Downregulation of EHT1 and EEB1 in Saccharomyces cerevisiae Alters the Ester Profile of Wine during Fermentation

EHT1 and EEB1 are the key Saccharomyces cerevisiae genes involved in the synthesis of ethyl esters during wine fermentation. We constructed single (Δeht1, Δeeb1) and double (Δeht1Δeeb1) heterogenous mutant strains of the industrial diploid wine yeast EC1118 by disrupting one allele of EHT1 and/or EEB1. In addition, the aromatic profile of wine produced during fermentation of simulated grape juice by these mutant strains was also analyzed. The expression levels of EHT1 and/or EEB1 in the relevant mutants were less than 50% of the wild-type strain when grown in YPD medium and simulated grape juice medium. Compared to the wild-type strain, all mutants produced lower amounts of ethyl esters in the fermented grape juice and also resulted in distinct ethyl ester profiles. ATF2, a gene involved in acetate ester synthesis, was expressed at higher levels in the EEB1 downregulation mutants compared to the wild-type and Δeht1 strains during fermentation, which was consistent with the content of acetate esters. In addition, the production of higher alcohols was also markedly affected by the decrease in EEB1 levels. Compared to EHT1, EEB1 downregulation had a greater impact on the production of acetate esters and higher alcohols, suggesting that controlling EEB1 expression could be an effective means to regulate the content of these aromatic metabolites in wine. Taken together, the synthesis of ethyl esters can be decreased by deleting one allele of EHT1 and EEB1 in the diploid EC1118 strain, which may modify the ester profile of wine more subtly compared to the complete deletion of target genes.

of wine aroma, superfluous higher alcohols have a negative impact on the taste [16]. For instance, acetate esters are frequently produced during alcoholic fermentation and impart fruity notes [5,17]. They are synthesized through acetyl transfer from acetyl-CoA to alcohol by the acetyltransferases ATF1 and ATF2 [4,18], and the deletion and overexpression of ATFs only affect acetate ester synthesis and not that of ethyl esters [4,19]. The interaction of MCFA ethyl esters and the acetate ester synthesis pathway remains unclear.
The genetic background of the yeast strains, fermentation substrate and environmental conditions markedly influence the final aroma profile of wine. Most studies have used haploid yeast strains to analyze the impact of their genes on fermentation and wine characteristics. However, the industrial strains of yeast are typically diploid, and produce esters and other flavor compounds more efficiently than laboratory haploid strains [11,13]. Therefore, we generated EHT1 or/and EEB1 deletion mutants of the industrial wine yeast EC1118, one of the most widely used strains for wine production globally, through homologous recombination. The wild-type and mutant strains were then used to ferment simulated grape juice to assess the influence of EHT1 and EEB1 on the production of esters during fermentation.

Construction of EHT1 and EEB1 Deletion Cassettes
Plasmid DNA was prepared using a FastPure Plasmid Mini Kit (Vazyme Biotech, China) and High Fidelity DNA polymerase (Vazyme Biotech) was used for PCR. The Cre-loxP system was used to inactivate one allele of the target locus with the kanMX marker [21,22]. The loxP-kanMX-loxP fragment was amplified from the pUG6 plasmid using the loxP-KanMXloxP-F1 and loxP-KanMX-loxP-R1 (EHT1) or the loxP-KanMX-loxP-F2 and loxP-KanMX-loxP-R2 (EEB1) primers. To construct the deletion cassette for EHT1 (EA1KEB1), its 5' and 3' flanking regions were amplified using the EA1-F/EA1-R and EB1-F/EB1-R primer pairs respectively and fused with loxP-kanMX-loxP by homologous recombination using the ClonExpress Entry One Step Cloning Kit (Vazyme Biotech). The EA2KEB2 deletion cassette was similarly constructed using the EA2-F/ EA2-R and EB2-F/ EB2-R primers that amplified the 5' and 3' flanking regions of EEB1, and the amplified fragment was fused with loxP-kanMX-loxP. The double deletion strain, Δeht1Δeeb1, was constructed from the Δeeb1strain by the above method. The deletion cassettes were verified by sequencing, and the primers used for cloning are listed in Table S1.

