An association study of SERPING1 gene and age-related macular degeneration in a Han Chinese population

Abstract

Purpose

Single nucleotide polymorphisms (SNPs) in the complement component 1 inhibitor (SERPING1) gene have been shown to be significantly associated with age-related macular degeneration (AMD) in Caucasian populations. A replication study of an association between these SNPs and AMD in a Chinese population is reported in this study.

Methods

Six SNPs, including rs2511990, rs1005510, rs11546660, rs2511989, rs2511988, and rs4926 in SERPING1 were genotyped in a Han Chinese subject group using the SNaPshot method of ABI. This subject group was composed of 194 patients with choroidal neovascularization (CNV or wet) AMD, 78 patients with soft drusen, and 285 matched controls. P values of the SNPs were calculated using an additive model. Haplotype frequencies between cases and controls were compared by χ2 analysis. The haplotype analysis was performed using Haploview 4.0.

Results

None of the six SNPs showed significant association with AMD. None of the major haplotypes were observed to be significantly associated with AMD or choroidal neovascularization AMD (CNV) after a stringent Bonferroni correction.

Conclusions

We demonstrate that SNPs in SERPING1 are not significantly associated with AMD in the mainland Han Chinese population.

Introduction

Age-related macular degeneration (AMD) is a leading cause of blindness in the elderly population, characterized as chronic and progressive degeneration of photoreceptors, the underlying retinal pigment epithelium (RPE), Bruch’s membrane, and possibly, the choriocapillaris in the macula [1,2]. AMD is divided clinically into dry and wet AMD. Patients with dry AMD present with cellular debris (drusen) in or under the retinal pigment epithelium (RPE), irregularities in the pigmentation of the RPE, or geographic atrophy (GA). Patients with exudative or wet AMD are characterized by serous detachment of the RPE or choroidal neovascularization (CNV), or both [1,2]. Advanced AMD, including geographic atrophy or exudative disease, can cause severe vision loss.

It is believed that AMD is a complex disorder caused by the interaction of multiple genetic and environmental risk factors [3–7]. Identification of AMD related genes has been tremendously successful. Complement pathway genes, including complement factor H (CFH) [2,8–13], C2/CFB [14–16], and C3 [15–17], have been confirmed by many replication studies. The LOC387715/HTRA1 gene has also been verified as a major AMD locus in different populations [18–24]. Recently, SNPs in the serpin peptidase inhibitor, clade G (C1 inhibitor) member 1 (SERPING1) gene showed highly significant genotypic association with age-related macular degeneration in two Caucasian populations [25]. Unfortunately, this finding could not be replicated by other studies [26–34].

To further analyze the association of SERPING1 and AMD, we investigated the association between SNPs in this gene and AMD in a mainland Han Chinese population.

Methods

Subjects

The Institutional Review Boards of the Sichuan Provincial People’s Hospital, Xinhua Hospital of Shanghai Jiao Tong University, and Zhongshan Ophthalmic Center, China approved this study. All subjects provided informed consent before participation in the study. AMD patients and normal age-matched controls, including individuals with a normal eye examination (individuals age 60 years or older with no drusen or RPE changes), were recruited in the ophthalmology clinic at Sichuan Provincial People’s Hospital, Xinhua Hospital of Shanghai Jiao Tong University, and Zhongshan Ophthalmic Center, China. All participants went through a standard examination protocol as in the previous description [19,24,27]. Grading was performed using a standard grid classification suggested by the International ARM Epidemiological Study Group for age-related maculopathy (ARM) and the age-related macular degeneration group [27]. All abnormalities in the macula were characterized according to the type, size and number of drusen, and hyperpigmentation or hypopigmentation, as well as AMD stages as defined by AREDS 1–5 stages. Patients with clinical features of AMD and CNV (CNV from other causes was excluded), with or without drusen, were diagnosed as wet AMD patients. Patients with only soft drusen were diagnosed as drusen (dry AMD) patients. In total, 194 wet AMD patients (Eight from Zhongshan Ophthalmic Center, 28 from Xinhua Hospital, 158 from of the Sichuan Provincial People’s Hospital), 78 soft drusen patients (all from of the Sichuan Provincial People’s Hospital), and 285 normal matched controls (Eight from Zhongshan Ophthalmic Center, 30 from Xinhua Hospital and 247 from of the Sichuan Provincial People’s Hospital) were recruited. In the normal matched controls, all individuals underwent an eye exam, no signs of early AMD, such as soft drusen or irregular pigmentations of the RPE in the macular area, were observed. Clinical information about the cases and controls is listed in Table 1.

Table 1. Characteristics of amd cases and controls matched for ages and ethnicity.

