Investigating the Potential of Division of Labor in Synthetic Bacterial Communities for the Production of Violacein

Abstract

With advancements in synthetic biology and metabolic engineering, microorganisms can now be engineered to perform increasingly complex functions, which may be limited by the resources available in individual cells. Introducing heterologous metabolic pathways introduces both genetic burden due to the competition for cellular transcription and translational machinery, as well as metabolic burden due to the redirection of metabolic flux from the native metabolic pathways. Division of labor in synthetic microbial communities offers a promising approach to enhance metabolic efficiency and resilience in bioproduction. By distributing complex metabolic pathways across multiple subpopulations, the resource competition and metabolic burden imposed on an individual cell are reduced, potentially enabling more efficient production of target compounds. Violacein is a high-value pigment with antitumor properties that exemplifies such a challenge due to its complex bioproduction pathway, imposing a significant metabolic burden on host cells. In this study, we investigated the benefits of division of labor for violacein production by splitting the violacein bioproduction pathway between two subpopulations of -based synthetic communities. We tested several pathway splitting strategies and reported that splitting the pathway into two subpopulations expressing VioABE and VioDC at a final composition of 60:40 yields a 2.5-fold increase in violacein production as compared to a monoculture. We demonstrated that the coculture outperforms the monoculture when both subpopulations exhibit similar metabolic burden levels, resulting in comparable growth rates, and when both subpopulations are present in sufficiently high proportions.

Introduction

Synthetic biology aims to engineer existing and develop new biological systems that can perform novel and useful functions. Recent progress in the field has vastly enhanced our capability to engineer organisms to perform desired functions. Historically, synthetic biology and biotechnology predominantly focus on modifying single organisms to perform desired functions in a consolidated bioprocess. However, engineering additional functionality into a living cell can increase genetic burden, the competition for cellular resources for the transcription and translation of heterologous proteins, as well as metabolic burden, the burden on the cell due to competition for energy (e.g., ATP), cofactors (e.g., NAD­(P)­H) and metabolites, which leads to reduced growth rates, reduced expression rates and provokes genetic instability making engineered cells more prone to negative selection than wild-type strains. − In nature, cells seldom grow in isolation, and rather exist in diverse interacting communities. This has been shown to confer many advantageous properties to the members of these consortia. Microbial communities demonstrate increased robustness to environmental perturbations and resilience to mutation. There is extensive exchange of metabolites between the members of the communities, with reduced metabolic burden on individual members and increased cooperation. , Inspired by these properties of natural microbial consortia, synthetic microbial cocultures are being developed to overcome the limitations of monocultures. − Synthetic microbial cocultures can mitigate the limitations of metabolic burden in engineered monocultures by performing division of labor, where complex pathways are distributed between different cell types, allowing each to specialize in a subset of reactions. Synthetic cocultures implementing division of labor also reduce the burden due to heterologous enzyme expression. Division of labor can also allow modulation of sections of the pathway, by controlling the activation of different reaction, expressing them in specialist organisms at desired growth rates and subpopulation ratios. ,

Violacein is a purple-hued pigment produced by a wide range of naturally occurring bacteria found in a variety of environments ranging from deep seas, to forests and even polar glacial reserves. − It has been demonstrated to possess anticancer and antibacterial properties particularly against Gram-positive bacteria, including some antibiotic resistant strains of and is also used to prevent and treat stomach ulcers. , Due to its strong coloration, it is also used as a biodye. Due to its extensive applications but limited yields in bioproduction, it is a high value product. Violacein is produced from tryptophan by a five step enzymatic pathway through the expression of the gene vioABCDE (Figure A). First, tryptophan is converted to 2-Imino-3-(indol-3-yl) propanoate (IPA Imine) by the action of the protein VioA. The IPA Imine then dimerizes into an IPA dimer catalyzed by the protein VioB. The IPA dimer is subsequently converted to protodeoxyviolaceinic acid (PDA) catalyzed by VioE, which is then converted to protoviolaceinic acid (PVA) catalyzed by VioD. Finally, PDA and PVA are converted into deoxyviolacein and violacein, respectively, by the action of VioC. The production of violacein in has been demonstrated to have significant growth deficits on the cell, showing that the heterologous gene expression of the pathway is burdensome on the cell.

