Though the incidence of gastric cancer (GC) is decreasing, GC still remains ranked the fifth for incidence of malignancy and the fourth leading cause of cancer death globally.¹ East Asia has the highest incidence rate of GC.¹ As GC has been detected early by nationwide cancer screening program especially in Korea² and Japan,³ endoscopic resection (ER) of indicated early GC (EGC) has set an alternative method to surgery for preserving stomach function. However, ER of EGC has a relative high risk of occurrence of metachronous GC (MGC) in remnant stomach.⁴

Helicobacter pylori is known to have a crucial role in progression of the Correa sequence.⁵ H. pylori-infected gastric mucosa can change from chronic gastritis to atrophic gastritis, intestinal metaplasia, dysplasia, and invasive carcinoma. As H. pylori eradication may reduce the incidence rate of EGC as well as MGC,⁶ there have been many studies about not only genetic but also epigenetic alteration on the mechanism of H. pylori-induced GC in the past decades.^7-9 In consequence, H. pylori infection has been thought to form aberrant DNA methylation by direct virulence factors and indirect inflammatory response including neutrophil, monocyte and cytokines. Accumulation of abnormal methylation contributes to an “epigenetic-field-defect” for cancerization which has an increased risk of developing GC.^9-11

Aberrant DNA methylation in cancer is divided into two categories: global DNA hypomethylation and regional hypermethylation.¹² Global hypomethylation occurs at CpG repetitive sequences throughout genome in normal tissue which induces carcinogenesis by relating to genomic instability and abnormal chromosomal structures. Regional hypermethylation is found in promotor CpG islands in H. pylori-induced nonneoplastic gastric mucosa that is susceptible to GC.

Cancer cells develop during losing intercellular contact and progressing to invasion. In this regard, Rho GTPases regulate the microtubule cytoskeleton, leading to crucial roles in cell cycle, polarity, migration to affect initiation and progression of cancer.^13-15 Most of the information offered about the function of Rho proteins and their regulators comes from the studies on the best-characterized members CDC42, Rac1, RhoA, ELMO1, and DOCK180.^14,16 Especially, disturbance of Rho GTPases signaling in cancer was caused by alteration of their regulators.¹³ A recent study presented seven novel epigenetic markers including ELMO1 as one of the regulator. Its methylation level was associated with GC risk.^12,17 A study about glioma reported that co-overexpression of ELMO1 and DOCK180 was related to Rac1 promoting cell migration and invasion.¹⁸

Meanwhile, H. pylori infection is known to be related with Rho GTPases via direct virulence factors such as vacuolating cytotoxin A and cytotoxin-associated gene A in cell-cell adhesion.¹⁹ These direct virulence factors influence oxidative stress with indirect inflammatory response such as tumor necrosis factor-α, interleukin-1β, interleukin-10, monocyte and neutrophil. The result of oxidative stress causes the change of genetic susceptibility, which influences DNA hypermethylation as a carcinogenic intracellular pathway.^20,21 It could reflect the transcriptional modulation called silencing. In chronic infection of H. pylori, several soluble factors interact aberrantly between H. pylori and microenvironment which causes bacterial persistent survival. This cascade leads to precancerous lesions and epithelial-mesenchymal transition.²² The epithelial-mesenchymal transition could be formed by loss of cell polarity like cell-cell adhesion. The regulatory complex including Rac1 and CDC42 influenced the disruption of tight junctions resulting in loss of cell polarity.¹⁶ H. pylori activated Rac1 and DOCK180 to interfere with focal adhesions of cells after increasing host-cell motility and elongation.^15,19 Besides, H. pylori-infected cells experienced the stimulation of cell motility in the pathway represented by RhoA and fibronectin.¹⁵

In this study, we investigated whether epigenetic fields related to Rho GTPases (CDC42, Rac1, and RhoA) and its regulators (ELMO1 and DOCK180) altered throughout H. pylori-related gastric carcinogenesis, which recovered after H. pylori eradication.

