There are two parts of optimization in the TPE project:

We use directed evolution to develop the process of synthesis from two-cell system to one-cell system, with an expected optimization controlling system of endogenous TnaA enzyme in E.coli.

We introduce a reversible glucose protection group to make the production of tyrian purple dye more stable and convenient for industrial scale-up.

The reported scheme of tyrian purple production in E.coli with the substrate of Trp is shown in Figure 1. We designed our optimization based on this scheme. As natural dye indigo and tyrian purple is hard to solute in water and cause difficulty in cloth dying. We applied the strategy that introduce a glucose moiety as a reversible protecting group to the reactive indigo and tyrian purple precursor indoxyl by glucosyltransferase PtUGT to form soluble indican as fermentation product (Figure 2). And β-glucosidase BGL is applied to remove the protecting group when cloth dying.

Also, for the steps using TnaA and MO (using one of special kinds named flavin-containing monooxygenase, MaFMO or FMO), we linked them with a flexible linker to form a chimeric enzyme for higher field and efficiency.

Thanks to 2021 iGEM team LINKS_China for providing us with their parts: part BBa_K4011004 (TnaA-FL-FMO) and part BBa_K4011003 (Fre-SttH)

We constructed plasmid pET28a-FMO-PtUGT to co-express PtUGT and FMO to convert indole or 6-Br-indole to indican or 6-Br-indican respectively. We also construct plasmid pET28a-BGL and pET28a-FMO to individually express BGL and FMO.

We transformed parts TnaA-FL-FMO and Fre-sttH into DH5α E.coli which endogenous TnaA gene have been knocked out. After culture and induced expression of these strains, we use the SDS-PAGE to separate and show the expression of TnaA-FL-FMO and Fre-sttH (Figure 3A). We found that we successfully express our fused proteins and these two proteins show high water solubility.

We then culture strain that has been induced to express Fre-sttH by NPB solution and 2.5mM trp and 300mM NaBr as substrate. During fermentation of 24h, we use LC-MS to obtained data of concentration of trp and 6-Br-trp by every 6h (Figure 3B). Strain that has been induced to express TnaA-FL-FMO also be cultured by NPB solution and 1mM trp or 6-Br-trp as substrate. After fermentationfor 24h, we get obvious blue and purple dye in our culture tubes (Figure 3C). Besides, we attempt to produce tyrian purple by TnaA-FL-FMO using the culture after fermentation of trp and NaBr with Fre-sttH and the result has no difference from fermentation using standard 1mM 6-Br-trp (Figure 3D).

Considering the bad solubility of indigo and tyrian purple, we apply 100% DMSO to solute the dye from cell precipitation and we get DMSO solution of indigo and tyrian purple (Figure 3E). Dye concentration is calculated by a standard curve measured by LINKS_China. We get a ~30% yield for indigo and ~20% yield for tyrian purple. We also use the tyrian purple we produce in cloth dying by put a piece of cloth into fermentation culture of Br-trp with TnaA-FL-FMO and we finally get the cloth in purple (Figure 3F).

Due to TnaA can react with both tryptophan and the first step product bromotryptophan, Fre-SttH and TnaA has to be in two different strains. So the two-cell system, ΔTnaA E. coli strain without endogenous TnaA expression and another ΔTnaA E. coli strain with extra TnaA expression, are used to produce tyrian purple with less byproduct indigo. That’s why Lee et al have to use two-cell system to complete the production of tyrian purple.

Considering the problem of tryptophanase TnaA lack of specificity, we want to use muted trp repressor that can only bind trp to sense the concentration of trp and 6-Br-trp in E. coli to control the expression of TnaA. We optimized two-cell sysytem to one-cell system with a spatio-temporal well-controlled TnaA synthesis (Figure 4). We applied a synthetic phylogeny of programmable Trp repressor with directed evolution to E.coli, which simplified the process of reacting with bromotryptophan not tryptophan.