Yeast Transformation and Screening
The deletion cassettes were transformed into the diploid EC1118 cells by the LiAc-PEG method [20], and the transformants were selected on YPD plates containing 400 μg/l G418 (Sigma) for 2-3 days at 30 o C. The Δeht1 mutants with the EA1KEB1 deletion cassette were verified by PCR using EA1-F and EB1-R primers, and the EA2-F and EB2-R primers were used to identify the Δeeb1 strains with the EA1KEB2 deletion cassette. The double mutants were verified likewise. The mutant strains were then transformed with the Cre-expressing pSH65 plasmid, and Cre recombinase was induced on YP+2% galactose to remove the kanMX label.

Microbial Strains, Plasmids, Media and Culture Conditions
The yeast and bacterial strains and plasmids used in this study are listed in Table S2. The yeast cells were inoculated into YPD medium (10 g/l yeast extract, 20 g/l peptone and 20 g/l glucose) and cultured at 30 o C with constant shaking at 200 rpm. The galactose-containing yeast peptone glucose (YPG) medium was used to induce the expression of Cre enzyme. The kanamycin and zeocin-resistant transformants were screened using G418 (400 μg/l) and Zeocin (2,500 μg/l) (Solarbio, China), respectively. Escherichia coli DH5α was grown in Luria-Bertani (LB) medium (5 g/l yeast extract, 10 g/l tryptone and 10 g/l NaCl) at 37 o C with constant shaking at 200 ×g. The clones expressing the gene deletion cassettes were selected on 1.5% agar media using kanamycin (50 μg/l) (Solarbio).

Real-Time Quantitative PCR
Total RNA was extracted from the yeast cells using the Yeast RNAiso Kit (Takara, China), and transformed into cDNA using a reverse transcription kit (Vazyme Biotech). Quantitative PCR (qPCR) was performed on a realtime PCR system (Bio-Rad iQ5) using the SYBR Green Kit (Vazyme Biotech). The thermal cycling conditions were as follows: 95 o C for 2 min followed by 40 cycles of 95 o C for 15 s and 56 o C for 10 s, and a melting cycle from 65 o C to 95 o C. The housekeeping gene actin was used as the internal control, and the transcript levels were quantified using the 2 −ΔCT formula, where ΔCT is the number of PCR cycles required for the log phase of amplification for the test gene minus that of actin [23]. The primer sequences are listed in Table S1.

Growth Curve and Fermentation
The yeast cells were pre-cultured in 10 ml YPD medium at 30 o C for 24 h. The OD 600 was measured, and the cells were respectively inoculated into fresh YPD medium and chemically defined grape (CDG) medium at the density equivalent of 0.1 OD 600 . The cells inoculated in the YPD medium were cultured at 30 o C and the cell density was measured every 2 h. The CDG medium was fermented at 25 o C for 12 days, and the loss of CO 2 was measured every 12 h. The concentrations of residual sugars, alcohol and total acids were determined at the end of the fermentation process.

Gas Chromatography-Mass Spectrometry
The fermented samples were analyzed qualitatively and quantitatively by gas chromatography-mass spectrometry (GC-MS) as described previously with some modifications [24]. Briefly, each sample was subjected to a topical solid-phase microextraction method using a 50/30 μm DVB/CAR/PDMS extracted head, following which a 10-ml sample, 1 g NaCl and 30 μl internal standard 4-methyl-2-pentanol (2 g/l) were mixed in a bottle.
After pre-heating at 40 o C for 10 min, the samples were agitated at 500 rpm and extracted at the requisite temperature for 60 min, followed by high-temperature resolution for 8 min. The chromatographic column temperature program was as follows: 40°C for 8 min, steady increase till 70°C with increments of 5°C per minute, hold at 70°C for 3 min, steady increase till 230°C at the rate of 5°C per minute, and hold at 2 min. For EI ion bombardment, the electron energy was 70eV and the ion source temperature was 200°C on full-scan mode. The results were preliminarily retrieved and analyzed using the computer spectrum library (NIST14). The standard internal method was used for quantitative analysis.

Statistical Analysis
All data were expressed as the mean ± standard deviation (SD) of three replicates. One-way analysis of variance (ANOVA) and LSD test (p < 0.05) were used to test for significant differences between the mean values. The SPSS19.0 software (SPSS Inc., USA) was used for all statistical analyses.

