CAUSATIVE GENES CONFERRING ACETIC ACID TOLERANCE IN YEAST
The present invention relates to genes conferring acetic acid tolerance in yeast. More specifically, the invention relates to the use of DOT5, preferably in combination with CUP2 and/or HAA1 to obtain acid tolerance in yeast. Even more preferably, the invention relates to specific alleles of said genes, and to yeast strains comprising said specific alleles.
Hydrolysates of lignocellulose are an interesting source for the production of bioethanol. However, one of the problems is the presence of toxic compounds such as acetic acid, furfural and lignin derivatives. Resistance against these inhibitors is essential for an efficient bioethanol production (Olsson and Hahn-Hagerdal, 1993). Especially acetic acid is known to have an inhibitory effect (Limtong et al., 2000). However, although overexpression of a single gene may improve acetic acid tolerance (Tanaka et al., 2012), it is important to understand the interplay of genes, proteins and other components that determine the physiological properties of a microorganism.
In the past, research focussed indeed primarily on the identification of single alleles or genetic loci that are involved in physiological traits (Glazier et al., 2002). However, in contrast to Mendelian traits (traits that are caused by one single locus), quantitative traits are caused by multiple genetic loci, which makes the unraveling of these complex traits rather difficult (Steinmetz et al., 2002). In addition, the genetic mapping of quantitative trait loci (QTL) is hampered by genetic heterogeneity, variable phenotypic contributions of each QTL, epistasis and gene-environment interactions (Flint and Mott, 2001 ). These limitations have facilitated the development of novel technologies to simultaneously identify genomic loci that are involved in complex traits. With these technologies, phenotypes like high-temperature tolerance, efficient sporulation and chemical resistance have been genetically unraveled (Steinmetz et al, 2002; Deutschbauer and Davies, 2005; Ehrenreich et al., 2010).
Recently, Swinnen et al. (2012) developed such a strategy, which was successfully employed to identify genetic determinants that are involved in high ethanol tolerance in the yeast Saccharomyces cerevisiae. In this strategy, called pooled-segregant whole-genome sequence analysis, it was demonstrated that QTLs underlying a complex trait can be mapped using small populations of segregants. However, the identification of causative mutations in these QTLs remains cumbersome since this method results in a relatively large size of the identified loci, which infers the analysis of a large number of genes. Reducing the size of QTLs can be achieved with inbreeding crosses, as was recently described by Parts et al (201 1 ). However, the use of very large pools makes it an extensive procedure, especially since phenotyping industrially relevant traits often requires elaborate procedures, making the use of large numbers of segregants undesirable. Furthermore, although inbreeding crosses can be used to decrease the size QTLs, it remains unknown how it influences the mapping of minor loci.
In order to investigate the effect of inbreeding crosses on QTL mapping of industrially relevant strains, we have applied the pooled-segregant whole-genome sequencing analysis methodology on F1 and F7 segregants of a cross between a yeast strain that is superior for acetic acid tolerance and an industrial strain that is inferior for the same trait. Acetic acid tolerance is an industrially important characteristic as yeast fermentation is severely inhibited by this weak organic acid. As mentioned above, the presence of acetic acid in lignocellulosic hydrolysate strongly affects the fermentative capacity of yeast (Casey et al., 2010; Huang et al., 201 1 ; Narendranath et al, 2001 ; Taherzadeh et al., 1997; Almeida et al., 2007). Especially the fermentation of pentose sugars suffers from the presence of acetic acid (Caseay et al., 2010; Bellissimi et al, 2009; Matsushika and Sawayama, 2012), emphasizing the importance of high acetic acid tolerance to enable efficient conversion of all sugars in lignocellulosic hydrolysate into ethanol. However, multiple attempts to rationally engineer increased acetic acid tolerance in yeast were met with limited success as a high number of genes is involved in the response to acetic acid stress (Abott et al., 2007; Mira et al., 2010 a & b; Li and Yuan, 2010, Hasunuma et al., 201 1 ; Zhang et al., 201 1 ) . Random approaches such as evolutionary engineering has rendered improved strains in terms of acetic acid tolerance (Koppram et al., 2012; Wright et al., 201 1 ), but this method leads to overselection of a single trait and to possible loss of other important properties.
We found for the first time that increased recombination frequency indeed results in the expected smaller loci, but also in unexpected appearance and disappearance of QTLs, compared to QTL mapping without inbreeding crosses. Furthermore, combining individual whole-genome sequencing data of acetic acid tolerant segregants with bioinformatics analysis enabled QTL mapping to single gene level. Surprisingly, DOT5 plays an important role in acetic acid tolerance, preferably in combination with CUP2 and HAA1. Even more surprisingly, we were able to identify superior alleles of these genes, which confer an even higher acetic acid tolerance to the yeast.
