Variant of EOMES Associated with Increasing Risk in Chinese Patients with Relapsing-remitting Multiple Sclerosis

Abstract

Background:

Multiple sclerosis (MS) is a common central nervous system autoimmune disorder. Increasing number of genome-wide association study (GWAS) analyses hint that MS is strongly associated with genetics. Unfortunately, almost all the GWAS analyses were Caucasian population based. Numbers of risk loci might not be replicated in Chinese MS patients. Hence, we performed a MassArray Assay to genotype the previously reported variants located in the transcription regulation genes in order to elucidate their role in the Chinese MS patients.

Methods:

One hundred and forty-two relapsing-remitting MS (RRMS) patients and 301 healthy controls were consecutively collected from September 2, 2008, to June 7, 2013, as stage 1 subjects. Eight reported transcription regulation-related single-nucleotide polymorphisms (SNPs) were genotyped using the Sequenom MassArray system. In stage 2, another 44 RRMS patients and 200 healthy controls were consecutively collected and Sanger sequenced from April 7, 2015, to June 29, 2017, for the validation of positive results in stage 1. Differences in allele and genotype frequencies between patients and healthy controls, odds ratios, and 95% confidence intervals were calculated with the Chi-square test or Fisher's exact test. Hardy-Weinberg equilibrium was tested also using the Chi-square test.

Results:

In stage 1 analysis, we confirmed only one previously reported risk variant, rs11129295 in EOMES gene. We found that the frequency of T/T genotype was much higher in MS group (χ² = 10.251, P = 0.005) and the T allele of rs11129295 increased the risk of MS (χ² = 10.022, P = 0.002). In stage 2 and combined analyses, the T allele of rs11129295 still increased the risk of MS (χ² = 4.586, P = 0.030 and χ² = 16.378, P = 5.19 × 10⁻⁵, respectively).

Conclusions:

This study enhances the knowledge that the variant of EOMES is associated with increasing risk in Chinese RRMS patients and provides a potential therapeutic target in RRMS.

INTRODUCTION

Multiple sclerosis (MS; MIM 126200) is a common, partially heritable central nervous system (CNS) autoimmune disorder being considered as one of the most common causes of neurological disability in young adults.[1,2] Relapsing-remitting MS (RRMS) is the most common form of MS which has a biphasic disease course characterized by alternating episodes of acute neurological deficits, followed by a complete or partial recovery. Although the exact pathogenesis of MS is still not well classified, it is no doubt that genetic factors are primarily responsible for the substantially increased frequency of the disease, especially in the relatives of affected individuals.[3,4,5] For years, the major histocompatibility complex (MHC) was the only known MS susceptibility region.[6] However, hundreds of variants outside MHC region had now been identified by several genome-wide association studies (GWASs).[7,8]

The major pathogenic process of MS mainly starts with the activation of autoreactive lymphocytes and their migration across the blood-brain barrier. Transcription factors (TFs) are a group of key regulators involving in controlling lineage differentiation and cell survival. Certainly, TFs are indispensable for the T helper (Th) cell fate determination and cytokine production. A number of previous studies have demonstrated that the dysregulation of TFs is deeply involved in the pathogenesis of MS.[9,10,11] In addition, a series of GWAS analyses had discovered several risk loci within transcription regulation genes. In 2011, a Caucasian population-based multicenter GWAS analysis reported at least eight transcription regulator-related risk loci among 29 novel MS susceptibility loci.[7] Later on, another GWAS analysis replicated these results in a larger cohort.[8] In addition, it has been indicated that the mRNA expression of one TF, EOMES, was significantly and consistently lower in MS patients comparing to healthy controls (HCs).[12]

However, to date, few analyses were conducted to identify the variants located in transcription regulation genes in Asian MS populations. Here, we performed a MassArray Assay to genotype the previously reported variants located in the transcription regulation genes in order to elucidate their role in the Chinese MS patients.

METHODS

Ethical approval

The written informed consent was signed by each participant before inclusion in the study. This study was approved by the Ethics Committee of Huashan Hospital of Fudan University, the Second Affiliated Hospital of Zhejiang University School of Medicine, and the First Affiliated Hospital of Fujian Medical University.

