Microbial Synthesis Powered Star Product for Vision Care

■ Substrate-independent de novo β-carotene biosynthesis: Achieved the CRISPR-Cas9-mediated, substrate-free de novo synthesis of β-carotene in Saccharomyces cerevisiae with high efficiency.
■ Standardized lipid droplet-localizing parts development: Established and characterized a set of well-defined lipid droplet-targeting biological parts, providing plug-and-play synthetic biology tools for global researchers.
■ Compartmentalization engineering boosts industrial production: Significantly enhanced β-carotene yield through innovative lipid droplet compartmentalization strategy, paving the way for industrial-scale manufacturing.

De novo Biosynthesis of β-carotene

We successfully established a de novo β-carotene biosynthetic pathway in S. cerevisiae(Figure 1). The metabolic engineering strategy involved:
(1) Construction of gene-integrated cassettes containing key enzymes in E. coli;
(2) CRISPR-Cas9-mediated genomic integration at the _POX_1 locus of S. cerevisiae BY4741; (3) Heterologous expression of carotenogenic genes including phytoene desaturase (carB, GenBank: AJ238028.1) from Mucor lusitanicus CBS 277.49 and bifunctional phytoene synthase/lycopene cyclase (carRP, GenBank: AJ250827.1) from Mucor circinelloides f. lusitanicus.
The engineered β-carotene-producing strain, designated SC-1, demonstrated successful pathway reconstitution.

Selection of Host Cells

Through comprehensive literature review and expert consultations, we identified Yarrowia lipolytica, Candida tropicalis, and S. cerevisiae as candidate chassis cells for β-carotene biosynthesis. For downstream applications in functional foods and animal feed production, biosafety considerations were prioritized. Consequently, S. cerevisiae — classified as Biosafety Risk Group 1 (RG1) — was selected as the optimal chassis organism, ensuring alignment with product safety requirements. Given that our objective is to synthesize β-carotene, and considering that terpenoids often inhibit cell growth, we supplemented the YPD medium with 0.2 g/L β-carotene to test the host strain’s tolerance to this terpenoid. Cell growth of S. cerevisiae was monitored by measuring optical density at 600 nm (OD₆₀₀) at various time points (Figure 2). The results demonstrate that S. cerevisiae was capable of growth in YPD medium supplemented with 0.2 g/L β-carotene, although it exhibited a mild reduction in growth rate compared to the control group without β-carotene supplementation.

Construction of Gene Integration Cassettes

We employed PCR to amplify the promoter, target gene, and terminator, followed by fusion PCR to generate Part1: PTDH3-carRP-TCYC1 (BBa_25I5R04N) and Part2 PTEF1-carB-TADH1 (BBa_25HTNRE3). These fragments were then cloned into a T-vector using DNA ligase and transformed into E. coli JM109 for plasmid extraction. In parallel, the genomic DNA of S. cerevisiae BY4741 was isolated, and the POX1 gene was amplified, ligated into a T-vector, and transformed into E. coli JM109 to obtain the recombinant plasmid. Subsequently, primers were designed to amplify: (1) the linear vector fragment containing POX1 homologous arms, and (2) Part1 and Part2. Finally, the complete gene-integrated cassette was constructed via Gibson assembly.

The construction procedures for Part1 and Part2 were performed as follows: First, the coding sequences of Phytoene desaturase (CarB, GenBank: AJ238028.1) from Mucor lusitanicus CBS 277.49 and CarRP (GenBank: AJ250827.1) from Mucor circinelloides f. lusitanicus were codon-optimized according to the codon bias of S. cerevisiae. Appropriate restriction sites were incorporated before chemical synthesis of these optimized genes. Subsequently, promoter regions (PTDH3 and PTEF1) and terminator sequences (TCYC1 and TADH1) were PCR-amplified from chromosomal DNA of S. cerevisiae BY4741 and inserted into the plasmid containing the codon-optimized CarRP and CarB genes. Using plasmid Ts-PTDH3-carRP-TCYC1 as template, we PCR-amplified the carRP open reading frame (ORF) cassette PTDH3-carRP-TCYC1 with designed primers. Similarly, the CarB ORF cassette PTEF1-carB-TADH1 was amplified from plasmid Ts-PTEF1-carB-TADH1. Finally, these two fragments were assembled by fusion PCR.

