Salmonella typhimurium may support cancer treatment: a review

Abstract

Antitumour treatments are evolving, including bacteria-mediated cancer therapy which is concurrently an ancient and cutting-edge approach. Salmonella typhimurium is a widely studied bacterial species that colonizes tumor tissues, showing oncolytic and immune system-regulating properties. It can be used as a delivery vector for genes and drugs, supporting conventional treatments that lack tumor-targeting abilities. This article summarizes recent evidence on the anticancer mechanisms of S. typhimurium alone and in combination with other anticancer treatments, suggesting that it may be a suitable approach to disease management.

Introduction

Bacteria-mediated cancer therapy (BMCT) is concurrently an ancient and cutting-edge approach to cancer treatment. The therapeutic effect of some infectious diseases on malignancies was reported in the 1700s [1]. Over a century ago, Coley, an orthopedic surgeon, injected heat-inactivated Streptococcus pyogenes into unresectable bone and soft tissue sarcomas (STS) and observed tumor regression, with some of the treated patients experiencing symptom relief and prolonged survival associated with what became known as Coley’s toxins [ 2, 3] . However, without a clear understanding of the mechanisms of BMCT at that time, disappointing results in clinical practice followed. However, with the development of biological, medical, and especially genetic engineering, many mechanisms of BMCT have been clarified. BMCT can not only destroy cancer tissue by bacteria itself but also trigger antitumour immune responses and even overcome some of the limitations of traditional cancer therapy. The number of papers about BMCT has increased since the mid-1990s, and an increasing number of studies suggest that BMCT may be a suitable approach to cancer treatment.

Salmonella typhimurium ( ST) is a parthenogenic anaerobic bacterium that is most commonly used in BMCT studies. This review aims to summarize the evidence on ST’s unique antitumour mechanisms, including tumor-targeting colonization, intrinsic oncolytic capacity, immune response regulation, and susceptibility to genetic modification, which make it an ideal vector for drug and gene delivery. Studies combining ST with traditional therapies are also summarized. We hope our work can provide a foundation for further research and clinical application.

Main Advantages of S. typhimurium in Cancer Treatment

As the most-studied bacterial strain of BMCT, ST is obviously favoured for a reason. Compared to other bacteria, Salmonella has several special antitumour properties, which is summarized in Table 1.

Table 1 Examples of BMCT combined with traditional anti-cancer therapy

Traditional therapy	Combined with	Tumor
Chemotherapy	CHOP	B-NHL
	Doxorubicin	Breast cancer
	Cisplatin	Primary and metastatic hepatocellular and lung cancers
	Gemcitabine	Pancreatic cancer
	Cyclophosphamide	Melanoma
	OMV-coated paclitaxel	Colon, breast, and liver cancer
	OMV-coated tegafur	Melanoma
Radiotherapy	Bacteria carrying gold nanoparticles	Colon cancer
	Bacteria carrying PDA	Melanoma
Immunomodulatory factors	IL-2	Unresectable hepatocellular carcinoma, osteosarcoma,
		malignant melanoma
	IFN-γ	Malignant melanoma
	IL-18	Colon cancer
	CCL-21	Colon cancer, and melanoma
	LIGHT	Breast cancer, and colon cancer
Immune checkpoint inhibitors	shIDO-ST	Non-small cell lung cancer, colon cancer
	Anti PD-L1 antibody	Melanoma
	IDO-shRNA, anti-CTLA-4, PD-1 antibodies	Lung cancer
	siRNA-PD-1, pimozide	Melanoma
	siRNA-PD-1, nifuroxazide	Colon cancer
	siRNA-PD-1, chloroquine	Colon cancer
	siRNA-PD-1, CpG ODN	Malignant melanoma

Tumor-targeting colonization

S. typhimurium colonizes tumor tissues at a rate that is over 1000-fold higher than that at which it colonizes normal tissues [ 4, 5] . It was believed that a hypoxic and poorly vascularized environment within tumor tissue is preferred by ST [6]. Additionally, no bacterial accumulation was observed in nontumor hypoxic tissues, supporting the role of other factors in ST tumor selectivity [7]. Several studies have shown that chemotaxis and motility are the key factors [ 8– 12] , but Stritzker et al. [13] pointed out that it is the tumor microenvironment (TME), bacterial metabolism, and host reticuloendothelial system that cause this tendency [13]. However, regardless of the exact mechanism involved, ST tumor-targeting can be used to support drug delivery in conventional anticancer therapy, helping minimize dosing frequency and toxicity.

