OUP user menu

Recombinant industrial brewing yeast strains with ADH2 interruption using self-cloning GSH1+CUP1 cassette

Zhao-Yue Wang, Jin-Jing Wang, Xi-Feng Liu, Xiu-Ping He, Bo-Run Zhang
DOI: http://dx.doi.org/10.1111/j.1567-1364.2009.00502.x 574-581 First published online: 1 June 2009


A self-cloning module for gene knock-out and knock-in in industrial brewing yeast strain was constructed that contains copper resistance and γ-glutamylcysteine synthetase gene cassette, flanked by alcohol dehydrogenase II gene (ADH2) of Saccharomyces cerevisiae. The module was used to obtain recombined strains RY1 and RY2 by targeting the ADH2 locus of host Y1. RY1 and RY2 were genetically stable. PCR and enzyme activity analysis of RY1 and RY2 cells showed that one copy of ADH2 was deleted by GSH1+CUP1 insertion, and an additional copy of wild type was still present. The fermentation ability of the recombinants was not changed after genetic modification, and a high level of glutathione (GSH) was secreted, resulting from GSH1 overexpression, which codes for γ-glutamylcysteine synthetase. A pilot-scale brewing test for RY1 and RY2 indicated that acetaldehyde content in fermenting liquor decreased by 21–22%, GSH content increased by 20–22% compared with the host, the antioxidizability of the recombinants was improved, and the sensorial evaluation was also better than that of the host. No heterologous DNA was harbored in the recombinants; therefore, they could be applied in the beer industry in terms of their biosafety.

  • industrial brewing yeast
  • self-cloning
  • acetaldehyde
  • glutathione
  • ADH2
  • GSH1


Self-cloning and non-self-cloning are two kinds of genetic modification techniques classified by the source of gene (Akada, 2002). As no heterologous DNA is carried into the self-cloned strains, the self-cloning technique has made it possible to construct new biosafe strains with improved fermentation capability.

Acetaldehyde is an intermediate compound in the formation of ethanol. As an off-flavor component, which also affects beer staling, the control of its production in beer is very important. The acetaldehyde content in Chinese beer is usually higher (3–8 mg L−1) than in overseas fine beer (<2 mg L−1) (Wang, 2005). This problem has been driving brewing research to produce flavorfully attractive beers with a lower acetaldehyde content. Previous research into decreasing the content of acetaldehyde in beer has mainly focused on the method of fermentation, heterologous microorganism contamination, brewer's yeast vitality, etc. Beers brewed with different yeast strains often contain different acetaldehyde contents, and therefore improvement of yeast strains that is especially based on the regulation of gene expression is also crucial to the beer quality. Several genes in Saccharomyces cerevisiae are related to acetaldehyde metabolism, such as pyruvate decarboxylase gene (PDC), aldehyde dehydrogenase gene (ALDH) and alcohol dehydrogenase (EC gene (ADH); of these, the ADH2 gene encodes for alcohol dehydrogenase II (YADH-2), catalyzing ethanol oxidation toward acetaldehyde (Russell, 1983). As ethanol is the metabolic end product, the interruption of the acetaldehyde pathway, which is derived from ethanol would have less effect on the metabolic balance in yeast cells, and therefore ADH2 gene deletion is appropriate for decreasing acetaldehyde. Some yeast strains with disruption of the ADH2 gene have been constructed, but they are generally auxotrophic strains or haploids that can only be used in laboratory (Shi, 2001; Saccharomyces Genome Deletion Project web page), or they were not safe for commercial application because of the bacteria gene brought into the cells during genetic modification (Wang, 2005; Guo, 2006; Jiang, 2007).

