The bacterial strains used in this study were Escherichia coli OmniMAX™ (Invitrogen) for cloning and Bacillus subtilis HT800F16 as an expression host. Competent cell preparation for both E. coli and B. subtilis followed established protocols17, 18. Cells were cultured in LB medium supplemented with antibiotics (100 µg/mL ampicillin for E. coli or 10 µg/mL chloramphenicol for B. subtilis) at 37°C with shaking. All plasmids and strains are listed in Table 1.

Table 1

Plasmids / Strains	Description	Reference
pHT2328	Template plasmid used to obtain the hlaH35LH48L gene	Our lab’s stock
pHT1796	The integration vector contains Pgrac100-MCS-yhcR118 and amyE locus, used as the backbone	Our lab’s stock
pHT2315	The integration vector contains Pgrac100-hlaH35LH48L-yhcR118 and amyE locus	This study
BsHT1796	B. subtilis strain with the integration of Pgrac100-MCS-yhcR118 into the chromosome at the amyE locus, used as the control	Our lab’s stock
BsHT2315	B. subtilis strain integration of Pgrac100- hlaH35LH48L-yhcR118 into the chromosome at the amyE locus	This study

Table 2

Primers	Oligonucleotide sequences (5’ – 3’)	Purpose	Amplicon length (bp)
ON2336 ON2335	ACTGTCGCTTCCAAGACGTCGTTTGTCATTTCTTCTTTC AAAGGAGGAAGGATCCATGAAAAC	Amplify hlaH35LH48L	993
ON2332 ON2335	GATCTTCTTTAATTGGGTCTTCCGTCCCCGGATCATCTC AAAGGAGGAAGGATCCATGAAAAC	Colony PCR for E. coli	1151
ON469 ON2137	GGCGTTCTGTTTCTGCTTCG CTGTTTGTGATGGTTATCATGCAGGATTG	Confirm integration at 5’amyE site	1097
ON925 ON2378	GAATTAGCTTGGTACCAAAGGAGGTAAGGATCACTAG CTCAACTGTCGCTTCCAAGACG	Confirm the presence of yhcR-hlaH35LH48L after integration	1122
ON2331 ON470	GACGGAAGACCCAATTAAAGAAGATCCAAGGCCAGGTGAAG AACCCGCTCCGATTAAAGCTAC	Confirm integration at 3’amyE site	1477

To facilitate the anchoring of heterologous protein on the B. subtilis cell wall, the vector pHT1796, containing the yhcR sequence, was used as a template in the study. The target gene hla_H35LH48L was initially amplified from the synthesized plasmid pHT2328 using the primers ON2336 and ON2335, as detailed in Table 2. After digestion with the two restriction enzymes, BamHI and AatII, both the target gene and the template pHT1796 were ligated using T4 DNA ligase to construct the plasmid pHT2315. The ligated plasmid pHT2315 was transformed into E. coli OmniMAX™, and cells containing the plasmid were screened on LB agar plates supplemented with ampicillin. The constructed plasmid was subsequently confirmed via colony PCR and sequencing.

Following cloning, the fusion gene yhcR-hla_H35LH48L from pHT2315 was naturally transformed into the chromosome of B. subtilis HT800F. The resulting strain, in which the fragment was integrated through double crossover at the amyE locus, was designated as BsHT2315. Bacterial cells were screened on LB agar plates containing chloramphenicol, and colonies were further validated by PCR.

For each colony, three pairs of primers were used to ensure correct integration. Details of all primers, including their nucleotide sequences and binding sites, are provided in Table 2. All bacterial strains, including the newly constructed BsHT2315, were stored at −80°C for further analysis.

The BsHT2315 bacterial cells encoding recombinant Hla_H35LH48L were pre-cultured in 4 mL of LB broth supplemented with chloramphenicol until the OD₆₀₀ reached approximately 2–3. The suspension cells were then sub-cultured into 10 mL of LB broth containing chloramphenicol and shaken at 37°C and 220 rpm until the OD₆₀₀ reached 0.8. Next, the suspension was sub-cultured into 10 mL of LB medium and incubated at 37°C with shaking at 220 rpm until an OD₆₀₀ of 0.8 was achieved. Subsequently, the culture was induced with IPTG at concentrations of 0 mM, 0.1 mM, and 1 mM and incubated for 20 hours at 30°C.

For Western blot preparation, cell pellets were collected by centrifugation at 10,000 rpm for 2 minutes. For whole-bacterial cell enzyme-linked immunosorbent assay (bactELISA), the culture was collected with 30% glycerol added as a preservative. All samples were stored at -80°C.