Compartmentalized partnered replication (CPR) is an emulsion-based directed evolution method based on a robust and modular phenotype–genotype linkage. In contrast to other in vivo directed evolution approaches, CPR largely mitigates host fitness effects due to a relatively short expression time of the gene of interest. CPR is based on gene circuits in which the selection of a 'partner' function from a library leads to the production of a thermostable polymerase. After library preparation, bacteria produce partner proteins that can potentially lead to enhancement of transcription, translation, gene regulation, and other aspects of cellular metabolism that reinforce thermostable polymerase production. Individual cells are then trapped in water-in-oil emulsion droplets in the presence of primers and dNTPs, followed by the recovery of the partner genes via emulsion PCR. In this step, droplets with cells expressing partner proteins that promote polymerase production will produce higher copy numbers of the improved partner gene. The resulting partner genes can subsequently be recloned for the next round of selection.

Iterative rounds of positive and negative compartmentalized partnered replication (CPR) led to the exponential amplification of variants that responded with high affinity and specificity to halogenated tryptophan analogs and novel operator sites. We empower cells to perceive the relative ratio of bromotryptophan to tryptophan to control TnaA synthesis.

We firstly constructed three plasmids pTrpR, pPOS and pNEG. pTrpR is induced to express muted trp repressors. Active trp repressors with its ligand can induce pPOS to express Taq polymerase and inactive trp repressors or trp repressor without its ligand can induce pNEG to express Taq polymerase (Figure 5).

For GFP test of CPR cycle, we constructed pNEG-sfGFP and pPOS-sfGFP which replace Taq polymerase gene with sfGFP to characterize the expression of Taq polymerase.

For the mock selection of CPR cycle, we inserted a DNA sequence of 200bp into wild type trp repressor to construct INAC-pTrpR (Figure 18). We mix plasmid pTrpR with wild type trp repressor and INAC-pTrpR with a mass ratio of 1:1 and transform the mixture into pNEG and pPOS competent cells. Induced cells were resuspend by CPR mix and emulsion droplets were constructed by mixing CPR mix and oil mix and vortex for 4 min.

(1)GFP test

The expression of Taq polymerase can not be monitored easily, we apply GFP assay to use GFP production to analyzing the gene circuits function. We transform pTrpR which express wild type trp repressor into competent cells that have been transformed with plasmid pNEG-sfGFP or pPOS-sfGFP. Then we culture and induce with different concentration of trp and 1mM IPTG to induce expression of sfGFP. We expect that as trp concentration increases, the expression of sfGFP (RPU/OD600) in pNEG-sfGFP gradually decreases, and the expression of sfGFP (RPU/OD600) in pPOS-sfGFP gradually increases (Figure 6A).

(2)Mock selection

In addition to GFP test, we also perform a mock selection to confirm the designed gene circuit function can actually select positive clone. At first, we mix culture with strains that has been induced to express sfGFP and oil mix to check the quality of emulsion droplet. We measure the diameter of droplet with a phase contrast microscopy. We found that the diameter of emulsion droplets are range about 2-10 μm and less than 20% emulsion droplets contain a GFP-expression cell (Figure 6B). Thus, we have checked any emulsion droplet can only contain a single cell. We insert a random sequence into wild type trp repressor to make it inactive to bind to trp, and we get plasmid INAC-pTrpR. We mix plasmid pTrpR with wild type trp repressor and INAC-pTrpR with a mass ratio of 1:1 and transform the mixture into pNEG and pPOS competent cells. A mock CPR seletion is next carried out. Induced strains is emulsified with emulsion mix and replicate positive clone by ePCR. The emulsion remains in a phase-independent state during ePCR (Figure 6C). After breaking the emulsion by adding ddH2O and chloroform, we apply a recovery PCR and perform a agarose electrophoresis to monitor the result of the mock selection. PNEG give an obvious amplification of inactive trp repressor and pPOS give an obvious amplification of active trp repressor (Figure 6D).

We found that high concentration of trp can inhibit expression of sfGFP in negative selection system. However, it can only provide about three times the inhibition effect which cannot achevie screen effect. Possibly, trp repressor has a low expression or trp repressor with trp cannot strongly bind to trp operator at DNA. Same experiments of positive selection system haven't be carry out.

Also, we have successfully make emulsion droplets of high quality. However, results of mock selection is not obvious which means selection system cannot distinguish wild type trp repressor with inactive trp repressor that consistent with results of GFP test.