A first aspect of the invention is the use of DOT5 to obtain acetic acid tolerance in yeast in yeast. Preferably, said use is the overexpression of the gene, and/or the use of a specific allele. Most preferably, said use is the use of a specific DOT5 allele. Preferably said specific DOT5 allele is encoding SEQ ID No. 2
Preferably, the use of DOT5 according to the invention is combined with the use of CUP2 and/or HAA1 and/or VMA7. Preferably, said use is the overexpression of the genes, and/or the use of specific alleles. Even more preferably, said CUP2 allele encodes SEQ ID No.4, said HAA 1 allele encodes SEQ ID No.6 and said VMA7 allele encodes SEQ ID No. 8. Preferably, said yeast is a xylose fermenting yeast. A xylose fermenting yeast, as used here, can be a yeast that is naturally producing ethanol on the base of xylose, or it can be a yeast that is mutated and/or genetically engineered to ferment xylose and to produce ethanol on the base of xylose. Even more preferably, said yeast is selected from the group consisting of Saccharomyces sp., Pichia sp., Candida sp., Pachysolen sp. and Spathaspora sp. Most preferably, said yeast is a Saccharomyces sp. preferably a Saccharomyces cerevisiae.
Another aspect of the invention is the use of a HAA1 allele encoding SEQ ID No. 6 for obtaining acetic acid tolerance in yeast. Preferably, said HAA1 allele comprises SEQ ID No. 5 as coding region. Preferably, said yeast is a xylose fermenting yeast. A xylose fermenting yeast, as used here, can be a yeast that is naturally producing ethanol on the base of xylose, or it can be a yeast that is mutated and/or genetically engineered to ferment xylose and to produce ethanol on the base of xylose. Even more preferably, said yeast is selected from the group consisting of Saccharomyces sp., Pichia sp., Candida sp., Pachysolen sp. and Spathaspora sp. Most preferably, said yeast is a Saccharomyces sp. preferably a Saccharomyces cerevisiae.
Another aspect of the invention is a yeast strain comprising SEQ ID No.1 in combination with SEQ ID No.3 and/or SEQ ID No.5 and/or SEQ ID No. 7. Preferably, said yeast is a xylose fermenting yeast. Even more preferably, said yeast is selected from the group consisting of Saccharomyces sp., Pichia sp., Candida sp., Pachysolen sp. and Spathaspora sp. Most preferably, said yeast is a Saccharomyces sp. preferably a Saccharomyces cerevisiae.
BRIEF DESCRIPTION OF THE FIGURES
Figure 1 . Fermentation profiles of the acetic acid susceptible segregant Ethanol Red 18 (A) and the acetic acid tolerant segregant 16D (B). Strains were inoculated in YPD medium with 2 % glucose at pH 4 and various concentrations of acetic acid; 0% (closed circle), 0.4% (closed square), 0.5% (closed upward triangle), 0.6% (closed downward triangle), 0.7% (closed diamond), 0.8% (open circle), 0.9% open square), 1 .0% (open triangle). Data points are the average of duplicate measurements, error bars represent the maximum deviation of the average.
Figure 2. QTL mapping of high acetic acid tolerance of pooled F1 segregants (green), pooled F7 segregants (red) and individual F7 segregants (black). Pooled F1 and pooled F7 segregants (27 segregants for both pools) were subjected to sequencing analysis utilizing the lllumina platform at BGI. Individual F7 segregants were sequenced with the lllumina platform at EMBL. Unselected pools consisting of 27 randomly selected segregants were also sequenced to eliminate linkage to inadvertently selected traits (BGI). P-values calculated using the individual sequencing data F7 segregants were plotted against the respective chromosomal position. P < 0.05 was considered statistically significant.
Figure 3. Fermentation profiles of the reciprocal hemizygosity analysis on HAA1. Fermentations were performed in YPD medium supplemented with 0.7% acetic acid at pH 4.0. Three diploid hybrid strains were tested and compared: ER18Ahaa1 x 16D (circle); ER18 x 16DAhaa1 (square); ER18 x 16D (triangle). Data points are the average of duplicate measurements. Error bars represent the maximum deviation of the average. Figure 4. Fermentation profiles of the reciprocal hemizygosity analysis on CUP2 (A), DOT5 (B) and VMA7 (C). Fermentations were performed in YPD medium supplemented with 0.7% acetic acid at pH 4.0. Three diploid hybrid strains were tested and compared: ER18A x 16D (circle); ER18 x 16DA (square); ER18 x 16D (triangle). Data points are the average of duplicate measurements. Error bars represent the maximum deviation of the average.
Figure 5. Fermentation profiles of the diploid strains GSE16-T18 and GSE16-T18-HAA1 * (carrying the unique HAA1 mutation of strain 16D in both HAA1 alleles). Fermentations were performed in YPD medium with 20% glucose and varying concentrations of acetic acid. Strains: GSE16-T18 (·) and GSE16-T18-HAA1 * (O). (A) No acetic acid; (B) 1 .0% acetic acid; (C) 1 .2% acetic acid; (D) 1 .4% acetic acid; (E) 1 .6% acetic acid; (F) 2.0% acetic acid.
