Global Regulation Module

To prevent the potential risk of overproliferation when engineered bacteria are applied in vivo, we designed a LuxR/TetR dual-factor cascade system. This system utilizes a quorum-sensing mechanism to monitor population density and incorporates a negative feedback loop to precisely regulate downstream gene expression, thereby achieving global control.

As shown in Figure 1.A, the amCyan gene was placed downstream of the P_tet. After the construct was successfully introduced into E. coli BL21(DE3), colony PCR was performed, and as shown in Figure 1.B, the bands matched the expected length. Sequencing analysis further confirmed that the recombinant plasmid was consistent with the design as shown in Figure 1.C.

To visually assess the expression level of the downstream gene, the recombinant plasmid was expressed in E. coli BL21(DE3). Once visible fluorescence was observed under ambient light, 10 mL of the bacterial culture was preserved in 50% glycerol. Subsequently, 100 μL aliquots were inoculated hourly into 10 mL of M9 medium supplemented with 50 μg/mL kanamycin. After 48 hours of incubation, samples from the experimental and empty-plasmid control groups were subjected to visual observation and imaging.

As shown in Figure 2, the AmCyan protein expression level in the engineered bacteria demonstrated a gradual increasing trend over time.

To further quantify the regulatory effect of the quorum-sensing circuit, the recombinant plasmid described above was expressed in E. coli BL21(DE3). When the culture reached an OD₆₀₀ of 0.4-0.6, a glycerol stock was prepared, and 1 μL of the glycerol stock was inoculated into 200 μL of LB medium for relative fluorescence intensity measurement. A preliminary experiment was designed to determine the total detection duration and time intervals. Based on the results, measurements were taken every 20 minutes starting from inoculation, for a total duration of 5.5 hours. Data analysis and ODE fitting were subsequently performed.

As shown in Figure 3, the growth kinetics of the engineered bacteria, measured by OD₆₀₀, exhibited a progressive increase that followed a logistic model (R² = 0.9943) over the course of the experiment.

As shown in Figure 4, the dynamic expression profile of the global regulation circuit was quantitatively monitored by measuring relative fluorescence units (RFU) over a 500-minutes cultivation period. The kinetic trajectory was fitted using a two-parameter model, which accurately captured the characteristic progression of circuit activation. The fitted curve demonstrates the density-dependent activation dynamics of the quorum-sensing regulatory system, showing an initial rapid increase to a peak RFU value, followed by a decline and eventual stabilization at a steady state.

Validation for BioPROTAC

To validate the functional capacity of bioPROTAC, the following experimental investigations were conducted:

1. Yeast two-hybrid (Y2H) and bacterial protein co-purification assays were employed to confirm that the fusion construct does not compromise VHH212 nanobody engagement with HIF-1α;
2. In vitro ubiquitination assays were performed to demonstrate that the VHL moiety within the fusion protein successfully mediates polyubiquitination of HIF-1α.

​​In the yeast two-hybrid system​​, all groups grew on SC-2 medium, confirming successful transformation. The experimental group co-transformed with pGADT7-vhh-vhl and pGBKT7-hif-1α plasmids, along with the positive control, exhibited robust growth on SC-4 medium, whereas the negative control showed slow or no growth[2]. These results verify that the interaction between the VHH nanobody and Hif-1α remains intact despite the fusion protein construction (Figure 5.A).

​​Similarly, in the bacterial protein co-purification assay​​, E. coli BL21(DE3) was co-transformed with pET-28b(+)-hif-1α-His and pUC57-vhh-vhl-Myc. After sonication, Hif-1α-His was purified using Ni-NTA affinity chromatography. If VHH-VHL interacts with HIF-1α, both proteins should be detectable by SDS-PAGE. Indeed, compared with the control, whole-cell lysates showed bands corresponding to Hif-1α-His (94.0 kDa) and VHH-VHL-Myc (34.6 kDa). Furthermore, the co-purification of VHH-VHL-Myc with Hif-1α-His was confirmed in the elution fractions, while neither band was detected in wash fractions containing non-specific proteins. This further demonstrates that the VHL ubiquitination domain does not interfere with the VHH module's ability to recognize and bind its target protein (Figure 5.B).

