Twenty-four male db/db mice (5 weeks old) were obtained from Japan SLC Inc. They were acclimatized for a week before initiation of the experiment (16). They were carefully handled and humanely maintained at the Gifu University Life Science Research Center in accordance with the Institutional Animal Care Guidelines. The mice were housed in cages and were provided basic diet CE-2 obtained from Oriental Yeast and sterilized tap water ad libitum. The environmental conditions inside the room were as follows: Temperature, 23±1°C; humidity, 50±20%; and 12-h light/dark cycle.

AOM was obtained from Wako Pure Chemical Co. and was intraperitoneally injected once weekly from weeks 1 to 4 after initiation of the experiment. AOM was diluted with ultra-pure water to administer a dose of 15 mg/kg body weight. Tofogliflozin was kindly provided by Kowa Co., Ltd. Tofogliflozin (1 mg/kg body weight and 10 mg/kg body weight) was dissolved in ultra-pure water. The solution was poured into the bottles and exchanged twice a week.

At five weeks of age, the mice were randomly divided into four experimental groups and treated as follows: No treatment (Group 1, n=5), AOM alone (Group 2, n=9), AOM followed by tofogliflozin (1 mg/kg body weight) in drinking water (Group 3, n=5), and tofogliflozin alone (Group 4, n=5). Group 2 and 3 mice were injected with AOM (15 mg/kg body weight) intraperitoneally once weekly from weeks 1 to 4 after initiation of the experiment. Mice were not anesthetized prior to AOM injection. At 9 weeks of age, the Group 3 and 4 mice received drinking water containing tofogliflozin. At 24 weeks of age, counting from 14 weeks of tofogliflozin treatment, all the mice were euthanized after 12 h of fasting. Then, blood samples were collected from the inferior vena cava for clinical chemistry and organs and tissues were removed for histopathological examination. For euthanasia, after inhalation of CO₂ the chest cavity was opened, and the heart was exposed and viewed directly to confirm death. The experimental protocol was approved by the Committee of Institutional Animal Experiments of Gifu University (authorization code: 30-7, dated 13 April 2018).

The livers, kidneys, white adipose tissues, and colorectal tissues were excised. The colon and rectum were opened longitudinally and fixed on a filter paper in 10% buffered formalin for more than 24 h. It was divided into a rectal portion (approximately 1 cm in length on the oral side from the dentate line) and colonic portion. Each segment and the other tissues were paraffin-embedded. From the rectal tissue blocks, two serial sections were subjected to hematoxylin and eosin (H&E) staining for histopathology and immunohistochemistry for β-catenin to evaluate the development of β-catenin accumulated crypts (BCACs) (16). Immunohistochemical analyses for β-catenin (1:1,000 dilution, BD Transduction Laboratories) was performed using a labeled streptavidin-biotin method (LSAB kit; DAKO; Agilent Technologies, Inc.). Immunohistochemical staining for F4/80 was also performed to examine macrophage infiltration in the adipose tissues, according to a previous study (28) with primary antibody (1:100 dilution; product code ab111101; Abcam). Immunoreactivity was considered as positive when apparent staining was detected in the cytoplasm and/or nuclei (28,29).

Serum was centrifuged (1,600 × g for 15 min) from the whole blood and was used for chemical analyses. Serum glucose, insulin, total cholesterol, free fatty acid (FFA), and triglyceride (TG) levels were determined by a commercial laboratory (SRL, Inc.). Serum TNF-α levels were estimated by mouse TNF-α ELISA kit (AKMTM-011; Shibayagi Co. Ltd.) in accordance with the manufacturer's protocol.

The mRNA expression in the colonic mucosa of the experimental mice was examined using quantitative real-time reverse transcription-polymerase chain reaction (qRT-PCR) analysis. Total RNA was isolated from the scraped colonic mucosa of all the experimental mice using PureLinkTM RNA Mini Kit (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions. The complementary DNA (cDNA) was amplified from 1.5 µg of total-RNA of each sample using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems; Thermo Fisher Scientific, Inc.). The primers used for the amplification of C-C motif chemokine-2 (CCL-2), tumor necrosis factor-α (TNF-α), F4/80, insulin-like growth factor-1 (IGF-1), insulin-like growth factor-2 (IGF-2), insulin-like growth factor binding protein-3 (IGFBP-3), and 18S specific genes were previously reported (30,31) or as follows: IGF-1 forward, 5′-TCGGCCTCATAGTACCCACT-3′ and reverse, 5′-ACGACATGATGTGTATCTTTATTGC-3′; IGF-2 forward, 5′-ACCTTCGGCCTTTGTCTGGTA-3′ and reverse, 5′-CGAAGGCCAAAGAGATGAGA-3′; and IGFBP-3 forward, 5′-GACGACGTACATTGCCTCAG-3′, and reverse, 5′-GTCTTTTGTGCAAAATAAGGCATA-3′. Real-time PCR was performed using a Light Cycler with FastStart Essential DNA Green Master (both from Roche Diagnostics) (Table III). The expression of these genes was normalized to 18S.