Material & Methods to the examples
Strains and growth conditions
Strains and plasmids used in this study are shown in Table 1
ypt A_ aa
Yeast cells were grown in a shaking incubator at 30°C and 200 rpm in YPD medium containing 1 % (w/v) yeast extract, 2 % (w/v) Bacto peptone, and 2 % (w/v) D-glucose. Acetic acid medium was prepared by adding acetic acid to YPD medium, after which the pH was adjusted to 4.0 with HCI or KOH. Subsequently, the acetic acid medium was filtersterilized using a 0.2 μηη filter. Antibiotics were added as required in the following final concentrations: geneticin, 400 μg ml; nourseothricin, 150 μg ml; chloramphenicol, 100 μg ml. For solid medium, 1 .5 % Bacto agar was added to the YPD medium.
Fermentation experiments were performed in batch under near-anaerobic conditions in straight glass tubes containing 100 ml YPD medium and various concentrations of acetic acid. The culture was stirred continuously at 120 rpm using magnetic stirrers. Fermentations were inoculated with 5 ml of late-exponential phase yeast cells that were pre-cultured in YPD- medium (30 °C, static incubation). The progress of the fermentation was monitored by measuring the decrease in weight of the fermentation tube + yeast cell culture. During fermentation, glucose present in the YPD medium is fermented, producing C02 that is emitted from the fermentation tubes. The emission of C02 is reflected by the loss in weight of the fermentation tube. General molecular biology techniques
Genomic DNA was extracted with PCI [phenol-chloroform-isoamyl-alcohol (25:24:1 )] as described by Hoffman and Winston (1987). PCR reactions were performed with ExTaq (TAKARA) for diagnostic purposes or Phusion® High-Fidelity DNA polymerase (New England Biolabs) for sequencing purposes, both according to manufacturer's protocols. Yeast was transformed using the LiAc/PEG method described by Gietz et al 27. Gene deletions were made using a PCR-based strategy 28-29. After transformation, gene deletions were verified with PCR. DNA isolation for whole-genome sequencing analysis
The two parent strains ER18 and 16D and all segregants displaying high acetic acid tolerance were individually grown to stationary phase in 50 ml YPD-medium. Segregants were pooled, based on OD600, such that the number of cells from every segregant in the pool was equal. The genomic DNA was extracted according to
Johnston30. At least 3 μg of genomic DNA was provided to the Beijing Genomics Institute (BGI) for sequencing analysis using the lllumina HiSeq2000 platform. Paired-end short reads of -100 base pairs were generated for four samples (ER18, 16D, selected pool and unselected pool). The assembled sequences had an average coverage of 37X (ER18), 32X (16D) and 36X (selected and unselected pool). Mapping of the short read sequences, variant calling and QTL analysis were performed as described previously by Swinnen et al. 6 and by Hubmann et al. 7.
The SNP variant frequencies were calculated by dividing the number of the alternative variant by the total number of aligned reads. A very high or a very low frequency was indicative of a one-sided SNP segregation preferentially coming from one parent, indicating a genetic linkage to the trait of interest. Statistical confirmation of genetic linkage was obtained using EXPIoRA (Duitama et al., in preparation) or by methods described earlier (Swinnen et al., 2012).
Scoring SNPs by mismatch mutation PCR
Individual SNPs were scored by PCR using forward and reverse primers that differ only at the 3' terminal nucleotide, based on the DNA sequence of ER18 or 16D. The optimal annealing temperature was determined by gradient PCR using DNA of ER18 and 16D. The optimal temperature is the annealing temperature at which only hybridization with primers containing an exact match was observed. Mating, sporulation, tetrad analysis
Mating, sporulation and tetrad analysis were performed by standard procedures 31. The mating type of the segregants was determined by diagnostic PCR for the MAT locus (Huxley et al., 1990).
Inbreeding crosses were performed by random spore isolation, followed by mass mating. Random spore isolation was done by resuspending sporulating cells in 25 ml sterile MQ-water supplemented with 10 μg ml zymolyase, 10 μΙ β-mercaptoethanol and glass beads. This cell suspension was incubated overnight in a shaking incubator (200 rpm). The cell suspension was subsequently vortexed for 5 minutes, followed by harvesting the spores with centrifugation (5 minutes, 3000 rpm, Beckman centrifuge). The spores were resuspended in 10 ml Nonidet P-40 (1 .5 % (v/v)) and put on ice for 15 minutes. After cooling, the suspension was sonicated four times (Amplitude = 75 %, cycle = 1 ) for 30 seconds with two minute intervals. The suspension was washed three times with Nonidet P-40 and again sonicated four times. Spores were pelleted, resuspended in 300 μΙ MQ-water and plated in serial dilutions for single colonies. The remaining solution was plated on a single YPD-plate and incubated at 30 °C for two nights to allow mass mating of the isolated spores. Confirming involvement of mutated alleles in superior phenotype
Confirming the involvement of mutated alleles in the superior phenotype was done by reciprocal hemizygosity analysis (RHA) 6. For RHA, diploid strains were constructed by crossing ER18 and 16D wild type or derived deletion strains such that the hybrid diploid strain carried only one allele (either from ER18 or 16D) of the candidate gene. Subsequent fermentation experiments were performed with two individual isolates of the constructed diploids.