​​For the in vitro ubiquitination assay​​, purified bioPROTAC protein (VHH-VHL) was incubated with truncated HIF-1α substrate (aa 1-365), E1 activating enzyme UBA1, E2 conjugating enzyme UBC8, ubiquitin-His, and ATP. Reactions were terminated at different time points (2 h, 4 h, 6 h) and analyzed by Western blot using an anti-ubiquitin antibody, with two replicates per group. It should be noted that, due to experimental equipment limitations, two nitrocellulose membranes were used for the assay.

Compared with the negative control lacking UBA1 and UBC8, the experimental group showed time-dependent enhancement of HIF-1α ubiquitination. In contrast to the blank control without VHH-VHL, the group supplemented with VHH-VHL exhibited significantly stronger ubiquitination signals(Figure 6.).

These results demonstrate that the bioPROTAC can efficiently mediate ubiquitination of HIF-1α in vitro, and strongly suggest its potential to promote ubiquitination of non-hydroxylated HIF-1α under hypoxic conditions in vivo.[3]

Through computational rational design and complementary in vivo/in vitro validation, the engineered bioPROTAC molecule VHH-VHL has demonstrated capacity to both retain the high-affinity binding of its VHH212 module to the target protein HIF-1α and functionally recruit the ubiquitination machinery to facilitate targeted polyubiquitination. These findings indicate that the designed bioPROTAC effectively binds to and mediates ubiquitin-dependent degradation of HIF-1α under hypoxic conditions in vivo, thereby attenuating hypoxia-driven angiogenesis at hemorrhoidal sites and therapeutically inhibiting disease progression.

Validation for Anti-VEGF Nanobody

To achieve the expression of P17-Anti-VEGF nanobody, we fused a 6×His-tag to the C-terminal of our PelB-P17-Anti-VEGF nanobody gene and cloned it into the pET-28b(+) plasmid for purification, and controlled the expression by the lac operon (Figure 7). Then the plasmid was transformed into E. coli SHuffle for expression. Transformants were selected on LB agar plates containing kanamycin, as only bacteria harboring the recombinant plasmid possess the kanamycin resistance gene. Positive colonies were picked for culture expansion.

Figure 7. Plasmid map used to express P17-Anti-VEGF nanobody.

The transformed bacteria were cultured at 37 °C until the OD₆₀₀ reached 0.5, at which point protein expression was induced with 0.5 mM IPTG for 16 hours. The bacterial cells were harvested by centrifugation, and the periplasmic proteins were extracted using the osmotic shock method. Successful expression of the P17-Anti-VEGF nanobody fusion protein was confirmed by SDS-PAGE and Western blot analysis of the obtained protein samples.(Figure 8.)

To further validate the anti-VEGF activity, cell proliferation was assessed in VEGFR2-positive HUVECs and VEGFR2-negative HEK293 cells using the MTT assay at 24 hours post-treatment. The results showed (Figure 9) that treatment with the nanobody for 24 hours significantly inhibited the proliferation of HUVECs in a dose-dependent manner, with inhibition rates ranging from 40% to 50% (P < 0.05) compared to the control. Conversely, no significant inhibitory effect was observed in HEK293 cells at any concentration tested, with inhibition rates below 5% (P > 0.05).

Figure 9. The antiproliferative activity of nanobody on HUVEC cells.

Anti-inflammatory Module

To alleviate hemorrhoid symptoms, treat local inflammation, and reduce swelling, pain, and effusion of hemorrhoidal tissue, we designed the anti-inflammatory module.

At the inflammation site of hemorrhoids, the concentration of reactive oxygen species (ROS) is very high, and we selected it as the signal trigger of the anti-inflammatory module. We chose SoxR/P_SoxS oxidative stress response promoter as the regulatory element. SoxR is a transcription factor containing [2Fe-2S] cluster, while P_SoxS is a promoter sequence. In a high ROS environment, SoxR binds between the -10 and -35 regions, and its oxidized form distorts the promoter conformation, promotes the binding of RNA polymerase (RNAP) and other transcription factor, and activates transcription.

To verify the function of the SoxR/P_SoxS oxidative stress response promoter, we cloned amCyan gene as the reporter gene into the pUC57 vector and constructed the corresponding reporter plasmid (Figure 10.).

We transformed the recombinant plasmid into E. coli BL21 (DE3) for subsequent functional verification. Before oxidant induction, we used 12% SDS-PAGE electrophoresis to verify that SoxR could be normally expressed in E. coli BL21 (DE3) (Figure 11).

