- Email: [email protected]
Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
Clinics and Research in Hepatology and Gastroenterology (2014) xxx, xxx—xxx
Available online at
ScienceDirect www.sciencedirect.com
ORIGINAL ARTICLE
Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis Misuk Yang a, Hyun-Soo Kim a,∗, Mee Yon Cho b a
Division of Gastroenterology, Department of Internal Medicine, Yonsei University Wonju College of Medicine, Wonju, Republic of Korea b Department of Pathology, Yonsei University Wonju College of Medicine, Wonju, Republic of Korea
Summary Objective: Gastric cancer (GC) is histologically classified into intestinal type and diffuse type, and diffuse type cancer can be further subdivided into poorly differentiated carcinoma (PDC) and signet ring cell carcinoma (SRCC). Recent evidence suggests that early SRCC is an initial, differentiated form of diffuse GC that may evolve into PDC. This study aimed at identifying the molecular features of epigenetic methylation changes in histologic differentiation status of GC. Methods: Included in this study are 149 samples of paraffin-embedded tissues and 115 fresh endoscopically biopsied tissues. Multiple paraffin tissues involving normal (n = 22), dysplasias (GDs, n = 39), differentiated cancers (DCs, n = 35), PDCs (n = 33) and SRCCs (n = 20) were included as an experimental group. For the validation group, endoscopically biopsied tissues of DCs (n = 50), PDCs (n = 31), and SRCs (n = 34) were analyzed. DNAs, isolated from each group were analyzed to determine the methylation status of 6 genes (GDNF, RORA, MINT25, KLF7, CDH1, LINE-1) using pyrosequencing. Results: LINE-1 was hypomethylated in GCs compared to normal and GD. GDNF, RORA and MINT25 were more hypermethylated in intestinal type GCs than those of diffuse type GCs, whereas CDH1 showed opposite patterns of methylation. Among diffuse type GCs, SRCCs showed lower level of methylation for GDNF, RORA, MINT25 and KLF7, and higher level for CDH1 compared to PDCs. Conclusions: In conclusion, intestinal type of GCs shows different epigenetic methylation profiles compared to the diffuse one. Moreover, SRCCs have different methylation profiles compared with PDCs, suggesting a unique molecular pathway in the gastric carcinogenesis. © 2014 Elsevier Masson SAS. All rights reserved.
∗ Corresponding author. Digestive Disease Center, Department of Internal Medicine, Yonsei University Wonju College of Medicine, 162, Ilsan-dong, 220-701 Wonju, South Korea. Tel.: +82 33 741 1229; fax: +82 33 741 1228. E-mail address: [email protected] (H.-S. Kim).
http://dx.doi.org/10.1016/j.clinre.2014.03.017 2210-7401/© 2014 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
2
M. Yang et al.
Introduction
Materials and methods
Gastric adenocarcinoma is one of the most common diseases in the world, and the main cause of cancer death in Asia [1]. In particular, the incidence rate of gastric cancer (GC; 41.1—44.3/100,000) was barely changed from 1999 to 2008, and the GC was the third leading cause of the cancer death in Korea [2]. Environmental factors including diet, nutritional conditions, Helicobacter pylori (H. pylori) infection and smoking are known as risk factors for GC development [3—5]. Furthermore, genetic and epigenetic factors also play a critical role in gastric carcinogenesis [6—10]. These genetic factors include single-nucleotide polymorphisms (SNPs) and DNA mutations [8—10], whereas epigenetic factors include posttranslational modifications, aberrant DNA methylation and histone modification [7,10]. Lauren classification is the most widely used method that distinguishes two histological types according to microscopic morphology of gastric adenocarcinoma [9—12]. Intestinal type is the well-differentiated (WD) tumor with cohesive tumor cells with gland-like tubular structures. Diffuse type is undifferentiated tumors that infiltrate through the stomach wall [9,10,13]. However, many gastric tumors are sorted as mixed type since some are not classified into either type of gastric tumors [10,14]. While intestinal type of GCs frequently metastasizes to the liver, diffuse type GCs can be easily disseminated to the peritoneum [10,15]. Gastric carcinogenesis progresses through serial steps involving atrophy, metaplasia, gastric dysplasia (GD) and carcinoma. These steps are more commonly seen in the intestinal type GC [7,10]. Molecular studies have provided evidence that GC results not only from the combined effects of environmental aspects and genetic modifications, but also from the cumulative genetic and epigenetic changes that play pivotal roles in tumorigenesis [8,10,16]. In GC, for instance, aberrant DNA methylations occur more frequently than mutations [12,17,18]. DNA methylation changes even occurring in nonneoplastic gastric epithelial cells, which suggest that the epigenetic methylation is an early event in GC [7,19—22]. Also, global DNA hypomethylation occurs in gastric or other tissues during the early stages of tumor growth [9,23]. GC proceeds through serial steps. Intestinal GC progresses through sequential steps after several genetic and epigenetic changes, while diffuse GC progresses directly to early cancer after similar genetic and epigenetic changes with intestinal GC [7,10,22]. Judging from the fact that the pathway is different for each type of cancer, genes associated with the cancer seem to affect different aspects of the changes involved in each pathway. Recent studies show that the pattern of lymph node metastasis is different between poorly differentiated carcinoma (PDC) and signet ring cell carcinoma (SRCC) [24,25]. Additionally, they also have different mechanisms of tumor invasion [26,27]. These reports have indicated that SRCC and PDC have different origins. In this study, therefore, we investigated differences of DNA methylation patterns at the molecular level especially in diffuse type GC, including SRCC and PDC.
Sample collection Gastric tumor tissues of the experimental group were obtained from Korean patients who had surgical or endoscopic resection at Yonsei University Wonju Severance Christian Hospital (Wonju, Korea) between May 2006 and December 2010. The 149 samples of paraffin-embedded tissues including 22 normal mucosa from healthy subjects, and 39 GDs, 35 differentiated gastric cancers (DGCs), 33 PDCs and 20 SRCCs from patients, were retrospectively examined. Among these samples recruited during the period, the pathologist randomly selected the patients who were diagnosed as GD or DGCs using the table of random sampling for numbers. Tissues were sent to pathologists for histopatologic diagnosis and fixed with paraffin. These paraffin-embedded tissue blocks were sliced for serial sections, and each section was microdissected by an expert pathologist to distinguish the tumor regions. The extracted DNA for each sample was therefore more than 70% homogeneous as described previously [18]. For the validation group, we collected 115 tumor tissues obtained between October 2008 and February 2011 at the same hospital. Because a large number of samples had been collected for intestinal type cancer, randomly selected samples were used. Due to the small number of available diffuse GC samples, however, all of them available were used. Small portions of these samples were used for histopathologic analysis and the rest of them were frozen immediately after biopsy. Serial sections of specimens were assessed for the presence of H. pylori by a modified Giemsa stain. Two biopsy specimens, one from the antrum and one from the corpus, were tested from each patient. The specimens were subjected to a rapid urease test (CLO test, Kimberly-Clark, USA), to detect the presence of H. pylori urease. A patient was defined as being H. pylori positive on the basis of a positive histological examination or a positive CLO test. Basic and clinicopathologic information about each patient were used as raw data. This study was approved by the institutional review board of Yonsei University Wonju Severance Christian Hospital (CR106021). We also notified all patients and received consents.