1. Gastric tissue samples

This study included 63 patients with H. pylori-positive GC and 15 H. pylori-negative controls. The H. pylori-positive GC group consisted of patients who underwent endoscopic submucosal dissection (ESD) for GC, and the control group included in subjects diagnosed as normal or gastritis without H. pylori infection by endoscopy at the outpatient clinic. All subjects in this study were older than 18 years and younger than 80 years. Exclusion criteria were as follows: history of other malignancy or gastrectomy; history of taking antibiotics or proton pump inhibitors within a month; previous history of H. pylori eradication. During endoscopy, we took two gastric biopsy samples from antrum and corpus of noncancerous tissues and stained them with hematoxylin-eosin and modified Giemsa stains for histological evaluation assessed by the updated Sydney System: neutrophils and mononuclear cells infiltration, atrophy, intestinal metaplasia, and H. pylori status.²³ H. pylori infection was defined as positive if the histologic stain or rapid urease test (CLOtest; Delta West Ltd., Bentley, Australia) was positive. Other two biopsy samples for promoter methylation and messenger RNA (mRNA) expression analyses were taken from antral and corpus mucosa and stored at –80^℃. Patients with H. pylori-positive GC were recommended eradication therapy after ESD. Eradication regimen was the standard triple therapy including 20 mg omeprazole, 1 g amoxicillin, and 500 mg clarithromycin twice daily for 1 week. We also obtained gastric mucosal tissues again after 1 year from index ESD in the same manner to evaluate the effect of H. pylori eradication on the epigenetic regulation and mRNA expression. This study was approved by the Institutional Review Board of Seoul National University Hospital (IRB number: H-1507-112-690). All participants were provided written informed consent before participation.

2. mRNA extraction and quantitative real-time reverse transcription-polymerase chain reaction

mRNAs were derived from gastric tissues using homogenizer and TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Complementary DNA was synthesized from reverse transcription of 2 μg total RNAs and 1 μL of resulting complementary DNA was amplified in 20 μL 2x AMPIGENE qPCR Green Mix Hi-ROX (Enzo Life Sciences, Farmingdale, NY, USA) using StepOnePlus^TM Real-Time PCR system (Thermo Fisher Scientific, Waltham, MA, USA). The human β-actin gene was used as an endogenous control and the relative expression levels of mRNAs were calculated by the comparative 2^-ΔΔCt method as described previously.^24-26 All samples were tested in duplicate.

3. DNA extraction, bisulfite modification, and methylation analysis

DNA was taken from gastric mucosa using the QIAamp DNA Mini Kit (Dawin Biotech, Seoul, Korea) by the manufacturer’s instructions. Unmethylated cytosine of 200 ng of extracted DNA switched to uracil as performing bisulfite modification using the EZ DNA Methylation Kit (Zymo Research, Orange, CA, USA).

The methylation status of the bisulfite-modified DNA was quantified by real-time polymerase chain reaction-based MethyLight assay as previously described.^27,28 MethyLight includes methylation-specific primers combined with methylation-specific fluorescent probes followed by highly sensitive detection and quantification of methylation.²⁸ Pairs of primers and probes (Table 1) were designed by using the software, Beacon Designer (PREMIER Biosoft International, Palo Alto, CA, USA). The level of DNA methylation was assessed by a percentage of methylated reference (PMR), which is calculated as PMR=100×[(methylated reaction/Alu)_sample/(methylated reaction/Alu)_{CpG methyltransferase}]. Each MethyLight for all samples was conducted in triplicate.

Table 1 . The Individual Primers/Probes for the MethyLight Assay and Corresponding Primers for RT-PCR Assay.