Example 1 : Screening for superior acetic acid tolerance
Ethanol Red is a diploid yeast strain that is being used for bio-ethanol production at high temperatures, showing ethanol yields of up to 18 %. However, the fermentation performance of this industrial yeast strain is severely affected by acetic acid, a weak organic acid present in high quantities in lignocellulosic hydrolysates. Haploid segregants were isolated from this yeast strain and scored on acetic acid tolerance by fermentation in YPD medium supplemented with various concentrations of acetic acid. It was observed that the maximum tolerance of Ethanol Red towards acetic acid was 0.6 % (v/v) in YPD medium at a pH of 4.0. However, the lag phase was significantly prolonged by adding acetic acid to the growth medium, with a lag phase of approximately 30 hours at concentrations of 0.5 % and 0.6 % (Figure 1 ). The haploid Ethanol Red segregant #18 (named ER18) showed similar tolerance to acetic acid and was therefore selected for further experiments.
In order to obtain a yeast strain with high acetic acid tolerance, the in-house yeast collection and the yeast collection from the Fungal Biodiversity Centre (CBS-KNAW, Utrecht, The Netherlands) were screened under acetic acid conditions. More than 1000 yeast strains were assessed, from which strain JT 22689 showed the best performance under fermentative conditions at high acetic acid concentrations, being able to ferment glucose in the presence of 0.9 % acetic acid without a lag phase (not shown). Also from this strain a haploid segregant, named 16D, could be isolated that showed a similar phenotype in terms of acetic acid tolerance.
Example 2: QTL mapping with pooled F1 segregants
Mapping the genetic determinants that are responsible for the high acetic acid tolerance of 16D was initiated by crossing the haploid segregants ER18 and 16D. The resulting hybrid strain was subsequently sporulated to obtain segregants that contain a mixture of the parental genomes. Obtained segregants were subsequently screened for high acetic acid tolerance, resulting in the identification of 27 (out of 288) segregants that were able to ferment glucose in the presence of 0.9 % acetic acid, which is comparable with the tolerance observed for the superior parent strain. These 27 segregants were therefore selected for pooled-segregant whole-genome sequencing analysis. Genomic DNA isolated from the two parent strains, a pool of the 27 selected segregants and a control pool of 27 randomly selected segregants was sent for custom sequencing analysis using the lllumina HiSeq2000 technology (BGI, Hong Kong, China). The sequence reads from parent strains ER18 and 16D were aligned with the reference sequence from strain S288C. A total number of 23,150 SNPs between ER18 and 16D could be identified, which were subsequently filtered according to the method described by Duitama et al. (2012). The SNP variant frequencies were calculated by dividing the number of the alternative variant by the total number of aligned reads. The calculated variant frequencies were subsequently plotted against the respective chromosomal positions. The underlying structure in the SNP variant frequencies scatterplot of a given chromosome was identified by fitting smoothing splines in the generalized linear mixed model framework, as described by Claesen et al. (2013). Variant frequencies that significantly deviate from 50% (random segregation) are indicative of genetic linkage to the phenotype.
The results from the QTL mapping (depicted in Fig. 2) show two loci on the genome with a strong linkage to the superior segregant 16D: QTL1 on chromosome XIII and a second QTL on chromosome XVI. The statistical significance of QTL1 was confirmed using the Hidden Markov Model described previously, stretching from position 181019 - 294166. Both QTLs were further investigated by scoring selected SNPs in the 27 individual segregants in order to precisely determine the SNP variant frequencies and the statistical significance of the genetic linkage. Using a binomial test previously described (Swinnen et al., 2012; Claessen et al., 2013), both loci were found to be statistically significant. Furthermore, the size of both QTLs could be decreased to regions stretching from roughly 224000 - 277000 for QTL1 on chromosome XIII, and 568000 - 615000 for QTL2 on chromosome XVI.
Example 3: Identification of causative genes by reciprocal hemizygosity analysis in F1 QTLs
For further analysis of the two identified QTLs, genes located within the linked regions were crosschecked with the Saccharomyces Genome Database. In QTL1 , none of the genes were previously shown to play an important role in acetic acid tolerance. However, in QTL2, the gene HAA1 was located, a transcriptional activator known to be involved in adaptation to weak acid stress (Mira et al., 2010b; Fernandez et al., 2005). This gene was therefore further tested using the reciprocal hemizygosity analysis (RHA) method (Steinmetz et al., 2002). We constructed hemizygous diploid ER18/16D hybrid strains which contained a single copy of the superior or the inferior allele of HAA1, while the other copy of the gene was deleted. When these RHA strains were tested in fermentations with acetic acid, a clear difference was observed between the reciprocal strains. The HAA1 allele from the superior parent strain 16D sustained a much faster fermentation rate than the allele from the inferior parent strain ER18 (Fig. 3). In addition, a strain was constructed by replacing the whole HAA1 allele (promoter + ORF + terminator) from ER18 with the allele from 16D. Fermentations performed with this strain in the presence of several concentrations of acetic acid showed that the acetic acid tolerance was improved (results not shown). These results identify HAA1 as a causative allele in QTL2.