Flavor stability is also very significant for evaluating beer quality; however, oxidation often occurs during beer storage, causing beer to age. As a result of oxidation, many aldehyde compounds, detrimental to beer flavor, are produced. If the antiaging compounds in beer were increased, the flavor stability might be improved. Glutathione (GSH) is an important antioxidant against the toxic effects of O2 and other oxidative compounds, and increasing the GSH content in brewer's yeast helps to retain the flavor of the beer (Li, 2004). GSH is synthesized by two sequential reactions in S. cerevisiae, and GSH1 gene encodes for γ-glutamylcysteine synthetase (EC, catalyzing the rate-limiting reaction for GSH synthesis (Ohtake, 1989). As a smaller peptide, GSH could be excreted to the outside of yeast cells; thus increasing the GSH1 gene copy in brewing yeast might lead to a higher GSH content in beer, and improved flavor stability. Although cloning of GSH1 gene and its expression in S. cerevisiae has been reported by some researchers (Yasuyuki & Seizou, 1991; Fan, 2004), most of the constructed transformants are hazardous for use in the beer industry because foreign DNA is harbored in the cells. In our previous studies, another copy of GSH1 gene was integrated into yeast proteinase A gene, PEP4, using a self-cloning technique, which not only ensured overexpression of GSH1, but also the safety of the engineered strains (Wang, 2007).

As an advantageous biotechnique, self-cloning genetic modification would be widely applied in many fields, especially in the food industry. Beer flavor, color and foam are the most significant indicators for evaluating beer quality, and the self-cloned strains with better beer foam have been constructed in our lab (Wang, 2007). However, as an important component for both beer flavor and its stability, until now, the acetaldehyde content in beer production has not been improved, based on the strain profile by self-cloning technique. In this study, recombinant industrial brewing yeast strains that are biosafe were constructed first by disrupting ADH2 gene and overexpressing GSH1 gene and the CUP1 gene, a copper-resistant gene for screening, for the purpose of decreasing acetaldehyde and increasing the antiaging ability of beer. In addition, both the genetic validity and the metabolic effect of the improved strains were assessed.

Materials and methods

The strains and plasmids used in this study are listed in Table 1. Escherichia coli DH5α was used as the host for plasmid construction. Industrial brewing yeast Y1 used as the host for yeast transformation was obtained from the Center of General Microbiology in China Committee (CGMCC).

View this table:
Table 1

Strains and plasmids used in this study

Relevant genotypeReferences or sources
Escherichia coli DH5αsupE44ΔlacU169(ϕ80lacZΔM15) hsdR17 recAl endAl gyrA96 thi-1 relAlStratagene
Y1Wild-type yeast strainCGMCC
RY1Recombined yeast strainThis work
RY2Recombined yeast strainThis work
YIp5Cloning vector, URA3 amp tetStinchcomb (1980)
pYCUPRecombined plasmid, URA3 ampZhang (2005)
pGF-2Recombined plasmid, URA3 ampFan (2004)
pYADHRecombined plasmid, URA3 ampThis work
pAGCRecombined plasmid, URA3 ampThis work

Cultivation conditions

Escherichia coli strain was grown at 37 °C in Luria–Bertani medium (Sambrook & Russell, 2001) supplemented with ampicillin (100 mg L−1) or tetracycline (50 mg L−1) when necessary.

The yeast strain for transformation was grown at 28 °C in YPD medium (1% yeast extract, 2% peptone and 2% glucose). The recombined strains were selected on YPD plate with 6 mM copper sulfate (CuSO4). YPD or 10°P wort was used as the medium for culture or fermentation of the yeast strains.

DNA manipulation and plasmid construction

Plasmid DNA was prepared from E. coli as described by Sambrook & Russell (2001). Genomic DNA of yeast strains was prepared as described by Burke (2000). The primers used in this study are listed in Table 2.

