Role of matrix metalloproteinase-9 gene polymorphisms in glaucoma: A hospital-based study in Chinese patients.

BACKGROUND: Glaucoma is the irreversible vision loss and contributes second leading cause of blindness worldwide. Matrix metalloproteinase-9 (MMP-9) is involved with remodeling and destruction of extracellular matrix. Elevated MMP-9 levels and various functional variants of MMP-9 have been associated with glaucoma in different population. In the current investigation, we tested association of MMP-9 common variants with different clinical categories of glaucoma in Chinese population.
MATERIALS AND METHODS: We enrolled total of 396 glaucoma patients those reported to hospital comprising of 212 primary angle closure glaucoma (PACG) cases and 184 primary open-angle glaucoma POAG patients. In addition, 329 normal individuals from similar geographical areas were enrolled as healthy controls. Five common single nucleotide polymorphisms (rs3918242, rs3918254, rs2250889, rs3918249, and rs17576) were genotyped by PCR-RFLP. Plasma levels of MMP-9 were quantified by ELISA.
RESULTS: Heterozygotes (GC) and allele "G" for rs2250889 polymorphism were more frequent in PACG cases compared with healthy controls (GC: P < .0001, OR = 2.26; G: P < .0001, OR = 1.19). Similarly, heterozygous mutant and minor allele for rs3918242 polymorphism were more prevalent in POAG in comparison with healthy controls. Interestingly, distribution of rs17576 variant was statistically higher in both PACG and POAG cases than healthy controls. Furthermore, analysis of plasma MMP-9 with MMP-9 polymorphisms revealed significant association of rs2250889, rs3918242, and rs17576 with plasma levels of the protein.
CONCLUSIONS: MMP-9 mutants are associated with elevated plasma MMP-9 and predisposed to development of glaucoma.

INTRODUCTION

Glaucoma is second most reason of complete blindness worldwide, and it is estimated that about 79.6 million of individuals will be affected by 2020 including 11.2 million of bilateral vision loss.1 Based on different clinical and morphological presentations, glaucoma is broadly divided into primary angle closure glaucoma (PACG) and primary open‐angle glaucoma (POAG). The PACG is characterized by reduced anterior and corneal chamber, closure of angle between iris and trabecular meshwork. Major clinical phenotype of POAG includes compression and posterior displacement of lamina cribrosa and increased coupling of optic disc. A recent report in Chinese demonstrated higher prevalence of PACG (54.42%) than POAG (39.79%) in the year 2015, and it is estimated to increase up to 25.16 million of glaucoma cased by 2050 including 14.49 million of PACG and 9.59 million of POAG patients.2 Matrix metalloproteinase is a group of enzymes mostly responsible for degradation of extracellular matrix. The role of matrix metalloproteinase‐9 in glaucoma is properly understood. In glaucoma patients, degradation of extracellular matrix and overexpression of MMP‐9 have been reported and believed to play a major role in pathogenesis of the disease.3, 4, 5 Altered levels of MMP‐9 presumably lead to abnormal ECM degradation and trabecular meshwork dysfunction.5