Subject	Total number	Male	Female	Average age
All AMD (CNV+drusen)	272	126	146	68.2±9.8
CNV	194	90	104	69.4±12.2
drusen	78	36	42	68.7±8.7
Controls	285	132	153	68.4±7.2

Selection of tag and functional SNPs

We used the data of the Han Chinese Beijing population in HapMap3 of the international HapMap project and previous studies to select tag SNPs or functional SNPs in SERPING1 [26,28] for this study. Six SNPs were selected for genotyping, including rs2511990 upstream of the transcription start site (−2877 bases), rs1005510 in intron 2, rs11546660 (A56V) in exon 3, rs2511989, rs2511988 in intron 6, and rs4926 (M480V) in exon 8.

Genotyping

Blood from each subject was drawn and collected in an EDTA-containing tube. Genomic DNA was extracted from the blood by a Gentra Puregene Blood DNA kit (Minneapolis, MN). SNP genotyping was performed by the dye terminator-based SNaPshot method (Applied Biosystems, Foster City, CA). SNP analysis was performed on the ABI 3130 genetic analyzer (Applied Biosystems). Genotypes of the SNPs were determined by Genemapper software (Applied Biosystems). All SNPs reported in this manuscript had a genotyping success rate >96 percent and accuracy as judged by random re-genotyping of 10 percent of the samples in the subject group. Six SNPs in SERPING1 were genotyped. The PCR and SNaPshot primers are listed in Table 2.

Table 2. Genotyping primers.

SNP	PCR Primer Forward	Snapshot Primer
rs2511989	F: TTCACAGCCTACCTTTCC	CCCTGGGTTTAATACAGGGGTTGTCAACTC
	R: CAGCCTCAATCATAATACCA
rs2511990	F: AAGCTGGAGCTGAAACTG	GTTCTCTTCCCACTGGGAGCAGGTCTAGGATTTCTC
	R: GGAAGAGGGATTCTGTGG
rs1005510	F: TTCTTACTACGAGGCACA	ATGTGGAAAATGTCCTGTACAAGAGAGTAATTTCTGACAGTGC
	R: TAAATCAAGGAGCACAAG
rs2244169	F: TGGCGTGAACCCTGGAGA	CAGCCAGAAAAGTTTTACAAAGCACGTATATGAC
	R: AGGTGGGAGGATTGCTTG
rs2511988	F: AGTGGGCTGGAACTTGGA	ATTGTGGGAGAGCTGCAGCTGCCCCACCTAGAAAATAAGAGATGCA
	R: GCATTGTGACAGAGGGTG

Haplotype analysis

Haplotype analysis was performed using Haploview 4.0. We performed the haplotype analysis following the instructions from the Broad Institute. If the genotype was not available, the genotype was set as 0.

Statistical analysis

Hardy–Weinberg equilibrium (HWE) for each SNP polymorphism was tested by the χ² test with df=1. P values of the SNPs were calculated using an additive model. Haplotype frequencies between cases and controls were compared by χ² analysis. The unadjusted odds ratios of the alleles and genotypes were estimated by the χ² test. All statistical analyses were performed using the software SPSS, (SPSS, Chicago, IL) version 10.0 [18–24].

Results

Single nucleotide polymorphism analysis

All six SNPs selected were successfully genotyped and all of these SNPs were within HWE in both case (p>0.001, Table 3) and control groups (p>0.05, Table 3). The SNP frequencies in this study were similar to those of Han Chinese Beijing (HCB) available in HapMap3 in the International HapMap Project. None of the six SNPs showed significant association with AMD or subphenotypes of AMD including wet AMD or soft drusen, which are landmarks of early AMD even before a stringent Bonferroni correction (p≥0.05, Table 3). SNP rs2511989 was reported to be the most significant association of SNP in the SERPING1 gene with AMD in previous studies [25]. Although rs2511989 showed high polymorphism, no association between this SNP and AMD was observed in the Chinese population (p=0.76 for all AMD; p=0.61 for CNV AMD; p=0.77 for soft drusen).

Table 3. Association between subphenotypes of AMD ANS SNPS in SERPING1 in the Han Chinese subject group.

