As a new type of anticancer drug carrier, bacteria will preferentially accumulate in tumors after entering the human body. Compared with healthy tissues, the accumulation of bacteria in tumor tissues is more than 1000 times higher [15]. Unlike liposomes, micelles, and NPs which can only passively target tumors by bloodstream transport, the bacteria have both passive and active targeting mechanisms and can actively penetrate deep into the tumor tissue [18,19]. The generation of the targeting property can be related to the suitable tumor microenvironment (TME) and the properties of bacteria. In addition, various modification methods can further improve the tumor-targeting ability of bacteria (Figure 1).

The external reasons for the tumor targeting of bacteria can be roughly divided into three points. Firstly, the hypoxia TME attracts obligate and facultative anaerobic bacteria to tumor tissue. Due to the rapid proliferation of tumor cells and incomplete vascular development, the supply of oxygen in solid tumors is insufficient, which eventually leads to the existence of hypoxic areas in tumors [20]. This is a necessary condition for the survival and reproduction of obligate and facultative anaerobic bacteria in tumors. Moreover, with further penetration of the tumor tissue, hypoxia tends to be more severe. This is very unfavorable for conventional treatment but is more conducive for bacterial tropism and depth of penetration [21,22]. Secondly, the rich nutrients in the tumor tissue are another important reason for attracting bacterial colonization. Kasinskas et al. demonstrated that aspartic acid, serine, ribose, and galactose in the tumor can help Salmonella chemotaxis [23]. Song et al. demonstrated that clusterin is one of the key biochemical molecules in E. coli chemotaxis of lung cancer cells [24]. Lastly, the Immunosuppressive properties of tumors benefit bacterial colonization [25]. Bacteria that colonize tumor sites in the early stages can avoid clearance by the immune system, while bacteria that travel to normal tissues will be cleared by the immune system. Therefore, the concentration of bacteria in the tumor site is often much higher than in other sites, and the bacteria do not colonize the non-tumor-related hypoxic or inflammatory lesions [26,27,28,29].

Some of the properties of bacteria can also help them to colonize tumors. It has been proposed that some chemical-specific receptors on the bacteria may sense chemicals secreted by cancer cells [19,23]. Kasinskas et al. proved that chemical receptors have an important effect on the tumor tropism of bacteria by knocking out different chemical receptors on the surface of Salmonella [23]. More interestingly, the accumulation of bacteria in different parts of the tumor can be controlled by selectively eliminating chemical receptor genes. In addition, some bacteria have special tendency properties. For example, magnetotactic bacteria have a tendency to a specific magnetic field [30]. By giving an auxiliary magnetic field in vitro, it can induce magneto-aerotactic bacteria to migrate to the tumor site.

There are two opposing views on the relationship between bacterial motility and tumor tropism. One side believes that motility is essential for the effective distribution and accumulation of bacteria in tumor tissues. Studies by Kasinskas et al. and Toley et al. both used in vitro models to demonstrate that motility is essential for the effective distribution of bacteria in tumors [23,31]. On the contrary, Stritzker et al. found that chemotaxis and motility did not seem to help the bacteria colonize the tumor by intravenously injecting E. coli and Salmonella strains into the BALB/c mice with 4T1 breast cancer [32]. However, this study only analyzed the distribution of bacteria within the tumor at 48h after injection. Later, Ganai et al. conducted a longer experimental study on the same animal model [33]. Interestingly, the results showed that the bacteria began to migrate from the tumor edge to the central core after 48 h, and began to move closer to the tumor transition zone after 96 h. This may indicate that the experiment of Stritzker et al. still has some flaws and, if they could do longer tests, they may give a different conclusion.

It is undeniable that the movement of bacteria appears to be very insignificant compared to the speed of blood flow [34]. Therefore, bacteria entering the animal’s body are mainly transported passively by blood flow to the tumor site. But, upon reaching the tumor, the motility of the bacteria plays a crucial role in helping the bacteria penetrate deeper into the core of the tumor. When bacteria penetrate the tumor tissue, they will multiply in the tumor. Combined with the previously described immune clearance phase in normal tissues, this results in a much larger number of bacteria at the tumor site than in normal tissues.

Gene editing and surface modification can further improve bacteria’s tumor-targeting ability. With the use of gene editing technology, it is possible to design modified auxotrophic bacteria corresponding to some specific purines, amino acids, and other nutrients within the tumor. Zhao et al. designed a leucine–arginine auxotroph S. typhimurium [35]. The engineered bacteria can only survive in tumors but not in normal tissues. In terms of surface modification, modifying tumor-homing peptides or tumor antibodies on the surface of bacteria can lead to better tumor-targeting ability. Park et al. improved the tumor tropism of bacteria by modifying an arginine–glycine–aspartate peptide on the outer membrane protein of S. typhimurium [36]. Massa et al. demonstrated that the tumor specificity of Salmonella can be significantly improved by surface-expressed antibodies against tumor-associated antigens [37].

