Introduction
Gastric cancer is the fifth most common cancer and the fourth cancer death cause globally, with over 1 million new cases diagnosed and about 777,000 deaths in 2020 [
1]. In China, the number of newly diagnosed with gastric cancer (GC) patients is nearly 500,000, and the morbidity and mortality of gastric cancer occupies the third position in malignant tumors [
2‐
3].
Over the past decade, some advancements have been made in the immunotherapy of gastric cancer, but the low immunogenicity of tumors, tumor-induced immunosuppression, drug resistance, and toxic effects all limited the application of immunotherapy in clinical practice [
4‐
5]. Therefore, it is important to identify more biomarkers and therapeutic targets that can be used as clinical markers for GC clinical diagnosis, prognosis, chemotherapy, and tumor immunotherapy.
APOBEC2 is the second member of APOBEC family, and the protein family is known to mediate editing of the C (cytidine)→U (uridine) by cytidine deaminating activity and contributes to the mutagenesis of genomic DNA [
6]. Recent observations have also linked the expression of APOBEC2 to tumors and indicated it in the regulation of tumor-related genes, including Eif4g2 and PTEN [
7‐
9]. Kostic et al. reported that the APOBEC2 gene is a target gene of TP53, while TP53 frequently mutates in GC [
10‐
11]. Based on genomic data of The Cancer Genome Atlas (TCGA), Shi et al. found that APOBEC2 transcriptional level was significantly correlated with the tumor mutational burden (TMB) in stomach adenocarcinoma (STAD) [
12]. In addition, APOBECs have been known to induce immunogenicity by mediating hypermutation and enhance immunological-related markers such as immune cell infiltration [
13‐
15]. However, the significance of APOBEC2 protein expression in GC, especially in its association with tumor-infiltrating immune cells is unknown. CD66b as a phenotype of high-density or low-density neutrophils, which has been found to expressed in human tumor-associated neutrophils [
16].
In this study, we investigated the APOBEC2 gene mutation in STAD in the publicly available clinicogenomic data on cBioportal. We also explored the transcription level of APOBEC2 in STAD and normal tissues from the GEPIA database. Western blot and qRT-PCR were used to detect APOBEC2 expression in STAD and normal tissues. Finally, we used immunohistochemistry to assess the expression of APOBEC2, CD66b and CD163 in tissue microarrays (TMAs) from 496 patients with STAD. The relationship of APOBEC2 with CD66b, CD163, and other clinicopathological characteristics was investigated. The associations between APOBEC2 expression and patient survival outcome were further analyzed.
Materials and methods
Identifying APOBEC2 genetic alteration in cBioportal
An open-access cancer genomics platform, the cBio Cancer Genomics Portal (
http://www.cbioportal.org/), gathers multidimensional cancer genomics data sets from the Gene Expression Omnibus (GEO), the International Cancer Genome Consortium (iCGC), TCGA and other databases [
17]. The cBio Cancer Genomics Portal empowers researchers to explore genomic data such as DNA copy number alterations (CNAs), DNA methylation and somatic mutations in an intuitive and rapid way. There are seven gastric adenocarcinoma-related studies including 1,512 samples at the time of this writing. Since some patients contributed multiple samples to these studies, we excluded 735 overlapping cases and explored DNA CNAs of APOBEC2 gene in the remaining 777 cases.
Exploring APOBEC2 gene expression in GEPIA
GEPIA (
http://gepia.cancerpku.cn/) is a publicly accessible website that offers visualization of gene differential expression, patient survival analysis, profiling plotting, and relevant genomic information based on Genotype Tissue Expression (GTEx) and TCGA data [
18]. We analyzed the expression of APOBEC2 gene in gastric adenocarcinoma (n = 408) and adjacent tissues (n = 211) using the function of box plots in the GEPIA database.
