t e r sequences (Ogden et al. 1986; Piper et al. 1988), deletions which cause widelydisparate levels of PGK transcription (Table 1). They lack F R T sites, the p r o b a b l e reason that no r e c o m b i n a t i o n was detected between these YEp vectors and the endogenous 2g D N A of BC3 in Southern blotting experiments, such as that in Fig. 1 b.
A PGK-bearing YEp vector can ensureplasmid maintenance on complex media, in a pgk- genetic background even cultures with low vector levels accumulating only small proportions of plasmid-deficient cells Y E P D cultures of P G K + BC3 transformants were plated on Y E P G E so as to determine the p r o p o r t i o n o f the cells that werepgk- due to mitotic loss o f the YEp plasmid (see Methods). These Y E P D cultures were maintaining their YEp vector at relatively low copy levels (Table 1) and would therefore be expected to show appreciable segregation of plasmid-minus cells. H o w e v e r they contained only a very small percentage of pgk- cells, invariably no more than 2% of the total viable cell count. This small pgk- p r o p o r t i o n must be due substantially to the g r o w t h arrest that accompanies plasmid loss. Even on Y E P G E the YEp plasmid confers P G K + cells with a m a r k e d growth advantage over pgk- ceils (see Methods). We f o u n d the p r o p o r t i o n ofpgk- cells in Y E P G E cultures of these transformants to remain stable at between 2 - 7 % . The use of a PGK-bearing plasmid in a pgk- strain therefore provides a self-regulating system ensuring maintenance o f the plasmid, plasmid-minus segregants arresting growth on fermentative media and growing at very m u c h reduced rate on glycerol plus ethanol. It is improbable that the results that follow were appreciably influenced by the presence of the small p r o p o r t i o n ofpgk- cells in the Y E P D cultures analysed.
122 When PGK is the sole plasmid function needed in fermentative growth its expression level influences vector copy levels On CMD medium supplemented with histidine and uracil, but lacking leucine, growth of the transformants in Table 1 requires two genes of the plasmid (LEU2d and PGK). Under these conditions the YEp plasmid was present at very high copy number, as is normal when selection is applied for LEU2d; also PGK was overexpressed relative to normal haploid cells with a single wild type PGK gene on chromosome III except in those BC3 transformants with the weakest PGK promoters (data not shown). This was consistent with LEU2d, not PGK expression, determining plasmid copy number on leucine-deficient medium. On rich glucose medium (YEPD), LEU2d is not required and the sole selective factor for plasmid maintenance is the need for PGK in fermentative growth. Under these conditions YEp vector sequences were found at relatively low copy numbers, yet not integrated into chromosomal sequences (Fig. l b). Also these copy numbers remained stable during more than 50 generations growth on YEPD (Table 1). Any effect of the level of expression of the plasmid PGK gene on growth rate was relatively small (Table 1), and the endogenous 2bt DNA of the cells was maintained at high copy levels (data not shown). The PGK promoter deletions of the transformants in Table 1 give rise to levels of PGK transcription varying from as little as 0.5-1% of that of the wild-type PGK gene, to the level displayed by wild-type PGK. These measurements of transcription level ("promoter activity" in Table 1) are taken from PGK mRNA levels when the same plasmids are present at higher copy in other yeast transformants (Ogden et al. 1986; Piper et al. 1988). Despite considerable variation in their PGK promoter activities the transformants were all able to maintain high PGK mRNA (Fig. 1c) and PGK enzyme (Table 1) levels during rapid fermentative growth. These levels did not directly reflecting the plasmid PGK promoter activity. Fig. 1b and Table 1 show that this was achieved, at least in part, by increasing plasmid copy number in response to weak PGK expression. Plasmid pMA777, whose PGK gene is transcribed at approximately the same levels as wild-type PGK, is maintained at about one copy per cell. Under identical culture conditions, plasmids lacking the promoter element that normally directs efficient PGK transcription (UASpGK; Ogden et al. 1986) are maintained at higher copy levels, the weakest PGK promoters (pMA767; p767\773) causing copy numbers in the region of 10-15. Unexpectedly, low PGK promoter activity in BC3 transformants is associated with PGK mRNA levels 2-3fold higher than normal To support rapid fermentative growth S. cerevisiae must maintain high glycolytic flux rates and glycolytic enzyme levels. However phosphoglycerate kinase levels in yeast
