Engineering iterations of the Light Control System

Design & Build

After determining the direction of the project, we selected Saccharomyces cerevisiae as the chassis, and built a light control system on it for the production of blended fragrance. As E.coli is a better system for protein engineering and there were studies on this system in prokaryotes[1] through a large number of literature research and information retrieval, so we selected E.coli for verification of function of the light control system. Considering the advantages of E.coli, such as shorter culture cycle, faster growth rate, and easy transformation and expression of heterologous genes, the team hopes to simplify the experimental operation, shorten the experimental time, and improve the effect of iteration number. We replaced the biosynthetic genes for terpenoid fragrances with fluorescent proteins, deleted signal amplification circuit involving GAL4, and replaced yeast promoters with bacterial ones, to generate the first designed gene circuit, as shown below.

Figure 1. Gene circuits in E. coli

We constructed plasmids according to sequences shown from the literature[1] and transformed them into E.coli TOP10 competent cells, as a chassis, for test. The construction of the engineered bacteria was verified by colony PCR, where primers specific for the three cassettes (PrrnBp1_rbs34_VP16_EL222, PBlindv1_ rbs34_eGFP, and PBLrep_rbs34_mCherry) were used. The amplification of DNA fragment corresponding to the expected size indicates successful construction of the strain.

Figure 2. E.coli transformation experimental verification result. M-marker. 1-rrnBp1_rbs34_VP16_EL222 (composite part K4616010, 937 bp). 2-PBlindv1_ rbs34_eGFP(composite part K4616011, 794 bp).
Figure 3. E.coli transformation experimental verification result. M-marker. 1-PBLrep_rbs34_mCherry (composite part K4616012, 765 bp).

Test

In order to verify the performance of the light control system, we cultured the engineered E.coli and illuminated the culture with pulsed light. After inspecting cells under fluorescence microscope, we verified that both GFP and RFP were expressed, indicating that the light control system worked as expected.

Figure 4. E.coli engineering bacteria expressing fluorescent protein under light and dark conditions.(A) Green fluorescent protein was induced under blue light.(B) Red fluorescent protein is produced in the dark.(C)E.coli engineering bacteria in the open field.

Since there are clear data in the literature[1], the literature data are used to model and guide the design of subsequent lighting experimental conditions. (More information please view our modeling page ->)


Learn & Redesign

We verified through experiment that this system is sensitive to blue light at 450nm, so we further integrated this system into S. cerevisiae to produce different proportions of products by controlling different light cycles. At the same time, considering that the Pc120 promoter in the gene line has different starting strength in different kinds of chassis, in order to make the gene circuit better adapt to the yeast cell, we introduced a set of gal transcription factor regulation system based on the original gene circuit, which amplified the signal and improved the expression controllability. Comparing to the final circuit, we replaced the biosynthetic genes with fluorescent proteins, to better visualize and quantify the expression level of the two cassettes. The second gene circuit is shown below.

Figure 5. Light control system in Saccharomyces cerevisiae

In the initial construction of engineered cells, the effect of fragment ligation using OE-PCR (overlap-extension PCR) technology was not satisfactory due to the excessive number of short gene fragments with different lengths. We switched to Gibson assembly to construct DNA fragments and plasmids.

Figure 6. Constructed light regulated plasmid.(A)Plasmid ez-l580 was purchased from addgene [2].(B)Finally, light controlled green fluorescent protein particles were constructed.
Figure 7. PCR result. (A)M-marker. 1-Pgal1-s (part K4616168, 719 bp). (B)M-marker. 1-egfp(750 bp).

Bulid & Test

To verify that this circuit can achieve our expected effect in yeast, we introduced the plasmid into yeast cells for expression and test.

Figure 8. Saccharomyces cerevisiae expressing fluorescent proteins regulated by light.(A)Saccharomyces cerevisiae producing green fluorescent protein was induced under blue light.(B)Saccharomyces cerevisiae without blue light induction.

We plan to use fluorescence intensity to express the yield of compounds, but due to time reasons, we only imported green fluorescent protein fragments into the light regulation line, which does not affect our detection of yeast yield changes under periodic pulsed light. We performed periodic light induction on Saccharomyces cerevisiae and detected it.

Figure 9. Temporal control of gene expression by blue light based light control system in a single cell. OFF–ON–OFF–ON cycle for every 8 h over a period of 52 h.

In general, the OD of yeast will reach a relatively stable plateau after 24h. Therefore, in order to reduce the impact of yeast growth on fluorescence intensity, we choose to start intermittent light detection after 24h.

As shown in the figure, we conducted intermittent light culture at an interval of 8h, and sampling and detection at an interval of 4H. That is, 28-36h is blue light culture, 36-44h is dark culture, and 44h-52h is blue light culture again.