Genes selection for methylation study We evaluated DNA methylation in the promoter of 6 genes for this study. Retinoic acid-related orphan receptor-alpha (RORA) may functionally associated with Wnt signaling pathway that is involved in oncogenic systems by attenuating the pathway [28,29]. Glial cell derived neurotrophic factor (GDNF) is a small protein and interaction of upregulated GDNF and RET ligand-receptor are possibly involved in the glucose-induced cancer progression [30,31]. The Kruppellike family (KLF) of transcription factors have been shown to play pivotal roles in tumorigenesis of various tissues [32]. The promotor of KLF7 was frequently methylated in the diffuse type gastric cancers (unpublished data). The expression of CDH1, an important e-cadherin protein for cell
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
Methylation profiles by gastric cancer subtypes adhesion, is frequently downregulated in sporadic tumors and is associated with the poorly differentiated phenotype. Hypermethylation of this gene leads to gene silencing in GC [11,33]. Global DNA hypomethylation of DNA repetitive elements within the human genome, such as long interspersed nuclear element-1 (LINE-1) is a common finding in cancer and is believed to cause chromosomal instability and protooncogene activation [12,19]. Three genes (RORA, GDNF and MINT25) of the six were identified as sensitive molecular markers of early GC [18,27]. In the previous study [18], we found several kinds of methylation markers according to stepwise gastric tumor progression; type 1 markers, showing consistently high levels of methylation in both gastric dysplasia and cancer (GDNF, MINT25); type 2 markers, showing high levels of methylation in early gastric cancer and gastric dysplasia but decreased levels in advanced gastric cancer (RORA); type 3 markers, revealing higher methylation as the gastric tumor progresses. For the experimental group, we would like to focus on the methylation profile changes from normal to gastric cancer progression and selected 3 genes (GDNF, RORA, and KLF7) and LINE-1 as representatives of above 3 types of markers. For the validation group, however, we focused on the comparison of methylation profiles between PDC and SRCC based on the preliminary dataset from the experimental group. Therefore, we added 2 more methylation markers (MINT25 and CDH1) as MINT25 was confirmed as a general hypermethylated marker for all stages of gastric tumor. Also, CDH1 was selected because this was already known to be functionally downregulated in the poorly differentiated phenotype GC.
Polymerase chain reaction Genomic DNA (gDNA) was extracted using QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). The methylation status of the gDNA was then characterized by first performing bisulfite conversion on the gDNA using EpiTect bisulfite kit (Qiagen). These two steps were carried out according to the supplied protocols. PCR was carried out twice except for LINE-1, with the PCR product of the first reaction being used as template for the second reaction. In the first PCR reaction, forward (F) and reverse (R) primers were used while in the second reaction F, reverse-universal (RU) and biotin-universal (BU) primers were used (Table 1, Supplementary data). The first PCR reaction conditions were 94 ◦ C for 30 s, 55—58 ◦ C for 30 s and 72 ◦ C for 30 s, 45 cycles. The second PCR was performed under the same conditions, except for the different annealing temperature (60 ◦ C for all primers). The reagents for PCR were HotStarTaq Plus PCR mixture (Qiagen).
Pyrosequencing Pyrosequencing was then performed on the PCR products according to the supplied protocols. For pyrosequencing, PyroMark Gold Q24 Reagents (Qiagen) and PyroMark Q24 machine (Qiagen) were used. The sequencing primers are listed with PCR primers in Table 1, Supplementary data. Through pyrosequencing, methylation levels were measured at 3—5 CpG sites of the promoter region in each gene and the results were shown as mean values.
3
Quantitative real-time PCR We measured mRNA expression levels of two genes, CDH1 and RORA by qPCR. These 2 genes were selected for Quantitative Real-Time PCR because RORA can be associated with Wnt signaling pathway and hypermethylation of CDH1 leads to gene silencing in GC [11,22]. Among the 5 genes studied, we supposed that these 2 genes were more likely functionally relevant in the gastric carcinogenesis, and examined mRNA expressions using cancerous and non-cancerous tissues. RNA was extracted using RNeasy Plus Mini kit (Qiagen) from tissue samples of several patients in the validation group, already used for pyrosequencing. We synthesized cDNA using QuantiTect Reverse Transcription Kit (Qiagen) and performed qPCR under the following conditions: denaturation step at 95 ◦ C for 15 s, annealing and extension step at 60 ◦ C for 60 s, 40 cycles. Pre-designed primer sets were used [29,34,35]. Paired samples were prepared from the same patient to compare the differences in expression levels between non-cancer (adjacent to cancer) versus tumor tissues. Real-time PCR was carried out using ABI 7900HT system (Applied Biosystems, Carlsbad, Cal). The SYBR green mixture used was made from KAPA SYBR FAST qPCR Kit (KAPA Biosystems, Woburn, MA). The data was analyzed using the comparative Ct method.
Statistical analysis All statistical analysis were carried out using SPSS for Windows, version 17 (SPSS, Inc, Chicago, IL) and PRISM software for Windows, version 5 (GraphPad Prism, Inc, San Diego, CA). Methylation level (in %), taken as a continuous variable was analyzed and associated with stages or differentiation of the different GC types (95% confidence intervals) for each gene. The z score was utilized to normalize the data of the methylation levels. The z score was calculated according to the following formula: z score = (methylation level of each sample — mean value of methylation level)/standard deviation of methylation level. Methylation levels of the experimental group were analyzed using one way ANOVA test according to cancer stage and type. Methylation levels of intestinal type and diffuse type including all types of each were compared using one way ANOVA test and t-test. As subtypes of diffuse GC, PDC and SRCC, were compared with t-test. Methylation levels in H. pylori positive and negative tissues were also compared using t-test. The results of qPCR were analyzed using paired sample t-test. Correlation between methylation patterns and mRNA expression for CDH1 and RORA was analyzed using Spearman’s correlation test.
Results Clinicopathologic characteristics of patients Clinical features of patients are summarized according to tumor types involving gastric dysplasia, early GC and advanced gastric cancer (AGC, Table 1). PDCs are associated with the patients with younger age. This means that undifferentiated cancers are more common in younger patients
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model
ARTICLE IN PRESS
CLINRE-583; No. of Pages 8
4
M. Yang et al. Table 1
Clinicopathologic characteristics of patients. Experimental group
Sex Male Female Age Mean ± SD Stage GD EGC AGC Location Antrum to angle Body Antrum to body HP infection Positive Negative
Validation group
GD
WD
MD
PD
SRC
WD
MD
PD
SRC
30 9 64 ± 8
10 4 67 ± 9
15 6 62 ± 15
24 9 63 ± 9
16 4 55 ± 11
12 13 70 ± 9
21 4 69 ± 11
26 5 66 ± 12
15 19 60 ± 15
39 —
— 12 2
— 10 11
— 9 24
— 4 16
— 18 7
— 12 13
— 5 26
— 9 25
33
11
12
17
13
19
21
13
12
4 2
3 —
9 —
16 —
7 —
5 1
4 —
14 4
22 —
26 13
5 9
9 12
18 15
8 12
12 13
5 20
12 19
14 20
GD: gastric dysplasia; WD: well-differentiated gastric cancer; MD: moderately differentiated gastric cancer; PDC: poorly differentiated gastric cancer; SRCC: signet ring cell carcinoma; EGC: early gastric cancer; AGC: advanced gastric cancer.
than that of older patients. H. pylori infection and differentiation of GC appears to be irrelevant.
Different methylation profiles of differentiated cancer from undifferentiated cancer in the experimental group Methylation levels were quantitatively measured and analyzed according to the status of tumor progression and differentiation. Data obtained from the experimental group showed that the methylation levels of normal mucosa were highly conserved in all genes (Fig. 1), whereas methylation patterns were changed as the disease progressed from gastric dysplasia to advanced GC. GDNF and RORA showed high methylation levels in precancerous lesion but the levels decreased as the cancer advanced. KLF had no significant changes but showed higher methylation levels in advanced GC compared with normal. In contrast, methylation levels of LINE-1 had no significant difference between normal and precancerous stage, but LINE-1 became hypomethylated as the cancer developed. GDNF and RORA showed significantly lower methylation levels in diffuse type GC compared with those of dysplasia and intestinal type GC, while KLF7 showed increased methylation levels in diffuse type GC compared to normal tissues only. LINE-1 was hypomethylated in both types of GC compared with normal and gastric dysplasia (Supplementary data, Fig. 1). We also compared the methylation levels between two types of the diffuse gastric cancer, PDC and SRCC (Fig. 2). There were no significant differences except for LINE-1, whose methylation level was higher in SRCC.
