Team:SUSTech Shenzhen/Engineering
Part1: Engineering bacteria with RcsB can produce colanic acid to remedy diarrhea-related dehydration
Design Part:
Colanic acid plays a vital role in preventing dehydration in microorganisms1, where its biosynthesis is regulated by Rcs (regulator of capsule synthesis) phosphorelay system2. Rcs is a complex system comprising the sensory RcsC, receiver/effector RcsB, regulator RcsA, and phosphotransmitter RcsD 2-5 . Leveraging on this microbial behaviour, we intended to adapt the Rcs system regulation to facilitate microbial cytoplasmic water retention to prevent water loss.
Systematic verification of the colonic acid biosynthesis is conducted using an IPTG (Isopropyl-beta-D-thiogalactopyranoside)-inducible promoter in the plasmid pBbE6k-RFP6 where the induced RFP-tagged RcsB expression is validated via western blot followed by colanic acid production quantification via staining method and HPLC(High-performance liquid chromatography). As a proof-of-concept of this component of the overall circuit, the RcsB expression is regulated using IPTG, where our findings suggest good water retention supported by tunable RcsB expression using the different inducer concentrations.
In our subsequent experiments, the water storage of polysaccharides is tested using THEMT (American Society of Experimental Materials) 570 standard and other methods in other studies7.
Build Part:
The codon-optimized sequence of the RcsB was synthesized and cloned into the pBbE6k vector using an Escherichia coli (E. coli) host BL21. The RcsB gene was cloned to make use of the LacI-promoter for gene transcription and translation.
Figure 1| Restriction enzyme digestion products of pBbE6k plasmid and RcsB in pUC57
Figure 2| RcsB gene cloned under lacI promoter in pBbE6K plasmid
Test Part:
1.The fluorescence curve shows that IPTG can successfully induce the expression of RFP. Each fluorescence level is measured three times in 96-well plates.
Figure 3| IPTG-induced RFP expression in E.coli
2.We detected the expressed RcsB protein via electrophoresis and western blot. The RcsB on the western blot is indicated by the red box, showing a size of approximately 24.5kDa.
Figure 4| Western blot of RcsB
3.To primarily verify whether our engineered bacteria can produce colanic acid, we conducted the staining test to detect the polysaccharide. Bacteria cells treated with the fuchsin solution are stained red and capsules of bacteria treated with the methylene blue are stained blue.
Figure 5| Staining BL21 cells without/with RcsB expressed
Figure 6| Staining Accela cells without/with RcsB expressed
4.4.Currently, we are quantifying the colanic acid yield using Ultra-High Pressure Liquid Chromatography to isolate and determine its concentration. Further studies on the output would help in finetuning the colanic acid production within the cells to facilitate proper dehydration prevention.
Learn Part:
Due to travel restrictions during the COVID-19 pandemic and limited laboratory time, we have not succeeded in quantifying the yield of colanic acid in probiotic strain E. coli Nissle 1917 and testing the water holding capacity. Furthermore, it is difficult to determine the exact concentration due to the overlapping emission spectra of the various sugar components that make up colanic acid. However, we intend to adopt an alternative approach to stain colanic acid for easier identification8, where we can precisely determine the colanic acid yield via UHPLC assay or other optimized alternatives.
Reference
[1]Holowko, M. B.; Wang, H.; Jayaraman, P.; Poh, C. L., Biosensing Vibrio cholerae with Genetically Engineered Escherichia coli. ACS Synth Biol 2016, 5 (11), 1275-1283.
[2]Higgins, D. A.; Pomianek, M. E.; Kraml, C. M.; Taylor, R. K.; Semmelhack, M. F.; Bassler, B. L., The major Vibrio cholerae autoinducer and its role in virulence factor production. Nature 2007, 450 (7171), 883-6.
[3]Pan, Y.; Chua, N.; Lim, K.; Ho, C. L., Engineering of Human Lactoferrin for Improved Anticancer Activity. ACS Pharmacology & Translational Science 2021.
[4]Azman, A. S.; Rudolph, K. E.; Cummings, D. A.; Lessler, J., The incubation period of cholera: a systematic review. J Infect 2013, 66 (5), 432-8.
[5]Wagner, M.; Peterson, C. G. B.; Stolt, I.; Sangfelt, P.; Agnarsdottir, M.; Lampinen, M.; Carlson, M., Fecal eosinophil cationic protein as a marker of active disease and treatment outcome in collagenous colitis: A pilot study. Scandinavian Journal of Gastroenterology 2011, 46 (7-8), 849-854.
