Results

1. Assemble target gene fragments

As indicated in Figure 1, BAT2 is the DNA strand that is responsible for producing the higher alcohol during the brewing of wine. A replacement of the BAT2 strand with ATF1 is strongly required in order to reduce the adverse effects after getting drunk. The BAT2-L-TEF1-ATF1-CYC1-TEFP-NrsR-TEFt-BAT2-R fragment was prepared to replace the BAT2 fragment in yeast.

Figure 1. The map of gene segment integrated into the yeast genome to replace BAT2 with ATF1

1.1 Preparation of five fragments of BAT2-L, TEF1, ATF1, CYC1-TEFP-NrsR-TEFt, BAT2-R

According to the experimental design, five fragments of BAT2-L, TEF1, ATF1, CYC1-TEFP-NrsR-TEFt and BAT2-R should be copied first. In figure 2, the gel electrophoresis results of the above five fragments are shown.

Figure 2. The gel electrophoresis results of the replicated five fragments.
A: Line: BAT2-R, 501 bp, correct;
Line: CYC1-TEFp-NrsR-TEFt , 1715 bp, correct;
Line: ATF1, 1575 bp, correct.
Line: BAT2-F, 503 bp, correct;
B: Line: TEF1, 400 bp, correct;

According to the comparison between the gel electrophoresis results and the fragment map, we analyzed that the lengths of the DNA in figure 2A, 2B were consistent with the length of the fragments we needed, so the replication of the fragments was successful.

1.2 fragment assembly

According to the experimental plan, after the five small fragments are cloned, they need to be assembled together. We combined the DNA strands with the final strands by overlap PCR. The results of gel electrophoresis of the final fragment BAT2-L-TEF1-ATF1-CYC1-TEFP-NrsR-TEFt-BAT2-R are shown in the figure.

Figure 3. The gel electrophoresis results of the final fragments BAT2-L-TEF1-ATF1-CYC1-TEFP-NrsR-TEFt-BAT2-R.
The red arrow indicates the correct position of the fragment.

After comparing with the marker, it can be found that the marker of the assembled fragment as indicated by the red arrow in the gel electrophoresis map is between the marker's 4000bp and 5000bp markers. The assembled fragment is 4,665bp long, which is within the expected range. The result shows the gene strands are correctly connected and replicated

2. Construction of plasmid pYES2-ATF1

2.1 Construction of ATF1 gene expression cassette

According to the experimental design, S. cerevisiae strain A integrates ATF1 into the genome and replaces the BAT2 gene. Different from strain A, S. cerevisiae strain B uses a plasmid to carry the ATF1 gene, and the plasmid version highly expresses the ATF1 gene. We called this plasmid pYES2-ATF1.

In order to ensure the successful construction of plasmid pYES2-ATF1, it is essential to construct the ATF1 gene expression cassette. Therefore, we still used PCR and gel electrophoresis to clone and verify whether the ATF1 gene expression cassette was successfully constructed.

Figure 4. The gel electrophoresis results of the ATF1 gene expression cassette.
The red arrow indicates the mark of the gene expression cassette in the figure.

In figure 4, the ATF1 gene expression cassette is 2,256 bp long. The length of DAN as indicated by the red arrow is between 1,000bp and 2,500bp; it is close to 2,500bp. This suggests that we have successfully constructed the ATF1 gene expression cassette. At this point, we can safely connect the ATF1 gene expression cassette to the pYES2 plasmid that has been digested.

2.2 Ligation of plasmid vector pYES2 and ATF1 fragment

After confirming that the ATF1 gene expression cassette has been constructed, we need to introduce it into the pYES2 plasmid that has been cut by Spe I and Sal I restriction enzymes, and ensure that the ATF1 gene expression cassette is connected to the pYES2 plasmid backbone and form a new pYES2-ATF1 plasmid. We need to use gel electrophoresis to verify that the S. cerevisiae strain SFA-1, that is, S. cerevisiae strain B, has a complete plasmid, so as to ensure the smooth progress of the subsequent strain growth curve and simulated brewing fermentation test.

Figure 5. The gel electrophoresis result of the ligated pYES2-ATF1 plasmid.
The red arrow indicates the mark of the plasmid in the figure.

As can be seen on the left side of the figure, there is a clearer mark. After comparing with the marker, we can prove that the length of plasmid pYES2-ATF1 in the figure is 6000bp to 8000bp. After confirmation with the map, plasmid pYES2-ATF1 is 7841 bp long, which proves that the marker pointed to by the red arrow is the pYES2-ATF1 plasmid, that is plasmid pYES2-ATF1. So the construction is successful.