View this table:
Table 2

Oligonucleotide primers used in PCR amplification

PrimersSequence 5′→3′

Alcohol dehydrogenase II gene, ADH2, was amplified from the genomic DNA of Y1 with primers ADH2-L and ADH2-R using PCR. The PCR product was purified and digested with restriction enzyme BamHI–EcoRI, and then cloned into YIp5 (Stinchcomb, 1980) to construct plasmid pADH. CUP1 gene was obtained by digesting plasmid pYCUP (Zhang, 2005) with HindIII–KpnI, and GSH1 gene was obtained by digesting plasmid pGF-2 (Fan, 2004) with KpnI–SacI. These two DNA fragments were then ligated into the HindIII and SacI sites of pADH by T4 DNA ligase to construct plasmid pAGC (Fig. 1), which was analyzed using different restriction enzymes (Fig. 2).

Figure 1

Construction of plasmid pAGC.

Figure 2

Analysis of plasmid pAGC by restriction enzyme digestion. Lanes 1–4, DNA fragments of pAGC digested by KpnI, SacI/KpnI, HindIII and SalI, respectively. Lane M, marker.

Yeast transformation and screening

The DNA fragment for yeast transformation was prepared by PCR using plasmid pAGC as template and ADH2-L/ADH2-R as primers (Fig. 3). The 6200-bp DNA product was purified according to the DNA Gel Extraction Kit, and used for yeast recombination using the lithium acetate method as described (Ito, 1983; Schiestl & Gietz, 1989). The recombinants were selected on YPD medium containing 6 mM CuSO4.

Figure 3

PCR products for ADH2 deletion. Lane 1, PCR products for ADH2 deletion. Lane M, marker.

Alcohol dehydrogenase II (YADH-2) activity and GSH content

The recombined strains and their host were cultivated in 50 mL YPD at 28 °C for 4 days, and the yeast cells and fermenting liquor were then used for preliminary determination of alcohol dehydrogenase II (YADH-2) activity and GSH content.

Crude enzyme was extracted using ultrasonication. YADH-2 activity was assayed as described by Blandino (1997). The reaction mixture contained 2.5 mL of 0.1 M Tris/HCl buffer (pH 8.8), 200 μL of 30 mM NAD+, 100 μL of 15 M ethanol and 10 μL of enzyme preparation, in a total volume of 3.0 mL. The enzyme reaction was carried out at 30 °C in a 1-cm cuvette followed by measurement of the increase in A340 nm. One unit of enzyme activity was defined as the amount of enzyme catalyzing the production of 1 mol of NADH per minute under these conditions.

Protein concentration in the crude enzyme extraction was assessed using the Bradford method (Bradford, 1976) with bovine serum albumin as a standard.

GSH content was determined using 5,5′-dithiobis (2-nitrobenzoic acid) (DTNB) (Fan, 2004). The GSH content was expressed as milligrams GSH per gram dried yeast cells or milligrams GSH per liter of fermenting liquor.

Genetic stability analysis

Yeast strains were successively transferred into YPD for 50 generations, then plate streaking of the 50th generation strains was performed for copper-resistance stability. After 2-day cultivation at 28 °C, 100 single-grown colonies were chosen randomly and transferred to 0.5 mL of sterilized water to starve for 4 h at room temperature. A Greiner inoculation loop of the starved yeast suspension was inoculated onto a YPD plate with 5 and 6 mM CuSO4, respectively, and kept at 28 °C for 2 days.

The stability of genome DNA of the first and 50th generations of the recombinants and their host was analyzed using PCR.

PCR and DNA sequence analysis

PCR for amplification of ADH2 gene was performed in 50 μL volume with 25 μL 2 × Pfu PCR MasterMix, 400 ng of template DNA and 0.4 μM primers. PCR for the DNA fragment for recombination was carried out in 100 μL volume with 50 μL 2 × Long Taq PCR MasterMix, 800 ng of template DNA and 0.4 μM primers. Cycle conditions were 94 °C for 5 min followed by 30 cycles of 94 °C for 40 s, 55 °C for 1 min, 72 °C for 4 min and finally 72 °C for 15 min. PCR analysis of yeast recombinants was carried out in 10 μL volume with 5 μL 2 × Long Taq PCR MasterMix, 160 ng of template DNA and 0.4 μM primers.

