Recombinant stable and mutant forms of human Gal-9 lacking the entire linker region were expressed and purified as described in our previous report (27). Lactose, sucrose, and fetal bovine serum (FBS) were purchased from Wako Chemicals. Cell Counting Kit-8 (CCK-8) was purchased from Dojindo Laboratories, and all other chemicals were obtained from Sigma Chemical.

The human colon cancer cell lines CACO-2, CW-2, COLO-320, LoVo, and the colorectal cancer cell line WiDr were obtained from the Japanese Cancer Research Resources Bank (Tokyo, Japan). All the cell lines were certified by STRA and mycoplasma testing was carried out for the cell lines used in our experiments. The cells were cultured in RPMI-1640 (Gibco; Invitrogen; Thermo Fisher Scientific, Inc.) supplemented with 10% heat-inactivated FBS, and penicillin-streptomycin (100 µg/ml; Invitrogen; Thermo Fisher Scientific, Inc.) in a humidified atmosphere with 5% CO₂ at 37°C.

We performed cell proliferation assays using the CCK-8 assay according to the manufacturer's instructions. Each cell line (1×10⁴ cells per well) was seeded into the wells of a 96-well plate and cultured in 100 µl of RPMI-1640 medium supplemented with 10% FBS. After incubation for 24 h, the cells were treated with 0.01, 0.03, 0.10, and 0.30 µM of Gal-9 added to the culture medium and subsequently cultured for an additional 48 h. Furthermore, 30 mM of lactose was added to inhibit the galactoside-binding of Gal-9, and sucrose was added as a control. CCK-8 reagent (10 µl) was added to each well, and the plates were incubated at 37°C for 3 h. The absorbance of each well was measured at 450 nm using an auto-microplate reader.

Cell apoptosis assays that measured the amounts of caspase-cleaved keratin-18 (CCK-18) were performed using the M30-ApoptoSensek ELISA kit (Peviva AB) (28). Each cell type (5×10³ cells per well) was seeded into a 96-well plate and cultured in 100 µl of culture medium for 24 h. Seeded cells were then treated with 0.3 µM of Gal-9. The rest of the experiments were carried out according to the manufacturer's instructions. The amount of antigen in the controls and samples was calculated by interpolation from a standard curve.

We conducted a flow cytometric analysis using the Cycle Phase Determination kit (Cayman Chemical Co.) to evaluate the mechanism of growth inhibition by Gal-9. CW-2 cells were digested with 0.25% trypsin and plated in 100-mm-diameter dishes at 1.0×10⁶ cells per dish. After incubation for 24 h without FBS, CW-2 cells were treated with 0.3 µM of Gal-9 or dimethyl sulfoxide (DMSO; control) for another 24 h, then harvested, washed with phosphate-buffered saline (PBS), suspended in 500 µl of PBS plus 10 µl of RNase A (250 µg/ml) and 10 µl of propidium iodide (PI) stain (100 µg/ml), and incubated for 30 min.

To determine the apoptosis rate of CW-2, CACO-2, and WiDr cells, we used flow cytometry and the Annexin V-FITC Early Apoptosis Detection Kit (Cell Signaling Technology, Inc.). CW-2, CACO-2 and WiDr cells were plated in 100-mm-diameter dishes at 1.0×10⁶ cells per dish and treated with 0.3 µM Gal-9 or DMSO control for 24 h. After incubation for 24 h, CW-2, CACO-2 and WiDr cells were harvested, and washed with PBS. Staining was performed according to the manufacturer's protocol. After adding Annexin V-FITC and PI, we analyzed apoptosis and necrotic cell death. Flow cytometry was conducted with a Cytomics FC 500 flow cytometer (Beckman Coulter) equipped with a 480-nm argon laser. Cell percentages were analyzed with Kaluza software version 2.1 (Beckmann Coulter).

The lysates were prepared according to the methods described in our previous report (29). All steps were carried out at 4°C. Protein concentrations were measured using a dye-binding protein assay based on the Bradford method (30).

