Introduction

The primary limitation of cancer therapy is lack of selectivity of therapeutic agents for tumor cells. Current efforts are focused on discovering and developing anticancer agents that selectively target only tumor cells and spare normal cells to improve the therapeutic index. The use of preferentially replicating bacteria as an oncolytic agent is one of the innovative approaches for the treatment of cancer. This is based on the observation that some obligate or facultative anaerobic bacteria are capable of multiplying selectively in tumors and inhibiting their growth. Salmonella typhimurium, a facultative anaerobe, has been employed as an antitumor agent that is capable of preferentially amplifying within tumors and inhibiting their growth^1,2,3. Furthermore, these tumor-targeting bacteria have been used to deliver genes encoding angiogenic inhibitors^4,5,6, prodrug-converting enzymes⁷, or cytokines⁸, aiming to enhance their oncolytic effects.

A hypoxic microenvironment is shared by many solid tumors. Hypoxia is also associated with a more malignant phenotype, which affects genomic stability, apoptosis, angiogenesis, and metastasis⁹. The hypoxic regions of tumors are less sensitive to ionizing radiation because its cell-killing effects depend on oxygen; they are also less sensitive to chemotherapeutic agents because drugs delivered to these regions may be suboptimal. The hypoxic areas of tumors are poorly vascularized and are not likely to be treated easily with conventional anticancer agents¹⁰.

Since hypoxia is a common characteristic of human tumors, which adversely affects the prognosis of cancer patients, targeting the hypoxic regions of tumors may increase the effectiveness of cancer treatment. We have used a vaccine strain of Salmonella choleraesuis as a live vector to carry DNA vaccines¹¹. Recently, we demonstrated its tumor-targeting potential in various murine tumor models^4,5. In this study, we exploited S. choleraesuis as a single tumor-targeting anticancer agent and as part of a combination therapy with chemotherapy for mice bearing syngeneic tumors. Our results indicate that the combination of S. choleraesuis and cisplatin exerts additive therapeutic effects in delaying tumor growth and prolonging survival of the tumor-bearing mice.

Results

Tumor-targeting potential of S. choleraesuis in immunocompetent mice and bacterial colonization in the hypoxic regions of tumors

We monitored the kinetics of the bacterial distribution after injection with 2 10⁶ colony-forming units (cfu) of S. choleraesuis into mice bearing syngeneic LL/2 lung tumors (Fig. 1A) or ML-1 hepatomas (Fig. 1B) implanted subcutaneously (s.c.). The bacterial amount was much higher in tumors than in livers and spleens in both strains of the mice at all the time points examined. Their amounts in the tumors continuously remained at a peak level for at least 4 weeks after bacterial inoculation and were approximately 3 to 5 orders of magnitude higher than those found in the livers or spleens. Whereas it was retained in the tumors at a constant level per gram of tumor, we detected S. choleraesuis to a much lesser extent in the spleens or livers at day 28. Notably, as tumors became larger at later time points examined, total bacterial number within the tumor actually increased with time, indicating that the bacteria not only preferentially accumulated but also amplified within the tumors. By contrast, as early as day 7, S. choleraesuis was undetectable in the blood from all the mice tested (data not shown). Determination of the LD₅₀ of the attenuated S. choleraesuis after intraperitoneal administration in mice revealed that its virulence was reduced by approximately 10⁶-fold, compared with that of a local virulent S. choleraesuis strain, Dadin, isolated from swine.

Figure 1.

Preferential accumulation of S. choleraesuis (S.C.) in the tumors and its colonization in the hypoxic regions of tumors from mice administered systemically with S. choleraesuis. Mice bearing (A) LL/2 or (B) ML-1 tumors were injected i.p. with 2 10⁶ cfu of S. choleraesuis at day 0. The amounts of accumulated S. choleraesuis in the tumors, livers, and spleens were determined at days 7, 14, 21, and 28. Each value represents the mean SD from five mice. (C) Detection of S. choleraesuis in the hypoxic regions of tumors. LL/2 tumor-bearing mice were injected i.p. with S. choleraesuis (2 10⁶ cfu), and at 28 days p.i. tumors and spleens were removed for immunofluorescence double staining with anti-Hypoxyprobe-1 antibody to visualize hypoxic regions by Texas red and with anti-S. choleraesuis serum to visualize the bacteria by fluorescein. Nuclei were counterstained with DAPI. The Merge column represents the superposition of the tumor sections stained with anti-Hypoxyprobe-1, anti-S. choleraesuis, and DAPI to visualize colocalization. Spleen was used as a negative control for hypoxic staining. Localizations of hypoxic regions and S. choleraesuis were observed under fluorescence microscope at an original magnification of 200.

Full figure and legend (196K)

To examine whether S. choleraesuis targeted the hypoxic regions in tumors, we injected S. choleraesuis into LL/2 tumor-bearing mice and observed the bacterial distribution and hypoxic regions within tumors. As shown in Fig. 1C, S. choleraesuis predominantly resided, not absolutely but very closely, in the hypoxic regions of the tumor, whereas it was scarcely detected in the spleen that was not hypoxic. In addition to being hypoxic, tumor microenvironment is very likely to be immunosuppressive, while the spleen is significantly more immunologically active. Therefore, the immunosuppressive microenvironment around the tumor may also contribute, in part, to the preferential accumulation of S. choleraesuis in tumors.

