Our experiments were aimed at the low-cost production of N-acetylneuraminic acid and human lactoferrin by microorganisms to solve the problem of nutritional incompleteness and high price of formula milk powder. To achieve this goal, we successfully produced N-acetylneuraminic acid and human lactoferrin in Saccharomyces cerevisiae BY4741 by inserting the N-acetylneuraminic acid and human lactoferrin production genes into pESC-UAR, constructing recombinant vectors, and introducing them into Saccharomyces cerevisiae BY4741 using Gibson assembly and T4 ligation. We also validated our yeast protein expression system using the fluorescent protein gene mCherry to ensure that our system is usable in Saccharomyces cerevisiae and further validated the product generation. Our project went through three cycles of DBTL altogether, and encountered numerous problems during this period. However, we ultimately succeeded in constructing four composite parts and proved that they could produce lactoferrin and N-acetylneuraminic acid as anticipated.

Previous iGEM teams have attempted to produce lactoferrin, however they have failed. Therefore, we are the first team to produce human-derived lactoferrin in iGEM. In addition, we have not reviewed any reports of N-acetylneuraminic acid production in Saccharomyces cerevisiae before us. This represents that our experiments could provide a good basis for subsequent production of these two substances in Saccharomyces cerevisiae. In addition, the constructed fluorescent expression system may also provide another solution for protein expression detection in Saccharomyces cerevisiae.

Design 1:

We chose HsLF, a lactoferrin synthesis gene from humans, as a way to produce a product that is more suitable for infants and children. We used pESC-URA as a vector and Gibson assembly to construct this recombinant plasmid.

Build 1.0:

After synthesis of the codon-optimized sequence of this gene by Gene Synthesis, we amplified it by PCR using primers HsLF-Gibson-F (CACTATAGGGCCCGGGGATGAAACTTGTTTTTTTTTAGTC) and HsLF-Gibson-R (GCTAGCCGCGGTACCAAGCTTTTAGTGGTGGTGATGGTGATG) for its PCR amplification.

Problems encountered:

Based on the electrophoresis results, however, our amplification was a failure, due to the fact that the electrophoresis results did not produce the expected DNA bands.
Failed

Learn 1.0:

We checked our primers and found that the primer HsLF-Gibson-F had excessive A-T content, which may have caused it to not bind well to the template, leading to amplification failure.

Design 2.0:

Therefore, we designed primer 2HsLF-Gibson-F (CGACTCACTATAGGGCCCGGGGATGAAACTTGTTTTTTTTTTTAGTCTTAC), which is a few base pairs longer than the original primer at both the 5‘ and 3’ ends, giving it a higher C-G content, easier binding to the template, and an end that avoids the continuous A-T region.

Build 2.0:

We re-amplified the fragment by PCR using new primers. The results were clearly superior to the previous ones. A band of about 2000bp appeared on the electropherogram, which is the size of the lactoferrin gene (2130bp).

After linearisation of the vector using SmaI and HindIII, we ligated the vector and target gene fragments using ClonExpress II One Step Cloning Kit, which was then transferred into E. coli DH5α competent cells and cultured in LB solid screening medium.

After 12h incubation at 37℃, we picked single colonies and performed colony PCR. For the strains that were verified to be correct, we sent them to a professional sequencing company to verify the correctness of their gene sequences, and the results were as follows:

After determining the correctness of the recombinant plasmid, we extracted the plasmid and transfected it into Saccharomyces cerevisiae BY4741 competent cells according to the method mentioned in the protocol, and transfected it into SD-URA solid medium for screening. We simultaneously transferred the pESC-URA plasmid into Saccharomyces cerevisiae as a control for subsequent validation experiments. Single colonies were successfully grown after 72 h of incubation at 30°C.

Test 2.1:

To test whether lactoferrin was successfully synthesized, following induction and fermentation, we extracted total protein from cells and examined it by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE).

Unfortunately, after SDS-PAGE, we did not see any obvious bands of lactoferrin appear.

Failed

Test 2.2:

By observing the protein bands, we found that the larger protein bands in our extracted Saccharomyces cerevisiae protein samples were significantly lighter compared to the standard protein samples. Through reviewing the literature and consulting experts, we learned that Saccharomyces cerevisiae itself is not suitable for expressing proteins that are too large, while lactoferrin has a size of 79 kDa.

In addition, Saccharomyces cerevisiae's thicker cell walls, which are difficult to lyse completely, may result in the larger proteins not being extracted completely, and therefore, if the protein is not expressed in sufficiently large quantities, this will likely lead to difficulties in the assay.

Due to the lack of more accurate testing equipment in our lab, we chose to measure the presence of lactoferrin in the fermentation broth using spectrophotometry. This utilizes the principle that lactoferrin produces a specific absorption peak at 475 nm.

