Our project aims to establish a cellular factory using E. coli by introducing enzymes to promote the efficient production of astaxanthin with high yield. We follow the Design-Build-Test-Learn (DBTL) cycle to guide this process. The DBTL cycle helps us continuously refine and optimize our ideas and procedures to enhance the production rate while increasing the yield of astaxanthin.
Figure 1：DBTL flow chart
To achieve astaxanthin production in E. coli through enzymatic action, we established a cellular factory by engineering a plasmid and then introducing it into E. coli, thereby constructing the astaxanthin biosynthesis pathway within the bacteria. Biosynthesis of astaxanthin in E. coli requires seven genes: CrtE, CrtB, CrtI, CrtY, CrtZ, CrtW, and idi. The T7 promoter offers extremely high transcription efficiency, driving abundant expression of the target proteins and boosting astaxanthin yield. Due to the large number of enzymes here, we employed a dual T7 promoter system (the expression plasmid pET-Duet-1, Figure 1), which helps reduce the accumulation of intermediate products. HpCrtE, PanCrtB, and PanCrtI were inserted into the multiple cloning site (MCS) of the first T7 promoter, and idi, PagCrtY, HpCrtZ, and BanCrtW were into the MCS of the second T7 promoter, thus to obtain the pET-ast plasmid (Figure 3).
Figure 2： The schematic diagram of pET-Duet-1 plasmid
The nucleotides for HpCrtE-PanCrtB-PanCrtI and idi-PagCrtY-HpCrtZ-BanCrtW cluster were chemically synthesized by Nanjing GenScript Biotech Co., Ltd (Nanjing, China). These two fragments were then inserted into the MCSs of pET-Duet-1, respectively, resulting in the pET-ast plasmid (Figure 3). This plasmid was transformed into E. coli TOP10 cells for proliferation and storage. The accuracy of gene sequences was verified by sequencing.
Figure 3：The schematic diagram of pET-ast plasmid
E. coli Top 10 cells harboring the pET-ast plasmid was inoculated into LB medium and cultured in a shaker at 37°C with 200 rpm for 12 hours. Cells were collected by centrifugation, and the pET-ast plasmid was extracted. The pET-ast plasmid was then transformed into E. coli BL21(DE3) via heat shock method for large-scale cultivation. IPTG was used for induction to promote astaxanthin synthesis in the E. coli.
The collected E. coli cells were freeze-dried under vacuum to obtain cell powder. We used 1 mL of methanol : isopropanol (8:2) solution to extract carotenoids. Subsequently, High-Performance Liquid Chromatography (HPLC) was employed for quantitative detection of carotenoids and LabSolution software was used to record the chromatograms and corresponding product peak areas. A calibration curve was generated by using known concentrations of commercial astaxanthin standards and their corresponding peak areas. The maximum absorption peak for β-carotene, canthaxanthin, and zeaxanthin was at a wavelength of 450 nm, while that for astaxanthin was at 475 nm. After obtaining the peak areas, they were applied to the standard curves to determine the astaxanthin concentration of unknown samples, which was then subjected to the final calculation of astaxanthin content in E. coli cells. The results indicated relative low yields of astaxanthin and a large amount of intermediate products (Figure 4). The overall suboptimal performance prompted us to proceed with further protein engineering.
Figure 4：Analysis of the Ability of pET ast to Synthesize Astaxanthin
Based on the experimental process, we can clearly conclude: In the study of using E. coli as a chassis, by introducing astaxanthin genes (such as CrtE, CrtB, CrtI, CrtY, CrtZ, CrtW, idi, etc.) and leveraging a highly efficient and controllable expression system (the combination of the T7 promoter and the BL21(DE3) strain) to build a cellular factory, we not only achieved stable maintenance of the astaxanthin biosynthesis pathway in E. coli but also experienced high-efficiency expression of target proteins through IPTG induction. Finally, HPLC analysis successfully detected the presence of astaxanthin in the products from E. coli and confirmed its accumulation to a certain level. This series of experimental results fully demonstrates the complete feasibility of establishing a cellular factory in E. coli for the targeted production of astaxanthin. It also lays a solid experimental foundation for further enhancing astaxanthin yield and promoting the industrial application of this technology.