Gene	Primer/probe	Sequence (5’→23’)	Product length, bp
Individual primers/probes for MethyLight assay
CDC42	Forward primer	TTTCGAATTTACGAGTTGCGTCG	92
	Reverse primer	AAAACTAAAACGAATCAAAACGATACCT
	Probe	6FAM-TACCCTAACTCCGCCCGCCTATCGC-BHQ1
DOCK180	Forward primer	TACGGCGTGAGCGTTAGTTTC	106
	Reverse primer	TCCCCTTCCTCTATCCCAATACG
	Probe	6FAM-CGCTCCGCACCCGAACACCCC-BHQ1
ELMO1	Forward primer	CGACGACGTAAATAACTCTACCTCT	110
	Reverse primer	TAGTCGGCGGTGTAGTAGTCG
	Probe	6FAM-ATACCCATCCTCGCCCGCTCCCG-BHQ1
Rac1	Forward primer	GGATGATAGACGTGAGTTATTGCG	134
	Reverse primer	AAACGATAATTTCAATAAACCCAAATAACG
	Probe	6FAM-ACTCCAACCTAAAATCAAAACGAAACCCCG-BHQ1
RhoA	Forward primer	TAGTAGTAGTAGTTTGAAGAGGGTTACG	140
	Reverse primer	AACCAATCATAACGAAATCCTAATCTAAAC
	Probe	6FAM-AACTCTACCGACCACGATACTAAACGCGC-BHQ1
Corresponding primers for quantitative RT-PCR assay
CDC42	Forward primer	GATGGTGCTGTTGGTAAA	195
	Reverse primer	TAACTCAGCGGTCGTAAT
DOCK180	Forward primer	TCAGACCATGAAGTTCAACC	223
	Reverse primer	ATTCACCCAATTAAAACCAG
ELMO1	Forward primer	CCGGATTGTGCTTGAGAACA	121
	Reverse primer	CTCACTAGGCAACTCGCCCA
Rac1	Forward primer	TTGCCAAAATACCTTCTGAACT	155
	Reverse primer	TGCTTTACGCATCTGAGAACTA
RhoA	Forward primer	CATCCGGAAGAAACTGGT	168
	Reverse primer	TCCCACAAAGCCAACTC
Actin	Forward primer	TTCGAGCAAGAGATGGCCAC	203
	Reverse primer	CGGATGTCCACGTCACACTT

RT-PCR, real-time reverse transcription-polymerase chain reaction; bp, base pair..

The annealing melting temperature of each reaction was 59℃..

4. Statistical analysis

The t-test was used for the overall comparison of continuous variables of the three groups (H. pylori-negative controls, H. pylori-eradicated GC, and H. pylori-persistent GC) for normally distributed data. Categorical variables were analyzed by the chi-square or Fisher exact tests. Generalized estimating equation was applied for pairwise group comparisons between ESD and 1-year follow-up for categorical variables. The Kruskal-Wallis test was used for methylation and Kruskal-Wallis test and Jonckheere-Terpstra test was performed for mRNA expression to examine differences between GCs and controls in the level at index ESD. We used the ranked analysis of covariance model to analyze differences in the level of methylation and mRNA expression after correcting baseline imbalances. Linear mixed model was used for comparison between H. pylori-eradicated and H. pylori-persistent groups over 1 year. We used the Mann-Whitney U test to analyze in subgroup analyses between “monocyte-sustained” and “reduced” groups, “absent to mild” and “moderate to severe” atrophic gastritis groups, or “absent to mild” and “moderate to severe” intestinal metaplasia groups. p<0.05 were considered significant. Statistical analyses were performed using SPSS, version 19.0 (IBM Corp., Armonk, NY, USA).

1. Baseline clinicopathological characteristics of subjects

The patients with H. pylori-positive GC were significantly older (median, 60.5 years vs 53.6 years, p=0.006) and predominantly male sex (79.4% vs 40.0%, p=0.002) than H. pylori-negative controls. In histologic evaluation, the degree of neutrophil, monocyte, mucosal atrophy, and intestinal metaplasia was more severe in H. pylori-positive GCs than H. pylori-negative controls (all p<0.001, OR=206.5 in monocyte) (Table 2).

Table 2 . Baseline Clinicopathological Characteristics of Subjects.

Characteristic	Hp-positive GCs (n=63)	Hp-negative controls (n=15)	p-value†	Hp-persistent GCs (n=34)	Hp-eradicated GCs (n=29)	p-value†
Age, yr	60.5 (54.5–67.5)	53.6 (45.0–61.0)	0.006	60.5 (54.5–66.5)	60.6 (54.0–70.0)	>0.05
Male sex	50 (79.4)	6 (40.0)	0.002	24 (70.6)	26 (89.7)	0.062
Mucosal atrophy*			0.003			>0.05
Absent to mild	29 (46.0)	14 (93.3)		17 (50.0)	12 (41.4)
Moderate to severe	23 (36.5)	0		11 (32.4)	12 (41.4)
Intestinal metaplasia			<0.001			>0.05
Absent to mild	23 (36.5)	15 (100)		10 (29.4)	13 (44.8)
Moderate to severe	40 (63.5)	0		24 (70.6)	16 (55.2)
Neutrophil			<0.001			>0.05
Absent to mild	7 (11.1)	15 (100)		5 (14.7)	2 (6.9)
Moderate to severe	56 (88.9)	0		29 (85.3)	27 (93.1)
Monocyte			<0.001			>0.05
Absent to mild	4 (6.3)	14 (93.3)		3 (8.8)	1 (3.4)
Moderate to severe	59 (93.7)	1 (6.7)		31 (91.2)	28 (96.6)