Bacterial accumulation and antitumor effects observed in both treated ipsilateral and untreated contralateral tumors following intratumoral injection of S. choleraesuis

We also examined whether S. choleraesuis could target the untreated tumor when injected intratumorally (i.t.) into one of the bilaterally implanted LL/2 tumors. Fig. 2A shows that S. choleraesuis, when injected i.t. into the left tumor, preferentially accumulated not only in the treated tumors but also in the untreated contralateral tumors at 7 days postinfection (p.i.), although the bacterial number in the treated tumor was slightly higher than that in the untreated one. The growth of both tumors ipsilateral (Fig. 2B) and contralateral (Fig. 2C) to the bacteria injection site was significantly retarded in the S. choleraesuis-treated mice compared with that in PBS-treated mice. We also noted preferential accumulation of S. choleraesuis in bilaterally implanted tumors after intraperitoneal (i.p.) injection of the bacteria (Fig. 2D), resulting in retarded bilateral tumors (Figs. 2E and 2F). Taken together, these results indicate that S. choleraesuis, administered via either intratumoral or systemic route, was able to accumulate in the tumors at distant sites and, as a consequence, conferred contralateral antitumor effect. Thus, this tumor-targeting property of S. choleraesuis provides the impetus to explore its use in inhibiting tumor growth at distant sites.

Figure 2.

Preferential accumulation of S. choleraesuis (S.C.) in the bilateral tumors from mice administered locally or systemically with S. choleraesuis. S. choleraesuis (2 10⁶ cfu) or PBS was injected i.t. into the left flank tumor or i.p. to C57BL/6 mice bearing bilateral LL/2 tumors at day 8. The amounts of accumulated S. choleraesuis (means SD, n = 4) in the left (L) and right (R) flank tumors, livers, and spleens of mice injected (A) i.t. or (D) i.p. with the bacteria were determined at day 15. Differences in the tumor volumes (means SEM, n = 4) of the (B, E) left and (C, F) right flank tumors between the mice treated with S. choleraesuis and PBS were compared at day 16. *P < 0.05.

Full figure and legend (117K)

Preferential accumulation of S. choleraesuis in orthotopic tumors

To ensure that the tumor-targeting potential of S. choleraesuis was not confined to the subcutaneous site, but rather was a general phenomenon, we investigated its tumor-targeting potential in more clinically relevant cancer models, namely the ML-1 model of orthotopic hepatoma and the highly aggressive LL/2 model of experimental metastasis. We inoculated mice bearing ML-1 liver nodules or LL/2 lung nodules, and normal healthy mice, i.p. with S. choleraesuis and determined the amounts of S. choleraesuis in the tumors, livers, and spleens after bacterial inoculation. We found a notable amount of S. choleraesuis in the orthotopic hepatoma, whereas S. choleraesuis was detected to a much lesser extent in the spleens or healthy livers (Fig. 3A). As shown in Fig. 3B, although there was no significant difference between tumor-bearing and normal healthy mice in the amount of S. choleraesuis that accumulated in the livers and spleens, where no tumors were found, the bacterial number per gram tissue of the lungs with LL/2 nodules was significantly higher than that from normal healthy mice. Taken together, S. choleraesuis accumulated in not only subcutaneous but also orthotopic tumors after systemic administration.

Figure 3.

Preferential accumulation of S. choleraesuis (S.C.) in the orthotopic tumors from mice administered systemically with S. choleraesuis. (A) BALB/c mice bearing orthotopic ML-1 tumors were injected i.p. with S. choleraesuis (2 10⁶ cfu) at day 18. The amounts of accumulated S. choleraesuis (means SD, n = 4) in the tumors, livers, and spleens were determined at day 38 (P < 0.001 for tumor versus liver or spleen). (B) C57BL/6 mice that had been injected with LL/2 cells (10⁵) via the tail vein at day 0 were injected i.p. with S. choleraesuis (2 10⁶ cfu) at day 15. A parallel experiment was performed using normal healthy mice injected with S. choleraesuis. The amounts of S. choleraesuis accumulated in the lungs, livers, and spleens from mice with pulmonary metastasis (means SD, n = 8) and normal healthy mice (means SD, n = 7) were determined at day 20. ^***P < 0.001.

Full figure and legend (51K)

Reduction of metastatic nodules and prolongation of survival time of the mice bearing experimental metastasis by systemic delivery of S. choleraesuis

Because inhibition of metastatic tumor growth is still a major challenge for cancer treatment, we next investigated whether S. choleraesuis could inhibit established pulmonary tumor nodules. We treated mice with the bacteria 15 days after establishment of pulmonary nodules and monitored their growth 5 days later. We found that tumor nodules with various sizes were disseminated and overlapped, making the quantification of tumor nodules difficult. To facilitate quantification of the metastatic tumor burden in the lungs, we induced pulmonary metastasis by using LL/2-Luc cells expressing luciferase. We injected mice with LL/2-Luc cells via the tail vein at day 0, treated them i.p. with S. choleraesuis or PBS at day 15, and killed them at day 20. A majority of the lungs from PBS-treated mice possessed numerous confluent pulmonary nodules, whereas lungs from the mice treated with S. choleraesuis exhibited smaller and fewer tumor nodules. To quantify further the tumor burden, we measured the wet lung weight and luciferase activity. The mice treated with S. choleraesuis had 24% less wet lung weight compared with those treated with PBS (Fig. 4A). Moreover, the level of luciferase expression in the lungs from S. choleraesuis-treated mice was approximately 75% less than that from PBS-treated mice (Fig. 4B). S. choleraesuis treatment also significantly prolonged the survival time of the mice with pulmonary metastasis compared with their PBS-treated counterparts (Fig. 4C). Collectively, these results indicate that systemic delivery of S. choleraesuis delayed tumor growth in the lungs and enhanced survival of the mice bearing pulmonary metastatic tumors.

Figure 4.

Antitumor effects of S. choleraesuis (S.C.) on mice bearing pulmonary metastatic tumors. Groups of eight C57BL/6 mice that had been inoculated with 10⁵ (A, B) LL/2-Luc or (C) LL/2 cells via the tail vein at day 0 were treated i.p. with S. choleraesuis (2 10⁶ cfu) at day 15. (A) The wet lung weight and (B) luciferase activity of the lungs were measured at day 20. Each value represents the mean SD (n = 8). (C) Kaplan–Meier survival curves at day 40 are shown. ^**P < 0.01.

