Abstract
Background/Aim: Gastric cancer (GC) is the third-leading cause of cancer-related deaths worldwide; thus, novel diagnostic and therapeutic biomarkers are needed. Annexin A10 (ANXA10) is a calcium- and phospholipid-binding protein. As far as we are aware, there are no reports describing the detailed functions of ANXA10 in GC. Therefore, we investigated the downstream mRNA variation and the effects of ANXA10 on chemoresistance in GC cell lines. Materials and Methods: ANXA10 knockout GC cell lines were generated, and we performed functional analyses, chemosensitivity drug testing, and microarray analyses. Additionally, immunohistochemistry for ANXA10 was performed on 40 patients with GC who had received 5-fluorouracil (5-FU)-based chemotherapy to compare their prognosis and clinicopathological factors. Results: ANXA10 knockout GC cells showed significantly increased proliferation, invasion, and sensitivity to 5-FU. The overall survival of ANXA10-positive cases was considerably lower than that of ANXA10-negative cases in GC patients who received 5-FU-based chemotherapy. Microarray analysis revealed candidate pathways regulated by ANXA10 and claudin 1 (CLDN1), keratin 80 (KRT80), RANBP2-type and C3HC4-type zinc finger containing 1 (RBCK1), and solute carrier family 7 member 5 (SLC7A5) genes. Conclusion: ANXA10 knockout increased the susceptibility of GC cell lines to 5-FU; ANXA10 may be a predictive indicator for response to 5-FU treatment in GC cases. ANXA10 may be involved in the pathogenesis of GC, in collaboration with CLDN1, KRT80, RBCK1, and SLC7A5.
Gastric cancer (GC) was the third leading cause of cancer-related deaths worldwide in 2020 (1). It is a heterogeneous disease with different phenotypes, morphologies, treatments, and prognoses (2). Multiple genetic and epigenetic alterations have been observed in carcinogenesis and progression of GC (3). Comprehensive next-generation sequence analysis, such as those of The Cancer Genome Atlas Research Network (4) and the Asian Cancer Research Group (5), has brought us much closer to the goal of providing personalised medicine. However, GC remains a complex disease, and to overcome this, it is necessary to elucidate the molecular mechanisms of carcinogenesis and progression.
Chemotherapy, combined with surgical and radiation therapies, is among the most effective treatments. In Asia, adjuvant chemotherapy is based on monotherapy with S-1 (a combination of tegafur, gimeracil, and oteracil). However, combination treatment with capecitabine and oxaliplatin or S-1 and docetaxel (6, 7) have shown that 5-fluorouracil (5-FU)-based chemotherapy provides consistent benefits compared with surgery alone (8). Despite the improvements observed from the use of numerous GC treatments, the 5-year survival rate for patients with advanced GC is approximately 20-30% (9), and this can be mainly attributed to chemotherapy-resistant tumour heterogeneity. Therefore, there is a need to explore biomarkers related to this resistance.
Annexin A10 (ANXA10) is a calcium- and phospholipid-binding protein that belongs to the annexin family. Annexin family proteins have an annexin repeat sequence (10) and play a significant role in cellular physiological processes (11). ANXA10 is expressed in the normal stomach, duodenum, kidney, and urinary bladder, and is a potential diagnostic marker for upper gastrointestinal cancer such as GC, ampullary carcinoma, biliary carcinoma, and pancreatic adenocarcinoma (12). The relationship between ANXA10 and cancer progression has been reported for various organs of the body (13–16). In GC, ANXA10 has been reported to contribute to GC progression through apoptosis and proliferation (17–19). Previously, we demonstrated that ANXA10 reflected prognosis of several types of gastrointestinal cancer: advanced GC (20), early GC, small bowel adenocarcinoma (21), and pancreatic ductal adenocarcinoma (22), and it is known that ANXA10 is correlated with the gastric mucin phenotype in GC (20). However, a comprehensive analysis of mRNA variation downstream of ANXA10 in GC is lacking.
