Abstract
Background/Aim: Previous studies from our research group have shown that trisomy 8 and the amplification of the 8q24.21 region is very frequent in gastric cancer (GC). Little is known about the role of most genes located in this region. Thus, the aim of this study was to understand the possible impact of transcriptional alterations and copy number variation (CNV) of four genes located in the 8q24.21 region – FAM49B, FAM84B, GSDMC and miR-5194 – in GC. Materials and Methods: Fifty-one to 85 matched pairs of tumoral and adjacent non-tumoral gastric tissues, from patients with primary GC, were used to analyze gene expression and CNV of the selected genes. We also included 29 H. pylori negative and gastritis negative gastric mucosa tissues from individuals without cancer obtained by endoscopy, as control samples. Results: The expression of FAM49B, GSDMC and miR-5194 was higher in both tumoral and adjacent non-tumoral samples compared to the negative control. The expression of FAM84B showed no significant difference between tumoral samples and negative controls. However, the expression of FAM84B in the adjacent non-tumoral samples was higher compared to negative control and tumoral samples. Moreover, the higher expression of GSDMC was associated with T3 and T4 tumors, with tumors on stage III and IV and with advanced tumors. Higher copy numbers of FAM49B and GSDMC were associated with intestinal tumor type and with moderately or well-differentiated tumors. Higher copy number of FAM84B was associated with moderately or well-differentiated tumors. Furthermore, the expression of all four genes was positively correlated. Conclusion: All four genes are upregulated in GC and may play an important role in these neoplasms. GSDMC expression was associated with more aggressive tumors.
Although incidence and mortality of gastric cancer (GC) have been globally decreasing in the last decades, it remains the fifth most common neoplasm and the fourth cause of cancer-related death worldwide (1). GC cause is multifactorial and it is strongly associated with Helicobacter pylori infection (2-5). Approximately 80% of the patients with GC are diagnosed in the advanced stages of this neoplasia due to the absence of specific symptoms in the early stages. Also, due to the late diagnosis, the average 5-years survival rate is less than 20%, emphasizing the urgency of new strategies for early diagnosis. Only a few countries, like Japan, adopt extensive programs for detecting early GC (6).
Trisomy 8 is one of the most common aneuploidies in GC primary tumors and GC cell lines (7-12) and mostly involves the amplification of 8q24 locus (13, 14). The 8q24.21 locus harbors genes related to cancer development process and bad GC prognosis, like the proto-oncogene MYC (8, 9, 13-19). Besides MYC, the 8q24.21 locus harbors 6 protein-coding genes and 19 non-coding genes, including 6 microRNAs and 13 long non-coding RNAs, but there are few studies examining these genes.
Therefore, based on the current literature, our group selected four genes located at 8q24.21 to be analyzed in GC. Three are protein-coding: Family with sequence similarity 49 member B (FAM49B), also known as CYRI; Family with sequence similarity 84 member B (FAM84B) and Gasdermin C (GSDMC), and one microRNA, miR-5194. To our knowledge, there are a few studies associating FAM49B (20), FAM84B (21-23) and GSDMC (24, 25) with GC, while there is no study associating miR-5194 with GC. Thus, the objective of this study was to analyze the RNA expression of FAM49B, FAM84B, GSDMC and miR-5194, as well as the copy number variation (CNV) of the former 3 genes.
Materials and Methods
Subjects and sampling. Depending on the gene studied, 51 to 85 matched pairs of tumoral and adjacent non-tumoral gastric tissues were obtained from patients with primary gastric adenocarcinoma who underwent gastric resection at Hospital São Paulo (HSP) and Hospital Universitário João de Barros Barreto (HUJBB). None of the patients had any history of exposure to either chemotherapy or radiotherapy prior to surgery, or other co-occurrence of diagnosed cancers. Samples were classified according to the Lauren classification; these data may be found in Table I. Furthermore, we included 29 H. pylori negative and gastritis negative gastric mucosa tissues as control samples; these were from individuals without cancer who underwent routine endoscopy at Centro de Gastroenterologia da Faculdade de Medicina de Marília (FAMEMA).