Validation of different methylation profiles according to the cancer differentiation In the validation group, the data were analyzed according to the differentiation status involving well-differentiated (WD), moderately differentiated (MD), poorly differentiated (PD) and SRCC. GDNF and RORA showed similar methylation patterns to the experimental group (Fig. 3). Their methylation levels were remarkably low in MD and SRCC. Specifically, the methylation levels of SRCCs were conspicuously low compared to other types. MINT25 had no statistical significance but showed a similar pattern and was significantly hypomethylated in SRCCs. However, CDH1 was hypermethylated in SRCCs. KLF7 had a significant difference between PDC and SRCC. Other types had no differences. LINE-1 showed different methylation levels between WD GCs and PD GCs. GDNF, RORA and MINT25 showed significantly lower methylation levels in diffuse type GCs (Supplementary data, Fig. 2). On the other hand, the methylation levels of CDH1 were significantly higher. KLF7 and LINE-1 showed inconsistent methylation pattern in the experimental group. Interestingly, CDH1 showed highly methylated in SRCC (Fig. 4). In another four genes, except LINE-1, the methylation levels of SRCCs were significantly decreased compared with PDCs.
mRNA expressions of CDH1 and RORA in non-cancerous and cancerous tissues We investigated mRNA expressions of CDH1 and RORA in the 19 pairs of samples that had been utilized for pyrosequencing in the validation group. We did not assess mRNA
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
Methylation profiles by gastric cancer subtypes
5
Figure 1 Methylation patterns in test group for each cancer stage. N: normal mucosa; GD: gastric dysplasia; EGC: early gastric cancer; AGC: advanced gastric cancer. (* P < 0.05).
Figure 2
Comparison of the differences in methylation patterns between PDC and SRCC in experimental group (* P < 0.05).
expressions of the following 3 genes because these GDNF and KLF7 are unlikely functionally relevant in gastric carcinogenesis, and MINT25 is an alternatively spliced form of the CABIN1 gene and was therefore not studied. Expression levels of mRNA were measured by quantitative real-time PCR and analyzed using comparative Ct method. The mRNA expressions in the cancer tissues were lower than those from the non-cancer tissues for both genes (Fig. 5). Although there was no significant difference in CDH1 expression levels between non-cancer and cancer tissues (P = 0.079), mRNA
expressions of RORA were significantly lower in cancer tissue (P = 0.041).
Helicobacter pylori infection and methylation pattern Histopathologic examination was used to determine if patients had H. pylori infection or not. Though many studies have reported that H. pylori infection is associated
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
6
M. Yang et al.
Figure 3
Figure 4
Methylation patterns upon four histologic differentiation status in the validation group (* P < 0.05).
Comparison of the differences in methylation patterns between PDC and SRCC in validation group (* P < 0.05).
Figure 5
mRNA expression of CDH1 and RORA (* P < 0.05).
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
Methylation profiles by gastric cancer subtypes with aberrant DNA methylation, we could not find any significant correlation between H. pylori infection and methylation pattern, except for GDNF in the validation group (Table 2, Supplementary data).
Discussion Recently, genome-wide epigenetic analysis have been performed to determine epigenetic mutations in diverse cancers [13,19]. Currently, much epigenetic research in gastric carcinogenesis is ongoing in Asia. Even though papers detailing the step-by-step methylation changes of GC have been on the increase in recent years, however, there are few large-scale studies regarding the role of gene methylation in carcinogenesis of SRCC. Particularly, cancer specific biomarkers are considered helpful for prognosis since some cancers, like undifferentiated GC have poor prognosis and these markers would help in distinguishing them from other cancers. In cancer cells, two conflicting epigenetic events can coexist together. These two are global hypomethylation and regional hypermethylation. Global hypomethylation frequently occurs in DNA repetitive sequences and is believed to cause chromosomal instability and protooncogene activation [17,19]. In this study, LINE-1 tends to be hypomethylated in advanced GC (Fig. 1). Specifically, hypomethylation of LINE-1 was frequently observed in PDC (Fig. 2). This would explain the chromosomal instability that is more commonly associated with PDC or undifferentiated cancers. A previous study determined 6 genes that are prone to hypermethylation (MINT25, GDNF, RORA, PRDM5, ADAM23, MLF1) in gastric neoplasia [18]. In particular, GDNF and MINT25 showed higher methylation patterns in gastric tumor samples regardless of tumor stage and can therefore be used as biomarkers for gastric tumors [18]. Similarly, in this study, methylation levels of GDNF and RORA are dramatically high in precancerous stages. In the advanced cancer stages, they showed decreased methylation levels as compared with precancerous stages but not with normal (Supplementary fig. 2). Therefore, these markers can be utilized to help in the diagnosis of early stage gastric neoplasia. Furthermore, many genes, including MINT25 are known to display specific methylation patterns [34]. In line with this result, GDNF, RORA and MINT25 show significant differences between PDC and SRCC in this study (Fig. 4). These results suggest that GDNF, RORA and MINT25 can be adopted as sensitive biomarkers to distinguish PDC from SRCC as well as biomarkers for the detection of early stage of gastric tumor. Interestingly, CDH1 also has significantly higher methylation levels in SRCCs compared to other cancers (Fig. 3), which is consistent with the previous results that CDH1 plays a key role as a tumor suppressor gene in hereditary diffuse type GCs [10,11,13,35,36]. In other words, the suppression of CDH1 gene expression by methylation can promote carcinogenesis [8,9,12,22]. This fact is more evident in undifferentiated GCs, especially SRCCs than in another subtypes of GCs. Therefore, the methylation of CDH1 is one of the molecular mechanisms for the development of undifferentiated cancers. Accordingly, CDH1 has the potential to be a specific biomarker for undifferentiated cancers and for
7 poor prognosis. Also, these findings strongly suggests that the different set of methylation genes are involved in the carcinogenesis pathway of the two subtypes of undifferentiated GCs. We examined relative mRNA expressions for CDH1 and RORA between non-cancerous and cancerous gastric tissues (Fig. 5). RORA showed a significantly different mRNA expression (P = 0.041), whereas CDH1 demonstrated only marginal difference (P = 0.079). These mean that hypermethylation of CDH1 may be related to downregulation of its mRNA expression which corresponds to the inactivation of e-cadherin genes that commonly occurs due to promoter methylation [19]. Moreover, RORA showed differential downregulation of mRNA in the tumor tissues, indicating this RORA methylation has a potential role in functional relevance of this gene. However, additional studies using in vitro cell lines and in vivo tissues are necessary, such as protein expression analysis and immunochemical staining to establish these results. Many papers report that H. pylori infection leads to gastric carcinogenesis by causing aberrant DNA methylation [9,19,21,22]. Our study, however, could not find any relationship between H. pylori infection and DNA methylation (Table 2, Supplementary data). There have been reported 2 types of methylation marker groups related to H. pylori infection; one group is H. pylori related methylation markers, which show common hypermethylation in the H. pylori positive chronic gastritis, and the other marker group is not dependent on H. pylori infection but more related to gastric tumor itself. Through this study, we found that the 5 methylation markers (GDNF, RORA, KLF7, CDH1 and MINT25) were not differentially methylated according to H. pylori infection status, suggesting these markers corresponded to the latter. Collectively, at the molecular level, the intestinal type of GCs have different epigenetic methylation profiles compared to the diffuse one. Furthermore, SRCCs have different methylation profiles compared with PDCs, suggesting a unique gastric carcinogenesis pathway. These results are consistent with the recent report that invasion and metastasis patterns are different between PDC and SRCC [24—26]. However, the results should be interpreted with caution because of the small number of samples and the retrospective nature of this study. Further studies are needed to assess the roles of these methylation markers in the prognosis prediction in undifferentiated GCs using follow-up cohort.