Cancer patients cannot clear cancer cells through their immune system, because cancer cells can develop multiple pathways to avoid being cleared by the immune system [38]. It has been found that bacteria inside the tumor can alter the immunosuppressive TME and stimulate the host immune system, thus enhancing the body’s immune system to clear the cancer cells (Figure 1).

The reasons for the inability of the anticancer immune response to eliminate tumors can be broadly classified as follows:

• Tumor cells themselves have developed specialized mechanisms to suppress immune responses, including downregulation of tumor antigen and major histocompatibility complex (MHC) class I expression [39,40], high expression of programmed death receptor-ligand 1 (PD-L1) to prevent T cell activation [41], and expression of various immunosuppressive cytokines and chemokines by themselves or induced tumor-infiltrating immune cells [38,42].
• The presence of a large number of immunosuppressive cells within the tumor, such as tumor-associated macrophages (TAMs), regulatory T cells (Tregs), and myeloid-derived suppressor cells (MDSCs), can significantly inhibit the infiltration of cytotoxic T lymphocytes [43,44,45].
• The patient’s immune function is so weak that the growth rate of tumor cells exceeds the clearance rate of the immune system [46]. The combination of these factors makes it difficult for the body to rely on its immune system to remove tumor cells, thus allowing the tumor to grow and spread.

Bacteria at the tumor site can stimulate the immune system to produce a series of anticancer immune responses is another advantage of bacterial carriers in cancer therapy. This is because bacteria carry a large number of antigens such as lipopolysaccharide and flagellin, which can bind to toll-like receptors and, thus, trigger a series of cellular signaling events [47,48,49]. The mechanisms by which bacteria activate anticancer immunity in the body are complex and interact with each other. For the sake of description, these mechanisms are roughly divided into two parts: promotion of anticancer immune response and reduction of tumor immune escape (Figure 2).

In enhancing the anticancer immune response, the recruitment of immune cells to reach the tumor site is the most fundamental step. Several studies have demonstrated that bacteria at the tumor site can induce a large number of immune cells to accumulate toward the tumor, including macrophages, NK cells, B cells, CD4+ T cells, CD8+ T cells, and so on [50,51,52]. When a large number of immune cells reach the tumor site, a series of anticancer immune responses are triggered. Secondly, bacteria can increase the number of TANs and, thus, enhance the innate immune response [53]. However, neutrophils show two sides in the development of tumors [54,55,56]. TAN1 exhibits anticancer activity through direct cytotoxic effects and activation of adaptive immune responses, while TAN2 exhibits tumor-promoting activity by promoting tumor cell proliferation, migration, and invasion, stimulating angiogenesis and mediating immunosuppression. Furthermore, bacteria promote gap junctions between DCs and cancer cells by upregulating connexin 43 to enhance the ability of dendritic cells to present tumor antigens [57,58]. This process leads to the secretion of large amounts of the proinflammatory cytokine IL-1β by DCs and subsequent activation of CD8+ T cells. Lastly, bacteria were found to activate the NF-κB pathway, which in turn upregulates the expression of various immunostimulatory and chemokines, including IL-6, IL-13, IL-17, IL-1β, G-CSF, GM-CSF, MIP-1α, TNF-α, and IFN-γ [59,60,61,62].

Reducing tumor escape is another important means to activate antitumor immunity. As mentioned above, tumor immune escape is associated with a large number of immunosuppressive cells, including TAMs, Treg, and MDSCs, which protect tumor cells in various ways to treat the activation of cytotoxic T cells [62]. Among them, TAMs are usually divided into two opposite subtypes, including M1 macrophages that exert antitumor activity and M2 macrophages that inhibit T-cell-mediated antitumor immune response. Both M1 and M2 macrophages have a high degree of plasticity, so they can be transformed into each other when the tumor microenvironment changes or therapeutic interventions [63]. It was found that S. typhimurium expressing flagellin promoted the M2–M1 transition of macrophages and increased the level of nitric oxide in tumors through the synergistic effect with TLR4 and TLR5 signaling pathways [64]. In addition, flagellin stimulates NK cells to produce IFN-γ, which can also promote the conversion of M2 to M1 [65]. Treg cells and MDSCs are the other two important immunosuppressive cells. Treg cells can inhibit the costimulatory signal of CD80 and CD86 expressed by dendritic cells, consume IL-2, secrete inhibitory cytokines, and directly kill effector T cells [66]. MDSCs play an important role in tumor angiogenesis, drug resistance and tumor metastasis [67]. Studies have found that some bacteria can effectively reduce Treg and DMSCs in tumor tissues. For example, attenuated Salmonella vaccine has been shown to induce immunosuppressive myeloid-derived suppressor cells to transform into TNF-αsecreting cells with neutrophil characteristics and significantly reduce Treg cells [68,69,70]. Secondly, Listeria was found to be delivered to metastatic and primary tumors by infecting DMSCs, significantly reduce the number of MDSC in blood and primary tumors, and transform the remaining MDSC subsets into an immunostimulatory phenotype that produces IL-12 [71]. In addition, bacteria downregulate the expression of immunosuppressive factors such as IL-4, ARG-1, and TGF-β [62,72].

