The bacterial strains, plasmids, and primers used in this study are described in Supplementary Table S1. Bacteria were grown aerobically at 37°C on LB agar, LB broth (prepared according to Sambrook and Russell, 2001 [27]), MacConkey (Kasvi, Brazil), and Salmonella-Shigella (SS) media (Oxoid, United Kingdom). Ampicillin (100 μg/mL; Sigma, Spain), Kanamycin (50 μg/mL; Sigma, Spain), chloramphenicol (25 μg/mL; USB, United Kingdom), or mannose (0.5%, Sigma, Spain) were used, as required. The bacterial strains used in this study were stored at −80°C in 20% glycerol in an LB medium.

Bacteria were seeded on LB agar and incubated at 37°C for 16–18 h. Colonies grown on LB agar were inoculated into LB broth and cultured for 16–18 h at 37°C under agitation (150 rpm). The next day, the culture was diluted 1:100 in fresh LB broth and grown to the exponential phase (∼10⁸ CFU/mL) under the above conditions. The culture was centrifuged at 4,000 × g for 5 min, and the pellet was suspended in PBS. Subsequently, the bacteria were diluted to the appropriate concentration for in vitro and in vivo experiments. In all experiments, mannose was added to the culture medium (LB agar or LB broth) for the growth of the ΔihfABpmi mutant.

The new triple mutant ΔihfABpmi was constructed using the λ Red recombination system followed by transduction with phage p22. We initially constructed the Δpmi:Kan mutant by recombination-mediated allelic exchange of the pmi gene with the kanamycin cassette. The deletion of the pmi gene was verified by PCR using the detection primers described in Supplementary Table S1. Subsequently, Δpmi:Kan was used as the donor strain for transduction with the p22 phage, and the ΔihfAB mutant was used as the recipient strain, resulting in the ΔihfABpmi mutant.

The double mutant ΔtolRA was constructed using the λ Red system as previously described [28]. We deleted the tolR and tolA genes from the ATCC 14028 chromosome in a single step, by red-mediated recombination and allelic exchange of the tolR and tolA genes with the kanamycin resistance gene. The gene deletion was confirmed by PCR using the tolRADT-F and tolRADT-R primers described in Supplementary Table S1.

Cell lines B16F10 (mouse melanoma) and 5,637 (human bladder cancer) were used in this study. Cells were grown at 37°C in humidified air with 5% CO_2. Dulbecco’s Modified Eagle medium (Thermo Fisher Scientific, Carlsbad, CA, United States) (DMEM) was used to grow B16F10 cells and RPMI 1640 (Thermo Fisher Scientific) medium was used to grow 5,637 cells. Both media were supplemented with 10% fetal bovine serum, penicillin (100 U/mL) and streptomycin (100 μg/mL) for cell growth.

The in vitro growth of the mutants was assessed in LB broth as previously described [29]. Briefly, a 16–18 h culture was diluted (1:1,000) in LB broth and incubated at 37°C and 150 rpm. Growth was monitored for 12 h employing hourly optical density readings and CFU determination by plating on LB-agar. Three independent experiments were performed.

Virulence attenuation of engineered mutants of S. enterica Typhimurium was evaluated in vivo in the Galleria mellonella infection model by inoculating 10 μL of a 1 × 10⁶ CFU/mL bacterial (mutant or wild-type strain) suspension in the last proleg of larvae between 200 and 250 mg. Ten larvae were used per group. Larvae survival was observed for 96 h after inoculation. The experiment was performed three times.