are normally sufficiently high as not to be a factor limiting glycolytic flux (Brindle 1988). When the only selection for YEp vector maintenance was the requirement for PGK expression, both PGK mRNA (Fig. lc) and PGK enzyme levels (Table 1) remained high irrespective of PGK promoter activity. Thus levels of PGK enzyme are sustained primarily through controls that ensure that the levels of its mRNA are kept high. In contrast to what was expected, these mRNA levels were > 2 - 3 times higher in those BC3 transformants with low activity PGK promoter plasmids as compared to either BC3 transformants with high activity PGK promoter plasmids or a strain (MD404c) with a single wild-type chromosomal PGK locus (Fig. 1 c and Table 1). Indeed, mRNA levels were approximately proportional to the copy-number of the PGK gene present in these growth conditions. If the lower mRNA levels characteristic of wild-type and high-activity promoters (Fig. 1 c) are sufficient for active growth on fermentable substrates, it is difficult to understand either what the selection pressures are that generate a higher copynumber of the PGK gene in the low-activity constructs or, on the other hand, what are the controls that apparently raise the specific activity of these promoters to that of the wild-type. One obvious possibility is that low PGK transcription may cause a marked increase in the stability of PGK mRNA. The unexpectedly high levels of PGK mRNA in transformants 3\67.73, 3\67 and 3\61 (Fig. 1c) would then be the result of weak PGK transcription causing a dramatic increase in the stability of the generated transcript. In strain MD40-4c, PGK mRNA has a halflife of about 30rain during conditions of rapid fermentative growth (Piper et al. 1986), yet this mRNA may be more stable under certain circumstances since other yeast mRNAs have halflives considerably in excess of this value (Santiago et al. 1986). We have not investigated if low levels of PGK transcription act to considerably increase PGK mRNA halflife, but suggest that this, and increased plasmid copy levels, may act together in ensuring that the transformants in Table 1 maintain PGK mRNA levels. The levels of PGK mRNA are therefore probably the subject of complex regulation, as also are the transcripts of at least one other yeast glycolytic gene (PYK1; Moore et al. 1989). In the latter case translation of PYK1 mRNA is inhibited as its abundance increases. Also the abundance ofPYK1 mRNA per gene copy decreases as copy number increases, an effect that also indicates controls acting either at transcription or at PYK mRNA turnover (Moore et al. 1989). Acknowledgements. This work was supported by the U.K. Science and Engineering Research Council. We thank Kevin Seward for technical assistanceand RajeeshPatel for the growth rate determinations.
References
Baldari C, Murray JAH, Ghiara P, Cesareni G, Galeotti CL (1987) EMBO J 6:229-234 Beggs JD (1978) Nature 275:104-109 Bergmeyer HU (ed) (1974) Methods of enzymatic analysis, vol 1. Academic Press, New York, pp 466-502
123 Brindle KM (1988) Biochemistry 27:6187-6196 Ciriacy M, Breitenbach I (1979) J Bacteriol 139:152-160 Dobson MJ, Futcher AB, Cox BS (1980) Curr Genet 2:201-205 Erhart E, Hollenberg CP (1983) J Bacteriol 156:625-635 Futcher AB (1988) Yeast 4:27-40 Futcher AB, Cox BS (1984) J Bacteriol 157:283-290 Harford MN, Peeters M (1987) Curr Genet 11:315-319 Hinnen A, Hicks JB, Fink GR (1978) Proc Natl Acad Sci USA 75:1929-1933 Moore PA, Bettany AJE, Brown AJP (1990) (submitted for publication) Murray JAH (1987) Mol Microbiol 1 : 1-4 Murray JAH, Cesareni G (1986) EMBO J 5:3391-3399
Ogden JE, Stanway C, Kim S, Mellor J, Kingsman A J, Kingsman SM (1986) Mol Cell Biol 6:4335-4343 Parent SA, Fenimore CM, Bostian KA (1985) Yeast 1:83-138 Piper PW, Curran B, Davies MW, Hirst K, Lockheart A, Ogden JE, Stanway CA, Kingsman A J, Kingsman SM (1988) Nucleic Acids Res 16:1333-1348 Piper PW, Curran B, Davies MW, Lockheart A, Reid G (1986) Eur J Biochem 161:525-531 Santiago TC, Purvis IJ, Bettany AJE, Brown AJP (1986) Nucleic Acids Res 14: 8347-8359 C o m m u n i c a t e d b y B. S. C o x