Full figure and legend (71K)

Additive antitumor effects of the combination therapy of S. choleraesuis and cisplatin

Although S. choleraesuis monotherapy was effective in retarding the growth of established lung cancer, complete tumor regression was not observed. We, therefore, determined whether it was more effective if administered in conjunction with chemotherapy. We chose to use cisplatin because this compound has been frequently used in the clinical treatment of hepatoma and lung cancer. We evaluated the antitumor effects of S. choleraesuis alone or combined with low-dose cisplatin in terms of tumor growth and survival of the mice bearing ML-1 or LL/2 tumors. In mice bearing either ML-1 or LL/2 tumors, treatment with S. choleraesuis alone significantly retarded tumor growth (Figs. 5A and 5C) and prolonged the survival time (Figs. 5B and 5D) compared with vehicle treatment. While treatment with cisplatin given in three doses exerted no antitumor effect, the combination treatment of S. choleraesuis plus cisplatin significantly retarded tumor growth and enhanced survival compared with S. choleraesuis treatment alone (Figs. 5A, 5B, 5C, and 5D). The mean tumor volume of ML-1 tumor-bearing mice treated with S. choleraesuis plus cisplatin was lowered by 55, 77, and 80% compared with those treated with S. choleraesuis, cisplatin, and PBS alone, respectively (Fig. 5A). Similarly, the mean tumor volume of LL/2 tumor-bearing mice treated with S. choleraesuis plus cisplatin was lowered by 46, 70, and 67% compared with those treated with S. choleraesuis, cisplatin, and PBS alone, respectively (Fig. 5C). Taken together, S. choleraesuis as a single-agent therapy could retard tumor growth and enhance survival in both the fast-growing LL/2 and the slow-growing ML-1 tumor models. More strikingly, additive antitumor effects could be achieved with the combination therapy of S. choleraesuis plus cisplatin. In accordance with the results shown in Fig. 5C, the tumor weight from LL/2 tumor-bearing mice treated with S. choleraesuis plus cisplatin was 62, 80, and 81% lower than that of those treated with S. choleraesuis, cisplatin, and PBS alone, respectively (Fig. 5E). Accordingly, the amount of S. choleraesuis (cfu/g tissue) in the tumors from mice treated with S. choleraesuis in combination with cisplatin was approximately 100-fold higher than that from mice treated with S. choleraesuis alone (Fig. 5F).

Figure 5.

Additive antitumor effects of S. choleraesuis (S.C.) in combination with cisplatin on subcutaneous ML-1 and LL/2 tumors. (A, B) Groups of eight BALB/c mice that had been inoculated s.c. with ML-1 cells (10⁶) at day 0 were treated i.p. with S. choleraesuis (2 10⁶ cfu) at day 18 followed by cisplatin (2 mg/kg) at days 25, 27, and 29 or with either treatment alone. (C, D, E, F) Groups of eight C57BL/6 mice that had been inoculated s.c. with LL/2 cells (10⁶) at day 0 were treated i.p. with S. choleraesuis (2 10⁶ cfu) at day 8 followed by cisplatin (2 mg/kg) at days 15, 17, and 19 or with either treatment alone. Vehicle control mice received PBS. Tumor volumes (means SEM, n = 8) among different treatment groups were compared in mice bearing (A) ML-1 (P < 0.001 for S.C. + cisplatin versus cisplatin or vehicle; P < 0.01 for S.C. + cisplatin versus S.C. and for S.C. versus vehicle) and (C) LL/2 tumors (P < 0.001 for S.C. + cisplatin versus cisplatin; P < 0.01 for S.C. + cisplatin versus S.C. or vehicle; P < 0.05 for S.C. versus vehicle). Kaplan–Meier survival curves of the mice bearing (B) ML-1 (P < 0.001 for S.C. + cisplatin versus cisplatin or vehicle; P < 0.05 for S.C. + cisplatin versus S.C. and for S.C. versus vehicle) and (D) LL/2 tumors (P < 0.01 for S.C. + cisplatin versus S.C., cisplatin, or vehicle; P < 0.05 for S.C. versus vehicle) with different treatments are shown. In another set of experiments in the LL-2 tumor model, (E) the tumor weight and (F) the amounts of S. choleraesuis accumulated in the tumors, livers, and spleens (means SD, n = 8) were measured at day 20. ^**P < 0.01; ^***P < 0.001.

Full figure and legend (132K)

Increased infiltrating neutrophils and CD8⁺ T Cells, as well as apoptotic cells, in the tumors following the combination treatment of S. choleraesuis and cisplatin