The integrity of LPS, which is associated with bacterial immunogenicity, was also analyzed by profiling LPS on a polyacrylamide gel. LPS from the mutants and wild-type strain was extracted as previously described [30]. Briefly, colonies grown on LB-agar plates were homogenized in deionized water until an OD₆₀₀ between 0.4 and 0.5 was reached. The bacterial suspension was centrifuged at 10,000 rpm for 2 min, and the supernatant was discarded. The pellet was homogenized in a Laemmli buffer (Tris-HCl, pH 6.8 0.625 M, 2% SDS, 25% glycerol, and 0.01% bromophenol blue) incubated at 100°C for 10 min. Then, 10 µL of Proteinase K solution (2.5 mg/mL) was added, and the suspension was homogenized and incubated at 60°C for 1 h. The LPS samples were analyzed by SDS-PAGE. Polyacrylamide electrophoresis was performed in a 5% stacking gel and in a 12% separating gel, with exposure to a constant voltage of 100 V for 2 h. The separated LPS was stained with silver stain [31], by initially incubating the gel in a fixative solution (40% ethanol and 5% acetic acid) for 16 h, followed by incubation in an oxidizing solution for 10 min. After three washes in deionized water, the gel was stained for 10 min with a silver solution (0.02M NaOH, 1.5% ammonium hydroxide, and 0.7% silver nitrate), followed by three wash-steps in deionized water. Bands were developed in 0.02% formaldehyde and 0.005% acetic acid solution. After the appearance of the bands, the gel was washed in deionized water and photographed. Pseudomonas aeruginosa (kindly donated by Dr. Regina Lúcia Baldini—Department of Biochemistry, Instituto de Química, Universidade de São Paulo) containing the wzz gene deletion was used as a control for the absence of LPS. The wzz gene is essential for LPS biosynthesis. In Pseudomonas aeruginosa, strains with a deletion of this gene cannot produce LPS.

The invasiveness and survival capabilities of the mutants were measured using gentamicin resistance, as previously described, with some modifications [22, 32]. Briefly, 1×10⁵ melanoma or bladder cancer cells per well were seeded in 24-well plates with antibiotic-free medium (RPMI for bladder cancer cells or DMEM for melanoma cells) supplemented with 10% fetal bovine serum and incubated at 37°C, 5% CO₂ for 20 h; two 24-well plates were prepared, one for the invasion assay and one for the survival assay. The next day, 2 × 10⁶ CFU of bacteria were added over the cells to reach a multiplicity of infection (MOI) of 10:1 and incubated at 37°C, 5% CO₂ for 1 h. After three washes with PBS, a medium (RPMI for bladder cancer cells or DMEM for melanoma cells) containing 100 μg/mL gentamicin was added to each well and incubated for 1 h to kill extracellular bacteria. Then, each well was rewashed with PBS. For the invasion assay, tumor cells from one of the plates were immediately lysed with 0.5% Triton X-100 solution in PBS, followed by CFU determination by plating on LB-agar medium. For the survival assay, the other plate was incubated with medium (RPMI or DMEM) containing 20 μg/mL gentamicin and incubated for 4 h. Then, the tumor cells were washed with PBS and lysed for CFU determination. We performed three independent experiments, with three replicates from each strain.

The MTT assay evaluated the cytotoxic effect of the mutants as previously described [33]. A total of 1×10⁴ cells were seeded in 96-well plates with antibiotic-free medium (RPMI for bladder cancer cells or DMEM for melanoma cells) supplemented with 10% fetal bovine serum. Cells were incubated for 20 h at 37°C and 5% CO₂. Then, the cells were washed with PBS and treated with 2 × 10⁶ CFU of bacteria to reach an MOI of 100:1. The plate was incubated at 37°C under 5% CO₂ for 1 h. Cells were rewashed with PBS and incubated with medium (RPMI or DMEM) containing 100 μg/mL gentamicin for 2 h at 37°C under 5% CO₂. Then, the cells were rewashed and incubated with a medium (RPMI or DMEM) containing 20 μg/mL of gentamicin for 24 h under appropriate culture conditions. After the incubation, the medium was removed, the cells were rewashed and the cells were re-incubated for 4 h with MTT (5 mg/mL) diluted in RPMI or DMEM medium. MTT diluted in the medium was discarded, and crystals were dissolved with dimethyl sulfoxide (DMSO-Sigma, Saint Louis, MO, United States). Absorbance was read at 570 nm. The viability of cells treated with the mutants was compared to that of untreated (control) cells. We performed three independent experiments, with three replications of each strain.