Figure 11. 12% SDS-PAGE results for verification of SoxR expression.

control:Total protein sample extracted from E. coli BL21 (DE3) carrying the empty plasmid. Lane 1: Total protein samples extracted from expanded cultures of one different single colony that were successfully transformed. A comparison of the two indicates that SoxR is successfully expressed.

After verifying the expression of SoxR, to verify the functions of the SoxR and P_SoxS, we set up a series of tert-butyl hydrogen peroxide (tBHP) concentration gradients for induction (0 μM, 20 μM, 40 μM, 60 μM, 80μM, 100 μM, 120 μM, 140μM, 160 μM, 180μM, 200μM). We added 200 μL of induction medium inoculated with engineered bacteria into a 96-well plate and used a microplate reader to detect the emission light intensity at 486 nm. Three parallel replicates were set for each induction concentration (Figure 12).

We found that the activity of the P_SoxS is extremely dependent on the ROS concentration , and the background expression of P_SoxS is low in the absence of ROS.

To identify a suitable anti-inflammatory effector, we reviewed extensive literature and data, and ultimately selected Di-melittin, designed by NKU-China, 2024. To facilitate expression and purification, we added an N-terminal pelB signal peptide and a C-terminal 6×His-tag.

Melittin itself is known for its strong anti-inflammatory activity through multiple mechanisms , making it highly attractive for our project. However, the clinical application of melittin is severely limited by its strong cytotoxicity. The Di-melittin variant designed by iGEM24_NKU-China significantly reduces this cytotoxicity by connecting two melittin monomers with a linker to form a unique hairpin structure, thereby making its application feasible.

After contacting NKU-China, we learned that their experiments involved eukaryotic expression of Di-melittin in yeast cells for subsequent validation. For biosafety considerations, however, we selected Escherichia coli Nissle 1917 (EcN) as our chassis strain. After careful discussion, we concluded that since Di-melittin consists of only 75 amino acids and has a relatively simple structure, it can be feasibly expressed in E. coli. Moreover, the modified Di-melittin shows greatly reduced cytotoxicity to host bacteria, making prokaryotic expression in E. coli achievable. Since NKU-China,2024 has already conducted comprehensive functional characterization and thoroughly validated the anti-inflammatory properties of Di-melittin, building upon their work, our work primarily focused on verifying its feasibility for prokaryotic expression. Therefore, we performed Di-melittin expression verification in E. coliBL21(DE3).

To this end, we engineered Di-melittin with an N-terminal PelB signal peptide for secretion into the periplasmic space and a C-terminal 6×His-tag for downstream purification. The 6×His-tag was connected to Di-melittin through a flexible linker (GGSSSGG) to minimize structural interference (Figure 13).

To achieve prokaryotic expression of Di-melittin, we constructed a recombinant plasmid based on pET-28b(+) through homologous recombination and transformed it into E. coli BL21(DE3) (Figure 14.A).The transformed colonies were cultured on LB plates containing kanamycin at 37 °C for 16 h. Colony PCR (Figure 14.B) and sequencing (Figure 14.C) confirmed the correct insertion of the target gene.

Positive clones were induced at OD₆₀₀≈0.6 with 1 mM IPTG and cultured at 16 °C for 18 h. Cells were harvested, lysed, and total protein was purified with Ni-NTA affinity chromatography. SDS-PAGE analysis showed a clear band of approximately 12 kDa in whole-cell lysates, while no band was detected from periplasmic extractions (Figure 15). Based on design, the expected size of Di-melittin after PelB cleavage was ~9.6 kDa. However, the observed 12 kDa band corresponded to the uncleaved PelB-Di-melittin, consistent with its theoretical molecular weight. Sequencing analysis verified the integrity of the PelB sequence, suggesting that the observed molecular weight discrepancy did not result from errors in the tag sequence.

Figure 15. The shown gel was obtained from electrophoresis on a 15% SDS-PAGE.

Lane 1-2: periplasmic extract after purification, showing no visible target band. Lane 3: purified product from whole-cell lysates, where a distinct band of 12 kDa was observed. Lane 4: total protein extract, also displaying a clear 12 kDa band. The observed size corresponds to uncleaved PelB-Di-melittin (12.1 kDa), rather than the expected 9.6 kDa after signal peptide cleavage. The 10 kDa marker band, highlighted in red, appeared faint during gel imaging and was specially indicated for clarity.

