Six human CRC cell lines were used in this study. HCT116 and LoVo cells were purchased from the Japanese Collection of Research Bioresources Cell Bank (JCRB; Tokyo, Japan), and HT29 and SW480 cells were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). DLD-1 cells were purchased from the Cell Resource Center for the Biomedical Research Institute of Development, Aging, and Cancer at Tohoku University. The KM12SM cells were the kind gift of Professor T. Minamoto (Kanazawa University, Ishikawa, Japan). Authenticity of all cells was confirmed. Cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum (FBS), 100 U/ml penicillin and 100 µg/ml streptomycin in a 37°C incubator supplied with 5% CO₂. Hypoxic culture conditions (1% O₂) were achieved in a multi-gas incubator (Panasonic MCO-5MUV; Panasonic, Osaka, Japan) using a continuous infusion of a gas mixture (95% N₂, 5% CO₂).

The use of clinical samples was approved by the Human Ethics Review Committee of the Graduate School of Medicine, Osaka University. Written informed consent was obtained from all patients included in the study.

For western blot analysis, 6 surgical CRC samples from patients with no preoperative chemotherapy or irradiation were randomly selected from samples collected at Osaka University Hospital in 2016. For immunohistochemical staining, 14 surgical samples of colorectal liver metastases from patients with no preoperative chemotherapy or irradiation were collected from 2012 to 2015 at the Osaka University Hospital.

Microarray analysis was performed as previously described (12,16,24). A total of 222 clinical samples were collected from consecutive patients who underwent curative (R0) resection for CRC from 2003 to 2006 at the Osaka University Hospital and its nine associated hospitals. The mean follow-up time was 40.2±26.4 months for patients with disease-free survival (DFS) and 52.7±18.5 months for all surviving patients (overall survival, OS). Table I shows the clinicopathological features of the patients, including gender, tumor location, depth of invasion, lymph node metastasis, histological grade, Dukes' stage and vessel invasion. None of the patients whose samples were analyzed in this study received preoperative chemotherapy or irradiation. After surgery, patients with Dukes' stage C tumors were generally treated with 5-fluorouracil (5-FU)-based chemotherapy.

mRNA expression was determined by real-time polymerase chain reaction (PCR). Total RNA was isolated using an RNeasy Maxi kit (Qiagen, Valencia, CA, USA), and reverse transcription was performed using the reverse transcription system (Promega, Madison, WI, USA) according to the manufacturer's instructions. The real-time PCR analysis was performed using the LightCycler 2.0 instrument system (Roche Diagnostics, Mannheim, Germany), based on Thunderbird SYBR qPCR Mix (Toyobo Co., Ltd., Osaka, Japan). The β-actin housekeeping gene was used as an internal control. The following primers were used in the present study: ALDOA: forward 5′-CTGCCAGTATGTGACCGAGA-3′ and reverse 5′-ACAGGAAGGTGATCCCAGTG-3′; HIF1A: forward 5′-GAAAGCGCAAGTCCTCAAAG-3′ and reverse 5′-TGGGTAGGAGATGGAGATGC-3′; vimentin: forward 5′-CCCTCACCTGTGAAGTGGAT-3′ and reverse 5′-TCCAGCAGCTTCCTGTAGGT-3′; and β-actin: forward 5′-GATGAGATTGGCATGGCTTT-3′ and reverse 5′-CACCTTCACC GTTCCAGTTT-3′. The analysis was performed using the standard curve method. For data analysis, raw counts were normalized to the housekeeping gene average for the same time-point and condition.