6–8-week-old C57BL/6JUnib females were used in this study. All animals were obtained from the Multidisciplinary Center for Biological Research (CEMIB—UNICAMP). The animal care committee of Universidade Estadual de Campinas approved all experiments with mice under protocol numbers 5769-1/2021 and 5895-1/2021. The animals were kept under specific pathogen-free conditions. The mice were acclimatized for 2 weeks before the start of the experiments. Mice were 8 weeks old at the time of the first inoculations. The mice were shaved on the right flank 2 days before starting the experiments to facilitate subcutaneous inoculations.

The animals were randomly divided into four groups of five and injected subcutaneously with 60 µL in the right flank. Group I: mice received 10⁵ CFU of ΔihfABpmi weekly for 2 weeks, two inoculations total. Group II: the animals received 10⁶ CFU of ΔtolRA weekly for 2 weeks. Group III: the animals received 10⁷ CFU of ΔtolRA weekly for 2 weeks. Group IV: mice received phosphate-buffered saline (PBS) weekly for 2 weeks. During the experiment, the animals were weighed thrice a week, and the appearance of any signs of disease (ocular discharge, piloerection, lethargy) was analyzed. They were euthanized 2 weeks after the last inoculation. In addition, parts of the liver, spleen, lung, and kidneys were collected and immediately fixed in 4% paraformaldehyde solution for organ damage analysis. Tissues were stained in H&E and analyzed under a Leica DM5500 B optical microscope (Wetzlar, Hesse, Germany). Two independent experiments were performed.

To quantify bacterial distribution and persistence at the end of the experiment, part of the liver, blood, and spleen were aseptically collected and homogenized in PBS with a tissue homogenizer (Omni Mixer Homogenizer, Vernon Hills, IL, United Stated), as previously described [29]. The homogenized tissues were plated on LB-agar, MacConkey, and SS with appropriate antibiotics to determine CFU. Two independent experiments were performed using 5 mice per group in each experiment.

The antitumor efficacy of the mutants was evaluated in a murine melanoma model. B16F10 cells were diluted in antibiotic-free and fetal bovine serum-free DMEM medium. B16F10 cells (3×10⁶) were inoculated subcutaneously into the dorsal flank region of C57BL/6JUnib mice. When the tumor reached 100 mm³ (10–12 days after tumor cell inoculation), the tumor-bearing animals were randomly assigned to groups (PBS, ΔihfABpmi, and ΔtolRA), seven mice per group. Mice were inoculated intratumorally with 60 µL of mutant strains suspensions (10⁵ CFU of ΔihfABpmi or 10⁶ CFU of ΔtolRA) or PBS once a week for 2 weeks. Tumor size was measured every 2–3 days, and tumor volume was calculated as previously described [19]. Two independent experiments were performed using 7 mice per group in each experiment.

Tumor-bearing mice that reached 2000 mm³ or showed signs of pain were euthanized to avoid suffering. The body weight of the mice was also measured every 2–3 days, and the mice were observed for any signs of disease throughout the treatment period. At the end of the experiments, the mice were euthanized by intraperitoneal injection of 5 mg/kg of xylazine and 40 mg/kg of ketamine, followed by cervical dislocation. Liver, spleen, kidney, and lung samples were collected, fixed in 4% formaldehyde, and stained with H&E for histopathological examination. The experiment was repeated twice.

To quantify bacterial distribution and persistence at the end of the experiment, tissues were weighed and homogenized with a tissue homogenizer (Omni Mixer Homogenizer, Vernon Hills, IL, United Stated), as previously described [29]. The mixed tissue suspension was plated in LB-agar, MacConkey, and SS with appropriate antibiotics to determine CFU. This experiment was performed twice.