Disclosure of interest The authors declare that they have no conflicts of interest concerning this article.
Acknowledgments This research was supported by Basic Science Research Program through the National Research Foundation (20120004734), and the research fund from Yonsei University Wonju College of Medicine.
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
8
M. Yang et al.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.clinre.2014.03.017.
References [1] Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2007. CA Cancer J Clin 2007;57:43—66. [2] Jung KW, Park S, Kong HJ, et al. Cancer statistics in Korea: incidence, mortality, survival, and prevalence in 2008. Cancer Res Treat 2011;43:1—11. [3] Asaka M, Dragosics BA. Helicobacter pylori and gastric malignancies. Helicobacter 2004;9(Suppl. 1):35—41. [4] Kang HY, Kim N, Park YS, et al. Progression of atrophic gastritis and intestinal metaplasia drives Helicobacter pylori out of the gastric mucosa. Dig Dis Sci 2006;51:2310—5. [5] Nardone G, Morgner A. Helicobacter pylori and gastric malignancies. Helicobacter 2003;8(Suppl. 1):44—52. [6] Choi IS, Wu TT. Epigenetic alterations in gastric carcinogenesis. Cell Res 2005;15:247—54. [7] Yasui W, Sentani K, Motoshita J, Nakayama H. Molecular pathobiology of gastric cancer. Scand J Surg 2006;95:225—31. [8] Resende C, Ristimaki A, Machado JC. Genetic and epigenetic alteration in gastric carcinogenesis. Helicobacter 2010;15(Suppl. 1):34—9. [9] Tamura G. Alterations of tumor suppressor and tumor-related genes in the development and progression of gastric cancer. World J Gastroenterol 2006;12:192—8. [10] Vogiatzi P, Vindigni C, Roviello F, Renieri A, Giordano A. Deciphering the underlying genetic and epigenetic events leading to gastric carcinogenesis. J Cell Physiol 2007;211:287—95. [11] Humar B, Blair V, Charlton A, More H, Martin I, Guilford P. Ecadherin deficiency initiates gastric signet ring cell carcinoma in mice and man. Cancer Res 2009;69:2050—6. [12] Ushijima T, Sasako M. Focus on gastric cancer. Cancer Cell 2004;5:121—5. [13] Panani AD. Cytogenetic and molecular aspects of gastric cancer: clinical implications. Cancer Lett 2008;266:99—115. [14] Kim WH, Park CK, Kim YB, et al. A standardized pathology report for gastric cancer. Korean J Pathol 2005;39:106—13. [15] Marrelli D, Roviello F, De Stefano A, et al. Risk factors for liver metastases after curative surgical procedures for gastric cancer: a prospective study of 208 patients treated with surgical resection. J Am Coll Surg 2004;198:51—8. [16] Crew KD, Neugut AI. Epidemiology of gastric cancer. World J Gastroenterol 2006;12:354—62. [17] Laird PW. The power and the promise of DNA methylation markers. Nat Rev Cancer 2003;3:253—66. [18] Watanabe Y, Kim HS, Castoro RJ, et al. Sensitive and specific detection of early gastric cancer with DNA methylation analysis of gastric washes. Gastroenterology 2009;136:2149—58. [19] Compare D, Rocco A, Liguori E, et al. Global DNA hypomethylation is an early event in Helicobacter pylori-related gastric carcinogenesis. J Clin Pathol 2011.
[20] Kang GH, Lee S, Kim JS, Jung HY. Profile of aberrant CpG island methylation along multistep gastric carcinogenesis. Lab Invest 2003;83:519—26. [21] Ksiaa F, Ziadi S, Amara K, Korbi S, Trimeche M. Biological significance of promoter hypermethylation of tumor-related genes in patients with gastric carcinoma. Clin Chim Acta 2009;404:128—33. [22] Yamamoto E, Suzuki H, Takamaru H, Yamamoto H, Toyota M, Shinomura Y. Role of DNA methylation in the development of diffuse-type gastric cancer. Digestion 2011;83:241—9. [23] Bariol C, Suter C, Cheong K, et al. The relationship between hypomethylation and CpG island methylation in colorectal neoplasia. Am J Pathol 2003;162:1361—70. [24] Kim JH, Lee YC, Kim H, et al. Endoscopic resection for undifferentiated early gastric cancer. Gastrointest Endosc 2009;69:e1—9. [25] Tong JH, Sun Z, Wang ZN, et al. Early gastric cancer with signet-ring cell histologic type: risk factors of lymph node metastasis and indications of endoscopic surgery. Surgery 2011;149:356—63. [26] Kim HM, Pak KH, Chung MJ, et al. Early gastric cancer of signet ring cell carcinoma is more amenable to endoscopic treatment than is early gastric cancer of poorly differentiated tubular adenocarcinoma in select tumor conditions. Surg Endosc 2011;25:3087—93. [27] Hiraki M, Kitajima Y, Koga Yet al. Aberrant gene methylation is a biomarker for the detection of cancer cells in peritoneal wash samples from advanced gastric cancer patients. Ann Surg Oncol 2011;18:3013—9. [28] Jetten AM. Retinoid-related orphan receptors (RORs): critical roles in development, immunity, circadian rhythm, and cellular metabolism. Nucl Recept Signal 2009;7:e003. [29] Lee JM, Kim IS, Kim H, Lee JS, Kim K, Yim HY, et al. RORalpha attenuates Wnt/beta-catenin signaling by PKCalpha-dependent phosphorylation in colon cancer. Mol Cell 2010;37:183—95. [30] Airaksinen MS, Saarma M. The GDNF family: signalling, biological functions and therapeutic value. Nat Rev Neurosci 2002;3:383—94. [31] Liu H, Ma Q, Li J. High glucose promotes cell proliferation and enhances GDNF and RET expression in pancreatic cancer cells. Mol Cell Biochem 2011;347:95—101. [32] Moore DL, Apara A, Goldberg JL. Kruppel-like transcription factors in the nervous system: Novel players in neurite outgrowth and axon regeneration. Mol Cell Neurosci 2011;47:233—43. [33] Humar B, Guilford P. Hereditary diffuse gastric cancer: a manifestation of lost cell polarity. Cancer Sci 2009;100:1151—7. [34] Odawara H, Iwasaki T, Horiguchi J, Rokutanda N, Hirooka K, Miyazaki W, et al. Activation of aromatase expression by retinoic acid receptor-related orphan receptor (ROR) alpha in breast cancer cells: identification of a novel ROR response element. J Biol Chem 2009;284:17711—9. [35] Watanabe T, Miura T, Degawa Y, Fujita Y, Inoue M, Kawaguchi M, et al. Comparison of lung cancer cell lines representing four histopathological subtypes with gene expression profiling using quantitative real-time PCR. Cancer Cell Int 2010;10:2. [36] Baylin SB. DNA methylation and gene silencing in cancer. Nat Clin Pract Oncol 2005;2(Suppl. 1):S4—11.