SNP (risk allele)	Physical location (Chr.11)*	Phenotype	Genotype count			Allele frequency	HWE	Trend p-value
rs2511990 (T)	57119284	CVN+Drusen	CC:211	CT:50	TT:10	0.13	0.003	0.59
		CNV	CC:148	CT:35	TT:10	0.14	0.002	0.3
		Drusen	CC:63	CT:15	TT:0	0.1	0.643	0.47
		Control	CC:226	CT:51	TT:8	0.12	0.068
rs1005510 (G)	57123798	CVN+Drusen	AA:140	AG:117	GG:15	0.27	0.135	0.44
		CNV	AA:101	AG:81	GG:12	0.27	0.724	0.49
		Drusen	AA:39	AG:36	GG:3	0.27	0.312	0.58
		Control	AA:134	AG:131	GG:16	0.29	0.087
rs11546660 (C)	57124043	CVN+Drusen	TT:249	CT:21	CC:0	0.04	0.506	0.05
		CNV	TT:178	CT:16	CC:0	0.04	0.836	0.1
		Drusen	TT:71	CT:5	CC:0	0.03	0.957	0.13
		Control	TT:245	CT:35	CC:1	0.07	0.978
rs2511989 (A)	57134901	CVN+Drusen	GG:198	AG:57	AA:5	0.13	0.706	0.76
		CNV	GG:147	AG:42	AA:5	0.13	0.644	0.61
		Drusen	GG:51	AG:15	AA:0	0.08	0.298	0.77
		Control	GG:215	AG:63	AA:3	0.12	0.791
rs2511988 (C)	57135746	CVN+Drusen	TT:155	CT:103	CC:14	0.25	0.557	0.99
		CNV	TT:110	CT:73	CC:11	0.24	0.971	0.89
		Drusen	TT:45	CT:30	CC:3	0.23	0.763	0.77
		Control	TT:155	CT:118	CC:9	0.24	0.025
rs4926 (A)	57138565	CVN+Drusen	GG:201	AG:65	AA:6	0.14	0.784	0.43
		CNV	GG:147	AG:42	AA:5	0.13	0.644	0.7
		Drusen	GG:54	AG:23	AA:1	0.16	0.689	0.25
		Control	GG:209	AG:63	AA:3	0.13	0.766

Haplotype association analysis

We then performed haplotype analysis using Haploview 4.0, and 14, 15, and 14 haplotypes were observed in the AMD-control, wet AMD-control, and drusen-control groups, respectively. We found that haplotype TGTGCG and haplotype CGCGCG were shown to have a significant difference between both AMD-control (p=0.0064, p=0.006, respectively, Table 4) and wet AMD-control groups (p=0.0042, p=0.025, respectively, Table 4). The haplotype CGTGCG was shown to have a significant difference between both AMD-control (p=0.0102, Table 4) and drusen-control (p=0.032, Table 4). In addition, the haplotype CGTGTA was shown to have a significant difference between wet AMD and controls (p=0.026, Table 4). But none of the haplotypes were shown to have a significant difference between cases and controls (p>0.05, Table 4) after a stringent Bonferroni correction. On the other hand, the haplotype CGTGCA was shown to be significantly associated with soft drusen in our subject group (p=7.87x10⁻⁵, Table 4) with frequencies of 0.11 in cases and 0.03 in controls, even after a stringent Bonferroni correction (p=0.0011, Table 4). This haplotype conferred a 3.72-fold (95% CI: 1.83–7.54) increased likelihood of dry AMD (Table 4). Additionally, the haplotype CGTACA was also shown to have a significant difference between both all AMD-control and wet AMD-control groups (p<0.05, Table 4) after a stringent Bonferroni correction. However, the frequency of this haplotype was low and it was absent in the controls.

Table 4. SERPING1 haplotype association with AMD in the Han Chinese subject group.

