MMP9 Rs3918242 Polymorphism Affects Tachycardia-Induced MMP9 Expression in Cultured Atrial-Derived Myocytes but Is Not a Risk Factor for Atrial Fibrillation among the Taiwanese

Matrix metalloproteinase (MMP) plays an important role in the pathogenesis of atrial fibrillation (AF). The MMP9 promoter has a functional polymorphism rs3918242 that can regulate the level of gene transcription. This study recruited 200 AF patients and 240 controls. The MMP9 rs3918242 was examined by polymerase chain reactions. HL-1 atrial myocytes were cultured and electrically stimulated. Right atrial appendages were obtained from six patients with AF and three controls with sinus rhythm undergoing open heart surgery. The MMP9 expression and activity were determined using immunohistochemical analysis and gelatin zymography, respectively. Rapid pacing induces MMP9 secretion from HL-1 myocytes in a time- and dose-dependent manner. The responsiveness of MMP9 transcriptional activity to tachypacing was significantly enhanced by rs3918242. The expression of MMP9 was increased in fibrillating atrial tissue than in sinus rhythm. However, the distribution of rs3918242 genotypes and allele frequencies did not significantly differ between the control and AF groups. HL-1 myocyte may secrete MMP9 in response to rapid pacing, and the secretion could be modulated by rs3918242. Although the MMP9 expression of human atrial myocyte is associated with AF, our study did not support the association of susceptibility to AF among Taiwanese subjects with the MMP9 rs3918242 polymorphism.


Introduction
Atrial fibrillation (AF) is the most prevalent arrhythmia in clinical practice, affecting about 46.1 million people worldwide [1], and it is associated with increased morbidity (especially stroke [2] and heart failure [3]) and mortality [4]. The mechanism of the pathogenesis of AF is complex, including electrical, contractile, and structural remodeling [5], which involves autonomic nervous system activation, genetic variants, renin-angiotensin-aldosterone (RAA) system activation, inflammation, and oxidative stress [6]. The longer the duration of AF, the more persistent it becomes as consequence Int. J. Mol. Sci. 2016, 17, 521 2 of 13 of atrial remodeling. Atrial fibrosis is the most remarkable feature of the structural remodeling, which plays a pivotal role in the development of a vulnerable substrate for AF. Atrial fibrosis was thought to result from disturbed extracellular matrix (ECM) metabolism with excessive fibrillar collagen deposition, generally in response to a cardiac insult [7,8].
Matrix metalloproteinases (MMPs) are a family of proteolytic enzymes, and they regulate the ECM metabolism and turnover in a harmony with its inhibitors, including tissue inhibitors of metalloproteinases (TIMPs) and reversion-inducing cysteine-rich protein with Kazal motifs (RECK) [9]. The ECM degradation by MMPs has been reported to be involved in the pathogenesis of cardiovascular diseases, including atherosclerosis, myocardial infarction, restenosis, and dilated cardiomyopathy [9,10]. The MMP-1, -2, -3, -9, -13, and -14 have been shown in the mammalian myocardium. Among the MMPs family, growing evidences have demonstrated that MMP9 play an important role in the pathogenesis of AF. Nakano et al. reported that the expression of MMP9 in the atrial tissue was increased in individuals with AF in comparison to individuals with sinus rhythm (SR) [11]. Higher circulating levels of MMP9 have been shown to correlate with elevated MMP9 activity [12]. A prospective case-cohort study demonstrated that, in individuals free of AF initially, elevated plasma levels of MMP9 were independently associated with increased risk of AF during an 11.8 years follow-up period [13]. The levels of MMPs have been found to differ according to whether the individual has paroxysmal or persistent AF, suggesting that the degree of atrial remodeling may be related to the burden or types of AF [14,15].
In a previous study, we have demonstrated that a single nucleotide polymorphism (SNP) rs3918242 in the promoter region of MMP9 gene is independently and significantly associated with plasma MMP9 level in a Taiwanese population [16]. SNP rs3918242 is a promoter polymorphism that is associated with the loss of binding of a nuclear repressor protein and elevated MMP9 expression [17]. Taken together, genetic polymorphisms that alter MMP9 production and thus affect atrial fibrosis/remodeling may modify the risk of AF. Furthermore, the major source of MMP9 in atria with AF remained unclear. In Nakano's study, double staining of the MMP9 and Mac-1 showed that the distribution of MMP9 did not always co-localize with that of Mac-1, suggesting that MMP9 in atria with AF may originate from other cells in addition to macrophage. The mechanism underlying the increasing expression of MMP9 in the fibrillating atria also remained unclear. Since "AF begets AF", whether electrical remodeling of atrial myocyte will increase its MMP9 expression and then propagate atrial interstitium ECM degradation with structural remodeling demands to be determined. The present study aims to test if rapid pacing would induce MMP9 secretion by the atrial myocyte, and to elucidate the association between the MMP9 promoter gene polymorphism, the MMP9 expression in atrial tissue, and the risk of AF.