Reports have shown that incidence rate of glaucoma is higher in siblings compared with unrelated controls, indicating an important role of host genetics in predisposition to glaucoma. Recent genome wide association studies reveled role of ATP‐binding cassette transporter A1, actin filament‐associated protein 1, GDP‐mannose 4,6 dehydratase, phosphomannomutase 2, transforming growth factor beta receptor III, fibronectin type III domain containing 3B, rho guanine nucleotide exchange factor 12, growth arrest‐specific protein 7, forkhead box C1, ataxin‐2, thioredoxin reductase 2 with POAG, and PACG showed to be linked to ependymin related 1, choline acetyltransferase, gli‐similar protein 3, fermitin family member 2, and dolichyl‐phosphate mannosyltransferase subunit 2.6, 7, 8, 9, 10 Furthermore, various candidate gene association studies presented association of single nucleotide polymorphisms with predisposition to development of POAG or PACG. Role of matrix metalloproteinase‐9 (MMP‐9) variants have been well investigated in different populations as MMP‐9 gene is responsible for remodeling of extracellular matrix. A recent study in North Indian patients showed 1.6 and 1.4 fold higher risk for development of PACG and POAG, respectively, in subject harbored with mutant for MMP‐9 (−1562 C > T) polymorphism.11 Meta‐analysis if earlier published reports revealed protective feature of rs17576 and rs3918249 variant against glaucoma development.12, 13 In Han Chinese population, variants of rs3918254 were associated with susceptibility to PACG.14 However, another independent study in Korean patients failed to demonstrate such association of MMP‐9 variants with susceptibility to POAG.15 Although various case‐control study has been reported worldwide, simultaneous investigation on POAG and PACG in a cohort is very limited. Various case‐control studies have focused on selective SNPs in smaller sample size. To have a definitive conclusion on role of MMP‐9 in glaucoma, study including both clinical phenotype of glaucoma in larger sample size is essential.

MMP‐9 gene is positioned on the longer arm (q11‐q13) of 20th chromosome. Several SNPs in MMP‐9 have been reported so far (https://www.ncbi.nlm.nih.gov/projects/SNP/snp_ref.cgi?geneId=4318), but functional variants those alter protein productions or affect transcription levels are of importance. Common polymorphisms both in promoter and coding region in MMP‐9 gene including rs3918242, rs3918254, rs2250889, rs3918249, and rs17576 have been widely investigated in different population. The SNP rs3918242 (−1562 C > T) is located at the promoter region and has been demonstrated to affect transcriptional rate of MMP‐9.16 Mutations in coding region believe to produce deform protein, hampering downstream signaling process, or degrade by ubiquitin mediated pathways. Three missense mutation in MMP‐9 gene viz. rs2250889 (Arg > Pro), rs3918252 (Asn > Lys), and rs17576 (Gln > Arg) are also presume to affect enzyme concentration and activities.

In the present hospital‐based case‐control study, we enrolled glaucoma patients and healthy controls from similar geographical area, genotyped common genetic polymorphisms MMP‐9 gene, and investigated possible association with development of glaucoma in Chinese population.

MATERIALS AND METHODS

Patients and controls

Glaucoma patients those reported to Department of Ophthalmology, Nanchong Central Hospital, Nanchong, Sichuan, China were recruited in the present study. Based on clinical representation, patients were grouped into PACG and POAG. Total of two hundred twelve PACG cases and 184 POAG patients were enrolled in the current investigation. A total of 329 healthy subjects from similar geographical areas were enrolled as healthy controls. About 5 mL of intravenous blood was collected from each participant with EDTA. Plasma was separated and stored at −80 degrees centigrade till further use. The study protocol was permitted by Institutional Ethical Committee of Nanchong Central Hospital, Nanchong and from all participant written informed permission was obtained.

DNA isolation and MMP‐9 typing

After plasma separation, 150 µL of residue was resuspended with 50 µL of sample dilution buffer and subjected to DNA isolation by SIGMA mini DNA isolation kit as per instruction from the manufacturer. MMP‐9 gene polymorphisms (rs3918242, rs3918254, rs2250889, rs3918249, and rs17576) were genotyped by polymerase chain reaction followed by restriction fragment length polymorphism (PCR‐RFLP) as described earlier.14 Details of primers, amplicon size, restriction enzyme used for genotyping, and differential banding pattern are shown in Table 1. In brief, primers for rs3918242 SNP produced an amplicon of 435 bp. The amplified products were digested with SphI restriction endonuclease and yielded various product sizes based on individuals genotype: CC: 435 bp, CT: 435 + 241+194 bp, and TT: 241 + 194 bp. Furthermore, for genotyping of rs3918252, region flanking the polymorphic site was amplified and digested with RsaI. As shown in Table 1, CC genotype remained uncut and yield 187 bp product, whereas subject with GG had 149 and 38 bp. Mutation introduces restriction site for certain endonuclease. Based on this, two common SNPs in MMP‐9 gene (rs2250889 and rs17576) were genotyped. As shown in Table 1, NlaIV and MspI were used for genotyping of rs2250889 and rs17576 polymorphism, respectively, and yield following fragments for respective genotypes (rs2250889: GG: 122 bp, GC: 122 + 63+59 bp, CC: 63 + 59 bp and rs17576: AA: 300 bp, AG: 300 + 170+130 bp, GG: 170 + 130 bp). BfaI enzyme was employed for genotyping of rs3918249 polymorphism. The PCR product of 153 bp was digested and yielded 96 + 57 bp (CC) or 153 + 96+57 bp (CT) or remained undigested of 153 bp (TT). About 20% of subjects were sequenced randomly and found concordant results with PCR‐RFLP typing data.

