Breast milk oligosaccharides, also known as human milk oligosaccharides (HMOs), are the third most abundant solid component in breast milk after lactose and fat. Over 200 sugar compounds were identified in HMOs, among which 2′-Fucosyllactose (2’-FL) has the highest proportion in HMOs (more than 30%). Studies have proved that 2'-FL not only promotes the development of the immune system and gut microbiota but also helps to prevent allergic disease and promotes brain function and cognitive development.
At present, there are three methods for synthesizing 2'-FL, including chemical synthesis, enzymatic synthesis, and whole-cell biosynthesis. However, there are some limiting factors for chemical synthesis and enzymatic synthesis including the toxic reagents and the cost of precursors. In contrast, the whole-cell biosynthesis method is relatively low-cost and mainly performed in engineered E.coli to produce 2'-FL. But, for large-scale fermentation, there is possible endotoxin contamination.
Thus, in this project, we designed and synthesized an engineered S. cerevisiae cellular factory producing 2'-FL from sweet potato residues by fermentation. On one hand, producing 2'-FL in S. cerevisiae is a relatively safe cellular factory to allow large-scale fermentation. On the other hand, it is promising that use sweet potato residues to further extend the industrial chain, improve the utilization value of inferior biomass resources, and cut the cost.
Our team aims to produce 2'-FL in S. cerevisiae for providing the economic strategy. In terms of raw material, there are two pathways to produce GDP-L-fucose GDP-D-mannose (called de novo pathway) or L-fucose (called salvage pathway). In the de novo pathway, the GDP-D-mannose converts 2'-FL and is secreted into a medium involving three enzymes gmd, wcaG, and DDD, as well as transporter lac12. In this study, industrial diploid Saccharomyces cerevisiae CCTCC M94055 was used as the host to express 2 ' -FL pathway genes by chromosome integration and complete molecular construction(Figure 1).

Firstly, we design the integration backbone plasmids XI-2 and X-3, which contain the up and down homology arms of XI-2 and X-3 genome site in yeast, respectively. Then, we designed the two genes WbgL and lac12 inserted into the integration backbone plasmids XI-2, and the other two genes gmd and wacG inserted into the integration backbone plasmids X-3. Those four genes are key to producing 2'-Fucosyllactose(2'-FL) in a yeast cellular factory. In addition, we added the promoter and terminator to flanking regions of these exogenous genes in order to facilitate expression in the engineered yeast. The map of these four plasmids are as figure 2.

XI-2 site target gene integration backbone plasmid, containing 200 bp upstream and downstream homology arms of chromosome XI-2 site. Randomly pick 10 transformants from the LB-Amp plate, and use the upstream and downstream primer pairs of the homology arm to verify whether the fragment inserted. The target band is about 450 bp. It’s possible that 2, 3, 4, 5, 9 and 10 transformants has inserted the DNA fragments as shown in Figure 3.

The plasmids of transformants 2, 3, 4, 5, 9, and 10 were extracted and further verified by NotI digestion (Figure 4). If the cut band size is correct as shown in Figure 4, we picked No. 10 to perform Sanger sequencing.

The sequence of the transformant plasmid No. 10 is no mutation and mismatch as shown in Figure 5, indicating we successfully constructed XI-2 site integration backbone plasmid.

The X-3 site contains 200 bp of homology arms, which on the upper and lower sides of the chromosome X-3 site. 10 transformants were randomly picked from the LB-Amp plate, and the upstream and downstream primer pairs of the homology arms were used to verify whether the fragment inserted. The target band was about 450 bp. As shown in Figure 6, except for 6 and 7, the rest of the transformants might insert fragments.

The plasmid of transformants 2, 3, 5, and 8 were extracted and further verified by by Not1 digestion (Figure 7). The target band was 3762+430 bp. Thus, we picked No. 5 for sequencing, and the results are shown in Figure 8.

The sequence alignment well matched, indicating that the X-3 site integrated backbone plasmid was constructed successfully.

Secondly, the wbgL-lac12 and gmd-wcaG incorporated into the integration plasmids XI-2 and X-3, respectively. The verification results were shown as follows.