Proteins were extracted from the cells and clinical samples in radio-immunoprecipitation assay (RIPA) buffer (Thermo Fisher Scientific, Waltham, MA, USA). Clinical samples were homogenized in RIPA buffer using an ultrasonic disruptor (Ultrasonic Disruptor UD-201; Tomy Seiko, Co., Ltd., Tokyo, Japan). Proteins were separated by electrophoresis on SDS-PAGE Tris-HCl gels (Bio-Rad Laboratories, Hercules, CA, USA). The separated proteins were transferred onto polyvinylidene difluoride membranes and blocked with 5% skim milk in Tris-buffered saline with Tween-20 (TBS-T) at room temperature for 1 h. The blots were incubated with primary antibodies overnight and then with HRP-linked anti-rabbit or anti-mouse IgG (GE Healthcare Biosciences, Piscataway, NJ, USA) at a dilution of 1:100,000 for 1 h at room temperature. The antigen-antibody complex was detected with the ECL Prime Western blotting detection kit (GE Healthcare Biosciences).

Western blot analyses were carried out using the following antibodies: anti-ALDOA (#3188; Cell Signaling Technology, Danvers, MA, USA); anti-E-cadherin (sc-7870; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA); anti-vimentin (#5741; Cell Signaling Technology) and anti-ACTB (A2066; Sigma-Aldrich, St. Louis, MO, USA).

Paraffin-embedded tissue sections (4-µm) were deparaffinized. After antigen retrieval treatment, immunohistochemical staining was performed using the Vectastain ABC Peroxidase kit (Vector Laboratories, Burlingame, CA, USA). The tissue sections were then incubated overnight at 4°C with antibodies at the following dilutions: anti-ALDOA, 1:100 (ab169544; Abcam, Cambridge, MA, USA) and anti-CA9, 1:100 (#5649; Cell Signaling Technology).

Cell lines were transfected with an ALDOA expression vector (pCMV6-XL5 vector, SC108418; Origene Technologies, Inc., Rockville, MD, USA) using the FuGENE^®6 transfection reagent (Promega) according to the manufacturer's recommended protocol. The same method was used to transfect cells with an empty vector as negative controls.

Cell lines were transfected with 20 nM siRNA targeting ALDOA expression (Silencer Select siRNA, siRNA ID s71, cat. no. 4390824; Life Technologies, Carlsbad, CA, USA: 5′-AGUCCCUCUUCGU CUCUAATT-3′ and 5′-UUAGAGACGAAGAGGGACUAG-3′) using Lipofectamine^® RNAiMAX transfection reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's recommended protocol. The same method was used to transfect cells with negative control siRNA (Silencer Select negative control #1 siRNA, cat. no. 4390843; Life Technologies).

The proliferation assay was performed as previously described (16). ALDOA overexpression cells and knockdown cells plus controls for each cell line were seeded and cultured in a 37°C incubator supplied with 5% CO₂. After 24, 48, 72 and 96 h of incubation, cell viability was measured using the Cell Counting kit-8 (CCK-8; Dojindo Laboratories, Kumamoto, Japan) according to the manufacturer's recommended protocol. CCK-8 solution was added to each well and incubated for 2 h.

The chemosensitivity assay was performed as previously described (16). Cell viability was measured after 48 h of incubation in drug-containing medium at IC₅₀ concentrations.

The radiosensitivity assay was performed as previously described (16). Cells were exposed to 8-Gy single-dose irradiation using a Gammacell 40 Exactor ¹³⁷Cs radiation source (Nordion, Inc., Ottawa, ON, Canada) and cell viability was measured after 48 h of incubation in normal medium.

The ability of cells to form spheres was evaluated in Dulbecco's modified Eagle's medium/Nutrient Mixture F-12 (DMEM/F12) supplemented with 20 ng/ml epidermal growth factor (Invitrogen), 20 ng/ml human platelet growth factor (Sigma-Aldrich) and 1% antibiotic-antimycotic solution (Invitrogen). Single cells were plated at a concentration of 100 cells/well in each well of a 96-well ultralow attachment plate (Corning Life Sciences, Acton, MA, USA) and cultured in a 37°C incubator supplied with 5% CO₂. The number of spheres ≥100 µm in each well was counted to evaluate the sphere-forming ability.

