{% extends "layout.html" %} {% block title %}Results{% endblock %} {% block lead %}You can describe the results of your project and your future plans here.{% endblock %} {% block page_content %}
To facilitate the secretion of functional peptides from Escherichia coli, we engineered a chimeric construct combining peptide sequences with Lpp'OmpA. Lpp'OmpA functions as a transport protein. This protein spans the outer membrane approximately five times. Peptides linked via an enterokinase cleavage site are appended to the C-terminus of Lpp'OmpA. The arrangement enables peptide transport across the bacterial outer membrane and exposure within the intestinal milieu. Subsequently, intestinal enterokinase cleaves at the designated site, allowing functional peptides to be absorbed by epithelial cells. To evaluate the system's efficiency, we constructed six variants of the Lpp'OmpA chimeric protein: Lpp'OmpA-lptE, LptE (no signal negative control lacking Lpp'OmpA), Lpp'OmpA-QEP-GST (QEP-GST), Lpp'OmpA-AQ-GST (AQ-GST), Lpp'OmpA-AQ-6xhis (AQ-6xhis) and Lpp'OmpA-QEP-6xhis (QEP-6xhis). We will present our findings in four dislevertinct sections.
First, we performed transformation experiments to establish our secretion system. We selected BL21(DE3) as the expression host, utilizing the T7-lac promoter in our plasmids. Following successful transformation, bacterial colonies were subjected to PCR, gel electrophoresis(Fig. 1), and verified by Sanger sequencing(Fig. 2).
To confirm the expression of Lpp'OmpA-peptides in bacteria from the transformed plasmids, we performed Western blot on bacterial lysates. We observed that induction temperature affects the stability of Lpp'OmpA-peptide expression under the T7-lac promoter. At temperatures of 37°C, 30°C, 25°C, and 16°C, the negative groups (0 mM IPTG) exhibited significant promoter leakage. Notably, lower temperatures resulted in more stable expression of the Lpp'OmpA-peptide system. At 37°C, which is close to human intestinal temperature (36.5-37.7°C), all six chimeras were successfully expressed (Fig. 3).
● Discussion:
After analysis, we hypothesized that lactose and its analogues in LB culture medium might induce expression even in the absence of IPTG. To address this problem, we replaced LB medium with a basic medium (without yeast extract) during the induction phases. However, comparing the induction results from basic medium with those from LB medium indicated that the yeast extract in LB medium may not be the primary cause of promoter leakage (Fig. 4).
The surface exposure of all constructs was verified by FACS analysis. Bacteria were incubated first with corresponding Mouse anti-tag antibodies and subsequently incubated with the FITC-conjugated Goat anti-Mouse IgG (H+L). No detergent was used in the staining process. Initially, we induced for 1h at 16°C, which showed display success in Lpp'OmpA_lptE, but not in the QEP and AQ sets. (Fig. 5)
We suspected that the induction time was not enough for the bacteria to complete the surface display. Therefore, we extended the induction time to overnight. This time we observed out that not only QEP and AQ sets had positive results, but also the lptE.
Lpp'OmpA_lptE incubated with mouse anti-6xhis tag and mouse anti-GST tag were used to confirm the effect of endogenous binding.
In contrast with Lpp'OmpA_lptE incubated with 6xhis tag, AQ_6xhis and QEP_6xhis showed higher fluorescence intensity. QEP_GST and AQ_GST showed similar results. Moreover, Lpp'OmpA_lptE incubated with the FLAG tag showed higher fluorescence intensity than the sample incubated with the 6xhis tag and GST tag. (Fig. 7)
To observe the surface display more clearly, we performed fluorescence microscopy on bacteria expressing the Lpp'OmpA chimeras. Bacteria were incubated first with corresponding Mouse anti-tag antibodies and subsequently incubated with the ABflo® 594-conjugated Goat anti-Mouse IgG (H+L) without detergent. The ABflo® 594 can be visualized in red, the DNA in blue (DAPI), and the membranes in green (Oregon green).
Similar results were observed:not only QEP and AQ sets had positive results, but also the lptE.
To obtain a better resolution ratio, we performed cryo-electron microscopy. Bacteria were incubated first with corresponding Mouse anti-tag antibodies and subsequently incubated with the 12nm Colloidal Gold-AffiniPure™ Goat Anti-Mouse lgG (H+L). Similar results were obtained. Both lptE and chimeras showed positive results.
From our results, we could confirm that Lpp'OmpA_lptE had been successfully displayed on the surface because when we induced for 1h at 37°C, FACS analysis showed a clear contrast between Lpp'OmpA_lptE and lptE. However, the surface display of QEP and AQ sets might not be verified, since lptE showed positive results in the same experiment. We hypothesized that during the long-term induction or pretreatment process, the polarity of the outer membrane wasn't maintained, which caused the LptE expressed in the negative control to move from the inner leaflet to the outer leaflet. Another hypothesis was that during the induction, the permeability of bacteria was affected. The antibody moved into the cells and interacted with the non-displaying proteins. More discussions can be seen on the engineering page.