The potential therapeutic antitumor of mutants was also evaluated intraperitoneally. Mice bearing B16F10 tumors (100 mm³) were treated intraperitoneally with 10⁵ CFU of ΔihfABpmi, 10⁶ CFU of ΔtolRA or PBS once a week for 2 weeks. Tumor size was measured every 2–3 days. The mice were euthanized 19 days after starting the treatments. One independent experiment was performed using 5 mice per group.

To understand the immune response involved in the antitumor effect mediated by the ΔihfABpmi mutant, mice bearing tumors of approximately 100 mm³ were inoculated intratumorally with 10⁵ CFU of ΔihfABpmi or PBS. Four days later, the mice were euthanized. The tumor mass was divided into two parts, one for macrophage analysis by flow cytometry and another for analysis of the expression of genes involved in the antitumor response.

A suspension of isolated cells was prepared from the tumor tissue. Tumor tissue was cut into small fragments and digested in 200 U/mL collagenase IV buffer at 37°C for 1 h, then passed through a 70 μm cell filter. The samples were incubated with fluorochrome-labeled antibodies [CD80-Pe, CD206-APC, CD11b-Percp, F4/80-Fitc (Elabscience. Houston, United States)] for 20 min at 4°C and 50,000 events were analyzed using the cytometer of NovoCyte flow using the following panel macrophage (F4/80+ CD11b+), M1 macrophage (F4/80+ CD80⁺) and M2 macrophage (F4/80+ CD260+). The data were analyzed with the software NovoExpress 1.5.0 software.

mRNA expression was analyzed by qRT-PCR. Total RNA extraction from tumor tissue was performed using a Direct-zol RNA MiniPrep Plus kit (Zymo Research, Irvine, CA, United States) and Trizol reagent (Invitrogen, United States) according to the manufacturer’s instructions. RNA integrity was confirmed by agarose gel followed by ethidium bromide staining. RNA purity and concentration were verified using NanoDrop 2000c (Thermo Scientific, United States). The RNA was treated with DNase I Amplification grade (Sigma-Aldrich. Louis, MO, United States) to eliminate genomic DNA contamination. As a contamination control, a PCR was performed on all samples to verify the absence of genomic DNA. Reverse transcription was performed to synthesize cDNA using the High-Capacity cDNA Reverse Transcription kit (Applied Biosystems. United States). Then, the real-time PCR reaction was performed using the 2x qPCRBioSyGreen Mix Separate-Rox kit (PCRBIOSYSTEMS. Wayne, Pennsylvania, United States), according to the manufacturer’s instructions, in a MicroAmp Optical 96-Well Reaction Plate microplate (Thermo Scientific, United States), using the StepOnePlus Real-Time PCR System (Applied Biosystems, United States). The dissociation Curve was made to verify the specificity and quality of the primers used. The mRNA expression was normalized to the expression of β-actin and GAPDH, which were used as endogenous controls. Relative mRNA expression levels were calculated using the 2^−ΔΔCT method. The sequences of the primers used are shown in Supplementary Table S2.

Statistical analysis was performed using GraphPad Prism version 8 (GraphPad, San Diego, CA, United States). One-way ANOVA followed by Dunnett’s test was used in experiments with three or more experimental groups. Student’s t-test was used to analyze data from experiments from two experimental groups. Data were represented as the mean ± Standard Deviation of the mean (SEM), and p < 0.05 was considered statistically significant.

First, we verified the potential effects of gene deletions on bacterial growth by comparing the mutants (ΔihfABpmi and ΔtolRA) with the parental strain 14028 WT. For in vitro growth in LB broth (37°C, 150 rpm), we observed similar OD₆₀₀ values between the mutants and the parental strain (Figure 1A). However, analysis of colony forming units (CFU) number counts showed that the growth rate of ΔihfABpmi was lower than that of 14028 WT (Figure 1B), and ΔihfABpmi had a lower number of CFUs in the stationary phase of growth as compared to 14028 WT. ΔtolRA showed a growth pattern similar to 14028 WT (Figure 1B).