Subjects

In the first stage, a total of 142 MS patients (59 males, 83 females; mean age 39.8 ± 12.3 years; range: 13.0–66.0 years) were consecutively collected from September 2, 2008, to June 7, 2013. All patients underwent detailed neurological examinations, laboratory tests, and magnetic resonance imaging scans for the brain and/or spinal cord. Patients were then diagnosed according to the revised McDonald criteria.[13,14] In addition, 301 healthy individuals (175 males, 126 females; mean age 37.0 ± 15.6 years; range 16.0–85.0 years) with no history of autoimmune diseases were recruited as HCs matched for case ethnicity and region. To validate the positive result from the first stage, another 44 RRMS patients and 200 HCs were consecutively collected from April 7, 2015, to June 29, 2017.

Genotyping

Genomic DNA was extracted from peripheral blood mononuclear cell using a QIAamp DNA Blood Minikit (QIAGEN, Hilden, Germany). Eight previously reported TF-related variants were genotyped using the Sequenom MassArray system. MassArray Assay Design 3.1 software (Sequenom, San Diego, USA) was used to design the polymerase chain reaction primers used in the genotyping [Supplementary Table 1]. Alleles were detected using a matrix-assisted laser desorption/ionization time of flight mass spectrometry platform (MassArray™, Sequenom Inc., San Diego, CA, USA) according to a previously described method.[15] In validation of the first stage, rs11129295 was further Sanger sequenced in subjects recruited for the second stage using a pair of in-house designed primer (Forward: 5'-TCTTGTTTTCTGGAGAGGAGC-3'; Reverse: 5'-ACCCACCTTCAGGAATTTCAAT-3').

Supplementary Table 1.

Primers for MassArray

SNP	Candidate gene	PCR primers	MassEXTEND primers
rs228614	NFKB1/MANBA	Forward: ACGTTGGATGTGCTTTTACTGTGTTCCTTC	GTCCCATTCAGTGCTTTC
		Reverse: ACGTTGGATGAGTCAGGCTTAAGCAACCAC
rs744166	STAT3	Forward: ACGTTGGATGACATTGAGAGGGCAATTGGG	gggcCTTGAGGGAATCGAGC
		Reverse: ACGTTGGATGTGGCTGTAATGTCTTGAGGG
rs2300603	BATF	Forward: ACGTTGGATGACATAGACTGATGCCGAGAG	cctctTCAGTATGAGGCTTTCATTC
		Reverse: ACGTTGGATGTTCTCTCTAAGCAGCCATCC
rs4410871	MYC	Forward: ACGTTGGATGTCTGCCGTGAATGAGAAACC	CCTCCCACACTGGAA
		Reverse: ACGTTGGATGGCAGTTACATCTGCAGTGTG
rs4902647	ZFP36L1	Forward: ACGTTGGATGTAAGCCTATAGCTCCCTTCC	caCCCGTCCCCTCTAAG
		Reverse: ACGTTGGATGGCTCCTTTGCAGAAAACCTC
rs9321619	OLIG3	Forward: ACGTTGGATGCATCTCTTGTAGTCTGGAGG	CAACTGGGCAGATGG
		Reverse: ACGTTGGATGGGGCAGGAAGAGCATTAAAG
rs11129295	EOMES	Forward: ACGTTGGATGGCTCATTTAATCTTCACAAC	cctcGGCCAGTTTTCTAACTTCT
		Reverse: ACGTTGGATGGTGACGTGGCCAGTTTTCTA
rs11154801	MYB/AHI1	Forward: ACGTTGGATGAGCTGTCATGTACCATGCAC	ccttaAGAAGGTCGAAACCTCAAGT
		Reverse: ACGTTGGATGCTCCTTCAGAAGGTCGAAAC

PCR: Polymerase chain reaction; SNP: Single-nucleotide polymorphisms.