Data are presented as median (interquartile range) or number (%)..

Hp, Helicobacter pylori; GC, gastric cancer..

*Exception where pathologic evaluation is inapplicable or medical record is absent; ^†t-test for continuous variable, Pearson chi-square test or Fisher exact test for the categorical variables..

2. Effect of H. pylori eradication on the environment of gastric mucosa in H. pylori-positive GCs

In 63 patients with H. pylori-positive GC, 29 patients were successfully eradicated (H. pylori-eradicated GCs) and 34 were remained as H. pylori-persistent GCs by non- or failed eradication. There were no significant differences in baseline clinicopathological characteristics between the two groups (Table 2). When compared at 1 year after eradication, the degree of mucosal atrophy or intestinal metaplasia between both GC groups were not statistically different over time (Table 3). Meanwhile, the degree of neutrophil (p<0.001) and monocyte (p<0.001) infiltration were significantly improved in the H. pylori-eradicated GCs over time.

Table 3 . Clinicopathological Difference Over Time between Hp-Eradicated and -Persistent Groups.

Variable	Baseline			1-yr follow-up		p-value†
	Hp-eradicated GCs (n=29)	Hp-persistent GCs (n=34)		Hp-eradicated GCs (n=29)	Hp-persistent GCs (n=34)
Age, yr	60.6 (54.0–70.0)	60.5 (54.5–66.5)
Sex
Male	26 (89.7)	24 (70.6)
Female	3 (10.3)	10 (29.4)
Mucosal atrophy*						0.546
Absent to mild	12 (41.4)	17 (50.0)		12 (41.4)	14 (41.2)
Moderate to severe	12 (41.4)	11 (32.4)		9 (31.0)	13 (38.2)
Intestinal metaplasia*						0.082
Absent to mild	13 (44.8)	10 (29.4)		12 (41.4)	9 (26.5)
Moderate to severe	16 (55.2)	24 (70.6)		16 (55.2)	25 (73.5)
Neutrophil*						<0.001†
Absent to mild	2 (6.9)	5 (14.7)		28 (96.6)	5 (14.7)
Moderate to severe	27 (93.1)	29 (85.3)		0	29 (85.3)
Monocyte*						<0.001†
Absent to mild	1 (3.4)	3 (8.8)		19 (65.5)	1 (2.9)
Moderate to severe	28 (96.6)	31 (91.2)		9 (31.0)	33 (97.1)

Data are presented as median (interquartile range) or number (%)..

Hp, Helicobacter pylori; GC, gastric cancer..

*Exception where pathologic evaluation is inapplicable or medical record is absent; ^†Generalized estimating equation..

3. Methylation and expression level of three Rho GTPases and two regulators in gastric mucosa according to H. pylori and disease status

We measured the levels of methylation in Rho GTPases and its regulators in normal or surrounding gastric tissue at index ESD. CDC42 as one of Rho proteins had no DNA methylation. Each DNA methylation level of two Rho proteins (RhoA and Rac1) in H. pylori-positive GCs was lower than the level of H. pylori-negative controls (RhoA mean, 38.6 vs 53.1, p=0.046; Rac1 mean, 58.3 vs 104.2, p<0.005) (Fig. 1A and B). DNA methylation levels of two regulators (DOCK180 and ELMO1) in both GCs were higher than those of H. pylori-negative controls (DOCK180 mean, 27.7 vs 2.1, p<0.005; ELMO1 mean, 34.4 vs 0.01, p<0.005) (Fig. 1C and D).