1.

(A) Violacein bioproduction pathway. The five gene pathway converts tryptophan to violacein and deoxyviolacein. Abbreviation: IPA: 2-Imino-3-(indol-3-yl)­propanoate. (B) Violacein pathway split strategies. The proposed pathway splitting strategies for violacein production are demonstrated in the two subpopulations. The green cell represents a tryptophan overproducing strain used to produce violacein precursors. Strategy 1 refers to the pathway splitting at tryptophan. Strategy 2 refers to the pathway splitting at PDA. Strategy 3 refers to the pathway splitting at PVA. Strategy 4 shown here is the production of violacein using a monoculture, with the entire violacein pathway expressed in the tryptophan overproducing strain, to be used as a control for coculture experiments. (C) Growth rates of the strains developed in this study, calculated mid log phase in a microplate reader. The monoculture has significant growth defects, compared to the wild-type strain. The VioABE and VioDC strains have similar growth rates, and show significant improvement in the growth rates compared to the monoculture. Relative growth rates are calculated at mid log phase of growth (OD600). Values shown are n = 3 biological replicates with mean value shown and error bars representing standard deviation. Statistically significant differences were determined using two-tailed Student’s t test (* represents p < 0.05, ** represents p < 0.01, *** represents p < 0.001, **** represents p < 0.0001, ns represents not significant). Symbols above the bars represent the statistical significance of the difference with the monoculture. OD600 measured using a microplate reader.

In this study, we investigated the advantages of division of labor in the context of the violacein biosynthetic pathway expressed in an based two member community. We investigated different pathway splitting points and different coculture compositions and identify conditions where the coculture performs better than the monoculture.

Materials and Methods

Bacterial Strains and Plasmids

strains DH10B (F- mcrA Δ­(mrr-hsdRMS-mcrBC) ϕ80lacZ ΔM15 ΔlacX74 recA1 endA1 araD139 Δ­(ara-leu)­7697 galU galK λ– rpsL­(StrR) nupG) and DH5α (F ϕ80lacZDM15 Δ­(lacZYA-argF) U169 recA1 endA1 hsdR17 (rk–, mk+) phoA supE44 l–thi–1 gyrA96 relA1) were used for cloning of all violacein constructs and grown in Luria–Bertani (LB) media at 37 °C with appropriate antibiotics (100 μg/mL Ampicillin, 10 μg/mL Tetracycline, 50 μg/mL Kanamycin). For expression of the violacein plasmids, BL21 (DE3) (F = ompT hsdSB (rB-mB-) gal dcm (DE3)) was used. For tryptophan overproduction, strain, B6, BL21 (DE3) (F–ompT hsdSB (rB- mB-) gal dcm (DE3) ΔtrpR ΔtnaA) was received from Prof. Zhang at Tsinghua University. The violacein genes were obtained from the Koffas Lab (pETM6-G6-vioABECD)­on (Addgene)­(Addgene plasmid # 66536; http://n2t.net/addgene:66536; RRID:Addgene_66536). The two populations for cocultures were tagged with sfGFP and mRFP1 expressed constitutively in the plasmids. The plasmids created in this study are detailed in Table S1. The bacterial strains used in this work are detailed in Table S2. The oligos used in this study are listed in Table S3. Polymerase chain reactions and Gibson and Golden Gate assemblies were used to build those plasmids. The cloning strategies used for the plasmids constructed in this study are shown in Figures S1–S7. All plasmid sequences were verified using Sanger sequencing and Oxford Nanopore whole plasmid sequencing. The plasmid for tryptophan over production, pHM068, and the TrpED genes were amplified from the MG1655 genome using the oligos TrpED_Fwd and TrpED_Rev. Two point mutations were introduced using around-the-world PCR with the oligo pairs Ser40LeuT_Fwd - Ser40LeuT_Rev, and Met293ThrC_Fwd - Met293ThrC_Rev. The plasmid was constructed with the TrpED cassette and a sfGFP cassette using Start–Stop Golden Gate assembly. The violacein genes were amplified along with the low strength T7 mutant promoter, G6, and RBS from the pETM6-G6-vioABECD plasmid using the oligos mentioned in Table S3. The plasmids were constructed by using restriction digestion and ligation. The plasmid pHM120 was constructed with mRFP using Start–Stop Golden Gate Assembly.