Table 1

Primers details and RFLP bands pattern for genotyping of MMP‐9 genotypes

SNP/ Chromosome position	Primer sequence	Amplicon size	Restriction enzyme	RFLP product and genotype	Reference
rs3918249 (C > T)/ Ch.20:46009497	F‐5′‐CCCTGGGTGGTCAGAAAG‐3′ R‐5′‐GGCTGGGCTCAAACTCCT‐3′	153 bp	BfaI	CC: 96 + 57 bp CT: 153 + 96 + 57 bp TT: 153	30
rs17576 (A > G)/Ch.20:46011586	F‐5′‐TCACCCTCCCGCACTCTGG‐3′ R‐5′‐CGGTCGTAGTTGGCGGCGGTGG‐3′	300 bp	MspI	AA: 300 bp AG: 300 + 170 + 130 bp GG: 170 + 130 bp	31
rs3918242 (C > T)/ Ch.20:46007337	F‐5′‐GCCTGGCACATAGTAGGCCC‐3′ R‐5′‐TTCCTAGCCAGCCGGCATC‐3′	435 bp	SphI	CC: 435 bp CT: 435 + 241 + 194 bp TT: 241 + 194 bp	32
rs3918252 (C > G)/ Ch.20:46010492	F‐5′‐CTCCTTCCTCTGGCTCTTACG‐3′ F‐5′‐CACGTTCTCACCCGCGACACC‐3′	187 bp	RsaI	CC: 187 bp CG: 187 + 149 + 38 bp GG: 149 + 38 bp	30
rs2250889 (G > C)/ Ch.20:46013767	F‐5′‐GCCCCTTCCTTATCGCCGACA‐3′ F‐5′‐CGGGGGAGGGAAAGTAGGTAA‐3′	122 bp	NlaIV	GG: 122 bp GC: 122 + 63 + 59 bp CC: 63 + 59 bp	30

Plasma MMP‐9 quantification

Based on availability of plasma samples (patients: 149 and controls: 78), MMP‐9 was measured for 227 subjects by enzyme linked immunosorbent assay (ELISA) (R&D system).

Statistical analysis

All analysis was conducted by Graphpad prism software (V.7). Genotype and allele frequencies were counted manually. Distribution of allele and genotype among different clinical categories were compared by Fisher exact test. P value, odds ratio (OR), and 95% confidence interval values were noted down. For genotypic and allelic comparison, a P value < .008 was considered as significant (Bonferroni correction 0.05/6 = 0.008). Mean plasma levels of MMP‐9 in healthy controls and different clinical condition of glaucoma were compared analysis of variance (ANOVA) and Tukey's post‐test. Association of MMP‐9 genotypes with plasma levels was analyzed ANOVA‐Tukey's post‐test. Distribution of MMP‐9 genotypes for Hardy‐Weinberg equilibrium (HWE) was investigated by formulated MS‐excel sheet. A P value < .05 was taken as statistically significant.

RESULTS

Baseline characteristics

In the present study, we enrolled 212 PACG cases and one hundred eighty‐four POAG Chinese patients. In addition, 329 healthy controls were enrolled hailing from similar geographical location. The mean age of PACG patients was 40.3 years, POAG was 42.5 years, and healthy controls were 41.3 years. Percentage of male included in the study was 72% in healthy controls, 75% in PACG and 80% in POAG cases.