In the present study, we examined the effects of ANXA10 knockout on the function and chemoresistance of GC cell lines. We also investigated the relationship between 5-FU and ANXA10 using clinical specimens from patients with GC who received 5-FU-based chemotherapy. In addition, we performed a microarray analysis to investigate the mRNA variation downstream of ANXA10. Finally, the genes that showed altered expressions after knockout of ANXA10 were examined for their association with prognosis.
Materials and Methods
Tissue samples. Forty primary GC tissue samples were collected from patients with stage II or III GC who underwent surgical resection between January 2012 and December 2014 at the Kure Medical Center and Chugoku Cancer Center (Hiroshima, Japan) and received adjuvant chemotherapy consisting of S-1. All patients uw R0 resection with more than D1 plus or D2 lymph node dissection. Patients were observed by their physicians until their death or the date of the last documented contact. For immunohistochemistry (IHC), the collected tissue samples were archival formalin-fixed and paraffin-embedded. A representative tumour block from each patient was evaluated using IHC. The tumour stage was determined according to the TNM classification system. Histological classifications were determined based on the Japanese Classification of Gastric Carcinoma, 14th edition (23) The Ethical Committee for Human Genome Research of Kure Medical Center and Chugoku Cancer Center approved the present study (2019-91), and all patient samples were obtained with consent.
IHC. IHC staining was performed and evaluated as described previously (12, 20). To detect ANXA10 expression, a rabbit monoclonal antibody to ANXA10 (NBP1-90156, 1:500; Novus Biologicals, Centennial, CO, USA) was used. Two surgical pathologists (A.I. and K.K.) assessed the IHC slides based on the guidelines of the previous reports.
Cell line and culture. A cell line derived from human GC (MKN-74) was purchased from RIKEN Cell Bank (Tsukuba, Japan). The cells were maintained in RPMI-1640 medium (Nacalai Tesque, Kyoto, Japan) supplemented with 10% foetal bovine serum (Corning, NY, USA) in a humidified atmosphere with 5% CO2 at 37°C.
Generation of ANXA10 knockout cells. CRISPR-Cas9 technology with a plasmid vector was used for the ANXA10 knockout cell lines KO-1 and KO-2. The web-based software CRISPRdirect was used to construct the expression plasmids for single guide RNA. Synthesised double-stranded oligonucleotides were inserted into pSpCas9(BB)-2A-Puro (Addgene plasmid 48139: PX459, Addgene, Cambridge, MA, USA). The single guide RNA sequence of the ANXA10-CRISPR vector was KO-1: TGG GAT TGA AAT TGG GAG CT; KO-2: AGC ATT TGG GCA TCC ATT AT. ANXA10 knockout MKN-74 cells were selected by passaging them into a medium containing 10 μg/ml puromycin.
Western blotting. Western blotting of parental and ANXA10 knockout cell lines was performed using a rabbit monoclonal antibody to ANXA10 (NBP1-90156, 1:500; Novus Biologicals) and beta-actin (Santa Cruz Biotechnology, Santa Cruz, CA, USA). The total amount of protein in each sample was measured using a Bicinchoninic Acid™ Protein Assay Kit (Thermo Fisher Scientific, Inc., Waltham, MA, USA). The protein concentration was adjusted to the same quantity for all samples. For the analysis, 30 μg of protein was loaded. The samples were boiled for 5 min with 50 mM 1,4-dithiothreitol and 0.025% bromophenol blue. The proteins were then separated via 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene difluoride membranes. The membranes were blocked with 5% (w/v) skim milk for 30 min. The membranes were then washed and incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature. Chemiluminescent detection was performed using ECL™ Prime Western Blotting Detection Reagent (Bio-Rad, Tokyo, Japan).