Human phosphorylated receptor tyrosine kinases (p-RTKs) were assayed using Human phospho-RTK Array Kits (R&D Systems), according to the manufacturer's instructions. Briefly, p-RTK array membranes were blocked with 5% bovine serum albumin/0.01 M Tris-HCl, pH 7.6 (TBS) for 1 h and then incubated with 2 ml of the lysate prepared from cell lines after normalization, so that the amounts of protein were equal. After three washes for 10 min each with TBS plus 0.1% v/v Tween-20 and two washes for 10 min with TBS alone to remove unbound material, the membranes were incubated with anti-phospho-tyrosine-horseradish peroxidase antibody for 2 h at room temperature. The unbound HRP antibody was washed out with TBS plus 0.1% Tween-20. Finally, each array membrane was exposed to an X-ray film using a chemiluminescence detection system (Perkin-Elmer Co.).

The RayBio Human Angiogenesis Antibody Array (RayBiotech Inc.) was used according to the manufacturer's protocol. This method is a dot-based assay enabling detection and comparison of 20 angiogenesis-specific cytokines. Each array membrane was exposed to an X-ray film using a chemiluminescence detection system (PerkinElmer Co.).

The samples of cancer cell lines were processed for total RNA extraction with the miRNeasy Mini Kit (Qiagen GmbH) according to the manufacturer's instructions. Typically, RNA samples showed A_260/280 ratios of between 1.9 and 2.1, using an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc.).

After RNA measurement with an RNA 6000 Nano kit (Agilent Technologies, Inc.), the samples were labeled using a miRCURY Hy3/Hy5 Power labeling kit and were hybridized on a human miRNA Oligo chip10, version 19.0 (Toray Industries). Scanning was conducted with the 3D-Gene Scanner 3000 (Toray Industries). 3D-Gene extraction version 1.2 software (Toray) was used to read the raw intensity of the image. To determine the change in miRNA expression between Gal-9-treated and control samples, the raw data were analyzed via GeneSpring GX v 10.0 (Agilent Technologies, Inc.). Samples were first normalized relative to 28S RNA, and the baseline was corrected to the median of all samples.

Replicate data were consolidated into two groups, those from Gal-9-treated cells and those from control cells, and were organized by using the hierarchical clustering and ANOVA functions in the GeneSpring GX v 10.0 software (Agilent Technologies, Inc.). Hierarchical clustering was performed by utilization of the clustering function (condition tree) and Euclidean correlation as a distance metric. Two-way ANOVA analysis and asymptotic P-value (<0.05) computation without any error correction on the samples were performed to search for the miRNAs that varied most prominently across the different groups. Only changes >50% for at least one of the time points for each sample were considered significant. All of the analyzed data were scaled by global normalization. The statistical significance of differentially expressed miRNAs was analyzed using Student's t-test.

All our micoroarray data in this study were submitted as a complete data set to the NCBI Gene Expression Omnibus (GEO), no. GSE163790 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE163790).

Animal experiments were performed according to the guidelines of the Committee on Experimental Animals of Kagawa University, Kagawa, Japan, following the National Institutes of Health guide for the care and use of laboratory animals. Approval Number of this animal experiment is HEISEI-25-164. We purchased 40 male athymic mice (BALB/c-nu/nu; 6 weeks old; 20–25 g) from Japan SLC (Shizuoka, Japan). The mice were maintained under specific pathogen-free conditions using a laminar airflow rack. The mice had continuous free access to sterilized (γ-irradiated) food (CL-2; CLEA Japan, Inc.) and autoclaved water. A total of 40 mice were equally divided into two groups (CACO-2 group and CW-2 group). Each mouse was subcutaneously inoculated with CACO-2 or CW-2 cells (3×10⁶ cells per animal) in the flank region. Subsequently, when the xenografts were identifiable as masses with a maximal diameter >3 mm, we randomly assigned the animals to the following groups: i) CACO-2 only (n=5), ii) CACO-2 treated with 90 µg of Gal-9 (n=5), iii) CW-2 only (n=7), and iv) CW-2 treated with 90 µg of Gal-9 (n=5). Our laboratory previously optimized the concentration of galectin-9 for the treatment of various gastroenterological cancers in mouse animal model (23,25,31). The Gal-9-treated groups were intraperitoneally (i.p.) injected with Gal-9 (90 µg) five times per week; the control groups were administered 5% DMSO alone. The tumor growth was monitored daily by the same investigators (AM and KN), and the tumor size was measured two times per week. The tumor volume (mm³) was calculated as the tumor length (mm) × tumor width (mm)²/2 (32). All animals were sacrificed on day 31 after the treatment, with all animals remaining alive until this time point. Tumor-bearing mice were euthanized using CO₂ (20% of the volume of the chamber per min) followed by cervical dislocation at the end of this experiment.