MMP‐9 variants are associated with predisposition to glaucoma

MMP‐9 gene polymorphisms (rs3918242, rs3918254, rs2250889, rs3918249, and rs17576) were genotyped PCR‐RFLP, and distribution of genotypes was investigated for HWE. Distribution of genotypes for four SNPs was in accordance with HWE (rs3918242: χ 2 = 0.52, P = .47; rs2250889: χ 2 = 1.74, P = .18, and rs17576: χ 2 = 1.69, P = .19). In contrast, other two polymorphisms were deviated from HWE (rs3918252: χ 2 = 8.04, P = .004 and rs3918249: χ 2 = 4.93, P = .02).

To investigate possible association of common MMP‐9 variants with glaucoma, both genotype and allele frequencies were compared among case and controls (Table 2). Prevalence of heterozygotes (GC) and minor allele (C) of rs2250889 polymorphism was more frequent in PACG cases in comparison with healthy controls (GC: P < .001, OR = 2.26; C: P < .0001, OR = 1.9). In addition, variant and minor allele for rs3918242 polymorphism were more frequent in POAG patients than healthy controls [CT: P = .0002, OR = 2.08; T: P = .002, OR = 1.63]. Interestingly, healthy controls had lower frequency of rs17576 mutant (AG) when compared to glaucoma patients (PACG: P = .003, OR = 1.70; POAG: P = .0005, OR = 1.96), indicating significant role of MMP‐9 polymorphisms in predisposition to glaucoma.

Table 2

MMP‐9 polymorphisms distribution in PACG, POAG, and healthy controls

Genotype/Allele	HC (n = 329)	PACG (n = 212)	POAG (n = 184)	HC vs PACG P value, OR (95% CI)	HC v s POAG P value, OR (95% CI)
rs3918242 ‐1562 C > T
CC	223(68)	150 (71)	94 (51)	1, ref	1, ref
CT	98(30)	55 (26)	86 (47)	0.37, 0.83 (0.56 to 1.22)	0.0002, 2.08 (1.43‐3.02)
TT	8 (2)	7 (3)	4 (2)	0.60, 1.30 (0.49 to 3.67)	0.73, 1.35 (0.43‐4.36)
C	544 (83)	355 (84)	274 (74)	1, ref	1, ref
T	114 (17)	69 (16)	94 (26)	0.67, 0.92 (0.66 to 1.28)	0.002, 1.63 (0.60‐2.23)
rs3918254
CC	168 (51)	100 (47)	90 (49)	1, ref	1, ref
CT	148 (45)	101 (48)	86 (47)	0.47, 1.14 (0.80 to 1.64)	0.70, 1.08 (0.74‐1.57)
TT	13 (4)	11 (5)	8 (4)	0.51, 1.42 (0.64 to 3.30)	0.81, 1.14 (0.47‐2.83)
C	484 (74)	301 (71)	266 (72)	1, ref	1, ref
T	174 (26)	123 (29)	102 (28)	0.36, 1.13 (0.86 to 1.48)	0.66, 1.06 (0.80‐1.41)
rs2250889 544 R > P, G1722C
GG	6 (2)	8 (4)	6 (3)	0.07, 2.92 (0.96 to 8.68)	0.35, 1.96 (0.60‐6.36)
GC	99 (30)	102 (48)	64 (35)	<0.0001, 2.26 (1.58 to 3.25)	0.23, 1.27 (0.85‐1.86)
CC	224 (68)	102 (48)	114 (62)	1, ref	1, ref
G	111 (17)	118 (28)	76 (21)	<0.0001, 1.9 (1.41 to 2.55)	0.20, 1.23 (0.89‐1.70)
C	547 (83)	306 (72)	304 (79)	1, ref	1, ref
rs3918249
CC	148 (45)	91 (43)	83(45)	1, ref	1, ref
CT	158 (48)	108 (51)	83 (45)	0.58, 1.11 (0.77 to 1.58)	0.77, 0.93 (0.64‐1.36)
TT	23 (7)	13 (6)	18 (10)	0.85, 0.91 (0.46 to 1.89)	0.38, 1.39 (0.70‐2.65)
C	454 (69)	290 (68)	249 (68)	1, ref	1, ref
T	204 (31)	134 (32)	119 (32)	0.84, 1.02 (0.78 to 1.33)	0.67, 1.06 (0.80‐1.39)
rs17576 279Q > R, A836G
AA	10 (3)	4 (2)	5 (3)	0.77, 0.76 (0.26 to 2.33)	0.77, 1.2 (0.44‐3.32)
AG	115 (35)	102 (48)	94 (51)	0.003, 1.70 (1.19 to 2.41)	0.0005, 1.96 (1.34‐2.82)
GG	204 (62)	106 (50)	85 (46)	1, ref	1, ref
G	135 (21)	110 (26)	104 (28)	0.04, 1.35 (1.02 to 1.81)	0.005, 1.52 (1.13‐2.04)
A	523 (79)	314 (74)	264 (72)	1, ref	1, ref