CRISPR-Cas9-mediated Gene Integration

The S. cerevisiae BY4741 host strain used in this study was a S288C-derivative laboratory strain, whose genotype is MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0. The knockout of URA3 gene (URA3 encodes orotidine-5’-phosphate decarboxylase, a key enzyme in the uracil biosynthesis pathway), resulting in an uracil auxotrophic mutant, which enables subsequent utilization of URA3 as a selectable marker. To integrate the CarB and CarRP genes into the POX1 locus of the genome, we first designed sgRNAs targeting POX1 using the CHOPCHOP online tool. Subsequently, the Cas9 gene and sgRNA were assembled into plasmid pRS426 via Gibson assembly, yielding the recombinant plasmid pRS426-Δ_POX1_. The gene integration cassettes were then co-electroporated with pRS426-ΔPOX1 into competent cells of S. cerevisiae BY4741. Following incubation in YPD liquid medium for 20 hours, the transformed cells were plated onto minimal medium (MM) for selection.

Screening of Recombinant Strains and Marker Gene Excision

The URA3 gene encodes orotidine-5’-phosphate decarboxylase (OMP decarboxylase). Loss of URA3 function renders mutant strains unable to grow on minimal medium (MM) lacking uracil, though they can grow on supplemented medium (SM) (MM containing 60 mg/L uracil) or SM containing 5-fluoroorotic acid (FOA medium). Successful transformants that regain OMP decarboxylase activity through gene integration can grow on MM. Notably, the introduced URA3 marker contains homologous sequences at its 5’ end that may facilitate self-excision during replication. Since OMP decarboxylase converts FOA into toxic 5-fluorouracil (5-FU), recombinant strains that excise the URA3 marker become FOA-resistant. To validate the engineered strains, 20 colonies grown on MM were initially screened by colony PCR followed by agarose gel electrophoresis analysis (Figure 3).

Positive transformants were inoculated into 20 mL YPD medium in 100 mL flasks and cultured at 30°C with shaking (200 rpm) for 48 h. Subsequently, 100 μL aliquots of the cultures were spread onto FOA selection plates and incubated at 30°C for 6 days to select for positive recombinants. Potential β-carotene-producing strains were further screened by replica-plating single colonies onto both MM and SM plates. Colonies exhibiting the desired MM- SM+ auxotrophic phenotype were streaked onto YPD plates for isolation. Final strain confirmation was achieved through colony PCR amplification and DNA sequencing of the integrated carotenoid biosynthesis pathway genes.

One validated recombinant strain, designated SC-1 (Figure 4), was selected as our optimal β-carotene-producing strain for subsequent characterization and fermentation studies.

Fermentation of β-Carotene de novo Synthesis Strains

Retrieve the β-carotene-producing strain SC-1 from the cryopreservation tube. Streak it onto a YPD agar plate and incubate upside-down at 30°C for 1-2 days until colonies reach a diameter >1 mm.Place the streaked YPD agar plate upside-down in a 30°C incubator. Incubate for 24-48 hours under static aerobic conditions. Inoculate a single colony into a 100 mL flask containing 20 mL of YPD medium. Incubate at 30 °C with shaking (200 rpm) for approximately 20 h. Transfer 40 μL of the culture to a 1.5 mL centrifuge tube and dilute 20-fold with deionized water. Measure the absorbance until the OD600 reaches 10-15. Inoculate 100-200 μL of the seed culture into a 250 mL flask containing 15 mL of YPD60 medium (initial OD600 ≈ 0.1). Ferment at 30 °C, 200 rpm for 3 days. As shown in Figure 5, the culture medium turned yellow after 1 day of fermentation (Figure 5A), with the yellow color intensifying by the second day (Figure 5B). By the third day, an orange hue became clearly visible (Figure 5C).