We analyzed the tumors from LL/2-bearing mice treated with S. choleraesuis or cisplatin alone, or in combination, for cell infiltrates by immunohistochemical staining and for apoptotic cells by the terminal deoxynucleotidyltransferase-mediated deoxyuridine triphosphate nick-end labeling (TUNEL) assay. Representative results for immunohistochemistry are shown in Fig. 6A. We observed notable increases of neutrophils and CD8⁺ T cells that infiltrated the tumors in the mice treated with S. choleraesuis and, in particular, in those treated with S. choleraesuis plus cisplatin. The numbers of infiltrating neutrophils and CD8⁺ T cells in the tumors treated with S. choleraesuis plus cisplatin were significantly increased compared with those in the tumors from the remaining three treatment groups, whereas no such difference was found between cisplatin- and PBS-treated groups (Figs. 6B and 6C). However, the levels of infiltrating CD4⁺ T cells, which were in a very small amount, were similar among the different treatment groups (data not shown). TUNEL assay showed an increase in the amount of cells undergoing apoptosis in the S. choleraesuis-treated tumors compared with cisplatin-treated or PBS-treated tumors (Figs. 7A and 7B). Notably, cisplatin alone failed to promote tumor apoptosis. Nevertheless, there was a twofold increase in the number of apoptotic cells induced by S. choleraesuis plus cisplatin compared with that induced by S. choleraesuis alone (Fig. 7B). Of interest is the finding that the mice treated with cisplatin alone, compared with those treated with PBS, revealed reduced intratumoral microvessel density (mean SEM 20.92 1.42/high-power field versus 26.92 1.51/high-power field, n = 4; P < 0.05). Taken together, these results indicate that the combination therapy with S. choleraesuis and low-dose cisplatin resulted in retarded tumor growth and an increase in infiltrating neutrophils and CD8⁺ T cells, as well as enhanced apoptosis in the tumors. Furthermore, the antiangiogenic effect exerted by low-dose cisplatin may have contributed, at least in part, to the enhanced antitumor effect of the combination regimen.

Figure 6.

Increases in neutrophil and CD8⁺ T cell infiltrates in the tumors from LL/2 tumor-bearing mice treated with S. choleraesuis (S.C.) in combination with cisplatin. Groups of four C57BL/6 mice that had been inoculated s.c. with LL/2 cells (10⁶) at day 0 were treated i.p. with S. choleraesuis (2 10⁶ cfu) at day 8 followed by cisplatin (2 mg/kg) at days 15, 17, and 19 or with either treatment alone. Vehicle control mice received PBS. (A) Tumors were excised at day 20 and immunostained with antibodies against Gr-1 or CD8⁺ (original magnification 400). (B) Neutrophils and (C) CD8⁺ T cells that infiltrated tumors were determined by averaging the cell numbers from three fields of the highest positive-stained cell density at 400 magnification in each section (means SEM, n = 4). ^*P < 0.05; ^**P < 0.01; ^***P < 0.001.

Full figure and legend (200K)

Figure 7.

Increase in tumor cells undergoing apoptosis in LL/2 tumor-bearing mice treated with S. choleraesuis (S.C.) in combination with cisplatin. Groups of four C57BL/6 mice that had been inoculated s.c. with LL/2 cells (10⁶) at day 0 were treated i.p. with S. choleraesuis (2 10⁶ cfu) at day 8 followed by cisplatin (2 mg/kg) at days 15, 17, and 19 or with either treatment alone. Vehicle control mice received PBS. (A) Tumors were excised at day 20, and TUNEL assay was used to detect apoptotic cells (original magnification 400). (B) TUNEL-positive cells were counted from three fields of the highest density of positive-stained cells in each section to determine the percentage of apoptotic cells (means SEM, n = 4). ^**P < 0.01; ^***P < 0.001.

Full figure and legend (115K)

Discussion

Bacteria have been exploited as an antitumor agent because they can be motile and grow in hypoxic areas of tumors¹. Not only attenuated bacterial strains, such as S. typhimurium^1,2,3, Vibrio cholerae, and Listeria monocytogenes, but also wild-type bacteria, such as Escherichia coli DH5 strain, exhibit tumor-targeting potential, indicating that no mutations affecting the survival of the bacteria are required for tumor-targeting potential¹². Previously, we exploited an attenuated S. choleraesuis as a delivery vehicle for DNA vaccine¹¹ and to carry eukaryotic expression vectors encoding antiangiogenic molecules as an anticancer agent^4,5. This rough variant of S. choleraesuis, designated vaccine 51, was obtained by spreading an 18-h broth culture of the virulent strain 188 of S. choleraesuis subsp. choleraesuis serovar Dublin over the surface of a dried nutrient agar plate and placing a drop of a suspension of Salmonella anti-O phage 1 and selecting for a phage-resistant colony after incubation at 37°C for 24 h¹³. In this study, we confirm and extend the findings regarding the usefulness of S. choleraesuis as a tumor-targeted anticancer agent. Our results demonstrate the capability of S. choleraesuis to target and multiply in primary tumors at both subcutaneous and orthotopic sites via systemic administration. The bacteria preferentially accumulated and amplified within implanted tumors in mice for at least 4 weeks, whereas in our previous studies S. choleraesuis carrying expression plasmids were retained within tumors for a shorter period of time^4,5. Because bacteria containing plasmids that confer antibiotic resistance have lower growth rates and tend to lose plasmids gradually in vivo, they appear to accumulate at lower levels and reside for a shorter period of time within tumors than bacteria without plasmids. The results shown in the present study, together with our previous findings⁴, demonstrate that S. choleraesuis predominantly resides in the hypoxic regions of tumors. Furthermore, by comparing the tumor-targeting ability of S. choleraesuis in immunocompetent and immunodeficient mice, we have shown previously that although the degrees of initial colonization and early replication of the bacteria within tumors are equivalent regardless of the immune status of the host, around 1 week, when adaptive immune responses are elicited, the bacteria are cleared more rapidly not only from the spleen and liver but also from the tumor of immunocompetent mice⁴. The bacteria may enter tumors through leaky vasculature, thereby escaping the immune surveillance of the host and finding sanctuary within tumor tissues¹². Therefore, Salmonella may be cleared more rapidly from livers and spleens, where abundant immune cells are present, than from tumors by the host immune surveillance, thereby persisting longer within tumors. In addition, hypoxic regions within tumors, where most necrotic regions are found, may provide more nutrients that selectively favor the growth of Salmonella.