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
ARTICLE IN PRESS
Clinics and Research in Hepatology and Gastroenterology (2014) xxx, xxx—xxx
Available online at
ScienceDirect www.sciencedirect.com
ORIGINAL ARTICLE
Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis Misuk Yang a, Hyun-Soo Kim a,∗, Mee Yon Cho b a
Division of Gastroenterology, Department of Internal Medicine, Yonsei University Wonju College of Medicine, Wonju, Republic of Korea b Department of Pathology, Yonsei University Wonju College of Medicine, Wonju, Republic of Korea
Summary Objective: Gastric cancer (GC) is histologically classified into intestinal type and diffuse type, and diffuse type cancer can be further subdivided into poorly differentiated carcinoma (PDC) and signet ring cell carcinoma (SRCC). Recent evidence suggests that early SRCC is an initial, differentiated form of diffuse GC that may evolve into PDC. This study aimed at identifying the molecular features of epigenetic methylation changes in histologic differentiation status of GC. Methods: Included in this study are 149 samples of paraffin-embedded tissues and 115 fresh endoscopically biopsied tissues. Multiple paraffin tissues involving normal (n = 22), dysplasias (GDs, n = 39), differentiated cancers (DCs, n = 35), PDCs (n = 33) and SRCCs (n = 20) were included as an experimental group. For the validation group, endoscopically biopsied tissues of DCs (n = 50), PDCs (n = 31), and SRCs (n = 34) were analyzed. DNAs, isolated from each group were analyzed to determine the methylation status of 6 genes (GDNF, RORA, MINT25, KLF7, CDH1, LINE-1) using pyrosequencing. Results: LINE-1 was hypomethylated in GCs compared to normal and GD. GDNF, RORA and MINT25 were more hypermethylated in intestinal type GCs than those of diffuse type GCs, whereas CDH1 showed opposite patterns of methylation. Among diffuse type GCs, SRCCs showed lower level of methylation for GDNF, RORA, MINT25 and KLF7, and higher level for CDH1 compared to PDCs. Conclusions: In conclusion, intestinal type of GCs shows different epigenetic methylation profiles compared to the diffuse one. Moreover, SRCCs have different methylation profiles compared with PDCs, suggesting a unique molecular pathway in the gastric carcinogenesis. © 2014 Elsevier Masson SAS. All rights reserved.
∗ Corresponding author. Digestive Disease Center, Department of Internal Medicine, Yonsei University Wonju College of Medicine, 162, Ilsan-dong, 220-701 Wonju, South Korea. Tel.: +82 33 741 1229; fax: +82 33 741 1228. E-mail address: [email protected] (H.-S. Kim).
http://dx.doi.org/10.1016/j.clinre.2014.03.017 2210-7401/© 2014 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
2
M. Yang et al.
Introduction
Materials and methods
Gastric adenocarcinoma is one of the most common diseases in the world, and the main cause of cancer death in Asia [1]. In particular, the incidence rate of gastric cancer (GC; 41.1—44.3/100,000) was barely changed from 1999 to 2008, and the GC was the third leading cause of the cancer death in Korea [2]. Environmental factors including diet, nutritional conditions, Helicobacter pylori (H. pylori) infection and smoking are known as risk factors for GC development [3—5]. Furthermore, genetic and epigenetic factors also play a critical role in gastric carcinogenesis [6—10]. These genetic factors include single-nucleotide polymorphisms (SNPs) and DNA mutations [8—10], whereas epigenetic factors include posttranslational modifications, aberrant DNA methylation and histone modification [7,10]. Lauren classification is the most widely used method that distinguishes two histological types according to microscopic morphology of gastric adenocarcinoma [9—12]. Intestinal type is the well-differentiated (WD) tumor with cohesive tumor cells with gland-like tubular structures. Diffuse type is undifferentiated tumors that infiltrate through the stomach wall [9,10,13]. However, many gastric tumors are sorted as mixed type since some are not classified into either type of gastric tumors [10,14]. While intestinal type of GCs frequently metastasizes to the liver, diffuse type GCs can be easily disseminated to the peritoneum [10,15]. Gastric carcinogenesis progresses through serial steps involving atrophy, metaplasia, gastric dysplasia (GD) and carcinoma. These steps are more commonly seen in the intestinal type GC [7,10]. Molecular studies have provided evidence that GC results not only from the combined effects of environmental aspects and genetic modifications, but also from the cumulative genetic and epigenetic changes that play pivotal roles in tumorigenesis [8,10,16]. In GC, for instance, aberrant DNA methylations occur more frequently than mutations [12,17,18]. DNA methylation changes even occurring in nonneoplastic gastric epithelial cells, which suggest that the epigenetic methylation is an early event in GC [7,19—22]. Also, global DNA hypomethylation occurs in gastric or other tissues during the early stages of tumor growth [9,23]. GC proceeds through serial steps. Intestinal GC progresses through sequential steps after several genetic and epigenetic changes, while diffuse GC progresses directly to early cancer after similar genetic and epigenetic changes with intestinal GC [7,10,22]. Judging from the fact that the pathway is different for each type of cancer, genes associated with the cancer seem to affect different aspects of the changes involved in each pathway. Recent studies show that the pattern of lymph node metastasis is different between poorly differentiated carcinoma (PDC) and signet ring cell carcinoma (SRCC) [24,25]. Additionally, they also have different mechanisms of tumor invasion [26,27]. These reports have indicated that SRCC and PDC have different origins. In this study, therefore, we investigated differences of DNA methylation patterns at the molecular level especially in diffuse type GC, including SRCC and PDC.
Sample collection Gastric tumor tissues of the experimental group were obtained from Korean patients who had surgical or endoscopic resection at Yonsei University Wonju Severance Christian Hospital (Wonju, Korea) between May 2006 and December 2010. The 149 samples of paraffin-embedded tissues including 22 normal mucosa from healthy subjects, and 39 GDs, 35 differentiated gastric cancers (DGCs), 33 PDCs and 20 SRCCs from patients, were retrospectively examined. Among these samples recruited during the period, the pathologist randomly selected the patients who were diagnosed as GD or DGCs using the table of random sampling for numbers. Tissues were sent to pathologists for histopatologic diagnosis and fixed with paraffin. These paraffin-embedded tissue blocks were sliced for serial sections, and each section was microdissected by an expert pathologist to distinguish the tumor regions. The extracted DNA for each sample was therefore more than 70% homogeneous as described previously [18]. For the validation group, we collected 115 tumor tissues obtained between October 2008 and February 2011 at the same hospital. Because a large number of samples had been collected for intestinal type cancer, randomly selected samples were used. Due to the small number of available diffuse GC samples, however, all of them available were used. Small portions of these samples were used for histopathologic analysis and the rest of them were frozen immediately after biopsy. Serial sections of specimens were assessed for the presence of H. pylori by a modified Giemsa stain. Two biopsy specimens, one from the antrum and one from the corpus, were tested from each patient. The specimens were subjected to a rapid urease test (CLO test, Kimberly-Clark, USA), to detect the presence of H. pylori urease. A patient was defined as being H. pylori positive on the basis of a positive histological examination or a positive CLO test. Basic and clinicopathologic information about each patient were used as raw data. This study was approved by the institutional review board of Yonsei University Wonju Severance Christian Hospital (CR106021). We also notified all patients and received consents.