Rapid Pacing Induces MMP9 Expression and Secretion of HL-1 Atrial Myocytes in a Time-and Dose-Dependent Manner
Previous studies have established that rapid stimulation of cultured HL-1 myocytes simulates the phenotype feature of tachycardia-induced atrial remodeling in vivo [18][19][20][21]. Accordingly, we applied this atrial-derived system to analyze the effect of tachypacing in vitro to study whether or not atrial myocytes express MMP9 and whether the tachypacing further affects its expression. As shown in Figure 1A, a dose-dependent manner between the MMP9 production of HL-1, quantified by Western blot analysis, and the frequency of the atrial pacing was observed. There was also a continuous increase in pacing-induced MMP9 production in parallel with the duration of pacing ( Figure 1B). Furthermore, the secreted MMP9 in the conditioned medium was quantified by gelatin zymography. As demonstrated in Figure 2, the MMP9 secretion was detected in the conditioned medium before pacing. The activity increased time-dependently in parallel with the duration of tachypacing. This effect is quite close to significance (test for trend, p = 0.052). These results suggest that atrial myocytes could express and secrete MMP9, and production could be augmented by rapid pacing. that atrial myocytes could express and secrete MMP9, and production could be augmented by rapid pacing.  Tachypacing of HL-1 induces MMP9 secretion. HL-1 cells were subjected to 4 Hz pacing for the indicated times, the secretion of MMP9 protein in collected medium was evaluated by gelatin zymography as described in Methods. (Bottom) The relative expression levels of MMP9 were quantified by densitometry and normalized to the control level, which was set at 1.0. Each value represents the mean ± SE of four independent experiments. p value: Dose-response relationships analyzed using one-way ANOVA with linear trend test. that atrial myocytes could express and secrete MMP9, and production could be augmented by rapid pacing.  Tachypacing of HL-1 induces MMP9 secretion. HL-1 cells were subjected to 4 Hz pacing for the indicated times, the secretion of MMP9 protein in collected medium was evaluated by gelatin zymography as described in Methods. (Bottom) The relative expression levels of MMP9 were quantified by densitometry and normalized to the control level, which was set at 1.0. Each value represents the mean ± SE of four independent experiments. p value: Dose-response relationships analyzed using one-way ANOVA with linear trend test. Tachypacing of HL-1 induces MMP9 secretion. HL-1 cells were subjected to 4 Hz pacing for the indicated times, the secretion of MMP9 protein in collected medium was evaluated by gelatin zymography as described in Methods. (Bottom) The relative expression levels of MMP9 were quantified by densitometry and normalized to the control level, which was set at 1.0. Each value represents the mean˘SE of four independent experiments. p value: Dose-response relationships analyzed using one-way ANOVA with linear trend test.

Effect of the Rs3918242 on the MMP9 Transcriptional Activity in Rapidly Stimulated HL-1 Myocytes
To investigate whether the SNP rs3918242 modulates the transcriptional regulation of the MMP9 promoter, human MMP9 promoter constructs containing the SNP rs3918242 C/T were created and transfected into HL-1 myocytes. As shown in Figure 3A, the relative luciferase activity of the constructs containing rs3918242 T was significantly higher than that of the wild-type (C) constructs (p < 0.05). The increasing promoter activity as a result from the rs3918242 C to T was further enhanced in HL-1 myocytes with rapid pacing ( Figure 3B). These results suggest that the SNP rs3918242 may enhance the transcriptional activity of MMP9 gene and its responsiveness to rapid pacing.

