Therapeutic efficacy of tumor-targeting Salmonella typhimurium A1-R on human colorectal cancer liver metastasis in orthotopic nude-mouse models.

Liver metastasis is the most frequent cause of death from colon and other cancers. Generally, liver metastasis is recalcitrant to treatment. The aim of this study is to determine the efficacy of tumor-targeting Salmonella typhimurium A1-R on liver metastasis in orthotopic mouse models. HT-29 human colon cancer cells expressing red fluorescent protein (RFP) were used in the present study. S. typhimurium A1-R infected HT-29 cells in a time-dependent manner, inhibiting cancer-cell proliferation in vitro. S. typhimurium A1-R promoted tumor necrosis and inhibited tumor growth in a subcutaneous tumor mouse model of HT-29-RFP. In orthotopic mouse models, S. typhimurium A1-R targeted liver metastases and significantly reduced their growth. The results of this study demonstrate the future clinical potential of S. typhimurium A1-R targeting of liver metastasis.


INTRODUCTION
Anecdotal records go back hundreds of years describing patients having their cancer go into remission after a bacterial infection [1]. In 1867, the German physician Busch reported that a cancer patient went into remission after contracting erysipelas, now known as Streptococcus pyogenes [2]. Bruns treated a cancer patient in 1888 with S. pyogenes and the tumor regressed [1]. Koch, Pasteur and von Behring recorded that cancer patients infected with S. pyogenes had tumor regression [1].
In the 1890s, William B. Coley of New York Cancer Hospital, which became Sloan-Kettering Memorial Cancer Center, treated cancer patients with S. pyogenes. Coley had excellent results infecting cancer patients with S. pyogenes. Hoption Cann et al. [3] compared Coley's bacterial treatment to current chemotherapy and found the 10-year survival rates of Coley's patients were comparable [4] to current conventional therapies [4]. Malmgren and Flanigan [5] demonstrated that anaerobic bacteria could survive and replicate in necrotic tumor tissue which had low oxygen content. Several other early approaches aimed at utilizing bacteria for cancer therapy in animal models were described [6][7][8][9][10][11][12][13][14][15][16].
The obligate anaerobes Bifidobacterium [17] and Clostridium [18], which replicate only in necrotic areas of tumors, have been tested for cancer therapy in mouse models. Bifidobacterium longum selectively localized in mammary tumors after systemic administration [17]. Spores of Clostridium novyi, without its lethal toxin (C. novyi no toxin [NT]). germinated within necrotic areas of tumors in mice and, in combination with chemotherapy, resulted in hemorrhagic necrosis and tumors regression [18]. Recently, C. novyi-NT was used in a patient with leiomyosarcoma and caused one metastatic lesion to regress [19].
Salmonella typhimurium (S. typhimurium) is a facultative anaerobe which, in contrast to obligate anaerobes, can grow in the viable regions as well as necrotic regions of tumors [20]. S. typhimurium with a lipid A-mutation (msbB) deletion along with purine auxotrophic mutations (purI) had antitumor efficacy in mice [21]. S. typhimurium (VNP20009), with msbB and purI mutations, was relatively safely administered to patients in a Phase I clinical trial on patients with metastatic melanoma and renal carcinoma. Overattenuation perhaps limited efficacy [22].
Liver metastases is the most frequent cause of death of patients with colorectal cancer [23]. Hepatectomy is the most effective treatment for liver metastasis from colorectal cancer, but the recurrence rate is over 50% after resection [24]. In addition, efficacy of chemotherapeutic agents are marginal [25]. Therefore, development of effective treatment for liver metastasis is urgently needed.
The present report demonstrates efficacy of S. typhimurium A1-R on liver metastasis of colon cancer in orthotopic mouse models.

S. typhimurium A1-R targeted HT-29 colon cancer cells in vitro
S. typhimurium A1-R-GFP infection of HT-29 cells expressing red fluorescent protein (HT-29-RFP) was observed beginning 1 hour after addition of bacteria to the cultures ( Figure 1A-1B). At 15 hours after addition, many S. typhimurium A1-R-GFP cells were observed inside the HT-29 cancer cells ( Figure 1C).

S. typhimurium A1-R inhibited HT-29 cancer cell proliferation in vitro
HT-29-RFP cancer cell colonies were significantly reduced by S. typhimurium A1-R-GFP both in area ( Figure 1D) and number ( Figure 1E) (P < 0.01 for both).