Data are shown in number (%).

Abbreviations: HC, healthy control; PACG, primary angle closure glaucoma; POAG, primary open‐angle glaucoma.

Plasma levels of MMP‐9 in glaucoma cases and healthy subjects

Based on availability of plasma samples, levels of MMP‐9 were quantified by ELISA in (HC: n = 78; PACG: n = 71; and POAG: n = 78). As shown in Figure 1, both PACG (P < .0001) and POAG (P < .0001) patients displayed significantly higher plasma MMP‐9 compared with healthy controls.

Figure 1

Plasma level of MMP‐9 in healthy control and glaucoma patients. Plasma levels of MMP‐9 was quantified by ELISA in controls (n = 78), PACG (n = 71), and POAG (n = 78) patients. Mean plasma MMP‐9 levels in different clinical categories were compared by ANOVA followed by Tukey's post‐test. A P value < .05 was taken as significant

Association of MMP‐9 gene variants with plasma MMP‐9 levels

We observed significant association of four SNPs in MMP‐9 gene with development of glaucoma. However, the possible mechanism of how these SNPs are linked with glaucoma development is not known. We theorized that MMP‐9 variants would be associated with plasma levels of MMP‐9. To test this, we compared mean plasma MMP‐9 levels among different genotypes of MMP‐9 polymorphisms. As shown in Figure 2, mutants of rs3918242, rs2250889, and rs17576 polymorphisms had elevated plasma MMP‐9 compared with their respective wild type. No genotype phenotype linked was observed for rs3918254 and rs3918249 polymorphism. Interestingly, similar trend was observed throughout irrespective of their respective clinical categories (HC: Figure S1; PACG: Figure S2; and POAG: Figure S3).

Figure 2

Association of plasma MMP‐9 levels with MMP‐9 variants. Mean plasma levels of MMP‐9 was compared among different genotypes of (A) rs3918242 C > T (B) rs3918254 C > T (C) rs2250889 G > C (D) rs3918249 C > T and (E) rs17576 A > G. Mean plasma MMP‐9 levels in different genotypes were compared by ANOVA followed by Tukey's post‐test. A P value < .05 was taken as significant

DISCUSSION

In the present report, we studied role of common MMP‐9 gene variants with predisposition to development of PACG or POAG. Out of five SNPs studies in Chinese population, three (rs3918242, rs2250889, and rs17576) revealed significant association with susceptibility to PACG or POAG. In addition, glaucoma patients displayed higher plasma MMP‐9 and correlated with gene polymorphisms.

MMP‐9 is an important member of endopeptidases group and has been shown to affect remodeling and disintegration of extracellular matrix.17 As the glaucoma is characterized by shorted axial length and improper eye growth, differential MMP‐9 level is believed to be associated with clinical phenotype of glaucoma.18 We observed elevated plasma MMP‐9 levels in both PACG and POAG patients compared to healthy controls in line with earlier reports.19, 20 Collectively these observations suggest an important role of MMP‐9 in glaucoma.

