RKO and SW480 colorectal cancer (CRC) cell lines used in this study were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA), and MKN45 gastric cancer cell line was obtained from the Japanese Cancer Research Resources Cell Bank (JCRB, Tokyo, Japan). In addition, we used MIA PaCa-2 pancreatic cancer cell line [17]. RKO and MIA PaCa-2 cells were cultured in Eagle’s minimum essential medium (Eagle’s MEM; FUJIFILM Wako Pure Chemical Corp., Osaka, Japan) supplemented with penicillin-streptomycin (Sigma-Aldrich, St. Louis, MO, USA) and 10% (v/v) foetal bovine serum (FBS; Regular; Collected in Brazil, Processed in UK, Corning, NY, USA). MKN45 and SW480 cells were cultured in Roswell Park Memorial Institute 1640 medium (RPMI; FUJIFILM Wako Pure Chemical Corp.), containing penicillin–streptomycin and 10% (v/v) FBS. All cells were cultured at 37˚C in a 5% CO₂ atmosphere.

To prepare LC water, we mixed LC powder (Wedge Co., Ltd., Fuji, Japan) with MilliQ water at a final concentration of 0.03%. The mixture was allowed to stand overnight at 4˚C, and then LC water was collected as supernatants following centrifugation of the mixture. The main elements present in the LC powder were calcium and magnesium. LC water, prepared from MilliQ water mixed with the low concentration of LC powder, as described above, contained mainly the three minerals (sodium 1.8 mg/L, magnesium 1.9 mg/L, and calcium 15 mg/L) and was weakly alkaline (around pH 10) [18]. ContM and LCM were prepared by dissolving medium powder (Eagle’s MEM or RPMI 1640 medium; Shimadzu Diagnostics Corp., Tokyo, Japan) in MilliQ water and LC water, respectively, according to the manufacturer’s protocol. The pH of the medium was adjusted by adding an 8% NaHCO₃ solution of the appropriate volume to each medium and further adding a small amount of 1 N HCl solution to LCM for the RPMI 1640 medium. We confirmed that the pH of each medium was maintained around 7.4 in a 5% CO₂ atmosphere during cell culture, and we examined the effects of culturing cancer cells in these media.

Cells pre-cultured in ContM or LCM for 5 days were harvested during the exponential phase of growth, seeded in 96-well plates (1.2 × 10³ cells/well) with the respective media, and thereafter cultured for 4 days. Assessments of the numbers of viable cells were performed daily based on the WST-8 [2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt] method using a Cell Counting Kit-8 (Dojindo Laboratories, Kumamoto, Japan), according to the manufacturer’s protocol. Briefly, WST-8 solution was added to the cultured cells in each well, and following incubation for 3 h, absorbance of the well contents was measured at 450 nm using a microplate reader(Synergy LX multi-mode plate reader; BioTek Instruments Inc., Winooski, VT, USA).

The migratory and invasive capacities of cancer cells were evaluated using Corning® BioCoat™ Transwell chambers (#354578 and #354480, respectively; Corning) with 24-well inserts and 8-mm pore-sized membranes, following the manufacturer’s protocol. Briefly, following pre-culture in ContM or LCM for 6 days, RKO and MKN45 cells were suspended in serum-free each medium and plated (RKO cells: 2.5 × 10⁵ cells/well; MKN45 cells: 5 × 10⁵ cells/well) on uncoated upper chambers for migration assay and on Matrigel-coated upper chambers for invasion assay. RKO cells were also cultured in ContM with or without 1 mM N-acetyl-L-cysteine (NAC) (FUJIFILM Wako Pure Chemical Corp.) for 3 days, and plated on both chambers, as described above. SW480 cells (1.3 × 10⁵ cells/well in both assays) and MIA PaCa-2 cells (0.5 × 10⁵ cells/well in the migration assay and 2.5 × 10⁵ cells/well in the invasion assay) were similarly plated following pre-culture. Each medium containing 20% serum was used as a chemoattractant in the lower chambers. Following incubation for 48 h, non-migrating and non-invading cells were gently removed from the upper surface of the membrane using a cotton-tipped swab. Migrating and invading cells through the membrane were fixed and stained with Diff-Quik staining solution (#16920; Sysmex, Hyogo, Japan) and the numbers of cells in each well were counted in approximately 10 randomly selected fields of view under a microscope using BZ-X analysis software (Keyence, Osaka, Japan). The mean of the numbers of migratory/invasive cells per field of view was taken to be indicative of the migratory and invasive capacities, respectively, and the capacities of the LCM-cultured cells were expressed as a ratio relative to those of the ContM-cultured cells.