Example 4: QTL mapping with pooled F7 segregants
It was observed that the two QTLs identified with the F1 pool were relatively large in size. Therefore, we attempted to narrow the QTLs by inbreeding F1 segregants obtained after sporulating the ER18/16D hybrid strain. Sporulating cells were harvested and haploid segregants were isolated using the random spore analysis method, after which the segregants were subjected to mass mating. A random amount of hybrid cells were subsequently sporulated again and the segregants isolated. This cycle of sporulation, isolating haploid segregants and mass mating was repeated six times, eventually resulting in the isolation of seventh generation (F7) segregants. These F7 segregants were screened for high acetic acid tolerance under the same conditions used for screening the F1 segregants. Out of 768 segregants assessed, 66 segregants showed good fermentation in the presence of 0.9% acetic acid. However, for statistical reasons, it was decided to construct a pool of similar size as the F1 pool to perform the pooled-segregant whole-genome sequencing analysis.
After analyzing and mapping the sequencing data, the data was compared with the mapping results obtained with the pool of F1 segregants (Fig. 2). As expected, the size of QTLs can be narrowed by increasing the recombination frequency through inbreeding crosses. The size of QTL1 on chromosome XIII obtained with F7 segregants was approximately 30 kb (position 247466 - 277019), which is 83 kb smaller than QTL1 obtained with F1 segregants. Moreover, the number of genes in the top of the locus significantly decreased compared to the F1 pool.
Surprisingly, increasing the recombination frequency also resulted in a number of unexpected outcomes. The second QTL identified with F1 segregants (chr. XVI) was no longer linked to the superior phenotype, as indicated by the decreased SNP variant frequencies calculated for this genomic region (Fig. 2). This indicates that the F7 segregants do not longer rely on the superior HAA1 gene from 16D to acquire higher acetic acid tolerance levels. It could therefore be expected that other alleles from 16D are causative for the superior phenotype. This notion was strengthened by the appearance of new peaks in the genetic mapping. Three new putative QTLs were identified, located on chromosome VII (QTL3), chromosome IX (QTL4) and chromosome X (QTL5).
Example 5: QTL mapping with individually sequenced F7 segregants
In an attempt to further enhance the resolution of QTL mapping, we sequenced the 27 segregants from the F7 pool individually. Genomic DNA samples were sent to the Genomic Core Facility of EMBL (Heidelberg, Germany) and the sequencing data was treated with the previously used scripts that were modified for this purpose (Duitama et al, unpublished). The main advantage of this approach is that the whole genome sequences of the 27 segregants can be compared with each other and that SNP variant frequencies can be calculated using the whole-genome sequences of the individual segregants, instead of calculating the frequencies from sequencing reads. Furthermore, by aligning the 27 whole-genome sequences we could score all SNPs in all single segregants compared to the inferior parent ER18 and calculate the statistical significance of every single SNP, using the binomial test described previously (Swinnen et al., 2012 Claessen et al., 2013).
Figure 2 shows that the same QTLs could be identified using sequencing data from either the pooled segregants or the individual sequences, indicating that sequencing individual segregants yields comparable genetic maps with sequencing pooled segregants. However, additional information could be gained from the calculated p-values. SNPs are considered statistically significant if the p-value is lower than 0.05, and by combining the genetic mapping with the calculated p-values, a number of regions could be pinpointed that might contain the causative genes for acetic acid tolerance. Using this approach, we were able to find five regions on the genome that were statistically linked: QTL1 on chromosome XIII (position 261255 - 271498), QTL3 on chromosome VII (position 471 171 - 554980), QTL4 on chromosome IX (position 335344 - 340345) and QTL5 on chromosome X (position 394527 - 451436). In addition, a new QTL could be identified on chromosome VII, stretching from position 107986 - 195096, making a total of 6 QTLs that contain potential superior alleles for acetic acid tolerance. Example 6: Identification of causative genes by RHA in F7 QTLs
The identification of the causative genes located in the QTLs found with F7 segregants was performed similarly to the approach described for the F1 segregants. A number of candidate genes were selected based on their statistical linkage to the phenotype predicted by the p- values and their previously predicted role in acetic acid tolerance. The selected candidate genes were: PMA1 and VMA7 (QTL3), DOT5 (QTL4), RAV1 and VTC4 (QTL5), TOS3 and CUP2 (QTL6). These genes were subsequently tested by the reciprocal hemizygosity analysis method. After constructing the necessary strains, fermentation experiments were performed to assess the effect of the candidate genes on high acetic acid tolerance. The results of these fermentations, shown in figure 4, indicated that VMA7, DOT5 and CUP2 play an important role in the high acetic acid tolerance of segregant 16D. Furthermore, these results indicate that the combination of mapping SNPs and calculating the statistical significance of every SIMP, using the whole-genome sequences of the individual segregants, strongly improves the resolution of QTL mapping. Example 7: Natural occurrence of mutations in the causative genes
The sequences of the identified causative genes were compared with 28 strains from which the whole-genome sequences have been published. Only mutations between strain ER18 and strain 16D were considered; additional mutations between the strains examined were left out of the comparison. The results, summarized in Table 2, show that most mutations are not uncommon to the other yeast strains. Within the open reading frame of VMA7, no differences could be found between the sequences of ER18 and 16D. Therefore, the promoter sequences of the gene were investigated. Multiple mutations were identified, all of which could be found in the other strains examined.