Relative Expression Levels of EHT1 and EEB1 in the EC1118 Mutants
We generated the Δeht1, Δeeb1 and Δeht1Δeeb1 mutant strains of the wine yeast EC1118 by replacing EHT1 and EEB1 genes with the homologous deletion cassettes EA1KEB1 and EA2KEB2, respectively. Given that EC1118 is a diploid strain and contains two alleles of EHT1and EEB1, the mutant strains were obtained by disrupting one allele of each target locus. To assess the effect of gene deletion, the relative expression levels of EHT1 and EEB1 were detected in the mutant strains grown in YPD medium at 30°C for 12 h. As shown in Fig. 1, EHT1 levels dropped by 59 and 66% in the Δeht1 and Δeht1Δeeb1 strains, respectively (Fig.1A), and that of EEB1 was reduced by 51 and 58% in the Δeeb1 and Δeht1Δeeb1 strains, respectively, compared to the wild type cells (Fig. 1B). Therefore, the mutants still retained ~50% of the expression level of target genes. Interestingly, neither EHT1 nor EEB1 was significantly upregulated in the Δeeb1 and Δeht1 strains, respectively (p > 0.05), suggesting that neither gene compensates for the loss-of-function of the other. Therefore, we hypothesized that despite incomplete deletion of EHT1 and EEB1, ethyl ester synthesis can still be downregulated in the relevant mutants, resulting in an altered ester profile.

Fermentation Characteristics of the Mutant Strains
The fermentation ability of the Δeht1, Δeeb1 and Δeht1Δeeb1 strains was assessed in terms of their growth curves and loss of CO 2 . The respective strains were inoculated into CDG medium, i.e., simulated grape juice medium with a nutrient composition similar to that of grape juice [25], which was supplemented with 100 mg/l yeast available nitrogen (YAN), 200 g/l of an equimolar mixture of glucose and fructose, and the pH was adjusted  to 3.5. Fermentation was performed in triplicates over a period of 12 days. As shown in Fig. 2, the lag phase of the three mutants was longer than that of wild-type EC1118, although the growth rate, final cell mass and CO 2 loss were similar. In addition, there were no significant differences in the contents of residual sugar, alcohol and total acids in the simulated grape juice after 12 days of fermentation by the mutant and control strains (Table S3). Taken together, deletion of one EHT1 or EEB1 allele did not significantly affect the fermentation ability of the mutant strains.

