Associations of Matrix Metalloproteinase-9 and Tissue Inhibitory Factor-1 Polymorphisms With Parkinson Disease in Taiwan

Abstract

Matrix metalloproteinases (MMPs) function in the degradation of extracellular matrix and are considered to play a role in the pathogenesis of neurodegenerative diseases including Parkinson disease (PD). MMPs activities are modulated by tissue inhibitors of metalloproteinases (TIMPs). This study examined whether the genetic polymorphisms of MMP-3, gelatinase (MMP-2 and MMP-9), TIMP-2, and TIMP-1 were associated with PD in Taiwan.

A total of 359 PD patients and 332 controls were enrolled. The candidate genetic variants included MMP-2 rs2285053 (−735 C > T), MMP-3 rs3025058 (−1171 5A > 6A), MMP-9 rs3918241 (−1831 T > A), rs17576 (G > A, R279Q), and rs3787268 (G > A, intron), TIMP-1 rs4898 (T > C, F124F), and TIMP-2 rs7503607 (−269 G > T). Associations were tested by logistic regression, adjusted with gender and age at onset.

Minor allele frequency of TIMP-1 rs4898 (36.0%) was significantly lower in the male PD patients than in the male controls (51.2%) (χ² test, P = 0.004). When adjusted with gender and age at onset, MMP-9 rs17576 AA genotype was associated with PD susceptibility in a recessive fashion (odds ratios [OR] = 2.28, 95% confidence intervals [95% CI] = 1.12–4.62, P = 0.02). In males, TIMP-1 rs4898 C allele was associated with a protective effect on PD (OR = 0.75, 95% CI = 0.60–0.94, P = 0.014). We did not find association between the examined genetic variants of MMP-2, MMP-3, and TIMP-2 and PD susceptibility.

This is the first study that demonstrated a protective effect of TIMP-1 rs4898 C allele on male PD and a modest association of MMP-9 rs17576 AA genotype with PD susceptibility in the Taiwan population. Further replication is needed for confirmation.

INTRODUCTION

Parkinson disease (PD) is the second most common neurodegenerative disorder. The pathological hallmarks of PD include progressive loss of nigro-striatal dopaminergic neurons and the presence of α-synuclein-containing Lewy bodies in the substantia nigra pars compacta (SNpc) and other sites of the brain.¹ The majority of PD cases are sporadic with only ∼10% identified as familial.² In contrast, mutations in different genes and environmental factors collectively account for most of the sporadic PD. There is ample evidence to suggest that it most likely results from an elaborate interplay of various factors: genetic predispositions, modifying effects by susceptible alleles, environmental exposures, gene–environment interactions, and their direct impact on the developing and aging brain.^1–3 Several pathways have been linked to PD pathogenesis including the presence of inflammation in the SNpc, oxidative stress, mitochondrial dysfunction, accumulation of atypical or misfolded protein, malfunction of ubiquitin-proteasome pathway, impairment of autophagolysosomes, and alterations of synaptic function and endosomal trafficking.^1,4–7

More recently, the role of inflammation in the pathogenesis of PD has gained rising attention.⁸ Pathology of substantia nigra of postmortem PD has shown CD8⁺ and CD4⁺ T-cell infiltration, accumulations of microglia cells and astrocytes, and alterations in glial cell morphology and function.⁹ Aggregated alpha-synuclein could activate microglia, which leads to disease progression in PD.¹⁰ Direct injection of α-synuclein into the substantia nigra resulted in the upregulation of mRNA expression of proinflammatory cytokines and microglial activation.¹¹ Microglia are the resident innate immune cells in the central nervous system and produce several factors (interleukins [ILs], tumor necrosis factor-alpha [TNF-α], nitric oxide [NO], prostaglandin E2 [PGE2], matrix metalloproteinases [MMPs], etc). Among these factors produced by activated microglia, MMPs are also proinflammatory factors that are toxic to neurons.¹² Accumulating evidence suggests that MMPs are involved in the neuropathological processes such as inflammation, blood–brain barrier (BBB) damage and neuronal cell death, which lead to central nervous system disorders such as PD.¹² Inducers of MMP expression and activity, such as cytokines, NO, and reactive oxygen species are implicated in the pathophysiology of PD. Tissue inhibitors of metalloproteases (TIMPs) have inhibitory actions against most MMPs with some predilections: TIMP-1 mainly inhibits MMP-9, whereas TIMP-2 inhibits MMP-2 and, paradoxically, contributes to activation of pro-MMP-2.