To validate the QEP sets and AQ sets, enzymatic cleavage experiment was performed.
To confirm that the enterokinase can target the DDDDK site and successfully cleave off the peptide into the environment, a series of enterokinase digestion experiments and western bolt experiments were conducted. We first designed experiments to investigate the minimum digestion concentration of enterokinase, digestion temperature, pH, and duration. Then, we expressed AQ-GST, QEP-GST, and lptE-GST under the optimal environmental factors that we explored before.
There are six possible circumstances we found in the following experiments: (1) bacteria membranes are intact and the proteins are presented on the surface, (2) bacteria membranes are not intact (bacteria are broken during induction or digestion), (3) bacteria membranes are intact and the proteins are not present on the surface, and the mixture of (1) and (2), (2) and (3), or (1) and (3) (Fig. 10). The bacteria culture was centrifugated after induction and digestion. The supernatant and precipitation are incubated with two different antibodies. To get a clearer picture, the hypothesized western blot results of different circumstances are demonstrated below (Fig. 11).
Phase 1: the exploration of digestion factors of enterokinase
To investigate the environmental factors affecting digestion—specifically the minimum enterokinase concentration, digestion temperature, time, and pH—we conducted a series of experiments. In each experiment, subsequent tests were adjusted based on the optimal conditions identified in the previous stage. By progressively aligning the experimental conditions with those of the intestinal environment, successful in vitro digestion may suggest similar effectiveness in vivo.
First, we tested the enterokinase concentration from 0 to 2U/50 μL in bacterial medium and digested the mixture at 25°C for 12 hours in pH 8.0, 25 mM Tris-HCl buffer (Fig. 12).
● Analysis
In the exploration of the minimum concentration of enterokinase, we first induced bacteria under 16℃ overnight with 1mM IPTG as our previous results indicated that the chimeras expressed and presented in a more stable way. To get a more specific picture of the digestion results, we incubated precipitation and supernatant with two different antibodies and repeated the experiment twice.
In the precipitation-FLAG blots, no signal was observed at 15KDa in both results, corresponding to Lpp'OmpA-FLAG. The degradation protein in precipitation-GST and supernatant-FLAG of two results indicate bacterial lysis. Although the results of the supernatant-GST show the successful digestion of Lpp'OmpA-FLAG-QEP-GST, the digestion occurring in the lysed bacteria did not confirm that the process was targeted to the bacterial surface. Although the two result have different signals, they both shows protein degradation. The reason for the differences may be because of the different condition of bacteria storage before induction (in 4℃ and -30℃) or the operation differences.
Although the bacterial lysis happened, there is still positive signal of QEP-GST which may means the 2U, 1.5U and 1U all can digest the protein. From the signal intensity at the three enzyme concentrations, we regard they all have similar and ability to digest when bacteria lysed. However, we cannot draw the inference about which concentration is optimal in the condition that the bacteria are intact.
Next, we tested the effect of digestion temperature (25°C and 37°C) and digestion time (3 and 5 hours) using 2U of enterokinase in the same buffer (Fig. 13).
● Analysis
In our exploration of optimal digestion temperature and time, we induced bacterial expression at 16°C overnight and at 37°C for 1 hour using 1mM IPTG, as bacterial lysis was observed in previous experiments. We hypothesized that the lysis was caused either by prolonged induction at low temperatures or by extended digestion in the nutrient-deprived buffer. In the precipitation-FLAG sample, no signal was observed at 15 kDa in plus enzyme group, corresponding to Lpp'OmpA-FLAG. Additionally, the absence of degraded proteins in the precipitation-GST and supernatant-FLAG samples suggests no bacterial lysis occurred. In the supernatant-GST sample, a comparison of plus and no enzyme group at 37°C for 1 hour showed successful digestion of Lpp’OmpA-FLAG-QEP-GST. However, the presence of degraded protein in the supernatant-GST sample from the 16°C, 12-hour induction with negative enzyme concentration supports the hypothesis that bacterial lysis was caused by prolonged low-temperature induction.
From the results, we can tell that the group with 37℃-1h induction and 25℃-5h and 37℃-3h digestion can not only keep the bacteria intact but also suceessfully digest chimeras as the AQ-GST signals are positive in 2 groups.
Following that, we varied the pH of the 25 mM Tris-HCl buffer to 6.5, 7.5, and 8.0 (the pH range of the intestine) and digested the mixture at 37°C for 3 hours with 2U of enterokinase (Fig. 14).