DNA sequence was analyzed using automated DNA sequencer.

Fermentation test and pilot-scale brewing

The yeast strains were first grown in 5 mL of 10°P wort at 25 °C for 24 h and the suspension inoculated into 60 mL of wort (10°P) at a ratio of 1 : 50. After cultivation at 25 °C for 60 h, the strains were inoculated into 480 mL of 10°P wort in conical flasks with fermentation bungs. The fermentation was carried out at 12 °C for 10 days. The fermentation flasks were weighed every day and the difference in weight on two adjacent days was the CO2 reduction. The yeast pellets from the above conical flasks were inoculated into a 6-L European Brewery Convention (EBC) tube with 5 L 10°P wort and fermented at 10 °C for 16 days.

Attenuation degree, alcohol content and real extract concentration in the filtrate of the fermenting liquor were measured using Beer Analyzer.

Acetaldehyde, diacetyl and pentanedione content were measured using GC.

GSH content was assayed using the DTNB method. The thiobarbituric acid (TBA) value was measured as described by Grigsby & Palamand (1976). The flavor freshness period was assessed using the resistance staling value (RSV): RSV=1/4 (12/ΔTBA12+24/ΔTBA24+36/ΔTBA36+48/ΔTBA48). ΔTBA12–48 is the difference in TBA between the control that is kept at 0 °C and the samples that were kept at 60 °C for 12, 24, 36 and 48 h, respectively, before TBA measurement.

Results and discussion

Construction of recombined plasmids

Part of ADH2 gene including its coding sequence was amplified using PCR; the result of agarose electrophoresis showed that the size of the PCR product was about 2500 bp, as anticipated (figure not shown). After digesting the DNA fragment with BamHI–EcoRI and cloning it into YIp5, plasmid pADH was constructed. By digesting pADH with HindIII–SacI and ligating copper-resistance gene CUP1 from pYCUP and GSH1 gene from pGF-2 at these sites, plasmid pAGC was generated (Fig. 1), in which a 1410-bp fragment of the coding region of ADH2 was disrupted by 5200-bp DNA fragment insertion, including CUP1 and GSH1. Using different restriction enzymes to digest pAGC, it was proved to be the correct recombined plasmid (Fig. 2).

Selection of recombined strains and alcohol dehydrogenase II (YADH-2) activity and GSH content assay

The host strain, industrial brewing yeast Y1, was transformed with the 6200-bp DNA product prepared by PCR using pAGC as template and ADH2-L/ADH2-R as primers (Fig. 3). Because of the homologous sequence in this DNA fragment, upstream and downstream of ADH2, recombination between the 6200-bp DNA and chromosomal DNA in Y1 occurred. The maximum growth concentration of the host was 5 mM CuSO4 in YPD medium; as one copy of CUP1 gene was added in the recombinants, they were screened on YPD with 6 mM CuSO4. These recombinants were designated as RY1, RY2, RY3, etc.

According to the results shown in Fig. 4a, alcohol dehydrogenase II (YADH-2) activity in all the recombined strains decreased by 20–33%, a significant difference at P<0.01, which meant that ADH2 gene in these strains might be disrupted. Figure 4b shows that the GSH content in RY1–RY8 cells increased significantly by 28–39% and the GSH content in their fermenting liquor increased by 27–37% (P<0.01) compared with that in the host, which might be caused by overexpression of GSH1 gene. The ratios between intracellular and excreted GSH were similar to each other among all the transformants and the host, which indicated that the genetic modification in this study did not change the intracellular and extracellular GSH balance of the yeast strain; these results also suggested that the strategy of GSH overexpression in yeast was helpful for GSH overproduction in beer. Both the YADH-2 activity and the GSH content of RY1 and RY2 were better than in other recombinants; hence, these two strains were chosen for further study.