Next, we investigated whether the deleted genes affected the virulence of 14028 WT by inoculating G. mellonella larvae with 1 × 10⁶ CFU of ΔihfABpmi, ΔtolRA, and 14028 WT (Figure 1C and Supplementary Figure S1). We observed lower mortality of larvae inoculated with the mutants than with the parental strain. Here, 90% and 50% of larvae inoculated with ΔtolRA and ΔihfABpmi, respectively, survived 96 h post-inoculation, while all larvae inoculated with 14028 WT died within 24 h of inoculation. These results suggest that the ΔihfABpmi and ΔtolRA mutants are virulence-attenuated.

Electroporation, used for constructing mutants, can result in the selection of mutant colonies with incomplete LPS. However, LPS is critical for bacterial immunogenicity. Therefore, we investigated the LPS integrity of the mutants by performing LPS preparations on polyacrylamide gels, followed by periodic acid oxidation and silver staining. We observed similar LPS profiles of the mutants and the wild-type strain, suggesting that the constructed mutants had a complete LPS layer (Figure 1D).

We determined whether gene deletion affected the invasiveness and survival of 14028 WT in tumor cells by infecting two cancer cell lines with the mutant and 14028 WT strains for 2 and 6 h, followed by determining the number of intracellular CFU after tumor cell lysis. In B16F10 murine melanoma cells, ΔtolRA showed a similar invasion and survival profile to the parental strain. However, the ΔihfABpmi mutant showed a statistically significant reduction in invasiveness and survival, and similar results were observed with bladder carcinoma 5,637 cells (Figures 2A, B).

In addition to invasion and survival, potential S. enterica Typhimurium mutants must also exert a cytotoxic effect. We assessed the in vitro antitumor potential of the mutants by performing the colorimetric assay with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) in B16F10 melanoma and 5,637 bladder carcinoma cells (Figures 2C, D). The MTT assay was performed with previously proposed modifications for the elimination of bacterial contribution to MTT reduction [33]. Cancer cells without the addition of bacteria were used as a negative control (100% viability), while cells treated with 14028 WT were used as a positive control. We observed that 14028 WT, ΔtolRA mutant and ΔihfABpmi mutant significantly decreased the viability of both cell lines (melanoma and bladder carcinoma cells). Treatment with 14028 WT decreased viability by 55%–68%, ΔihfABpmi mutant treatment decreased viability by 36%–54%, while the ΔtolRA mutant decreased viability by 32%–40%. B16F10 melanoma cells were more resistant to the cytotoxic effect of both mutants and the parental strain, likely due to their aggressive and metastatic phenotype.

We assessed treatment tolerance with attenuated mutants of S. enterica Typhimurium by inoculating healthy mice subcutaneously with the mutants once a week for 2 weeks, for a total of two doses (the experimental design is shown in Figure 3A). The mutant ΔihfABpmi dose (10⁵ CFU) was chosen based on our previous studies that showed that the double mutant ΔihfAB is an attenuated strain and 10⁵ CFU of the ΔihfAB mutant is a safe and efficient dose to treat bladder cancer in mice (unpublished data). Based on these results, we chose 10⁵ CFU of the triple mutant ΔihfABpmi as the experimental dose here. A previous study reported that the ΔtolRA mutant is highly attenuated [28], and our results in G. mellonella further confirmed this attenuation (Figure 1C and Supplementary Figure S1). Thus, we tested two doses of 10⁶ and 10⁷ CFU of the ΔtolRA mutant in safety analysis experiments.

The body weight of the mice was determined as a sign of general health. After the first inoculation, we observed that mice inoculated with 10⁷ CFU of the ΔtolRA mutant lost ±2 g (12% of body weight) with signs of disease, such as eye discharge, piloerection, and lethargy. However, 1 week after the first inoculation with 10⁷ CFU of the ΔtolRA mutant, the mice regained weight, and the disease symptoms disappeared (Figure 3B). In mice inoculated with PBS, 10⁶ CFU of the ΔtolRA mutant, or 10⁵ CFU of the ΔihfABpmi mutant, we did not observe weight loss or signs of systemic disease such as ocular discharge, piloerection, lethargy throughout the experiment (Figure 3B). Despite the symptoms, no mortality was observed in the mice throughout the experiment (Table 1).