Pellets containing cell wall–bound proteins were incubated at 37°C for 30 minutes in 100 μL of lysis buffer (0.25 M sucrose, 25 mM Tris-HCl, pH 7.2) supplemented with 3 μL of 50 mg/mL lysozyme. The samples were then combined with 5× loading buffer and heated at 95°C for 5 minutes. After separation by SDS-PAGE, proteins were transferred to a nitrocellulose membrane.

For Western blot analysis, reactive bands were detected using an anti-Hla primary antibody (raised in mice) at a dilution of 1:20,000, followed by incubation with a rabbit anti-mouse IgG–HRP conjugate (Sigma) at a dilution of 1:40,000. Signal detection was performed using Pierce™ ECL Western Blotting Substrate (Thermo Fisher Scientific). The BsHT1796 strain and the attenuated toxin Hla_H35LH48L were utilized as controls in the assay.

Whole-bacterial cell ELISA (bactELISA) was employed to verify the anchoring of Hla_H35LH48L on the B. subtilis cell wall. This technique was referenced in the study by Cumming et al.19. Briefly, BsHT1796 and BsHT2315 vegetative cells, serving as analytes, were resuspended in 200 μL of coating buffer (100 mM NaHCO₃, pH 9.6) and directly coated onto a microplate (Thermo Scientific™ Nunc™ MicroWell™ 96-Well Microplates) overnight at 4°C. The samples were washed twice with 100 μL of 0.1% PBS-Tween and then incubated in blocking buffer, consisting of PBS-Tween supplemented with 5% (w/v) skim milk, at room temperature for 1 hour. After washing twice, 50 μL of anti-Hla primary antibody (diluted 1:10,000) was applied to each well and incubated for 2 hours at room temperature. Next, 50 μL of rabbit anti-mouse IgG–HRP conjugate (Sigma), diluted 1:40,000, was added. The plates were incubated with the conjugate for 2 hours at room temperature, washed, and then incubated with 50 μL of TMB Liquid Substrate for ELISA (Sigma) for 20 minutes. Finally, 50 μL of 1N HCl was added to stop the reaction. Absorbance was measured at OD₄₅₀ using a BMG Labtech CLARIOstar plate reader. Each sample was replicated three times, and statistical analysis was performed using a one-way ANOVA test in GraphPad Prism software (USA).

Each experimental group consisted of five 6-week-old female Swiss mice. Oral immunizations were administered on days 0, 14, and 28 using 250 µL of B. subtilis vegetative cells (strains BsHT1796 and BsHT2315) at an optical density at 600 nm (OD₆₀₀) of 60, diluted in 1X PBS. The control groups comprised mice receiving either BsHT1796 or PBS via oral gavage. Blood and stool samples were collected on days 21, 35, and 42, while small intestine samples were collected on day 42. For analysis, 0.6 g of feces or small intestine tissue was treated with 500 µL of 1X PBS containing 0.2 mg/mL PMSF, homogenized, and centrifuged to obtain the supernatant.

A 96-well plate (Thermo Scientific™ Nunc™ MicroWell™) was coated with 50 μL of Hla_H35LH48L protein (produced by our lab) at a concentration of 5 μg/mL. Subsequently, 50 μL of either serum, fecal, or intestinal extract, serving as the primary antibody, was added at dilutions of 1:250 for serum and 1:50 for fecal/intestinal extract, followed by overnight incubation at 4°C. The wells were then incubated with either anti-mouse IgG (whole molecule) peroxidase-conjugated antibody produced in rabbits (Sigma, A9044) at a dilution of 1:40,000 or anti-mouse IgA (α-chain specific) peroxidase-conjugated antibody produced in goats (Sigma, A4789) at a dilution of 1:10,000 for 2 hours. All samples were analyzed in triplicate, and absorbance at 450 nm (OD₄₅₀) was measured using a CLARIOstar plate reader. Statistical analysis was performed using one-way analysis of variance (ANOVA) followed by Tukey’s post-hoc test in GraphPad Prism software (USA).

Figure 1

The target gene hla_H35LH48L, with an expected size of 993 bp, was amplified from pHT2328 via PCR and subsequently fused with the C-terminal yhcR-encoding sequence of pHT1796, resulting in the construction of pHT2315. The structure of vector pHT2315 is illustrated in Figure 1A. Following transformation, E. coli colonies harboring the recombinant plasmids were selected on LB agar plates containing ampicillin, and colony PCR was performed. The colony PCR results revealed a band at 1151 bp (Figure 1B), confirming the presence of the fusion gene hla_H35LH48L-yhcR in pHT2315. The sequence of hla_H35LH48L-yhcR was further verified by DNA sequencing (data not shown).