In the XI-2-wbgL-lac12 plasmid, the wbgL and lac12 gene expression cassettes are inserted between the XI-2 homology arm, in different orientations. The wbgL+lac12 gene expression plasmid for XI-2 site integration was constructed by a two-step digestion cloning method. After wbgL gene expression cassette was integrated, lac12 inserted though Xho1 and BamH1 digestion sites.
Plasmids of 10 transformants were extracted and verified by Xho1+BamH1 double-enzyme digestion (Fig. 9). The positive transformant band was 6007+2635 bp, and the correct NO.11 was selected for sequencing. The results were shown as Figure 10.

The sequence alignment results showed well matched, indicating that the XI-2-wbgL-lac12 plasmid was constructed successfully.

In the X-3-gmd-wcaG plasmid, the gmd and wcaG gene expression cassettes are inserted between the upstream and downstream of the X-3 homology arm, in the same direction. The gmd+wcaG gene for X-3 site integration was constructed by a two-step digestion cloning method. After the gmd gene expression cassette was integrated, the wcaG was introduced using Xho1 digestion site.
Plasmids of 12 transformants were extracted and verified by Xho1+Mss1 double-enzyme digestion (Fig.11). The positive transformant band was 6220+1896 bp, and the No.11 plasmid with correct digestion was randomly selected and sent for sequencing.

The details of sequence alignment showed there is no mutation, indicated that the X-3-gmd-wcaG plasmid was constructed successfully.

Then, we used CRISPR-Cas9 technology to integrate the target four genes WbgL, lac12, gmd, and wacG into the genome of S. cerevisiae. The constructed X-3-gmd-wcaG integration plasmid and XI-2-wbgL-lac12 integration plasmid were digested with NotI, respectively. The large fragments were transformed into the yeast competent cell that already contains the Cas9 expression plasmid by the lithium acetate transformation method, together with the gRNA. Finally, colony PCR was used to verify the integration of exogenous genes.

Colony PCR was used to verify whether the X-3 and XI-2 loci gene fragments were integrated into S. cerevisiae strains. The results are shown in Figure 13. The integrated copy number of the four transformants was verified using four different primer pairs. If the internal primers (X-3 inner-primer pairs and XI-2 inner-primer pairs) can amplify the target band, and the outer primers (X-3 outer-primer pairs and XI-2 outer-primer pairs) cannot amplify the target band with the size of the integration homology arm (4346bp and 4880bp), it means that 2 copies have been integrated. If the primers outside the site amplify the target band of the size of the integration homology arm, it means that 1 copy of the target gene has been integrated. Based on the above analysis, we judged that the middle transformants 2 and 3 have clearly integrated one copy of the target gene, and can be tested for subsequent experiments.

The productivity of 2'-FL by recombinant yeast in the synthetic medium YPD30L2 was tested. As shown in Figure 14, 30 g/L glucose was completely consumed within the initial 4 h, and then the strain began to use the produced ethanol as a carbon source, showing a secondary growth state (Figure 14A). The initial addition of 2 g/L of lactose was undetectable after 48 h, resulting in about 0.7 g/L of 2'-FL (Figure 14B). The theoretical conversion rate of lactose to 2'-FL was 100%. Part of 2'-FL accumulated intracellularly and failed to be effluxed into the medium.

The productivety of 2'-FL by recombinant yeast from sweet potato residues was tested, as shown in Figure 15. Generally, the growth of the strain was slightly worse than that of the synthetic medium (Figure 15A), which may be due to the presence of some inhibitor for yeast growth in the sweet potato residues. The presence of lactose was not detected at 48 h, and the final yield was about 0.6 g/L 2'-FL (Figure 15B).

By expressing exogenous WbgL, lac12, gmd, and wacG genes, we constructed the metabolic pathway of 2'-FL and obtained a yeast cell factory with high production of 2'-FL. Our fermentation experimental data show that the gene-edited yeast cells can produce 2'-FL at a high level of 0.6-0.7g/L. At the same time, we integrated the exogenous gene lac12 into yeast, so that yeast cells have the ability to efflux 2'-FL into the medium. This is in line with the fact that the fermentation of high-yield 2'-FL by Saccharomyces cerevisiae, which is mainly based on food safety, has become a research hotspot and difficulty, and it is also the key to solving practical social problems. Last but not least, the CRISPR-Cas9 gene editing technology we used in this project is a good case. In the future, we can use this technology to continue editing yeast genomes to create more diverse yeast cell factories and produce more meaningful metabolites.