Associations between clinicopathological features with FAM49B. FAM84B. GSDMC and miR-5194 expression.
H. pylori presence was detected by PCR for amplification of the 150 bp fragment, referring to the 16S subunit fragment of bacterial rRNA, according to the methodology realized by our research group. Sequence of the s1/m1 allele of the VacA gene, of the 60190 strain of H. pylori (GeneBank access n°. U05676). Sequence of the s2/m2 allele of the VacA gene, of the Tx30a strain of H. pylori (GeneBank access n°. U29401) (26). The oligonucleotides Cag1 (ATGACTAACGAAACTA TTGATC) and Cag2 (CAGGATTTTTGATCGCTTTATT) were used for detection of CagA gene (27).
Written informed consent with the approval of the ethics committee of HSP and HUJBB was obtained from all patients before and after surgery (CAAE 32833514.9.1001.5505).
All procedures of the research project (CAAE 32833514.9. 1001.5505) followed the ethical standards of the Ethics Committee of Hospital São Paulo, Federal University of São Paulo – UNIFESP, according to the Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans. All invited patients and their legal representatives agreed to participate on the research and signed the Informed Consent Form before the evaluation.
DNA, RNA extraction and cDNA synthesis. Total DNA and RNA were extracted from frozen gastric tissue using the AllPrep DNA/RNA/miRNA kit (Qiagen, Hilden, North Rhine-Westphalia, Germany), according to the manufacturer’s instructions. The quality and concentration of the DNA and RNA were measured using NanoDrop ND-1000 (Thermo Fisher Scientific, Waltham, MA, USA). The complementary DNA (cDNA) was synthetized using High-Capacity cDNA Reverse Transcription (Thermo Fisher Scientific), according to the manufacturer’s instructions.
Quantitative real-time polymerase chain reaction (qPCR). Gene expression detection and quantification was carried out using TaqMan assays-on-demand probes (Thermo Fisher Scientific) on the Viia 7 real-time PCR system (Thermo Fisher Scientific). Each qPCR was performed in triplicate. For gene expression standardization we used GAPDH+B2M, as determined by our research group (28). Gene expression analysis was performed with values of ΔCrt and ΔΔCrt, which have inverse relationships to the transcriptional levels of each gene. The Crt method sets a threshold for each curve individually based on their shape, unlike the Ct method, that considers all the curves for a specific target to determine the threshold (29, 30).
Copy number variation analysis. Copy number variation analysis was realized using TaqMan assays-on-demand probes (Thermo Fisher Scientific). All the qPCR reactions were performed in quadruplicates using genomic DNA (gDNA) from GC tissue on the Viia 7 real-time PCR system. As internal control, we analyzed the copy numbers of the RNAse P gene (Thermo Fisher Scientific). The human gDNA G1471 and G1521 (Promega) were used for calibration. The CopyCaller software (Applied Biosystems, Waltham, MA, USA) was used to predict the copy number of each gDNA sample, according to software instructions.
Statistical analysis. Statistics were performed using Jamovi version 2.2 (31). The statistical power of the tests was calculated by post hoc analysis using GPower 3.1. The data were first analyzed for their distribution using the Shapiro-Wilk test. The data didn’t follow a normal distribution and were thus analyzed by non-parametric tests. The Wilcoxon test was used to compare the gene expression between tumoral and adjacent non-tumoral samples. Mann-Whitney U-test was used to associate gene expression with patients’ clinicopathological features. The Mann-Whitney U and Kruskal-Wallis tests were used to associate the CNV of each gene with their gene expression. The chi-square test or Fisher’s exact test were used to associate the CNV with the clinicopathological characteristics. The Spearman correlation test was used to correlate the gene expression of each gene in tumoral gastric tissue. Cox Regression was used for survival analysis, model metric designed by Jamovi project (32-35). All confidence intervals were set at the 95% level.