Owing to time constraints, we were not able to perform further experiments to reach a definitive conclusion. We propose several possible explanations:

1. Interference from protein folding. The target protein may adopt secondary structures or fold during intracellular synthesis or membrane translocation, thereby hindering the access of signal peptidase to the cleavage site.
2. Degradation of periplasmic peptides. Anti-inflammatory peptides are relatively small and may be unstable during periplasmic extraction or subsequent handling. Endogenous periplasmic proteases could degrade the peptide, and although PMSF was applied as an inhibitor, the concentration may have been insufficient to prevent degradation.
3. Inclusion body formation. The protein may aggregate in the cytoplasm as inclusion bodies, producing strong bands in whole-cell lysates but yielding little or no soluble protein in the periplasmic fraction.

While the functional activity of Di-melittin had been comprehensively verified by iGEM24_NKU-China, our results further demonstrated that the peptide can be efficiently expressed in a prokaryotic system, paving the way for scalable production and in vivo applications.

This experiment engineered an E. coli Nissle 1917 strain with the SoxR/P_SoxS oxidative stress-responsive promoter to express the anti-inflammatory peptide Di-melittin specifically in high-ROS hemorrhoidal environments, thereby alleviating swelling, pain, and exudation.

OMV Delivery Module

To achieve precise targeted delivery of Escherichia coli outer membrane vesicles (OMVs) to the hemorrhoid site, we selected the OmpA-CAR fusion protein as the targeting element. The α-helix of the OmpA signal peptide efficiently inserts into membrane structures, while CAR specifically recognizes angiogenic endothelial cells abundant at the hemorrhoid site. Thus, the OmpA-CAR fusion protein effectively mediates targeted OMV transport to the hemorrhoid site.

We constructed expression vector containing ompA-CAR, ompA-amCyan, and srp-TagRFP on pET-28b(+), respectively. We co-transformed two plasmid pairs, ompA-amCyan with srp-TagRFP and ompA-CAR with srp-TagRFP, into E. coli Nissle 1917 (ΔnlpI), a mutant known for its increased OMV yield. PCR appraisal of the transformed colonies showed that both groups exhibited double positive bands, indicating that both sets of plasmids had been successfully double transformation into the EcN strain (Figure 16,17).

Subsequently, we employed the differential centrifugation method (DC) to isolate the OMV precipitate to verify the extraction efficiency, we further utilized the OMV marker protein OmpA (approximately 38.2 kDa) as the core validation indicator and performed a 12% SDS-PAGE experiment.[1] The electrophoresis results showed a clear characteristic OmpA protein band in the extract, directly confirming the success of the OMV extraction procedure (Figure 18).

We designed and constructed the OmpA-CAR fusion peptide, whose N-terminus retains the positively charged-hydrophobic amphiphilic region characteristic of the OmpA signal peptide, enabling stable insertion and anchoring into the OMV lipid bilayer membrane; the C-terminus is linked to the CAR sequence with specific recognition functionality, exposing it on the outer surface of the vesicles. This structure simultaneously achieves dual functions of membrane anchoring and targeted recognition, guiding drug-loaded OMV to precisely accumulate at the hemorrhoid lesion site, thereby laying the molecular and carrier foundation for the subsequent directional delivery of the antiangiogenic module.

Limited by the current experimental progress, the loading efficiency and targeting effect of OMV have not yet been clarified. In subsequent studies, we will first measure the Zeta potential of the extract to clarify the stability of its dispersion system and preliminarily confirm the structural characteristics of OMV; then, through fluorescence confocal microscopy experiment, observe the co-localization of AmCyan and TagRFP to verify whether the loading of Engineered outer membrane vesicle is successful; finally, conduct cell scratch assay to test the promoting effect of CAR on cell migration, thereby verifying the targeting efficacy of engineered OMV.

Suicide and Medicine-Food Collaboration Module

To ensure the safety of engineered bacteria, we designed a suicide module that kills bacterial strains both in the external environment and in the digestive tract once the food-derived small molecules are no longer supplied, thereby avoiding potential leakage of engineered bacteria.