Culture Conditions

All experiments described in this article were conducted in Lysogeny Broth, with appropriate antibiotics (100 μg/mL Ampicillin, 10 μg/mL Tetracycline, and 50 μg/mL Kanamycin). The experiments were conducted at two different volumes, 5 mL of liquid culture in 14 mL round-bottom culture tubes and 10 mL of liquid culture in 50 mL conical flasks. From overnight liquid cultures, fresh subcultures were made at OD600 of 0.01 and inoculated at 37 °C for 5 h. After 5 h, 0.1 M Isopropyl β-d-1-thiogalactopyranoside (IPTG) was added to the cultures, and for the 5 mL cultures in 14 mL tubes, the lids were replaced by autoclaved foam stoppers for consistent and sufficient aeration. The cultures were then incubated at 20 °C for 19 h.

The growth experiments were conducted in clear flat-bottomed 96 well plates in a TECAN Spark microplate reader. From overnight liquid cultures of the strains, fresh subcultures were made in M9 medium supplemented with 10% casamino acids and 0.8% glucose. To each well 200 μL of culture volumes were added with starting OD600 of 0.025 and the plate was covered with a Breath-Easy membrane (Diversified Biotech). The plates were grown in the plate reader for 16 h with readings every 15 min at 37 °C with double orbital shaking and 1.5 mm amplitude. For each strain, three biological replicates were tested, and growth curves were fitted to a Gompertz curve, which was used to calculate the growth rate at mid log phase.

Metabolite Purification and Quantification

For the quantification of violacein and deoxyviolacein, at the end of the experiment after the cultures were incubated at 20 °C for 19 h, the metabolites were extracted in absolute methanol. One mL of liquid culture was pelleted in a 1.5 mL microcentrifuge tube in a benchtop centrifuge, and the supernatant was discarded. The pellet was resuspended in 1 mL of absolute methanol and transferred to a 2 mL screw cap microtube with glass beads. The tubes were shaken at 6000 rpm for 5 min, with 15 s breaks every minute, using a Precellys Evolution Homogenizer. The samples were spun down in a benchtop centrifuge, and the supernatant was transferred to a HPLC vial. The samples were analyzed in a Vanquish Core HPLC system with the YMC Carotenoid C30 column (150 × 4.6 mmI.D. S-3 μm). The mobile phases used were acetonitrile (A) and water (B), both containing 0.1% formic acid. The mobile phase was run with a flow rate of 1 mL/min with the following gradient: 0 min, 5% A; 1 min, 5% A; 5 min, 35% A; 7 min, 55% A; 9 min, 95% A; 10 min, 5% A; 12 min, 5% A. Violacein, with a retention time of 2.63 min and deoxyviolacein, with a retention time of 3.03 min were analyzed by peak area integration at 565 nm using a standard curve.

Flow Cytometry

Flow cytometry was used to determine the composition of the cocultures at different time points. Cell fluorescence was measured using an Attune NxT flow cytometer (Thermo Scientific) using the following parameters: FSC 660 V, SSC 500 V, violet laser VL1 (405 nm ex./440(50) nm em.) 420 V, blue laser BL1 (488 nm ex./530(30) nm em.) 450 V, yellow laser YL2 (561 nm ex./620(15) nm em.) 560 V. 10,000 cells were counted for each sample and the data was analyzed using FlowJo. The population was gated using the FSC-H and SSC-H channels, and singlets were identified using the FSC-H and FSC-A channels. The GFP and RFP tagged populations were separated by plotting the BL1-H and YL2-H channels. The gating used for determining the coculture composition is shown in Figure S11.