In the 4 main categories of MMP family, MMP-3 (one of the stromelysins) has been reported to influence pathogenesis of PD by generation of specific aggregation-enhancing α-synuclein fragments resulting from limited proteolysis.¹³ MMP-3 was induced and activated in dopaminergic cells upon stress conditions.¹⁴ In the postmortem brains of PD patients, α-synuclein and MMP3 were found to be co-localized in Lewy bodies.^14,15 MMP-3 contributes to the loss of dopaminergic neurons in a mouse model of PD with BBB damage and infiltration of peripheral immune cells.¹⁶ In addition, gelatinases (MMP-9 and MMP-2) have been shown to be related to PD. Reduced MMP-2 and increased TIMP-1 levels were shown in substantia nigra of postmortem brain of PD.¹⁷ Increased TIMP-1 levels in cerebrospinal fluid (CSF) of PD patients were also shown.¹⁸ Although these findings pointed towards a possible link between MMPs and TIMPs and pathogenesis of PD, the associations between genetic polymorphisms of the MMP/TIMP pathway and the susceptibility to PD are largely unknown in the Taiwan population. A prior report showed no association between MMP-3 − 1171 (5A/6A) and PD risks in Finns;¹⁹ however, the sample number was small. A significant association between the −1562 C > T polymorphism in the MMP-9 gene and risk of PD has been reported in Chinese.²⁰ The study herein is a case–control study that examines whether the single nucleotides polymorphisms (SNPs) of MMP-3, gelatinase (MMP-2 and MMP-9), and their counter-proteins (TIMP-2 and TIMP-1) are associated with PD susceptibility in the Taiwan population.

Methods

Subjects

A total of 359 PD patients and 332 controls were enrolled in this study. Patients (mean age at onset 61.0 ± 11.5 years, age at recruitment 68.4 ± 10.8 years, 49.6% women) diagnosed with PD were recruited from the neurology clinics of Chang-Gung Memorial Hospital (CGMH). All patients were diagnosed with probable sporadic PD according to the United Kingdom PD Society Brain Bank clinical diagnostic criteria.²¹ Controls were recruited from unrelated healthy adult volunteers (age at recruitment 67.4 ± 8.1 years, 50.3% women) matched for age, gender, and ethnicity. This study was approved by the Institutional Review Board of CGMH. All subjects gave informed consent for the study.

Polymorphism Selection and Detection

The cytogenetic location of MMP-2 is at 16q12. We selected a promoter variant rs2285053 (−735 C > T) which increased MMP-2 transcription.²² The cytogenetic location of MMP-3 is at 11q22. The common polymorphism rs3025058 (−1171 5A/6A) in the promoter of MMP-3 gene affected the MMP-3 gene expression, in which the 5A allele was associated with higher transcriptional activity compared to 6A allele.²³ The SNPs of MMP-9, located at 20q13, were selected based on international HapMap data on NCBI Build 36 assembly for Asian population and our prior experience,²⁴ including rs3918241 (−1831 T > A, promoter), rs17576 (G > A, missense variant R279Q, exon6), and rs3787268 (G > A, intron). For the TIMP-1 gene (cytogenetic location at Xp11.23), we selected rs4898 (T > C, synonymous, F124F, exon5) which is a strong tag SNP for Han Chinese. For the TIMP-2 gene (cytogenetic location at 17q25.3), we selected TIMP-2 rs7503607 (−269 G > T) which is 8 base pairs away from −261 G > A and fits the condition for TaqMan^® SNP Assays. Genomic DNA was extracted from peripheral blood mononuclear cells by standard protocols. Polymorphisms were genotyped by the TaqMan^® SNP Assays using the ABI Prism 7900HT Sequence Detection System (Table 1). All the SNPs are in Hardy–Weinberg equilibrium with significance level set at 0.05 to control SNP quality.

TABLE 1.

Primers for Genotyping of Single Nucleotides Polymorphisms (SNPs) Using the TaqMan SNP Assays

Statistical Analysis and Power Estimation

The differences in allele frequencies of SNPs between PD patients and controls were analyzed by the χ² test and Fisher exact test where appropriate. Multivariable logistic regression was used to analyze the phenotype-genotype associations of PD first under the additive genetic model and then the recessive model based on the distribution of genotypes. Covariables included age and gender.

Given the observed allele frequency in the present case–control study, at the 0.05 significance level, we had power >0.8 to identify an association of each genetic variant with PD susceptibility when the per-allele genetic effect was greater than an odds ratio of 1.4. Analyses were performed using SAS software version 9.1.3 (SAS Institute, Cary, NC).

RESULTS

The allele frequency distributions of the examined polymorphisms in PD patients and controls are displayed in Table 2. Minor allele frequency (MAF) of TIMP-1 rs4898 (36.0%) were significantly lower in the male PD patients, compared with that in the male controls (51.2%) (χ² test, P = 0.004). The frequency of minor allele in all of the other examined genetic variants was not different between PD patients and controls. The distributions of genotypes of the candidate variants and the associations of individual genetic variant with PD risk are displayed in Table 3. When adjusted with gender and age of onset, we found that MMP-9 rs17576 AA genotype was associated with PD susceptibility in a recessive fashion (odds ratios [OR] = 2.28, 95% confidence intervals [95% CI] = 1.12–4.62, P = 0.02). In men, TIMP-1 rs4898 C allele was associated with a protective effect on PD (OR = 0.75, 95% CI = 0.60 to 0.94, P = 0.014). The association between TIMP-1 rs4898 and PD was not observed in the women. We did not find association between the examined variants of MMP-2, MMP-3, and TIMP-2 and PD susceptibility. In addition, pairwise haplotype analysis of the 3 polymorphisms of MMP-9 showed no additional information regarding PD susceptibility.