Detecting of APOBEC2 protein in cancer and paracancerous tissue microarrays
The medical records of patients with gastric cancer who received gastrectomy at the First Affiliated Hospital of the Second Military Medical University from December 2006 to July 2011 were evaluated [
19]. Patients who received preoperative cancer treatment and were diagnosed with autoimmune diseases or other malignant tumors were excluded. Finally, 496 Patients with gastric adenocarcinoma were enrolled in this study, and their formalin-fixed paraffin-embedded (FFPE) tissue blocks were collected and made into TMAs. The TMAs contained 496 cancer tissues and 48 non-malignant tissues (constructed by Outdo Biotech, shanghai, China). The hematoxylin-eosin (H&E) slides of 496 patients were also collected. The clinical and pathological information of 496 patients, including age, gender, serum CA199 (carbohydrate antigen 199) level, serum CEA (carcinoembryonic antigen) level, tumor size, differentiation, WHO histological classification, Lauren’s histological type, and TNM stage, was gathered from the electronic medical record and histologic sections (Table
1). The tumors were classified as follows according to the 7th edition of the American Joint Committee on Cancer Staging Manual: Stage I (n = 160), Stage II (n = 134), Stage III (n = 195), and Stage IV (n = 7). Of these, 339 patients received postoperative chemotherapy. In this research, the median follow-up duration was 60 (1-101) months. Overall survival (OS) was defined as the date of surgery to the last follow-up date or death, and disease-free survival (DFS) was defined as the date of surgery to the time of recurrence or metastasis.
Immunohistochemical staining and scoring
The antibodies we used as the following: APOBEC2 (dilution 1:100, rabbit anti-human APOBEC2 monoclonal antibody, ab170859, Abcam), CD66b (dilution 1:400, no.555,723, BD Biosciences, NJ), CD163 (ready-to-use, MAB-0206, Maxim, Fuzhou, Republic of China), secondary antibodies (MaxvisionTM2 HRP-polymer anti-mouse/rabbit IHC kit, Fuzhou, China).
The immunohistochemical EnVison two-step method was applied to detect the expression of APOBEC2. Details of Immunohistochemical APOBEC2 staining were presented in Fig.
S1 (see Appendix). The positive expression of APOBEC2 was localized to cytoplasm. The staining intensity was scored as follows: 0 point for negative staining, 1 point for light yellow, 2 points for moderate yellow, 3 points for brown. If the percentage of cells is more than 75%, it scores 3 points; 51–75%, 2 points; 10–50%, 1 point; and below 10%, nil. An immunohistochemical score was calculated as the score for staining intensity plus the score for percentage of cells. The final score of IHC greater than 1 was regarded as positive, and a score equal to or less than 1 was regarded as negative. Immunohistochemistry protocols for CD66b and CD163, and quantitative evaluation of immunostaining, were presented in a previous study [
19]. For survival and correlation analysis, patients were divided into CD66b
low, CD66b
high, CD163
low and CD163
high groups, using the median value as a cut-off.
Western bolt and real-time quantitative reverse transcription PCR (qRT-PCR)
Homogenates of skeletal muscle, STAD and adjacent normal tissues were diluted in 5×loading buffer and boiled for 5 min. 20 µg/lane protein samples were electro-transferred onto polyvinylidene fluoride after being resolved by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The primary antibodies used as the following: APOBEC2 (dilution 1:400, rabbit anti-human APOBEC2 monoclonal antibody, ab170859, Abcam) and α-Tubulin (dilution 1:10000, mouse monoclonal antibody, ab7291, Abcam).
Total RNA was extracted from cell (gastric cancer cell line HGC27 and AG5), STAD and normal gastric tissues using Trizol (Takara, Japan) and the concentration was determined by Nanodrop. RNA was reverse transcribed into cDNA using PrimeScript RT kit with gDNA Eraser (Takara, Japan) at room temperature for 5 minutes, 37˚C for 30 minutes and 85˚C for 5 seconds. Using Applied Biosystems 7500 Real Time PCR system, real-time PCR was conducted as follows: 95˚C for 2 minutes; 40 cycles of 95˚C for 10 seconds and 60˚C for 30 seconds; 95˚C for 15 seconds; 60 ˚C for 1minutes; 95˚C for 15 seconds at the last step. The primer sequences we used were as follows: APOBEC2, forward 5’-CCTCTCTCCTCTCCCTCAGT-3’, reverse 5’-TTTCAGCTTCTCAGGGTCGT-3’; CYCLOPHILIN, forward 5’-TGGTGTTTGGCAAAGTGAAA-3’, reverse 5’-TCGAGTTGTCCACAGTCAGC-3’ (Sangon Biotech, China).