Proliferation assay. Proliferation of parental and ANXA10 knockout cell lines was assessed using Cell Counting Kit-8 (CCK-8; Dojindo, Kumamoto, Japan). Cells were plated at a density of 5,000 cells/well in a 96-well plate, and cell growth was monitored on days 0, 1, 2, and 4.
Wound-healing assay. Wound-healing assays were performed using the CytoSelect 24-well wound healing assay (Cell Biolabs, Inc., San Diego, CA, USA). Parental and ANXA10 knockout cells were plated onto fibronectin-coated 24-well plates and incubated at 37°C for 48 h. A clear area was then scraped using a plastic tip. The migration of cells into wounded areas was evaluated using an inverted microscope 2 days later.
Drug sensitivity test. Oxaliplatin was purchased from Yakult (Tokyo, Japan); irinotecan was purchased from Towa (Osaka, Japan); and 5-FU was purchased from Towa (Osaka, Japan). Cells were plated at a density of 5,000 cells/well in a 96-well plate. After 24 h, cells were untreated (negative control) or treated with escalating doses of drugs to assess cell viability. The CCK-8 assay was performed 72 h after chemotherapy.
RNA isolation and microarray analysis. Total RNA was extracted from cell lines using Isogen (Nippon Gene, Tokyo, Japan), and the quality and quantity of the extracted RNA samples were evaluated using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Inc.). A total of six RNA samples were sent to KURABO industries ltd. (Osaka, Japan) to conduct certain technical procedures, including quality control. The samples were applied to a Clariom S Human Array (Affymetrix, Santa Clara, CA, USA). The expression value for each gene was determined by calculating the average of the differences in intensity between the probe pairs and was stored in CEL data files. iDEP (http://bioinformatics.sdstate.edu/idep/) was used to normalise and filter the gene-expression profile, Gene Ontology analysis (24), Kyoto Encyclopedia of Genes and Genomes (25) analysis and parametric gene set enrichment analysis (Broad Institute, Cambridge, MA, USA) (26). Probe sets with a fold change of ≥2 or <0.5 with adjusted values of p≤0.05 were selected, and p-values were adjusted for multiple comparisons using the false-discovery rate. Probe sets with more than a two-fold change in expression with false-discovery rate-adjusted values of p≤0.05 were considered statistically significant.
Kaplan–Meier analysis. Kaplan–Meier analysis was performed for ANXA10 using KMplot software from a database of public microarray datasets (http://kmplot.com/analysis) (27). To analyse the prognostic value of the probe, samples were split into two groups based on the cut-off value stipulated by the software program. Hazard ratios and log-rank p-values were derived for each survival analysis.
Statistical analysis. Statistical significance was set at p<0.05. Fisher’s exact test was used to examine the correlation between clinicopathological parameters and ANXA10 expression. To test the statistical differences between the survival curves, the log-rank test was employed. Differences between groups in the CCK-8 assay, wound-healing assay, and invasion assays were evaluated using Student’s t-test. These assays involved three independent experiments, and the mean±standard deviation was calculated for each of these experiments. GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA) was used for all analyses.
Results
Effect of ANXA10 knockout on GC cells. ANXA10 knockout cell lines was generated to identify the genes altered by ANXA10 knockout. In our previous study, we found that ANXA10 expression was high in the MKN-74 cell line (20). ANXA10 knockout was confirmed by a western blot analysis (Figure 1A). To further investigate the effect of ANXA10 knockout, a CCK-8 assay was performed, and MKN-74 cells with ANXA10 knockout showed significantly increased cell growth relative to parental MKN-74 cells (Figure 1B). The migration activity of ANXA10 knockout cells was then examined using a wound-healing assay, and the wound-healing distance of ANXA10 knockout MKN-74 cells was found to be significantly reduced compared to that of the negative control MKN-74 cells (Figure 1C). A chemosensitivity assay was then conducted to investigate the effect of ANXA10 knockout on chemosensitivity, and a CCK-8 assay was used to measure the cell viability of the ANXA10 knockout and negative control MKN-74 cells under different concentrations of chemotherapeutics: oxaliplatin (Figure 1D), 5-FU (Figure 1E), and irinotecan (Figure 1F). The viability of MKN-74 ANXA10 knockout cells was found to be significantly lower than that of the negative control MKN-74 cells when cells were exposed to 5-FU (Figure 1E). These results suggested that ANXA10 knockout enhanced the proliferative and invasive ability of cells and enhanced their sensitivity to 5-FU.