Results
Gene expression in GC and controls. The expression of FAM49B, FAM84B and miR-5194 was higher in the adjacent non-tumoral samples compared to the gastric tumoral samples (Z=−2.304, p=0.021; Z=−4.176, p<0.001; Z=−3.922, p<0.001, respectively). On the other hand, the expression of GSDMC was higher in the gastric tumoral samples when compared to the adjacent non-tumoral samples (Z=−2.213, p=0.027). Furthermore, we observed that, in both tumoral and adjacent non-tumoral samples, the expression of FAM49B (U=252 and 158, respectively, p<0.001 for both comparisons), GSDMC (U=82.5 and 212, respectively, p<0.001 for both comparisons) and miR-5194 (U=652 and 393, respectively, p<0.001 for both comparisons) was higher compared to the control samples. However, we observed a significant increase of FAM84B expression only in the adjacent non-tumoral samples compared to controls (U=379, p<0.001; Figure 1).
Analysis of FAM49B, FAM84B, GSDMC and miR-5194 gene expression. (a) Expression of FAM49B in 85 pairs of tumoral and adjacent non-tumoral gastric samples and 29 gastric control samples. (b) Expression of FAM84B in 51 pairs of tumoral and adjacent non-tumoral gastric samples and 29 gastric control samples. (c) Expression of GSDMC in 52 pairs of tumoral and adjacent non-tumoral gastric samples and 26 gastric control samples. (d) Expression of miR-5194 in 77 pairs of tumoral and adjacent non-tumoral gastric samples and 29 gastric control samples. Data are expressed as median±95% confidence interval. *p<0.05 by Mann-Whitney test. **p<0.05 by Wilcoxon test.
Gene expression, survival analysis and correlations in GC. The COX Regression for survival analysis showed no increased hazard between high or low expression of FAM49B, FAM84B, GSDMC and miR-5194. These data may be found in Figure 2.
Survival analysis. (a) Survival analysis of GC patients with high (N=27) and low (N=25) expression of FAM49B. (b) Survival analysis of GC patients with high (N=16) and low (N=10) expression of FAM84B. (c) Survival analysis of GC patients with high (N=8) and low (N=22) expression of GSDMC. (d) Survival analysis of GC patients with high (N=20) and low (N=26) expression of miR-5194. Data are expressed as 2−ΔCRT.
Our results showed that the transcriptional levels of FAM49B, FAM84B, GSDMC and miR-5194 are positively correlated to each other in tumoral gastric samples, except between FAM84B and GSDMC (Table II).
Correlation between the transcriptional levels of FAM49B, FAM84B, GSDMC and miR-5194 in tumoral gastric samples.
Gene expression and clinicopathological associations. Increased expression of GSDMC in tumoral gastric samples was associated with more invasive tumors (U=201, p=0.043), with tumors in stage III and IV (U=217, p=0.044) and with advanced tumors (U=118, p=0.033) (Figure 3). The transcriptional levels of FAM49B, FAM84B and miR-5194 in gastric tumoral samples were not significatively associated with any clinicopathological feature. All the clinicopathological data may be found in Table I.
Analysis of gene expression and clinicopathological associations. (a) Association of higher GSDMC transcriptional levels with tumors T3 and T4 (N=34) compared to tumors T1 and T2 (N=18). (b) Association of higher GSDMC transcriptional levels with tumors on stage III and IV (N=21) compared to tumors on stage I and II (N=31). (c) Association of higher GSDMC transcriptional levels with advanced tumors (N=42) compared to early tumors (N=10). Data are expressed as median±95% confidence interval. *p<0.05 by Mann-Whitney test.
