- Research article
- Open Access
Transcriptome analysis of a respiratory Saccharomyces cerevisiae strain suggests the expression of its phenotype is glucose insensitive and predominantly controlled by Hap4, Cat8 and Mig1
© Bonander et al; licensee BioMed Central Ltd. 2008
Received: 13 March 2008
Accepted: 31 July 2008
Published: 31 July 2008
We previously described the first respiratory Saccharomyces cerevisiae strain, KOY.TM6*P, by integrating the gene encoding a chimeric hexose transporter, Tm6*, into the genome of an hxt null yeast. Subsequently we transferred this respiratory phenotype in the presence of up to 50 g/L glucose to a yeast strain, V5 hxt1-7 Δ, in which only HXT1-7 had been deleted. In this study, we compared the transcriptome of the resultant strain, V5.TM6*P, with that of its wild-type parent, V5, at different glucose concentrations.
cDNA array analyses revealed that alterations in gene expression that occur when transitioning from a respiro-fermentative (V5) to a respiratory (V5.TM6*P) strain, are very similar to those in cells undergoing a diauxic shift. We also undertook an analysis of transcription factor binding sites in our dataset by examining previously-published biological data for Hap4 (in complex with Hap2, 3, 5), Cat8 and Mig1, and used this in combination with verified binding consensus sequences to identify genes likely to be regulated by one or more of these. Of the induced genes in our dataset, 77% had binding sites for the Hap complex, with 72% having at least two. In addition, 13% were found to have a binding site for Cat8 and 21% had a binding site for Mig1. Unexpectedly, both the up- and down-regulation of many of the genes in our dataset had a clear glucose dependence in the parent V5 strain that was not present in V5.TM6*P. This indicates that the relief of glucose repression is already operable at much higher glucose concentrations than is widely accepted and suggests that glucose sensing might occur inside the cell.
Our dataset gives a remarkably complete view of the involvement of genes in the TCA cycle, glyoxylate cycle and respiratory chain in the expression of the phenotype of V5.TM6*P. Furthermore, 88% of the transcriptional response of the induced genes in our dataset can be related to the potential activities of just three proteins: Hap4, Cat8 and Mig1. Overall, our data support genetic remodelling in V5.TM6*P consistent with a respiratory metabolism which is insensitive to external glucose concentrations.
We engineered the first respiratory Saccharomyces cerevisiae strain, KOY.TM6*P, by integrating the gene encoding a chimeric hexose transporter, Tm6*, into the genome of an hxt null yeast . Subsequently we demonstrated the transferability of this respiratory phenotype to a yeast strain, V5 hxt1-7 Δ, in which only HXT1-7 had been deleted . The resulting V5.TM6*P strain produced only minor amounts of ethanol when cultured on external glucose concentrations as high as 5% . Despite the fact that the glucose flux was reduced to 30% in the V5.TM6*P strain compared with that of its parental V5 strain, the V5.TM6*P strain produced biomass at a specific rate as high as 85% of that produced by the V5 wild-type strain, the yield itself increasing by 50% in the mutant compared to its parental strain.
Having performed this initial physiological characterization of the V5.TM6*P strain, we wanted to examine more thoroughly the basis of its respiratory phenotype by comparing its transcriptome with that of its wild-type parent strain, V5. This approach has been widely used to obtain a global picture of the cellular responses of yeast to a wide range of physiological changes, including those experienced at the diauxic shift [3–5]. Recently Ohlmeier and colleagues  demonstrated that major changes at the transcriptional level are not reflected at the protein level. Indeed, focusing on the mitochondrial proteome following a diauxic shift, they showed that the levels of only 18 out of 253 identified proteins had changed (17 increased, and 1 decreased). Among them were proteins involved in the tricarboxylic acid (TCA) cycle (Sdh1, Sdh2, and Sdh4) and the respiratory chain (Cox4, Cyb2, and Qcr7). This seeming disagreement with their observations of heterogeneous changes in the transcriptome (where more than 4,000 changes were recorded) is also consistent with our own prior data where we showed that the protein patterns obtained from a 2D-PAGE analysis of the two strains (grown under the same conditions used in this study) were not substantially different . We did note, however, that the levels of the upper-part glycolytic intermediates and ATP did differ between the two strains, and that Cdc19 is present in lower amounts in V5.TM6*P. This discrepancy between proteome and transcriptome regulation has also been observed by others  and is also consistent with the fact that there is an acknowledged lack of relationship between the levels of glycolytic proteins and the glycolytic flux [8–10]. Nonetheless, researchers continue to depend on mRNA as an indicator of cellular state, which is a situation that will continue while methods for the global analysis of protein expression are improved .
In this respect, Brauer and colleagues [3–5] have used transcriptome analysis to examine adjustments and remodelling of the metabolism of glucose-limited yeast cultures. In that study, the authors constructed metabolic models consistent with their cDNA array data. In this work, the V5.TM6*P strain was therefore characterized at the transcriptome level in comparison with the parental V5 strain in order to obtain a more global understanding of any genetic remodelling underlying the physiological differences between them. Both strains were cultured in 50 g/L glucose with six samples being taken at defined points between 38 and 6 g/L glucose for cDNA array analysis in two independent fermentations for each strain. Consequently, we generated a highly redundant dataset (n = 12) where 922 genes were either induced or repressed when comparing V5 and V5.TM6*P under all glucose concentrations. Using in vivo studies together with documented binding specificities of three transcription factors known to have a role in the regulation of respiratory genes (the Hap complex, Cat8 and Mig1), it was possible to correlate the up-regulation of all but 18 of the 146 genes induced by a factor ≥2.0 with the presence of one or more of their binding sites. This means that 75% of the transcriptional response of the genes in our dataset could be related to the activities of just three transcription factors, which in turn would account for 88% of the response of the induced set. Finally, we made the unexpected observation that both the induction and repression of many of the genes in our dataset had a clear glucose dependence in the parent V5 strain that was not present in V5.TM6*P. This important result indicates that the relief of glucose repression in the wild-type is already operable at much higher glucose concentrations than is currently widely accepted and that the phenotype of the respiratory strain is glucose insensitive by comparison.
