Comprehensive analysis of the expression of SLC30A family genes and prognosis in human gastric cancer

The solute carrier 30 (SLC30) family genes play a fundamental role in various cancers. However, the diverse expression patterns, prognostic value, and potential mechanism of SLC30A family genes in gastric cancer (GC) remain unknown. Herein, we analyzed the expression and survival data of SLC30A family genes in GC patients using multiple bioinformatic approaches. Expression data of SLC30A family genes for GC patients were extracted from the Cancer Genome Atlas (TCGA) and genetic alteration frequency assessed by using cBioportal database. And validated the expression of SLC30A family genes in GC tissues and corresponding normal tissues. The prognostic value of SLC30A family genes in gastric cancer patients were explored using Kaplan–Meier plotter database. Functional enrichment analysis performed using DAVID database and clusterProfiler package. And ssGSEA algorithm was performed to explore the relationship between the SLC30A family genes and the infiltration of immune cells. We found that the median expression levels of SLC30A1-3, 5–7, and 9 were significantly upregulated in gastric cancer tissues compared to non-cancerous tissues, while SLC30A4 was downregulated. Meanwhile, SLC30A1-7, and 9 were significantly correlated with advanced tumor stage and nodal metastasis status, SLC30A5-7, and 9–10 were significantly related to the Helicobacter pylori infection status of GC patients. High expression of five genes (SLC30A1, 5–7, and 9) was significantly correlated with better overall survival (OS), first progression survival (FPS), and post progression survival (PPS). Conversely, upregulated SLC30A2-4, 8, and 10 expression was markedly associated with poor OS, FP and PPS. And SLC30A family genes were closely associated with the infiltration of immune cells. The present study implied that SLC30A5 and 7 may be potential biomarkers for predicting prognosis in GC patients, SLC30A2 and 3 play an oncogenic role in GC patients and could provide a new strategy for GC patients treatment.


Materials and methods
Patients and samples. The present study was performed using data obtained from 40 consecutive patients from Cixian and Cixian, a region in Hebei Province with a high rate of epidemiologically and histologically confirmed GC 9,11 . All patients were surgically treated at The Fourth Hospital of Hebei Medical University from January 1, 2017 to December 31, 2018. All patients have received pathological diagnosis of primary GC (Supplementary Table 1).

RNA isolation and reverse transcription-quantitative polymerase chain reaction (RT-qPCR).
Total RNA was extracted from frozen tumor and corresponding non-tumorous tissues using TRIzol reagent (Invitrogen, Thermo Fisher Scientific, Inc.). After the concentration and purity of the total RNA were determined by ultraviolet absorbance spectroscopy, RNA was reverse transcribed into cDNA using Rever-tAid First Strand cDNA Synthesis Kit (Thermo Scientific, Lithuania). qRT-PCRs using SuperReal PreMix Plus (SYBR Green) (TianGen, Beijing, China) were performed on ABI7500 Real-Time System (Life Technologies Corp., Foster City, CA, USA). The PCR cycling parameters were as follows: 95 ℃ for 10 min, and 40 cycles of 95 ℃ for 15 s, 60 ℃ for 30 s and 72 ℃ for 30 s. The samples were run in triplicate and the mean value was calculated for each case. The primers for SLC30A family genes are listed in Supplementary Table 2. The human GAPDH gene was employed as an internal control. The relative expression of SLC30A family genes was calculated using the 2 −ΔΔCT method according to the previously described protocol 25 . TCGA database. TCGA is a large repository of high throughput data of human carcinomas, containing over 30 human tumor cohort studies 26 . The expression profiling of SLC30A family genes were retrieved from the TCGA-STAD database. In addition, the clinicopathological parameters of GC were downloaded from TCGA in order to assess the diagnostic value of SLC30A family genes in GC patients using receiver operating characteristic (ROC) curve. UALCAN database. UALCAN is a web resource that provides comprehensive cancer transcriptome data (https ://ualca n.path.uab.edu/) 27 . The expression level of SLC30A family genes in GC tissues and normal gastric tissues were assessed using the UALCAN database. TIMER database analysis. TIMER (https ://cistr ome.shiny apps.io/timer /) is an a comprehensive and userfriendly online tool to systematically investigate and visualize the correlation between immune infiltrates and a wide spectrum of factors, including gene expression, clinical outcomes and somatic mutations over 10,897 tumors from 32 cancer types 28,29 . The differential expression of SLC30A family genes between tumor and normal tissues could be evaluated using Diff Exp module across all the TCGA database tumors and the results were shown with boxplots. cBioportal database. cBioportal is an interactive open-source platform, that provides large scale cancer genomics data sets (https ://www.cbiop ortal .org/) 30,31 . The frequency of SLC30A family gene alterations (ampli-Kaplan-Meier plotter database. ROC curve analysis was conducted using the pROC 37 package in R software to explore the sensitivity and specificity of using the SLC30A family genes to distinguish GC patients from healthy individuals. Kaplan-Meier plotter (https ://kmplo t.com/) is an online database containing microarray gene expression data and survival information extrcated from Gene Expression Omnibus and TCGA database, which contain the gene expression and survival data of 1065 GC patients 38 . 631 GC patients were included in this study (Supplementary Table 3). Patients missing expression values or lacking complete clinical data, Single-sample gene set enrichment analysis (ssGSEA). The infiltration levels of immune cell types were quantified by ssGSEA method using gsva package 39 in R software. The ssGSEA applies gene signatures expressed by immune cell populations to indivadual cancer samples 40 . The deconvolution approach used in our study including Consent for publication. All authors have reviewed the manuscript and consented for publication.