The mice were euthanized 21 days after the initial inoculation, and the organs were collected for bacterial persistence and histopathological analysis. To analyze bacterial distribution and persistence at the end of the experiment, homogenized or macerated blood samples from the liver and spleen were diluted and plated on LB agar, SS, and MacConkey with appropriate antibiotics for subsequent CFU counting of our mutants. However, no bacterial colonies were detected in any of the three media tested or any tissue group, even when samples were plated without dilution. The fact that no mutants were isolated suggests that the bacteria cannot persist in these organs 21 days after the first inoculation and 15 days after the second inoculation.

Macroscopic examination of the organs showed hepatomegaly and splenomegaly in the group inoculated with 10⁷ CFU of the ΔtolRA mutant. Significant differences in spleen and liver weights were observed in mice inoculated with 10⁷ CFU of the ΔtolRA mutant, as compared to the group inoculated with PBS (Figures 3C, D). However, we did not observe significant differences between organ weights of mice inoculated with 10⁶ CFU of the ΔtolRA mutant or 10⁵ CFU of the ΔihfABpmi mutant compared with mice inoculated with PBS (Figures 3C, D).

Microscopic analysis of spleen, liver, lung, and kidney samples from mice inoculated with PBS revealed typical histological structures (Supplementary Figure S2). The spleen, liver, kidney, and lung of mice inoculated with the ΔihfABpmi mutant showed no signs of tissue damage. However, histological changes related to mild inflammation were observed in the liver, lung, and spleen. Interestingly, we also observed an increase in the megakaryocyte number in the spleen of mice inoculated with the ΔihfABpmi mutant as compared to that in mice inoculated with PBS. Megakaryocytes have previously been related to inflammation [34]. Despite these effects, no histological architectural changes were observed in the evaluated organs, suggesting that subcutaneous inoculation of 10⁵ CFU of the ΔihfABpmi mutant is safe for use as a treatment.

Based on the safety analysis described above, we performed the antitumor efficacy tests with 10⁶ CFU of the ΔtolRA mutant and 10⁵ CFU of the ΔihfABpmi mutant since, at these concentrations, no weight loss, splenomegaly, or hepatomegaly were observed, suggesting that these mutants are well-tolerated in mice at these concentrations. The antitumor efficacy of the mutants was evaluated in terms of the survival rate of the mice and the ability to reduce tumor mass. B16F10 cells (3 × 10⁶) were implanted subcutaneously in the dorsal flank region of female C57BL/6JUnib mice. When the tumor reached 100 mm³ (10–12 days after tumor cell inoculation), mice bearing B16F10 tumors were treated intratumorally or intraperitoneal with 10⁶ CFU of the ΔtolRA mutant or 10⁵ CFU of the ΔihfABpmi mutant. Mice bearing B16F10 tumors inoculated with PBS were used as a negative control. As shown in Figure 4, both the ΔtolRA mutant and the ΔihfABpmi mutant treatments reduced tumor growth compared with the PBS-treated group (control), and treatment with the ΔihfABpmi mutant was more efficient in reducing melanoma tumors than treatment with the ΔtolRA mutant.

Figures 4C, I show the tumor size at the end point of the experiment (19 days after the first dose of treatment with the mutants). Intratumoral treatment with the ΔihfABpmi mutant eliminated the tumor from all mice, leaving only a scar at the tumor site (Figure 4C). Moreover, we also observed that the ΔihfABpmi mutant completely reduced the tumor mass of all mice 6 days after the first dose of treatment, which did not grow back in the following 15 days (Figure 4E). Intraperitoneal treatment with the ΔihfABpmi mutant eliminated the tumor in two of the five treated mice (Figure 4I). On the other hand, although intratumor or intraperitoneal treatment with the ΔtolRA mutant inhibited tumor growth compared to the PBS control, it did not eliminate tumors in mice.