Content of Esters and Higher Alcohols and the Expression Profiles of Relevant Genes During Fermentation
The content of esters and higher alcohols in the CDG medium was also measured on day 12 of fermentation with mutants and control strains (Table 1). In addition, we also analyzed the expression profiles of EHT1, EEB1, and of ATF1 and ATF2, in the three mutants at different time points during fermentation (Fig. 3). The results are discussed in the following sections.  Ethyl Esters. MCFA ethyl ester biosynthesis in S. cerevisiae is catalyzed by the EEB1 and EHT1-encoded Oacyltransferases [10]. We observed a decrease in the production of ethyl esters by all three mutants compared to the wild-type strain (Table 1). Furthermore, EEB1 expression was consistently lower in the Δeeb1 and Δeht1Δeeb1 strains compared to that in the control and Δeht1 strains during fermentation, and decreased in all strains after day 4. Similar trends were seen with EHT1 as well. However, studies show that EEB1 plays the dominant role in ethyl ester production [4]. Consistent with this, the amount of ethyl esters was reduced by ~36% in both the Δeeb1 and Δeht1Δeeb1 strains, as opposed to only 29% in the Δeht1 strain (Table 1) compared to the wild-type strain. Nevertheless, the decrease in EHT1 was slightly more in the Δeht1 strain compared to that of EEB1 in the Δeeb1 strain (Figs. 3A and 3B). In addition, ethyl ester production did not drop further in the double mutant compared to that in the Δeeb1 strain, which raises the possibility of a limited compensatory effect of other minor alcohol acyltransferases such as MR210w [10] during simultaneous downregulation of EEB1 and EHT1.
Ethyl hexanoate levels were the lowest in the Δeeb1 strain and showed a 52% reduction compared to those in the wild-type strain (Table 1), which is consistent with a previous finding that the maximum expression level of EEB1 correlates significantly with ethyl hexanoate concentration (Fig. 3) [13,26]. Since EHT1 catalyzes ethyl ester synthesis as well as hydrolysis, some studies on EHT1-overexpressing yeast strains proposed a strong negative correlation between EHT1 expression and the final concentrations of ethyl octanoate and decanoate [24,26]. In our study, however, EHT1 expression was negatively correlated to only the final concentration of ethyl decanoate in the mutant strains (Table 1, Fig. 3). In line with the EHT1 expression level in the Δeeb1 and Δeht1 strains, the content of ethyl octanoate was slightly higher in the former (Table 1). This indicates that when EHT1 is expressed at lower levels, the content of ethyl octanoate may be largely determined by the esterase and not the synthesis activity of EHT1.
Acetate esters. Acetate esters are an important group of wine esters that are synthesized by alcohol acetyltransferases from acetate and alcohols. In S. cerevisiae cells, alcohol acetyltransferases are encoded by the ATF genes [17,27]. Three acetate esters were detected in this study, including ethyl acetate, isoamyl acetate and phenethyl acetate (Table 1). Compared to the wild-type strain, the content of acetate esters did not change significantly in Δeht1, but showed a slight increase in the Δeeb1 and Δeht1Δeeb1 strains. We also analyzed the expression levels of ATFs during fermentation, and found that irrespective of the strain, both ATF1 and ATF2 peaked on day 4 and decreased thereafter. However, ATF2 expression levels were significantly higher in the Δeeb1 and Δeht1Δeeb1 strains compared to that in wild-type and Δeht1 strains on day 4, whereas ATF1 was expressed at similar levels in all strains at the same time point (Fig. 3C). A previous study showed that the deletion of ATF2 resulted in only a minor decrease in acetate ester production as opposed to that of ATF1 [19]. Likewise, we found that the upregulation of ATF2 in the Δeeb1 and Δeht1Δeeb1 strains caused only a slight increase in the content of acetate esters (Table 1, Fig. 3D). It is worth noting that ATF2 expression may be affected by EEB1 across the different strains grown under the same culture conditions, although the underlying mechanisms are unclear. In contrast, deletion or overexpression of ATFs does not affect ethyl ester synthesis [19]. Altogether, we proposed that in contrast to EHT1, regulating EEB1 expression may modify the profile of ethyl esters and acetate esters more effectively during fermentation.
Higher alcohols. Higher alcohols are volatile compounds that are mainly derived from amino acid degradation via the Ehrlich pathway during fermentation [14,28]. They are the precursors of acetate esters, and therefore influence the ester profile of the fermented product [29,30]. The Δeeb1 and Δeht1Δeeb1 strains significantly increased the content of higher alcohols (Table 1), indicating that EEB1 activity influences their biosynthetic pathway. Given that MCFA ethyl esters are the product of ethanol O-acyltransferase-catalyzed condensation of MCFA acyl-CoA and ethanol, downregulation of EEB1 results in the accumulation of the precursor MCFAs (Fig. 4). The MCFAs are then fed into the elongation reaction pathway to form complete long-chain fatty acids (FAs) in order to alleviate their toxicity in yeast cells [4,31]. The FAs can also upregulate the genes encoding for transaminases (BAT1 and BAT2) and decarboxylases (PDC1, PDC5, PDC6, and ARO10), which increases the production of higher alcohols via the Ehrlich pathway [14,24]. On the other hand, during fermentation under limited oxygen levels, the long-chain saturated FAs accumulate and inhibit acetyl-CoA carboxylase, which blocks fatty acid synthesis [32,33]. In addition, the TCA cycle cannot be (fully) used under anaerobic conditions [4,34] (Fig. 4), which renders all regular acetyl-CoA-consuming pathways inactive and leads to the accumulation of acetyl-CoA. Since EEB1 downregulation led to an increase in ATF2 expression level (see previous section), we surmise that the excess amount of acetyl-CoA or/and higher alcohols produced during fermentation by the Δeeb1 and Δeht1Δeeb1 strains induce ATF2 to produce more ethyl acetate (Figs. 3C and 3D). In our next study, we will analyze the content of FAs and the expression levels of genes involved in higher alcohol synthesis in the mutant yeast strains during fermentation, and determine the regulatory effect of ATF2.
To summarize, deletion of one allele of EEB1 or/and EHT1 in industrial yeast EC1118 decreased production of ethyl esters. Furthermore, the partial downregulation of EEB1 altered the expression level of ATF2 and affected the acetate ester profile of wine. Although deleting one allele of the target genes might have a weaker effect on the ester profile of wine as opposed to complete deletion, it might be a more feasible strategy for modifying wine aroma. Modest changes in the concentration of these secondary metabolites can cumulatively affect the final sensorial quality of wine [13]. Therefore, sensory test of wine samples should be performed to evaluate the effect of these mutant strains with greater accuracy.