In the work described here, we employed LL/2 lung tumor models of bilaterally implanted subcutaneous lung tumors and experimental metastatic nodules to investigate the tumor-targeting potential of S. choleraesuis via intratumoral or systemic administration. Our results show that intratumoral injection of S. choleraesuis not only replicated within the treated tumor but also spread to and accumulated in the untreated contralateral tumor at distant sites. Furthermore, S. choleraesuis accumulated and replicated not only in subcutaneous LL/2 tumors, but also in metastatic small nodules after systemic administration. More importantly, we also demonstrated preferential accumulation of S. choleraesuis in a more clinically relevant orthotopic hepatoma model in immunocompetent mice. Therefore, the ability of S. choleraesuis to target multiple tumors from distant sites makes it an ideal anticancer agent over some other cancer therapeutic agents limited to local administration. The propensity of S. choleraesuis to accumulate and amplify within primary tumors and metastatic nodules may have contributed to retarding tumor growth and metastasis, as well as prolonging the survival of tumor-bearing mice. Recently, we showed the therapeutic efficacy of S. choleraesuis carrying eukaryotic expression vectors encoding antiangiogenic molecules, namely endostatin or thrombospondin-1, which exhibits a dual effect involving tumoricidal and antiangiogenic activities on tumor growth, for the treatment of syngeneic murine bladder tumor and melanoma^4,5. In the present study, we demonstrate that the combination of S. choleraesuis with low-dose cisplatin induced an additive effect on retarding tumor growth and prolonging the survival time in the highly aggressive LL/2 lung tumor and slowly growing ML-1 hepatoma models.

In the present study, despite the failure of cisplatin to exhibit antitumor effects, its combination with S. choleraesuis enhanced the antitumor efficacy of S. choleraesuis and promoted accumulation of the bacteria within tumors in the murine lung tumor model. Notably, while cisplatin administered alone neither increased cell infiltrates nor induced apoptosis in the tumors, S. choleraesuis treatment alone significantly increased infiltrating neutrophils and CD8⁺ T cells as well as apoptotic cells within tumors, compared to vehicle treatment. Surprisingly, the combination treatment of S. choleraesuis and cisplatin resulted in a further twofold increase in cell infiltrates and apoptosis in the tumors. Tumors from the mice treated with cisplatin alone revealed a significant decrease in microvessel density compared with those from the control mice. However, this activity appeared insufficient to inhibit tumor growth or enhance survival of LL/2 tumor-bearing mice. In the context of tumor angiogenesis, cisplatin may have pleiotropic activities that can affect angiogenic processes directly and indirectly. In this regard, previous reports have documented that cisplatin in combination with some antiangiogenic agents decreases angiogenesis and increases apoptosis in tumors, resulting in enhanced antitumor effects in murine tumor models, compared with either treatment alone^14,15,16. To this end, some low-dose chemotherapeutic agents also exhibit antiangiogenic activities, which are mediated by a systemic increase in thromobospondin-1 levels^17,18. In addition to its antiangiogenic effect, cisplatin enhances expression of the cell death receptor, Fas, in various cancer cells, including LL/2 lung tumor cells. This effect may, in turn, augment cytotoxic T-cell-mediated antitumor immunity¹⁹. In this study, immunohistochemical studies reveal that combined therapy with S. choleraesuis plus cisplatin or S. choleraesuis monotherapy increased neutrophils and CD8⁺ T cells that infiltrated tumors. The persistence of high levels of Salmonella within tumors may have induced inflammatory responses, leading to the recruitment of immune cells, such as macrophages, neutrophils, and lymphocytes to the tumor site. Since bacterial replication in tumors and subsequent lysis of tumor cells may induce cell-mediated immune responses to tumor cells, higher oncolysis could account, in part, for an increased infiltrate of CD8⁺ T cells in S. choleraesuis-treated tumors. The cytotoxic T cell response against tumor cells may enhance the antitumor efficacy of S. choleraesuis. Antitumor effects of neutrophils, in particular, after being activated by substances derived from microorganisms, have also been demonstrated in various tumors^20,21,22. Moreover, lipopolysaccharide from Salmonella may induce apoptosis of tumor and endothelial cells²³. The potentiation of antiangiogenesis and upregulation of cytotoxic T cell responses by cisplatin may contribute to the enhanced antitumor effect of S. choleraesuis in conjunction with cisplatin.

In this study, we exploited a vaccine strain of S. choleraesuis as a tumor-targeted anticancer agent in murine syngeneic tumor models. However, the mutations responsible for attenuation in this vaccine strain have not been genetically defined¹³. VNP20009, an attenuated strain of S. typhimurium²⁴, and its derivative, TAPET-CD, which expresses an E. coli cytosine deaminase (CD)⁷, have been evaluated as anticancer agents in clinical trials for cancer patients^25,26. VNP20009 is modified from a wild-type strain of S. typhimurium by partial deletion of the msbB gene, which results in the generation of a lipid A mutant with diminished ability to induce TNF--mediated septic shock³. It is also a purine auxotroph and cannot metabolize xylose. Recently, the targeting and replication of VNP2009 expressing herpes simplex virus thymidine kinase in murine tumors in vivo have been noninvasively detected by positron emission tomography imaging²⁷. It has been shown that S. typhimurium has limited ability to adhere to tumor vasculature and migrate within tumors and survives only in tissue that becomes necrotic²⁸. As S. typhimurium accumulates in necrotic regions of tumors that are distant from tumor vasculature, systemic delivery of the prodrug 5-fluorocytosine to the necrotic regions of tumors may be suboptimal in tumor-bearing mice treated with S. typhimurium expressing CD, which may contribute to the failure of the treatment in eradicating tumors. Furthermore, S. typhimurium has been demonstrated to grow in the necrotic and relatively hypoxic foci within tumors, but not in well-oxygenated tumors at the rim of the growing nodules²⁹. The limited ability of S. typhimurium to disperse throughout the tumor may be the most important shortcoming in its use as an anticancer agent²⁹. Regarding the importance of the dispersal capability of bacteria for tumoricidal activity, Dang et al.³⁰ showed that Clostridium novyi, an obligate anaerobe, is able to disperse evenly and eradicate tumors in mice when combined with an antivascular agent. Therefore, a genetically modified Salmonella strain capable of dispersing more homogeneously throughout tumors would be a more desirable tumoricidal agent or transgene delivery vector. Our previous results suggest that production of antiangiogenic agents via S. choleraesuis-mediated gene transfer increases S. choleraesuis accumulation in tumors and, as a consequence, enhances tumoricidal effects. In this study, our data suggest that the capability of S. choleraesuis to disperse within tumors and hence to delay tumor growth was augmented when combined with low-dose cisplatin. An additional strategy to enhance antitumor efficacy would involve inhibiting the viable rim of tumor growth or modifying the tumor matrix that would facilitate the motility and penetration of Salmonella.