Genes selection for methylation study We evaluated DNA methylation in the promoter of 6 genes for this study. Retinoic acid-related orphan receptor-alpha (RORA) may functionally associated with Wnt signaling pathway that is involved in oncogenic systems by attenuating the pathway [28,29]. Glial cell derived neurotrophic factor (GDNF) is a small protein and interaction of upregulated GDNF and RET ligand-receptor are possibly involved in the glucose-induced cancer progression [30,31]. The Kruppellike family (KLF) of transcription factors have been shown to play pivotal roles in tumorigenesis of various tissues [32]. The promotor of KLF7 was frequently methylated in the diffuse type gastric cancers (unpublished data). The expression of CDH1, an important e-cadherin protein for cell
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
Methylation profiles by gastric cancer subtypes adhesion, is frequently downregulated in sporadic tumors and is associated with the poorly differentiated phenotype. Hypermethylation of this gene leads to gene silencing in GC [11,33]. Global DNA hypomethylation of DNA repetitive elements within the human genome, such as long interspersed nuclear element-1 (LINE-1) is a common finding in cancer and is believed to cause chromosomal instability and protooncogene activation [12,19]. Three genes (RORA, GDNF and MINT25) of the six were identified as sensitive molecular markers of early GC [18,27]. In the previous study [18], we found several kinds of methylation markers according to stepwise gastric tumor progression; type 1 markers, showing consistently high levels of methylation in both gastric dysplasia and cancer (GDNF, MINT25); type 2 markers, showing high levels of methylation in early gastric cancer and gastric dysplasia but decreased levels in advanced gastric cancer (RORA); type 3 markers, revealing higher methylation as the gastric tumor progresses. For the experimental group, we would like to focus on the methylation profile changes from normal to gastric cancer progression and selected 3 genes (GDNF, RORA, and KLF7) and LINE-1 as representatives of above 3 types of markers. For the validation group, however, we focused on the comparison of methylation profiles between PDC and SRCC based on the preliminary dataset from the experimental group. Therefore, we added 2 more methylation markers (MINT25 and CDH1) as MINT25 was confirmed as a general hypermethylated marker for all stages of gastric tumor. Also, CDH1 was selected because this was already known to be functionally downregulated in the poorly differentiated phenotype GC.
Polymerase chain reaction Genomic DNA (gDNA) was extracted using QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). The methylation status of the gDNA was then characterized by first performing bisulfite conversion on the gDNA using EpiTect bisulfite kit (Qiagen). These two steps were carried out according to the supplied protocols. PCR was carried out twice except for LINE-1, with the PCR product of the first reaction being used as template for the second reaction. In the first PCR reaction, forward (F) and reverse (R) primers were used while in the second reaction F, reverse-universal (RU) and biotin-universal (BU) primers were used (Table 1, Supplementary data). The first PCR reaction conditions were 94 ◦ C for 30 s, 55—58 ◦ C for 30 s and 72 ◦ C for 30 s, 45 cycles. The second PCR was performed under the same conditions, except for the different annealing temperature (60 ◦ C for all primers). The reagents for PCR were HotStarTaq Plus PCR mixture (Qiagen).
Pyrosequencing Pyrosequencing was then performed on the PCR products according to the supplied protocols. For pyrosequencing, PyroMark Gold Q24 Reagents (Qiagen) and PyroMark Q24 machine (Qiagen) were used. The sequencing primers are listed with PCR primers in Table 1, Supplementary data. Through pyrosequencing, methylation levels were measured at 3—5 CpG sites of the promoter region in each gene and the results were shown as mean values.
3
Quantitative real-time PCR We measured mRNA expression levels of two genes, CDH1 and RORA by qPCR. These 2 genes were selected for Quantitative Real-Time PCR because RORA can be associated with Wnt signaling pathway and hypermethylation of CDH1 leads to gene silencing in GC [11,22]. Among the 5 genes studied, we supposed that these 2 genes were more likely functionally relevant in the gastric carcinogenesis, and examined mRNA expressions using cancerous and non-cancerous tissues. RNA was extracted using RNeasy Plus Mini kit (Qiagen) from tissue samples of several patients in the validation group, already used for pyrosequencing. We synthesized cDNA using QuantiTect Reverse Transcription Kit (Qiagen) and performed qPCR under the following conditions: denaturation step at 95 ◦ C for 15 s, annealing and extension step at 60 ◦ C for 60 s, 40 cycles. Pre-designed primer sets were used [29,34,35]. Paired samples were prepared from the same patient to compare the differences in expression levels between non-cancer (adjacent to cancer) versus tumor tissues. Real-time PCR was carried out using ABI 7900HT system (Applied Biosystems, Carlsbad, Cal). The SYBR green mixture used was made from KAPA SYBR FAST qPCR Kit (KAPA Biosystems, Woburn, MA). The data was analyzed using the comparative Ct method.
Statistical analysis All statistical analysis were carried out using SPSS for Windows, version 17 (SPSS, Inc, Chicago, IL) and PRISM software for Windows, version 5 (GraphPad Prism, Inc, San Diego, CA). Methylation level (in %), taken as a continuous variable was analyzed and associated with stages or differentiation of the different GC types (95% confidence intervals) for each gene. The z score was utilized to normalize the data of the methylation levels. The z score was calculated according to the following formula: z score = (methylation level of each sample — mean value of methylation level)/standard deviation of methylation level. Methylation levels of the experimental group were analyzed using one way ANOVA test according to cancer stage and type. Methylation levels of intestinal type and diffuse type including all types of each were compared using one way ANOVA test and t-test. As subtypes of diffuse GC, PDC and SRCC, were compared with t-test. Methylation levels in H. pylori positive and negative tissues were also compared using t-test. The results of qPCR were analyzed using paired sample t-test. Correlation between methylation patterns and mRNA expression for CDH1 and RORA was analyzed using Spearman’s correlation test.
Results Clinicopathologic characteristics of patients Clinical features of patients are summarized according to tumor types involving gastric dysplasia, early GC and advanced gastric cancer (AGC, Table 1). PDCs are associated with the patients with younger age. This means that undifferentiated cancers are more common in younger patients
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model
ARTICLE IN PRESS
CLINRE-583; No. of Pages 8
4
M. Yang et al. Table 1
Clinicopathologic characteristics of patients. Experimental group
Sex Male Female Age Mean ± SD Stage GD EGC AGC Location Antrum to angle Body Antrum to body HP infection Positive Negative
Validation group
GD
WD
MD
PD
SRC
WD
MD
PD
SRC
30 9 64 ± 8
10 4 67 ± 9
15 6 62 ± 15
24 9 63 ± 9
16 4 55 ± 11
12 13 70 ± 9
21 4 69 ± 11
26 5 66 ± 12
15 19 60 ± 15
39 —
— 12 2
— 10 11
— 9 24
— 4 16
— 18 7
— 12 13
— 5 26
— 9 25
33
11
12
17
13
19
21
13
12
4 2
3 —
9 —
16 —
7 —
5 1
4 —
14 4
22 —
26 13
5 9
9 12
18 15
8 12
12 13
5 20
12 19
14 20
GD: gastric dysplasia; WD: well-differentiated gastric cancer; MD: moderately differentiated gastric cancer; PDC: poorly differentiated gastric cancer; SRCC: signet ring cell carcinoma; EGC: early gastric cancer; AGC: advanced gastric cancer.
than that of older patients. H. pylori infection and differentiation of GC appears to be irrelevant.
Different methylation profiles of differentiated cancer from undifferentiated cancer in the experimental group Methylation levels were quantitatively measured and analyzed according to the status of tumor progression and differentiation. Data obtained from the experimental group showed that the methylation levels of normal mucosa were highly conserved in all genes (Fig. 1), whereas methylation patterns were changed as the disease progressed from gastric dysplasia to advanced GC. GDNF and RORA showed high methylation levels in precancerous lesion but the levels decreased as the cancer advanced. KLF had no significant changes but showed higher methylation levels in advanced GC compared with normal. In contrast, methylation levels of LINE-1 had no significant difference between normal and precancerous stage, but LINE-1 became hypomethylated as the cancer developed. GDNF and RORA showed significantly lower methylation levels in diffuse type GC compared with those of dysplasia and intestinal type GC, while KLF7 showed increased methylation levels in diffuse type GC compared to normal tissues only. LINE-1 was hypomethylated in both types of GC compared with normal and gastric dysplasia (Supplementary data, Fig. 1). We also compared the methylation levels between two types of the diffuse gastric cancer, PDC and SRCC (Fig. 2). There were no significant differences except for LINE-1, whose methylation level was higher in SRCC.