Statistical analysis
To compare the APOBEC2 expression between cancer samples and adjacent gastric samples, Kruskal–Wallis test or paired t-test was performed with GraphPad Prism 8.2. By using SPSS21.0 software (IBM SPSS Statistics for window, Version 21.0), the Mann-Whitney U test and Pearson Chi-square test were used to analyze the association between APOBEC2 expression and clinicopathological variables of patients. The Kaplan-Meier curves and Log-rank test was performed to analyzed differences in survival outcomes using with GraphPad Prism 8.2. To identify prognostic factors of OS and DFS, the univariate and multivariate Cox proportional hazards regression models were applied using SPSS21.0 software. P value < 0.05 was regarded as statistical significance.
Discussion
In this study, we noticed the frequency of APOBEC2 gene alteration in stomach adenocarcinoma using the database from cBioportal wanderer platform. It showed that the proportion of APOBEC2 gene alteration was about 5%, among these only 1 of 89 Asian patients existed APOBEC2 gene alteration. Similarly, a few previous studies have reported the APOBEC2 gene is conserved [
20]. These findings seemed to be consistent with our observation. Furthermore, we performed the analysis with cancer tissues in GEPIA database and revealed APOBEC2 gene expression is commonly down regulated in cancer tissues compared with normal tissues, suggesting APOBEC2 may be associated with the development of stomach adenocarcinoma.
Next, we investigated the expression of APOBEC2 protein in gastric cancer tissue and non-malignant tissues. We noticed that the obvious difference of distribution pattern of APOBEC2 between in gastric cancer tissue and non-malignant tissues. Notably, APOBEC2 was positively stained in the fundic gland, whereas stain of pyloric gland was negative (Fig.
2). Our proteomics and microarray data were consistent with the finding that the fundus and body of stomach had a certain level of APOBEC2 mRNA [
21]. Gastric fundus glands, mostly composed of parietal cells that generate hydrochloric acid and chief cells that secrete pepsinogen, are distributed at the gastric body and fundus. These products of the gastric mucosa such as stomach acid and pepsinogen have been confirmed to have important functions in screening, diagnosis and prognosis of gastric disease. To prove that the expression of APOBEC2 in gastric fundus glands might play an important role in the physiological functions of stomach. As for the result of western bolt, we can’t find the significant APOBEC2 protein expression in fresh gastric tissues, even we can show the APOBEC2 expression samples in striated muscle tissues, it hints us that APOBEC2 protein expression levels are relatively low in clinic samples, especially considering the detected mixture samples including the uncertain exact quantity and ratios cancer tissues and the non-cancer tissues. To research the expression of APOBEC2 protein in STAD and normal gastric tissues better requires us to use laser microdissection.
We showed that the expression of APOBEC2 was associated with the infiltration of CD66b
+TANs, differentiation grade, TNM stage, histological classification, and gender in STAD. In particular, the percentage of APOBEC2-positive patients in the well differentiation grade was high. Furthermore, the Kaplan-Meier analyses in 496 patients with STAD also indicated that APOBEC2-negative patients had higher risk of death and recurrence compared with APOBEC2-positive patients. The above results indicated that APOBEC2 may be involved in the progression and prognosis of gastric adenocarcinoma. Although APOBEC2 expression was associated with the infiltration of CD66b
+ TANs in STAD, there was no correlation between the infiltration of CD163
+ TAMs. Consequently, there are questions to be answered about the exact mechanism of correlation between APOBEC2 and tumor infiltrating immune cell. Some studies suggested that the APOBEC2 might be involved in immune reactions to infection or inflammation. For example, it has been reported that elevated APOBEC2 was negatively related with serum IgG level in the tonsils with IgA nephritic patients [
22]. Moreover, Matsumoto et al. reported that pro-inflammatory cytokines (TNF-α and IL-1β) can promote the recruitment of NF-kB, which is known to have a complicate relationship with carcinogenesis among several functional NF-kB binding sties in the APOBEC2 promoter region [
23]. Research showed that neutrophils can produce TNF-α and IL-1β involved in in the inflammatory response and immune regulation [
24]. Further studies are warranted to clarify associations between APOBEC2 and tumor microenvironment including immune cell infiltration.