Copy number variation and clinicopathological associations. The CNV of FAM49B, FAM84B and GSDMC was not significatively associated with the respective gene expression, but the increase of FAM49B and GSDMC copy number was associated with intestinal type tumors (χ2=10.2, p=0.001 and χ2=5.86, p=0.05, respectively). Moreover, the increase of FAM49B, FAM84B and GSDMC copy number was associated with well or moderately differentiated tumors (χ2=14, p<0.001; χ2=7.82, p=0.005 and χ2=8.37, p=0.015, respectively) (Figure 4). The study of miR-5194 CNV was not performed because this gene is located in a homologous region to FAM49B, therefore, the amplification of the region of interest would be nonspecific to miR-5194. All the CNV and clinicopathological data may be found in Table III.
Analysis of copy number variation and clinicopathological associations. (a) Association of increased FAM49B copy number with intestinal type tumors (N=82). (b) Association of increased FAM49B copy number with well/moderately differentiated tumors (N=81). (c) Association of increased FAM84B copy number with well/moderately differentiated tumors (N=49). (d) Association of increased GSDMC copy number with intestinal type tumors (N=50). (e) Association of increased GSDMC copy number with well/moderately differentiated tumors (N=50). Data are expressed as % of samples. *p≤0.05 by Chi-squared test.
Associations between clinicopathological features with FAM49B, FAM84B and GSDMC copy number.
Discussion
The 8q24.21 region is known as a “gene desert” as there is a lack of protein-coding genes. In addition, it is frequently altered in several different tumors, being responsible for the susceptibility to cancers mainly by the amplification of the proto-oncogene MYC, as previously observed by our group (36-41). Thus, we selected three protein-coding genes and one microRNA to analyze their transcriptional levels and CNV in gastric cancer: FAM49B, FAM84B, GSDMC and miR-5194.
FAM49B acts on cytoskeletal remodeling, which leads to bacterial protection, T cell activation and mitochondrial regulation (42-46). In this study we found an overexpression of FAM49B and increased copy numbers in GC samples, suggesting that this gene may be an oncogene in GC, similarly to breast and gallbladder cancer (47, 48). Conversely, this gene has been described as a tumor suppressor in pancreatic ductal adenocarcinoma (42). In the literature, FAM84B was associated with the progression of esophageal squamous cell carcinoma, glioma, prostate and breast cancers (23, 49-53). Moreover, the long non-coding RNA FAM84-AS promotes resistance to platinum drugs in GC (22), suggesting that FAM84B acts as an oncogene in these neoplasms. As in the literature, our results showed that FAM84B may also be an oncogene in GC. MiR-5194 has been reported to be downregulated in colorectal cancer and glioblastoma (54, 55), but we found miR-5194 upregulated in GC samples. The expression levels of miR-5194 in GC were unknown until this study and its role is still unclear. This microRNA may be part of the network between MYC oncogene and microRNAs in GC, leading to the development and progression of this neoplasm (56).
Interestingly, the expression of FAM49B, FAM84B and MIR5194 was higher in the adjacent non-tumoral samples when compared to tumoral samples; this result showed that the adjacent tissue, despite not showing macroscopic signs of neoplasia, carries significant genetic alterations. This may be evidence that FAM49B, FAM84B and MIR5194 play more significant roles in the beginning of the carcinogenesis process than in tumor maintenance.
In the literature, Saeki et al. reported that GSDM is mainly expressed in the epithelium of gastric tissues. Also, although GSDM gene is amplified in GC cells, its transcriptional levels are suppressed (57, 58). Moreover, the GSDMC gene is reported as an oncogene in lung, colorectal, serous ovarian, kidney and intestinal type gastric cancer, but not in diffuse type GC (24, 59-62). In this study, we showed that GSDMC is overexpressed in GC regardless of the histological type, and it may also be an oncogene in GC. Interestingly, GSDMC is a member of the gasdermins family, which induces inflammation-mediated programmed cell death (pyroptosis). Thus, it would be expected to have an anti-tumor action, but only a few studies have associated the expression of this gene with a suppressive activity (25, 63, 64). Yang et al. reported that GSDMC is upregulated in GC tissues but it is associated to better prognosis due the role of pyroptosis (25).