We compared the transcriptional profiles of the V5 and V5.TM6*P strains cultured in 50 g/L glucose, with samples extracted between 38 and 6 g/L residual glucose. We identified 190 genes with adjusted p-values < 0.05 that changed their expression levels either up by a factor of ≥2.0 (146 genes) or down by a factor of ≤0.5 (44 genes) when comparing the two strains at all three sampling points. These genes are presented by functional category of their gene products in Additional file 1 and are discussed below. Additional file 2 provides the information in Excel format and also includes other relevant literature data. A further 636 genes changed their expression levels either up by a factor of <2.0 (317 genes) or down by a factor > 0.5 (319 genes) and are presented as part of Additional file 3. The remaining 96 genes were either dubious, Ty-transposable elements, or were detected due to cross-hybridisation, as indicated in the legend to Additional file 1.
The majority of genes that have an altered expression in V5.TM6*P compared with its parental strain also have an altered expression following diauxic shift
We compared our dataset with three sets of expression data for yeast undergoing a diauxic shift, where the samples had been taken at least 2 hours after the glucose is depleted [3–5]. Of the 190 genes in our dataset that changed their expression levels either up or down by a factor of 2 or more, 154 of these could be identified as changing in the same direction in at least one of these three datasets, with 124 having at least two matches. Only 8 genes (ERG3, RPS27A, YHB1, TRX1, YER188w, YKL066w, YAR068w and Y0L150c) had changes that were in the opposite direction in at least two of the datasets. This significant overlap between our data set and the data sets of the three previous studies is entirely consistent with the fact that there is a transition from a respiro-fermentative metabolism in V5 to a respiratory metabolism in V5.TM6*P.
Glycolytic and gluconeogenic genes have substantially altered expression levels in V5.TM6*P
Alternative carbon source usage genes are induced in V5.TM6*P compared with V5, consistent with a reduced glucose flux
An adaptation to enable growth on alternative carbon sources is seen in the induction of the plasma membrane transporter genes MAL31 and STL1 as well as the genes SUC2, GUT1, GUT2, ALD6 and BDH1 encoding proteins for the utilization of sucrose, glycerol, acetaldehyde and butanediol, respectively. GUT1 in particular is strongly induced by a factor of 6.2, and STL1 is the most strongly induced of all genes on the cDNA array (21.7), suggesting that V5.TM6* has undergone a substantial genetic adjustment compared to its parent. Furthermore, CYB2 (10.1) and DLD3 (0.5), whose gene products are involved in the conversion of lactate to pyruvate, and in the case of CYB2 feeding the released electrons into the electron transport chain, are oppositely regulated. Biochemical data and sequence patterns for YGL157w (0.3) , YAL061w (2.9) and DSF1 (9.4, homologous to D-mannitol 2-dehydrogenase from Rhodobacter sphaeroides) suggest that these ORFs encode an oxidoreductase, a polyol dehydrogenase and a putative sugar dehydrogenase, respectively, which could be involved in channelling alternative carbon sources into the TCA cycle. The carbon-deficient state of V5.TM6*P  could also be consistent with the induction of CRC1 and CAT2 whose gene products are involved in β-oxidation and carnitine-dependent transport of acetyl-CoA from peroxisomes to mitochondria as a way of using carbon from sources other than glucose . The induction of SFC1, JEN1, FBP1 and ATO2 seen in our study has previously been reported to be a specific identifier of carbon-limited growth .
All TCA- and glyoxylate cycle genes have increased expression in V5.TM6*P compared with V5
The genes encoding the enzymes of the TCA cycle (CIT1, ACO1, IDH1, KGD1, KGD2, LSC2, SDH1-4, FUM1, MDH1) all have a significant (adjusted p-value < 0.05) increase in expression when comparing V5 and V5.TM6*P. In addition, the gene encoding an isoform of aconitase (ACO2) has its expression level reduced by a factor of 0.5. Fermentation is almost completely abolished in the V5.TM6*P strain, and we have previously shown that the TCA cycle has a higher carbon flux  to harness the electron flow into the respiratory chain and generate a proton gradient using F0F1-ATPase for ATP synthesis (several respiratory chain genes are induced, as discussed below). Interestingly, all genes of the TCA-cycle are induced by greater than a factor of 3.5 except for CIT1/CIT2 and KGD1/KGD2 where the factor change is 2.2 – 3.5 for each gene of the pair, presumably yielding the same overall transcriptional response. Consistent with this, Ohlmeier and colleagues have shown that Sdh1, 2 and 4 all have increased expression after a diauxic shift .
Growth characteristics of V5 and V5.TM6*P at 30°C starting at 50 g/L glucose.