Validation of different methylation profiles according to the cancer differentiation In the validation group, the data were analyzed according to the differentiation status involving well-differentiated (WD), moderately differentiated (MD), poorly differentiated (PD) and SRCC. GDNF and RORA showed similar methylation patterns to the experimental group (Fig. 3). Their methylation levels were remarkably low in MD and SRCC. Specifically, the methylation levels of SRCCs were conspicuously low compared to other types. MINT25 had no statistical significance but showed a similar pattern and was significantly hypomethylated in SRCCs. However, CDH1 was hypermethylated in SRCCs. KLF7 had a significant difference between PDC and SRCC. Other types had no differences. LINE-1 showed different methylation levels between WD GCs and PD GCs. GDNF, RORA and MINT25 showed significantly lower methylation levels in diffuse type GCs (Supplementary data, Fig. 2). On the other hand, the methylation levels of CDH1 were significantly higher. KLF7 and LINE-1 showed inconsistent methylation pattern in the experimental group. Interestingly, CDH1 showed highly methylated in SRCC (Fig. 4). In another four genes, except LINE-1, the methylation levels of SRCCs were significantly decreased compared with PDCs.
mRNA expressions of CDH1 and RORA in non-cancerous and cancerous tissues We investigated mRNA expressions of CDH1 and RORA in the 19 pairs of samples that had been utilized for pyrosequencing in the validation group. We did not assess mRNA
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
Methylation profiles by gastric cancer subtypes
5
Figure 1 Methylation patterns in test group for each cancer stage. N: normal mucosa; GD: gastric dysplasia; EGC: early gastric cancer; AGC: advanced gastric cancer. (* P < 0.05).
Figure 2
Comparison of the differences in methylation patterns between PDC and SRCC in experimental group (* P < 0.05).
expressions of the following 3 genes because these GDNF and KLF7 are unlikely functionally relevant in gastric carcinogenesis, and MINT25 is an alternatively spliced form of the CABIN1 gene and was therefore not studied. Expression levels of mRNA were measured by quantitative real-time PCR and analyzed using comparative Ct method. The mRNA expressions in the cancer tissues were lower than those from the non-cancer tissues for both genes (Fig. 5). Although there was no significant difference in CDH1 expression levels between non-cancer and cancer tissues (P = 0.079), mRNA
expressions of RORA were significantly lower in cancer tissue (P = 0.041).
Helicobacter pylori infection and methylation pattern Histopathologic examination was used to determine if patients had H. pylori infection or not. Though many studies have reported that H. pylori infection is associated
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
6
M. Yang et al.
Figure 3
Figure 4
Methylation patterns upon four histologic differentiation status in the validation group (* P < 0.05).
Comparison of the differences in methylation patterns between PDC and SRCC in validation group (* P < 0.05).
Figure 5
mRNA expression of CDH1 and RORA (* P < 0.05).
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
Methylation profiles by gastric cancer subtypes with aberrant DNA methylation, we could not find any significant correlation between H. pylori infection and methylation pattern, except for GDNF in the validation group (Table 2, Supplementary data).
Discussion Recently, genome-wide epigenetic analysis have been performed to determine epigenetic mutations in diverse cancers [13,19]. Currently, much epigenetic research in gastric carcinogenesis is ongoing in Asia. Even though papers detailing the step-by-step methylation changes of GC have been on the increase in recent years, however, there are few large-scale studies regarding the role of gene methylation in carcinogenesis of SRCC. Particularly, cancer specific biomarkers are considered helpful for prognosis since some cancers, like undifferentiated GC have poor prognosis and these markers would help in distinguishing them from other cancers. In cancer cells, two conflicting epigenetic events can coexist together. These two are global hypomethylation and regional hypermethylation. Global hypomethylation frequently occurs in DNA repetitive sequences and is believed to cause chromosomal instability and protooncogene activation [17,19]. In this study, LINE-1 tends to be hypomethylated in advanced GC (Fig. 1). Specifically, hypomethylation of LINE-1 was frequently observed in PDC (Fig. 2). This would explain the chromosomal instability that is more commonly associated with PDC or undifferentiated cancers. A previous study determined 6 genes that are prone to hypermethylation (MINT25, GDNF, RORA, PRDM5, ADAM23, MLF1) in gastric neoplasia [18]. In particular, GDNF and MINT25 showed higher methylation patterns in gastric tumor samples regardless of tumor stage and can therefore be used as biomarkers for gastric tumors [18]. Similarly, in this study, methylation levels of GDNF and RORA are dramatically high in precancerous stages. In the advanced cancer stages, they showed decreased methylation levels as compared with precancerous stages but not with normal (Supplementary fig. 2). Therefore, these markers can be utilized to help in the diagnosis of early stage gastric neoplasia. Furthermore, many genes, including MINT25 are known to display specific methylation patterns [34]. In line with this result, GDNF, RORA and MINT25 show significant differences between PDC and SRCC in this study (Fig. 4). These results suggest that GDNF, RORA and MINT25 can be adopted as sensitive biomarkers to distinguish PDC from SRCC as well as biomarkers for the detection of early stage of gastric tumor. Interestingly, CDH1 also has significantly higher methylation levels in SRCCs compared to other cancers (Fig. 3), which is consistent with the previous results that CDH1 plays a key role as a tumor suppressor gene in hereditary diffuse type GCs [10,11,13,35,36]. In other words, the suppression of CDH1 gene expression by methylation can promote carcinogenesis [8,9,12,22]. This fact is more evident in undifferentiated GCs, especially SRCCs than in another subtypes of GCs. Therefore, the methylation of CDH1 is one of the molecular mechanisms for the development of undifferentiated cancers. Accordingly, CDH1 has the potential to be a specific biomarker for undifferentiated cancers and for
7 poor prognosis. Also, these findings strongly suggests that the different set of methylation genes are involved in the carcinogenesis pathway of the two subtypes of undifferentiated GCs. We examined relative mRNA expressions for CDH1 and RORA between non-cancerous and cancerous gastric tissues (Fig. 5). RORA showed a significantly different mRNA expression (P = 0.041), whereas CDH1 demonstrated only marginal difference (P = 0.079). These mean that hypermethylation of CDH1 may be related to downregulation of its mRNA expression which corresponds to the inactivation of e-cadherin genes that commonly occurs due to promoter methylation [19]. Moreover, RORA showed differential downregulation of mRNA in the tumor tissues, indicating this RORA methylation has a potential role in functional relevance of this gene. However, additional studies using in vitro cell lines and in vivo tissues are necessary, such as protein expression analysis and immunochemical staining to establish these results. Many papers report that H. pylori infection leads to gastric carcinogenesis by causing aberrant DNA methylation [9,19,21,22]. Our study, however, could not find any relationship between H. pylori infection and DNA methylation (Table 2, Supplementary data). There have been reported 2 types of methylation marker groups related to H. pylori infection; one group is H. pylori related methylation markers, which show common hypermethylation in the H. pylori positive chronic gastritis, and the other marker group is not dependent on H. pylori infection but more related to gastric tumor itself. Through this study, we found that the 5 methylation markers (GDNF, RORA, KLF7, CDH1 and MINT25) were not differentially methylated according to H. pylori infection status, suggesting these markers corresponded to the latter. Collectively, at the molecular level, the intestinal type of GCs have different epigenetic methylation profiles compared to the diffuse one. Furthermore, SRCCs have different methylation profiles compared with PDCs, suggesting a unique gastric carcinogenesis pathway. These results are consistent with the recent report that invasion and metastasis patterns are different between PDC and SRCC [24—26]. However, the results should be interpreted with caution because of the small number of samples and the retrospective nature of this study. Further studies are needed to assess the roles of these methylation markers in the prognosis prediction in undifferentiated GCs using follow-up cohort.