β-Carotene Extraction

Cell pellets were obtained by centrifuging 1.5-2.0 mL of fermentation broth (Figure 6A) at 12,000 × g for 3 min (4°C) (Figure 6B), followed by two washes with distilled water. Acid hydrolysis was performed by resuspending pellets in 1 mL of 3 M HCl, vortexing vigorously for 15 sec, and incubating at 99°C for 5 min in a dry bath. Reactions were quenched immediately on ice for 3-5 min. Hydrolyzed samples were centrifuged (12,000 × g, 4°C, 3 min), washed once with distilled water to remove residual acid, and extracted with acetone (Figure 6C). Combined acetone fractions were stored in light-protected glass vials at −20°C to prevent degradation.

HPLC Analysis

Acetone extracts were filtered through 0.22 μm PTFE membranes (Millipore) into HPLC vials prior to analysis. Quantification was performed using an Agilent 1260 HPLC system equipped with a XDB-C18 column (5 μm, 4.6 × 250 mm; Agilent Technologies) maintained at 25°C. The mobile phase (35% acetonitrile:35% methanol:30% ethyl acetate) was delivered isocratically at 1.0 mL/min. β-Carotene was detected at 450 nm with a 20 μL injection volume (20 min run time). Calibration curves (10-50 mg/L) were generated from freshly prepared acetone standards (100 mg/L stock, stored at −20°C in amber vials). The HPLC analysis was performed to detect β-carotene and lycopene standards, which were used to identify the retention times of the synthesized products (Figure 7). Based on the standard calibration curve, the β-carotene titer produced by SC-1 in shake-flask fermentation was calculated to reach 129.1 mg/L.

Enhanced β-Carotene Production via Lipid Droplet Compartmentalization Strategy

Compartmentalization Strategy

We first identified and functionally characterized lipid droplet (LD)-targeting signal peptides, then successfully constructed integration cassettes encoding key enzymes fused with these LD-directing signals. Building upon the SC-1 strain, we spatially redirected the β-carotene biosynthetic pathway to LDs (Figure 8). Shake-flask fermentation demonstrated that this compartmentalization strategy significantly improved β-carotene production, yielding our second-generation engineered strain SC-2 with enhanced productivity.

Identification of Lipid Droplet-Targeting Signal Peptides

The substantial synthesis of terpenoids often exerts cytotoxic effects due to intracellular accumulation. To further improve β-carotene production, we implemented a subcellular compartmentalization strategy by targeting its biosynthetic pathway to lipid droplets (LDs). As dynamic organelles, LDs provide an ideal hydrophobic environment within their neutral lipid core for the synthesis and storage of nonpolar molecules, including key intermediates (GGPP, phytoene, and lycopene) and the final product β-carotene. This spatial organization offers two critical advantages: (1) preventing cytotoxic aggregation of hydrophobic metabolites in the cytosol, and (2) enabling direct storage of β-carotene within LDs, thereby avoiding its disruptive accumulation in membrane systems (e.g., endoplasmic reticulum) that could interfere with essential metabolic processes.

To verify the subcellular localization function of the selected signal peptide HD1, HD2, HD3 and HD4, we constructed a fusion protein by linking HDn to the C-terminus of the green fluorescent protein (GFP) reporter via a flexible linker (BBa_243ZON2D). The expression cassette was subsequently integrated into the POX1 genomic locus of S. cerevisiae using CRISPR-Cas9-mediated genome editing. Positive transformants were isolated and subjected to fermentation for functional characterization.

To verify the LD-targeting capability of the HDn signal peptide, we performed colocalization studies using Nile Red staining and GFP fluorescence. Cells in mid-log phase (OD600 = 2.0) were harvested and stained with 1 μL Nile Red (5 μmol/L in DMSO) for 10 min at room temperature in the dark. After three washes with PBS (pH 7.4), the cells were resuspended in 200 μL PBS.