Generation time during glucose growth (h)
3.7 (0.2; n = 2)
4.5 (0.3; n = 4)
Biomass yield (g/g)
0.21 (0.02 ; n = 2)
0.32 (0.01; n = 4)
Glucose consumption (mmolg dry weight) -1 h -1
9.00 (0.70; n = 2)
2.70 (0.20; n = 4)
Ethanol yield (g/g)
0.33 (0.01; n = 2)
0.05 (0.01; n = 4)
Glycerol yield (g/g)
0.04 (0; n = 2)
Genes encoding respiratory chain enzymes are induced in V5.TM6*P compared with V5
Changes of expression in genes encoding mitochondrial and plasma membrane transporters are consistent with a respiratory phenotype in V5.TM6*P
The genes encoding the hexose transporters, Hxt1 to Hxt7, have very high sequence identities  and thus cross reactivity between them is expected on the array. This explains the observed signal for HXT1, 4 and 6, which have been deleted in V5.TM6*P. HXT13, HXT15 and HXT16 are known to be repressed by high levels of glucose and are consequently induced as expected in V5.TM6*P. A carbon source deficiency triggers an up-regulation of genes encoding plasma membrane proteins responsible for the import of alternative carbon sources, as already noted: STL1 (21.7; glycerol); JEN1 (10.2; lactate), PUT4 (5.8; proline – the main nitrogen source in grape juice) and MAL31 (3.1; maltose). A gene associated with nutrient deficiency in yeast , YGP1, is induced by a factor of 2.6, again supporting the notion that V5.TM6*P has undergone a genetic adjustment compared to its parent.
It seems that V5.TM6*P also has an altered requirement for metal ions in that genes encoding three metal ion channels have altered expression: CTR3 (copper transport) is induced by a factor of 3.0, while ZRT1 (zinc transport) and ENB1 (iron transport) are repressed by a factor of 0.4 – 0.5. It is likely that copper import is increased so that copper can be transferred to cytochrome c oxidase, which is the main copper-requiring enzyme in Additional file 1, with 3 copper ions per complex and a copy number of 5,000–12,000 per cell. There are only 45 genes in Additional file 1 with greater than 5,000 copies per cell and none of these has copper as a prosthetic group except cytochrome c oxidase.
The amino acid uptake systems encoded by BAP3 and GNP1 are repressed while the proline permease encoded by PUT4 is induced. PUT4 has a carbon source responsive element (CSRE)  that is expected to be induced at the low glucose flux experienced in the V5.TM6*P strain, while BAP3 and GNP1 are regulated by a sensor of extracellular amino acid concentration, where Ssy5  is a vital component. It is likely that in V5.TM6*P the repression of BAP3 and GNP1 is a result of the repression (by a factor 0.65; Additional file 3) of SSY5 seen on the cDNA array. Interestingly, induction of genes encoding the ammonia transport system in the V5.TM6*P strain (ATO2, MEP1) could suggest that the strain is experiencing general nutrient limitation. It has previously been reported for example that the induction of another MEP family member, MEP2, is associated with nitrogen limitation .
88% of genes induced in V5.TM6*P compared to its parent have transcription factor binding sites for one or more of the Hap complex, Cat8 or Mig1
Genome-wide studies have yielded a wealth of information on which promoters are bound by different transcription factors . In order to rationalise these resources, we began by looking for biologically-verified data on transcription factors that might control the repression of the subset of 44 genes of Additional file 1. Since only limited information was publicly available, we did not pursue this part of the dataset further. In contrast, for the 146 induced genes in Additional file 1, our observation of the similarity between the transcriptional changes in the V5 to V5.TM6*P transition and those in the transition from glucose to ethanol growth lead us to analyse the roles of Hap4, Cat8 and Mig1 which have been previously been verified to be crucial to the expression of genes in the diauxic shift [16, 19, 20]: Hap4 (in complex with Hap2, 3 and 5)  and Cat8 (optionally in complex with Sip4 and Adr1)  are transcriptional activators, whereas Mig1 (optionally with Mig2/3) is a transcriptional repressor that works in concert with Tup1-Cyc8 .
Biologically-verified data on transcription factors that act on genes also found in Additional file 1: transcript dependence on HAP4 over-expression.
FACTOR CHANGE IN GENE EXPRESSION FROM V5 TO V5.TM6*P
FACTOR CHANGE IN GENE EXPRESSION FROM WT TO A HAP4 OVER-EXPRESSION STRAIN
Alternative carbon source utilization
Plasma membrane transport
Fatty acid metabolism
Ribosomal proteins in the mitochondria or cytosol
Biologically-verified data on transcription factors that act on genes also found in Additional file 1: transcript dependence on Cat8 and Adr1.
FACTOR CHANGE IN GENE EXPRESSION FROM V5 TO V5.TM6*Pa
Alternative carbon source utilization
Plasma membrane transport
2.7b; 6.1c; 2.0d
3.6b; 10c; 10.5d
Biologically-verified data on transcription factors that act on genes also found in Additional file 1: transcript dependence on Mig1.