In addition to activating the immune system to kill cancer cells, bacteria colonizing tumors also have a certain anticancer effect, which can help drugs better play their anticancer role. This oncolytic activity can be achieved through a variety of pathways, including the induction of tumor cell death, inhibition of tumor angiogenesis, inhibition of tumor metastasis, and reduction of tumor drug resistance (Figure 1).

Bacteria induce cancer cell death through multiple pathways, including the induction of apoptosis, the release of bacterial toxins, and competition for nutrients. For example, Listeria can induce tumor cell apoptosis by activating nicotinamide adenine dinucleotide phosphate (NADPH) oxidase and increasing intracellular calcium levels to increase ROS levels in tumors [73]. Salmonella induces autophagy and caspase-mediated apoptosis in tumor cells by downregulating the AKT/mTOR pathway [74], and metabolizes nitrate to nitrite via nitrate reductase and further converts it to nitric oxide in tumors to induce tumor cell apoptosis [75]. Secondly, bacteria can release toxins to kill tumor cells. For example, Colicins have anticancer activity against breast cancer, colon cancer, bone cancer, and other human tumor cell lines [76]. However, bacterial toxins are a double-edged sword. Strong toxicity may lead to damage to other normal tissues while showing good tumor killing. Therefore, it is necessary to select the appropriate intensity of toxicity as well as the dose. Lastly, the growth and reproduction of bacteria at the tumor site can consume a large number of nutrients, and by competing with tumor cells for nutrients it can promote apoptosis of some tumor cells [77].

Tumor growth is closely linked to tumor angiogenesis. The rapid growth of tumor cells requires a large amount of oxygen and nutrients; therefore, a large number of blood vessels are generated inside the tumor to transport the materials required by the tumor cells. If angiogenesis fails, the lack of oxygen and nutrients limits the tumor diameter to 2–3 mm [78]. The hypoxic TME can stimulate the tumor to produce some adaptive responses. At this time, the tumor cells will activate the hypoxia-inducible factor (HIF) signal and upregulate the expression of vascular endothelial growth factor (VEGF) and other proangiogenic factors, thereby increasing tumor angiogenesis [79,80]. Salmonella was found to downregulate HIF-1α expression via downregulation of the AKT/mTOR pathway, which in turn inhibited tumor VEGF expression and angiogenic signaling [81].

The metastasis of malignant tumors is frequently the main cause of tumor treatment failure, and the degradation of extracellular matrix caused by matrix metalloproteinase-9 (MMP-9) plays a key role in this process. Cancer cells can break through the physical barrier of the extracellular matrix by influencing host cells to secrete MMP-9 to degrade multiple collagen proteins in the basement membrane to metastasize and invade other tissues [82]. Tsao et al. demonstrated in a mouse model of melanoma and lung cancer that Salmonella inhibited MMP-9 expression by downregulating the AKT/mTOR signaling pathway, resulting in the inhibition of tumor cell migration and reduction of nodule production in vivo [83].

Reducing the drug resistance of tumor cells is also one of how bacteria achieve oncolysis. p-glycoprotein (P-gp), also called multidrug resistance protein, is a protein that pumps certain intracellular chemicals out of the cell and reduces their concentration in the cell for cell protection [84]. The P-gp is also distributed on the surface of tumor cells, which allows the chemotherapeutic drugs inside the tumor cells to be excreted extracellularly, thus causing the tumor cells to develop drug resistance. In mouse melanoma and mammary tumor models, Salmonella reduced the expression of P-gp by inhibiting the expression levels of phosph-protein kinase B, phosph-mammalian targets of rapamycin, and phosphate-p70 ribosomal s6 kinase in tumor cells., thus improving the sensitivity of tumors to chemotherapy [85].