The constructed vector pHT2315 was successfully delivered into Bacillus subtilis HT800F through the process of natural transformation. During the transformation, DNA uptake occurred in some bacterial cells, leading to the integration of the heterologous fusion gene hla_H35LH48L-yhcR into the B. subtilis chromosome at the homologous amyE locus. Bacterial cells that did not incorporate the target fusion gene were eliminated through a preliminary screening on chloramphenicol LB agar plates.

PCR analysis was performed on the newly generated strain to confirm the integration of the hla_H35LH48L-yhcR fusion gene into the B. subtilis chromosome via a double-crossover recombination event. This analysis employed three specific primer pairs with distinct functions: one pair verified the presence of the gene of interest, while the other two targeted the integration sites at the 3′ and 5′ ends of amyE, respectively (Figure 1C). Gel electrophoresis results showed visible bands of varying sizes for all colonies with the three primer pairs (Figure 1D), matching the predicted sizes as outlined in Table 2, thereby confirming the successful creation of the integrated strain. The newly engineered strain was designated BsHT2315.

Figure 2

In this experiment, the vegetative cells of BsHT1796 and BsHT2315 strains were lysed after resuspension in a lysis buffer containing lysozyme. Proteins released from both strains, along with the control (purified Hla produced by our lab), were loaded onto an SDS-PAGE gel, separated, and transferred to a nitrocellulose membrane. Western blot analysis was then conducted using a primary antibody raised in mice against Hla and a secondary rabbit anti-mouse IgG conjugated with HRP. Finally, the interaction between the ECL substrate and the HRP enzyme produced a chemiluminescent signal, visualizing the proteins on the membrane.

As shown in Figure 2A, compared to the BsHT1796 strain, which lacks the hla_H35LH48L protein-encoding gene, BsHT2315 under inducible conditions with 0.1 and 1 mM IPTG displayed visible bands of relatively equal intensity. Meanwhile, the BsHT2315 sample without IPTG induction also exhibited a band, but of lighter intensity. The difference in band thickness corresponding to the inducer concentration indicated that expression of the target fusion protein was regulated by the correct Pgrac212 promoter and lacI operon system in the presence of the inducer. Additionally, the larger molecular size of the fusion protein produced by the strain BsHT2315, compared to the purified Hla_H35LH48L, is likely due to the contribution of the yhcR C-terminal region, which increased the molecular weight.

To confirm that Hla_H35LH48L was anchored on the B. subtilis cell wall after being linked to YhcR, a bactELISA assay was performed. In this assay, only samples induced with 0.1 mM IPTG were evaluated. As antigens, B. subtilis vegetative cells were directly coated onto a microplate. Cells with target proteins anchored on the cell surface would subsequently bind to anti-Hla and secondary antibodies. The coated cells remained stable during a procedure that allowed them to retain their shape, clearly demonstrating the display of the fusion protein on the cell wall. Figure 2B demonstrated a significant two-fold increase in the signal intensity for Hla_H35LH48L on the cell wall of BsHT2315 compared to the control (0.7260 ± 0.0074 vs. 0.3447 ± 0.0097, p = 0.0009). This indicated successful surface display of the target protein on the cell wall of B. subtilis.

Figure 3

We confirmed the successful surface display of alpha toxin on B. subtilis cells. To assess the immunogenicity of cell-surface displayed Hla_H35LH48L and its oral delivery efficacy, Swiss mice were orally immunized with three doses of B. subtilis vegetative cells on days 0, 14, and 28. IgG and IgA antibody levels were evaluated in serum, fecal, and small intestinal samples.

ELISA results in Figure 3B revealed a significant increase in anti-Hla_H35LH48L intestinal IgA levels in mice immunized with BsHT2315 cells on day 42 (OD450 = 0.135 ± 0.004, p = 0.0005), which was two-fold higher than those immunized with BsHT1796 (OD450 = 0.066 ± 0.008). Similarly, a significant difference (p = 0.0012) was observed between BsHT2315 and PBS groups (OD450 = 0.076 ± 0.017). As expected, no significant differences were observed between the two control groups (BsHT1796 and PBS), as BsHT1796 lacks the target protein sequence. There was no significant increase in fecal IgA levels in the BsHT2315 group compared to the control (data not shown). In addition, serum IgG levels of BsHT2315 exhibited a slight increase on days 21 (0.194 ± 0.030, p = 0.0239), 35 (0.164 ± 0.019, p = 0.8436), and 42 (0.203 ± 0.045, p = 0.0151) compared to the control group BsHT1796. However, no significant differences were observed between dosage regimens throughout the immunization process (Figure 3A).

These results suggest that B. subtilis cells displaying Hla_H35LH48L on their surface via YhcR were successfully delivered to the mouse gut and elicited a local immune response at mucosal sites following oral administration.