Figure 1. The levels of promoter DNA methylation and corresponding mRNA expression in the study groups. (A, B) In RhoA and Rac1, the methylation levels of the Hp-negative controls were higher than in Hp-positive GCs. Their expression levels in Hp-negative controls were higher than in Hp-positive GCs. (C, D) In DOCK180 and ELMO1, the methylation levels of in Hp-negative controls were lower than Hp-positive GCs. Their expression levels of Hp-negative controls were lower than in Hp-positive GCs. (E) In CDC42, the expression levels of Hp-negative controls were lower than in Hp-positive GCs.
PMR, percentage of methylated reference; Hp, Helicobacter pylori; mRNA, messenger RNA; GC, gastric cancer. *The asterisks mean the outlier that is the data were over 1.5 interquartile range above 75th percentile. The number of specimen calculating the level of DNA methylation was 29, 34, and 14 cases in Hp-eradicated group, Hp-persistent group and Hp-negative control. The number of specimen calculating the level of mRNA expression was 29, 33 and 15 cases in Hp-eradicated group, Hp-persistent group and Hp-negative control; ^†p<0.05 using Kruskal-Wallis test for PMR and Kruskal-Wallis test and Jonckheere-Terpstra test for mRNA relative expression. Data was partly supported by D&P Biotechnology in Kyungpook National University Chilgok Hospital (Daegu, Republic of Korea).

Similar to their methylation pattern, quantitative real-time reverse transcription-polymerase chain reaction of RhoA and Rac1 revealed that the expression level of H. pylori-positive GCs was lower than that of H. pylori-negative controls (RhoA mean, 0.66 vs 1.00, p<0.005; Rac1 mean, 0.99 vs 1.00, p=0.05) (Fig. 1A and B). Conversely, the expression level of GCs in DOCK180, ELMO1, and CDC42 was higher than H. pylori-negative controls (DOCK180 mean, 1.27 vs 1.00; ELMO1 mean, 4.55 vs 1.00; CDC42 mean, 1.78 vs 1.00, all p<0.005) (Fig. 1C-E).

Ranked analysis of covariance model was used to examine the differences in the level of methylation and mRNA expression after correcting baseline imbalances of sex and age. It showed that the differences of methylation and expression levels between both GCs and H. pylori-negative controls kept statistical significance except the levels of methylation in Rho A and mRNA expression in Rac1 (Table 4).

Table 4 . Differences in the DNA Methylation and Expression Levels of mRNAs between Groups after Adjustment of Covariates.

Groups after adjustment of covariates	DNA methylation				mRNA expression
	df	F	p-value		df	F	p-value
RhoA
Age	1	0.013	0.908		1	0.049	0.826
Sex	1	1.850	0.178		1	1.726	0.193
Group	1	1.360	0.247		1	3,423.5	<0.001*
Rac1
Age	1	0.033	0.856		1	0.007	0.932
Sex	1	0.781	0.380		1	3.732	0.057
Group	1	8.050	0.006*		1	0.355	0.553
DOCK180
Age	1	0.015	0.902		1	0.692	0.408
Sex	1	0.232	0.631		1	1.026	0.314
Group	1	4.345	0.041*		1	11.082	0.001*
ELMO1
Age	1	1.852	0.178		1	0.014	0.907
Sex	1	1.521	0.221		1	0.804	0.373
Group	1	4.857	0.031*		1	57.282	<0.001*
CDC42
Age					1	0.030	0.863
Sex					1	0.852	0.359
Group					1	117.151	<0.001*

mRNA, messenger RNA; df, degree of freedom..

*p<0.05 using analysis of covariance..

4. Effect of H. pylori eradication on methylation and expression of three Rho GTPases and two regulators

DNA methylation and expression levels in noncancerous gastric mucosa of H. pylori-positive GCs were compared between after 1 year and index ESD using linear mixed model. In RhoA and Rac1 as Rho protein, promoter methylation level of noncancerous mucosa at index ESD had no difference between H. pylori-eradicated and persistent GCs (p>0.05) (Fig. 1A and B). There was also no difference in the degree of change over time (p>0.05) (Fig. 2A and B). Meanwhile, the expression level in H. pylori-eradicated GCs of RhoA and Rac1 at index ESD was higher than that of H. pylori-persistent GCs (p<0.005) (Fig. 1A and B). Although the similar pattern in expression level of H. pylori-eradicated and H. pylori-persistent groups, the expression level in H. pylori-eradicated GCs of these two Rho proteins after 1 year kept higher than that of H. pylori-persistent GCs (p<0.005, p=0.02, respectively) (Fig. 2A and B).