Results and Discussion

Identifying Pathway Splitting Candidates

The violacein bioproduction pathway is a five-gene pathway downstream of tryptophan (Figure A). As tryptophan is vital for the production of violacein, we constructed a plasmid with the genes TrpED expressed under the control of a pTet promoter (pHM068), in a tryptophan-overproducing strain of BL21­(DE3), referred to in this text as the B6 strain (Figure S8). We introduced the genes vioA, vioB, vioC, vioD, and vioE expressed under the control of a low strength mutant T7 inducible promoter, G6, on a plasmid, along with pHM068 into the B6 strain. On performing growth culture experiments, we found that there is significant growth defect (p < 0.0001) on the expression of the pathway as compared to the wild-type BL21­(DE3) strain (Figure C). This demonstrates the genetic and metabolic burden on the cell on introduction of the pathway. To tackle this metabolic burden, we selected three different pathway splitting points where the first strain produces an intermediate, which is transported out of the first strain and is taken up by the second strain expressing the rest of the pathway to produce the final product, violacein. In each of these cocultures, the first strain is the Trp strain with the TrpED_sfGFP plasmid (pHM068) and the second strain is the BL21­(DE3) strain with the mRFP plasmid (pHM120) and the violacein genes are expressed on a plasmid.

The three splitting strategies are Trp:VioABEDC, TrpVioABE:VioDC, and TrpVioABED:VioC (Figure B). We used the monoculture that expressed the entire pathway in the tryptophan overproducing Trp strain as a control. The strains were grown in a microplate reader to measure their growth behavior­(Figure C). We saw that while the introduction of the entire violacein pathway significantly reduces the growth rate of the monoculture strain, splitting the pathway alleviates this growth defect, resulting in less pronounced growth reductions compared to the monoculture. The Trp:VioABEDC split was selected because tryptophan is a widely used metabolite in the cell and an important branching point through which metabolic flux is diverted into the violacein pathway. Tryptophan has also been shown to traverse the cell membrane via well-characterized transporters in . The TrpVioABE:VioDC split was selected because in this split both strains were expected to have similar metabolic burden levels and hence similar growth rates. We also chose the TrpVioABED:VioC split due to the nature of the pathway, if the first cell produces PVA which is transported to the second strain, this would allow for the production of pure violacein, without deoxyviolacein, which is proven to be challenging to produce using microbial bioproduction.

Investigation of Different Pathway Split Mechanisms

We then constructed the coculture strains with the violacein pathway genes expressed under the control of the inducible low strength mutant T7 promoter, G6. To monitor the composition of the coculture, the two strains were also tagged with sfGFP and mRFP respectively, expressed constitutively on a separate plasmid. The coculture strains constructed were then tested in comparison to those of the monoculture strain for violacein production. The three cocultures developed were inoculated with an initial ratio of 1:1 for the two subpopulation strains and cultured for 24 h, and violacein was extracted. This experiment was conducted at two different tryptophan expression levels, with and without the overexpression of TrpED. For inducing the tryptophan overproducing genes, i.e., for tryptophan overexpression, the culture was supplemented with 100 nM aTc. We found that the TrpVioABE:VioDC split without TrpED overexpression produced the highest violacein titers (25.4 mg/L) among the cocultures and achieved slightly higher titers (not statistically significant) as compared to the monoculture (17.3 mg/L) (Figure A). Furthermore, we found that TrpVioABE:VioDC had similar final cell concentrations as the monoculture, which suggests that the productivity of the violacein producing cells is higher than that of monocultures (Figure B). We found that the final titers of the monoculture with the overexpression of TrpED were comparable to the case without induction of TrpED, but there were significant growth defects for the monoculture (p < 0.05) on the overexpression of TrpED. This suggests that tryptophan concentration may not be the limiting factor for violacein production, as overexpressing TrpED does not lead to increased violacein titers. This may be due to the increased metabolic burden caused by TrpED overexpression, which could negatively impact violacein production more than the benefit gained from the additional tryptophan. Hence, as the violacein titers as well as the cell concentrations were lower for the cultures with TrpED overexpression, for the following section we proceeded with the cultures without TrpED overexpression.