The invasion assay was performed using Corning^® BioCoat™ Matrigel^® Invasion Chambers (Corning Life Sciences). Briefly, 25,000 cells were seeded in triplicate onto the Matrigel-coated membrane. After 48 h, cells that had invaded the undersurface of the membrane were fixed and stained with Diff-Quik (Sysmex Internal Reagents, Co., Ltd., Kobe, Japan). Three microscopic fields were randomly selected for cell counting.

Lactate production was measured using the Lactate Colorimetric Assay kit II (BioVision, Inc., Milpitas, CA, USA). Cells were seeded in 12-well plates (Corning Life Sciences) at densities of 100,000 cells (DLD-1) and 150,000 cells (HT29 and KM12SM) per well. After 24 h, the cells were transfected with siRNA targeting ALDOA or with an ALDOA expression vector, and then 48 h after transfection, the subconfluent cells were washed three times with phosphate-buffered saline (PBS). Krebs-Ringer bicarbonate buffer containing 10 mM glucose was added, and the amount of lactate was assessed 30 min later in aliquots of media from each well. The absorbance at 450 nm was then measured using the Bio-Rad Model 680 XR microplate reader (Bio-Rad Laboratories).

Apoptosis was measured by flow cytometry using the Annexin V-FITC apoptosis kit (BioVision) according to the manufacturer's recommended protocol.

The samples were analyzed using BD FACSAria IIu™ instrument (BD Biosciences, San Jose, CA, USA). Both early apoptotic (Annexin V-FITC-positive, PI-negative) and necrotic/late apoptotic (Annexin V-FITC-positive, PI-positive) cells were quantified as apoptotic cells.

ROS production was measured using the CellROX^® Deep Red reagent (Invitrogen) according to the manufacturer's recommended protocol. Fluorescence was measured using the BD FACSAria IIu instrument (BD Biosciences).

Cell cycle phase distribution was analyzed by flow cytometry using PI staining. Briefly, cells were fixed in 70% ethanol on ice for 30 min. Fixed cells were incubated in a solution containing 0.1 mg/ml RNase A at 37°C for 20 min and then in a solution containing 25 µg/ml PI on ice for 2 min. Fluorescence was measured immediately using the BD FACSAria IIu instrument. Flow cytometry data were analyzed using FlowJo software (Tree Star, Inc., Ashland, OR, USA).

Data management and statistical analysis were performed using the JMP software (version 11.0; SAS Institute, Inc., Cary, NC USA). Tests for associations were performed using the Student's t-tests and Chi-squared tests. The cumulative patient OS and DFS were compared using the Kaplan-Meier technique, and log-rank tests and Cox's model were used to assess the risk ratio for multivariate analysis. A P<0.05 was considered statistically significant.

Western blot analysis indicated that in 5 out of 6 randomly selected CRC surgical samples, ALDOA expression was higher in tumor tissue than in normal colon mucosa (Fig. 1A). Then, the 222 CRC patients were divided into two groups according to the ALDOA mRNA expression level in the tumor samples. The relationships between ALDOA mRNA expression and clinicopathological features were analyzed (Table I). The high ALDOA expression group included more patients with lymph node metastasis than the low expression group (P=0.016), as for other clinicopathological factors, there were no significant differences.

DFS and OS were significantly longer in the low expression group than in the high expression group (DFS: P=0.0008, OS: P=0.0072; Fig. 1B). Univariate and multivariate analysis showed that ALDOA mRNA expression was a significant prognostic factor for DFS (P=0.0013; Table II) and for OS (P=0.0124; Table III).

In all of the CRC cell lines we examined, ALDOA mRNA expression was significantly increased after 48 to 72 h of culture in hypoxic conditions compared with the levels in cells cultured in normoxic conditions (Fig. 1C). Immunohistochemical staining of CRC liver metastases showed that the regions stained by the anti-ALDOA antibody were almost the same as the regions stained by an antibody to carbonic anhydrase IX (CA9), which is an endogenous hypoxia marker (Fig. 1D).