The mutations identified in the ORFs of HAA1, DOT5 and CUP2 are commonly found the other strains. However, the mutation at position 574534 in HAA1 from 16D was not found in the other strains and may therefore be a novel and rather unique mutation. To confirm that this mutation is a causative allele in conferring high acetic acid tolerance, this point mutation was introduced into the two copies of the HAA1 allele of the industrial strain GSE16-T18 (which has the HAA1 allele of Ethanol Red). Fermentations were performed with YP + 20% glucose in the presence of 1 .0, 1 .2, 1 .4, 1 .6 and 2.0% acetic acid at pH 5.2. In the absence of acetic acid there was no difference in the fermentation performance, while in the presence of whatever acetic acid concentration the performance of the GSE16-T18 HAA1 * strain was consistently better than that of the GSE16-T18 strain. Especially the lag phase was strongly reduced by the HAA1 * mutation, while the actual fermentation rate was not much affected. See Figure 5.
Acetic acid is one of the major inhibitors in lignocellulose hydrolysates used for the production of second-generation bioethanol. Although several genes have been identified in laboratory yeast strains that are required for tolerance to acetic acid, the genetic basis of the high acetic acid tolerance naturally present in some Saccharomyces cerevisiae strains is unknown. Identification of its polygenic basis may allow improvement of the acetic acid tolerance of yeast strains used for second-generation bioethanol production by precise genome editing, minimizing the risk of negatively affecting other industrially-important properties of the yeast. Haploid segregants of a strain with unusually high acetic acid tolerance and a reference industrial strain were used as superior and inferior parent strain, respectively. After crossing of the parent strains, QTL mapping, using the SNP variant frequency determined by pooled- segregant whole-genome sequence analysis, revealed two major QTLs. All F1 segregants were then submitted to multiple rounds of random inbreeding and the superior F7 segregants were submitted to the same analysis, further refined by sequencing of individual segregants and bioinformatics analysis taking into account the relative acetic acid tolerance of the segregants. This resulted in disappearance in the QTL mapping with the F7 segregants of a major F1 QTL, in which we identified HAA1, a known regulator of high acetic acid tolerance, as a true causative allele. Novel genes determining high acetic acid tolerance, DOT5, CUP2, and a previously identified component, VMA7, were identified as causative alleles in the second major F1 QTL and in three newly appearing F7 QTLs, respectively. The superior HAA1 allele contained a single point mutation that was able by itself to improve acetic acid tolerance when inserted into an industrial yeast strain.
This work reveals the polygenic basis of high acetic acid tolerance in S. cerevisiae in unprecedented detail. It also shows for the first time that a single strain can harbour different sets of causative genes able to establish the same polygenic trait. The superior alleles identified can be used for improvement of acetic acid tolerance in industrial yeast strains. Table 2. Occurrence of SNPs in the identified genes HAA 1, VMA7, DOT5 and CUP2.
ORF ORF ORF ORF ORF ORF ORF ORF
573259 573609 57371 1 574042 574109 574147 574276 574534
ER18 T G A G TC T c G
16D C A G A AT C A A
S288C C A G A AT C A G
AWRI 1631 c A G A AT c A G
AWRI796 c A G A AT A G
BY4741 c A G A AT c A G
BY4742 c A G A AT c A G
CBS7960 c A G A AT c A G
CEN.PK1 13 c A G A AT c A G
CLI B215 c A G A AT c A G
EC11 18 c A G Δ AT c A G
EC9-8 c A G A AT c A G
FL100 c A G A AT c A G
FostersB - - - R WY Y M G
FostersO ¾3lf:'¾¾! A TC c C G
JAY291 c G G A TC c A G
Kyokai7 T G A G TC T C G
LalvinQA23 - - - - WY c A G
PW5 c G A A TC T C G
R l l-l a c A G A AT c A G
Sigma 1278b c A G A AT c A G
T7 T G A G TC T C G
UC5 c G A G TC T G
VL3 - A G A AT c A G
Vinl3 c R G A WY c A G
W303 c A G A AT c A G
YJM269 c G A G TC T C G
YJ 789 c G A G TC T C G
YPS163 c G A G TC T 5¾J¾ §S¾I G
ZTW1 T G A A TC T c G
VMA7 DOT5 CUP2
Prom. Prom. Prom. Prom. Prom. ORF ORF ORF ORF
527390 527396 527439 527467 527618 334506 335344 191489 191625
ER18 G C G T G G c G T
16D T T C A A A T A C
S288C T T C A A A c A c
AWRI 1631 T T A A A T A. c
AWRI796 T T c A A A c A c
BY4741 T T c A A A c A c
BY4742 T T c A A A T A c
CBS7960 T T c A. A A c A c
CEN.PK1 13 T T c A A A T A c
CLIB215 T T c A A A T A c
EC1 1 18 T T c A A A T A c
EC9-8 T T .A A. A c A c
FL100 T T c A A A c A c
FostersB T T A A A R c
FostersO T T c A A A T R c
JAY291 T T c A A. A T A c
Kyokai7 G " c G T G c G T LalvinQA23 C T T Λ A A Λ C
PW5 T c G A A A C G T
RMl l -l a T T A A A. T A c
Sigma 1278b T T c A A G c A c
T7 T c G A A A c G T
UC5 G c G T T A T G T
VL3 T T c A G G c A c
Vin l 3 T T c A A A T A c
VV303 T T c A A A T A c
YJ 269 G c G A A A c G T
YJM789 T T C A A A A c
Y PS 163 T c G A T A T G T
ZTW1 G c G T G G c G T
The SNPs present in 16D compared to ER18 were checked in 28 strains of which the whole genome sequence has been published. SNPs present in the other strains compared to ER18, but not in 16D, are not presented.