Wide tumor range coverage

ST cancer treatment has been proven effective in a variety of murine tumor studies, including melanoma [ 14, 15] , osteosarcoma [16], STS [17], non-small cell lung carcinoma [18], colorectal cancer [19], cervical carcinoma [20], prostate cancer [21], T-cell lymphoma [22], and follicular dendritic cell sarcoma [23]. Metastasis inhibition by ST was also observed in osteosarcoma [24], breast cancer [25], and spinal cord glioma [26].

Oncolytic capacity

ST achieves tumor lysis by inducing apoptosis in tumor cells. Nutritional deficiency and bacterial toxin release may promote apoptosis [ 27, 28] . ST can induce autophagy by downregulating the AKT/mTOR pathway in a time- and dose-dependent manner in melanoma cells [29]. The downregulation of this pathway inhibits the expressions of HIF-α and vascular endothelial growth factor (VEGF), interfering with intratumor angiogenesis, which slows tumor growth [ 30– 32] . In addition, ST lysis releases nitrate reductase, which converts nitrite and nitrate to nitric oxide, thereby inducing tumor cell apoptosis [ 33, 34] . Additionally, the expression of some oncoproteins is downregulated by ST in tumor cells, such as p-glycoprotein (p-gp) [35] which causes drug resistance, and matrix metalloproteinase 9 (MMP-9) [36] which promotes tumor metastasis.

Immunomodulatory effects

Salmonella infection induces a series of immune responses ( Figure 1). Therefore, since ST tends to colonize tumor tissues, the immune status in the TME will be changed. The TME contains immune cells, cancer-associated fibroblasts, and structural elements, such as collagen and hyaluronic acid, which form the extracellular matrix. They directly surround tumor tissue, promoting tumor growth, survival, immune escape, and metastasis [37]. Changes in the TME can affect the proliferation and metastatic capacity of tumors, limiting or promoting disease progression [38]. The immunosuppressive TME may limit the efficacy of conventional cancer therapies. Poor prognosis of cancer patients is associated with an immunosuppressive TME [ 39– 41] .

Figure 1 .

Main advantages in cancer treatment of Salmonella typhimurium

ST tends to colonize in multiple tumor tissues and can induce a series of anti-cancer immune responses. ST is also an ideal tumor-targeting vector which can deliver almost everything to tumor tissues.

The formation of an immunosuppressive TME can be attributed to tumor molecular heterogeneity and tumor-intrinsic dysregulated signaling pathways. By activating Wnt/β-catenin proteins in tumor cells, tumor-specific T cells can be removed from the TME [42]. By activating STAT3 signaling, the expression of tumor-secreted proinflammatory mediators can be significantly reduced [43]. Activation of the PI3K/PTEN/AKT pathway may negatively affect immune responses; for example, mutations such as PIK3CA activation or PTEN loss of function can activate the PI3K/AKT signaling, which increases the level of immunosuppressive tumor-associated macrophage (TAM) infiltration [44]. In addition, activating the PI3K/AKT signaling axis can upregulate indoleamine-2,3-dioxygenase (IDO), a tryptophan-degrading enzyme, in tumor cells and the TME [ 45, 46] . With high IDO activity, the availability of tryptophan to effector immune cells is deprived, which traps T cells in the stationary cell cycle (G1 phase). T cells sensing tryptophan deficiency trigger general control nonderepressible 2 (GCN2) kinase release, which mediates proliferative arrest and increases Treg populations [ 47, 48] . Moreover, kynurenine, which is produced in tryptophan catabolism, induces apoptosis of T cells [49]. It is believed that abnormal and high expression of IDO leads to poor clinical prognosis [ 50, 51] .