Effect of the Rs3918242 on the MMP9 Transcriptional Activity in Rapidly Stimulated HL-1 Myocytes
To investigate whether the SNP rs3918242 modulates the transcriptional regulation of the MMP9 promoter, human MMP9 promoter constructs containing the SNP rs3918242 C/T were created and transfected into HL-1 myocytes. As shown in Figure 3A, the relative luciferase activity of the constructs containing rs3918242 T was significantly higher than that of the wild-type (C) constructs (p < 0.05). The increasing promoter activity as a result from the rs3918242 C to T was further enhanced in HL-1 myocytes with rapid pacing ( Figure 3B). These results suggest that the SNP rs3918242 may enhance the transcriptional activity of MMP9 gene and its responsiveness to rapid pacing. Figure 3. Effect of the rs3918242 on the MMP9 transcriptional activity in HL-1 cells with and without rapid pacing. (A) HL-1 cells were transfected with plasmids containing rs3918242 C or T allele in the MMP9 promoter for 6 h. The luciferase activity was assayed as described in Methods; (B) HL-1 cells were transfected with plasmids containing rs3918242 C or T allele in the MMP9 promoter for 24 h and subsequently received tachypacing (4 Hz) for 6 h. Each value (mean ± SE, n = 4) is expressed as a fold change of luciferase activity relative to the control condition. * represents p < 0.05. Table 1 shows a summary of the clinical and demographic characteristics of the study population. Of the 440 subjects, 200 had AF and 240 were controls. No statistical differences were found between the two groups regarding age, gender, BMI, hypertension, diabetes mellitus, smoking, hypercholesterolemia, and history of coronary artery disease (CAD). The proportion of subjects using angiotensin receptor blockers, β-blockers, calcium antagonists, diuretics, statins, digoxin, aspirin, and oral anticoagulants, were significantly higher in patients with AF than in control subjects. The genotype and allele frequencies of the MMP9 SNP rs3918242 in the control and AF groups are shown in Table 2. No significant deviation from the Hardy-Weinberg equilibrium was observed for the MMP9 SNP rs3918242 in either the cases or the controls. Both the distribution of rs3918242 genotypes and allele frequencies were not significantly different between the control and AF groups (p = 0.190, and 0.424, respectively; Table 2). After adjustment for age and gender, the presence of the rs3918242 T allele was still not associated with the risk of AF using an additive inheritance model (odds ratio = 0.84, 95% CI = 0.54-1.30; p = 0.430).  Table 1 shows a summary of the clinical and demographic characteristics of the study population. Of the 440 subjects, 200 had AF and 240 were controls. No statistical differences were found between the two groups regarding age, gender, BMI, hypertension, diabetes mellitus, smoking, hypercholesterolemia, and history of coronary artery disease (CAD). The proportion of subjects using angiotensin receptor blockers, β-blockers, calcium antagonists, diuretics, statins, digoxin, aspirin, and oral anticoagulants, were significantly higher in patients with AF than in control subjects. The genotype and allele frequencies of the MMP9 SNP rs3918242 in the control and AF groups are shown in Table 2. No significant deviation from the Hardy-Weinberg equilibrium was observed for the MMP9 SNP rs3918242 in either the cases or the controls. Both the distribution of rs3918242 genotypes and allele frequencies were not significantly different between the control and AF groups (p = 0.190, and 0.424, respectively; Table 2). After adjustment for age and gender, the presence of the rs3918242 T allele was still not associated with the risk of AF using an additive inheritance model (odds ratio = 0.84, 95% CI = 0.54-1.30; p = 0.430).  The genotype data of 4 controls and 2 AF patients were missing.