Figure 2. Change in promoter DNA methylation and corresponding mRNA expression according to the Hp eradication status. (A, B) The methylation of RhoA and Rac1 had no difference between Hp-eradicated and Hp-persistent GCs over 1 year, and their expression in Hp-eradicated GCs was slightly higher than it was in Hp-persistent GCs over time. The number of specimens for methylation was 34 Hp-persistent GCs and 29 Hp-eradicated GCs for RhoA and Rac1. The number of specimens for expression was 34 Hp-persistent GCs and 29 Hp-eradicated GCs for RhoA and 34 Hp-persistent GCs and 28 Hp-eradicated GCs for Rac1. (C, D) Although the methylation of DOCK180 and ELMO1 had no difference between Hp-eradicated and Hp-persistent GCs at the index endoscopic resection, the gap of methylation between Hp-persistent GCs and Hp-eradicated GCs increased over time. Their expression of Hp-eradicated GCs was much lower than that of Hp-persistent GCs over time. The number of specimens for methylation was 34 Hp-persistent GCs and 29 Hp-eradicated GCs for DOCK180 and 27 Hp-persistent GCs and 28 Hp-eradicated GCs for ELMO1. The number of specimens for expression was 34 Hp-persistent GCs and 28 Hp-eradicated GCs for DOCK180 and 34 Hp-persistent GCs and 29 Hp-eradicated GCs for ELMO1. (E) The expression in CDC42 of Hp-eradicated GCs was much lower than that of Hp-persistent GCs over time. The number of specimens for expression was 34 Hp-persistent GCs and 29 Hp-eradicated GCs for CDC42.
PMR, percentage of methylated reference; Hp, Helicobacter pylori; mRNA, messenger RNA; GC, gastric cancer. *p<0.05 using Linear mixed model. Data was partly supported by Seoul National University Cancer Research Institute (Seoul, Republic of Korea).

Conversely, DOCK180 and ELMO1 as regulators had other patterns with two Rho proteins. Though methylation levels at index ESD had no difference between H. pylori-eradicated and persistent GCs (p>0.05) (Fig. 1C and D), the level of H. pylori-eradicated GCs decreased and that of H. pylori-persistent GCs increased over time (p=0.039, p=0.031, respectively) (Fig. 2C and D). The methylation of DOCK180 and ELMO1 in H. pylori-persistent group were remarkably higher than H. pylori-eradicated group by crossing within 1 year.

Whether the degree of change over time by linear mixed model was statistically significant or not (p=0.034, p>0.05, respectively) (Fig. 2C and D), the level of expression in H. pylori-persistent GCs was higher than that in H. pylori-eradicated GCs throughout 1 year. Though the expression of RhoA, Rac1, and CDC42 had not much difference between H. pylori-persistent and -eradicated groups, we considered that more difference could be expected in a longer follow-up period as they were statistically significant in linear mixed model.

The subgroup analysis was performed to confirm whether the decrease in methylation level of two regulators after H. pylori eradication was due to reducing inflammation or not. We named the two groups monocyte-sustained group and monocyte-decreased group. Twenty-eight cases of H. pylori-eradicated GCs had moderate to severe monocytes at index ESD. Nineteen cases among them had reversed absent to mild monocytes, “monocyte-decreased group” and nine cases had kept the severity of monocytes after 1 year. The methylation levels of ELMO1 and DOCK 180 at index ESD and after 1 year had no difference between monocyte-sustained group and monocyte-decreased group in H. pylori-eradicated GCs (ELMO1, p=0.117; DOCK180, p=0.438, Mann-Whitney U test) (Fig. 3).

Figure 3. The difference in methylation levels of ELMO1 and DOCK180 between the monocyte-sustained group and the monocyte-decreased group. The methylation levels of ELMO1 and DOCK180 at index endoscopic submucosal dissection and after 1 year had no difference between these groups of H. pylori-eradicated gastric cancers (Mann-Whitney U test). (A, B) The methylation levels of ELMO1 and DOCK180 had no difference between the monocyte-sustained group and monocyte-decreased group.
PMR, percentage of methylated reference.