● Analysis
In our exploration of optimal digestion pH, we induced bacterial expression at 16°C overnight and at 37°C for 3 hours using 1mM IPTG, as the 37℃ 1h induction signal is quite weak. We hypothesized that the weak signal is due to the incomplete presentation of chimeras on the outer membrane.
In the precipitation-FLAG sample, no signal was observed at 15 kDa in plus enzyme group, corresponding to Lpp'OmpA-FLAG. Additionally, the absence of degraded proteins in the precipitation-FLAG and supernatant-GST samples suggests bacterial lysis occurred. Although the results of the supernatant-GST in positive enzyme group in both induction group showed the successful digestion of Lpp'OmpA-FLAG-QEP-GST, the digestion occurring in the lysed bacteria in both groups did not confirm that the process was targeted to the bacterial surface. However, the presence of weak signals in the supernatant-GST sample from the 37°C, 3-hour induction does not support the hypothesis that the weak signal is due to incomplete presentation because of short induction time.
From the results of group with 37℃-3h induction and pH 7.5 digestion and group with 16℃-12h induction and pH 6.5 digestion shows that positive pH digestion range can cover the pH range of intestine environment.
Phase 2: expression of AQ-GST, QEP-GST, and lptE-GST under optimal digestion conditions
For lptE-GST and QEP-GST, the degradation protein signal in supernatant-GST and in supernatant-FLAG indicates bacterial lysis. In supernatant-GST, without enterokinase, the Lpp'OmpA-FLAG-lptE-GST signal indicates the incomplete digestion of lptE-GST. While QEP-GST only shows degradation protein signal and Lpp'OmpA-FLAG-QEP-GST signals, it may indicate the inclusion body formation.
For AQ-GST, no degradation protein signal indicates no bacterial lysis. In supernatant-GST, without enterokinase, the Lpp'OmpA-FLAG-AQ-GST signal also indicates the incomplete digestion.
● Discussion:
We also conducted an experiment to investigate whether there will be some bacteria resuspended in the supernatant, which may cause the same results as bacterial lysis. In this experiment, there is no difference in filter used and no filter used sample preparation. It may mean that there is no bacteria in the supernatant without the filter(Fig. 16).
● Future:
All Lpp'OmpA-FLAG signal is missing in the experiments. The reason may be that the FLAG tag is in an inherently undetectable structural domain after digestion or the antibody affinity is not good enough to generate detectable signal. To further investigate the protein we can conduct TEM experiment to detect Lpp'OmpA-FLAG which is anchored on outer membrane after digestion.
To evaluate the functions of the target proteins, we separately transformed the pET-28a-[HSP60-Flag], pET-28a-[LAP-His], pET-28a-[Neae-Flag] into strain E. coli BL21 (DE3). We conducted colony PCR to verify the transformation(Fig. 17). Moreover, The Sanger sequencing conducted by Tsingke ensureed the specific sequences (Fig. 18).
To ensure the correct expressions of target proteins, we utilized protein purification resin(Fig. 19) and Western Blot(WB)(Fig. 20) to detect the cellular products.
It's demonstrated that our engineered bacteria underwent cellular metabolism properly and synthesize desired proteins.
The expression levels of HSP60, LAP, and Neae were better at higher temperatures, which may correspond with the optimal growth temperature of E. coli and indicate that our proteins can function better at a temperature closer to the human body.
However, the expression level of HSP60 has been consistently concerning, and its content in the lysate supernatant is relatively low. The same situation happened to protein LAP as well. Both of the circumstances pose challenges for subsequent experiments like co-immunoprecipitation (Co-IP).
To determine whether the surface display protein, Neae, anchors on the outer membrane or not, we conducted a range of experiments.
Firstly, we employed indirect immunofluorescence(IF) to confirm the appearance of Neae. The result demonstrated that Neae presented a stronger fluorescence at 30℃(Fig. 21).
Furthermore, we also conducted WB to ensure the expression level of Neae on the outer membrane(Fig. 22). The result indicates that Neae is successfully present on the outer membrane as well and yields more at a higher temperature, which corresponds to the IF result.
Apart from Neae, we also examined the expression level of LAP in cell supernatant and outer membrane(Fig. 23). Both of the component contained LAP though the concentration of the protein remained low.
As a cell membrane protein that enables Listeria to adhere to the intestine epithelium, the LAP we chose from non-pathogenic Listeria (L. innocua) also presents itself on the cell membrane of our engineered E. coli. L. innocua is able to take advantage of LAP and binds with HSP60 in the small intestine, occupying the binding site of pathogenic Listeria and avoiding intestinal infection induced by pathogenic Listeria[1].
We constructed pET-28a-[Neae-HSP60-FLAG] and pET-28a-[Neae-LAP] after the confirmation of the target proteins(Fig. 24). Moreover, we detected the expression levels of the fusion proteins.