At present, adjuvant chemotherapy is usually recommended for patients with advanced gastric cancer, according to the guidelines of the National Comprehensive Cancer Network (NCCN) [
25]. In addition, a certain number of trials around the world have shown that adjuvant chemotherapy contributes to reducing the rate of recurrence and improving survival rates in gastric cancer [
26]. However, given the adverse reactions and resistance issue, it is critical to look for biomarkers that might predict adjuvant chemotherapy success. In our study, Kaplan-Meier survival curves for the patients receiving postoperative chemotherapy (n = 339) showed a difference between the APOBEC2-negative arm and the APOBEC2-positive arm. Here, we showed that APOBEC2-negative was related to poor survival outcomes in patients receiving postoperative adjuvant chemotherapy. Furthermore, we also illustrated the remarkable survival difference in stage III patients who received postoperative chemotherapy (n = 171). As for the results of stage II patients, we speculated that the small sample size was under-powered to show statistical differences between groups. To reduce the impact of age, we performed survival analyses on patients larger than the median age (n = 160) and on patients below the median age (n = 179) who received postoperative chemotherapy, respectively. Then the remarkable survival difference could still be found. So, we presume that detection of APOBEC2 expression could predict the efficacy of postoperative chemotherapy in patients. TMB was negatively correlated with APOBEC2 expression in STAD [
12]. Patients with TMB-low benefited better from postoperative chemotherapy in terms of OS and DFS in some cohorts [
27]. Combine our findings, it may serve as an inspiration for future research on the influence of postoperative chemotherapy, APOBEC2, and TMB on the prognosis of STAD patients.
Additionally, we also evaluated the influence of postoperative chemotherapy on prognosis of gastric cancer. Multivariate analysis revealed that received postoperative chemotherapy was associated with poor prognosis in that patients with adjuvant chemotherapy had higher risk of death and recurrence compared patients without adjuvant chemotherapy (Tables
2 and
3). These unexpected results were similar to those earlier published studies that revealed a greater survival rate in the surgery-only control group and no statistically significant benefit from adjuvant chemotherapy. For example, the median overall survival was 57.6 months and 56.7 months, respectively, in a study by the Italian Oncology Group for Cancer Research that compared surgery alone vs. adjuvant cisplatin chemotherapy [
28]. In another randomized trial designed by the European Organization for Research and Treatment of Cancer (EORTC) institutions, there were no significant differences between the postoperative chemotherapy arm and control arm (treated with surgical operation alone) for either OS or DFS [
29]. They reported 5-year DFS was 41% in the treatment arm and 42% in the control arm, and 5-year OS was 43% and 44%, respectively. Furthermore, we speculated that the unexpected results of our cohort may be because of the different chemotherapy regimen, poor compliance to adjuvant chemotherapy or elderly patients combined with poor chemotherapy tolerance.
In our study, univariate Cox analysis (Table
2) revealed that negative APOBEC2 was associated with poor prognosis in that APOBEC2-negative patients had higher risk of death and recurrence compared with APOBEC2-positive patients. When used together as a two-marker predictor, APOBEC2 and CD66b were able to separate 496 patients with STAD into groups with different prognoses. Univariate Cox analysis revealed that patients with APOBEC2
−CD66b
low had the worst OS and DFS among all four subgroups (Table
3). Nevertheless, it is noted that when adjusted for other factors (e.g., TNM stage) in the multivariate Cox analysis, APOBEC2 was not an independent prognostic factor for STAD. And multivariate Cox analysis demonstrated that the two-marker classifier (APOBEC2 and CD66b) couldn’t independently predict the prognosis of STAD, when the APOBEC2
+CD66b
high subgroup was used to a reference. In a word, we failed to show APOBEC2 or the combination of APOBEC2 and CD66 was an independent prognostic factor for stomach adenocarcinoma. There are several explanations for the unexpected results. First, we speculated that the relatively low proportion of early disease in our cohort might affect the prognostic evaluation values of APOBEC2. Second, there are other important factors that influence the recurrence and the survival of STAD. Thirdly, it might be due to a lack of statistical power resulted from limited sample size (n = 496, particularly, there were some samples with censored data).
It should be noted that our study has certain limitations. More verification and molecular mechanisms are needed even we have made the connections that patients with positive APOBEC2 may benefit from postoperative chemotherapy to improve OS and DFS, and the APOBEC2 expression profiles were evaluated by TMAs, immunohistochemical, conventional pathological methods. High-throughput sequencing, such as single-cell RNA sequencing (scRNA-seq), has become a powerful tool in the field of cancer research and has provided novel insights into the cellular and molecular features of tumors [
30‐
32]. Further studies are needed to clarify the expression of APOBEC2 in specific cells of gastric cancer based on scRNA-seq.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.