CDC42 had a pattern similar to that of two regulators in mRNA expression. The level of its expression in H. pylori-eradicated GCs was lower and much lower over time than H. pylori-persistent GCs (all p<0.005) (Figs 1E and 2E).

We performed the subgroup analysis for the difference of methylation level according to the degree of atrophic gastritis or intestinal metaplasia in H. pylori-positive GC group at 1 year after ESD (Table 2). The methylation levels in four target genes were not different between “absent to mild” (26/63) and “moderate to severe” atrophic gastritis group (22/63) (all of four genes p>0.1, Mann-Whitney test). The methylation level of ELMO1 unlike others was significantly different between “absent to mild” (21/63) and “moderate to severe” intestinal metaplasia groups (37/63) (ELMO1, p=0.003, Mann-Whitney test). While the degree of methylation in ELMO1 had the difference between both “absent to mild” and “moderate to severe” intestinal metaplasia groups in persistent GCs (p=0.002, Mann-Whitney test), that was not differ “absent to mild” from “moderate to severe” intestinal metaplasia groups in eradicated GCs (p=0.352, Mann-Whitney test).

H. pylori infection is considered to affect aberrant DNA methylation. Assembly of these methylation forms to epigenetic-field-defect which have an increased risk of susceptible GC.^10,11,29 This study showed that ELMO1 and DOCK180 as regulators of Rho GTPases had higher methylation levels in H. pylori-positive GCs than H. pylori-negative controls. Regardless of decrement in highly-methylated inflammatory cells, H. pylori eradication reduced methylations of two regulators and persistency of H. pylori infection increased the degree of methylation. Following the result of a recent study about ELMO1 as a novel methylation marker of GC,^12,17 present study suggested that two regulators could be the great markers of H. pylori-related methylation status in gastric carcinogenesis.

This study showed that the degree of inflammatory cells decreased significantly after 1 year of H. pylori eradication. H. pylori infection is related to not only secretion of pro-inflammatory cytokines but also activation of monocyte and neutrophil via chemotactic factors.²¹ This infection-associated inflammatory response is a critical role for H. pylori infection to induce aberrant DNA methylation in gastric carcinogenesis.^30,31 This study showed that eradication of H. pylori leaded to decrease inflammation in gastric mucosa and further reduced aberrant methylation in DOCK180 and ELMO1 as regulators of Rho GTPases.

There were few studies on the epigenetic field of noncancerous gastric mucosa of GC patients compared with normal gastric mucosa of controls. This study could suggest the risk of MGC occurrence after ER of index cancer and the meaning of gastric atrophy or intestinal metaplasia as a precancerous lesion. Predicting epigenetic-field-defect as a risk developing GC in residual stomach has an important meaning in that ER is one standard treatment for EGC. Although precancerous lesions like atrophic gastritis and intestinal metaplasia had no change in severity after H. pylori eradication in this study, it was noteworthy that gastric mucosa with moderate to severe intestinal metaplasia had higher methylation level of ELMO1 than absent to mild. A recent study also showed that the methylation level of intestinal metaplasia had a quantitative meaning as precancerous lesions.¹² H. pylori eradication removed the difference of methylation levels by severity in mucosal metaplasia in this study. We suggested the reason of these results why precancerous lesions have the time for reversal of inflammation at least 3 to 5 years after eradication,³² and the decrease of methylation in ELMO1 precede it.