Intracellular ROS levels were assessed using CellROX Green Reagent (C10444, Invitrogen, Waltham, MA, USA) probe, according to the manufacturer’s instruction.Briefly, following pre-culture in ContM or LCM for 6 days, the cells were suspended in the respective media, seeded in 96-well plates (1 × 10⁴cells/well), and cultured for 24 h. H₂O₂ solution (FUJIFILM Wako Pure Chemical Corp.) was added at the indicated doses followed by incubation for 15 min, after which, CellROX green reagent was added at 6 µM, with further incubation for 25 min. Following incubation, the cells were washed with PBS, and fluorescence (Ex/Em = 485/528 nm) was analysed using a fluorescence microplate reader (Synergy HTX multi-mode plate reader; BioTek Instruments, Inc.). The cells in each well were also stained with a crystal violet dye, and absorbance was measured at 595 nm using the Synergy LX microplate reader. The intensity of ROS fluorescence was normalized with respect to the absorbance at 595 nm, and the normalized ROS values were considered indicative of ROS levels. The levels obtained for each treatment were expressed relative to those of the ContM culture condition without H₂O₂ treatment, as the relative ROS intensity. No decrease in cell viability was observed, even under the H₂O₂existence, at least during the measurement period.

RNA was extracted from cancer cells cultured in ContM or LCM for 6 days using a Direct-zol^TM RNA Miniprep Kit (Zymo Research, Irvine, CA, USA) following the manufacturer’s protocol, and gene expression was examined based on RNA-seq analysis. Libraries for RNA-seq samples were prepared using a TruSeq Stranded mRNA Sample Prep Kit (Illumina Inc., San Diego, CA, USA) following the manufacturer’s protocol. Deep sequencing was performed using the Illumina NovaSeq 6000 platform. More than 5 million reads per sample were acquired through the RNA-seq analysis. STAR (version 2.6.0c) [19] was used to map FASTQ reads to the hg38 reference genome and RSEM (v1.3.1) [20] was used for transcript abundance quantification. Gene expression levels are expressed as transcripts per million mapped sequence reads (TPM). With respect to TPM fold-change determination, TPM with values <1.0 were all designated ‘1.0’, given that these values are considered to be less reliable. Genes down- and up-regulated in response to culturing in LCM were selected based on log₂ TPM fold-change (LCM versus ContM) <−0.25 and >0.25, respectively, in both RKO and MKN cells, and the four cell lines examined were sorted into two groups, LCM-responsive cells and LCM-non-responsive cells, by two-way hierarchical clustering of the genes selected using the cutoff range.

Data are presented as mean ± standard deviation (SD). Differences among groups were assessed using Student’s t-test, with statistical significance set at P < 0.05. Heat maps were drawn using Java TreeView software (http://jtreeview.sourceforge.net/), and gene ontology (GO) enrichment analysis was performed using the gene annotation tool in the Metascape web application [21] (http://metascape.org).

To examine effects of LCM culture on the proliferation of cancer cells, we pre-cultured RKO and MKN45 cells in ContM or LCM for 5 days, and thereafter assessed cell growth under each culture condition using WST-8 assay over a 4-day time course. During the experimental period, we detected no significant differences between the assessed culture conditions with respect to the rates of cancer cell proliferation (Figure 1), thereby indicating that LCM has no appreciable effects on cell proliferation. Similar results were obtained for SW480 and MIA PaCa-2 cells (Figure S1). In addition, for the pre-culture period, growth rates determined by cell number counting were not significantly altered by LCM culture in all four assessed cancer cell lines (data not shown).

Figure 1. The effects of LCM culture on cancer cell proliferation.

Following pre-culture in ContM or LCM for 5 days, the rates of RKO and MKN45 cell proliferation were estimated using WST-8 assay over a 4-day time course. Data are presented as mean ± standard deviation (SD) of three wells of the 96-well plate, n= 3.