Statistical analyses

Quantitative measures were summarized with descriptive statistics, such as mean ± standard deviation (SD) and 95% confidence interval (CI) of mean. Difference in ages was measured using the Student's t-test. Differences in genders, allele and genotype frequencies between MS and HC, odds ratios (OR s), and 95% CI s were calculated using the Chi-square test or Fisher's exact test. Hardy-Weinberg equilibrium (HWE) was tested using the Chi-squaretest. Statistical analyses were performed using SPSS 19.0 software (SPSS Inc., Chicago, IL, USA) and GraphPad Prism 6.02 (GraphPad Inc, San Diego, CA, USA). The criterion for a signifcant difference was a value of P < 0.05.

RESULTS

Demographic data

The demographic data are summarized in Table 1. In stage 1, the average ages of the MS patients and HCs were 39.8 ± 12.3 and 37.0 ± 15.6 years, respectively (t = 1.840, P = 0.070). In stage 2, the average ages of the MS patients and HCs were 39.0 ± 13.8 and 33.3 ± 11.3 years, respectively (t = 2.740, P = 0.007). In total, the average ages of the MS patients and HCs were 39.6 ± 12.5 and 35.5 ± 14.2 years, respectively (t = 3.310, P = 0.001).

Table 1.

Demographic and clinical characteristics of RRMS patients and healthy controls

Characteristics	MS	HC	Statistics	P
Stage 1, N	142	301
Male/female, n	59/83	175/126	10.656*	0.001
Age (year), mean ± SD	39.8 ± 12.3	37.0 ± 15.6	1.840†	0.070
Stage 2, N	44	200
Male/female, n	17/27	98/102	1.555*	0.210
Age (year), mean ± SD	39.0 ± 13.8	33.3 ± 11.3	2.740†	0.007
Stage 1 + Stage 2, N	186	501
Male/female, n	76/110	273/228	10.083*	0.001
Age (year), mean ± SD	39.6 ± 12.5	35.5 ± 14.2	3.310†	0.001

*χ²; ^†t. MS: Multiple sclerosis; RRMS: Relapsing-remitting MS; HC: Healthy controls; SD: Standard deviation.

EOMES variant rs11129295 associated with increasing risk in Chinese multiple sclerosis patients

The average genotyping success rate across the single-nucleotide polymorphisms (SNPs) was 97.5% using the Sequenom MassArray system. As shown in Supplementary Table 2, most of the SNPs in each group were under the HWE except STAT3 variant rs744166 in MS patients (P = 0.020). Eventually, as summarized in Table 2, we found that the frequency of T/T genotype in EOMES variant rs11129295 was much higher in MS group with a significant difference (χ² = 10.251, P = 0.005). In addition, the T allele of rs11129295 increased the risk of MS (OR = 1.764, P = 0.002). Besides, when analyzed in a dominant model, the genotype T/T and C/T together can increase the risk of MS (OR = 4.076, P = 0.019). Hence, it was a recessive model (OR = 1.776, P = 0.007). Furthermore, we replicated these findings in another group of population using Sanger sequencing. In this stage, we still found that the frequency of T/T genotype in rs11129295 was much higher in MS group with a significant difference (P = 0.020). Besides, the T allele increased the risk of MS (OR = 1.830, P = 0.030). However, we did not find significance in recessive model in stage 2 [Table 2].

Supplementary Table 2.

Hardy-Weinberg equilibrium tests for all Chinese Han participants in this study

SNP	Candidate gene	MS		HC
		χ2	P	χ2	P
Stage 1
rs228614	NFKB1/MANBA	1.69	0.19	0.07	0.79
rs744166	STAT3	5.95	0.02	2.66	0.10
rs2300603	BATF	1.23	0.27	0.62	0.43
rs4410871	MYC	1.09	0.30	1.82	0.18
rs4902647	ZFP36L1	0.02	0.90	1.42	0.23
rs9321619	OLIG3	0.48	0.49	0.01	0.92
rs11129295	EOMES	0.77	0.38	0.13	0.72
rs11154801	MYB/AHI1	1.72	0.19	0.92	0.34
Stage 2
rs11129295	EOMES	2.91	0.09	2.16	0.14
Stage 1 + 2
rs11129295	EOMES	2.67	0.10	1.64	0.20

MS: Multiple sclerosis; HC: Healthy controls.

Table 2.