The gene transfection miR-1237 inhibitor and negative control miRNA were obtained from Thermo Fisher Scientific, Inc. CW-2 and WiDr cells were seeded into 6-well plates. After 24 h, the CW-2 and WiDr cells were transfected with the miR-1237 inhibitor or negative control miRNA at a final concentration of 0.3 µM using Lipofectamine RNAiMAX (Invitrogen; Thermo Fisher Scientific, Inc.). After 48 h of incubation, the cells were harvested and washed with ice-cold PBS for subsequent analysis. The integrity of miR-1237 transfection was confirmed using real-time RT-PCR (Fig. S1). However, no miR-1237 reduction was detected in both cancer cell lines after microRNA inhibitor transfection. This inhibitor just binds to specific microRNA and stops translation of target gene and therefore, no reduction in specific miRNA was detected using real-time RT-PCR. In addition, no obvious miR-1237 target was published (no obvious downstream pathway). We demonstrated the differences in phenotype using MTT assay after miR-1237 inhibition.

All analyses were conducted using the GraphPad Prism 8.4.2 software (GraphPad Software, Inc.). Two-tailed paired or unpaired analysis between the groups was conducted using Student's t-test. For comparison of multiple groups, Kruskal-Wallis test was first performed, and then Dunnett test was conducted as a post hoc test. P<0.05 was considered to indicate a significant difference between groups.

To evaluate the effects of Gal-9 on the growth of human colon and colorectal cancer cells in vitro, we examined the effects of Gal-9 on the proliferation of the five colon cancer and colorectal cell lines CACO-2, CW-2, COLO-320, LoVo, and WiDr. Cells were grown in 10% FBS and treated with 0.01, 0.03, 0.10, and 0.30 µM of Gal-9 with untreated cells as controls. The cell proliferation assay was conducted 48 h after the addition of Gal-9. As shown in Fig. 1A and B, Gal-9 led to a dose-dependent and strong inhibition of cell proliferation in CACO-2 and CW-2, which are Gal-9-sensitive colon cancer cell lines. However, a significant antiproliferative effect of Gal-9 on COLO-320, LoVo, and WiDr cell lines was not detected (Fig. 1C-E). These results are also expressed as percentages of viable cells compared to control 48 h after Gal-9 treatment (Fig. 1F).

CCK-18 is produced specifically in apoptosis. We performed ELISAs to determine the levels of CCK-18 in colorectal cancer cells treated with 0.3 µM Gal-9. We found that Gal-9 significantly increased the CCK-18 levels in Gal-9-sensitive CACO-2 and CW-2 cells, but not in WiDr cells (Fig. 2). To further examine whether Gal-9 promotes apoptosis, we also assessed an apoptosis marker in CW-2, CACO-2, and WiDr cells treated for 24 h with 0.3 µM Gal-9. In flow cytometry measurements, the percentages of Annexin V⁺ cells were significantly increased in Gal-9-treated CW-2 and CACO-2 cells, but not in WiDr cells (Fig. 3). This suggests that Gal-9 administration induced apoptosis in Gal-9-sensitive cell lines.

We next used a p-RTK array system to identify ‘key RTKs’ associated with the antitumor effect of Gal-9. By using the antibody array (Fig. 4), we simultaneously screened the expression of 42 different activated RTKs in CACO-2, CW-2, and WiDr cell lines with and without the addition of 0.3 µM Gal-9. It was noted that Gal-9 enhanced the expression of ALK, DDR1, and EphA10 in Gal-9-sensitive CW-2 cells (Fig. 4). By contrast, CACO-2 cells and Gal-9-resistant WiDr cells did not exhibit a change in the expression of the activated RTKs.