• DOT5: SEQ ID No. 1/SEQ ID No. 2 · CUP2: SEQ ID No. 3/SEQ ID No. 4
• HAA1 : SEQ ID No. 5/SEQ ID No. 6
• VMA7: SEQ ID No. 7/SEQ ID No. 8
• YPT7: SEQ ID No. 9/SEQ ID No. 10
Abbott, D.A. et al. (2007). Generic and specific transcriptional responses to different weak organic acids in anaerobic chemostat cultures of Saccharomyces cerevisiae. Ferns Yeast Research 7, 819-833.
Almeida, J.R.M. et al. (2007). Increased tolerance and conversion of inhibitors in lignocellulosic hydrolysates by Saccharomyces cerevisiae. Journal og Chemical Technology and Biotechnology 82, 340-349.
Baudin, A., Ozier-Kalogeropoulos, O., Denouel, A., Lacroute, F. & Cullin, C. (1993). A simple and efficient method for direct gene deletion in Saccharomyces cerevisiae. Nucleic Acids Res 21 , 3329-3330.
Bellissimi, E., van Dijken, J. P., Pronk, J.T. & van Maris, A.J.A. (2009). Effects of acetic acid on the kinetics of xylose fermentation by an engineered, xylose-isomerase-based Saccharomyces cerevisiae strain. Ferns Yeast Research 9, 358-364.
Casey, E., Sedlak, M., Ho, N.W. & Mosier, N.S. (2010). Effect of acetic acid and pH on the cofermentation of glucose and xylose to ethanol by a genetically engineered strain of Saccharomyces cerevisiae. Ferns Yeast Research 10, 385-393.
Claesen, J., Clement, L, Shkedy, Z., Foulquie-Moreno, M.R. & Burzykowski, T. (2013). Simultaneous mapping of multiple gene Loci with pooled segregants. PLoS One 8, e55133.
Deutschbauer, A.M. & Davis, R.W. (2005). Quantitative trait loci mapped to single- nucleotide resolution in yeast. Nat Genet 37, 1333-1340.
Duitama, J. et al. (2012). Fosmid-based whole genome haplotyping of a HapMap trio child: evaluation of Single Individual Haplotyping techniques. Nucleic Acids Res 40, 2041 -53. Ehrenreich, I.M. et al. (2010). Dissection of genetically complex traits with extremely large pools of yeast segregants. Nature 464, 1039-1042.
Fernandes, A.R., Mira, N.P., Vargas, R.C., Canelhas, I. & Sa-Correia, I. (2005). Saccharomyces cerevisiae adaptation to weak acids involves the transcription factor Haal p and Haal p-regulated genes. Biochem Biophys Res Commun 337, 95-103.
Flint, J. & Mott, R. (2001 ). Finding the molecular basis of quantitative traits: successes and pitfalls. Nat Rev Genet 2, 437-445.
Gietz, R.D., Schiestl, R.H., Willems, A.R. & Woods, R.A. (1995). Studies on the transformation of intact yeast cells by the LiAc/SS-DNA PEG procedure. Yeast 1 1 , 355- 360.
Glazier, A.M., Nadeau, J.H. & Aitman, T.J. (2002). Finding genes that underlie complex traits. Science 298, 2345-2349. Hasunuma, T. et al. (201 1 ). Metabolic pathway engineering based on metabolomics confers acetic and formic acid tolerance to a recombinant xylose-fermenting strain of Saccharomyces cerevisiae. Microb Cell Fact 10, 2.
Hoffman, C.S. & Winston, F. (1987).A ten-minute DNA preparation from yeast efficiently releases autonomous plasmids for transformation of Escherichia coli. Gene 57, 267-272.
Huang, H. et al. (201 1 ). Identification of crucial yeast inhibitors in bio-ethanol and improvement of fermentation at high pH and high total solids. Bioresour Technol 102, 7486-7493.
Hubmann, G. et al. (2013). Quantitative trait analysis of yeast biodiversity yields novel gene tools for metabolic engineering. Metab Eng 17, 68-81 .