Lack of Association of SNP rs3918242 with MMP9 Expression in Atrial Tissues
To assess whether the SNP rs3918242 is associated with the MMP9 expression in atrial tissues of AF, six AF patients, three with CC homozygotes and three with CT heterozygotes for rs3918242, were elected for comparison. Three SR patients (all CC homozygotes) were assigned as controls. Table 3 lists the baseline characteristics of these nine patients. As shown in Figure 4, MMP9 expression was significantly increased in the atrial myocytes (determined by co-localization with α-actin) of AF patients than of sinus controls. However, no significant difference was observed for the MMP9 expression in myocytes in the atria of AF patients between the SNP rs3918242 CT heterozygotes and the wild-type CC homozygotes.   . At least five random fields were chosen to observe >30 myocytes with scanning and averaging. Data are expressed as mean ± SE; * represent p < 0.05, the significant difference compared with SR controls.

Discussion
Hoit et al. demonstrated that rapid atrial pacing-induced atrial failure is associated with a significant increase in atrial MMP9 activity in the dog model [22]. Chen et al. reported that the expression of the atrial MMP9 level was increased by rapid atrial pacing in pigs, and the expression of MMP9 was mainly localized in the interstitium of atrial myocardium [23]. In addition to macrophage known to produce MMP9 in atrial myocardium [11], our study demonstrated that rapid pacing of HL-1 atrial myocytes also expressed and secreted MMP9 in a time-and dose-dependent manner. Notably, SNP rs3918242 might affect the transcriptional activity of the MMP9 gene and its responsiveness to rapid pacing. The increasing promoter activity as a result from the rs3918242 C to T was further enhanced in HL-1 myocytes with rapid pacing. These findings suggest that atrial electrical remodeling could result in structural remodeling, which form the vulnerable substrates for the initiation and maintenance of AF, as the well-known concept of "AF begets AF". It is possible that the MMP9 SNP rs3918242 may result in different susceptibility of the atrial tissue to electrical stimuli, and participate in the interaction of electrical remodeling and structural remodeling.
In a previous study, the rs3918242-TT genotypes in the promoter region of MMP9 gene is independently and significantly associated with high plasma MMP9 level in the same ethnic population [16]. A prospective cohort study suggested that each one standard deviation increase in plasma MMP9 levels was associated with 27% increase in risk of AF [13]. Thus, the rs3918242-T allele

Discussion
Hoit et al. demonstrated that rapid atrial pacing-induced atrial failure is associated with a significant increase in atrial MMP9 activity in the dog model [22]. Chen et al. reported that the expression of the atrial MMP9 level was increased by rapid atrial pacing in pigs, and the expression of MMP9 was mainly localized in the interstitium of atrial myocardium [23]. In addition to macrophage known to produce MMP9 in atrial myocardium [11], our study demonstrated that rapid pacing of HL-1 atrial myocytes also expressed and secreted MMP9 in a time-and dose-dependent manner. Notably, SNP rs3918242 might affect the transcriptional activity of the MMP9 gene and its responsiveness to rapid pacing. The increasing promoter activity as a result from the rs3918242 C to T was further enhanced in HL-1 myocytes with rapid pacing. These findings suggest that atrial electrical remodeling could result in structural remodeling, which form the vulnerable substrates for the initiation and maintenance of AF, as the well-known concept of "AF begets AF". It is possible that the MMP9 SNP rs3918242 may result in different susceptibility of the atrial tissue to electrical stimuli, and participate in the interaction of electrical remodeling and structural remodeling.
In a previous study, the rs3918242-TT genotypes in the promoter region of MMP9 gene is independently and significantly associated with high plasma MMP9 level in the same ethnic population [16]. A prospective cohort study suggested that each one standard deviation increase in plasma MMP9 levels was associated with 27% increase in risk of AF [13]. Thus, the rs3918242-T allele was expected to associate with increased risk of AF. However, our study did not confirm the hypothesis. The present case-control study showed that the presence of the rs3918242 T allele was not significantly associated with the risk of AF. Previous genome wide association studies (GWAS) enrolling thousands of individuals only identified a few susceptible loci for AF with small odds ratios (1.1-1.84) [24][25][26][27][28]. The sample size of our study could only determine a relative risk of >1.85 for the rs3918242 T allele carriers being a risk factor for AF with a power of 80% at an alpha level of 0.05 in a population with an incidence of the T allele carriers of 23.7% [29]. Thus, it is not unexpected that current study did not find significant association between rs3918242 and AF. Therefore, our results still do not exclude the possibility of a small relative risk of AF associated with the rs3918242 T allele given this small sample size. In the other hand, plasma MMP9 is not specific to the atrial origin. Addition to macrophage and myocytes, fibroblasts, smooth muscle cells, and endothelial cells all secreted MMPs. The association between MMP9 and AF may not be a causal relation, but rather a marker of the ongoing ECM metabolism in cardiovascular system as a result from upstream insults like RAA activation, inflammation and oxidative stress.
In accordance with most of the previous reports [11,30,31], our study demonstrated that the expression of MMP9 in the atrial tissue of AF was greater than that in atrial tissue of SR, suggesting that MMP9 may contribute to the pathogenesis of AF through atrial structural remodeling and atrial dilation. Consistent to the result of genetic association study, the rs3918242 T allele was not significantly associated with the MMP9 expression in atrial myocytes of AF patients. The difference may be explained by the heterogeneity of the studied atrial tissue, which included patients with CAD and valvular heart disease. Anne et al. [32] reported that MMP9 was down-regulated in the mitral stenosis patients compared to coronary bypass patients both in the right and left atria, and without a difference between the SR and AF groups. Another possibility is that other functional SNPs in the MMP9 promoter may affect the transcriptional activity of the MMP9 gene. Furthermore, only nine samples of the atrial tissue were studied in the present study. The influence on the expression of MMP9 in atrial tissue by acquired factors, such as mitral valve disease, heart failure, ischemia and medication may outweigh the effect of a single genetic variant [33][34][35][36][37][38].
Some limitations exist in the present study. The main limitations were its relatively small sample size and lack of plasma MMP9 information. Second, we did not assess the effect of expression of TIMPs in either rapid-pacing HL-1 cells or atrial tissue. Third, we did not have atrial tissues from AF patients with rs3918242 TT homozygotes. Finally, the examined subjects were ethnically Chinese, and hence, it is unclear whether our results can be generalized to other ethnic groups.