In present study, DNA methylation level of two Rho proteins (RhoA and Rac1) in surrounding mucosa in H. pylori-positive GCs was lower and methylation level of two regulators (DOCK180 and ELMO1) was higher than H. pylori-negative controls. Corresponding mRNA expression of two Rho proteins had the same pattern as DNA methylation. As previously mentioned, aberrant DNA methylation in cancer has two manners called “global DNA hypomethylation” and “regional hypermethylation.”^10,12 While two Rho proteins with global DNA hypomethylation might be related to genomic instability and alteration of chromosomal structures, two regulators would play a key role by DNA hypermethylation. The present study showed that the hypermethylation in regulators of Rho GTPases occurring H. pylori-infected nonneoplastic gastric mucosa could be evaluated as a risk marker for GC.³³

When H. pylori-eradicated and persistent GCs were compared, the methylation level had no difference in both groups over time and H. pylori-eradicated GCs had slightly more expressed than H. pylori-persistent GCs in RhoA and Rac1. Meanwhile, the methylation of H. pylori-persistent GCs increased much more than H. pylori-eradicated GCs and leading mRNA expression level of H. pylori-persistent GCs was much higher than H. pylori-eradicated GCs in DOCK180 and ELMO1. CDC42 had a similar expression pattern as two regulators. Therefore, RhoA and Rac1 of Rho proteins might work in different ways with two regulators and CDC42 in H. pylori-related carcinogenesis. This phenomenon is correlated with a proceeding review related to cell migration and invasion. While DOCK180 and ELMO1 activated CDC42, they inactivated RhoA, and Rac1 was affected by other pathway.¹⁵

When Rho GTPases regulate several biological processes relevant to cancer including cell cycle control, epithelial cell polarity, cell migration, cell survival and angiogenesis, the activation state of Rho, Rac1, and CDC42 depends on balance with their regulators.³⁴ Besides, ELMO1 and DOCK180 have a synergistic role of carcinogenesis with interacting each other.¹⁸ For instance, ELMO1 interacts with DOCK180 by inhibiting ubiquitylation of DOCK180.³⁵ It could be noted that regulators played an important role in a diverse Rho GTPase-related mechanism.¹⁵ The present results had the important meaning in that the methylation and expression pattern of these regulators could be explained by the concept of epigenetic-field-defect. Epigenetic field cancerization of DOCK180 and ELMO1 could be used for methylation markers in risk stratification of GC.³⁶

The present study has several merits. First, this is the first study to evaluate DNA methylation and corresponding expression of Rho GTPases and its regulators in nonneoplastic gastric mucosa of H. pylori-positive GC patients. Even after the development of GC, this epigenetic change of regulators (guanine nucleotide exchange factors) could be recovered by H. pylori eradication. Second, we suggested that Rho GTPases and its regulators had two different mechanisms of global hypomethylation and regional hypermethylation. The regulators functioning the latter method could play the crucial role in both Rho GTPase-related and H. pylori-induced carcinogenesis. Third, this study has a meaning that methylation markers not as tumor suppressor genes but as oncogenes. Fourth, quantitative real-time reverse transcription-polymerase chain reaction and MethyLight analysis are sensitive and quantitative to analyze mRNA expression and promoter methylation. As relative degree of methylation was especially detected from small amounts of DNA, it could be useful as tumor marker of GC by using less invasive method such as blood or gastric juice than gastric tissue.¹⁰

This study has several limitations. First, the number of subjects included in the analysis was small. However, this number was enough to get statistical significance. Second, as we could not include GC tissue, we did not analyze the difference between neoplastic and nonneoplastic gastric tissue. Third, there was a lack of long-term follow-up data. It could be possible to analyze more detailed effects about atrophic gastritis and intestinal metaplasia and MGC occurrence in H. pylori-induced gastric carcinogenesis over long period.

In conclusion, we demonstrated that aberrant DNA methylation of DOCK180 and ELMO1 was associated with H. pylori infection and the formation of an epigenetic field for GCs. These epigenetic alterations could be recovered by H. pylori treatment.

This work was supported by grant number (03-2019-0060) from the SNUH Research Fund and a grant from the Liver Research Institute, Seoul National University College of Medicine.

Data was partly supported by D&P Biotechnology in Kyungpook National University Chilgok Hospital (Daegu, Republic of Korea) and Seoul National University Cancer Research Institute (Seoul, Republic of Korea).

No potential conflict of interest relevant to this article was reported.

Study concept and design: all authors. Data acquisition: J.L.K., E.N., S.G.K. Data analysis and interpretation: J.L.K. Drafting of the manuscript: J.L.K. Critical revision of the manuscript for important intellectual content: all authors. Statistical analysis: J.L.K. Obtained funding: S.G.K. Administrative, technical, or material support; study supervision: S.G.K. Approval of final manuscript: all authors.