3. Growth Curve

We obtained a total of two genetically modified yeast strains, named SF-1 and SFA-1. SF-1 is a strain that integrates ATF1 into the BAT2 gene of the yeast genome, which results in reduced production of higher alcohols, thereby reducing drunkenness. SFA-1 is to transfer plasmid pYES2-ATF1 into yeast, and overexpress ATF1 in the form of plasmid, so as to achieve the purpose of increasing the aroma characteristics of wine. The third fermenting strain was the wide type (WT) yeast, which served as a negative control without any genetic modification of yeast.

Subsequently, we performed fermentation experiments based on three yeast strains, SF-1, SFA-1, and WT (wild type). Therefore, obtaining data between 2, 4, 7, 16, 24, and 32 hours respectively would allow the targeted organisms to grow and at every time node, we tested the OD₆₀₀ absorbance for each culture three times using a spectrophotometer, while each time we blanked the machine using ddH2O, thus, drew the Growth Curve using R program.

To be more specific, in order to make sure we have grown yeast equally, whenever finish testing the OD₆₀₀ absorbance value we put them back into the cradle for mixing the nutrients and organisms comprehensively so that the nutrient level won't be a limiting factor for yeast growths. Considering using the spectrophotometer, each time, we extracted 1.5ml of the solution from each of the three cultured strains: SF-1, SFA-1, and WT. Consequently, we transferred the solution into the transparent cuvette one at a time, while blanking the machine with another previously filled cuvette with double distilled water in it. Through the testing process, which we tried three times with each strain solution, the machine gave us three OD₆₀₀ absorbance numbers.

Thus, drew the Growth Curve using the R program. This diagram uses LOG2 to fitting the curve and we calculated the p-values when comparing with the WT using ANOVA Test as more than two groups were considered.

Figure 6. The curve of OD₆₀₀ absorbance of each strain that R program generated.
Time is on the x-axis and relative absorbance calculated from logarithum is the y-axis, with three keys SFA-1, SF-1 and Wildtype (WT). p-value is calculated when comparing with WT.

After analysis, we can find that although the growth amount and efficiency of Saccharomyces cerevisiae strains A and B do not reach or exceed the level of wild-type Saccharomyces cerevisiae strains. However, the slopes of cerevisiae strain SF-1 and SFA-1 can be relatively equal to those of wild-type cerevisiae strains at different test time points. This means that the growth efficiency of cerevisiae strains A and B is relatively consistent with that of the wild-type cerevisiae strain, which improves the competitiveness of our strain products. But we have at least demonstrated that our Saccharomyces cerevisiae strain can survive, grow and ferment.

4. Laboratories fermentation tests

The purpose of the simulated alcohol fermentation test is to verify that the strain can produce ethanol normally. In this project, this test also proves whether the simple strain we constructed can effectively reduce the production of higher alcohol.

We use gas chromatography to measure the content of higher alcohols. The following table shows the experimental data of various alcohols produced by wild-type AQ yeast and SF-1 yeast. We did two sets of experiments for each yeast. The experimental data showed that the ethanol content of SF-1 yeast was similar to that of wild-type AQ yeast, indicating that the knockout of BAT2 gene and integration of ATF1 gene had little effect on the production of ethanol.

We averaged the two sets of experimental data and calculated the percentage of each higher alcohol production, as shown in the following figures.

Figure 7. Fitting result of Wild-type AQ yeast and engineered yeast SF-1.

We found that both sets of data of strain SFA-1 were abnormal, and the initial judgment was that the strain was contaminated by controllable unknown organisms. This is undoubtedly very negative news, we did not do a good job of protecting the samples during the experiment, which led to controllable contamination.

But the good news is that the higher alcohol production of strain SF-1 decreased by about 152.6 mg per liter compared to the wild-type strain. is 38.000% of the wild-type strain. This proved that strain A achieved the experimental assumption and successfully reduced the production of higher alcohols.

However, the ethanol production of strain A was slightly lower than that of the wild-type strain by about 0.15% vol. is about 87.234% of the wild-type strain. This result is estimable and acceptable because the growth curve data for strain A is the worst of the three strains.

Since strain B has only one more ATF1 gene expression cassette than strain A, it is not difficult to speculate the test result of strain B. It is speculated that the ethanol production of strain B may be reduced by about 0.1-0.2% vol compared to the wild-type strain, but the production of higher alcohols may be reduced more than that of strain A, To pessimistic estimate, it is possible to reduce about 155 mg per liter to 170 mg per liter compared to the wild type strain. If optimistic estimates, strain B can reduce about 180 mg per liter or even 200 mg per liter of higher alcohol compared to the wild-type strain. Although this prediction is radical, it cannot be ruled out.

In a nutshell, strain A successfully achieved our experimental goal, and although strain B did not prove to be achieved, it was predicted to be achieved with certainty.