[6]Rivera, F. P.; Medina, A. M.; Bezada, S.; Valencia, R.; Bernal, M.; Meza, R.; Maves, R. C.; Ochoa, T. J., Bovine lactoferrin decreases cholera-toxin-induced intestinal fluid accumulation in mice by ganglioside interaction. PLoS One 2013, 8 (4), e59253.
[7]Moon, T. S.; Lou, C.; Tamsir, A.; Stanton, B. C.; Voigt, C. A., Genetic programs constructed from layered logic gates in single cells. Nature 2012, 491 (7423), 249-53.
[8]Shong, J.; Collins, C. H., Quorum sensing-modulated AND-gate promoters control gene expression in response to a combination of endogenous and exogenous signals. ACS Synth Biol 2014, 3 (4), 238-46.
Part2: Engineering pathway to detect Vibrio cholerae’s autoinducer molecule, CAI-1
Design Part
In efforts to develop an input sensor that is responsive to the specific Vibrio cholerae quorum sensing molecule (CAI-1), we engineered the metabolic pathways of E. coli to establish an optimized relationship between the sensing the cell density of V. cholerae and the promotor response used to regulate the E. coli cellular behaviour.
Previously, a team led by Poh CL., Holowko MB., and Wang H.1 identified three critical phosphorylation proteins involved in the CAI-1 signalling pathway. To further optimize the phosphorylation cascade, closer studies on the interaction between CAI-1 concentration and the promotor strength are used in activating the downstream pathway within an E. coli chassis. Through modelling and calculating the required expression levels of the three phosphorylation proteins, we generated an optimized sequence to facilitate the E. coli cellular response to CAI-1.
Figure 1| CAI-1 detecting sequence design
Build Part:
We cloned the optimized sequence containing the three phosphorylation proteins into the pBbE6k plasmid that contains a KanR antibiotic resistance gene as a selection marker. Following the original CqsS, LuxU, and LuxO protein sequences from Vibrio cholerae, we codon-optimized the sequences according to the codon utilization in the E. coli cell. We paired the different genes with constitutive promotors with differing expression strengths to minimize the metabolic burden imposed on the E. coli host. Additionally, we included a strep-tag sequence on either the N- or C- terminus of CqsS, LuxU, LuxO to track the protein expression levels in the E. coli cytoplasm. We specifically added the strep-tag to the N-terminal of CqsS as the C-terminal comprises the transmembrane domain. We were concerned that the inclusion of an N-terminus strep-tag might interfere with the protein folding and function.
As an output response to test the CAI-1 promotor strength, we cloned the mRFP downstream of the CqsS gene to monitor the cellular response to CAI-1.
Test Part：
To evaluate the E. coli response to the CAI-1 quorum sensing molecules, we monitored the RFP expression via the fluorescence output when tested under different V. cholerae cellular densities. We tested the different cellular densities to evaluate the CAI-1 production in V. cholerae under the different stages of cellular growth. In the article published by MB Holowko2, they discovered that an overnight culture of V. cholerae can have an accumulated concentration of CAI-1 as high as 1.25μM. Thus, through testing the supernatant using different dilution factors, we can establish a linear correlation between CAI-1 concentration and the gene expressions.
Learn Part：
In the initial design of our work, we used the same promoter sequence to regulate the expression of various proteins in the E. coli host. However, we observed gene truncation or gene loss in the extracted plasmids, indicating that similar sequences resulted in gene recombination.
Additionally, in adapting exogenous gene expression in our E. coli chassis, such as the CqsS system in Vibrio cholerae, all regulatory proteins need to be codon optimised to ensure a rapid response rate. The codon optimization is vital to prevent translational slowdown resulting from the rate-limiting presence of tRNA carrying the desired amino acids.
Reference
[1]Holowko, M. B.; Wang, H.; Jayaraman, P.; Poh, C. L., Biosensing Vibrio cholerae with Genetically Engineered Escherichia coli. ACS Synth Biol 2016, 5 (11), 1275-1283.
[2]Higgins, D. A.; Pomianek, M. E.; Kraml, C. M.; Taylor, R. K.; Semmelhack, M. F.; Bassler, B. L., The major Vibrio cholerae autoinducer and its role in virulence factor production. Nature 2007, 450 (7171), 883-6.