Results
Relative transcriptional expression of SLC30A family genes in GC patients using the UALCAN database. Comparison of the transcriptional expression of SLC30A family genes in gastric tumor tissues and normal tissues indicated that mRNA expression of SLC30A1-3, 5-7, and 9 was significantly upregulated in cancer tissues compared to non-cancerous tissues in GC patients, while SLC30A4 was downregulated in the former compared to the latter ( Fig. 1A and Figure S3). Moreover, assessment of the correlation between SLC30A family genes expression levels and the tumor stages of GC patients indicated that the expression levels of most SLC30A family genes, including SLC30A1, 5-7, and 9, were significantly and positively associated with tumor stage in GC patients. Nevertheless, SLC30A8 and 10 expression had no statistical significance (Fig. 1B). We also analyzed the relationship between the expression level of SLC30A family genes and the nodal metastasis status of GC patients. Five genes were positively associated with nodal metastasis for GC patients (SLC30A1, 5-7, and 9). However, SLC30A4 was significantly and negatively correlated with nodal metastasis for GC patients (Fig. 1C). The expression level of most SLC30A family genes was significantly associated with the Helicobacter pylori infection status of GC patients, but the most significant correlation occurred for SLC30A5-10 ( Fig. 1D). Furthermore, we validated the expression of SLC30A family genes in 40 GC patients. Most of the expression levels of SLC30A family genes were consistent with those of previous studies, but the expression levels of SLC30A8 and 9 had no significant differences between GC tissues and corresponding non-cancerous tissues (Fig. 1E).  (Table 1). Using a forest plot to investigate the potential prognostic value of SLC30A family genes, to reveal the correlation between OS, FPS, PPS, and mRNA expression of SLC30A family genes in GC patients ( Fig. 3D-F). The results showed that the high expression of five genes, (SLC30A1, 5-7, and 9), had a positively significant correlation with improved FPS, and PPS. In contrast, upregulated SLC30A2-4, 8, and 10 expression was negatively correlated with favorable FPS, and PPS.
Association of SLC30A family genes prognostic values in GC patients with different clinicopathological features. Investigation of the correlation between clinicopathological features such as gender, clinical stage, Lauren classification, differentiation, HER2 status, treatment types, and perforation and mRNA expression level of SLC30A family genes showed that all SLC30A family gene expression was significantly correlated with gender in GC patients ( Table 2). Five genes were promising positive prognostic factors in both male and female patients, including SLC30A1, 5-7, and 9. Nevertheless, SLC30A2-4, 8, and 10 were significantly correlated with poor prognosis in both male and female patients. Upregulated expression of SLC30A1, 5-7, and 9 predicted a favorable prognosis in GC patients with stage III/IV, I/III/IV, I/III/IV, I/III/IV, and I/III/IV, respectively (Table 3). High expression of SLC30A2-4, 8, and 10 was significantly associated with an unfavorable prognosis in stage I/III/IV, III, III/IV, I/III/IV, and I/III GC patients, respectively. SLC30A1, 3, 5-7, and 9 were promising favorable prognostic factors in both intestinal and diffuse type GC patients, and high SLC30A5 expression was also significantly correlated with mixed type patients (Table 4). Besides, SLC30A2 and SLC30A8 predicted poorer prognosis in both intestinal and diffuse type patients and high expression of SLC30A3 and SLC30A10 correlated with poor prognosis in intestinal, mixed type GC patients, respectively. High expression of SLC30A2, 4, and 9 were correlated with the improved prognosis in poorly differentiation GC patients (Table 5). Nevertheless, SLC30A1 and 6 were significantly associated