Several case‐control studies on association of MMP‐9 polymorphisms and predisposition to glaucoma have been conducted worldwide and revealed possible link between glaucoma and MMP‐9. Variants of MMP‐9 such as rs391825414 and rs225088921 are associated with predisposition to development of PACG in Chinese cohort. However, other independent study including Chinese glaucoma patients failed to demonstrate such association.22 Furthermore, association of MMP‐9 mutants with susceptibility to PACG was deciphered in Australian,23 Taiwanese,24 and Singaporeans.25 Remarkably, a report including Pakistanis glaucoma patients showed distinct association of rs17576 with vulnerability to POAG and PACG, respectively.26 In addition, a recent meta‐analysis revealed role of rs17576 and rs3918249 against glaucoma development in Caucasian.12 In the present investigation, we observed strong association of rs2250889 with susceptibility to PACG and other variant (rs3918242) with predisposition to development of POAG. In addition, heterozygotes of rs17576 polymorphism were more frequent in glaucoma patients (both in PACG and in POAG) compared with healthy controls. These observations suggest a significant role of MMP‐9 variants with development of glaucoma in Chinese cohort.

The mechanism of how common variants of MMP‐9 gene are linked with glaucoma predisposition is not known. It is believed that MMP‐9 mutants possibly affect MMP‐9 levels in terms of protein or mRNA and ultimately alter optimum function of MMP‐9. To test functional relevance of MMP‐9 variants, we compared plasma levels of MMP‐9 in different genotype groups of MMP‐9 polymorphism. We observed elevated plasma MMP‐9 levels in mutants of rs3918242, rs2250889, and rs17576 polymorphisms compared with their respective wild type. Interestingly, similar association was noticed when patients and controls were investigated separately. Earlier studies on genotype‐phenotype association of MMP‐9 variants are very limited. TT genotype of rs3918242 polymorphism has been linked with elevated mRNA and MMP‐9 plasma levels when compared to CC and CT.27 However, no significant association of rs3918254 and rs3918249 polymorphisms with plasma levels of MMP‐9 was observed in the present study, in line with an earlier report suggesting these variants may not have functional role in determination of plasma levels of MMP‐9. Mechanism of how some MMP‐9 variants affect plasma levels is not well understood. Polymorphisms in promoter region possibly enhancing binding of various transcription factors or in coding region enhancing stability of MMP‐9 protein, in turns increasing plasma levels of MMP‐9 in mutant subjects.

Although in the present study, we enrolled larger number of samples and investigated possible association of MMP‐9 with predisposition of glaucoma, the report has several limitations. Firstly, various reports have demonstrated positive correlation between plasma levels of MMP‐9 and activity MMP‐9 studied by zymography, and in the present study, we have not quantified active MMP‐9 levels. Furthermore, the used kit (R&D systems) quantifies pro‐MMP‐9 (92 kDa) and active MMP‐9 (82 kDa) levels. Secondly, in the present study we quantified free MMP‐9 levels that is both pro‐MMP‐9 and active MMP‐9 levels. However, we have not measured MMP‐9/NGAL complex concentrations, which is believed to have an additive functional affect.28 Anticoagulants used for collection of blood samples were shown to altered activity of MMP‐9.29 In the current report, we have used EDTA as anticoagulant and samples collected in EDTA has been shown optimal activity of both pro and mature MMP‐9.29

In conclusion, glaucoma cases displayed higher plasma MMP‐9 in Chinese cohort. MMP‐9 gene variants are associated with higher plasma MMP‐9 and predisposed to POAG or PACG development. These observations indicate possible exploration of MMP‐9 inhibitors for treatment of glaucoma. However, further studies enrolling larger sample size and include different populations and possible clinical trials of MMP‐9 inhibitors are required to validate our observations.

Click here for additional data file.

Click here for additional data file.

Click here for additional data file.