Part3: Engineering dehydration detection component and cellular-cellular signaling in Lactobacillus reuteri
Design part:
This experiment aims to create an upstream diarrhea prevention sensor and signal production tool, which is used to sense the environmental changes caused by diarrhea and convey the signal to the downstream E. coli engineering bacteria. A good portion of diarrhea results in an excessive increase in osmotic pressure (mainly due to chloride ion concentration) within the intestine, causing water sequestration, resulting in watery stools11. Therefore, in the presence of high salt concentrations, our L. reuteri produces the gram-negative bacterial quorum sensing molecule, acylated homoserine lactones (AHL) and subsequently activating autolysis. The autolysis releases the AHL into the intestinal tract that will activate our engineered probiotic E. coli in the large intestine to facilitate water retention.
Build Part:
We achieved this cellular behaviour using two independent cellular functions. The first function produces AHL, and the second function facilitates cellular autolysis in the presence of elevated salt concentrations.
The first route uses the luxI gene regulated by the ermB promoter in the lactic acid bacteria3to produce AHL used in E. coli quorum sensing2. The second route uses chlorine ion regulatory promoters rrnhB and gadR4, and downstream-regulated genes are the automatic gene acmA of Gram-positive bacteria5 from the previous iGEM design (Part: BBa_K3142013 from team iGEM19_SZPT-CHINA). We selected L. reuteri as our desired chassis due to its GRAS status and its ability to colonize the small intestine upstream of the large intestine, essential for delivering the AHL to the localized engineered probiotic E. coli. As the backbone of our genetic engineering component, we decided on the shuttle vector pTRKH3 that can replicate in both gram-negative and gram-positive bacteria.
Figure 1| The shuttle plasmid is pTRKH3-ermGFP from from Michela Lizier16
Test Part：
Due to limited wet laboratory time during the COVID-19 pandemic and various gene synthesis issues encountered in our work, we failed to complete the testing component on time. However, based on the experimental design, we intend to use different concentrations of NaCl solution and IPTG, then test its OD600 in 60 hours to compare its gene expression rate. By combining the AHL synthase and cellular autolysis function in the plasmid, we intend to determine the concentration of AHL released in the supernatant through the use of UHPLC by comparing it to commercially available standards.Learn Part：
The gene synthesis encountered some issues due to the gadR terminator that forms a stable hairpin during gene synthesis. Therefore, we replaced the gadR terminator with a double-series terminator, which has successfully prevented the formation of a stable hairpin structure.
Reference
[1]Berkes, J.; Viswanathan, V. K.; Savkovic, S. D.; Hecht, G., Intestinal epithelial responses to enteric pathogens: effects on the tight junction barrier, ion transport, and inflammation. Gut 2003, 52 (3), 439.
[2]Lizier, M.; Sarra, P. G.; Cauda, R.; Lucchini, F., Comparison of expression vectors in Lactobacillus reuteri strains. FEMS microbiology letters 2010, 308 (1), 8-15.
[3]Prindle, A.; Samayoa, P.; Razinkov, I.; Danino, T.; Tsimring, L. S.; Hasty, J., A sensing array of radically coupled genetic 'biopixels'. Nature 2011, 481 (7379), 39-44.
[4]Sanders, J. W.; Leenhouts, K.; Burghoorn, J.; Brands, J. R.; Venema, G.; Kok, J., A chloride-inducible acid resistance mechanism in Lactococcus lactis and its regulation. Molecular Microbiology 1998, 27 (2), 299-310.
[5]Steen, A.; Buist, G.; Horsburgh, G. J.; Venema, G.; Kuipers, O. P.; Foster, S. J.; Kok, J., AcmA of Lactococcus lactis is an N-acetylglucosaminidase with an optimal number of LysM domains for proper functioning. The FEBS Journal 2005, 272 (11), 2854-2868.
Part4: Programming E. coli antimicrobial response to combat Vibrio cholera
Design Part:
On top of preventing dehydration to treat diarrhea, a more direct approach to eliminating the diarrhea-causing pathogen can treat the symptoms. Thus, in this section, we aimed to target diarrhea-causing pathogen, V. cholerae, using antimicrobial proteins.
Human lactoferrin1 (flHLF) has antibacterial properties that are regarded as safe for host consumption. The antimicrobial properties are often attributed to the antagonism of flHLF to iron ions that are essential for bacterial growth1. We intend to use flHLF as our output to target the V. cholerae pathogenic infection, where the protein is expressed and secreted in the presence of CAI-1. We are also extending the study to use an improved variant of flHLF, rtHLF4, which is more effective and stable than conventional flHLF1.