Figure 4

Comparison of YADH-2 activity and GSH content of different recombinants and their host. (a) YADH-2 activity; (b) GSH content. Values represent the means of three replications. Significance of difference P<0.01.

Genetic stability

All of the 100 single-grown colonies of the 50th generation recombinants RY1 and RY2 could grow on YPD+6 mM CuSO4, which suggested that CUP1 gene might have been inserted internally in ADH2 gene in the recombinants, resulting in the increase of copper resistance, and that these two strains were genetically stable.

PCR verification and sequence analysis

Four primers were designed in this study (Table 2) to show whether disruption of ADH2 and insertion of GSH1 and CUP1 occurred in the chromosomes of recombinants. Primers ADH2-L and ADH2-R were, respectively, located in the 5′- and 3′-ends of ADH2 gene. GSH1-R was the 3′-end sequence of GSH1 gene, and CUP1-L the 5′-end sequence of CUP1 gene. The PCR products using CUP1-L/ADH2-R and ADH2-L/GSH1-R as primer pairs and genome DNA of the recombinants as template were equal to the theoretical value in length, which were, respectively, 1500 and 4700 bp (Fig. 5, lanes 1 and 2). Sequence analysis showed that the ORF of CUP1 downstream of ADH2 and upstream of ADH2, and the ORF of GSH1 were included in the 1500- and 4700-bp DNA fragments, respectively, which displayed identity to reported sequences, suggesting that part of ADH2 gene in the recombined strains was knocked out with CUP1 and GSH1 gene knock-in. As ADH2 in the host was integral, there were no PCR products for CUP1-L/ADH2-R and ADH2-L/GSH1-R primer pairs (Fig. 5, lanes 4 and 5). Two bands of PCR products, 2500- and 6200-bp DNA, were seen when using ADH2-L/ADH2-R as primers and transformant DNA as template (Fig. 5, lane 3). Considering the results of sequence assay, synthesis of 6200-bp DNA indicated that one copy of ADH2 gene had been successfully disrupted by GSH1+CUP1 insertion; however, an additional copy of wild-type ADH2 was still present in RY1 or RY2 cells as reflected by the PCR products of 2500-bp DNA, which were the same as that of the host (Fig. 5, lane 6).

Figure 5

PCR analysis of genome DNA of yeast strains. Lane M, marker. Primer pairs: lanes 1–3 and lanes 4–6 were, respectively, CUP1-L/ADH2-R, ADH2-L/GSH1-R and ADH2-L/ADH2-R. Templates: lanes 1–3, genome DNA of RY1 or RY2; lanes 4–6, genome DNA of Y1.

The results of PCR verification for genome DNA of the 50th generation of the recombinants was in accordance with that of the first generation – further evidence of the genetic stability of these strains.

Fermentation test and pilot-scale brewing

CO2 reduction of RY1 and RY2 from conical flask fermentation was analogous to that of the host, indicating that the fermentation ability of these two strains did not change after the genetic modification. The attenuation results further indicated that the recombinants and the host presented a similar degree of fermentation (data not shown), suggesting that disruption of ADH2 and overexpression of GSH1 did not affect the fermentation performance of the recombinants. The acetaldehyde content in the fermenting liquor of RY1 and RY2 from conical flasks was 11.0 and 11.5 mg L−1, respectively, which decreased by 22–25% compared with that of the host. In comparison with the host, the GSH content of RY1 and RY2 (8.51 and 8.76 mg L−1, respectively) increased by 25–29% after conical flask fermentation.

During pilot-scale brewing, EBC tube fermentation, both acetaldehyde and GSH content on the first day were similar to each other in the recombinants and the host. However, when the fermentation was going on, differences appeared: the acetaldehyde content of the recombinants became lower and the GSH higher than that of the host. When fermentation was finished, the acetaldehyde content of RY1 and RY2 was decreased by 21–22%, and the GSH content increased by 20–22% compared with the host (Fig. 6). These results indicated that ADH2 deletion did indeed cause reduction of acetaldehyde content, and GSH1 overexpression ensured enhancement of the GSH content in the fermenting liquor of the recombinants.