with poor OS in moderately differentiation GC patients. Analysis of HER2 status and expression of SLC30A family genes revealed that upregulated expression of SLC30A1, 5-6, and 9 predicted favorable OS in both HER2-positive and HER2-negative patients, while SLC30A2-3, 8, and 10 were associated with a worse prognosis. High expression of SLC30A4 and 7 were significantly associated with unfavorable OS in HER2-negative and improved prognosis  (Table 6). SLC30A6-7 and 9 were strongly related to favorable OS in GC patients based on a surgery only treatment. SLC30A1 and 9-10 were positively associated with other adjuvant treatments, while high expression of SLC30A2 predicted better prognosis in 5 fluorouracil (FU)-based adjuvant treatment. Nevertheless, overexpression of SLC30A2-3 and 8 were correlated with poor prognosis in patients that received surgery alone. SLC30A3, 8, and 10 were strongly negatively associated with OS in patients that received 5-FU based adjuvant treatment (Table 7). Furthermore, analysis of the correlation between mRNA expression of SLC30A family genes and prognosis in patients with no perforation showed that SLC30A9 was a favorable factor in patients without perforation, while overexpression of SLC30A1 and 8 were significantly associated with poor prognosis (Table 8). Taken together, all SLC30A family genes were strongly correlated with clinical characteristics including gender, clinical stage, Lauren classification, differentiation, HER2 status, perforation, and treatment method (Fig. 4).
Genetic alteration differences of SLC30A family genes in GC patients. To explore the roles of SLC30A family genes in GC patients, genetic alteration of 10 genes was performed using the cBioportal database. A total of 1443 patients from seven GC studies were analyzed. As results showed that mRNA mutation, amplification and deep deletion were the most important factors for alteration in different GC subtypes, including tubular stomach adenocarcinoma, mucinous stomach adenocarcinoma, intestinal type stomach adenocarcinoma, stomach adenocarcinoma, signet ring cell carcinoma of the stomach, diffuse type stomach adenocarcinoma, papillary stomach adenocarcinoma and esophagogastric adenocarcinoma (Fig. 5A). As Fig. 5B shows that SLC30A family genes were altered in 269 samples of 1443 GC patients (19%). The genetic alteration percentages of SLC30A family genes for GC varied from 1.1% to 7% for individual genes (SLC30A1, 2.1%; SLC30A2, 1.1%; SLC30A3, 3%; SLC30A4, 1.6%; SLC30A5, 2.1%; SLC30A6, 1.9%; SLC30A7, 1.8%; SLC30A8, 7%; SLC30A9, 1.9%; SLC30A10, 1.3%). The results of Kaplan-Meier plotter and log-rank test showed no significantly statistical difference in overall survival (OS) and disease-free survival (DFS) in cases with and without SLC30A family genes alterations (P-value was 0.331 and 0.0915, respectively. Figure 5C,D).
Correlation and functional enrichment analysis of SLC30A family genes. To further reveal the potential functional mechanisms in GC patients, we constructed the correlation between the expression of SLC30A family genes, protein-protein interaction (PPI) network, gene ontology (GO) term analysis, and Kyoto  (Fig. 6). The individual mRNA expressions of SLC30A family genes in GC patients were weakly correlated (Fig. 6B). The PPI network showed that 30 genes including XPA, FARSB, DACH1, and DACH2 participated in PPI networks through multiple pathways, physical interactions, genetic interactions, shared protein domains and co-expression (Fig. 6A). SLC30A family genes and their neighboring genes were mainly involved in the zinc transport, cellular zinc ion homeostasis, zinc ion homeostasis, cellular transition metal ion homeostasis, and transition metal ion transport, which are mineral transport related biological processes and mineral absorption pathways analyzed by GO term analysis and KEGG pathway enrichment analysis (Fig. 6C-F).