ST modulation of the immune system can convert the TME from immunosuppressive to immunogenic by increasing multiple innate and adaptive immune cells, such as CD4+ helper T cells [ 52– 54] , CD8+ cytotoxic T cells [ 53– 55] , B cells [56], macrophages [ 52, 54] , and NK cells [53], enhancing the anticancer immune response. Meanwhile, through TLR5 signaling, ST flagellin can reduce the number of CD4+CD25+ Treg cells in the TME, thus promoting immunosuppression disruption in the TME [ 57, 58] . By activating the NF-κB pathway, the expression of proinflammatory cytokines and chemokines, such as IL-1α, IL-1β, IL-6, IL-13, IL-17, G-CSF, GM-CSF, TNF-α, IFN-γ, MIP-1α, and iNOS [ 15, 59– 61] , is upregulated [ 62– 65] ; at the same time, immunosuppressive factors, such as ARG-1, IL-4, and TGF-β, are downregulated [ 15, 66] . In addition, phenotypic and functional activation of intratumor myeloid cells may be induced, reducing the immunosuppressive effect of the TME. The significant reduction in IL-4-IL-13/ARG1 axis activity may explain the phenomenon, which is a marker of the activity of immunosuppressive tumor-associated macrophages [15]. The activation of inflammatory vesicle pathways accompanied by increased levels of IPAF, NLRP-3, and caspase-1 has also been observed, alongside the enhancement of tumor antigen expression by dendritic cells [67]. Reduced IDO expression and activity, associated with inhibited AKT/mTOR/p70S6K signaling, have been reported [46]. The inhibition of this pathway also downregulates the expression of programmed death ligand 1 (PD-L1), which is a transmembrane receptor for PD-1 [68]. High PD-L1 expression on the tumor surface inhibits the activation of effector T cells, the production of cytokines, and TCR-mediated proliferation [69], which can be reversed by ST [68]. ST can also alter immune check-point expression (seen in “Immune checkpoint inhibitors”).

Deliver everything: DNA, RNA, and protein

Because of its cancer-targeting property, ST has become an ideal delivery system that can carry DNA, RNA, and protein to cancer tissues ( Figure 2). Compared with viral vectors, ST is a cheaper, safer and more accurate way to deliver therapeutic genes [70]. Some concepts and efforts are shown as follows.

Figure 2 .

Examples of different mechanisms followed by Salmonella typhimurium for cancer therapy

ST can increase multiple innate and adaptive immune cells like CD8+ T cells, NK cells and macrophages, it also reduces Treg cells in TME. Upregulation and activation of Cx43 expression by ST enhances antigen delivery in dendritic cells and facilitates the delivery of chemotherapeutic drugs and apoptotic signals within cancer cells. Using gene-modified strains, drug-loaded liposomes or OMVs carrying drugs/pro-drugs, anti-tumor genes/drugs can be delivered directly into tumor tissues by ST. Plasmid carrying cytosine deaminase can convert non-toxic 5-fluorocytosine (5-FC) to 5-fluorouracil (5-FU). This approach limits the toxicity of chemotherapeutic drugs to the tumor area.

DNA vaccines are a method of delivering plasmids encoding tumor-expressed antigens into tumor cells to induce a host immune response. By the control of a eukaryotic promoter, bacteria deliver plasmids encoding therapeutic genes into tumor cells. Once entered, bacteria release the plasmids into the cytoplasm and transfer them to the nucleus, where the therapeutic genes are expressed by the transcriptional and translational systems of the host cell. This process is called bactofection. Several trials have reported the inhibition of tumor growth and metastasis using oral attenuated ST bactofection of vascular endothelial growth factor receptor 2 (VEGFR2) into tumor cells, thus inhibiting intratumoral vascular growth [ 71– 73] . Similar oncogenic effects have been observed with IL-18 [74], survivin [75], and tumor endothelial marker 8 (TEM8) [76].

RNA interference (RNAi) can regulate mRNA stability and translation in almost all human cells. Short hairpin RNA (shRNA) molecules can effectively silence specific genes [77]. ST is a safe, effective, and inexpensive vector for delivering shRNAs. ShRNAs targeting STAT3 [78], multidrug resistance protein 1 (MDR1) [79], INHA [ 80, 81] , and Bcl-2 [82] have been tested in various tumor-bearing mouse models using attenuated ST, resulting in tumor growth inhibition and prolonged survival.

The use of bacteria to deliver proteins is a promising approach to cancer treatment, which uses bacteria to express therapeutic proteins that have therapeutic effects. Unlike bactofection, this treatment requires the continuous and stable presence of bacteria in the target tissue, e.g., the enzyme-prodrug approach. Attenuated ST can deliver cytosine deaminase to tumor cells, which converts nontoxic 5-fluorocytosine (5-FC) to 5-fluorouracil (5-FU), associated with antitumour effects. This approach limits the toxicity of chemotherapeutic drugs to the tumor area [ 83, 84] .