Allele and genotype distributions of rs11129295 between MS and healthy controls

SNP	Region	Candidate Gene	Genotype/Allele	MS, n (%)	HC, n (%)	χ2	OR (95% CI)	P
Stage 1
rs11129295	Intergenic	EOMES	TT	91 (65.9)	157 (52.2)	10.251		0.005
			CT	44 (31.9)	119 (39.5)
			CC	3 (2.2)	25 (8.3)
			T	226 (81.9)	433 (71.9)	10.022	1.764 (1.238–2.514)	0.002
			C*	50 (18.1)	169 (28.1)
			CT + TT†	135 (97.8)	276 (91.7)	5.958	4.076 (1.209–13.739)	0.019
			CC	3 (2.2)	25 (8.3)
			TT‡	91 (65.9)	157 (52.2)	7.313	1.776 (1.169–2.769)	0.007
			CT + CC	47 (34.1)	144 (47.8)
Stage 2
rs11129295	Intergenic	EOMES	TT	26 (59.1)	97 (48.5)	6.336		0.020
			CT	18 (40.9)	78 (39)
			CC	0 (0)	25 (12.5)
			T	70 (79.5)	272 (68.0)	4.586	1.830 (1.046–3.200)	0.030
			C*	18 (20.5)	128 (32.0)
			CT + TT†	44 (100)	175 (87.5)	6.128	1.143 (1.085–1.204)	0.011
			CC	0 (0)	25 (12.5)
			TT‡	26 (59.1)	97 (48.5)	1.618	1.534 (0.791–2.973)	0.203
			CT + CC	18 (40.9)	103 (51.5)
Stage 1 + 2
rs11129295	Intergenic	EOMES	TT	117 (64.3)	254 (50.7)	17.452		4.17×10−5
			CT	62 (34.1)	197 (39.3)
			CC	3 (1.6)	50 (10.0)
			T	296 (81.3)	705 (70.4)	16.378	1.834 (1.363–2.466)	5.19×10−5
			C*	68 (18.7)	297 (29.6)
			CT + TT†	179 (98.4)	451 (90.0)	12.947	6.615 (2.037–21.481)	8.12×10−5
			CC	3 (1.6)	50 (10.0)
			TT‡	117 (64.3)	254 (50.7)	9.932	1.750 (1.233–2.484)	0.002
			CT + CC	65 (35.7)	247 (49.3)

*The C allele is the ancestral allele according to dbSNP build 141; ^†Analysis in a dominant model; ^‡Analysis in a recessive model. SNP: Single-nucleotide polymorphisms; MS: Multiple sclerosis; HC: Healthy controls. CI: Confidence interval; OR: Odds ratio.

In combined analyses of the two stages, we found that the frequency of T/T genotype in rs11129295 was much higher in MS group with a significant difference (P = 4.17 × 10⁻⁵), as well as the T allele increased the risk of MS (OR = 1.834, P = 5.19 × 10⁻⁵) and the higher frequency of T/T plus C/T in dominant model (OR = 6.615, P = 8.12 × 10⁻⁵) and T/T genotype in recessive model (OR = 1.750, P = 0.002; Table 2). As shown in Table 3, we did not find any other significant differences between MS and HC groups among other previously reported variants within transcription-related genes.

Table 3.

Other variants involved in transcription regulation between MS and healthy controls

SNP		Location*	Region		Candidate gene
rs228614		Chr4: 102657480	Intronic		NFKB1/MANBA
rs11154801		Chr6: 135418217	Intronic		MYB/AHI1
rs9321619		Chr6: 137553271	Intergenic		OLIG3
rs4410871		Chr8: 127802783	Intronic		MYC
rs4902647		Chr14: 68787474	Downstream		ZFP36L1
rs2300603		Chr14: 75539214	Intronic		BATF
rs744166		Chr17: 42362183	Intronic		STAT3
SNP	MS MAF (allele)	HC MAF (allele)	χ2	P	OR (95% CI)
rs228614	0.48 (A)	0.48 (G)	1.005	0.361	0.864 (0.649–1.150)†
rs11154801	0.36 (A)	0.35 (A)	0.067	0.796	1.040 (0.771–1.404)
rs9321619	0.36 (A)	0.39 (A)	0.559	0.455	0.894 (0.665–1.200)
rs4410871	0.30 (T)	0.34 (T)	1.205	0.272	0.842 (0.620–1.144)
rs4902647	0.30 (C)	0.33 (C)	0.648	0.421	0.881 (0.648–1.199)
rs2300603	0.31 (C)	0.27 (C)	2.139	0.144	1.260 (0.924–1.719)
rs744166	–	–	–		–