Effects of the knockout of annexin A10 (ANXA10) on function and chemosensitivity in gastric cancer cells. A: Western blot analysis of ANXA10 in ANXA10 knockout (KO-1/KO-2) and parental MKN-74 cell lines. B: Effect of ANXA10 knockout on MKN-74 cell growth. Cell growth was evaluated by CCK-8 assay. Data are the mean±standard deviation of three independent experiments. Chemosensitivity assays using CCK-8 kit were performed to investigate the inhibitory effect of oxaliplatin (L-OHP) (C), 5-FU (D), and irinotecan (E). NC: Negative control. *Significantly different at p<0.05.
Relationship between ANXA10 and 5-FU-based chemotherapy in patients with GC. To confirm sensitivity to 5-FU observed in the GC cell line, we recruited 40 patients diagnosed with GC who had received 5-FU-based therapy. As reported previously in the literature (20, 28), we performed ANXA10 immunohistochemical staining, and ANXA10 expression was found in the nuclei of tumour cells (Figure 2A). We thus investigated the relationship between ANXA10 staining and the clinicopathological features (Table I). We then conducted Fisher’s exact test, which revealed no significant difference in any factors, including age (p=0.101), sex (p=0.281), pT stage (p=0.640), pStage (p=0.089), vascular invasion (p=0.501), lymphatic invasion (p=1.000), and histology (p=0.601) according to ANXA10 expression, except for pN stage (p<0.01).
Immunohistochemical and Kaplan–Meier analysis of annexin A10 (ANXA10) in patients with gastric cancer (GC) who received 5-fluorouracil-based chemotherapy. A: Representative image of ANXA10 staining in GC. Original magnification: 100× (left). B: Overall survival probability of patients with GC according to ANXA10 expression from a public database. C: Probability of overall survival of 40 GC cases based on immunohistochemical expression.
Relationship between annexin A10 expression and clinicopathological features in patients with gastric cancer who received 5-fluorouracil-based chemotherapy.
We also examined the relationship between ANXA10 expression and the prognosis of patients with GC who received 5-FU-based therapy using a public database. We found that there was no significant trend in overall survival when patients with GC were stratified by the expression of ANXA10 (p=0.39; Figure 2B). However, ANXA10-positive GC cases had a significantly lower survival probability than ANXA10-negative cases (p=0.011; Figure 2C). Therefore, in patients with GC who had received 5-FU-based chemotherapy, high expression of ANXA10 was associated with a worse prognosis.
Overview of genes differentially expressed after ANXA10 knockout. To analyse ANXA10 knockout, the gene-expression profile of parental MKN-74 cells was compared to that of ANXA10 knockout MKN-74 cells. Box and density plots show the normalization results of the microarray data for a total of six samples (Figure 3A and B), and the principal component analysis plot is shown in Figure 3C, where the volcano plot shows the expression distribution of mRNAs (Figure 3C).
Gene-expression profiles of annexin A10 (ANXA10) knockout (KO) and parental cell lines. A: Box plot of ANXA10 expression after normalisation, which shows the maximum, upper quartile, median, lower quartile, and minimum of different samples. B: Density plot of expression after normalization, which shows the probability distribution at different ANXA10 mRNA expression levels. C: Principal component (PC) analysis plot of each sample. D: Analysis of ANXA10 expression in each sample using a volcano plot. E: Heat map of 100 genes with differential expression between control and knockout cells. FDR: False-discovery rate. Ctrl: Control.