Most Commonly Used Attenuated Salmonella Strains

As mentioned above, ST is an ideal living anticancer nanomachine. However, ST is also the most common pathogen of food poisoning. It is obviously unwise and unfeasible to directly use the wild-type strains of ST for cancer treatment. Thus, gene-edited attenuated Salmonella strains were born. Here are the most-used attenuated Salmonella strains.

VNP20009

VNP20009 is derived from the ST ATCC14028 wild-type (WT) strain. It is a genetically attenuated strain with an excellent safety profile due to the absence of the purl and msbB genes, which reduces its bacterial virulence and potential to induce infectious shock while maintaining its antibiotic sensitivity. The strain is also genetically stable, and after multiple generations of in vitro and in vivo propagation, it still maintains its genetic and phenotypic stability.

In vivo experiments in a monkey model showed that the strain appeared in the bloodstream at 24 h after infection and at day 41 and was totally cleared from all organs [5]. VNP20009 exhibited significant tumor-targeting properties and colonized transplanted mouse tumors and various patient-derived organoid-based xenografts (PDOX) at a ratio higher than 1000:1 in tumor-bearing mice [5].

However, VNP20009 shows unsatisfactory efficacy in clinical trials [85]. In a trial of 25 patients (24 metastatic melanoma and 1 metastatic renal cell carcinoma) treated with VNP20009 (1×10 ⁶ to 1×10 ⁹ cfu/m ²) intravenously, the maximum tolerated dose was 3×10 ⁸ cfu/m ². In this trial, one patient (dose of 1×10 ⁹ cfu/m ²) presented with toxic reactions (dose-limited), including anemia, thrombocytopenia, persistent bacteremia, hyperbilirubinemia, nausea, vomiting, diarrhea, hypophosphatemia, and alkaline phosphatase elevation. In addition, VNP20009 increased the blood levels of multiple proinflammatory cytokines, such as TNF-α, IL-1β, IL-6, and IL-12. In three patients, VNP20009 tumor colonization was detected; however, no significant tumor regression was observed in any patient. Although intratumoral colonization was observed, VNP20009 did not show significant anticancer effects. The authors suggested that further studies may focus on reducing dose-related toxic effects or increasing tumor localization for therapeutic purposes [85]. However, excessive toxicity reduction of VNP20009 may result in its unsatisfactory efficacy. Some studies have shown that the addition of VNP20009 to LPS reduces its toxicity while decreasing its anticancer effects [86]. Therefore, VNP20009 may be more suitable as a delivery system. Some studies have attempted to enhance the anticancer properties of the strain. Cheng et al. [87] deleted phoP and phoQ genes from VNP20009 and obtained a strain characterized by reduced colonization in normal liver and spleen and increased tumor localization ability. The VNP20009 genome has been sequenced; in addition to the genes identified in knock-out studies, purM deletions, 108-kbp Suwwan deletions, and 50 nonsynonymous SNPs have been identified [88].

A1/A1-R

Strain A1 is derived from ATCC14028wt, which is a nutrient-deficient strain (leucine/arginine-dependent). This strain may grow aggressively in tumor xenografts, likely due to amino acid availability in tumor tissue that supports bacterial growth. It can also colonize normal tissues but can be cleared even in immunodeficient mice (excised thymic tissue). In vitro, A1 grows in PC-3 human prostate cancer cells and eventually destroys the nucleus. In vivo, intravenous injection of A1 into a murine model of PC-3 human prostate cancer cells showed that bacteria invaded tumor cells and sustained replication, arresting tumor growth. The tumor/liver bacterial ratio was approximately 2000–10,000:1 on day 4, and the bacteria could not be detected in normal tissues such as lung, liver, spleen, or kidney by day 15 after the injection. In contrast, the tumor regressed completely by day 20 after a direct injection of A1 into the tumor. Neither route of administration caused adverse host effects. Safety testing showed that under the administration of the 1×10 ⁷ cfu/m ² A1 strain, no mouse deaths were observed, whereas mice receiving ATCC14028wt all died within 3 days [89]. A1-R is a modified A1 strain with high antitumour virulence. Compared with A1, the adhesion ability is approximately 6 times stronger for A1-R to HT-29 human colon cancer cells [90]. In vitro, within 200 min, all A1-R-infected PC-3 cells were ablated; meanwhile, in vivo, A1-R colonized PC-3 tumor cells at 100-fold higher concentrations than A1 [91]. Studies involving PDOX models have shown that this strain performs well in malignant and metastatic tumors, such as prostate cancers, breast cancers, spinal tumors, pancreatic cancers, malignant melanoma, glioma, STS, and cervical and ovarian cancers [ 92– 96] . Meanwhile, A1-R may induce the differentiation of cell cycle quiescent tumor cells to the S/G2/M phase to enhance their response to chemotherapy [97].