2.

Violacein production by monoculture versus different pathway split coculture with and without overexpression of tryptophan. (A) Violacein and deoxyviolacein titers for the monocultures and cocultures. (B) Final cell concentrations attained by the cultures at the end of experiment at 24 h. For tryptophan induced case, the culture was supplemented with 100 nM aTc at the time of subculture. All experiments were conducted in 5 mL liquid culture in 14 mL culture tubes. All cocultures started with 1:1 initial composition at the time of subculture. Values shown are n = 3 biological replicates with mean value shown and error bars representing standard deviation. Statistically significant differences were determined using two-tailed Student’s t test. The differences between violacein titers of the monoculture and VioABE + DC cocultures in the tryptophan uninduced condition is not significant. The differences between violacein titers of the monoculture and Trp + VioABEDC cocultures in the tryptophan induced condition is not significant.

Coculture Ratios Optimize Violacein Production

In order to identify the optimal coculture conditions for violacein production, we tested several coculture inoculation ratios for the three coculture systems. We tested the following initial coculture compositions for the three coculture systems: 1:9, 1:3, 1:1, 3:1 and 9:1. We found that, as with 1:1 inoculation ratio, the TrpVioABE:VioDC split performed the best of the three splits (Figure A). Moreover, the TrpVioABE:VioDC split coculture starting at 3:1 ratio, had the highest violacein titers (30.3 mg/L), producing significantly higher violacein titers (p < 0.05) than the monoculture (17.3 mg/L).

3.

Coculture composition controls the violacein titers. (A) Three cocultures Trp:VioABEDC, VioABE + VioDC, and VioABED + VioC cultured with five initial inoculation ratios of 1:9, 1:3, 1:1, 3:1 and 9:1. We see that the VioABE + VioDC split produces significantly more violacein than the monoculture. Values shown are mean violacein titers of three biological replicates. (B) Violacein titers plotted against the percentage of the VioABE subpopulation in the VioABE:DC coculture. The violacein titer shows a nonlinear (third order polynomial) correlation with R ² value = 0.74. (C) Composition of the VioABE + VioDC coculture trajectories over the course of the experiment. The colors of the trajectories represent the final violacein titers­(mg/L) of the coculture measured after 24 h of culture. The cocultures with starting compositions between 40:60 and 65:35 produce higher titers, depicted here in green. Statistically significant differences were determined using two-tailed Student’s t test (* represents p < 0.1).

Based on the results for the TrpVioABE:VioDC coculture, we proceeded to identify the ideal coculture compositions for maximum violacein titers. Figure B shows the correlation between the final violacein titers and the final coculture composition after 24 h of culturing. We found that the final coculture composition has a very strong impact on the final violacein titers. We observed that the final composition of 60:40 ratio produced the highest violacein titers for the TrpVioABE:VioDC coculture split. This outcome may result from a balanced conversion of intermediates and effective metabolite exchange. At lower ratios, insufficient intermediate accumulation might limit the VioDC strain’s ability to convert intermediates into violacein, while at higher ratios, the reduced number of VioDC cells could limit maximum conversion efficiency. Furthermore, we saw that the ideal coculture composition for maximum violacein production is different for different splits (Figure S10). We also observed that the coculture composition changes through the course of the experiment (Figure C). We found that the cocultures diverge from the inoculation composition and very similar starting compositions can lead to different final compositions as well. For the TrpVioABE:VioDC coculture, we see that for the maximum violacein titers, the most suitable range of initial coculture composition was found to be between 40:60 and 65:35 (Figure C).