It is evident that upon the initiation of blue light exposure, the fluorescence intensity of the yeast initially decreases before subsequently rising. This observed phenomenon could be attributed to the impact of blue light on certain yeast pathways, resulting in a reduction in the fluorescent protein content within the yeast cells. During this phase, the newly synthesized fluorescent protein has not yet reached sufficient levels, leading to the initial decrease in fluorescence intensity. As the blue light exposure continues, it triggers the activation of the promoter, resulting in an increase in green fluorescent protein production, subsequently leading to an augmentation in the overall bacterial fluorescence intensity.

When the cell is in the dark, the promoter in the metabolic pathway of green fluorescent protein production stops functioning, resulting in the inability to synthesize green fluorescent protein, and then the fluorescence intensity of the cell will continue to decline.

The second time of blue light irradiation, the fluorescence intensity also showed a first decline and then rise, the same as the first time, which also verified our speculation.


Learn

From the results of Test, it can be seen that our light control components can achieve the light control effect well, so we decided to take the next step of design. On one hand, Professor Zhao from the school of chemistry and chemical engineering, Beijing Institute of Technology gave the following suggestions for the particularity of the fermenter we need: for large-scale industrial equipment, light cannot completely penetrate the entire medium, but this characteristic can be used to calculate and design, so that the yeast in the upper part of the culture medium in the constructed container can be exposed to light to produce biomass a, and the lower part cannot be exposed to light to produce biomass B. This breaks our original idea that every yeast is exposed to light. For specific amplification design, please refer to relevant HP links.On the other hand, the experimental group plans to carry out promoter enhancement engineering, by strengthening the strength of Pgal promoter and enhancing its response to transcription factors, to weaken the problem of yield decline when light penetration decreases. Through literature research, the team decided to select a specific promoter Pgal1-s for gene circuit construction [2], and the new gene circuit is shown in the figure below.

Figure 10. Light control system after promoter modification

Redesign

Through literature research, we combined the production of sclareol and santalol with light control, so that the cell factory can produce a specific proportion of mixed fragrance. The circuit is still under construction, with all the biosynthetic genes amplified (shown in Figure 13 and 14) and we hope we can show you the result in the near future.

Figure11. Gene circuit

In addition, in order to verify that the product will not affect the cell growth, we also carried out the toxicity experiment of the products of sclareol and santalol on cells[3,4]. View the experiment page for details.


Bulid & Test

(1)The toxicity test results of the product on cells are shown in the figure below, indicating that the growth of yeast was not negatively affected by santalol or sclareol under their relatively high concentrations'.

Figure 12. Growth chart of cytotoxicity test

(2)We have amplified the gene fragment of the synthetic product by PCR, and the result is shown in the figure below. The target fragment was obtained, but it was not integrated into the yeast genome due to time reasons.

Figure 13. Genes related to the synthesis of sclareol PCR result. M-marker. 1-tps_lpps (part K4616888, 4132 bp).
Figure 14. Genes related to the synthesis of santalol PCR result. (A)M-marker. 1-sass (part K4616003, 1757 bp). 2-cprs (part K4616009, 2157bp). (B)M-marker. 1-cyp76 (part K4616004, 1546 bp).

(3) In order to determine the proportion of products we synthesized under specific light conditions, we used GC-MS for detection and quantification. We have analyzed a series of standard samples with known concentrations, and by integrating the peak areas and plot against their concentrations, we built standard curves for the two products in our project, as shown in the figure below. After the synthesis, we would perform GC-MS analysis under the same conditions, and calculate the product concentration using standard curve.

Figure 15. The standard curve of sclareol
Figure 16. The standard curve of santalol

Prospect

Our work can provide guiding help for industry to produce relevant products through light control system, and provide ideas for industrial production.

In the actual production in the future, our engineered cells can reduce the environmental pollution and related labor and equipment costs of traditional production methods. In the future, we will continue to improve our yeast strain, which is controlled by light to produce mixed fragrance, and improve the yield and control accuracy. Thereafter, we hope to expand the product category and apply it to more substances produced in proportion. We also plan to use PYB-PIF3 type red light inducible promoter together with existing lines to build three product controllable gene production lines.

Iteration on microbial chassis (Yeast.) for optimizing production effects

Design & construction

In order to reduce the basic metabolic pathways in yeast cells to consume extra energy and material, and thus to increase the productivity of our designed production and regulation lines, we analyzed the metabolic flow of yeast, applied OptKnock to select the metabolic pathways that do not affect the growth of yeast for genome knockdown,[5,6] and used homologous recombination technology to knock down some of the genes as shown in the following Fig. 17 Fig. 18, and constructed a first version of chassis cells in the final form as shown in Fig. 19.

Figure 17. Metabolic pathways related to products formation and candidate genes to knockout.
Figure 18. Metabolic pathways related to products formation and candidate genes to knockout.
Figure 19. Metabolic pathways that should be on the relevant metabolic pathway after knockout

Test

We want to verified that the knocked-out gene did not affect the normal growth of the cells, so we designed an experiment to detect the change of OD600, which reflects the growth of the cells, over time, and used the software (Origin) to fit the growth curves using the logistic growth model, which proved that the growth of the yeast (strain Δ236W) was affected by the knocked-out gene, which was unsatisfactory in comparison with our expected effect.