FACTOR CHANGE IN GENE EXPRESSION FROM V5 TO V5.TM6*Pa
Alternative carbon source utilization
Plasma membrane transport
We noted that HAP4 (factor change 1.7, Additional file 3) is induced and CAT8 (0.5, adjusted p value < 0.06; Additional file 5), MIG2 (0.5, Additional file 1) and MIG3 (0.3, Additional file 1) are down-regulated in V5.TM6*P compared to V5. In our dataset, the specific change in array signal for MIG1 had an adjusted p-value > 0.05 over the complete experimental range. The result for CAT8 was unexpected as it contains an upstream Mig1 binding site similar to HAP4 (which is induced) and Cat8 target genes such as FBP1 and PCK1 are themselves induced. It is possible that the transcript was not probed accurately on the array either due to cross-hybridisation between closely related DNA sequences or on account of its low base expression value (0.24). A future real time Q-PCR experiment would probe the individual transcript.
Transcripts dependent on external glucose concentration are not found in the respiratory V5.TM6*P strain, even though several glucose- or carbon-source-associated transcripts are glucose dependent in the parental wild-type strain
In order to examine whether the glucose dependence of genes in V5 affected our data in Additional file 1, we excluded the data points at 13.5 and 5.6 g/L glucose for V5.TM6*P, and 10.5 and 7.4 g/L glucose for V5 (which gave the highest array signals) and compared the cDNA array data in this case with each other exactly as for the full data set. Following this procedure, only a small subset of genes that are highly induced had their factor changes increased substantially, but this did not affect the fact that they remain within the top 30 most induced genes in Additional file 1. They are HXK1 (9.0 in Additional file 1, 19.3 following re-calculation; the gene contains binding sites for the Hap complex, Cat8 and Mig1); STL1 (21.7, 27.5; Hap complex, Cat8); DSF1 (9.4, 15.1; Hap complex, Mig1); CYB2 (10.1, 13.4; Hap complex, Mig1) and FBP1 (5.3, 7.4; Cat8, Mig1). Fourteen of the seventeen genes that were most affected by this recalculation were found to contain a Mig1 binding site. It is therefore possible that Mig1 is also involved in relieving glucose repression in the 10 to 5 g/L range, and not only at the depletion of glucose at the diauxic shift. Overall, we found that the amplitudes of the factor changes varied by only 9.3% in this new calculation compared with the original one, which indicated that the glucose dependence of the V5 strain does not have a major influence on the way we generated the data presented in Additional file 1.
The main physiological change in both the diauxic shift and the transition from a respiro-fermentative to a respiratory metabolism is that glycolysis is channelled to support respiration: this occurs upon glucose depletion during the diauxic shift and throughout the growth curve in glucose medium in V5.TM6*P. Although previously-published genomic analyses of yeast's response to the diauxic shift transition have all been sampled under different conditions – rich (YPD) medium with 2% glucose , minimal defined medium in chemostat cultures at different dilution rates , and minimal defined media in batch cultures sampled 4 h after the diauxic shift  – common themes emerge with changes on going from V5 to V5.TM6*P. Overall 81% of genes which are induced or repressed on going from V5 to V5.TM6*P in our study were found in at least one of these three datasets for the diauxic shift. Furthermore, we also find that for all proteins induced in a study of the diauxic shift by 2D-gel electrophoresis , the corresponding genes are induced on going from V5 to V5.TM6*P except for the alcohol dehydrogenase, Adh2. There is a hence close agreement between induction of the following genes and the levels of the corresponding proteins: FBP1, ICL1, PCK1, IDP2, MLS1, DLD1, ALD6, SDH1 and CIT2. This strong correlation with previously-published data and our own observed correlation of data for Cdc19 and CDC19 supports the phenotypic similarity between the V5 to V5.TM6*P transition and the diauxic shift. It is worth noting that there is no clear correlation of our data in Additional file 1 with any possible differences in growth rate between the two strains. In a study by Regenberg and colleagues , 180 genes changed in expression by more than a factor of 4 when comparing growth rates. Of these 180, 3% of all genes are found in Additional file 1 providing no significant overlap in data.
It is also possible to see parallels between our data set and a study of nitrogen deprivation and stationary phase growth for wild-type yeast by Gasch and colleagues . This is supported by the fact that ten of the 15 genes that were found to be strongly induced upon nitrogen deprivation and growth in stationary phase in that study are also induced in the V5 to V5.TM6*P transition. Only two genes, YGR067c (likely to encode a transcription factor as it has high partial sequence homology with MIG3 and ADR1 and it fine-tunes the response to nutrient limitation) and ECM13, are not induced in the V5.TM6*P strain, but were highly induced in the Gasch study  upon nitrogen deprivation and stationary phase growth. In addition, we find that the genes encoding Mep1 and Ato2, which are involved in ammonia transport are also strongly induced in V5.TM6*P in our study. Such an induction is seen in both the diauxic shift studies performed in defined media [4, 34], but not in rich medium , suggestive of nutrient limitation. In fact, V5.TM6*P also appears carbon limited, not least as a result of the fact that amongst the most strongly induced genes are FBP1 and PCK1, which encode gluconeogenic enzymes. STL1, MAL31 and SUC2 (encoding proteins involved in usage and/or transport of alternative carbon sources: glycerol, maltose and sucrose respectively) are induced in V5.TM6*P, also consistent with its low glycolytic flux and suggestive of an adaptation of metabolism ('discontinuous modelling) as seen by Brauer and colleagues [4, 34] and supported by our previous work . Several of these genes including FBP1 and STL1, as well as JEN1, ODC1, SFC1 and ATO2 have already been identified as being either induced under carbon limitation or being specific indicators of carbon limitation in yeast cultures , consistent with our own data.