A total of 530 differentially expressed genes featured a log2 fold-change of more than +0.5 or less than −0.5. The heatmap demonstrated that 379 genes were up-regulated, and 151 genes were down-regulated in response to ANXA10 knockout in MKN-74 cells (Figure 3D and E).
Top genes differentially expressed after ANXA10 knockout and enrichment analysis. To determine the biological functions of the differentially expressed genes, GO analyses were performed. Significant enrichment of the top highest gene counts were found in the molecular function of Gene Ontology analysis (such as “Cadherin binding”, “Enzyme binding”, and “Cell adhesion molecule binding”) (Figure 4A) and in the Kyoto Encyclopedia of Genes and Genomes analysis (“MAPK signalling pathway”, “Proteoglycans in cancer”, and “Focal adhesion”) (Figure 4B). The top 10 most significant differentially expressed genes are listed in Table II. We then evaluated the prognostic effects of representative genes in patients with GC. The Kaplan–Meier curves demonstrated that high expression of claudin 1 (CLDN1) (Figure 5A), keratin 80 (KRT80) (Figure 5B), RANBP2-type and C3HC4-type zinc finger containing 1 (RBCK1) (Figure 5C), and solute carrier family 7 member 5 (SLCA5) (Figure 5C) mRNA were associated with a poor prognosis in patients with GC.
Functional and signalling pathway analyses of differentially expressed genes: Molecular function (A) and Kyoto Encyclopedia of Genes and Genomes (B) analysis.
The top 10 differentially expressed genes identified from microarray analysis between parental cells and annexin A10 knockout cells.
Analysis of survival according to differentially expressed genes in a database of public microarray datasets. Prognostic value of claudin 1 (CLDN1) (218182_s_at) (A), keratin 80 (KRT80) (231849_at) (B), RANBP2-type and C3HC4-type zinc finger containing 1 (RBCK1) (207713_s_at) (C) and solute carrier family 7 member 5 (SLC7A5) (201195_s_at) (D).
Discussion
In this study, we investigated the effect of ANXA10 knockout in GC cell lines, and the mRNA variation downstream of ANXA10 was comprehensively analysed. To the best of our knowledge, this is the first study to show that ANXA10 is correlated with resistance to 5-FU.
We first examined the function of ANXA10 using ANXA10 knockout GC cell lines and found that ANXA10 knockout increased the susceptibility to 5-FU. Previous reports have found that low expression of ANXA10 in cell lines promoted proliferative (19, 20) and invasive capacities (17), and these findings were confirmed in this study. However, the relationship between ANXA10 and chemoresistance had not been previously investigated.
Oxaliplatin, irinotecan, and 5-FU, as well as S-1 (29) and capecitabine, are the most commonly used drugs in chemotherapy for GC (6, 8, 30). Among these, 5-FU is converted into an active metabolite, fluorodeoxyuridine monophosphate, which acts as an inhibitor of DNA synthesis (31). In our pathway analysis, DNA-binding transcription activator activity was found to be increased in ANXA10 knockout cell lines. These findings suggest that sensitivity to 5-FU may have increased because molecules downstream of ANXA10 are involved in nucleic acid synthesis.
Through a comparative analysis of microarrays for ANXA10 knockout GC cell lines, we found that ANXA10 is associated with the mitogen-activated protein kinase (MAPK) and epidermal growth factor receptor (ERBB) signalling pathways. Previous reports have found that ANXA10 is involved in chromosomal deletion (17) and reduces apoptosis (19) to promote tumour progression in GC. Although reports have revealed that ANXA10 enhances the MAPK signalling pathway in oesophageal (13) and oral cancer (14), the significance of this finding here may differ because this pathway was enhanced by ANXA10 knockout. A past pathway analysis using a microarray analysis of ANXA10 knockdown in cholangiocarcinoma cell lines found that pathways such as the phospholipase D signalling pathway, vascular smooth muscle contraction, and carbon metabolism were regulated (32). These findings suggest that ANXA10 has an organ-specific function and may have different functions in GC that depend on the phenotype.