Figure 20. Growth rate curves
Figure 21. Growth curves

Learn

Through communication with our instructor, we believe that we need to consider the effect of NADPH in the metabolic flow reaction and that the original model should be modified to re-screen the genes that need to be knocked out.


Redesign & Test

Considering the effect of NADPH, we redesigned the Optknock model and came up with genes of metabolic pathways that need to be knocked out. In addition, considering the effect of GAL80 and GAL4 genes present in yeast, we also decided to knock them out. The pathways that need to be knocked out and the metabolic pathways after knockout are shown below.

Figure 22. Updated metabolic pathways related to products formation and new candidate gene to knockout
Figure 23. Metabolic pathways related to products formation after gene knockout

We knocked out the above genes (gal4, gal80, spe4), cultured the knocked-out chassis cells, measured their growth curves, still fitted them using the logistic growth model, and used Origin for graphing, and the results are shown below.

Figure 24. Growth curves fitted to OD600 data

As shown in the figure, the growth was not significantly affected or even better. Hence the envisioned results were achieved. Constructing our gene line on this chassis could increase the yield of our expected product.

More information please view our modeling page ->

The actual experimental work is often slow and tortuous, and we inevitably encountered various problems in our work, and the results were not always as good as we expected, requiring repeated thinking and attempts. Through the process of engineering iterative DBTL, which combines the depth of modeling and experimentation, we have achieved the optimization of metabolic flow, the optimization of chassis selection, the satisfaction of production conditions, and the analysis of production estimation, which provides the basis for the product to be put into production in the future.

Reference

  1. [1]Jayaraman P, Devarajan K, Chua TK, Zhang H, Gunawan E, Poh CL. Blue light-mediated transcriptional activation and repression of gene expression in bacteria. Nucleic Acids Res. 2016 Aug 19;44(14):6994-7005. doi: 10.1093/nar/gkw548. Epub 2016 Jun 28. PMID: 27353329; PMCID: PMC5001607.
  2. [2]Zhao EM, Lalwani MA, Chen JM, Orillac P, Toettcher JE, Avalos JL. Optogenetic Amplification Circuits for Light-Induced Metabolic Control. ACS Synth Biol. 2021 May 21;10(5):1143-1154. doi: 10.1021/acssynbio.0c00642. Epub 2021 Apr 9. PMID: 33835777; PMCID: PMC8721662.
  3. [3]Wang Y, Gong X, Li F, Zuo S, Li M, Zhao J, Han X, Wen M. Optimized biosynthesis of santalenes and santalols in Saccharomyces cerevisiae. Appl Microbiol Biotechnol. 2021 Dec;105(23):8795-8804. doi: 10.1007/s00253-021-11661-9. Epub 2021 Nov 5. PMID: 34738171.
  4. [4]Cao X, Yu W, Chen Y, Yang S, Zhao ZK, Nielsen J, Luan H, Zhou YJ. Engineering yeast for high-level production of diterpenoid sclareol. Metab Eng. 2023 Jan;75:19-28. doi: 10.1016/j.ymben.2022.11.002. Epub 2022 Nov 9. PMID: 36371032.
  5. [5]Lu H, Li F, Sánchez BJ, Zhu Z, Li G, Domenzain I, Marcišauskas S, Anton PM, Lappa D, Lieven C, Beber ME, Sonnenschein N, Kerkhoven EJ, Nielsen J. A consensus S. cerevisiae metabolic model Yeast8 and its ecosystem for comprehensively probing cellular metabolism. Nat Commun. 2019 Aug 8;10(1):3586. doi: 10.1038/s41467-019-11581-3. Erratum in: Nat Commun. 2020 Oct 22;11(1):5443. PMID: 31395883; PMCID: PMC6687777.
  6. [6]Heirendt L, Arreckx S, Pfau T, Mendoza SN, Richelle A, Heinken A, Haraldsdóttir HS, Wachowiak J, Keating SM, Vlasov V, Magnusdóttir S, Ng CY, Preciat G, Žagare A, Chan SHJ, Aurich MK, Clancy CM, Modamio J, Sauls JT, Noronha A, Bordbar A, Cousins B, El Assal DC, Valcarcel LV, Apaolaza I, Ghaderi S, Ahookhosh M, Ben Guebila M, Kostromins A, Sompairac N, Le HM, Ma D, Sun Y, Wang L, Yurkovich JT, Oliveira MAP, Vuong PT, El Assal LP, Kuperstein I, Zinovyev A, Hinton HS, Bryant WA, Aragón Artacho FJ, Planes FJ, Stalidzans E, Maass A, Vempala S, Hucka M, Saunders MA, Maranas CD, Lewis NE, Sauter T, Palsson BØ, Thiele I, Fleming RMT. Creation and analysis of biochemical constraint-based models using the COBRA Toolbox v.3.0. Nat Protoc. 2019 Mar;14(3):639-702. doi: 10.1038/s41596-018-0098-2. PMID: 30787451; PMCID: PMC6635304.