Ethics Statement
All participants provided written informed consent. The study was approved by the Human Research Ethics Committee at Chang Gung Memorial Hospital (Chang Gung Medical Foundation Institutional Review Board 100-2922B and 100-3196C1) and was conducted in concordance with the Declaration of Helsinki principles.

Study Population
The study recruited patients who were <65 years of age and had unknown causes of AF. Patients who had a history of hyperthyroidism, severe valvular heart disease (>grade II mitral regurgitation and/or aortic regurgitation), or congestive heart failure (left ventricular ejection fraction <50%) were excluded. The control group has sinus rhythm (SR) with comparable age and gender, and was enrolled from a population receiving routine health examinations and a population receiving regular hypertension treatment in an outpatient clinic. The demographic details of the AF patients and control subjects have been reported elsewhere [39,40].

Clinical Assessment
The presence of AF was documented by patient history, serial electrocardiograms (ECGs), and/or ambulatory ECG recording. Echocardiography was performed to measure left atrial and left ventricular functions and to detect severe valvular diseases. Left atrial enlargement and left ventricular dysfunction were defined as diameter >40 mm and ejection fraction <50%, respectively. Hypertension was defined as a systolic blood pressure (BP) of ě140 mmHg and/or a diastolic BP of ě90 mmHg, or receiving antihypertensive drugs. Hypercholesterolemia and diabetes mellitus were defined in accordance with the third report of the National Cholesterol Education Program and the guidelines of the American Diabetes Association, respectively.

Genomic DNA Extraction and Genotyping of the Rs3918242
Genomic DNA was extracted from peripheral blood leukocytes by use of the Puregene DNA Isolation Kit (Qiagen, Minneapolis, MN, USA). Genotyping of the SNP rs3918242 was performed by polymerase chain reaction (PCR) followed by restriction enzyme digestion, as described previously [16].

Human Samples
Tissue samples from the right atrial appendages were obtained from 6 patients with AF and 3 controls with SR undergoing open heart surgery. After excision, atrial tissues were immediately frozen in liquid nitrogen and stored at´85˝C. Subsequently, genomic DNA from each patient was subjected to genotype analysis as described above.