Type of AMD	Haplotype	Frequency		Haplotype association (p value)	Bufferoni correction	Odds ratio (95% CI)
		Case	Control		(p value)
All AMD	H1:CATGTG	0.63	0.57	0.0609
	H2:CGTGTG	0.04	0.06	0.1091
	H3:TGTGCG	0.06	0.02	0.0064	0.0900
	H4:TGTACG	0.03	0.04	0.3301
	H5:CGTGCA	0.04	0.03	0.2641
	H6:CGTGCG	0.02	0.05	0.0102	0.1428
	H7:CATGTA	0.02	0.03	0.4105
	H8:CATATG	0.02	0.03	0.5689
	H9:TATGTG	0.02	0.01	0.3679
	H10:CGCGCA	0.02	0.01	0.5900
	H11:CGTACA	0.03	0.00	0.0008	0.0112
	H12:CGTACG	0.01	0.02	0.4070
	H13:CGTGTA	0.01	0.02	0.0628
	H14:CGCGCG	0.00	0.02	0.0060	0.0840
Wet AMD	H1:CATGTG	0.63	0.57	0.0717
	H2:CGTGTG	0.03	0.06	0.0946
	H3:TGTACG	0.04	0.04	0.7668
	H4:TGTGCG	0.06	0.02	0.0042	0.0630
	H5:CGTGCG	0.02	0.05	0.0620
	H6:CATGTA	0.03	0.03	0.6366
	H7:CGTGCA	0.02	0.03	0.5447
	H8:CATATG	0.02	0.03	0.2741
	H9:CGCGCA	0.02	0.01	0.5848
	H10:TATGTG	0.02	0.01	0.3616
	H11:CGTACG	0.01	0.02	0.6028
	H12:CGCGCG	0.00	0.02	0.0256	0.3840
	H13:CGTACA	0.03	0.00	0.0003	0.0045
	H14:CATGCG	0.01	0.01	0.3513
	H15:CGTGTA	0.00	0.02	0.0263	0.3950
Drusen AMD	H1:CATGTG	0.62	0.57	0.2702
	H2:CGTGTG	0.04	0.06	0.5894
	H3:CGTGCA	0.11	0.03	7.87E-05	1.10E-03	3.72 (1.83–7.54)
	H4:CGTGCG	0.01	0.05	0.0317	0.4438
	H5:TGTACG	0.01	0.04	0.0909
	H6:CATATG	0.04	0.03	0.6489
	H7:TGTGCG	0.05	0.02	0.1275
	H8:CATGTA	0.02	0.03	0.3929
	H9:CGTGTA	0.01	0.02	0.6998
	H10:CGCGCG	0.00	0.02	0.0894
	H11:TATGTG	0.02	0.01	0.3297
	H12:CGCGCA	0.01	0.01	0.8976
	H13:CGTACG	0.01	0.02	0.4630
	H14:CATGCG	0.00	0.01	0.2656

Discussion

Although genes in complement pathways, including CFH, C2/BF, and C3 [2,8–17] and chr.10q26 (LOC387715/HTRA1) [18–24], have been identified as related to AMD, these loci could not explain all genetic contributions to AMD, suggesting that additional genetic variants related to AMD have not yet been found. Based on the candidate gene approach, Ennis et al. [25] reported that SNPs in SERPING1 were significantly associated with AMD in two Caucasian populations. Additional evidence for SERPING1 involving AMD includes: 1) SERPING1 gene encoding C1INH plays an important role in complement pathways, which have been confirmed to participate in the pathogenesis of AMD; and 2) SERPING1 was expressed in both retinal and RPE-choroid layers in RT–PCR and immunofluorescence studies [25,29]. AMD affection status was correlated with increased abundance of choroidal C1INH [29]. Complement activation pathways include lectin, classical and alternative pathways. SERPING1 encodes C1INH, an inhibitor of the classical and lectin pathways of complement activation. The classical complement pathway is initiated by the C1 complex, which comprises a C1q hexamer complex with a zymogenic (C1r)₂-(C1s) ₂). SERPING1 irreversibly inhibits C1r and C1s, MASP-1 (mannan-binding lectin serine peptidase 1), and MASP-2 (mannan-binding lectin serine peptidase 2, the C1s ortholog in the lectin pathway), as well as modulating the complement activation through inhibition unrelated to proteases [30–33]. However, Park et al. [26] were unable to replicate the association between the genetic variation in SERPING1 and AMD in two large and well characterized Caucasian subject groups, and Allikmets et al. [34] were also unable to replicate the association between rs2511989 in SERPING1 and AMD. Additional replication studies, especially of a different ethnicity, are important to determine if SERPING1 is really associated with AMD. None of the six SNPs showed significant association with AMD and none of the major haplotypes were observed to be significantly associated with AMD or choroid neovascularization AMD (CNV) after a stringent Bonferroni correction in our study, suggesting that SERPING1 may not be related to AMD in the Han Chinese population. In the haplotype analysis, none of the SNPs tagged the significant haplotypes. Because half of samples’ genotype data for rs11546660 and rs4926 was not available in the HapMap3 for the Chinese, we cannot compare the haplotype frequencies to those in the HapMap. Although four haplotypes including TGTGCG, CGCGCG, CGTGCG, and CGTGTA were shown to have significant associations with different subphenotypes of AMD, anymore after a stringent Bonferroni correction, the significant associations no longer existed, suggesting that these haplotypes were not specifically associated with AMD. Since the haplotype CGTACA was rare in all AMD (3%) and wet groups (3%), and absent in the drusen group and controls, we think that the significant association between this haplotype and AMD is not reliable. The haplotype CGTGCA was shown to be significantly associated with soft drusen in the subject group even after a stringent Bonferroni correction (p=0.0011, Table 4). Further replication studies are needed to clarify the current situation because of the limited number of soft drusen samples in this study.