The analysis of transcription factor binding sites present in the genes of Additional file 1 revealed that 127 genes had Hap complex binding sites, the majority of which encode mitochondrial proteins. Of genes that only had Hap complex binding sites present, thirteen had only one and of these only five were induced leading us to hypothesise that multiple Hap complex binding sites are correlated with induction of the Additional file 1 genes (Fig. 3A). Interestingly, in a microarray screen of a hap1Δ mutant for Hap1-dependent genes , 7/24 genes that were induced in the hap1Δ mutant are also induced in Additional file 1 (CYB2, MLS1, HMX1, PUT4, STL1, HPA2, PUT1), and in each case also have a Hap complex binding site. In addition, 19 genes containing a Cat8 site and 31 genes containing a Mig1 site were predicted, with many of these genes being experimentally-verified in previous in vivo studies. Our analysis of the regulation of 88% of the induced genes of our unique respiratory strain (Fig. 3) is further supported by a novel computational method  that predicts the Hap complex and Mig1 control the genes of Additional file 1; Cat8 is not included in that database.
A quite unexpected result from comparing the transcriptional profiles of V5 with V5.TM6*P is the data set in Figure 4, showing genes from the parent strain that are clearly glucose dependent and that the glucose dependence observed begins at concentrations (6 – 36 g/L glucose) higher than the range generally regarded as the glucose-repressed region. Strain DBY7286 , which is unrelated to V5, previously showed a gradual relief of glucose repression starting at 20 g/L glucose for GLK1, HXK1 and FAL1. These results for V5 and DBY7286 suggest that gradual relief is a general S. cerevisiae phenomenon. Future studies will show whether the gradient of the slope is different for different strains. That such glucose dependence requires an intact glucose uptake system is suggested by the fact that it is not seen in the V5.TM6*P strain: the array signal is not changed significantly for any of the genes in the V5.TM6*P strain at the different glucose concentrations sampled. This implies that in V5.TM6*P, glucose sensing is independent of external glucose concentration. The wild-type yeast, on the other hand, shows an extensive modulation of its expression profile at high glucose concentrations despite the fact that its physiological response is largely unaffected. The insensitivity of the V5.TM6*P strain therefore makes it an extremely valuable biotechnological tool as it can be cultured in a wide range of external glucose concentrations, whilst maintaining the same respiratory phenotype .
In this study, we have been able to characterize the transcriptome of a unique respiratory yeast strain. We have been able to identify highly complete collections of known genes in the TCA cycle, glyoxylate cycle and respiratory chain that are consistent with a respiratory metabolism. Our results suggest that there has been genetic remodelling predominantly through the activity of Hap4, Cat8 and Mig1, and that the gene expression profile of V5.TM6*P during growth on glucose resembles wild-type S. cerevisiae cells in the diauxic shift.
The auxotrophic wine yeast strain V5 (MAT a ura3 gal) and the V5 hxt1–hxt7 Δ mutant strain (MAT a ura3 gal hxt514 Δ::loxP hxt367 Δ::lox P hxt2 Δ::loxP) were kindly provided by the Professor Bruno Blondin, Montpellier, France. The construction of V5.TM6*P was previously described by Henricsson and colleagues  where the expression cassette from KOY.TM6*P  was transferred into the genome of the V5 hxt1–hxt7 Δ strain by using the primers PROHXT3 (TCAAATGGCGGTGTAGTTTGAAAAG) and TERHXT7 (TTAAGTGACGGGCGATGAGTAAGAA). Transformants were selected by growth on glucose. The resulting ura3 strain is referred to as V5.TM6*. The auxotrophic V5 wild type and V5.TM6* used in this study were made prototrophic by integration of URA3  and are referred to as the V5 wild-type strain and V5.TM6*P, respectively.
For both pre-cultures and main cultures, 5 × defined minimal medium  with 5% glucose as the sole carbon and energy source was used. A two-step pre-cultivation was performed on a rotary shaker. First a 10 mL culture was grown for 72 h, and used to inoculate 100 mL medium, which after 30 h at 30°C was inoculated to give a final OD610 of 0.05 in a bioreactor (BRO2, Belach Bioteknik AB, Sweden) in a volume of 2.5 L. Polypropylene glycol P2000 was added as antifoam (100 μ L/L). The cultivation conditions were 30°C, 1,000 rpm, pH 5.0 with an airflow of 1.25 L/min. Gas evolution was monitored on-line (type CP460 O2/CO2, Belach Bioteknik AB). Cultivations were performed at least in duplicate.
Yeast cells (2 – 20 mL) were directly transferred into twice their volume of ice-cold water. The cells were collected by centrifugation at 0°C for 5 min at 3000 × g, frozen in liquid nitrogen and stored at -80°C. Total RNA was then prepared using the RNeasy kit from Qiagen, following the manufacturer's instructions.
cDNA array production
PCR products for the yeast ORFs were generated from yeast DNA using the ORF-specific primer set from Research Genetics and then re-amplified with the Resgen universal forward and reverse primers (Catalogue number 40612). PCR products were prepared by Randy Strich (Fox Chase Cancer Centre, USA) as part of a collaboration with the Wistar Institute, Philadelphia, USA. PCR products were spotted with a BioRobotics Microgrid TASII (Genomic Solutions) by the Wistar Genomics Facility. Each array contained 6,319 individual yeast genes and 1,169 gene repeats. The complete list of arrayed genes on array YA04 is listed under GEO Series accession number GPL4423.