Efficacy of S. typhimurium A1-R on subcutaneous HT-29 tumor growth
Subcutaneous tumor growth was significantly inhibited after the 3 rd i.v. administration of S. typhimurium A1-R. The tumor volume ratio at day 22 compared to day 1 in the control group was 6.17 ± 1.16 and 2.68 ± 0.73 in the S. typhimurium A1-R treatment group, P < 0.05 (Figure 2A-2D). Resected specimens showed S. typhimurium A1-R induced more extensive necrosis in the tumors compared to those without treatment ( Figure 2E-2H).

Efficacy of S. typhimurium A1-R on orthotopic liver metastasis mouse models
S. typhimurium A1-R-GFP targeted the liver metastasis 3 days after the 2 nd i.v. administration of S. typhimurium A1-R ( Figure 3). S. typhimurium A1-R treatment significantly suppressed metastatic progression compared to the control group at day 22. The metastasis fluorescent area ratio on day 22 compared to day 1 was 5.69 ± 0.83 in the S. typhimurium A1-R treatment group and 12.96 ± 1.49, in the untreated control group (P < 0.01; Figure 4).
The present study demonstrated that tumor-targeting S. typhimurium A1-R has significant efficacy on liver metastasis in orthotopic mouse models, suggesting clinical activity for patients with colorectal liver metastasis. The present study suggests S. typhimurium A1-R may also be useful in the neo-adjuvant setting to reduce the liver metastasis that would make previously inoperable liver metastasis resectable.
Spleen injection selects colon cancer cells which can grow in the liver. However, this leads to multiple metastatic sites, which were harvested to directly implant a single metastases on the liver of additional mice, which was the desired model. Future experiments will treat models with multiple liver metastases with S. typhimurium A1-R.
This is the first study to use S. typhimurium A1-R to treat colon cancer liver metastasis, which occurs at high frequency and is usually the cause of lethality of this disease. The significant efficacy of S. typhimurium A1-R on the liver metastasis was comparable to the efficacy of S. typhimurium A1-R on other tumor types .
We have previously demonstrated efficacy of S. typhimurium A1-R in combination with anti-angiogenesis therapy in pancreatic cancer mouse models [37]. Future experiments will also test this combination in colon-cancer liver-metastasis models.
S. typhimurium A1-R in combination with chemotherapy was also active against stomach and pancreatic cancer models [35,40]. Such strategies will also be tested on colon-cancer liver-metastasis in future experiments.
The present study used GFP and RFP to image S. typhimurium A1-R and the cancer cells, respectively, in vitro and in vivo. Genetic reporters have an important advantage over injectable probes in that the label is permanent with genetic reporters. This is important for long-term monitoring of bacterial targeting as well as tumor growth, including metastasis as well as recurrence [48]. www.impactjournals.com/oncotarget Bacteria have numerous advantages over other biological therapy in that bacteria actively invade tumors, even when vascularity is poor and have a large genome to manipulate for effective and selective tumor targeting [2].
Tissue-selective therapy focuses on unique properties of normal tissues and how therapy can target a property of a tissue that kills the cancer tissues that arise from the normal tissue without affecting other tissues [49,54]. S. typhimurium A1-R is an example of tissueselective therapy.
De-differentiation of a tumor leading to resistance to chemotherapy is a limitation of tissue-selective therapy, since the targeted protein or pathway may no longer be expressed in the de-differentiated tumor [54]. However, S. typhimurium A1-R does not seem to depend on such targets and should be active against de-differentiated tumors. Specific caspase inhibitors may protect normal cells which may not be present in drug-resistant cancer cells and those could provide further protection to normal cells during S. typhimurium A1-R therapy [49,51]. S. typhimurium A1-R may also be effectively combined with teratogens which could selectively effect cancer cells that are de-differentiated [50]. Since S. typhimurium A1-R can decoy quiescent cancer cells to begin to cycle, S. typhimurium A1-R could be effectively combined with agents which selectively target proliferating cancer cells [52]. In this regard, normal cells could be protected by agents which induce wild type p53 [53].

Establishment of RFP-labeled HT29
The pDsRed-2 vector (Clontech Laboratories Inc., Palo Alto, CA) expressing RFP and neomycin resistance gene were used to transfect HT-29 to stably express RFP. For RFP gene transfection, 25% confluent HT-29 cells were incubated with a mixture of retroviral supernatants of PT67-RFP packaging cells and DMEM for 24 h. Fresh medium was replenished at this time, and cells were allowed to grow in the absence of retrovirus for 12 h. This procedure was repeated until high levels of RFP expression were achieved. Cells were then harvested with trypsin-EDTA and subcultured into selective medium that contained 200 μg/ml G418 (Invitrogen Corp., Carlsbad, CA). The level of G418 was increased to 2, 000 μg/ml stepwise. HT-29 clones expressing high levels of RFP were isolated and were amplified and transferred using conventional culture methods. High RFP-expression clones were subsequently isolated in the absence of G418 for 10 passages to select for stable expression of RFP [55,57,58].