ÄppGpp

The pathogenicity of Salmonella is associated with certain genes known as Salmonella Pathogenicity Island (SPI) [98]. At least five SPIs have been discovered. One of the most studied SPIs, SPI-1, which encodes type III secretion system 1 (T3SS-1), is related to host cell infection. The transcription of SPI-1 is regulated by various factors and is especially regulated by a transcription regulator, hilA, which is also encoded within SPI-1 [ 99, 100] . It was found that the expression of genes encoded by SPI-1 and hilA depends on a stringent signal molecule, ppGpp [ 101, 102] . ppGpp is synthesized by two enzymes, RelA and spoT [101]. Thus, when RelA and spoT are deleted, attenuated ÄppGpp strains are generated. All SPI-1 genes, including hilA, will not be expressed in this attenuated strain. Studies have shown that in female BALB/c mice, the LD ₅₀ of the attenuated strain is approximately 10,000-fold higher than that of the WT strain (oral, 3×10 ⁹ CFU/m ² vs 4×10 ⁴ CFU/m ²; intraperitoneal, 3×10 ⁶ CFU/m ² vs<10 CFU/m ²), regardless of the route of administration. At the same time, the attenuated strain can also effectively induce cellular and humoral immunity [61]. All of these factors make ÄppGpp an excellent vector for targeting tumors. Tan et al. [103] used engineered ÄppGpp ST expressing cytolysin A (ClyA) to treat pancreatic cancer-bearing nude mice and observed that ST markedly accumulated and proliferated in tumor tissues and that tumor growth was successfully inhibited. A previous study compared ÄppGpp and VNP20009 in MC38 xenograft mice. They showed similar effects in cancer suppression, while the ÄppGpp group had a higher rate of complete tumor eradication and longer survival time; and compared with the VNP20009 group, the levels of inflammatory cytokines of the ÄppGpp group were much higher in the TME, while those in circulation were lower [104]. Therefore, ÄppGpp may promote the selection of BMCT.

Combination with Traditional Anticancer Therapy

Some studies suggest that the antitumour effect of Salmonella is related to its toxicity [ 105, 106] , and the results of several clinical trials support this view [ 85, 107] . How to balance the antitumour efficacy and toxicity of Salmonella is a topic that needs further research. In addition to modifying the strain itself, combining existing anticancer treatments seems to be a more convenient and feasible approach. Traditional cancer treatment methods have been proven to be effective in many clinical applications but at the same time bring serious and even fatal toxic side effects to patients. Poor tumor targeting is one of the main reasons. The tumor-targeting property of ST can compensate for this shortcoming. At present, a considerable number of studies have combined ST with traditional anticancer treatments to evaluate their anticancer effects ( Table 1). The results are inspiring. Combined treatment not only exerts the advantages of ST and traditional treatment methods but also has a sensitizing effect on each other.

Chemotherapy

Currently, chemotherapy is still the first-line anticancer therapy for most cancers. However, the associated systemic side effects and tumor resistance limit its efficacy. Many studies have examined the use of ST in combination with chemotherapeutic agents. Bascuas et al. [108] used attenuated ST combined with CHOP (cyclophosphamide, adriamycin, vincristine, and prednisone), which is the first-line chemotherapy regimen for B-cell non-Hodgkin’s lymphoma (B-NHL), and observed tumor growth-slowing and longer survival time in mice. This study also mentioned that the combination treatment is more advantageous than any single treatment, given that the bacterial vector acts as an immunotherapeutic agent, and fewer systemic toxicities of chemotherapy were observed with intratumoral administration. This work successfully demonstrated that combination chemotherapy with ST is a safe and effective cancer treatment [108]. In a study of breast cancer, mice receiving intravenous attenuated ST and low-dose chemotherapy (doxorubicin, 1.25 mg/kg) showed an acceptable rate of tumor growth (<3-fold volume at day 35, superior to low-dose doxorubicin monotherapy) and 5% weight loss. Although mice given an extreme dose of doxorubicin intravenously (5 mg/kg) showed a more pronounced slowing of tumor growth (1.4-fold volume), toxicities were significant in this group (25% weight loss) [109]. A separate study demonstrated the efficacy of attenuated ST carrying an anti-vascular growth gene in combination with cisplatin in primary and metastatic hepatocellular and lung cancers [110]. ST A1-R may improve gemcitabine efficacy in pancreatic cancer [111]. ST VNP20009 in combination with cyclophosphamide achieved good results in mice with B16F10 melanoma [112].