In order to examine the relationship between angiogenesis and Gal-9, an angiogenesis antibody array system was used to identify key angiogenesis-related molecules associated with the antitumor effect of Gal-9 (Fig. 5). By using the antibody array, we simultaneously screened the expression of 20 different angiogenesis-associated molecules in CACO-2, CW-2, and WiDr cells with or without Gal-9 administration. The expression levels of interleukin-8 (IL-8) and tissue inhibitor of metalloproteinases-2 (TIMP-2) were induced by Gal-9 treatment in Gal-9-resistant WiDr cells as detected by the protein array (Fig. 5). In Gal-9-sensitive CACO-2 and CW-2 cells, the expression of angiogenesis molecules did not change following treatment with Gal-9 (Fig. 5).

To examine whether growth inhibition was due to cell cycle changes, we investigated the cell cycle profiles of CW-2 cells 24 h after treatment, with or without 0.3 µM Gal-9, using flow cytometry. Treatment with 0.3 µM Gal-9 significantly increased the percentage of cells in the G₀/G₁ phase and significantly decreased the percentage of cells in the S phase at 24 h after treatment (Fig. 6). This result indicates that Gal-9 blocked the cell cycle progression from the G₀/G₁ phase to the S phase and may induce apoptosis.

To investigate the effects of Gal-9 administration on Gal-9-sensitive cells, we examined the cell cycle-related protein expression in CACO-2 and CW-2 cells using western blotting. Cells were treated with 0 or 0.3 µM Gal-9 for 24–48 h. We observed no obvious reduction in cyclin D1, which is one of the key proteins involved in the transition from the G₀ to the G₁ phase, due to Gal-9 treatment in CACO-2 and CW-2 cells (Fig. 7). Additionally, the analysis of other proteins associated with G₀/G₁ transition indicated that Cdk4 and Cdk6, the catalytic subunits of cyclin D1, were not decreased at any time point in CACO-2 cells after Gal-9 treatment (Fig. 7). Cyclin E, which is the key cyclin for G₁/S transition, was unchanged, and no pRB reduction was detected in CACO-2 and CW-2 cells after Gal-9 treatment (Fig. 7).

These findings suggest that Gal-9 is not involved in the regulation of cell cycle-related molecules in Gal-9-sensitive cells.

To determine whether Gal-9 can affect tumor growth in vivo, we subcutaneously inoculated nude mice with CACO-2 and CW-2 cells, followed by i.p. injection of Gal-9. The Gal-9 treatment at a dose of 90 µg significantly inhibited tumor growth in both CACO-2 and CW-2 implanted tumors compared to untreated control mice, as determined by integrated tumor growth curves (Student's t-test, *P<0.05; Fig. 8). Gal-9 had no apparent toxic effects on the mice including their body weight during the study (data not shown). Furthermore, all animals survived until the end of the experiment.

Using a custom microarray platform, we analyzed the expression levels of 2,555 human miRNA probes in colon cancer cell lines in vitro and in vivo with and without Gal-9 treatment. As shown in Fig. 9, when the expression of miRNAs was studied in CACO-2 cells with and without Gal-9 treatment in vitro and in vivo, only hsa-miR-1246 was found to be commonly upregulated 24 h after Gal-9 treatment (P=0.00000689 and P=0.000804, respectively), whereas hsa-miR-15b-5p and hsa-miR-1237 expression levels were significantly changed in CW-2 cells in vitro (P=0.00956 and P=0.0033, respectively) and in vivo (P=0.00582 and P=0.000289, respectively). In addition, only hsa-miR-1237 was commonly altered in CACO-2 cells in vitro, CW-2 cells in vitro, and CW-2 cells in vivo.

Unsupervised hierarchical clustering analysis with Pearson's correlation showed that CACO-2 cells in vitro, CW-2 cells in vitro, and CW-2 cells in vivo treated with Gal-9 clustered together and separately from untreated CACO-2 and CW-2 cells (Fig. 9).

Our in vitro findings led us to hypothesize that Gal-9 might inhibit the proliferation of colon and colorectal cancer cells via miR-1237 downregulation. To examine this hypothesis, we examined CW-2 and WiDr cells transfected with miR-1237 inhibitor or negative control miRNA. After 24 h of incubation, we examined the cell proliferation with or without miR-1237. Interestingly, the percentages of viable cells were significantly diminished 24 h after miR-1237 inhibitor transfection in the colon cancer CW-2 cells, but not in the colorectal cancer WiDr cells (Fig. 10). This revealed that miR-1237 inhibited by Gal-9 suppressed the cell proliferation in one of the Gal-9-sensitive colon cancer cell lines.