Immune infiltrates in correlation with SLC30A family genes in GC. The complex interactions
between solid tumors and their microenvironment remain unclear, and previous studies had shown that immune infiltrates were significantly related to the progression and prognosis of GC [41][42][43] . We conducted the ssGSEA algorithm to deconvolve the relative abundance of each cell type based on expression profiling data obtained from GSE62254. The immune phenotype landscape as shown in Fig. 7A. We get further explored the coefficient of the association of SLC30A family genes in immune cell subsets (Fig. 7B). The results showed that SLC30A family genes were closely associated with the infiltration of immune cells, indicating that SLC30A family genes play an important role in GC partly because of immune infiltration.

Discussion
In the present study, ROC analysis suggested that most SLC30A family genes had high diagnostic value for distinguishing GC patients from healthy individuals and could play an important role in GC diagnosis. Furthermore, univariate survival analysis showed that upregulated SCL30A1, 5-7, and 9 expression was positively associated with favorable OS, FPS, and PPS. On the contrary, high expression of SLC30A2-4, 8, and 10 were significantly correlated with poor OS, FPS, and PPS in GC patients. Moreover, all SLC30A family genes were Zinc is an indispensable trace element that is crucial for the proper function of various cellular proteins and essential for key physiological processes including nucleic acid metabolism, regulation of gene expression, cell division 44,45 . Furthermore, cancer cells may extract zinc from circulation to promote cancer growth 46,47 . In this study, to our best knowledge and for the first time, we used various large database, including TCGA, GEO, UALCAN, cBioPortal, STRING, and Kaplan-Meier Plotter, to systematically analyzed the expression level of SLC30A family genes, prognostic values, genetic alterations, and functional enrichment analysis in GC patients.
Aberrant zinc expression levels and regulation of SLC30A family genes have been reported in various kinds of cancer. SLC30A1 is upregulated in bladder cancer and negatively targeted by miR-411 to inhibit the growth and metastasis of bladder cancer cells 48 . Upregulated SLC30A1 expression of could lead to cytotoxic cell death in human ductal adenocarcinoma cell lines 49 . Meanwhile, SLC30A1 has high expression in ovarian cancer (OC) cell lines and tissues and a recovery experiment revealed that upregulated SLC30A1 counteracts the effect of miR-8073 mimics on OC cell proliferation and apoptosis to affect the malignant progression of OC 50 . SLC30A2 is dysregulated in breast cancer lines and SLC30A2-mediated Zn accumulation in mitochondria is associated with increased mitochondrial oxidation 51 . Meanwhile, SLC30A2 over-expression leads to Zn vesicularization, shifts in cell cycle, enhanced apoptosis, and reduced proliferation and invasion in breast cancer 52 . SLC30A2overexpression represses the cytotoxic effects of zinc hyper-accumulation in malignant metallothionein-null T47D breast tumor cells 53 . SLC30A4 is significantly overexpressed in prostate cancer compared to normal tissues from other organs 22 . SLC30A5-7 and 9 are significantly upregulated in colorectal cancer and SLC30A9 is involved in the canonical Wnt pathway 24 . Overexpressed SLC30A7 in esophageal squamous cell carcinoma could be a mechanism adapted by tumor cells to maintain the basal zinc requirement for carrying out vital functions during zinc deficiency 54 . SLC30A7 is also significantly upregulated in hepatocellular carcinoma 55 . SLC30A8 is aberrantly www.nature.com/scientificreports/ expressed in breast cancer and glioblastoma tumors, and decreased expression of SLC30A8 could contribute to the uncontrolled growth, proliferation, and tumor maintenance of glioblastoma multiforme cells 56,57 . SLC30A9 expression is significantly higher in hepatocellular carcinoma tissues than adjacent non-cancerous tissues, but is not correlated with survival in hepatocellular carcinoma patients 58 . SLC30A10 is aberrantly expressed in colorectal cancer and is significantly related to the methylation epigenotype and molecular genesis of colorectal cancer 59,60 . In the present study, mRNA expression of SLC30A1-3, SLC30A5-7, and 9 was significantly upregulated in gastric cancer tissues compared to non-cancer tissues in GC patients, while SLC30A4 was downregulated in cancer tissues.
To further clarify the genetic alteration and carcinogenic mechanism of SLC30A family genes, we found that the percentages of genetic alterations in SLC30A family genes for GC varied from 1.1 to 7% for individual genes. Furthermore, the results of Kaplan-Meier plotter and log-rank test showed no significantly statistical differences in OS and DFS in cases with and without SLC30A family gene alterations. Consistent with previous research, GO term analysis and KEGG pathway enrichment analysis showed that SLC30A family genes contributed to mineral transport related biological processes, including zinc transport, cellular zinc homeostasis, cellular transition metal ion homeostasis, and the mineral absorption pathway and our results showed that SLC30A family genes were closely associated with the infiltration of immune cells,. Therefore, we hypothesized that the action mechanism of SLC30A family genes induced tumorigenesis and progression by regulating zinc homeostasis in tumor cells and partly because of immune infiltration. This may provide a new insight in diagnosis and treatment of GC patients, especially in areas with zinc deficiency such as Cixian and Linxian.

Conclusions
In conclusion, SLC30A family genes were aberrantly expressed in GC tissues. High expression of SLC30A1, 5-7, and 9 as well as low expression of SLC30A2-4, 8, and 10 were significantly associated with favorable prognosis in GC patients. High SLC30A2 expression was significantly correlated with poor OS, FPS, and PPS in in all of Table 4. The relationship between SLC30A family genes and OS in different Lauren classification of GC patients (Kaplan-Meier Plotter). a: The P-value was set at 0.05 and the * indicate that the results are statistically significant.    www.nature.com/scientificreports/ GC patients indicating that these genes play an oncogenic role in GC and are markers for improved GC survival and prognostic accuracy.
Scientific Reports | (2020) 10:18352 | https://doi.org/10.1038/s41598-020-75012-w www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.