We examined the migratory and invasive capacities of the four cancer cell lines, using Transwell migration and invasion assays, which revealed that compared with MKN45 cells, RKO cells were characterized by higher capacities of both invasion and migration (Figure 2(A), Figure 2(B)); for RKO and MKN45 cells, when cultured in ContM, we obtained mean values (from four wells of the Transwell chamber) of 1530 ± 249 and 25.8 ± 9.83 for the numbers of migratory cells and 88.3 ± 22.8 and 32.4 ± 14.0 for the numbers invasive cells per field of view, respectively. Notably, for both cell lines, the migratory and invasive capacities decreased significantly in the LCM-cultured cells compared to the ContM-cultured cells (Figure 2(A), Figure 2(B)); the counts of migratory cells decrease to approximately 50% and 35%, and those of invasive cells decreased to approximately 20% and 35%, in RKO and MKN45 cells, respectively, thereby providing evidence for the suppression of cancer cell migration and invasion by LCM in both cell lines. Contrastingly, in the case of SW480 and MIA PaCa-2 cells, there was no significant differences between the ContM and LCM culture conditions in the either migratory or invasive capacity (Figure 2(C)).

Figure 2. The effects of LCM culture on cancer cell migration and invasion. RKO and MKN45 cells were pre-cultured in ContM or LCM for 6 days, after which, the migratory (A) and invasive (B) capacities were evaluated as described in the Materials and Methods section. Left; representative images of the lower surface of the Transwell membranes are shown (×10 magnification; scale bars, 200 μm). Right; the numbers of migrating (A) and invading (B) cells under both culture conditions were counted in selected fields of view under the microscope and statistically analysed. The migratory and invasive capacities of LCM-cultured cells are expressed as the ratio relative to those of ContM-cultured cells. In both assays, four wells of the Transwell chamber were used for the respective cells. Data are presented as mean ± standard deviation (SD), n= 4. **P< 0.01, ***P< 0.001 for LCM culture vs. ContM culture. SW480 and MIA PaCa-2 cells were similarly pre-cultured in ContM and LCM, and the migratory (C, left) and invasive (C, right) capacities were evaluated as described above. In both assays, two wells of the Transwell chamber were used. Data are presented as mean ± standard deviation (SD),n= 2.

With respect to oxidative stress, we initially examined whether NAC, a typical antioxidant, could inhibit the migration and/or invasion of RKO cells, which we established to have higher migratory/invasive capacities compared with MKN45 cells. NAC inhibited both of the intracellular accumulation of ROS associated with cell metabolism in the absence of H₂O₂treatment and H₂O₂(50 μm and 100 μm)-induced ROS production (Figure 3(A)), and also suppressed migration and invasion in the cells (Figure 3(B)). These findings show that under the assessed assay conditions, antioxidants could suppress the migration and invasion of the cancer cells.

Figure 3. The effects of NAC on ROS levels and the migratory and invasive capacities of RKO cells. RKO cells were pre-cultured for 3 days in ContM or the medium containing 1 mM NAC (ContM+NAC), after which, ROS levels were measured as described in the Materials and Methods section, and the levels of ROS are expressed relative to those in the cells cultured in ContM without H₂O₂ treatment, as the relative ROS intensity (A). Three wells of the 96-well plate were used for each concentration of H₂O₂. Data are presented as mean ± standard deviation (SD), n= 3. *P< 0.05, **P< 0.01 for ContM+NAC vs. ContM. The migratory and invasive capacities were analysed for each culture condition (B). Upper; representative images of the lower surface of the Transwell membranes are shown (×10 magnification; scale bars, 200 μm). Lower; the migratory and invasive capacities of the cells cultured in ContM+NAC are expressed as the relative ratio to those of ContM-cultured cells. In both assays, three wells of the Transwell chamber were used. Data are presented as mean ± standard deviation (SD), n= 3. *P< 0.05, **P< 0.01 for ContM+NAC vs. ContM.

Subsequently, we compared the accumulation of ROS in the LCM- and ContM-cultured cells.In the absence of H₂O₂ treatment, the levels of ROS in RKO and MKN45 cells cultured in LCM were found to be less than 50% of those detected in ContM-cultured cells (Figure 4 (upper)), thus tending to indicate an inhibition of ROS production in both cell lines following exposure to LCM. Similarly, when treated with H₂O₂at 50 μm and 100 μm, the ROS levels increased in both cell lines when cultured in ContM, but the increased levels were also significantly suppressed in the LCM-cultured cells (Figure 4 (upper)). These findings thus indicate that culturing in LCM can inhibit not only the intracellular accumulation of ROS but also ROS produced in response to the oxidative stress by the H₂O₂ treatment in both cancer cells. On the other hand, in SW480 cells the intracellular accumulation of ROS was only weakly inhibited by LCM culture, whereas culturing under these conditions had no significant effects regarding H₂O₂-induced ROS production (Figure 4 (lower)). Moreover, in the case of MIA PaCa-2 cells, culturing in LCM inhibited neither intracellular ROS accumulation nor H₂O₂-induced ROS production (Figure 4 (lower)).