Immunoblotting
For the isolation of proteins from HL-1 myocytes, the cells were washed twice in in phosphate-buffered saline (PBS) and then lysed by the addition of lysis buffer containing 5 mM EDTA, 250 mM NaCl, 50 mM HEPES, 0.1% NP40. Equal amounts of protein in SDS-PAGE sample buffer were subjected to electrophoresis on 8% SDS-polyacrylamide gels. After transfer to nitrocellulose membranes (PerkinElmer, Waltham, MA, USA), proteins were incubated with primary antibodies against anti-MMP9 antibody (ab119906; Abcam, Cambridge, UK) for 24 h, and then membranes were incubated with an anti-mouse horseradish peroxidase antibody for 1 h. Signals were detected by ECL-detection (Amersham, Roosendaal, The Netherlands) and quantified by densitometry. Signal-bands were in the linear immunoreactive range and expressed normalized to tubulin.

MMP9 Gelatin Zymography
HL-1 cells were incubated in serum-free medium at different conditions as described above. Conditioned medium supernatant was concentrated using Vivaspin 500 (GE healthcare, Buckinghamshire, UK). Electrophoresis was performed on a 10% polyacrylamide/SDS gel containing 1 mg/mL gelatin. The gel was washed with 2.5% Triton X-100 for 1 h at room temperature to remove SDS and incubated overnight at 37˝C in enzyme buffer containing 50 mM Tris, pH 7.5, 200 mM NaCl, 5 mM CaCl 2 and 0.02% Brij-35. Areas of gelatin degradation, identified as MMP activity, appeared as distinct white bands after staining the gels with 0.5% Coomassie brilliant Blue and destained with 10% acetic acid.

MMP9 Promoter Activity Assay
Fragments of genomic MMP9 promoters containing the SNP rs3918242 C/T were extracted from patients. A «2 kb MMP9 promoter (´2181 to +11) was amplified by PCR using primers as described previously [17]. The PCR product was subcloned into the pGL3-Basic vector (Promega, Madison, WI, USA) at SacI and XhoI. For transient transfection assays, HL-1 myocytes grown to 50%-60% confluence were transfected with indicated plasmids using LipofectAMINE 2000 reagent (Life Technologies, Gaithersburg, MD, USA) according to the manufacturer instructions. This method was resulted in approximately 60% transfection efficiency. The cells transfected with MMP9 promoters containing´1562 C or´1562 T were paced at 4 Hz for 6 h. Luciferase activities were measured with a luminometer (GloMax 20/20 Luminometer, Promega, Madison, WI, USA) and normalized by cellular protein concentrations.

Immunohistochemical Analysis
Immunohistochemical analysis was conducted using MMP9 and α-actin, primary antibodies (Abcam, Cambridge, MA, USA) followed by either fluorescein isothiocyanate (FITC)-conjugated or Cy3-conjugated secondary antibodies (Chemicon, Temecula, CA, USA). Nuclei were stained with DAPI. The expression levels of target proteins were calculated as protein-occupied area in the tissue divided by the nuclear area. For each analysis, at least 5 random fields were chosen to observe >30 myocytes.

Statistical Analysis
The clinical characteristics of the continuous variables are expressed as means˘standard deviation (SD) and tested using the two-sample t-test. The chi-square test was used to examine the differences in categorical variables and to compare the genotype and allele frequencies. Binary logistic regression analysis was used to evaluate the independent association of SNP rs3918242 with AF after adjustment for age and gender. Unpaired Student's t-test and one-way ANOVA with linear trend tests were applied for 2 groups and dose-response relationships comparisons, respectively. All calculations were performed with standard statistical SPSS software (version 20 for Windows, Inc., Chicago, IL, USA). Values of p < 0.05 using a two-sided test were considered statistically significant.

Conclusions
Our study demonstrated that rapid pacing induced MMP9 secretion of HL-1 atrial myocytes in a time-and dose-dependent manner, which suggested that atrial electrical remodeling might contribute in structural remodeling through the increased secretion of MMP9 by atrial myocyte. Although the MMP9 expression of human atrial myocyte is associated with AF, our study did not support the association of susceptibility to AF among Taiwanese with the MMP9 gene functional polymorphism rs3918242.