Female Balb/c mice aged 6–8 weeks old with an average weight of 20.52±0.83 g were obtained from the Laboratory Animal Center of Sun Yat-sen University (Guangzhou, China). The mice were housed in a special pathogen-free facility with free access to drinking water and a pellet-based diet, and were quarantined for 7 days prior to surgery. The mice were kept in a room with controlled temperature (22±1°C) and humidity (50-70%), and a 12-h light/dark cycle. All animal experiments were approved by the Ethical Committee of Sun Yat-Sen University and were conducted in accordance with guidelines of the Institutional Animal Care and Use Committee of Sun Yat-Sen University and the Committee for Animal Experiments of Sun Yat-Sen University.

On day 1, 20 mice were randomly assigned into experimental and control groups, with 10 mice in each group. Mice were anesthetized using an intraperitoneal injection of 4% chloral hydrate (400 mg/kg). To verify anesthesia, it was ensured that the mice were unresponsive when pinching the skin or passively opening the eyes. After local disinfection of the skin, an abdominal incision was made along the left subcostal margin, and 1×10⁵ CT-26 cells in 50 µl phosphate-buffered saline (PBS) were injected into the spleen of the mice in the experimental group using a 32-G needle. The control mice received the same volume of sterile PBS. The abdominal wall was closed using a two-layer technique and non-absorbable sutures. The mice were monitored, with body weight and general health indicators being recorded every 3 days over a period of 3 weeks. No mouse exhibited signs of peritonitis in the present study following the use of chloral hydrate. If a mouse could not ambulate or was unable to intake food or water independently during the experiment, it would be euthanized ahead of schedule; however, no mouse was euthanized or found dead ahead of schedule in the present study. The mice were sacrificed on day 22 using an intraperitoneal injection of 4% chloral hydrate (400 mg/kg), followed by cervical dislocation. To verify death, breath and heartbeat were observed. Briefly, the mice exhibited no abdominal or thoracic respiration and no observable throbbing in the precordium. After verification of death, the liver and spleen were removed, and the number of macroscopic metastatic nodules in the liver was counted. Liver tissues were fixed for 24 h at room temperature using 10% neutral buffered formalin for subsequent paraffin embedding, or stored in RNA stabilization solution at −80°C (RNA later; Ambion; Thermo Fisher Scientific, Inc.) for further analysis.

The undifferentiated murine colon adenocarcinoma cell line, CT-26, was obtained from the Chinese Academy of Science Cell Bank (Shanghai, China) and cultured in RPMI-1640 medium with 10% fetal calf serum and antibiotics (1% penicillin and streptomycin) (all Gibco; Thermo Fisher Scientific, Inc.) at 37°C in a humidified incubator (Thermo Fisher Scientific, Inc.) with 5% CO₂. The cultures were routinely tested once a week for mycoplasma contamination using MycoProbe (R&D Systems, Inc.), and the medium was regularly replaced every 2–3 days. The cells were harvested using 0.25% trypsin (Gibco; Thermo Fisher Scientific, Inc.) for 2 min at room temperature.

The spleen and liver were homogenized by trituration (300 × g at 4°C for 10 min) and suspended in PBS, and single-cell suspensions were prepared. After RBC lysis (cat. no. 555899; BD Biosciences) with 0.17 M ammonium chloride at 4°C for 5 min, the cells were washed in PBS and counted using the cellometer Auto T4 (BD Biosciences). Aliquots of 1×10⁶ cells were stained using anti CD4-FITC antibody (1:10; cat. no. 561833; BD Biosciences) and/or anti Foxp3-PE-Cy5 antibody (1:10; cat. no. 563101; BD Biosciences) at 4°C for 30 min, according to the manufacturer's instructions. The stained cells were added into a FACS canto flow cytometer (BD Biosciences), and the T-cell subsets were analyzed in samples of 1×10⁴ cells each using Cell Quest software (v5.1; BD Pharmingen). Samples from all mice were analyzed using flow cytometry, which was repeated three times.

Tissues were fixed with 10% formalin at room temperature for 24 h and embedded in 10% paraffin for 24 h, cut into 4-µm thick sections and placed on salinized glass slides. Hematoxylin and eosin staining was performed as per the standard protocol. For immunohistochemistry, the sections were deparaffinized with dimethylbenzene and rehydrated through an alcohol gradient, followed by antigen retrieval in a hot sodium citrate buffer at 96–98°C for 15 min. After quenching the endogenous peroxidase activity with 0.3% hydrogen peroxide solution and blocking using 5% goat serum (cat. no. 16210064; Gibco; Thermo Fisher Scientific, Inc.) at room temperature for 30 min, the sections were incubated overnight with anti-α-SMA (1:1,000; cat. no. ab124964), anti-HGF (1:1,000; cat. no; ab83760) and anti-c-Met (1:1,000; cat. no. ab51067) primary antibodies (all Abcam) at 4°C. The sections were then incubated with a Qdot 655-conjugated goat anti-chicken IgY (H+L) secondary antibody (1:50; cat. no. Q14421MP; Invitrogen; Thermo Fisher Scientific, Inc.) at room temperature for 60 min, washed and stained with diaminobenzidine in an Envision System (Dako), followed by counterstaining with hematoxylin at room temperature for 2 min. Slides were observed under an inverted phase contrast light microscope (magnification, ×200; Leica Microsystems GmbH).

Sample size in each group was determined according to the study by Syed Khaja et al (17) and the power test using G power (t-test, difference between two independent means). Briefly, the mean and standard deviation in each group was used to calculate the effect size d, and the power was >90% for all tests using G power software. All data are expressed as mean ± standard deviation (n=3). The weight of the mice was compared using a repeated-measures analysis of variance and the expression of T cell markers was compared using an unpaired Student's t-test in SPSS v21 software (IBM Corp.). Two-sided P-values of <0.05 were considered to indicate a statistically significant difference.

All tumor-bearing and control mice survived until the end of the experiment. However, unlike the control mice, the tumor-bearing mice exhibited signs of illness, such as reduced activity, slow response, lackluster fur, loss of appetite, emaciation and a distended abdomen, which were accompanied by significant loss in body weight (P=0.013; Fig. 1A). The maximum percentage of body weight loss was 8.86%.

At 3 weeks post-tumor cell injection, the mice were sacrificed. The liver and spleen of the control mice was soft and smooth without any neoplasms (Fig. 1B). By contrast, both organs in the CT-26-injected mice had a rough and uneven surface, while macroscopic tumors were also observed (Fig. 1C). An average of 20.5 metastatic nodules was observed on the liver of the CT-26-injected mice, and pathological evaluation using hematoxylin and eosin staining confirmed the nodules as liver metastases of colon carcinoma (Fig. 1D).

Single-cell suspensions of the spleen and liver were prepared and analyzed using flow cytometry. As indicated in Fig. 2, the numbers of CD4⁺ T cells in the spleen and liver of the tumor-bearing mice were significantly lower compared with those in the control group (P<0.001 and P=0.003, respectively). By contrast, the proportion of CD4⁺FOXP3⁺ Tregs among the entire CD4⁺ T cell population was significantly higher in the tumor-bearing group compared with that in the control group (P<0.001 and P=0.026 in the spleen and liver, respectively; Fig. 3).

The liver tissues of the control mice were negative for α-SMA and expressed only low levels of HGF and c-Met. By contrast, α-SMA, HGF and c-Met levels were significantly upregulated in the liver metastatic nodules (Fig. 4).