By taking advantage of the tumoricidal effect of Salmonella and pleiotropic activities of cisplatin, we conclude that S. choleraesuis in combination with cisplatin appears to hold promise for the treatment of solid tumors. However, further work is warranted to elucidate the underlying mechanism of antitumor effects of the combination therapy of S. choleraesuis and cisplatin.

Materials and methods

Bacteria, cell lines, and mice

A vaccine strain of S. choleraesuis [S. choleraesuis subsp. choleraesuis (Smith) Weldin serovar Dublin (ATCC 15480)] was obtained from the Bioresources Collection and Research Center (Hsinchu, Taiwan)¹³. The LL/2 mouse Lewis lung carcinoma cell line was obtained from the American Type Culture Collection. The ML-1 mouse hepatoma cell line was originally established from hepatocytes of BALB/c mice³¹. LL/2 and ML-1 cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 1% glutamine, and 50 g/ml gentamicin at 37°C in 5% CO₂. LL/2 cells cultured in six-well plates were transfected with 2 g of pTCYLuc⁴, and clonal derivatives were isolated by G418 (400 g/ml) selection and expanded to independent LL/2-Luc clones. Male C57BL/6 and BALB/c mice (6–8 weeks of age) were obtained from the Laboratory Animal Center of the National Cheng Kung University. The experimental protocol adhered to the rules of the Animal Protection Act of Taiwan and was approved by the Laboratory Animal Care and Use Committee of the National Cheng Kung University.

Assay of tumor-targeting potential of S. choleraesuis and detection of hypoxic areas within tumors

C57BL/6 and BALB/c mice were inoculated s.c. with 10⁶ LL/2 and ML-1 cells, respectively. When the tumors had grown to 100 to 450 mm³, which took around 8 days for LL/2 tumor and 18 days for ML-1 tumor, the mice were injected i.t. or i.p. with 2 10⁶ cfu of S. choleraesuis. At various time points p.i., five mice in each group were sacrificed, and the numbers of S. choleraesuis in the tumors, livers, and spleens were determined on LB agar plates and expressed as cfu per gram of tissues as previously described⁴. Meanwhile, at 28 days p.i., we examined the locations of S. choleraesuis and hypoxic regions within LL/2 tumors as described⁴. In the bilateral tumor model, LL/2 cells (10⁶) were inoculated s.c. into both flanks of C57BL/6 mice at day 0. Subsequently, S. choleraesuis (2 10⁶ cfu) or PBS was injected i.t. into the left flank or i.p. into the mice at day 8. The amounts of accumulated S. choleraesuis in the left and right tumors, livers, and spleens of four mice in each group were determined at day 15. In the orthotopic ML-1 tumor model, groups of four BALB/c mice were anesthetized with 90 mg/kg ketamine hydrochloride, their abdominal cavities opened, and liver lobes exposed, and ML-1 cells (5 10⁵) were injected into the left liver lobe at day 0 using a 10-l syringe (Terumo, Elkton, MD, USA). Notably, the needle hole was sealed with an electric coagulator immediately after withdrawal of the needle to avoid leakage of the injected fluid. The incision was subsequently sutured. In the murine model of experimental pulmonary metastasis, groups of eight C57BL/6 mice were injected with LL/2 cells (10⁵) via the tail vein at day 0. Subsequently, S. choleraesuis (2 10⁶ cfu) was injected i.p. into the ML-1 tumor-bearing mice at day 18 or into the LL/2 tumor-bearing mice at day 15, and S. choleraesuis that accumulated in the tumors or lungs, as well as livers and spleens, were quantified at day 38 (for ML-1 tumor) or day 20 (for LL/2 tumor). A parallel experiment was performed using normal healthy mice inoculated with S. choleraesuis, and the bacteria in the respective organs were also assessed.

Treatment of primary lung tumor and experimental pulmonary metastasis with S. choleraesuis

S. choleraesuis (2 10⁶ cfu) or PBS was injected i.t. into the left flank tumor or i.p. into groups of four C57BL/6 mice bearing bilateral LL/2 tumors. Palpable tumors in both sides were measured as described previously⁵. To enhance the ability to quantify tumor burden in the lungs, pulmonary metastasis was induced by inoculation of LL/2-Luc cells (10⁵) via the tail vein into C57BL/6 mice to evaluate the antimetastatic effect of S. choleraesuis. The bacteria (2 10⁶ cfu) or PBS was then inoculated i.p. into groups of eight mice on day 15, and the wet lung weight and luciferase activity of the lungs were measured at day 20⁵. In another set of experiments, LL/2 cells rather than LL/2-Luc cells were used to induce pulmonary metastasis and the same treatment regimen was applied. Survival of the mice in the treated and control groups was monitored daily.