We used pESC-URA / BY4741 as a negative control to exclude measurement errors caused by metabolites in the medium and during fermentation of the strain. 48 hours after the addition of the inducer galactose, we measured the absorbance of the fermentation supernatant at 475 nm. As shown, the absorbance of pESC-HsLF/BY4741 was significantly increased compared to the negative control.

Learn 2.0:

However, we are still not sure that the results of this experiment confirm the production of lactoferrin due to the susceptibility of the spectrophotometric method to errors due to other components in the fermentation broth and manipulation. According to literature review and consultation with experts, adding a fluorescent protein gene after the target protein can enable the localization and quantification of target proteins that are difficult to detect.

Design 3.0:

Therefore, we designed a system to verify whether the proteins in the recombinant plasmids can be expressed properly. For this purpose, two plasmids, pESC-mCherry and pESC-HsLF-mCherry, were constructed using Gibson assembly technology, and the latter was used to detect the expression of lactoferrin.

Build 3.0:

The construction process used the following primers mCherry-Gibson-F, mCherry-Gibson-R and HsLF-mCherry-F, HsLF-mCherry-R to amplify the mCherry gene fragments, respectively, which were later assembled using Gibson. The recombinant vector was also amplified using DH5α. The construction process and results are as follows.

Primer	Sequences (5’-3’)
mCherry-Gibson-F	CACTATAGGGCCCGGGCGTCGACatggtgagcaagggcgaggag
mCherry-Gibson-R	GCTAGCCGCGGTACCAAGCTTttacttgtacagctcgtcc
HsLF-mCherry-F	CACCATCACCACCACTAATAAAAGCTTatggtgagcaagggcgaggag
HsLF-mCherry-R	GGATCTTAGCTAGCCGCGGTACCttacttgtacagctcgtccatg

Test 3.0:

After the correctness of the recombinant plasmid was verified, it was transferred into BY4741 for culture. After 48 hours of adding the inducer galactose, the results are shown.

The rightmost figure shows from left to right: negative control, negative control, pESC-mCherry, pESC-HsLF-mCherry.

It can be seen that after induction, the strain shows a pink color when illuminated by UV light. However, this color can only be seen after centrifugation of the collected bacteria and cannot be observed directly by the naked eye in the liquid. In particular, pESC-HsLF-mCherry/BY4741 was lighter in color compared to pESC-mCherry/BY4741. This confirms our previous idea that the expression of HsLF in Saccharomyces cerevisiae is indeed low.

Learn 3.0:

After these two means of verification, we believe that we did successfully synthesize lactoferrin in Saccharomyces cerevisiae, which was not done by the previous iGEM team. Meanwhile, this work also offers experience for our subsequent projects. That is, we should employ multiple means and methods to verify whether our experiments succeed.

Team	Target	Results
2020 BJ101HS	Artificial breast milk: production of β-casein, κ-casein, lactoferrin, α-lactalbumin	Failed
2023 PIHS-BEIJING	Lactoferrin production in Escherichia coli	Failed

Design 1.0:

Since we found no precedent for synthesizing N-acetylneuraminic acid in Saccharomyces cerevisiae when reviewing the literature, we referred to reports of the synthesis of this substance in Escherichia coli and Pichia pastoris in the hope of introducing the genes gnal, yqaB, age, and neuB into Saccharomyces cerevisiae in order to complete the de novo synthesis of N-acetylneuraminic acid.

Build 1.0:

We first verified the synthetic pathway from N-acetylglucosamine to N-acetylneuraminic acid, for which we needed to construct the plasmid pESC-neuB-age. The N-acetylglucosaminidase differential isomerase gene (age) is from Anabaena sp. CH1 and the N-acetylneuraminic acid synthesis gene (neuB) is from Escherichia coli K1, synthesized by Gene Synthesis.

We amplified neuB and age fragments using neuB-Spe-F, neuB-EcoR-R and age-Xho-F, age-Hind-R respectively. The specific sequences of the primers are listed below.

Primer	Sequences (5’-3’)
neuB-Spe-F	GGACTAGTTCATTCACCTTGATTCTTGAACTC
neuB-EcoR-R	CCGGAATTCATGTCCAACATCTACATTGTTG
age-Xho-F	CCGGAATTCATGTCCAACATCTACATTGTTG
age-Hind-R	CCCAAGCTTTTAAGACAAAGCCTCAAATTGTTG

After electrophoretic detection, it was confirmed that both fragments were amplified successfully.

Subsequently, we enzymatically digested the vector and neuB gene fragments with SpeI and EcoR I, respectively, and ligated them using T4 ligase before transferring them into E. coli DH5α competent cells. After screening by transferring into Amp-added LB medium, single colonies were successfully grown after 12h. Colony PCR and sequencing were used to verify whether the gene was successfully inserted into the plasmid.