After testing the CPR circuit works well, we are going to introduce directed evolution to trp repressor. Our libraries are designed in which residues (I57, V58, R54, L60, G85, T81, I82) that might contact trp at its 6C positions of the indole ring due to their proximity. Considering several position overlapped with those involved in the formation of the HTH motif, we initially choose I57, V58 and T81 as our mutant site for directed evolution. We perform site-directed saturation mutation on these three residues by completely randomized through re-synthesis of the gene with NNS codons (N = A, T, G, C; S = G, C). The library (32768 variants) are transformed into competent cells that have been transformed with pNEG or pPOS beforehand and carry out several circle of CPR to select positive clone we want (Figure 7). Due to the restriction of time, we have not perform a selective circle in our project, we plan to simulate this process using mathematical modeling.

More details in Drylab

At first, we modeling the process of Transcription and translation process. We used the ODE solver “deSolve” in R to simulate our positive selection system based off the 9 ODEs we wrote and “ggplot2” to plot the dynamics of molecules of interest, and obtained the results below: As shown by the graph (Figure 8A), within 2 hours, TrpR reaches its plateau at about 150 uM and at the same time TrpR gets rapid polymerization formation into TrpR2, which peaks at Time = 20h, 75 uM. Compared with TrpR and TrpR2, TrpR2-T's formation process is much lower, reaching it saturated at Time = 40h, 25 uM. In the CI, CI2 graph (Figure 8B), CI quickly reaches its peak at Time = 8h, 3.8 uM and then it go through a steady but slow decreasing process to about 1.7 uM. While for the CI2 protein, it has the similar trend with CI and its peak and steady state concentration is half of that of CI. While for the GFP graph (Figure 8C), we could see in the first day (within 24h), GFP quickly peaks at about 2 uM, while after that it experiences a significant drop to no more than 0.5 uM. Soon, GFP quickly get translation.

Then, we model the influence of Trp's concentration in directed evolution system X aix means Trp increase from 10-4 to 1, Y aix means relatively expression level (Figure 8D). In the left figure, we can see, when Trp's concentration is under 20 × 10-3 uM. CI2 protein is at its saturate state. When Trp increases to 30 × 10-1 uM, CI2 experiences a significant drop, from 2.0 uM to 0.5 uM. In the right figure, GFP's relative expression process is contrary to CI2 trend. Below Trp concentration 20 × 10-3 uM, GFP's expression is close to 0. But as trp's concentration increases.

With the help of modeling, we know better about this directional evolution and selection system. As seen in the results, our modeling shows how the positive screening system works. TrpR and TrpR2 quickly reach its saturate state with proper Trp concentration, and TrpR2 and Trp's polymer can generate, which can inhibit the expression of CI2. Also, Taq enzyme can be expressed. With increasing Trp concentration, the relative expression of GFP has a similar trend with experiment data in wetlab. According to this, we could estimate the combined rate between TrpR2 and Trp, which has a huge benefit to our gene design.

[1] Lee, J., Kim, J., Song, J. E., Song, W. S., Kim, E. J., Kim, Y. G., Jeong, H. J., Kim, H. R., Choi, K. Y., & Kim, B. G. Production of Tyrian purple indigoid dye from tryptophan in Escherichia coli. Nature chemical biology, 2021, 17(1), 104-112. DOI: https://doi.org/10.1038/s41589-020-00684-4
[2] Hsu, T. M., Welner, D. H., Russ, Z. N., Cervantes, B., Prathuri, R. L., Adams, P. D., & Dueber, J. E. Employing a biochemical protecting group for a sustainable indigo dyeing strategy. Nature chemical biology, 2018, 14(3), 256–261. DOI: https://doi.org/10.1038/nchembio.2552
[3] Ellefson, J. W., Ledbetter, M. P., & Ellington, A. D. Directed evolution of a synthetic phylogeny of programmable Trp repressors. Nature chemical biology, 2018, 14(4), 361–367. DOI: https://doi.org/10.1038/s41589-018-0006-7
[4] Abil, Z., Ellefson, J. W., Gollihar, J. D., Watkins, E., & Ellington, A. D. Compartmentalized partnered replication for the directed evolution of genetic parts and circuits. Nature protocols, 2017, 12(12), 2493–2512. DOI: https://doi.org/10.1038/nprot.2017.119