Given the prevalence of skin problems caused by environmental irritants, we aim to design an eco-friendly, effective, and affordable skincare product. To achieve this, we decided to modify metabolic pathways in yeast to produce key cosmetic ingredients using synthetic biology strategies. After thorough research, we selected ceramides and bisabolol as the primary active components for our product.

Our design involves overexpressing the tsc10 gene to produce excess ceramides and introducing the fused protein ERG20-MrBBS into yeast. We can monitor the expression of TSC10 and ERG20-MrBBS through the green and blue fluorescence emitted by EGFP and BFP, respectively[1][2].

Figure 2 Overexpressing MrBBS-ERG20 and TSC10 to produce bisabolol and ceramide

Upon further literature review, we discovered that excess ceramides and bisabolol can have adverse effects on yeast[3][4].

Figure 3 Adverse effects of excess bisabolol and ceramide

Therefore, we assessed the feasibility and limitations of our initial design.

We decided to proceed with overexpressing tsc10 and introducing ERG20-MrBBS, but with a greater focus on protecting yeast to enhance ingredient production. We conducted further literature reviews to develop strategies for controlling production within safe and effective ranges.

Inspired by Zhou et al. (2023), we explored the use of a genome oscillator, specifically the SIR2-HAP4 oscillator, to extend the lifespan of yeast. We realized that this system could be adapted to protect yeast and enhance the production of our target ingredients.

Our design involves inserting tsc10 and ERG20-MrBBS downstream of the sir2 and hap4 genes, respectively.

Figure 4 Inserting ERG20-MrBBS and TSC10 into yeast genome to regulate bisabolol-ceramide oscillation

This setup allows the oscillation of sir2 and hap4 to cause fluctuations in the production of ceramides and bisabolol.

After consulting with our instructors, we evaluated the feasibility and limitations of this approach:

We recognize the genome longevity oscillator as an effective method to enhance yeast production in the presence of toxic products. However, we decided to redesign the system to better integrate cosmetic ingredient production into the yeast genome oscillator.

We adopted a modular design approach to rebuild our production system, achieving oscillation without inserting genes into the yeast genome. This modular design facilitates the parallel assembly of multiple plasmids during experiments, allows for easy modification in case of issues, and enables the future expansion of cosmetic ingredient production.

Instead of directly inserting tsc10 and ERG20-MrBBS into the genome, we maintained these genes on plasmids and swapped the promoters to link the production system with SIR2 and HAP4.

Figure 5 Modular design to separate oscillation and production.

We successfully transferred the pRS42H-TSC10 plasmid and the pESC-ERG20-MrBBS plasmid into yeast with the genome longevity oscillator, significantly enhancing the production of ceramides and bisabolol. For more details, please refer to our Results section.

Our future work will focus on the following aspects:

1. Yeast Suicide Module: To further ensure the safety of our product, we plan to introduce a promoter (activated by small molecules) - SIR2 plasmid into yeast. This will allow us to control yeast quantity by adding small molecules to induce cell senescence when there are excess yeasts.

2. Precise Manipulation of Cosmetic Ingredient Production: After simulating an improved model, we will modify the promoters and add other regulatory elements accordingly. This will enable precise control over the production of cosmetic ingredients, ensuring that yeast produces ceramides and bisabolol at levels that meet industry standards.

First, we decided to use a restriction enzyme to cut the vector, preparing it for subsequent insertion steps. Next, we utilized PCR to obtain the CYC1 promoter and the TSC10 gene fragment. We then connected these fragments using Gibson assembly, ensuring they were correctly integrated into the vector. This process will help us construct the desired recombinant vector for future experiments.

(1) To build the vector, we used BplⅠ and BamHⅠ to cut the vector and disrupt the TEF promoter, allowing us to insert the CYC1 promoter and TSC10 into the plasmid.

Figure 6 BplⅠ and BamHⅠare used to cut the vector, but did not delete the whole TEF promoter

(2) To obtain CYC1 promoter and TSC10, we designed relevant primers and conducted colony PCR (click on protocol for more information)

(1) After cutting the vector, we realized that using BplⅠ and BamHⅠ might not completely disrupt the TEF promoter, potentially overriding the regulative function of CYC1 promoter. Thus, we decided to look for other restriction enzymes and cutting sites.

(2) Using colony PCR, we observed PCYC1 electrophoretic band, but no TSC10 bands.

Figure 7 There are PCYC bands but no TSC10 bands in colony PCR results

(1) Since the TEF promoter is stronger than the CYC1 promoter, it is crucial to completely remove the TEF promoter sequence. Therefore, we need to find better cutting sites.

(2) To determine the cause of the experiment's failure, we checked our primers, PCR programs, and DNA templates, and decided to change our DNA extraction method and improve the PCR programs.

Following the last iteration of the Design-Test-Build-Learn cycle, we revised our design and tried a different PCR program. We decided to continue using restriction enzymes to linearize the vector due to its simplicity and speed. However, we switched to different restriction enzymes to ensure the TEF promoter was fully disrupted. To address the poor TSC10 PCR results, we adopted touchdown PCR and extracted yeast genomic DNA instead of using colony PCR.