After confirming the success of the construction, we extracted the plasmid and used it as a vector, which was digested with target genes age using Xho I and Hind III, respectively, and then ligated using T4 ligase. The subsequent steps were the same as before. The correctness of the recombinant plasmid was also determined by colony PCR and sequencing.

Test 1.0:

Subsequently, the plasmid pESC-neuB-age was transferred into the Saccharomyces cerevisiae BY4741.

For the detection of N-acetylneuraminic acid in the fermentation broth, a commercial kit was chosen. This kit is based on the principle that N-acetylneuraminic acid forms a purplish-red complex with 5-Methylresorcinol in the presence of an oxidising agent, and the absorbance follows the colourimetric law. The amount of N-acetylneuraminic acid can be calculated by measuring the absorbance of the complex and comparing it to a standard.

In order to avoid errors caused by other components in the medium, we transfected the pESC-URA plasmid into BY4741 as a negative control, which was cultured and induced in exactly the same way as the engineered bacterium pESC-neuB-age/BY4741.

Compared to the negative control tubes, we found that the colour of the engineered bacterium pESC-neuB-age/BY4741 appeared significantly deeper and the absorbance increased. From the measurement of absorbance and calculation, we can see that after 48h fermentation, the yield of N-acetylneuraminic acid was about 0.7 g/L.

Learn 1.0:

Based on our experimental results, we have successfully completed the synthesis of N-acetylneuraminic acid in Saccharomyces cerevisiae for the first time. However, due to the time problem, we could not complete the de novo synthesis of this product. And, from the results, our current yield is relatively low. Combined with the expression of lactoferrin and fluorescent protein, we speculated that this might be affected by our yeast expression system. Therefore, discarding the plasmid expression vector and switching to overexpression of the relevant genes on the genome might be a way out to solve this problem.

Although we successfully produced lactoferrin and N-acetylneuraminic acid in Saccharomyces cerevisiae, we still hope to increase the product yield and reduce the cost in our future work.

Firstly, the current yield of HsLF is still difficult to detect by SDS-PAGE, and although there are many reasons for this result, the lower yield and the toxicity of human lactoferrin to Saccharomyces cerevisiae seem to be of the greatest concern.

Through expert interviews and review of the literature, we learnt that human lactoferrin consists of two parts, the N-lobe and the C-lobe, which can perform their physiological functions separately. Its absorption in the human body also requires breaking down into small molecule peptide chains first. Therefore, we will subsequently split lactoferrin into two structural domains around 37 kDa, which are more suitable for Saccharomyces cerevisiae to express separately. We will also try to modify Saccharomyces cerevisiae to enhance its resistance to HsLF using a directed evolutionary approach.

For N-acetylneuraminic acid, our initial plan was to achieve its de novo synthesis in Saccharomyces cerevisiae. However, due to time constraints, we only verified its ability to convert N-acetylglucosamine into N-acetylneuraminic acid. Considering the limited ability of plasmid vectors to express protein, we will subsequently integrate the relevant genes into the genome of Saccharomyces cerevisiae and explore the best way to combine them. In addition, we will try to knock down the competing pathways to further increase the metabolic flux of the N-acetylneuraminic acid synthesis pathway.

1. Liang Q, Richardson T. Expression and characterization of human lactoferrin in yeast Saccharomyces cerevisiae[J]. Journal of Agricultural and Food Chemistry, 1993, 41(10): 1800-1807.
2. Yen C C, Wu P Y, Ou-Yang H, et al. Production of bioactive porcine lactoferrin through a novel glucose-inducible expression system in Pichia pastoris: unveiling antimicrobial and anticancer functionalities[J]. International Journal of Molecular Sciences, 2024, 25(3): 1818.
3. 涂茂林,杨戬,刘猛,等. 乳铁蛋白检测方法研究进展[J]. 乳业科学与技术,2015,38(1):32-37. DOI:10.7506/rykxyjs1671-5187-201501008.
4. Thomassen E A J, Veen H A, Berkel P H C, et al. The protein structure of recombinant human lactoferrin produced in the milk of transgenic cows closely matches the structure of human milk-derived lactoferrin[J]. Transgenic research, 2005, 14: 397-405.
5. 赵小敏. 代谢工程改造毕赤酵母生产N-乙酰神经氨酸[D]. 北京:中国农业科学院,2022.
6. Gietz R D, Schiestl R H. High-efficiency yeast transformation using the LiAc/SS carrier DNA/PEG method[J]. Nature protocols, 2007, 2(1): 31-34.
7. Conesa C, Calvo M, Sánchez L. Recombinant human lactoferrin: a valuable protein for pharmaceutical products and functional foods[J]. Biotechnology Advances, 2010, 28(6): 831-838.
8. Pang Q, Han H, Liu X, et al. In vivo evolutionary engineering of riboswitch with high-threshold for N-acetylneuraminic acid production[J]. Metabolic engineering, 2020, 59: 36-43.