Figure 6

Acetaldehyde and GSH content during EBC tube fermentation. Values represent the means of three replications. *Data for GSH content.

After EBC tube fermentation, real degree of fermentation and real extract of the recombinants were similar to the host, and alcohol content was a little higher than Y1 (Table 3). Although the genetic operation in this study was not directly targeted to the metabolism of diacetyl and pentanedione, the content of these two off-flavor components decreased more than that in the host (Table 3). This might be the result of the overexpression of GSH1 gene, because GSH could act as a key redox-dependent signaling molecule to mediate glutathionylation of several proteins during post-translational modification, in analogy with protein phosphorylation (Fratelli, 2005); consequently, the synthetic pathway of diacetyl and pentanedione might be impaired and their content reduced.

View this table:
Table 3

Test of performance parameters in fermenting liquor from EBC fermentation tube

Performance parametersStrains
Real degree of fermentation (%)66.5266.3466.41
Alcohol content (v/v, %)5.155.305.33
Real extract (%)4.564.484.60
Diacetyl (μg L−1)284227215
Pentanedione (μg L−1)207155160
TBA (OD530 nm)0.480.390.41
  • Values represent the means of three replications.

The TBA value is a universal method to determine the degree of beer staling during beer production, and RSV reflects the flavor freshness period of beer. The smaller the TBA is, the stronger the antioxidizability. The higher the RSV is, the longer the flavor freshness period lasts. According to the results shown in Table 3, the TBA value of the recombinants decreased, suggesting that the antioxidizability of fermenting liquor brewed with RY1 and RY2 was increased. The increased RSV value of the recombinants (Table 3) was further evidence that the flavor freshness performance of the improved strains was better than that of the original strain. As the TBA and the RSV methods are based on the measurement of oxycarbonyl compounds, the changes in both the TBA and the RSV value should be the global reflection of GSH1 overexpression and ADH2 deletion in the recombinant strains. As a major antioxidant, higher GSH could play an important role in preventing beer oxidation forming more oxycarbonyl compounds; as a crucial redox mediator, increased GSH could regulate several metabolisms by protein glutathionylation, and the content of some oxycarbonyl compounds might decrease, such as diacetyl and pentanedione, as shown in this study. As acetaldehyde is one of the most numerous oxycarbonyl compounds in beer, lower acetaldehyde levels could surely cause the decrease in TBA (Li, 2005). Our results further proved that the effect of GSH and acetaldehyde on beer staling was significant.

To evaluate comprehensively the sensorial characteristics of the fermenting liquor, six experts were invited to judge the taste; all considered that the beer brewed from the recombinants tasted better than the host.

In summary, several parameters of RY1 and RY2 were not affected after genetic modification, such as fermentation degree, whereas other parameters were improved, for example, the main indexes for maturation degree including acetaldehyde, diacetyl and pentanedione all decreased compared with the host, and GSH content and the antioxidizability increased. As both GSH and acetaldehyde are crucial to beer flavor and its stability, the upregulation for GSH1 and downregulation for ADH2 in further investigations might result in more advantageous characteristics for the engineered strains.


The ORF of ADH2 gene in industrial brewing yeast was disrupted by integration of GSH1 gene, and self-cloning recombined strains were constructed. These strains were genetically stable and their fermentation ability was not affected after homologous recombination. The performance of recombinants was better than the host in terms of lower acetaldehyde, higher GSH content and increased antioxidizability of fermenting liquor brewed with recombined strains RY1 and RY2. As the heterologous DNA fragment was not imported into the recombined strains, the brewed beer would be safe for public use and would be easily accepted by consumers.


This work was supported by Tsingtao Brewery Co. Ltd.


  • Editor: Isak Pretorius


View Abstract