Preparation of S. typhimurium A1-R
GFP-expressing S. typhimurium A1-R bacteria (AntiCancer Inc.,) were grown overnight on LB medium (Fisher Sci., Hanover Park, IL, USA) and then diluted 1:10 in LB medium. Bacteria were harvested at late-log phase, washed with PBS, and then diluted in PBS [28].

Clonogenic assay
HT-29-RFP cells (2.0 × 10 2 ) were seeded in 35 mm dishes. S. typhimurium A1-R-GFP (5 × 10 5 CFU/ml) was added to the cancer cells, which were incubated at 37 o C for 40 minutes. The cells were rinsed and cultured in medium containing gentamycin (10 μg/ml). After 9 days culture, the colonies were fixed with methanol and stained with crystal violet. The Image J program v1.49f (National Institutes of Health) was used to quantify the colonies.

Confocal imaging of cancer cells infected by S. typhimurium A1-R
HT-29-RFP cells (4.0 × 10 4 ) were seeded in 35 mm dishes. S. typhimurium A1-R-GFP (5 × 10 5 CFU/ml) was added to the cancer cells, which were incubated at 37 o C for 40 minutes. The cells were rinsed and cultured in medium containing gentamycin (10 μg/ml). The interaction between S. typhimurium A1-R expressing GFP and HT-29 cancer cells expressing RFP was imaged in real time by confocal microscopy (Fluoview FV1000, Olympus Corp., Tokyo, japan) [59] before, and 1 hour and 15 hours after administration of S. typhimurium A1-R.

Efficacy of S. typhimurium A1-R on HT-29 subcutaneous tumors
HT-29-RFP cells were harvested by trypsynization and washed with serum-free medium. A cell suspension (2 × 10 6 cells/100 μl in medium with 50% Matrigel) was injected subcutaneously in the right flank of nude mice. Two weeks after cell injection, established subcutaneous tumors were measured weekly and treated. Fourteen mice were randomized into a control group (n = 7) and an S. typhimurium A1-R treatment group (n = 7). The first treatment day was defined as day 1. Mice in the S. typhimurium A1-R group were treated with S. typhimurium A1-R (5 × 10 7 CFU/body) 3 times at days 1, 8, and 15. All tumors in both groups were observed weekly with the OV100 Small Animal Imaging System (Olympus) [60] and harvested at day 22 for tissue evaluation. Each tumor was measured individually in each mouse. Tumor volume was calculated with the following formula: Tumor volume = 1/2 × Length × Width 2 . Treatment efficacy was presented as a ratio of the tumor volume at each time point compared to the tumor volume at the beginning of the treatment. Dosing was determined from efficacy in other tumor models of S. typhimurium A1-R .

Initial establishment of liver metastases
HT-29-RFP cells were harvested by trypsinization and washed twice with serum-free medium. HT-29-RFP cells (5 × 10 5 in 50 μl serum-free medium with 50% Matrigel) were injected into the superior and inferior pole of the spleen in nude mice. Three weeks after injection, liver metastases were established ( Figure 5A).

Surgical orthotopic implantation of liver metastasis
Liver metastases, as described above, were resected and cut into block (8 mm 3 ) ( Figure 5B). A single tumor fragment was orthotopically implanted into the left lobe of the liver in other nude mice ( Figure 5C). Four weeks later, liver metastasis were observed at the implanted site ( Figure 5D).

Efficacy of S. typhimurium A1-R on liver metastasis
Four weeks after orthotopic implantation of HT-29-RFP to the liver, 10 mice were randomized into 2 groups: untreated control group (weekly, 3 weeks, n = 5) and the S. typhimurium A1-R treatment group (S. typhimurium A1-R, 5 × 10 7 CFU/body, iv, weekly, 3 weeks, n = 5). The left lobe of the liver with metastasis was exposed before (at day 1) and after treatment (at day 22) for observation with the OV100. The tumor fluorescence area was analyzed with ImageJ software. Treatment efficacy in each mouse was compared as a ratio of the tumor volume at each time point compared to the tumor volume at the beginning of treatment. Liver metastasis in the S. typhimurium A1-R treatment group was imaged with the FV1000 confocal microscope at day 11 to observe S. typhimurium A1-R-GFP targeting the RFP-expressing HT-29 liver metastasis.

Histology of tissue
Fresh tumor samples were fixed in formalin (10%) and embedded in paraffin before sectioning and staining. Tissue sections (3 mm) were deparaffinized in xylene and rehydrated in an ethanol series. Hematoxylin and eosin (H & E) staining was performed according to standard protocols.