Expression of SGLT2 was confirmed by western blot for the HT29, HCT116, SW837, and SW480 human CRC cell lines, and Huh7 human hepatoma cell line. Huh7 was used as a positive control (25). These cell lines were obtained from the JCRB Cell Bank (Osaka, Japan). All cell lines were authenticated via short tandem repeat analysis by PCR and were certified to be free of bacteria, fungi and mycoplasmas by the cell bank. These cell lines were grown in appropriate media according to the instructions of cell bank. Cells were subjected to experiments within 6 months after receipt. All the cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin (all from Sigma-Aldrich; Merck KGaA) in an incubator at 37°C under a humidified atmosphere containing 5% CO₂.

Total protein was extracted from the cultured cells and equivalent amounts of protein (20 µg/lane) were examined by western blot as previously reported (32). The primary antibodies against SGLT2 (cat. no. 24654-1-AP) were obtained from ProteinTech Group, Inc. and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; product no. 2118) was obtained from Cell Signaling Technology, Inc. GAPDH served as a loading control.

SW480 and HT29 cells were seeded in 96-well plates. On the following day, the cells were treated with indicated concentrations (0, 0.05, 0.5, 5, and 50 µM) of tofogliflozin for 48 h. The cell proliferation assays were performed by 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay (Promega Corporation) in accordance with the manufacturer's protocol.

The data are represented as the mean ± standard error (SE). Differences between groups were analyzed using the Kruskal-Wallis test, then the Steel-Dwass test was performed between the groups on data to confirm statistical significance. A P-value of <0.05 was considered to indicate a statistically significant difference.

There was no significant difference in relative kidney weight, and white adipose tissue weight among all the groups (Table I). The body weight of tofogliflozin-treated mice was relatively increased compared to the control, which is consistent with a previous study demonstrating the body weight gain in SGLT2 inhibitor-treated mice (33). During the experiment, there were no clinical symptoms in all the groups. Histopathological examination did not reveal toxicity of tofogliflozin in the livers and the kidneys (data not shown).

The serum concentrations of glucose, insulin, total cholesterol, FFA, and TG in each group are listed in Table II. There were no significant differences in the results between the groups of mice treated with AOM alone and AOM followed by tofogliflozin. In this study, no adverse effect on serum parameters was observed after tofogliflozin administration in the experimental mice.

Colorectal pre-neoplastic lesions BCAC (34,35) were developed in the colons of AOM-injected mice with and without administration of tofogliflozin. Optical microscopic representative images of AOM-induced BCACs are presented in Fig. 1A. BCACs developed in the colon and rectum of all the mice injected with AOM (Groups 2 and 3), but not in the Group 1 and 4 mice that did not receive AOM. As revealed in Fig. 1B, the number of BCACs was significantly less in the mice treated with AOM plus tofogliflozin as compared to the ones treated with AOM alone (72% reduction, P<0.05).

TNF-α is an essential tumor promoter involved in obesity, inflammation, and carcinogenesis (36,37). Previous studies have indicated that chronic inflammation and activation of the IGF/IGF-IR axis are involved in colorectal carcinogenesis (16,38,39). Therefore, the serum levels of TNF-α were estimated by enzyme-linked immunosorbent assay (ELISA). The mRNA expression of the specific molecules associated with chronic inflammation and IGF signals, such as TNF-α, F4/80, CCL2, IGF1, IGF2, and IGFBP3, in colonic mucosa, was evaluated by qRT-PCR. Serum TNF-α levels tended to be lower in the AOM plus tofogliflozin-treated group as compared to the AOM-treated group, however the change was not statistically significant (Fig. 2A). There were also no statistically significant differences in the expression of the analyzed genes between the AOM plus tofogliflozin-treated group and the AOM-treated group, however, IGF1 was decreased by treatment with tofogliflozin. In addition, although there were no significant differences, the expression levels of TNF-α and F4/80 in the colon mucosa were also decreased in the tofogliflozin-treated group compared to the no treatment group (Fig. 2B and C).

Macrophages play significant roles in pro-inflammation in obese adipose tissues (40,41). In the present study, immunohistochemical staining for F4/80 was performed, which revealed significant higher macrophage infiltrations in the white adipose tissues of the AOM-treated mice as compared to the mice without AOM treatment (Fig. 3A). However, the number of F4/80 positive cells were decreased 0.25-fold in the mice treated with AOM/tofogliflozin as compared to those treated with AOM alone (Fig. 3B). In addition, it was investigated whether tofogliflozin administration attenuated chronic inflammation and decreased macrophage infiltration in the white adipose tissues. The expression of the genes associated with chronic inflammation, such as TNF-a and CCL2, analyzed by qRT-PCR in the white adipose tissues tended to be lower in the AOM plus tofogliflozin-treated group as compared to the AOM-treated group, although the differences were not statistically significant (Fig. 3C). These results indicated that administration of tofogliflozin attenuated the inflammatory response in the white adipose tissues of the diabetic mice.

Recent studies have demonstrated that several types of cancer cells exhibit SGLT2 expression (24–26). Therefore, the expression of SGLT2 in four human CRC cell lines, HT29, HCT116, SW837, and SW480 cells were examined. As revealed in Fig. 4A, western blot analysis revealed that SGLT2 was detected in these cells as well as the hepatoma cell line Huh7 (positive control) (25). To evaluate whether tofogliflozin could directly inhibit the growth of CRC cells, a cell proliferation assay was performed using SGLT2-expressing HT29 and SW480 cells. The MTS assay indicated that treatment with tofogliflozin exhibited no effects on the proliferation of these cells (Fig. 4B).