*Position is based on GRCh38.p10 and dbSNP Build 141; ^†For Allele A. MS: Multiple sclerosis; HC: Healthy controls; SNP: Single-nucleotide polymorphisms; CI: Confidence interval; –: Not available; MAF: Minor allele frequency; OR: Odds ratio.

DISCUSSION

MS is an autoimmune demyelinating disease of CNS. Increasing number of evidences hint that MS is strongly associated with genetics. The MHC loci made the early success in demonstrating the important role of genetic factors. To date, 13 MHC loci had been identified.[6] Nevertheless, little progress was made in unraveling the non-MHC genes underlying susceptibility to MS until the advent of GWAS technology. Over the last decade, over 200 non-MHC genes had been identified using GWAS analyses in large cohorts.[7,8,16] Unfortunately, almost all the GWAS analyses were Caucasian population based. A number of risk loci might not be replicated in Chinese MS patients. In our previous studies, Liu et al.[17] found that the variants of interferon regulatory factor 5 were not associated with MS in the southeastern Han Chinese population. Moreover, Cai et al.[18] found the association between autophagy-related gene 5 (ATG5) and neuromyelitis optica, but failed in MS. Fortunately, Zhuang et al.[19,20] identified that the variants in interleukin 7 (IL7) and CYP27B1 were associated with MS. Similarly, in the present study, we confirmed only one previously reported risk variant, rs11129295 in EOMES gene. We found that the frequency of T/T genotype was much higher in the MS group and the T allele of rs11129295 increased the risk of MS.

TFs are playing critical roles in the differentiation of Th cell. As was demonstrated previously, T-bet is a major factor for Th1 cell differentiation and IFN-γ production.[21] Similarly, Foxp3 and RORγt are the master TFs of nTreg cell and Th17 cell, respectively.[9,22] EOMES, also termed as TBR2, encodes a TF which is crucial for embryonic development of the CNS in vertebrates.[23] Besides, multiple lines of evidences have demonstrated that EOMES deeply involves in defense against viral infections.[24,25] Its function in CD4⁺ Th cell differentiation was remarkably noted recently. A functional experiment revealed that EOMES expression directly suppresses the Rorc and IL-17a expressions through binding to the promoter regions of these genes, which results in suppression of Th17 cell differentiation.[26]

Meanwhile, EOMES expression itself is suppressed by transforming growth factor beta via a Smad-independent pathway in autoimmune disorders.[26] In addition, a Caucasian population-based mRNA sequencing study further indicated the negative role of EOMES dysregulation played in MS progression.[12] However, more recent functional studies were questioning the role of EOMES in some specific conditions. One study even revealed that the higher expression of EOMES in Th cell could result in the occurrence of secondary-progressive MS.[27] Another study conducted by Lupar et al.[28] revealed that the expression of EOMES limits the Foxp3 induction in a cell-intrinsic way. However, more recently, Zhang et al.[29] reported that EOMES promotes the development of type 1 regulatory T cells, a Foxp3-negative, IL-10-producing T cell subset, which has potent immunosuppressive functions in autoimmunity. Thus, the role of EOMES playing in the pathogenesis of MS seems rather complicated and might be divided into two phases, acute phase and chronic phase. The more specific mechanisms of EOMES in these two phases remain to be clarified further.

In summary, we have identified a transcription regulation-related variant in Chinese MS patients. This variant is associated with increasing risk. The findings in the current study, together with previous studies, enhanced the knowledge that EOMES low expression in the acute phase of RRMS could promote the disease progression. As well, it possibly hints that overexpression of EOMES in the acute phase of RRMS might be a potential therapeutic target in RRMS.

Financial support and sponsorship

This study was supported by the grants from the National Natural Science Foundation of China (No. 81371414 and No. 81125009).

Conflicts of interest

There are no conflicts of interest.