To validate the activity of the thiM riboswitch and its response range to TPP, we placed the AmCyan protein downstream of the riboswitch and introduced the construct into E. coli BL21(DE3). As shown in Figure 19.A, the amCyan reporter was placed downstream of the riboswitch. As shown in Figure 19.B, after the construct was successfully introduced into E. coli BL21(DE3), colony PCR was performed and the bands matched the expected size. As shown in Figure 19.C, sequencing analysis further confirmed that the recombinant plasmid was consistent with the design.

When cultures reached OD₆₀₀ = 0.4-0.6, the bacteria were transferred to M9 minimal medium. We set up a preliminary experiment (TPP concentrations of 0, 10 μM, 100 μM, 1 mM, and 5 mM) to determine the approximate range in which TPP exerts its inhibitory effect (Figure 20.A). As shown in Figure 20.A, functional testing demonstrated that the TPP riboswitch gradually repressed downstream gene expression as TPP concentration increased, with obvious inhibition observed at concentrations above 1 mM. These results confirmed the successful construction of the expression vector and defined the approximate inhibitory range of the riboswitch ligand TPP.

The samples were divided into groups and supplemented with 0-30.0 μL of TPP stock solution in 2.0 μL increments, corresponding to final concentrations of 0-15 mM in 1 mM increments. Fluorescence intensity (Ex: 453 nm, Em: 486 nm) and OD₆₀₀ were measured every 30 minutes for 5 hours at 37℃ with shaking at 200 rpm. Relative fluorescence intensity was calculated as fluorescence divided by OD₆₀₀.

As shown in Figure 20.B, the thiM riboswitch gradually repressed downstream gene expression as the TPP concentration increased, with the inhibitory effect reaching a plateau above 13 mM. Data fitting using the four-parameter logistic (4PL) model yielded an IC₅₀ = 6.81 ± 1.77 mM. This experiment validated that the TPP riboswitch can nearly completely suppress gene expression under high TPP concentrations.

In our experiment, we constructed a plasmid vector with CcdB toxin protein connected after the lactose promoter and introduced it into E. coli BL21(DE3). IPTG was used to induce expression and OD₆₀₀ was detected continuously for 4 hours. Samples were taken every 30 minutes, diluted and plated to determine viable cell counts.

As illustrated in Figure 21.A, the ccdB gene was inserted downstream of the T7 promoter. Subsequent transformation of this construct into E. coli BL21(DE3) enabled verification by colony PCR, which yielded bands corresponding to the anticipated molecular size (Figure 21.B). Additionally, sequencing analysis further confirmed that the recombinant plasmid conformed precisely to the designed specifications (Figure 21.C).

As illustrated in Figure 22.A, OD₆₀₀ analysis revealed that the CcdB protein induced a potent bactericidal effect, leading to rapid bacterial cell death within one hour post-induction. Figure 22.B further demonstrated a drastic decline in viable cell counts and a significant reduction in colony diameters over time. Consistent with these observations, Figure 23.A-C confirmed a sharp decrease in both viable cell numbers and colony sizes with prolonged incubation. These findings collectively establish that CcdB exhibits a pronounced lethal activity against bacterial cells, in contrast to weakly toxic proteins that solely inhibit growth.

To validate the functionality of the thiM riboswitch and the toxin protein CcdB, a genetic circuit was constructed utilizing the pUC57 plasmid as the backbone. The ccdB gene was inserted downstream of the thiM riboswitch, with the entire circuit regulated by the P_J23100 promoter, and subsequently transformed into E. coli BL21(DE3). As illustrated in Figure 24.A, the design positioned ccdB under the control of the riboswitch. Figure 24.B demonstrates that colony PCR performed after successful transformation yielded bands corresponding to the expected size. Additionally, sequencing analysis, as depicted in Figure 24.C, confirmed that the recombinant plasmid conformed to the intended design.

Cultures were grown to an OD₆₀₀ of 0.4-0.6, transferred to M9 medium, and induced with 7 mM TPP (corresponding to the half-maximal inhibitory concentration, IC₅₀). Toxicity was assessed in 96-well plates by monitoring OD₆₀₀ using a Synergy™ H1 microplate reader, with comparisons made between experimental and control groups. As shown in Figure 25, at the IC₅₀ concentration of the riboswitch ligand TPP, the suicide system elicited a pronounced response, effectively suppressing bacterial growth and confirming the potent bactericidal activity of the CcdB protein under riboswitch regulation.

Figure 25. OD₆₀₀ measurement under 7 mM TPP induction.