Treatment of subcutaneous lung tumor and hepatoma with S. choleraesuis in combination with low-dose cisplatin

We used subcutaneous LL/2 and ML-1 tumor models to evaluate the antitumor efficacy of S. choleraesuis (2 10⁶ cfu) in combination with low-dose cisplatin. C57BL/6 and BALB/c mice were inoculated s.c. with 10⁶ LL/2 or ML-1 cells at day 0, and visible nodules developed at all injection sites with approximate tumor volumes of 100 and 450 mm³ at days 8 and 18, respectively. Groups of eight LL/2 tumor-bearing mice were injected i.p. with S. choleraesuis on day 8 followed by cisplatin (2 mg/kg) treatment on days 15, 17, and 19 or with either treatment alone. Similarly, groups of eight ML-1 tumor-bearing mice were treated with S. choleraesuis on day 18 followed by cisplatin treatment on days 25, 27, and 29 or with either treatment alone. All of the mice were monitored for tumor growth and survival as previously described⁵.

Immunohistochemistry and apoptotic assays

To analyze cell infiltrates in the tumors, groups of four C57BL/6 mice that had been inoculated s.c. with 10⁶ LL/2 cells at day 0 were injected i.p. with either 2 10⁶ cfu of S. choleraesuis at day 8, or low-dose cisplatin (2 mg/kg) at days 15, 17, and 19, or both. Control mice received PBS. The tumors were excised and snap frozen at day 20. Cryostat sections (4 m) were prepared, fixed, and incubated with rat anti-mouse Ly-6G (Gr-1) (RB6-8C5; PharMingen, San Diego, CA, USA), rat anti-mouse CD4 (L3T4) (GK1.5; PharMingen), or rat anti-mouse CD8a (Ly-2) (53-6.7; PharMingen) antibody. After sequential incubation with appropriate peroxidase-labeled secondary antibody and aminoethyl carbazole as substrate chromagen, the slides were counterstained with hematoxylin. The infiltrating cells were quantified by averaging the number of each cell type in three areas of highest cell density at 400 magnification in each section. Tumor vascularization was determined by counting factor VIII-positive vessels with immunohistochemical analysis, and microvessel density was determined by averaging the number of vessels in three areas of highest vessel density at 400 magnification in each section⁵. TUNEL assay was used to detect cell apoptosis within tumors and was performed according to the manufacturer's instructions (Promega, Madison, WI, USA). TUNEL-positive cells (brown staining) were counted under the microscope. We counted three high-power (400) fields with approximately 200–250 cells that showed the highest density of positive-stained cells per field to determine the average percentage of apoptotic (TUNEL-positive) cells in each section.

Statistical analysis

The unpaired, two-tailed Student t test was used to determine differences between groups for the comparisons of tumor volume, tumor weight, lung weight, luciferase activity, and numbers of S. choleraesuis, neutrophils, CD8⁺ T cells, factor VIII-positive cells, and apoptotic cells. The survival analysis was performed using the Kaplan–Meier survival curve and log-rank test. Any P value less than 0.05 was regarded as statistically significant.