Several possible mechanisms may account for these efficacy-enhancing effects of the bacterial-chemotherapy combination, including ST anticancer and immune-mediating properties. Tumor-targeting and colonization abilities of bacteria-chemotherapy combinations that are greater than those of each agent alone may be attributed to the disruption of tumor microcirculation and creation of a local hypoxic tumor environment associated with the combination therapy [ 59, 110, 112] . Chang et al. [113] suggested that the upregulation of connexin 43 (Cx43) expression by ST may increase treatment efficacy. Reduced Cx43 expression in a variety of cancer cells, which attenuates gap junction intracellular communication, makes tumor cells less responsive to treatment [114]. In contrast, upregulation and activation of Cx43 expression by ST enhances antigen delivery in dendritic cells and facilitates the delivery of chemotherapeutic drugs and apoptotic signals within cancer cells [ 67, 113] . Yang et al. [115] found that Salmonella cholerae inhibited the expression of p-gp on tumor cells, which is also known as MDR1, associated with removing foreign substances from the cell. Therefore, decreased expression of p-gp decreases the pumping ability of tumor cells, thus reducing tumor cell chemoresistance. In addition, the expressions of p-p70S6K, p-AKT, and p-mTOR were reduced, while high expression of p-AKT reversed S. cholerae-mediated downregulation of p-gp. Overall, S. cholerae may enhance the sensitivity of tumor cells to chemotherapeutic drugs by downregulating p-gp expression levels [35]. It can also push quiescent tumor cells, which remain in G0/G1 phase, into the S/G2/M phase to enhance tumor cell drug sensitivity [ 116, 117] . Studies have shown that most cancer cells (nearly 80%) are in the stationary cell cycle phase (G0/G1 phase), and even more (up to 90%) are in tumor centers, while most current drugs effectively kill proliferating cancer cells but not stationary cancer cells [118]. Using CDDP-resistant osteosarcoma lung metastasis PDOX model mice, Hoffman et al. [119] compared the efficacy of CDDP alone, the ST A1-R+CDDP combination, and the ST A1-R+rMETase+CDDP combination. Among them, rMETae, a recombinant methionine enzyme, selectively traps S/G2 phase tumor cells and causes cell cycle arrest. Significant tumor suppression was observed in all conditions, except single CDDP, creating a novel approach to tumor therapy called “induce, trap, and kill” [ 16, 116] .

Bacterial outer membrane vesicles (OMVs) brought new ideas to this combination. OMVs are released from the membrane of gram-negative bacteria during their natural growth [ 120– 122] . They contain most of the immunogenic components of their parent bacteria, so OMVs have the capacity to stimulate anticancer immune responses [123]. Additionally, like their parent bacteria, they prefer to colonize tumor tissues, so they seem to be ideal carriers for tumor-targeting drug delivery [124]. Because they cannot self-replicate, it seems safer to use attenuated bacteria directly. Rasha et al. [125] demonstrated the safety of ST-OMVs. Additionally, they used OMVs combined with paclitaxel to treat HTC116, MCF-7 and HepG2 tumor-bearing mice and observed a significant decrease in tumor volume, a significant increase in the tumor growth inhibition rate, and a series of tumor-suppression immune responses. Chen et al. [126] coated tegafur (prodrug of 5-FU) in OMVs, prepared engineered OMV-coated polymeric nanoparticles, and used them to treat melanoma-bearing mice, successfully inhibiting tumor growth and prolonging survival. In addition, they found that the nanoparticles effectively inhibited the lung metastasis of melanoma. These attempts provide a new approach to tumor-targeted chemotherapy.