All mice inoculated with the mutants survived until the endpoint of the experiment. However, the mice in the PBS group were euthanized 1 week before the end point of the experiment to avoid further suffering due to substantial tumor growth. We did not observe any significant differences between the weight of the mice inoculated with the mutants and the PBS group (Figures 4B, H). We also did not observe any signs of disease in mice intraperitoneally or intratumorally treated with the ΔihfABpmi mutant.

In the endpoint of the experiment intratumorally, blood, liver, spleen, and tumor samples were homogenized or macerated and plated on LB agar, SS, and MacConkey for subsequent CFU counting. However, no bacterial colonies were detected in the analyzed groups (mice treated with PBS, ΔihfABpmi, or ΔtolRA), even when samples were plated without dilution. The non-isolation of the mutants suggests that the bacteria either cannot persist in these organs or get eliminated by the immune system of the mice. It would be interesting to determine the colonization efficiency of the two mutants tested in the tumor microenvironment to see if differences in antitumor activity can be attributed to bacterial fitness. Future analyses will include the analysis of bacterial colonization after 24, 48 and 72 h of treatment with the mutants.

Hematoxylin and Eosin (H&E) staining of tissues revealed inflammatory foci (infiltration of mononuclear and polymorphonuclear cells) in the liver and spleen of mice treated with the ΔihfABpmi mutant (Figure 5). However, we did not observe pathological changes in the architecture of the analyzed organs.

The immunological responses underlying the antitumor effects of S. enterica Typhimurium remain poorly understood. Recent evidence suggests the involvement of cells of the innate immune system, such as macrophages, and the production of pro-inflammatory cytokines in tumor elimination [15]. We used the ΔihfABpmi mutant to analyze the immune responses involved in its antitumor response because treatment with the ΔihfABpmi mutant showed a more significant tumor inhibition than treatment with the ΔtolRA mutant. Mice were euthanized, and tumor tissue was collected 4 days after treatment with 10⁵ CFU of the ΔihfABpmi mutant, as the ΔihfABpmi mutant rapidly inhibited tumor growth.

Tumor cell suspensions prepared from tumor tissue were labeled and analyzed using the following panel macrophage (F4/80+ CD11b+), M1 macrophage (F4/80+ CD80⁺) and M2 macrophage (F4/80+ CD260+). Flow cytometric analysis showed significant difference between the proportion of macrophages from mice treated with the ΔihfABpmi mutant and mice treated with PBS (Figure 6B). Next, we examined the phenotype of intratumoral macrophages, and observed that treatment with the ΔihfABpmi mutant significantly reduced the proportion of M2-type macrophages and increased the proportion of M1-type macrophages compared to PBS-treated mice (Figure 6C). This suggests that the tumor elimination ability of ΔihfABpmi is associated with the induction of the accumulation of macrophages with antitumor phenotype and the reduction of pro-tumor immunosuppressive macrophages.

We evaluated the expression of cell proliferation genes (Ki-67), angiogenesis (VEGF), apoptosis (Bax), and pro-inflammatory (IL-6, TNF-α, and iNOS) markers to further investigate the immune responses underlying the antitumor effects of ΔihfABpmi. Tumor tissue from mice treated with the ΔihfABpmi mutant or PBS was collected after 4 days to measure the levels of mRNA expression using qRT-PCR. We observed that treatment with the ΔihfABpmi mutant induced a statistically significant upregulation of the mRNA levels of Bax, IL-6, TNF-α, and iNOS (Figures 7A, B, D, F, respectively). Thus, pro-inflammatory cytokines secreted by immune cells such as macrophages can contribute to tumor cell death and trigger a strong tumor-specific immune response [35].