A comparative gene-expression analysis of ANXA10 knockout cell lines identified CLDN1, KRT80, RBCK1, and SLC7A5, which showed significant differences in prognosis in public databases. CLDN1 is a claudin protein that promotes GC progression (33, 34), and KRT80 encodes keratin 80, which is an epithelial keratin. In GC, KRT80 is activated by hsa_circ_0014130 (circPIP5K1A), and a high expression of KRT80 promotes tumor (35), but its relationship with prognosis has not been investigated. In renal cell carcinoma, the high expression of RANBP2-type and C3HC4-type zinc finger containing 1 (RBCK1) is associated with poor outcomes (36). In other carcinomas, the role of RBCK1 is unknown. SLC7A5 is classified as a transporter of large neutral amino acids (for example, as an L-type amino acid transporter 1, and its effect has already been reported in numerous cancer types (37). However, as the relationship between these molecules and ANXA10 has not been investigated, its investigation may lead to the elucidation of the pathogenesis of GC in the future.
In the present study, we investigated the relationship between ANXA10 and sensitivity to chemotherapeutic agents using samples of GC resected from patients who had undergone 5-FU-based chemotherapy. We found that a high ANXA10 expression was associated with a significantly poorer prognosis. The expression of ANXA10 is a biomarker of a good (20, 28, 38) or poor (13, 16, 21, 39, 40) prognosis, depending on the carcinoma type. In GC, prognosis differs depending on the morphological phenotype: a diffuse type has a good prognosis, while an intestinal type has a poor prognosis (18). We demonstrated that in patients overall, a low ANXA10 level was a marker of poor prognosis (20, 28). However, the prognosis was reversed for patients with GC who had received 5-FU-based chemotherapy. Thymidylate synthase has been reported as a biomarker for the efficacy of 5-FU (41). ANXA10 might similarly be an additional, new predictive biomarker for 5-FU chemotherapy. To validate these results, prospective studies using a larger number of patients are required.
In conclusion, we demonstrated that ANXA10 knockout increased sensitivity to 5-FU in the GC cell line. In patients with GC who received postoperative 5-FU-based chemotherapy, ANXA10-negative cases had a significantly better prognosis than ANXA10-positive cases. In addition, a microarray analysis using ANXA10 knockout GC cell lines revealed that ANXA10 identified candidate pathways regulated by ANXA10. We also identified certain genes (CLDN1, KRT80, RBCK1, and SLC7A5) that might be involved in the pathogenesis of GC in collaboration with ANXA10.
Acknowledgements
The Authors would like to express their sincere gratitude to N. Yasumura, A. Kan, H. Fujisawa, K. Iwahiro and F. Kimura for their excellent technical assistance and support. We thank all the staff at the Institute for Clinical Research, National Hospital Organization Kure Medical Center/Chugoku Cancer Center for their willingness to use their facilities.
This study was supported by a KAKENHI Grant in Aid for Early-Career Scientists (grant Number: JP21K15394) from the Japan Society for the Promotion of Science.
Footnotes
Authors’ Contributions
A. Ishikawa designed the study. T. Kuwai, H. Tazawa, T. Suzuki, and H. Tashiro collected and analysed the patient clinical data. A. Ishikawa, K. Kuraoka, J. Zaitsu, A. Saito, and K. Taniyama performed the experiments and collected, and analysed the data. A. Ishikawa, K. Kuraoka, T. Kuwai, T. Suzuki, H. Tashiro, and W. Yasui interpreted and analysed the results. A. Ishikawa, K. Kuraoka, and W. Yasui drafted and edited the article. All Authors read and approved the article and agreed to be accountable for all aspects of the research to ensure that the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Conflicts of interest
The Authors declare no conflicts of interest.
- Received February 5, 2022.
- Revision received February 25, 2022.
- Accepted March 2, 2022.
- Copyright © 2022 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.