RNA amplification, hybridization, and scanning
Amplified RNA (aRNA) was prepared from 1 μ g of total RNA for each sample tested using the RiboAmp kit (Arcturus). Labelled targets were prepared from 1.6 μ g aRNA with Superscript II reverse transcriptase (Invitrogen), in the presence of 3,000–5,000 Ci/mM [α-33P] dCTP (Amersham Pharmacia Biotech), 1 mmol/L dATP, 1 mmol/L dTTP, 1 mmol/L dGTP, 67 ng/μ g oligo-dT (Promega Biosciences), and 0.65 × random decamer primers (Ambion). Labelled targets were hybridized to individual arrays at 42°C for 18 h in 3 mL Microhyb buffer (Invitrogen). Arrays were washed twice in 2 × standard saline citrate (SSC)/1% SDS solution for 30 min at 50°C, once in 0.5 × SSC/1% SDS and once in 0.1 × SSC/0.5% SDS for 30 min at 55°C. The arrays were exposed to phosphor screens (Amersham Biosciences) for 6 days and scanned in a Storm 820 PhosphorImager (Molecular Dynamics). Quantitation of each spot was assessed by Imagene 5.0 software (BioDiscovery Incorporated) using manual spot alignment, measuring median pixel intensity for each spot and subtracting the local background. The data have been deposited in the NCBI Gene Expression Omnibus and are accessible through GEO Series accession number GSE11799.
cDNA array data analysis
Observed values less than 0.1 (5% (8752 values) of the 176640 raw data values) were set to 0.1, in accordance with information from Wistar that variations in values below this threshold should be regarded as measurement noise. After converting to a logarithmic scale, average expression values were computed over spots with the same ORF, and arrays were scale normalized by subtracting the median for each of the 24 arrays. Averages were computed over repeated measurements with the same conditions producing 6 values at different glucose concentrations for each of the V5 (38.4, 35, 25.5, 24.4, 10.5, 7.4 g/L glucose) and V5.TM6*P (36.5, 34.8, 26.3, 25.1, 13.5, 5.6 g/L glucose) strains for each gene. T-tests were then performed, comparing the two strains, using Empirical Bayes robustification of variance estimates [39, 40].
The T-tests were adjusted for multiple testing using the false discovery rate method (disregarding correlations between genes). Essentially, then, p-values were adjusted so that when selecting all genes with p-values less than a threshold q, a proportion of q false positives would be expected amongst these genes. The estimated effects are shown as factors in Additional file 1, as the linear modelling was performed on a log scale. The p-values in Additional file 1 are the adjusted values, so that selecting genes with p-values < 0.05 gives an expected false discovery rate of 5%. Calculations were performed using the BioConductor package limma . To present the data, a tabular format was chosen over heatmaps in an attempt to make it as accessible as possible to the scientific community.
When producing data for Figure 4, a similar linear model was used, but with a different design matrix: Essentially, for each gene, 2 straight lines were simultaneously fitted, one for each of the strains, to plots of the 6 log-expression values versus glucose concentration. Empirical Bayes robustification and adjustment for multiple testing was included even in this analysis, using the limma tool . Lines with slopes significantly different from zero were found for 20 genes for the V5 strain, but for no genes from the V5.TM6*P strain.
This work was supported by the European Commission via contracts LSHG-CT-2004-504601 (E-MeP) and LSHG-CT-2006-037793 (OptiCryst) to RMB, and QLG2-CT-2002-00988 (SPINE) to LG and RMB. The Wistar Genomics facility is supported in part by NCI P30 CA10815-34S3. The BBSRC supports bioreactors in the RMB laboratory through an REI award.
- Otterstedt K, Larsson C, Bill RM, Stahlberg A, Boles E, Hohmann S, Gustafsson L: Switching the mode of metabolism in the yeast Saccharomyces cerevisiae. EMBO Rep. 2004, 5 (5): 532-537.PubMedPubMed CentralView ArticleGoogle Scholar
- Henricsson C, de Jesus Ferreira MC, Hedfalk K, Elbing K, Larsson C, Bill RM, Norbeck J, Hohmann S, Gustafsson L: Engineering of a novel Saccharomyces cerevisiae wine strain with a respiratory phenotype at high external glucose concentrations. Appl Environ Microbiol. 2005, 71 (10): 6185-6192.PubMedPubMed CentralView ArticleGoogle Scholar
- Bonander N, Hedfalk K, Larsson C, Mostad P, Chang C, Gustafsson L, Bill RM: Design of improved membrane protein production experiments: quantitation of the host response. Protein Sci. 2005, 14 (7): 1729-1740.PubMedPubMed CentralView ArticleGoogle Scholar