Build Part:
We cloned the flHLF and rtHLF4 genes into a pET28 vector for expression in E. coli. The recombinantly expressed protein was purified via metal ion affinity chromatography and size exclusion chromatography before evaluating the anti-V. cholerae properties. In the antimicrobial screening, we tested the purified proteins against the attenuated Vibrio cholera strain, BNCC 2320302 ( BeNa Culture Collection).
Figure 1| Size exclusion chromatography of rtHLF4
Test Part：
In testing for the antimicrobial activity of the flHLF and rtHLF4 proteins, V. cholerae cultures were inoculated in fresh media containing an increasing concentration of the proteins. The changes in the growth trend observed over absorbance at 600nm help in determining the bactericidal or bacteriostatic properties of these proteins. Based on our assay, we discovered that flHLF and rtHLF4 are bacteriostatic and bactericidal, respectively (Fig. 2). Our results showed a prominent decrease in V. cholerae growth with the increasing concentrations of both flHLF and rtHLF4. flHLF showed a delay in growth, indicating the flHLF interferes with the pathogen’s growth cycle but is insufficient to repress the pathogenic proliferation over a duration of time.
Figure 2| Antibacterial test of the HLF
Learn Part：
The antimicrobial properties of flHLF require higher concentrations of the purified proteins to exert the antimicrobial activities and can only slow down the V. cholerae growth rate. Further, flHLF requires over 7 hours to have observable antimicrobial properties, rendering the protein unsuited for treating V. cholerae infections. Clinical data reported that severe cases of V. cholerae infections could be lethal within 24 hours of infection, where the flHLF’s antimicrobial activity is ill-suited to combat the pathogen (Fig. 2).
In order to resolve this issue, we substituted the lactoferrin protein with an alternative antibacterial peptide found from the iGEM database. A potential antibacterial candidate is the Eosinophil cationic protein that has antibacterial properties suited for use in the distal region of the human large intestine3-4. Further studies on the antimicrobial properties of the cationic protein will be tested in the nematode Caenorhabditis elegans to observe the antibacterial properties in a gut-like environment.
Reference
[1]Pan, Y.; Chua, N.; Lim, K.; Ho, C. L., Engineering of Human Lactoferrin for Improved Anticancer Activity. ACS Pharmacology & Translational Science 2021.
[2]Azman, A. S.; Rudolph, K. E.; Cummings, D. A.; Lessler, J., The incubation period of cholera: a systematic review. J Infect 2013, 66 (5), 432-8.
[3]Wagner, M.; Peterson, C. G. B.; Stolt, I.; Sangfelt, P.; Agnarsdottir, M.; Lampinen, M.; Carlson, M., Fecal eosinophil cationic protein as a marker of active disease and treatment outcome in collagenous colitis: A pilot study. Scandinavian Journal of Gastroenterology 2011, 46 (7-8), 849-854.
[4]Rivera, F. P.; Medina, A. M.; Bezada, S.; Valencia, R.; Bernal, M.; Meza, R.; Maves, R. C.; Ochoa, T. J., Bovine lactoferrin decreases cholera-toxin-induced intestinal fluid accumulation in mice by ganglioside interaction. PLoS One 2013, 8 (4), e59253.
Part5：Gene circuit design to regulate antimicrobial properties of E. coli to target V. cholerae infection
Design Part:
We regulate the expression of the antimicrobial protein using an AND gate to have a precise release of the antimicrobial agents using the two inputs are the transcriptional repressor protein PhlF—its promoter (Pphlf)1, and transcriptional activator protein EsaR—Esa box2. The EsaR-Esa box is regulated using the AHL produced by our engineered L. reuteri upon sensing changes in the salt concentrations in the small intestine, whereas the PhlF-pPhlf is regulated by sensing the V. cholerae quorum sensing molecule CAI-1. The programmed AND gate is used to produce the antimicrobial agents to kill V. cholerae in the presence of dehydrating conditions and CAI-1, ensuring that the produced antimicrobial agent does not cause unnecessary perturbation to the host microbiome.
EsaR protein is constitutively expressed, functioning as a transcription repressor to the Esa box to inhibit downstream expression2. The EsaR protein interacts with the AHL quorum-sensing molecules, causing it to be released from the Esa box and restoring the expression of genes regulating colanic acid biosynthesis and the conditional expression of the antimicrobial agent.
Many studies have demonstrated the use of the PhlF inverter system in E. coli1. PhlF functions as a transcription repressor, inhibiting the downstream gene expression regulated by the pPhlF promoter. The repressed levels of PhlF release the imposed inhibition on pPhlF, leading to the expression of the protein sequences regulated by pPhlF.