TABLE 2.

Comparison of Minor Allelic Frequency Between Parkinson Disease (PD) Cases and Controls

TABLE 3.

The Distributions of Genotypes of the Candidate Variants and the Associations of Individual Genetic Variant With PD Risk

DISCUSSION

This is the first study showing associations of MMP-9 and TIMP-1 variants with PD susceptibility. For MMP-9, a prior study showed a significant association between −1562 C > T (rs3918242) polymorphism in the MMP-9 gene and PD risk in Chinese.²⁰ Although we did not find association between PD risk and rs3918241, the selected polymorphism at MMP-9 promoter, we discovered association between rs17576 AA genotype (missense variant R279Q, exon 6) and PD risk. For the TIMP1 gene, the tag SNP rs4898 is considered to represent a complete haplotype block covering the whole TIMP1 gene, based on international HapMap data on NCBI Build 36 assembly for Asian population.²⁴ We found that carriers of minor allele T of TIMP1 rs4898 might be protected from PD risk in the male group. For MMP-2 and TIMP-2, we did not find associations between the selected polymorphisms and PD. For MMP-3, this study supported the prior Finns research which demonstrated no association between MMP-3 and PD.¹⁹ Because the significance of the discovered associations is weak, further replication study with larger sample size is needed to confirm our findings.

Recently, increasing attention has been drawn to the role of inflammation in the pathogenesis of PD. Metalloproteinases are a large family of important proteases that include MMPs and proteins with a disintegrin and metalloproteinase domains (ADAMs). MMPs and ADAMs play an important role in neuroinflammmation. MMPs produced by activated microglia are proinflammatory factors and have been implicated to contribute to the pathogenesis of neurodegenerative diseases including PD.²⁴ Microglia stimulated by alpha-synuclein induced the expression of MMP-1, -3, -8, and -9 and inhibition of MMP-3, -8, or -9 suppresses the production of NO and other proinflammatory cytokines in primary microglia.²⁵ Overexpression of alpha-synuclein in rat primary astrocytes or glia increased MMP-9 activity.²⁶ 6-Hydroxydopamine (6-OHDA) and 1-methyl-4-phenylpyridinium (MPP⁺) increased MMP-9 gene expression by inducing nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) binding to the MMP-9 promoter in human neuroblastoma cells.²⁷ All of these suggest the expression of MMPs may contribute the pathogenesis of PD and expression of MMPs may influence the risk of PD.

The promoter regions of the inducible genes encoding MMPs generally contain binding sites for transcription factors such as activator proteins and NF-κB.²⁸ Activation of the NF-κB that controls target genes encoding proinflammatory cytokines, adhesion molecules, chemokines, and inducible enzymes, has been shown in PD brain.²⁹ Reduced MMP-2 and increased TIMP-1 have been shown in substantia nigra of postmortem brain from PD patients.¹⁷ Increased TIMP-1 levels in CSF of PD patients were also reported.¹⁸ Our study further provides evidence of link between MMPs/TIMPs and PD, although this study demonstrates only modest significant association of MMP-9 and TIMP-1 polymorphisms with PD susceptibility. This study identified rs17576 AA genotype of MMP-9 as a risk of PD susceptibility. Missense variation of rs17576 causes change in the catalytic domain and pexin-like domain of the MMP-9 enzyme, leading to polarity and functional change, which may thus contribute to enhanced inflammatory process.³⁰ There are limitations in our study. First, the protective genetic effect of rs4898 C allele of TIMP1 and the risk imposed by rs17576 AA genotype of MMP-9 on PD were not strong, and it is worth mentioning that the weak significance of the discovered associations might not survive during statistic correction of multiple testing. In addition, this study does not exclude association of PD with other SNPs within genes MMP2, MMP3, and TIMP2. Finally, the biological relevancies of the rs4898 C allele of TIMP1 and AA genotype of MMP-9 are not clear, which remains to be investigated. Further replicated studies with large sample size are needed to confirm our results before these candidate SNPs can be viewed as independent predictors of PD.

In summary, this is the first study that demonstrated modest association of MMP-9 rs17576 AA genotype with PD susceptibility and a protective effect of TIMP-1 rs4898 C allele on PD. The MMP/TIMP pathway posing a risk factor for PD may not be disease-specific, given that these polymorphisms could be risk factors for other medical conditions.^17,20,24,28 Further functional studies are needed to clarify the pathophysiology underlying the association between MMP-9 and TIMP-1 and PD risk and provide the additional functional information to support our findings.