Aldolase A as a prognostic factor and mediator of progression via inducing epithelial–mesenchymal transition in gastric cancer

Abstract Glycolysis is regarded as the hallmark of cancer development and progression, which involves a multistep enzymatic reaction. This study aimed to explore the clinicopathological significance and potential role of glycolytic enzyme aldolase A (ALDOA) in the carcinogenesis and progression of gastric cancer (GC). ALDOA was screened from three paired liver metastasis tissues and primary GC tissues and further explored with clinical samples and in vitro studies. The ALDOA protein level significantly correlated with a larger tumor diameter (P = .004), advanced T stage (P < .001), N stage (P < .001) and lymphovascular invasion (P = .001). Moreover, the expression of ALDOA was an independent prognostic factor for the 5‐year overall survival and disease‐free survival of patients with GC in both univariate and multivariate survival analyses (P < .05). Silencing the expression of ALDOA in GC cell lines significantly impaired cell growth, proliferation and invasion ability (P < .05). Knockdown of the expression of ALDOA reversed the epithelial–mesenchymal transition process. Mechanically, ALDOA could affect the hypoxia‐inducible factor (HIF)‐1α activity as demonstrated by the HIF‐1α response element–luciferase activity in GC cells. Collectively, this study revealed that ALDOA was a potential biomarker of GC prognosis and was important in the carcinogenesis and progression of human GC.

help explain the biochemical mechanism underlying the processes of metastasis and progression as well as therapeutic sensitivity.
Tumour cells proliferate at a rapid rate and are always in a hypoxic microenvironment. Cancer cells develop adaptive responses to hypoxia by activating the hypoxia-inducible genes to overcome this harsh environment. [8][9][10] Under this circumstance, hypoxia is a major characteristic of cancer; in hypoxic conditions, cancer cells develop metabolism reprogram from oxidative phosphorylation to glycolysis. 11,12 A feature of this phenomenon is increased lactate production even at normal oxygen concentrations. This phenomenon is named as the Warburg effect or aerobic glycolysis. The Warburg effect was considered to be at the root of carcinogenesis and progression. 3,13,14 It not only provides cancer cells with adenosine triphosphate (ATP) and nutrients but also leads to an acidic environment that facilitates metastasis. 3 Glycolysis represents a 10-step metabolic process that involves multiple enzymes. 15 Studies have shown that some glycolytic enzymes are more complicated and multifaceted proteins than simple components of the glycolytic enzymes. 16 These glycolytic enzymes have acquired additional nonglycolytic functions in many aspects. 17 In this study, quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) arrays of glycolysis-related enzymes were performed in three paired liver metastasis tissues and primary GC tissues. Aldolase A (ALDOA) was chosen for further study. This study mainly investigated the expression of ALDOA in GC tissues and its clinical significance and biological function in GC cells and also elucidated the mechanisms underlying the regulatory role of ALDOA in GC cells.

| Patients and samples
A series of 252 patients with pathologically confirmed GC was used for the immunohistochemical (IHC) study to explore the clinical significance of the expression of ALDOA in GC. All patients who had no neoadjuvant chemotherapy and received radical gastrectomy were randomly selected from samples collected between 2008 and 2012. The median follow-up time was 46 months. The characteristics of the samples are shown in Table 1

| Cell lines
Human gastric adenocarcinoma cell lines, AGS and MGC-803, were originally obtained from the Institute of Biochemistry and Cell Biology at the Chinese Academy of Sciences (Shanghai, China). The cells were cultured in the Dulbecco's modified Eagle's medium containing 10% foetal bovine serum, 100 U/mL penicillin and 100 lg/mL streptomycin in a 37°C incubator supplied with 5% CO 2 .