While the TrpVioABE:VioDC coculture split produced violacein titers comparable to the monoculture, the other two splits Trp:VioABEDC and TrpVioABED:VioC have lower violacein titers. For the Trp:VioABEDC split, the highest violacein production is observed for the starting composition ratio of 1:9 (Figure A). This can be explained by the fact that as all of the violacein producing genes are expressed in the second strain, there needs to be sufficient concentrations of the second strain for maximum violacein production. Moreover, the overproduction of tryptophan from the first strain did not compensate for the reduced concentration of the second strain in the coculture. We also observed that as the growth rate of the Trp strain is higher than the growth rate of the VioABEDC strain, over time, the concentration of the Trp strain increases in the coculture (Figure S10). In the case of the TrpVioABED:VioC coculture, for violacein production by the VioC strain, the TrpVioABED strain needs to produce sufficient PDA and it needs to be exported out of the TrpVioABED strain and imported into the VioC strain. This might be the limiting step here, leading to inefficient violacein production by the coculture. The results from the coculture experiments with the overexpression of TrpED can be found in Figure S9. We observed that the overexpression of TrpED also led to similar results, with the TrpVioABE:VioDC producing the highest violacein titers. Interestingly, we found that the highest titers were produced for cocultures with the initial coculture composition of 9:1. Figure S10 shows the final violacein titers observed for the different coculture split strategies, with and without TrpED overexpression. We observed that for the TrpVioABE:VioDC split, the curve for violacein titers against final coculture composition with TrpED overexpression is shifted to the right compared to the culture without TrpED overexpression. This suggests that in this case the ideal coculture composition for maximum violacein production is 70:30. We demonstrate that while the violacein titers are dependent on the coculture composition, the ideal coculture composition for maximum violacein production is different for different cocultures under different growth conditions.

Conclusions

With the advancement in the use of synthetic microbial communities, there is a rise in the development of control systems that allow the tuning of coculture composition. There is very limited study of the need of dynamic composition control for applications in bioproduction using microbial communities. In this study, we demonstrated the effect of the coculture composition on the final product titers and demonstrated the need for dynamic composition control for reproducible robust bioproduction of high value compounds. Recent studies have explored coculture-based violacein production by splitting the pathway at tryptophan and optimizing culture conditions or population ratios. − In contrast, this study investigates the role of division of labor in increasing product titers by investigating several pathway splitting strategies and different population ratios to improve titers.

In this study we systematically screened the solution space for division of labor for violacein production. Here, we examined key variables for optimizing division of labor, i.e., different pathway splitting points and coculture compositions to identify conditions where the coculture is optimized and produces the highest titers of Violacein. Among the tested splitting strategies, the pathway split TrpVioABE:VioDC emerged as the most effective, producing significantly higher violacein titers compared to other splits and the monoculture. The coculture produced the highest titers (30.3 mg/L) at a final composition around 60:40, which is significantly higher (p < 0.05) than the titers achieved for the monoculture (17.3 mg/L). We also found that different coculture splits exhibit different trends of violacein titers at different compositions. We demonstrated that a coculture performing division of labor can be more advantageous than a monoculture when two conditions are satisfied. First, that the burden is well distributed between the two subpopulations such that the strains have similar growth rates and both the subpopulations are present at high enough concentrations, and second, that the intermediate where the pathway is split is efficiently exchanged between the two subpopulations. We demonstrate the value of division of labor as a strategy to improve product yields for violacein, and other high-value compounds with complex biosynthesis pathways. This strategy can be combined with other pathway modifications and culture condition optimizations to improve production titers beyond the highest titers obtained yet.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acssynbio.5c00120.

• Lists of plasmids, strains, and oligos used in this study; cloning strategies used for the construction of the plasmids in this study; tryptophan titers for the overproducing strain; results for experiments conducted with tryptophan overproduction; and flow cytometry analysis method (PDF)

H.M., R.L.A., and G.-B.S. designed the study. H.M. performed the experiments and collected the data. H.M. analyzed the data. H.M. wrote the original draft. All authors reviewed and approved the final manuscript.

The authors declare no competing financial interest.