In total, the EsaR system responds to the AHL concentration, and PhlF system respond to CAI-1 concentrations. Thus, combining the two systems using an AND gate would promote the downstream expression only in the presence of CAI-1 and AHL.
Build Part：
The PhlF expression and the pPhlF + Esa box-regulated GFP expression were synthesized on opposite strands of the double-stranded DNA (fig. 1). This design was intended to prevent transcriptional spillover resulting from two adjacent genes being close to each other. Thus, arranging the genes on the opposing strands can increase the output stability of the experimental system.
Figure 1| PhlF + GFP + PphlF sequence design
We similarly applied the same approach to studying the EsaR-Esa box (fig. 2).
Figure 2| EsaR controlling part
We integrated the EsaR and PhlF regulatory systems on a single plasmid to ensure no loss of gene function due to the plasmid loss (fig. 3). We included RFP and GFP as our output for both the gene circuit response, where individual regulatory proteins were tagged with His-tag and Strep-tag to monitor its expression rate via western blot (fig. 1 & fig. 3). We selected different terminators and RBSs to prevent gene recombination (fig. 3).
Figure 3| Total plasmid design
In the specific selection of transformation, electrical stimulation rather than thermal transformation was used in the experiment to E.coil Nissle 1917. According to existing experimental data, thermal transformation can lead to irreversible denature of E.coil Nissle 1917 protein and death of strain, so the required transformation cannot be completed3 .To simplify the experiment's procedure, E.coli Acella was used to test some of the parts by the transformation method of heat shock.
Test Part：
The plasmid was initially transformed into E.coli Acella (fig. 4) and selected for Kanamycin resistance. We found that the design plasmid can be transformed into the E. coli host while persevering the antibiotic resistance, as shown in figure 4. The negative control bacteria without plasmid transformation on the left and the right are heat-shock bacteria.
Figure 4| Transformation of the plasmid to E.coil Acella
Western blot was used to detect the normal expression of PhlF (fig. 5). Using anti-His tag antibodies conjugated with Horseradish peroxidase, we found that the induced expression of PhlF accumulates at the inclusion body, whereas uninduced cells showed no expression of PhlF. This result suggests a tightly regulated expression of PhlF, facilitating the following step of the study in identifying the strength of PhlF in inhibiting pPhlF. A purified His tagged-GFP was used as a positive control in the study.
Figure 5| Western blot of PhlF (E is experiment; C is control; S is supernatant; P is precipitate; PC is positive control)
We used fluorescence tests to verify the entire pathway. AHL and IPTG with different concentrations were added to the E. coli chassis harbouring the lac+ESAR+phlf+RFPv1.1 plasmid shown in fig. 3. The RFP fluorescence should be inversely proportional to the concentration of AHL, whereas GFP fluorescence should be inversely proportional to the AHL and IPTG concentration. However, our results so far do not support our hypothesis (fig. 6).
Figure 7| Fluorescence test of GFP with different concentrations of IPTG and same concentration of AHL. We have conducted this experiment at a series of AHL concentrations: 0nM, 10nM, 100nM, 1000nM. But the result is similar and do not show obvious inhibitory effects, thus we does not show all of them here.
Learn Part：
Due to the time limitation and travel restrictions, we could not complete the validation of our designed circuit. Despite repeating the experiment numerous times using different inducer concentrations and reaction conditions, the change in fluorescence level over time does not match the profile we generated in silico. We hypothesize that this might be due to various factors, including imposed metabolic stress from the protein expression in the host chassis. Thus, we intend to regulate the expression of these key proteins using RBS with differing strengths in order to limit the stress imposed on the host cell.
Reference
[1]Moon, T. S.; Lou, C.; Tamsir, A.; Stanton, B. C.; Voigt, C. A., Genetic programs constructed from layered logic gates in single cells. Nature 2012, 491 (7423), 249-53.
[2]Shong, J.; Collins, C. H., Quorum sensing-modulated AND-gate promoters control gene expression in response to a combination of endogenous and exogenous signals. ACS Synth Biol 2014, 3 (4), 238-46.
[3]Hancock, V.; Vejborg, R. M.; Klemm, P., Functional genomics of probiotic Escherichia coli Nissle 1917 and 83972, and UPEC strain CFT073: comparison of transcriptomes, growth and biofilm formation. Mol Genet Genomics 2010, 284 (6), 437-54.