Radiotherapy

Radiation therapy is an anticancer modality prescribed to approximately 50% of cancer patients [127]. However, radiotherapy does not target the tumor and may cause tumor resistance, limiting its efficacy. A cumulative therapeutic effect of attenuated ST combined with radiotherapy has been reported in various tumor cells; bacteria may also increase the radiosensitivity of tumor cells [ 128– 130] . The increased radiosensitivity of tumors may be caused by changes to the immune components of the TME, including increased count and activity of dendritic cells, recruitment and activation of effector T cells, suppression of T-cell depletion, elimination of inhibitory signals, and release of cytokines (IL-2, GM-CSF, and IL-12) and chemokines (CCL3 and CCL5) [131]. Meanwhile, Kefayat et al. [132] used attenuated ST Ty21a as a vector for delivering gold nanoparticles to the central hypoxic radiation-resistant region of CT26 colon cancer. Gold nanoparticles are an excellent radiosensitizer. The bacteria-sensitizer combination enhances the efficacy of radiotherapy in the central hypoxic region of the tumor [133]. Chen et al. [134] used ST VNP20009 covered with polydopamine (PDA), a photothermal agent that converts near-infrared light into heat, to enhance the radiotherapy efficacy in mice bearing melanoma. VNP20009 delivers PDA to the hypoxic part of the tumor before radiotherapy. The additional heat generated by PDA locally raises the temperature and helps kill the surrounding cancer cells. This combination treatment helped eliminate melanoma without evidence of recurrence or distant metastasis [134].

Immunomodulatory factors

Immunomodulatory factors have been used in cancer treatment. The United States Food and Drug Administration (FDA) approved IL-2 for the treatment of metastatic kidney cancer and metastatic melanoma [135]; meanwhile, in Europe, IFN-α has been used in hairy cell leukemia [136]. However, due to limited efficacy and severe toxicity associated with systemic administration, immunomodulatory factors have been withdrawn from clinical use [ 137, 138] . ST can be used as a vector for targeted delivery of immunomodulatory factors to tumor tissues, helping maximize therapeutic effects and minimize toxicity. In a study using ST carrying the IL-2 gene to treat mice with unresectable hepatocellular carcinoma, the combination therapy showed antitumour and antimetastatic effects superior to those of ST monotherapy. This study also demonstrated that CD8+ T cells and NK cells have anticancer activity [139]. Similar results were reported in trials of this combination in osteosarcoma [140] and malignant melanoma [32]. In a clinical trial (phase I) in 2020, patients bearing metastatic gastrointestinal tumors treated with attenuated ST carrying the IL-2 gene showed no significant toxicity alongside a significant NK cell and NK-T-cell increase in the circulatory system. However, patient prognosis and survival were not improved [107]. Yoon et al. [141] used ST carrying the IFN-γ gene to treat mice with B16F10 malignant melanoma cells and successfully inhibited tumor development and prolonged survival by activating NK cells. However, the authors concluded that the oncogenic effect of ST was due to direct killing of tumor cells rather providing a stable anticancer immune response. Another study showed that ST carrying TNF-α had similar anticancer effects and could enhance the efficacy of other treatments ( e.g., chemotherapy) [142]. Several other studies using bacteria carrying other cytokines (e.g., CCL21, LIGHT, IL-18, among others) yielded good results [ 74, 143, 144] . Overall, this evidence suggests that ST is an effective and safe carrier of immunomodulatory factors.

Immune checkpoint inhibitors

Immune checkpoint inhibitors changed cancer immunotherapy. By inhibiting negative regulators of T-cell function and restoring T-cell activity, agents such as CTLA-4, PD-1, and PD-L1 kill tumor cells. Great efficacy has been observed by using immune checkpoint inhibitors in the treatment of different cancer types [ 145– 148] . However, severe, even fatal, immune-related adverse reactions have been observed in some patients [ 149, 150] .