| qPCR
The total RNA was extracted from cell lines using the TRIzol

| IHC staining
Immunohistochemical staining was performed for ALDOA. Briefly, paraffin sections were baked for 60 minutes at 70°C, deparaffinized in xylene and rehydrated in gradually varied alcohol. Then, the sections were managed with 3% H 2 O 2 to neutralize endogenous peroxidase for 30 minutes. The antigen retrieval was processed with citrate buffer (pH = 6.0) in a pressure cooker. After antigen retrieval, the sections were incubated with ALDOA primary antibody (11217-1-AP; 1:100; Proteintech) and secondary antibody. The sections were then stained with 3,3-diaminobenzidine, terminated in phosphate-buffered saline (PBS), and counterstained with haematoxylin. Based on the staining intensity of ALDOA in each case, the grading was as follows: 0, negative; 1, weak; 2, moderate; and 3, strong. The scores 0 and 1 were regarded as low expression and 2 and 3 as high expression. Two observers graded the score of staining intensity independently.

| Stable transfection of GC cells
Biologically active short hairpin RNAs (shRNA) were generated using the lentiviral expression vector pLKO.1-puro. The shRNA target sequence for human ALDOA was 5 0 -CCATGCTTGCACTCAGAAGTT-

| Cell Counting Kit-8 assays
The

| Colony formation assays
The cells were trypsinized and plated on six-well plates (200 cells/ well) and cultured for 2 weeks. The colonies were stained with 1% crystal violet for 30 seconds after fixation with 4% paraformaldehyde for 5 minutes. The colonies were counted and defined as >50 cells/colony. Three independent experiments were performed.
The data were calculated using paired t test.

| Transwell migration
Cell invasion assays were performed using 6.5-mm Transwell chambers (8-lm pore size, BD). The cells were seeded at a density of 50 000 cells/well into Transwell chambers coated with Matrigel for assays. The wells were washed with PBS after 36 hours of seeding.
F I G U R E 1 Heatmap of differentially expressed genes in three paired primary gastric cancer tissues and liver metastasis using qRT-PCR. The results were normalized by Z-score. Each column represented a specimen (denoted at the top), and each row represented a gene (denoted to the right). Red colour indicates genes that were up-regulated, and green colour indicates genes that were down-regulated. P represents primary gastric cancer tissue, and M represents corresponding liver metastasis The cells that had migrated to the basal side of the membrane were fixed and stained with crystal violet, visualized and photographed using a CKX41 microscope (Olympus, Japan) at 2009 magnification.
Images of three random fields from three replicate wells were obtained, and the cells that had migrated were counted.

| Statistical analysis
Comparisons between groups for statistical significance were performed with a two-tailed paired Student's t test. The relationship between the expression of ALDOA and clinicopathological

| Glycolytic enzyme ALDOA might promote tumor metastasis in GC
Glycolysis is a process involving a series of glycolytic enzymes. The qRT-PCR assay was performed and the transcriptional expression levels of glycolytic enzymes in three paired GC tissues, and their paired synchronous liver metastasis tissues were measured to identify the glycolytic enzymes involved in gastric cancer metastasis. All patients underwent synchronous primary tumour and liver metastases resection. The results indicated that the up-regulated expression of ALDOA and down-regulated FBP1 were most significant in liver metastases compared with its primary tumour ( Figure 1). As FBP1 has been studied previously, 18 ALDOA was chosen for further study.

| ALDOA was overexpressed in GC tissues
Quantitative reverse transcriptase-polymerase chain reaction and IHC staining were performed on 30 pairs of GCs (T) and matched adjacent gastric tissues (N) to explore whether ALDOA was dysregulated in human GC. As shown in Figure 2A, with their matched adjacent cervical tissues. As expected, the protein expression of ALDOA was also markedly up-regulated in the GC tissues as judged by IHC ( Figure 2B,C).

| Expression of ALDOA correlated with inferior clinicopathological parameters
Immunohistochemistry

| Expression of ALDOA significantly correlated with survival outcomes in GC
This study further evaluated the relationship between the expression of ALDOA protein and the 5-year overall survival (OS) or

| ALDOA affected hypoxia-inducible factor-1a activity in GC
Hypoxia-inducible factor (HIF)-1a is critical for tumour growth and metastases, in part by inducing glycolytic enzymes and EMT process. Glycolysis is also necessary for maintaining HIF-1a activity.
Therefore, it is possible that ALDOA can affect the HIF-1a activity in GC. HIF-1a acts by binding to the HIF-1a response element (HRE) upon hypoxia. Therefore, this study used HRE-luciferase reporter to examine whether ALDOA affected HRE-luciferase activities. As expected, ALDOA could increase HRE-luciferase activity in a dose-dependent manner ( Figure 6A). Moreover, HRE activity could be inhibited by reducing the expression of ALDOA using its specific shRNA ( Figure 6B), implying that ALDOA indeed affected the HIF-1a activity.