Figure 4. The effects of LCM culture on the accumulation of ROS. Following pre-culture in ContM or LCM for 6 days, levels of intracellular ROS in the four assessed cancer cell lines, with and without H₂O₂ treatment, were determined based on measurements of fluorescence intensity using CellROX Green Reagent. The levels of ROS are expressed in terms of fluorescence intensity relative to those in the ContM culture condition without H₂O₂ treatment, as the relative ROS intensity. Data are presented as mean ± SD of three wells of the 96-well plate, n= 3. *P< 0.05, **P< 0.01, ***P< 0.001 for LCM culture vs. ContM culture.

To determine the effects of LC water on gene expression, we analysed changes in gene expression in the four assessed cancer cell lines (RKO, MKN45, SW480, and MIA PaCa-2), cultured in LCM or ContM based on RNA-seq. Among 17,129 genes expressed in the four cell lines, 111 and 83 genes were selected as down- and up-regulated, respectively, in response to culturing in LCM culture in both of RKO and MKN45 cells (Figure 5). Heatmaps generated to visualize patterns in gene expression change revealed certain differences between the two cell line groups (Figure 5 (left panel)); the group consisting of RKO and MKN45 cells, being responsive to the migration-/invasion-suppressive and antioxidative effects of culturing in LCM, and the other group consisting of SW480 and MIA PaCa-2 cells, being either weakly responsive or non-responsive to the effects.

Figure 5. The effects of LCM culture on gene expression in cancer cells. RNA was extracted from the four assessed cancer cell lines cultured in ContM or LCM for 6 days, and gene expression was analysed based on RNA-seq, as described in the Materials and Methods section. Genes down- and up-regulated in response to culturing in LCM were selected on the basis of a log₂ TPM fold-change (LCM versus ContM) <−0.25 and >0.25, respectively, in both of RKO and MKN45 cells. In heatmaps, changes in the levels of 111 down-regulated and 83 up-regulated genes in the four assessed cell lines are presented in order of the rate of change in RKO cells in left upper and lower panels, respectively. GO groups selected based on GO analysis from the down- and up-regulated genes with highest significance are shown in right upper and lower panels, respectively, with genes classified in the respective GO groups. No other GO groups with high significance were selected from either down-regulated genes or up-regulated genes.

The 111 down-regulated genes were most significantly enriched in the GO term ‘Recognition and association of DNA glycosylase with site containing an affected pyrimidine’ (P = 1 ×10^-4.2) (Figure 5 (right upper panel)). The functions of the GO group (Dn-GO) include base-excision repair involved in the repair of oxidative DNA damage and oxidative stress-induced cell cycle arrest, and NEIL1 and CDKN2D in Dn-GO are the typical genes playing roles in the response to oxidative stress (Table 1). These findings indicate that in the two LCM-responsive cancer cells, there was a reduction in the response to oxidative stress in the culture condition of LCM compared with that of ContM, which is taken to be indicative of antioxidative effects by LCM culture in the two cancer cells. Furthermore, some of genes classified in the Dn-GO group, namely,SNAI1, TUBA3D,GTF2H2C, and SYCE2, have been reported to be up-regulated in certain cancer cells and function in cancer progression, including the promotion of cancer cell proliferation, metastasis, migration/invasion, or epithelial-mesenchymal transition (EMT) (Table 1). In addition, we identified the top 10 down-regulated genes in RKO (Dn-RK) and MKN45 (Dn-MK) cells, respectively, some of which are also associated with the progression of cancer (Table 1).

The 83 up-regulated genes were most significantly enriched in the GO term ‘RMTs methylate histone arginines’ (P = 1 × 10^−8.7) (Figure 5 (right lower panel)), thereby tending to indicate that LCM-induced gene expression involves epigenetic modification. Some of genes in the GO group (Up-GO), including TLE4, PML, CDKN1C, and U2AF1, have been reported to be down-regulated in certain cancer cell lines, including CRC cells, and have been established to play roles in cancer suppression, such as the inhibition of cancer cell proliferation, metastasis, migration/invasion, or EMT (Table 2). Other genes among the top 10 up-regulated genes in RKO (Up-RK) and MKN45 (Up-MK) cells have also been established to function in cancer suppression (Table 2).