- Brauer MJ, Saldanha AJ, Dolinski K, Botstein D: Homeostatic adjustment and metabolic remodeling in glucose-limited yeast cultures. Mol Biol Cell. 2005, 16 (5): 2503-2517.PubMedPubMed CentralView ArticleGoogle Scholar
- DeRisi JL, Iyer VR, Brown PO: Exploring the metabolic and genetic control of gene expression on a genomic scale. Science. 1997, 278 (5338): 680-686.PubMedView ArticleGoogle Scholar
- Ohlmeier S, Kastaniotis AJ, Hiltunen JK, Bergmann U: The yeast mitochondrial proteome, a study of fermentative and respiratory growth. J Biol Chem. 2004, 279 (6): 3956-3979.PubMedView ArticleGoogle Scholar
- Hack CJ: Integrated transcriptome and proteome data: the challenges ahead. Brief Funct Genomic Proteomic. 2004, 3 (3): 212-219.PubMedView ArticleGoogle Scholar
- Davies SE, Brindle KM: Effects of overexpression of phosphofructokinase on glycolysis in the yeast Saccharomyces cerevisiae. Biochemistry. 1992, 31 (19): 4729-4735.PubMedView ArticleGoogle Scholar
- Schaaff I, Heinisch J, Zimmermann FK: Overproduction of glycolytic enzymes in yeast. Yeast. 1989, 5 (4): 285-290.PubMedView ArticleGoogle Scholar
- van Hoek P, van Dijken JP, Pronk JT: Regulation of fermentative capacity and levels of glycolytic enzymes in chemostat cultures of Saccharomyces cerevisiae. Enzyme Microb Technol. 2000, 26 (9-10): 724-736.PubMedView ArticleGoogle Scholar
- Katz M, Frejd T, Hahn-Hagerdal B, Gorwa-Grauslund MF: Efficient anaerobic whole cell stereoselective bioreduction with recombinant Saccharomyces cerevisiae. Biotechnol Bioeng. 2003, 84 (5): 573-582.PubMedView ArticleGoogle Scholar
- Swiegers JH, Dippenaar N, Pretorius IS, Bauer FF: Carnitine-dependent metabolic activities in Saccharomyces cerevisiae: three carnitine acetyltransferases are essential in a carnitine-dependent strain. Yeast. 2001, 18 (7): 585-595.PubMedView ArticleGoogle Scholar
- Boer VM, de Winde JH, Pronk JT, Piper MD: The genome-wide transcriptional responses of Saccharomyces cerevisiae grown on glucose in aerobic chemostat cultures limited for carbon, nitrogen, phosphorus, or sulfur. J Biol Chem. 2003, 278 (5): 3265-3274.PubMedView ArticleGoogle Scholar
- Kruckeberg AL: The hexose transporter family of Saccharomyces cerevisiae. Arch Microbiol. 1996, 166 (5): 283-292.PubMedView ArticleGoogle Scholar
- Destruelle M, Holzer H, Klionsky DJ: Identification and characterization of a novel yeast gene: the YGP1 gene product is a highly glycosylated secreted protein that is synthesized in response to nutrient limitation. Mol Cell Biol. 1994, 14 (4): 2740-2754.PubMedPubMed CentralView ArticleGoogle Scholar
- Haurie V, Perrot M, Mini T, Jeno P, Sagliocco F, Boucherie H: The transcriptional activator Cat8p provides a major contribution to the reprogramming of carbon metabolism during the diauxic shift in Saccharomyces cerevisiae. J Biol Chem. 2001, 276 (1): 76-85.PubMedView ArticleGoogle Scholar
- Eckert-Boulet N, Regenberg B, Nielsen J: Grr1p is required for transcriptional induction of amino acid permease genes and proper transcriptional regulation of genes in carbon metabolism of Saccharomyces cerevisiae. Curr Genet. 2005, 47 (3): 139-149.PubMedView ArticleGoogle Scholar
- Harbison CT, Gordon DB, Lee TI, Rinaldi NJ, Macisaac KD, Danford TW, Hannett NM, Tagne JB, Reynolds DB, Yoo J, Jennings EG, Zeitlinger J, Pokholok DK, Kellis M, Rolfe PA, Takusagawa KT, Lander ES, Gifford DK, Fraenkel E, Young RA: Transcriptional regulatory code of a eukaryotic genome. Nature. 2004, 431 (7004): 99-104.PubMedPubMed CentralView ArticleGoogle Scholar
- Lascaris R, Bussemaker HJ, Boorsma A, Piper M, van der Spek H, Grivell L, Blom J: Hap4p overexpression in glucose-grown Saccharomyces cerevisiae induces cells to enter a novel metabolic state. Genome Biol. 2003, 4 (1): R3-PubMedPubMed CentralView ArticleGoogle Scholar
- Lutfiyya LL, Johnston M: Two zinc-finger-containing repressors are responsible for glucose repression of SUC2 expression. Mol Cell Biol. 1996, 16 (9): 4790-4797.PubMedPubMed CentralView ArticleGoogle Scholar
- Roth S, Kumme J, Schuller HJ: Transcriptional activators Cat8 and Sip4 discriminate between sequence variants of the carbon source-responsive promoter element in the yeast Saccharomyces cerevisiae. Curr Genet. 2004, 45 (3): 121-128.PubMedView ArticleGoogle Scholar
- Treitel MA, Carlson M: Repression by SSN6-TUP1 is directed by MIG1, a repressor/activator protein. Proc Natl Acad Sci U S A. 1995, 92 (8): 3132-3136.PubMedPubMed CentralView ArticleGoogle Scholar
- McNabb DS, Xing Y, Guarente L: Cloning of yeast HAP5: a novel subunit of a heterotrimeric complex required for CCAAT binding. Genes Dev. 1995, 9 (1): 47-58.PubMedView ArticleGoogle Scholar