PCR and western blot analysis demonstrated high transfer efficiency of the ALDOA expression vector and of the siRNA targeting ALDOA (data not shown). In the CRC-derived cell lines DLD-1 HT29 and KM12SM, upregulation of ALDOA expression induced significant cell proliferation compared to controls at 72 h (P<0.05; Fig. 2A), and downregulation of ALDOA expression reduced cell proliferation compared to controls at 72 h (P<0.05; Fig. 2B).

Upregulation of ALDOA significantly reduced chemosensi-tivity to 5-FU in DLD-1 and HT29 cells and to oxaliplatin in all three cell lines compared to controls (P<0.05; Fig. 2C). In contrast, ALDOA downregulation significantly induced chemosensitivity to 5-FU in DLD-1 and HT29 cells and to oxaliplatin in DLD-1 and KM12SM cells compared to controls (P<0.05; Fig. 2D). Upregulation of ALDOA significantly reduced radiosensitivity in DLD-1 and HT29 cells and that downregulation significantly induced radiosensitivity in DLD-1 and KM12SM cells compared to controls (P<0.05; Fig. 2C and D).

DLD-1, HT29 and KM12SM cells that overexpressed ALDOA were more tumorigenic than controls (P<0.05; Fig. 2E). In DLD-1 and HT29, ALDOA knockdown cells were less tumorigenic than controls (P<0.05; Fig. 2F).

DLD-1 cells that overexpressed ALDOA invaded the Matrigel-coated membrane significantly faster than control cells (P<0.05) and ALDOA knockdown cells invaded significantly more slowly than control (P<0.05; Fig. 2G).

To examine the role of ALDOA in hypoxia, the proliferation and sphere formation assay were performed in hypoxic conditions. DLD-1 cells transfected with siRNA targeting ALDOA barely proliferated and formed very few spheres compared to controls (P<0.05; Fig. 3A and B). Similarly, in the other cell lines, ALDOA knockdown cells formed hardly any spheres compared to controls (P<0.05; Fig. 3B), collectively indicating that ALDOA is necessary for proliferation and tumor formation in hypoxic conditions. In addition, PCR analysis showed that HIF1A expression is upregulated by knockdown of ALDOA, probably because of a negative feedback loop (Fig. 3C).

In DLD-1, HT29 and KM12SM cells, the knockdown of ALDOA resulted in a decrease in lactate production compared to controls, and, in contrast, its overexpression resulted in an increase in lactate production (P<0.05; Fig. 3D). This means that the expression of ALDOA and the activity of the glycolytic pathway were positively correlated.

ALDOA knockdown assay was performed under hypoxia to assess the function of ALDOA in hypoxic conditions. To focus on the relationship between ALDOA and epithelial-mesenchymal transition (EMT), we performed western blot analysis to assess the expression of EMT markers in ALDOA knockdown DLD-1 and HT29 cells cultured under normoxic and hypoxic conditions. We found that E-cadherin expression was reduced by hypoxia and induced by knockdown of ALDOA both in normoxia and hypoxia. As for mesenchymal markers, vimentin expression level was too low to detect on western blots (Fig. 4A). PCR revealed that vimentin mRNA expression was reduced by ALDOA knockdown in both normoxia and hypoxia (P<0.05; Fig. 4B).

The rate of apoptosis was higher in hypoxic DLD-1 and HT29 cells than in normoxic cells. It also showed that the rate of apoptosis was higher in ALDOA knockdown cells than in controls in both hypoxia and normoxia (P<0.05; Fig. 5A). In addition, the ROS-positive cell rate of ALDOA knockdown cells was higher than controls both in hypoxia and in normoxia (P<0.05; Fig. 5B). Cell cycle analysis showed that the G2/G1 ratio was higher in ALDOA knockdown cells than in controls in both hypoxia and normoxia (P<0.05; Fig. 5C).