Simultaneously, we observed a significant accumulation of β-carotene and canthaxanthin in addition to the targeted astaxanthin, revealed by HPLC qualitative and quantitative analysis of intracellular products in the E. coli. This phenomenon clearly indicates insufficient activity of the β-carotene hydroxylase, functioning to catalyze canthaxanthin into astaxanthin. This not only reduces the conversion rate to the target product astaxanthin, but may also adversely affect the host cell's membrane structure and metabolic homeostasis due to intermediate product buildup. Therefore, we decided to proceed with protein engineering in the next step, aiming to make the cellular factory more efficient and robust, and further increase astaxanthin yield.
Since the existing β-carotene hydroxylase is insufficient to support more efficient astaxanthin synthesis, we proposed the idea of designing mutants with higher catalytic activity.
Using AlphaFold and PrankWeb, we obtained model files of β-carotene hydroxylase and its two substrates (β-carotene and canthaxanthin). Molecular docking simulations were performed on the CB-Dock2 website to identify the amino acid distribution in the binding pocket of β-carotene hydroxylase. Each amino acid in the pocket region was mutated to alanine, and the docking interactions between the mutants and canthaxanthin were analyzed individually. The results indicated that mutations at ILE102ALA, SER96ALA, CYS191ALA, and THR213ALA may enhance the catalytic efficiency of β-carotene hydroxylase for canthaxanthin. Based on these findings, we proceeded to the second round of DBTL (Design-Build-Test-Learn).
The side chain of alanine is a simple methyl group (-CH₃), which is small and lacks other functional groups. Mutating to alanine may enlarge the pocket structure, facilitating better binding between the substrate and the enzyme, thereby promoting enzymatic reactions. In this phase of mutant improvement, we introduced alanine mutations targeting the pocket structure. By analyzing changes in the binding energy between the alanine mutants and canthaxanthin/β-carotene, we screened for β-carotene hydroxylase mutants with potentially enlarged pockets that could enhance astaxanthin production. The candidate mutation sites identified were ILE102ALA, SER96ALA, CYS191ALA, and THR213ALA.
We selected four sites for site-directed mutagenesis: ILE at position 102 mutated to ALA, SER at position 96 mutated to ALA, CYS at position 191 mutated to ALA, and THR at position 213 mutated to ALA. The alanine mutants were obtained through chemical synthesis, synthesized by Nanjing GenScript Biotechnology Co., Ltd., and then used to replace the HpCrtZ gene on the pET-ast plasmid, resulting in pET-astILE102ALA,pET-astSER96ALA，pET-astCYS191ALA， and pET-astTHR213ALA, respectively. The synthesized plasmids were verified again by sequencing and then transformed into E. coli.
Figure 6：Verification of β-carotene hydroxylase mutants by sequencing
E. coli strains containing pET-astILE102ALA, pET-astSER96ALA， pET-astCYS191ALA, and pET-astTHR213ALAwere inoculated in LB medium and cultured at 37°C with shaking at 200 rpm for 12 hours. The bacterial cells were collected by centrifugation, and plasmids were extracted using a ? kit. The plasmids were then transformed into E. coli BL21 (DE3) via heat shock method for expansion culture. IPTG was used to induce expression and promote astaxanthin synthesis in E. coli. Subsequently, the bacterial cells were harvested, intracellular pigments were extracted, and astaxanthin production was analyzed by HPLC. The results showed that the astaxanthin yield of pET-astILE102ALA was 1.13 times that of the wild-type, pET-astSER96ALA was 1.24 times, pET-astCYS191ALA was 1.8 times, and pET-astTHR213ALA was 1.27 times that of the wild-type. All these mutants exhibited a significant increase in astaxanthin production.
Figure 2：Effect of β-carotene hydroxylase alanine mutants on astaxanthin production
According to the docking simulation results, a significant improvement in the binding energy score was observed. This demonstrates that modifying the pocket region of the protein facilitates substrate binding, enhances the activity of β-carotene hydroxylase, and enables the identification of the most efficient mutant through functional analysis.