For microscopy observation, 10 μL of cell suspension was mounted on glass slides and covered with coverslips sealed with nail polish. Samples were immediately analyzed using a Zeiss LSM 980 confocal laser scanning microscope (Carl Zeiss AG, Germany) with the following settings: GFP channel (excitation 488 nm, emission 510 nm) and Nile Red channel (excitation 569 nm, emission 593 nm). By comparing the green fluorescent localization of GFP with the red fluorescence localization of Nile Red staining, we determined that HD2 effectively localizes to lipid droplets.

Lipid Droplet-Targeted β-Carotene Biosynthetic Pathway

We engineered SC-2 by fusing the HD2 signal peptide to the C-termini of CarB and CarRP via GGGGS linkers, constructing integration cassettes (BBa_259LM0HD) that were subsequently targeted to the POX1 locus using our established CRISPR-Cas9 protocol. Transformants were validated through diagnostic PCR and Sanger sequencing of junction regions, confirming successful assembly of a lipid droplet-localized β-carotene pathway.

Fermentation and β-Carotene Determination of Engineered SC-2

The engineered β-carotene-producing strain SC-2 was cultured in YPD60 medium (30°C, 200 rpm, 3 days). After fermentation, cells were harvested, acid-hydrolyzed (3 M HCl, 99°C, 5 min), and extracted with acetone. The extract was filtered (0.22 μm PTFE) and analyzed via HPLC (C18 column, 35% acetonitrile, 35% methanol, 30% ethyl acetate, 1.0 mL/min, 450 nm detection). Compared to the first-generation strain SC-1 (129.1 mg/L), SC-2 achieved a significantly higher β-carotene titer of 314.8 mg/L, demonstrating a 2.4-fold improvement in production (Figure 10).

Scale-Up Fermentation in a 5-Liter Bioreactor

The engineered strain SC-2 was cultivated in a 5 L bioreactor under fed-batch fermentation conditions (30°C, pH 5.5, aeration rate 2-4 vvm, agitation speed 400-800 rpm, DO maintained above 20%). We conducted the first scale-up fermentation of SC-2, employing an online monitoring device to track real-time color changes in the fermentation broth (https://2025.igem.wiki/nacis-shanghai/hardware/). After 240 hours of fermentation, β-carotene production reached its peak concentration of 1.3 g/L at the 192-hour (Figure 11).

Based on the correlation established between color and yield, we conducted the second scale-up fermentation. During the fermentation, We optimized the dodecane addition timing and improved the feeding strategy, and β-carotene production was estimated using the predictive model. The fermentation process was ultimately terminated at 312 hours. HPLC analysis revealed that the peak β-carotene concentration occurred at 288 hours. As shown in Figures 9, rapid cell growth was observed within the first 84 h. Subsequently, OD600 remained stable between 110-140. From 48 h to 288 h, β-carotene production increased steadily, reached 1.8 g/L by 288 h, representing a 5-fold increase compared to shake-flask fermentation (Figure 12). Overall, this color-yield model significantly reduced the time required and provided an accurate prediction for the optimal fermentation termination timing.

Reference

1. Guo, Q., Peng, Q. Q., Li, Y. W., Yan, F., Wang, Y. T., Ye, C., & Shi, T. Q. (2024). Advances in the metabolic engineering of Saccharomyces cerevisiae and Yarrowia lipolytica for the production of β-carotene. Critical Reviews in Biotechnology, 44(3), 337-351.
2. Dasso, M. E., Centola, C. L., Galardo, M. N., Riera, M. F., & Meroni, S. B. (2025). FSH increases lipid droplet content by regulating the expression of genes related to lipid storage in Rat Sertoli cells. Molecular and Cellular Endocrinology, 595, 112403.
3. Zhao, Y., Zhang, Y., Nielsen, J., & Liu, Z. (2021). Production of β-carotene in Saccharomyces cerevisiae through altering yeast lipid metabolism. Biotechnology and Bioengineering, 118(5), 2043-2052.
4. Greenspan, P., Mayer, E. P., & Fowler, S. D. (1985). Nile red: a selective fluorescent stain for intracellular lipid droplets. The Journal of cell biology, 100(3), 965-973.