- Romer KA, Kayombya GR, Fraenkel E: WebMOTIFS: automated discovery, filtering and scoring of DNA sequence motifs using multiple programs and Bayesian approaches. Nucleic Acids Res. 2007, 35 (Web Server issue): W217-20.PubMedPubMed CentralView ArticleGoogle Scholar
- Tachibana C, Yoo JY, Tagne JB, Kacherovsky N, Lee TI, Young ET: Combined global localization analysis and transcriptome data identify genes that are directly coregulated by Adr1 and Cat8. Mol Cell Biol. 2005, 25 (6): 2138-2146.PubMedPubMed CentralView ArticleGoogle Scholar
- Lutfiyya LL, Iyer VR, DeRisi J, DeVit MJ, Brown PO, Johnston M: Characterization of three related glucose repressors and genes they regulate in Saccharomyces cerevisiae. Genetics. 1998, 150 (4): 1377-1391.PubMedPubMed CentralGoogle Scholar
- Mukherjee S, Berger MF, Jona G, Wang XS, Muzzey D, Snyder M, Young RA, Bulyk ML: Rapid analysis of the DNA-binding specificities of transcription factors with DNA microarrays. Nat Genet. 2004, 36 (12): 1331-1339.PubMedPubMed CentralView ArticleGoogle Scholar
- Lundin M, Nehlin JO, Ronne H: Importance of a flanking AT-rich region in target site recognition by the GC box-binding zinc finger protein MIG1. Mol Cell Biol. 1994, 14 (3): 1979-1985.PubMedPubMed CentralView ArticleGoogle Scholar
- Yin Z, Wilson S, Hauser NC, Tournu H, Hoheisel JD, Brown AJ: Glucose triggers different global responses in yeast, depending on the strength of the signal, and transiently stabilizes ribosomal protein mRNAs. Mol Microbiol. 2003, 48 (3): 713-724.PubMedView ArticleGoogle Scholar
- Larroy C, Pares X, Biosca JA: Characterization of a Saccharomyces cerevisiae NADP(H)-dependent alcohol dehydrogenase (ADHVII), a member of the cinnamyl alcohol dehydrogenase family. Eur J Biochem. 2002, 269 (22): 5738-5745.PubMedView ArticleGoogle Scholar
- Yin Z, Smith RJ, Brown AJ: Multiple signalling pathways trigger the exquisite sensitivity of yeast gluconeogenic mRNAs to glucose. Mol Microbiol. 1996, 20 (4): 751-764.PubMedView ArticleGoogle Scholar
- Regenberg B, Grotkjaer T, Winther O, Fausboll A, Akesson M, Bro C, Hansen LK, Brunak S, Nielsen J: Growth-rate regulated genes have profound impact on interpretation of transcriptome profiling in Saccharomyces cerevisiae. Genome Biol. 2006, 7 (11): R107-PubMedPubMed CentralView ArticleGoogle Scholar
- Gasch AP, Spellman PT, Kao CM, Carmel-Harel O, Eisen MB, Storz G, Botstein D, Brown PO: Genomic expression programs in the response of yeast cells to environmental changes. Mol Biol Cell. 2000, 11 (12): 4241-4257.PubMedPubMed CentralView ArticleGoogle Scholar
- Brown AJ, Planta RJ, Restuhadi F, Bailey DA, Butler PR, Cadahia JL, Cerdan ME, De Jonge M, Gardner DC, Gent ME, Hayes A, Kolen CP, Lombardia LJ, Murad AM, Oliver RA, Sefton M, Thevelein JM, Tournu H, van Delft YJ, Verbart DJ, Winderickx J, Oliver SG: Transcript analysis of 1003 novel yeast genes using high-throughput northern hybridizations. EMBO J. 2001, 20 (12): 3177-3186.PubMedPubMed CentralView ArticleGoogle Scholar
- Ter Linde JJ, Steensma HY: A microarray-assisted screen for potential Hap1 and Rox1 target genes in Saccharomyces cerevisiae. Yeast. 2002, 19 (10): 825-840.PubMedView ArticleGoogle Scholar
- Ernst J, Vainas O, Harbison CT, Simon I, Bar-Joseph Z: Reconstructing dynamic regulatory maps. Mol Syst Biol. 2007, 3: 74-PubMedPubMed CentralView ArticleGoogle Scholar
- Wieczorke R, Krampe S, Weierstall T, Freidel K, Hollenberg CP, Boles E: Concurrent knock-out of at least 20 transporter genes is required to block uptake of hexoses in Saccharomyces cerevisiae. FEBS Lett. 1999, 464 (3): 123-128.PubMedView ArticleGoogle Scholar
- Verduyn C, Postma E, Scheffers WA, Van Dijken JP: Effect of benzoic acid on metabolic fluxes in yeasts: a continuous-culture study on the regulation of respiration and alcoholic fermentation. Yeast. 1992, 8 (7): 501-517.PubMedView ArticleGoogle Scholar
- Breitling R: Biological microarray interpretation: The rules of engagement. Biochim Biophys Acta. 2006, 1759 (7): 319-327.PubMedView ArticleGoogle Scholar
- Smyth GK: Limma: linear models for microarray data. Bioinformatics and Computational Biology Solutions using R and Bioconductor. Edited by: Gentleman R, Carey V, Dudoit S, Irizarry R, Huber W. 2005, New York , Springer, 397-420.View ArticleGoogle Scholar
- Velculescu VE, Zhang L, Zhou W, Vogelstein J, Basrai MA, Bassett DE, Hieter P, Vogelstein B, Kinzler KW: Characterization of the yeast transcriptome. Cell. 1997, 88 (2): 243-251.PubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.