References

1. Pawelek, J. M., Low, K. B. and Bermudes, D. (1997). Tumor-targeted Salmonella as a novel anticancer vector. Cancer Res. 57: 4537–4544. | PubMed | ISI | ChemPort |
2. Pawelek, J. M., Low, K. B. and Bermudes, D. (2003). Bacteria as tumour-targeting vectors. Lancet Oncol. 4: 548–556. | Article | PubMed | ISI |
3. Low, K. B. et al. (1999). Lipid A mutant Salmonella with suppressed virulence and TNF- induction retain tumor-targeting in vivo. Nat. Biotechnol. 17: 37–41. | Article | PubMed | ChemPort |
4. Lee, C. H., Wu, C. L. and Shiau, A. L. (2004). Endostatin gene therapy delivered by Salmonella choleraesuis in murine tumor models. J. Gene Med. 6: 1382–1393. | Article | PubMed | ChemPort |
5. Lee, C. H., Wu, C. L. and Shiau, A. L. (2005). Systemic administration of attenuated Salmonella choleraesuis carrying thrombospondin-1 gene leads to tumor-specific transgene expression, delayed tumor growth and prolonged survival in the murine melanoma model. Cancer Gene Ther. 12: 175–184. | Article | PubMed | ISI | ChemPort |
6. Li, X. et al. (2003). Bifidobacterium adolescentis as a delivery system of endostatin for cancer gene therapy: selective inhibitor of angiogenesis and hypoxic tumor growth. Cancer Gene Ther. 10: 105–111. | Article | PubMed | ISI | ChemPort |
7. King, I. et al. (2002). Tumor-targeted Salmonella expressing cytosine deaminase as an anticancer agent. Hum. Gene Ther. 13: 1225–1233. | Article | PubMed | ISI | ChemPort |
8. Yamada, H., Matsumoto, S., Matsumoto, T., Yamada, T. and Yamashita, U. (2000). Murine IL-2 secreting recombinant Bacillus Calmette–Guerin augments macrophage-mediated cytotoxicity against murine bladder cancer MBT-2. J. Urol. 164: 526–531. | Article | PubMed | ISI | ChemPort |
9. Graeber, T. G. et al. (1996). Hypoxia-mediated selection of cells with diminished apoptotic potential in solid tumours. Nature 379: 88–91. | Article | PubMed | ISI | ChemPort |
10. Brown, J. M. and Giaccia, A. J. (1998). The unique physiology of solid tumors: opportunities (and problems) for cancer therapy. Cancer Res. 58: 1408–1416. | PubMed | ISI | ChemPort |
11. Shiau, A. L., Chen, Y. L., Liao, C. Y., Huang, Y. S. and Wu, C. L. (2001). Prothymosin enhances protective immune responses induced by oral DNA vaccination against pseudorabies delivered by Salmonella choleraesuis. Vaccine 19: 3947–3956. | Article | PubMed | ChemPort |
12. Yu, Y. A. et al. (2004). Visualization of tumors and metastases in live animals with bacteria and vaccinia virus encoding light-emitting proteins. Nat. Biotechnol. 22: 313–320. | Article | PubMed | ISI | ChemPort |
13. Smith, H. W. (1965). The immunization of mice, calves and pigs against Salmonella dublin and Salmonella cholera-esuis infections. J. Hyg. (London) 63: 117–135. | ChemPort |
14. Tan, G. H. et al. (2004). Combination of low-dose cisplatin and recombinant xenogeneic edoglin as a vaccine induces synergistic antitumor activities. Int. J. Cancer 112: 701–706. | Article | PubMed | ISI | ChemPort |
15. Mishima, K. et al. (2000). A peptide derived from the non-receptor-binding region of urokinase plasminogen activator inhibits glioblastoma growth and angiogenesis in vivo in combination with cisplatin. Proc. Natl. Acad. Sci. USA 97: 8484–8489. | Article | PubMed | ChemPort |
16. Schiller, J. H. and Bittner, G. (1999). Potentiation of platinum antitumor effects in human lung tumor xenografts by the angiogenesis inhibitor squalamine: effects on tumor neovascularization. Clin. Cancer Res. 5: 4287–4294. | PubMed | ChemPort |
17. Bocci, G., Francia, G., Man, S., Lawler, J. and Kerbel, R. S. (2003). Thrombospondin 1, a mediator of the antiangiogenic effects of low-dose metronomic chemotherapy. Proc. Natl. Acad. Sci. USA 100: 12917–12922. | Article | PubMed | ChemPort |
18. Klement, G. et al. (2000). Continuous low-dose therapy with vinblastine and VEGF receptor-2 antibody induces sustained tumor regression without overt toxicity. J. Clin. Invest. 105: R15–R24. | PubMed | ISI | ChemPort |
19. Merritt, R. E., Mahtabifard, A., Yamada, R. E., Crystal, R. G. and Korst, R. J. (2003). Cisplatin augments cytotoxic T-lymphocyte-mediated antitumor immunity in poorly immunogenic murine lung cancer. J. Thoracic Cardiovasc. Surg. 126: 1609–1617. | ChemPort |
20. Kasai, S., Fujimoto, S., Nitta, K., Baba, H. and Kunimoto, T. (1991). Antitumor activity of polymorphonuclear leukocytes activated by a -1,3-D-glucan. J. Pharmacobiodyn. 14: 519–525. | PubMed | ChemPort |
21. Fujii, Y., Kimura, S., Arai, S. and Sendo, F. (1987). In vivo antitumor effect of lymphokine-activated rodent polymorphonuclear leukocytes. Cancer Res. 47: 6000–6005. | PubMed | ChemPort |
22. Yang, K. D., Stone, R. M., Lee, C. S., Chao, T. Y., Cheng, S. N. and Shaio, M. F. (1992). Effect of picibanil (OK432) on neutrophil-mediated antitumor activity: implication of monocyte-derived neutrophil-activating factors. Cancer Immunol. Immunother. 35: 277–282. | Article | PubMed | ChemPort |
23. Bannerman, D. D., Tupper, J. C., Ricketts, W. A., Bennett, C. F., Winn, R. K. and Harlan, J. M. (2001). A constitutive cytoprotective pathway protects endothelial cells from lipopolysaccharide-induced apoptosis. J. Biol. Chem. 276: 14924–14932. | Article | PubMed | ISI | ChemPort |
24. Clairmont, C. et al. (2000). Biodistribution and genetic stability of the novel antitumor agent VNP20009, a genetically modified strain of Salmonella typhimurium. J. Infect. Dis. 181: 1996–2002. | Article | PubMed | ChemPort |
25. Toso, J. F. et al. (2002). Phase I study of the intravenous administration of attenuated Salmonella typhimurium to patients with metastatic melanoma. J. Clin. Oncol. 20: 142–152. | Article | PubMed | ISI |
26. Nemunaitis, J. et al. (2003). Pilot trial of genetically modified, attenuated Salmonella expressing the E. coli cytosine deaminase gene in refractory cancer patients. Cancer Gene Ther. 10: 737–744. | Article | PubMed | ISI | ChemPort |
27. Soghomonyan, S. A. et al. (2005). Positron emission tomography (PET) imaging of tumor-localized Salmonella expressing HSV1-TK. Cancer Gene Ther. 12: 101–108. | Article | PubMed | ChemPort |
28. Forbes, N. S., Munn, L. L., Fukumura, D. and Jain, R. K. (2003). Sparse initial entrapment of systemically injected Salmonella typhimurium leads to heterogeneous accumulation within tumors. Cancer Res. 63: 5188–5193. | PubMed | ChemPort |
29. Rosenberg, S. A., Spiess, P. J. and Kleiner, D. E. (2002). Antitumor effects in mice of the intravenous injection of attenuated Salmonella typhimurium. J. Immunother. 25: 218–225. | Article | PubMed | ChemPort |
30. Dang, L. H., Bettegowda, C., Huso, D. L., Kinzler, K. W. and Vogelstein, B. (2001). Combination bacteriolytic therapy for the treatment of experimental tumors. Proc. Natl. Acad. Sci. USA 98: 15155–15160. | Article | PubMed | ChemPort |
31. Hsieh, J. L., Wu, C. L., Lee, C. H. and Shiau, A. L. (2003). Hepatitis B virus X protein sensitizes hepatocellular carcinoma cells to cytolysis induced by E1B-deleted adenovirus through the disruption of p53 function. Clin. Cancer Res. 9: 338–345. | PubMed | ISI | ChemPort |

Acknowledgements

This work was supported by grants (NSC 92-3112-B-006-007 and NSC 93-3112-B-006-008) from the National Science Council, Taiwan.