GSDMC proved to be the most promising of the four studied genes in this study. Unlike the other three genes, GSDMC showed a higher expression in tumoral samples compared to adjacent non-tumoral samples. Furthermore, this gene showed a higher expression in both tumoral and adjacent non-tumoral samples compared to control samples. This result indicates that this gene may play important roles both in the beginning of the carcinogenesis process and in tumor maintenance. Moreover, the expression of GSDMC was significantly higher in more invasive, advanced and in stage III and IV tumors, emphasizing the importance of more detailed studies about this gene in GC. The results of all the clinicopathological analyses may be seen in Table I.
We also sought to elucidate the deregulation of FAM49B, FAM84B and GSDMC by analyzing their copy number. But the results of this CNV study did not show an effect of the copy number on the transcriptional levels of each gene. Therefore, the mechanism that causes the alteration of the expression of these genes still needs to be better studied and understood. As the 8q24.21 region is a “gene desert”, it carries several long non-coding RNAs and microRNAs that are frequently altered in cancer and play roles from transcription to translation (65), so this could be the mechanism by which the genes in this region are deregulated. Another possibility is that there is an association among MYC and the genes in 8q24.21, as reported with FAM84B (23).
Besides that, the increase in FAM49B, FAM84B and GSDMC copy number was associated with intestinal type and moderately/well differentiated tumors, which are complementary results, since one of the main features of the intestinal type tumors is to present a higher cellular differentiation compared to diffuse type GC (66).
The genes FAM49B, FAM84B, GSDMC and miR-5194 are genetically linked, and their expression in tumoral samples were positively correlated. These data reinforce the study of Du et al. (67), which shows the region 8q24 as a regulatory hub of genes that interacts with each other and with multiple genomic loci across the genome in a prostate cancer cell line.
Conclusion
According to our results, we can conclude that FAM49B, FAM84B, GSDMC and miR-5194 are upregulated in GC. Therefore, these transcripts are potential diagnostic biomarkers in GC.
Trisomy 8 is often found in primary GC and GC cell lines, with the amplification of 8q24 locus being the most frequent. Our results confirmed these data. FAM49B, FAM84B and GSDMC have alterations in their copy number. In addition, their increased copy number is associated with intestinal type and moderately or well differentiated tumors.
Finally, GSDMC seems to have a more relevant impact on GC progression, since its expression is associated with more aggressive cancers, i.e., advanced and more invasive tumors.
Acknowledgements
We would like to thank Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) and Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) for the financial support and fellowships. We would also like to thank the staff of the Hospital São Paulo (HSP), Hospital Universitario João de Barros Barreto (HUJBB) and Centro de Gastroenterologia da Faculdade de Medicina de Marília (FAMEMA) for their contribution with the sample collection. This study was supported by the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP, Process n° 2016/25562-0 to MACS) and the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq; to MACS).
Footnotes
Authors’ Contributions
Conceptualization: Marília Arruda Cardoso Smith, Fernanda Wisniesk; writing-original draft preparation: Brunno dos Santos Pereira; writing: Brunno dos Santos Pereira; review and editing: Marília Arruda Cardoso Smith, Fernanda Wisniesk and Bruno Takao Real Karia; sample collection: Brunno dos Santos Pereira, Leonardo Caires dos Santos, Fernanda Wisniesk, Renata Sanches Almeida, Camila Albuquerque Pinto, Elizabeth Suchi Chen, Rommel Rodríguez Burbano; bench work: Brunno dos Santos Pereira, Camila Albuquerque Pinto, Renata Sanches Almeida; statistical analysis: Brunno dos Santos Pereira and Fernanda Wisniesk; surgery: Paulo Pimentel Assumpção, Carlos Haruo Arasaki, Laercio Gomes Lourenço; supervision: Marília Arruda Cardoso Smith. All Authors have read and agreed to the published version of the manuscript.
Conflicts of Interest
The Authors declare that they have no conflicts of interest.
- Received May 11, 2022.
- Revision received July 6, 2022.
- Accepted July 11, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.