For pET-28a-[Neae-HSP60-FLAG], we employed PCR to introduce restriction enzyme cleavage sites for EcoRI and HindIII. After performing DNA purification, restriction enzyme digestion, and T4 ligation, we successfully constructed the target plasmid within E. coli DH5α and BL21(DE3)(Fig. 25)(Fig. 26).
The expression level of Neae-HSP60 was detected by WB (Fig. 27). The expression characteristics of protein Neae-HSP60 were quite similar to protein Neae and HSP60. All of them exhibited a better yield at a higher temperature.
For pET-28a-[Neae-LAP], we constructed plasmids using similar methods, except using EcoRI and XhoI as restriction enzymes. We successfully transferred the plasmid into E. coli BL21(DE3) and received expected sequencing result(Fig. 28).
The lack of soluble proteins in the supernatant and non-specific binding of the tag has always been a persistent challenge for us. Under this circumstance, we tried to replace the promoter and protein tag for WB and add a solubilizing protein tag for the original part (See more in
Engineering page).
Due to time constraints, not all genetic circuits were completed. We successfully built up SUMO-HSP60 circuits in E. coli BL21 and other three fruitful circuits are presented below (Fig29).
We introduced the soluble protein tag SUMO into the HSP60 construct to increase the amount of soluble protein in the supernatant, and we replaced the 6xHis tag with GST to improve antibody binding specificity. Furthermore, we tried to introduce SUMO-HSP60 and LAP fragment into plasmid backbone pBAD to change the promoter, and to replace LAP-6x His with LAP-GST at the same time.
For SUMO-HSP60, we added the soluble protein tag SUMO into the circuit of HSP60 to enhance soluble protein expression in the supernatant. This is also a successfully constructed genetic circuit;
For LAP-GST, we replace the 6xHis with GST to achieve better antibody binding specificity and protein solubility.
Gene segments SUMO, HSP60 and LAP are cloned and inserted into the pBAD backbone, forming pBAD-SUMO-HSP60-FLAG and pBAD-LAP.
For the genetic circuit SUMO-HSP60, DNA segments SUMO and HSP-pET-backbone were amplified successfully (Fig. 30). Then, we ligated the two DNA segments and constructed plasmid SUMO-HSP60 in E. coli BL21(DE3) with single colonies containing the target band (Fig 31).
Apart from the SUMO-HSP60 genetic circuit, we also amplified segments GST, LAP, HSP60, and pBAD backbone for other genetic circuits. However, they haven't been ligated to form complete plasmids yet due to time constraints(Fig. 32).
We first verified the quorum sensing system. For more details on the plasmid, visit the Engineering page PLux-deGFP and PLas-deGFP were transformed into E. coli BL21 separately. Sequencing verifies transformation success.
To confirm that pLux is induced by the combination of LuxR and 3OC6HSL, and pLas is induced by the combination of LasR and 3OC12HSL, we conducted induction experiments to measure fluorescence intensity at different AHL concentrations. The results show a clear dose-dependent pattern, with fluorescence intensity increasing as the concentration of AHL rises, indicating a well-functioning system (Fig. 34).
Initially, we performed transformation experiments to establish the CcdB-CcdA system. We selected E. coli Tuner(DE3)pLysS as the expression host, utilizing the T7-lac promoter in our plasmids. Following successful transformation, bacterial colonies were subjected to PCR, gel electrophoresis, and verified by Sanger sequencing. The sequencing results confirmed the correct sequence (Fig. 35).
After the transformation, we conducted 2 sets of experiments: CcdB toxicity verification and CcdA antioxidation verification. In the former set, different concentrations of IPTG were added to the transformant, which would induce the expression of CcdB and kill bacteria. In the latter set, the same concentration of IPTG and different concentrations of L-arabinose were added to the transformant. L-arabinose would trigger the expression of CcdA and detoxify CcdB. We measured OD600 changes over time using the plate reader as an indication of the viability of bacteria. (Fig. 36)
Compared with the no-induction group, the cells in the L-arabinose gradient showed relatively high density in the LB medium, which indicated that the L-arabinose triggered the expression of CcdA and rescued the cells. In the IPTG gradient, Growth was significantly inhibited from the concentration of 104 μM to the 100μM, which fully induced CcdB expression. However, when the concentration reached 10μM and below, growth was back to normal. Overall, the toxicity and detoxification of the CcdB-CcdA set were verified.
Drolia, R., Amalaradjou, M. A. R., Ryan, V., Tenguria, S., Liu, D., Bai, X., Bhunia, A. K. (2020). Receptor-targeted engineered probiotics mitigate lethal Listeria infection. Nature Communications, 11(1), 6344. doi:10.1038/s41467-020-20200-5