Disclosure of interest The authors declare that they have no conflicts of interest concerning this article.
Acknowledgments This research was supported by Basic Science Research Program through the National Research Foundation (20120004734), and the research fund from Yonsei University Wonju College of Medicine.
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017
+Model CLINRE-583; No. of Pages 8
ARTICLE IN PRESS
8
M. Yang et al.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.clinre.2014.03.017.
References [1] Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2007. CA Cancer J Clin 2007;57:43—66. [2] Jung KW, Park S, Kong HJ, et al. Cancer statistics in Korea: incidence, mortality, survival, and prevalence in 2008. Cancer Res Treat 2011;43:1—11. [3] Asaka M, Dragosics BA. Helicobacter pylori and gastric malignancies. Helicobacter 2004;9(Suppl. 1):35—41. [4] Kang HY, Kim N, Park YS, et al. Progression of atrophic gastritis and intestinal metaplasia drives Helicobacter pylori out of the gastric mucosa. Dig Dis Sci 2006;51:2310—5. [5] Nardone G, Morgner A. Helicobacter pylori and gastric malignancies. Helicobacter 2003;8(Suppl. 1):44—52. [6] Choi IS, Wu TT. Epigenetic alterations in gastric carcinogenesis. Cell Res 2005;15:247—54. [7] Yasui W, Sentani K, Motoshita J, Nakayama H. Molecular pathobiology of gastric cancer. Scand J Surg 2006;95:225—31. [8] Resende C, Ristimaki A, Machado JC. Genetic and epigenetic alteration in gastric carcinogenesis. Helicobacter 2010;15(Suppl. 1):34—9. [9] Tamura G. Alterations of tumor suppressor and tumor-related genes in the development and progression of gastric cancer. World J Gastroenterol 2006;12:192—8. [10] Vogiatzi P, Vindigni C, Roviello F, Renieri A, Giordano A. Deciphering the underlying genetic and epigenetic events leading to gastric carcinogenesis. J Cell Physiol 2007;211:287—95. [11] Humar B, Blair V, Charlton A, More H, Martin I, Guilford P. Ecadherin deficiency initiates gastric signet ring cell carcinoma in mice and man. Cancer Res 2009;69:2050—6. [12] Ushijima T, Sasako M. Focus on gastric cancer. Cancer Cell 2004;5:121—5. [13] Panani AD. Cytogenetic and molecular aspects of gastric cancer: clinical implications. Cancer Lett 2008;266:99—115. [14] Kim WH, Park CK, Kim YB, et al. A standardized pathology report for gastric cancer. Korean J Pathol 2005;39:106—13. [15] Marrelli D, Roviello F, De Stefano A, et al. Risk factors for liver metastases after curative surgical procedures for gastric cancer: a prospective study of 208 patients treated with surgical resection. J Am Coll Surg 2004;198:51—8. [16] Crew KD, Neugut AI. Epidemiology of gastric cancer. World J Gastroenterol 2006;12:354—62. [17] Laird PW. The power and the promise of DNA methylation markers. Nat Rev Cancer 2003;3:253—66. [18] Watanabe Y, Kim HS, Castoro RJ, et al. Sensitive and specific detection of early gastric cancer with DNA methylation analysis of gastric washes. Gastroenterology 2009;136:2149—58. [19] Compare D, Rocco A, Liguori E, et al. Global DNA hypomethylation is an early event in Helicobacter pylori-related gastric carcinogenesis. J Clin Pathol 2011.
[20] Kang GH, Lee S, Kim JS, Jung HY. Profile of aberrant CpG island methylation along multistep gastric carcinogenesis. Lab Invest 2003;83:519—26. [21] Ksiaa F, Ziadi S, Amara K, Korbi S, Trimeche M. Biological significance of promoter hypermethylation of tumor-related genes in patients with gastric carcinoma. Clin Chim Acta 2009;404:128—33. [22] Yamamoto E, Suzuki H, Takamaru H, Yamamoto H, Toyota M, Shinomura Y. Role of DNA methylation in the development of diffuse-type gastric cancer. Digestion 2011;83:241—9. [23] Bariol C, Suter C, Cheong K, et al. The relationship between hypomethylation and CpG island methylation in colorectal neoplasia. Am J Pathol 2003;162:1361—70. [24] Kim JH, Lee YC, Kim H, et al. Endoscopic resection for undifferentiated early gastric cancer. Gastrointest Endosc 2009;69:e1—9. [25] Tong JH, Sun Z, Wang ZN, et al. Early gastric cancer with signet-ring cell histologic type: risk factors of lymph node metastasis and indications of endoscopic surgery. Surgery 2011;149:356—63. [26] Kim HM, Pak KH, Chung MJ, et al. Early gastric cancer of signet ring cell carcinoma is more amenable to endoscopic treatment than is early gastric cancer of poorly differentiated tubular adenocarcinoma in select tumor conditions. Surg Endosc 2011;25:3087—93. [27] Hiraki M, Kitajima Y, Koga Yet al. Aberrant gene methylation is a biomarker for the detection of cancer cells in peritoneal wash samples from advanced gastric cancer patients. Ann Surg Oncol 2011;18:3013—9. [28] Jetten AM. Retinoid-related orphan receptors (RORs): critical roles in development, immunity, circadian rhythm, and cellular metabolism. Nucl Recept Signal 2009;7:e003. [29] Lee JM, Kim IS, Kim H, Lee JS, Kim K, Yim HY, et al. RORalpha attenuates Wnt/beta-catenin signaling by PKCalpha-dependent phosphorylation in colon cancer. Mol Cell 2010;37:183—95. [30] Airaksinen MS, Saarma M. The GDNF family: signalling, biological functions and therapeutic value. Nat Rev Neurosci 2002;3:383—94. [31] Liu H, Ma Q, Li J. High glucose promotes cell proliferation and enhances GDNF and RET expression in pancreatic cancer cells. Mol Cell Biochem 2011;347:95—101. [32] Moore DL, Apara A, Goldberg JL. Kruppel-like transcription factors in the nervous system: Novel players in neurite outgrowth and axon regeneration. Mol Cell Neurosci 2011;47:233—43. [33] Humar B, Guilford P. Hereditary diffuse gastric cancer: a manifestation of lost cell polarity. Cancer Sci 2009;100:1151—7. [34] Odawara H, Iwasaki T, Horiguchi J, Rokutanda N, Hirooka K, Miyazaki W, et al. Activation of aromatase expression by retinoic acid receptor-related orphan receptor (ROR) alpha in breast cancer cells: identification of a novel ROR response element. J Biol Chem 2009;284:17711—9. [35] Watanabe T, Miura T, Degawa Y, Fujita Y, Inoue M, Kawaguchi M, et al. Comparison of lung cancer cell lines representing four histopathological subtypes with gene expression profiling using quantitative real-time PCR. Cancer Cell Int 2010;10:2. [36] Baylin SB. DNA methylation and gene silencing in cancer. Nat Clin Pract Oncol 2005;2(Suppl. 1):S4—11.
Please cite this article in press as: Yang M, et al. Different methylation profiles between intestinal and diffuse sporadic gastric carcinogenesis. Clin Res Hepatol Gastroenterol (2014), http://dx.doi.org/10.1016/j.clinre.2014.03.017