| DISCUSSION
Great attention has been paid to glycolytic enzymes as potential therapeutic targets because the cancer-specific metabolism depends more on the glycolytic pathway than on aerobic respiration, a phenomenon called the Warburg effect. 8 This is a novel study that characterized the importance of ALDOA in GC. First, ALDOA was screened out by performing qRT-PCR arrays of a panel of glycolysis-related genes in three paired liver metastasis tissues and primary GC tissues. Then, the clinical significance and potential mechanisms by which ALDOA mediated GC metastases and progression were explored in detail. In fact, ALDOA has been validated as an oncogene in some solid tumours. In the pancreas, it promotes tumorigenesis and progression into a highly metastatic pancreatic cancer by regulating the expression of E-cadherin. 19 It regulates the cell cycle in non-small cell lung cancer and several solid tumours. 20 One of the most important findings in the present study was a direct relationship between the expression of ALDOA and the inferior clinicopathological features, and the high expression of ALDOA was an independent prognostic factor for both OS and DFS. The functional study indicated that silencing the expression of ALDOA impaired the proliferation and invasion abilities of GC cells. As EMT is the initial step of cancer metastases, the relationship between the expression of ALDOA and EMT-related marker was further validated, which suggested that ALDOA might promote tumour metastases by inducing the EMT process.
HIF-1a is critical for cancer cell survival and metastases in the solid tumour under stressed hypoxic tumour environment, in part by inducing glycolytic enzymes and EMT process. [21][22][23] It could up-regulate a series of glycolytic genes during anaerobic glycolysis and then increase the glycolysis rate and ATP production. A previous study also suggested that glycolysis was necessary for maintaining HIF-1a activity. 24 In the present study, the glycolytic enzyme ALDOA was also found to increase HIF-1a activity. Glycolysis and HIF-1a formed a feed-forward loop that stimulated tumour growth and metastases.
F I G U R E 6 Aldolase A (ALDOA) affected the HIF-1a activity in gastric cancer. A, It could increase the HRE-luciferase activity in a dose-dependent manner (P < .05). B, Silencing the expression of ALDOA decreased the HRE activity (P < .05) When silencing the expression of ALDOA, the anaerobic glycolysis was inhibited, ATP levels were reduced, the feed-forward loop was broken and tumour proliferation and metastases were inhibited. Conversely, the high expression of ALDOA could increase the HIF-1a activity, and HIF-1a could, in turn, promote glycolysis and EMT process. Hence, ALDOA may serve as a promising biomarker and target therapy for GC.
Several studies focused on the relationship between EMT and glycolysis. A previous study demonstrated that FBP1 was a negative regulator of EMT in GC. 18 Loss of FBP1 by Snail-mediated repression provided metabolic advantages in basal-like breast cancer. 25 The ectopic expression of aldolase B was associated with poor prognosis and promotes tumour progression by EMT in colorectal adenocarcinoma. 17 Some oncogene and tumour suppressors had a crosstalk between EMT and glycolysis in cancer progression and metastases, such as HIF-1a, 23 c-myc, 26 FOXM1, 27 Gas1, 28 and so on. The present study suggested a novel function of ALDOA other than a glycolytic enzyme.
Collectively, this study provided firm evidence that ALDOA was crucial in GC by inducing the EMT pathway and affecting HIF-1a activity in tumour progression and metastasis. The expression level of ALDOA was an independent adverse prognostic factor for both OS and DFS, providing additional information for guiding therapeutic strategies.

CONF LICTS OF INTEREST
None.