A study by Ebelt et al. [151] on attenuated ST containing an IDO-targeted shRNA plasmid (shIDO- ST) in suppressing the expression of CTLA-4, PD-1, and PD-L1 in different splenic immune cells highlighted the ability of ST to change the immune components of the TME. By using ST A1-R carrying an anti-PD-L1 antibody, Binder et al. [152] successfully rescued the function of tumor-specific CD8+ T cells against melanoma and enhanced cancer rejection. Phan et al. [153] used attenuated shIDO- ST to treat two murine models of colon cancer (CT26 and MC38), showing delayed tumor growth; however, a combination with anti-PD-1 did not show additional tumor-suppressive effects. A study using IDO-shRNA- ST in combination with anti-CTLA-4 and PD-1 antibodies to treat lung cancer-bearing mice reported a delay in tumor growth associated with the combination treatment. The increased infiltration of CD4+ and CD8+ T cells may be the reason [151]. Zhao et al. [154] used attenuated ST carrying siRNA-PD-1 plasmids (PD-1-siRNA- ST) in combination with pimozide to treat B16 melanoma mice and observed tumor suppression and prolonged survival of the mice, suggesting that this combination stimulates the immune response and enhances the antimelanoma effect of pimozide. Similar results were obtained using PD-1-siRNA- ST in combination with nifuroxazide (Stat3 inhibitor) to treat CT26 colon cancer mice [155], combination regimens with chloroquine for colon tumors [156] and with CpG ODN for malignant melanoma [157] have been used with good results. Overall, this evidence suggests that an ST-immune checkpoint inhibitor combination may effectively treat tumors, while ST may be a high-quality tumor-targeting delivery vector.

Conclusions and Prospects

ST’s anticancer effects are associated with its oncolytic properties and abilities to change the immunosuppressive TME and to trigger an immune response. ST can also colonize and proliferate in tumor tissues and can be genetically modified to enhance its colonization ability and reduce bacterial virulence, making it an ideal delivery vector for anticancer compounds, which have been used in studies on combinations with chemotherapy, radiotherapy, and immunotherapy.

However, clinical trials have yielded less favorable results, including reports of tumor progression and metastasis [ 85, 107] , suggesting that ST technology needs to be improved before it can be safely and effectively used in BMCT. Some of the clinical trials of ST-based cancer therapy are listed in Table 2. ST is a pathogenic bacterium, and excessive attenuation may compromise its oncolytic and immune system-stimulating properties. Identifying the clinically optimal strain dose requires further investigation. As with conventional anticancer agents, the balance between toxicity and therapeutic efficacy is critical and challenging in BMCT. Oral administration tends to be safe and convenient; however, it may reduce treatment efficacy due to exposure to the acidic environment of the gastrointestinal tract, which kills most bacteria. Meanwhile, intravenous administration may cause toxicity and infection, while direct intratumoral injection seems to be the most ideal method but still requires more evidence. Further studies are required to identify the optimum administration route. Overall, ST is a promising potential contributor to cancer therapy; however, further evidence is required before it can be used in the clinic.

Table 2 Examples of clinical trials for Salmonella-based cancer therapy

Strain	Cargo	Cancer type	Phase	Status	Enrollment	NCT No.	Website
VNP20009	Bacteria alone	Neoplasm metastasis	I	Completed	45	NCT00004988	https://clinicaltrials.gov/ct2/show/NCT00004988
VNP20009	Bacteria alone	Unspecified solid tumor	I	Completed	45	NCT00006254	https://clinicaltrials.gov/ct2/show/NCT00006254
VNP20009	Bacteria alone	Unspecified solid tumor	I	Completed	40	NCT00004216	https://clinicaltrials.gov/ct2/show/NCT00004216
x4550	IL-2	Liver cancer, biliary cancer	I	Completed	22	NCT01099631	https://clinicaltrials.gov/ct2/show/NCT01099631
VXM01	VEGFR-2	Stage IV pancreatic cancer	I	Completed	72	NCT01486329	https://clinicaltrials.gov/ct2/show/NCT01486329
TXSVN	Survivin	Multiple myeloma	I	Recruiting	18	NCT03762291	https://clinicaltrials.gov/ct2/show/NCT03762291
SS2017	Tyrosine hydroxylase Phox2B, Survivin, MAGEA1, MAGEA3, and PRAME	Relapsed Neuroblastoma	I	Recruiting	12	NCT04049864	https://clinicaltrials.gov/ct2/show/NCT04049864
Saltikva	IL2	Metastatic pancreatic cancer	II	Recruiting	60	NCT04589234	https://clinicaltrials.gov/ct2/show/NCT04589234
SGN1	L-Methioninase	Advanced solid tumor	I	Not yet recruiting	50	NCT05038150	https://clinicaltrials.gov/ct2/show/NCT05038150
SGN1	L-Methioninase	Advanced solid tumor	I	Not yet recruiting	50	NCT05103345	https://clinicaltrials.gov/ct2/show/NCT05103345

COMPETING INTERESTS

The authors declare that they have no conflict of interest.

Funding Statement

This work was supported by the grant from RuiJin Hospital Luwan Branch, Shanghai Jiaotong University School of Medicine.