(1) After consulting our instructors, we used NaeⅠ and BamHⅠ this time. Although NaeⅠ also cuts the f1 ori region on the plasmid, it completely removes the TEF promoter.

Figure 8 NaeⅠ and BamHⅠare used to cut the vector and deleted the whole TEF promoter

(2) We used commercial kits to extract yeast genome DNA instead of colony PCR. Also, we changed the PCR program into touchdown PCR. For more information, please click on protocol.

After using NaeⅠ and BamHⅠto cut the vector and connect to CYC1 promoter and TSC10, we successfully built pRS42H-TSC10 plasmid:

Figure 9 After updating vector-cutting and PCR programs, our PCR achieved positive results.

Moving forward, we need to measure the level of ceramides after transferring the pRS42H-TSC10 plasmid and adjust promoters or other regulatory elements of gene expression to ensure that yeast produces ceramides at levels that meet industry standards.

To meet our need of observing fluorescent intensity of yeast cells in single-cell resolution and a real-time manner, a microfluidic chip was designed. The microstructures anchoring yeast cells and dispatching cell suspension and culturing media were designed referring to (5). PDMS was chosen as the material for the chip for transparency and stability. The outer profile of the chip was designed to fit into the slot of culturing chamber of the confocal microscope that we chose for observation.

Figure 10. The microfluidic chip structure.

The design was sent to Zhongxinhengqi Co. for manufacture. The chip was bond to 1mm thick glass slide as carrying plate. The media supplying tubes were inserted into the chip via straight stainless-steel tubes. Culturing media were contained in a 50ml syringe with piston removed raised to 90cm above the chip for gravity feeding. Outlet tube was inserted in the same way as the supplying tube, with its end placed at same level as the chip to avoid excess positive/negative pressure. Check hardware for detailed design.

The chip with accessories were installed as designed. Yeast cells suspension was filled into the 50ml syringe and raised to 60mm above the chip, trying to load cells through gravity. No cells were observed to be loaded into chip. So it was attempted to be forced into the chip by pushing the piston of the syringe. Though successful, it was met with significant resistance. The 60x lens that was planned for observation could not properly focus, so it was switched to 20x lens. The tubes connected to the chip were forced to bend to side to enable the cover of culturing chamber to close up. After testing, when disassembling and cleaning the chip, we discovered cracks on the PDMS chip around the pores with reduced tightness under hydraulic pressure, which was most likely caused by stress from stainless-steel tubes inserted. We also discovered clogs in channels in the chip, making it unable to be reused.

The resistance was caused by air in the chip, which should be expelled from the whole system in further tests, from the inlet tubes, chip, to the outlet tube.

Inability of 60x lens to properly focus was caused by their limited working distance. Thinner carrying glass plate should be applied to address this issue.

The clogging of the chip could be avoided by filtering all acellular fluids through 0.22μm filters and washing all tubes with filtered ddH2O.

The stress on PDMS chip by stainless-steel tubes was mainly caused by the height limit imposed by culturing chamber. Design choices like 90 degrees bended tubes and supports of the tubes could be adopted.

Figure 11. The microfluidic with. Red box indicates the stress from the stainless steel tube.

Based on what were learned from the last cycle, the carrying glass plate was switched to a 0.17mm thick cover slip. The stainless-steel tubes connecting the chip and inlet and outlet tubes changed from straight ones to 90 degrees bended ones.

Figure 12. The stainless steel was changed to the 90 degree bended stainless-steel tube.

The chip and tubes were pre perfused with filtered ddH2O. The chip was perfused in vacuum, with 0.0075% Tween-20 dropped on pores of the chip. The tubes were perfused with syringe. The whole chip and accessories assembly was built according to design.

The YPD media were filtered before usage. A time-lapse microscopy was set up to test the system in real observation. Focus shift that was not compensated by auto focus was discovered when inspecting pictures taken.

Figure 13. The 0.17mm coverslip enable us to test cell under the 60x microscope.

The focus shift was caused by deformation of chip significant enough to be uncorrectable by auto focus. This could be caused by two factors. One was shifting to thinner carrying glass plates, which were not rigid enough to resist minor forces imposed on them. Another was the stainless-steel tubes inserted into the chip, which were likely to be the sources of subtle forces. A support or jig for both the chip itself and the stainless-steel tubes could probably solve the problems of focus shift.

A jig to support the chip itself and the stainless-steel tubes inserted into the chip was designed. The bottom piece and the mainframe supported the chip, while a piece of drilled acrylic screwed to mainframe supported the stainless-steel tubes without significantly affecting observation. Check hardware for detailed design.

Figure 14. The jig of the chip.

The bottom plate and midframe were 3D printed in light curing resin. The piece of acrylic was hand drilled to ensure matching precision to the chip. The acrylic piece was screwed to the midframe with M2 screws. The jig and the chip were secured using paper tapes. Check hardware for detailed information.

The chip-jig assembly with accessory plumbing systems were tested with time-lapse microscopy under 60x lens. No obvious focus shift was observed.

This design was effective. It helped support the jig and reduce deformation of the chip, hence less focus shift.

Figure 15. The focus shift was largely reduced by using the chip jig.