Susceptibility of MMP3 gene polymorphism to coronary artery disease: A meta-analysis

Background Conclusions on susceptibility of MMP3-1612 5A/6A to morbid risk of coronary artery disease (CAD) are controversial. This meta-analysis aims to obtain the accurate relationship between them. Methods Relevant literatures on susceptibility of MMP3-1612 5A/6A to morbid risk of CAD published before July 2019 were searched in PubMed, Web of Science, Cochrane Library, CNKI, VIP and Wanfang. Data were extracted from eligible literatures and analyzed by RevMan5.3 and STATA12.0 for calculating OR and corresponding 95% CI. Study selection: A total of 18 literatures reporting MMP3-1612 5A/6A and CAD were enrolled. Data extraction was conducted by two researches independently. Any disagreement was solved by the third research.


Introduction
Coronary artery disease (CAD) is a heart disease caused by stenosis or obstruction of the coronary artery lumen due to coronary atherosclerosis, as well as functional changes that lead to myocardial ischemia, hypoxia, or necrosis.CAD is also referred to as ischemic heart disease, and it is the most common subtype of atherosclerosis-related lesions that significantly impact human health (1).CAD primarily affects individuals over 40 years of age, and it is more prevalent in men than in women.In recent years, the incidence of CAD has markedly increased in our country (1).
The matrix metalloproteinase (MMP) family consists of more than 20 secretases or ectocellular enzymes that degrade extracellular matrix proteins, coagulation factors, lipoproteins, latent growth factors, chemokines, and cell adhesion molecules (2,3).Dysregulated extracellular matrix (ECM) metabolism is of significance in vascular remodeling during the progression and complication period of atherosclerosis (4).MMP3 not only degrades ECM components, but also activates MMP1 and other family members (5).With the advanced progression achieved on DNA technologies, MMP3 gene polymorphisms in the promoter region have been identified.In particular, MMP3 5A/6A is closely related to multiple pathological states (6).Therefore, we speculated that MMP3 gene polymorphisms may be related to susceptibility of CAD.
Numerous studies have evaluated the involvement of MMP3 gene polymorphisms in CAD; however, the results have often been non-replicable (7)(8)(9).This may be attributed to racial-specific genetic characteristics, insufficient sample sizes, improper selection of patients and controls, and a lack of adjustments for confounding factors.In this study, we extracted data from eligible literature to assess the involvement of MMP3 -1612 5A/6A in the susceptibility of CAD.Additionally, we explored the heterogeneity among enrolled literature and the existence of publication bias.

Materials and Methods
The meta-analysis of observational research poses a particular challenge due to the inherent biases and differences in research design.Therefore, we performed this analysis in accordance to the guiding principles of the meta-analysis of observational studies in epidemiological statements (10).

Searching strategy
Relevant literatures on susceptibility of MMP3 -1612 5A/6A to morbid risk of CAD published before July 2019 were searched in PubMed, Web of Science, Cochrane Library, CNKI, VIP and Wanfang.Key words were searched as follows: MMP3, coronary heart disease/coronary syndrome/myocardial infarction/atherosclerosis and gene/allele/genotype/polymorphism/variation.

Study selection
Eligible full-text literatures were searched after reviewing titles and abstracts.Citations were manually searched.Data were extracted from recent or most complete publications.When an article contained more than one homogenous features (i.e. the endpoints of CAD and MI), each feature was separately analyzed.

Inclusion/exclusion criteria
Inclusion criteria included: 1) Independent casecontrol studies analyzing hospital-based or population-based subjects without any language limitations.2) Data were complete to obtain or calculate genotype frequencies.3) Literatures reporting susceptibility of MMP3 -1612 5A/6A to morbid risk of CAD. 4) Genotype distribution in the controls was in accordance to HWE (Hardy-Weinberg equilibrium).5) None of repeated published data.
Flow diagram of the publication selection was depicted in Figure 1.

Data extraction
Data were independently extracted and analyzed by two researchers.Any disagreement was solved by the third researcher.The following data were extracted from each literature: 1) Baseline characteristics of enrolled literature, including published magazine, first author, publish time, etc. 2) Baseline characteristics of subjects, including case number, country of publication, genotype number and distribution, HWE, etc.

Statistical analysis
Published data were synthesized when three or more studies reporting the same gene polymorphism.Once genotype data were provided, c 2 test was applied in the control group.
Heterogeneity test was conducted by calculating OR and the corresponding 95% CI with the I 2 test and the Q test.The pooled OR in studies lacking the heterogeneity was calculated by the fix-effects model; otherwise, a random-effects model was used.Publication bias was evaluated by depicting funnel plots and quantified by Egger's test.Data analyses were conducted using RevMan 5.3 and STATA12.0 (London, UK).

Discussion
MMPs are a family of Zn 2+ -dependent neutral proteases that degrade and reshape ECM, thus maintaining their homeostasis (27).MMP3 has a proteolytic activity on many ECM macromolecules (including different types of collagen) and it is able to promote the transformation of other MMPs.It is reported that MMP1 is upregulated in human atherosclerotic plaques, and ECM degradation activity is also enhanced, suggesting the involvement of MMP1 in CAD and MI (27,28).A common deletion/insertion polymorphism (rs3025058) exists in promoter region of human MMP3, which is characterized as 5 or 6 adenosines at 1612 bp upstream of transcription initiation site (29).Multiple studies have reported that MMP3 rs3025058 is closely linked to coronary atherosclerotic diseases (30)(31)(32).
Previous studies generally believed that MMP3 5A/6A is associated with AS, and the 5A allele is associated with plaque rupture.In this paper, CAD risk in subjects carrying MMP3 -1612 5A homozygous was 30% higher than those carrying MMP3 -1612 6A homozygous.In over-dominant genetic model, MMP3 -1612 5A/6A was a risk factor for CAD.Inconsistent with our findings, a relevant study reported that morbid risk and progression risk are higher in subjects carrying MMP3 -1612 6A homozygous than those carrying MMP3 -1612 5A homozygous (33).Xu et al. (34) suggested that MMP3 and MMP9 mutations could be utilized as hallmarks for predicting susceptibility to CAD.Furthermore, subgroup analyses uncovered that MMP3 -1612 6A/5A posed a huge impact on MI or East Asian population, which was consistent with the findings proposed by Want et al. (35) and Niu et al. (4).Moreover, taken into consideration of CAD as endpoint, MMP3 -1612 6A/5A was significantly related to CAD.Some limitations in this analysis should be highlighted.First of all, MMP3 gene polymorphism could be interacted with other known or unknown CADassociated risk factors.Secondly, we have ruled out many relevant literatures because genotype distribution was failed to be calculated using the inadequate data.Thirdly, sample size was relatively small.Fourthly, additional analyses on baseline characteristics of CAD patients were lacked.

Conclusions
Collectively, our findings demonstrated that MMP3 -1612 6A /5A was involved in CAD under recessive model and over-dominant model.In East Asian population (mainly Chinese population), MMP3 -1612 6A /5A was susceptibility to CAD in the four genetic models.Our findings required to be validated in a multi-center hospital with larger sample size.

Figure 1
Figure 1 Flow diagram of the publication selection process.

Table I
Baseline characteristics of enrolled literatures.

Table II
Subgroup analysis based on population.