Huxley, C, Green, E.D. & Dunham, I. (1990). Rapid assessment of S. cerevisiae mating type by PCR. Trends Genet 6, 236.
Johnston, J.R. (1994). Molecular Genetics of Yeast: A Practical Approach, (Oxford University Press Inc., New York.
- Koppram, R., Albers, E. & Olsson, L. (2012). Evolutionary engineering strategies to enhance tolerance of xylose utilizing recombinant yeast to inhibitors derived from spruce biomass. Biotechnol Biofuels 5, 32.
Li, B.Z. & Yuan, Y.J. (2010).Transcriptome shifts in response to furfural and acetic acid in Saccharomyces cerevisiae. Appl Microbiol Biotechnol 86, 1915-1924.
- Lumtong, S., Sumpradit, T., Kitpreechavanich, V., Tuntirungkij, M., Seki, T. and Yoshida, T.
(2000). Effect of Acetic acid on growth and ethanol fermentation of xylose fermenting yeasts and Saccharomyces cerevisiae. Kastaert J. (Nat. Sci) 34, 64-73.
Matsushika, A. & Sawayama, S. (2012). Characterization of a Recombinant Flocculent Saccharomyces cerevisiae Strain That Co-Ferments Glucose and Xylose: II. Influence of pH and Acetic Acid on Ethanol Production. Appl Biochem Biotechnol.
Mira, N.P., Palma, M., Guerreiro, J.F. & Sa-Correia, I. (2010a). Genome-wide identification of Saccharomyces cerevisiae genes required for tolerance to acetic acid. Microb Cell Fact 9, 79-91 .
Mira, N.P., Becker, J.D. & Sa-Correia, I. (2010b). Genomic expression program involving the Haal p-regulon in Saccharomyces cerevisiae response to acetic acid. OMICS 14, 587- 601 .
- Narendranath, N.V., Thomas, K.C. & Ingledew, W.M. (2001 ). Effects of acetic acid and lactic acid on the growth of Saccharomyces cerevisiae in a minimal medium. J Ind Microbiol Biotechnol 26, 171 -177.
- Olsson, L. and Hahn-Hagerdal, B. (1993). Fermentative performance of bacteria and yeast in lignocellulose hydrolysates. Process Biochem. 28, 249-257. Parts, L. et al. (201 1 ). Revealing the genetic structure of a trait by sequencing a population under selection. Genome Res.
Sherman, F. & Hicks, J. (1991 ). Micromanipulation and dissection of asci. Methods
Enzymol 194, 21 -37.
- Steinmetz, L.M. et al. (2002). Dissecting the architecture of a quantitative trait locus in yeast. Nature 416, 326-330.
Swinnen, S. et al. (2012). Identification of novel causative genes determining the complex trait of high ethanol tolerance in yeast using pooled-segregant whole-genome sequence analysis. Genome Res 22, 975-984.
- Taherzadeh, M.J., Niklasson, C. & Liden, G. (1997). Acetic acid - friend of foe in anaerobic batch conversion of glucose to ethanol by Saccharomyces cerevisiae. Chemical Engineering Science 52, 2653-2659.
Tanaka, K., Ishii, Y., Ogawa, J. And Shima, J. (2012). Enhancement of acetic acid tolerance in Saccharomyces cerevisiae by overexpression of the HAA1 gene, encoding a transcription factor. Appl. Environ. Microbiol. 78, 8161 -8163.
Wach, A. (1996). PCR-synthesis of marker cassettes with long flanking homology regions for gene disruptions in S. cerevisiae. Yeast 12, 259-265.
Wright, J. et al. (201 1 ). Batch and continuous culture-based selection strategies for acetic acid tolerance in xylose-fermenting Saccharomyces cerevisiae. Ferns Yeast Research. - Zhang, J.G. et al. (201 1 ). Improvement of acetic acid tolerance and fermentation performance of Saccharomyces cerevisiae by disruption of the FPS1 aquaglyceroporin gene. Biotechnol Lett 33, 277-284.
1 . The use of DOT5 for obtaining acetic acid tolerance in yeast.
2. The use of DOT5 according to claim 1 , wherein said DOT5 is an allele encoding SEQ ID No.2
3. The use of DOT5 according to claim 1 or 2, wherein said use is combined with the use of CUP2 and/or HAA1.
4. The use of DOT5 according to claim 3, wherein CUP2 is an allele encoding SEQ ID No.4.
5. The use of DOT5 according to claim 3, wherein HAA 1 is an allele encoding SEQ ID No.6.
6. The use of a HAA1 allele encoding SEQ ID No. 6 for obtaining acetic acid tolerance in yeast.
7. The use of a HAA1 allele according to claim 6, wherein said allele comprises SEQ ID No. 5.
8. The use according to any of the preceding claims, wherein said use is an overexpression.
9. The use according to any of the preceding claims, wherein said yeast is a Saccharomyces sp.
10. A yeast strain, comprising SEQ ID No.1 , in combination with SEQ ID No.3 and/or SEQ ID No.5.
1 1 . The yeast strain according to claim 10, wherein said yeast is a Saccharomyces sp.