- Email: [email protected]
Selective Induction of Matrix Metalloproteinases and Tissue Inhibitor of Metalloproteinases in Atrial and Ventricular Myocardium in Patients With Atrial Fibrillation
Selective Induction of Matrix Metalloproteinases and Tissue Inhibitor of Metalloproteinases in Atrial and Ventricular Myocardium in Patients With Atrial Fibrillation Rupak Mukherjee, PhDa,*, Amanda R. Herron, BSa, Abigail S. Lowry, BSa, Robert E. Stroud, MSa, Martha R. Stroud, MSa, J. Marcus Wharton, MDb, John S. Ikonomidis, MD, PhDa, A. Jackson Crumbley III, MDa, Francis G. Spinale, MD, PhDa, and Michael R. Gold, MD, PhDb Atrial fibrillation (AF) produces changes in atrial structure and extracellular matrix composition, which is regulated by matrix metalloproteinases (MMPs). Moreover, AF often occurs in the setting of congestive heart failure (CHF), which also affects MMPs. Whether changes in MMPs or the tissue inhibitors of metalloproteinases (TIMPs) within atrial and ventricular myocardium are differentially regulated with AF remains unclear. Myocardium from the walls of the right atrium, right ventricle, left atrium, and left ventricle was obtained from the explanted hearts of 43 patients with end-stage CHF. AF was present in 23 patients (duration 1 to 84 months). The remaining 20 patients served as non-AF controls. The groups were well matched clinically, but left atrial (LA) size was increased in the AF cohort (5.5 ⴞ 0.8 vs 4.9 ⴞ 0.7 cm, p <0.05). Myocardial collagen content and levels of MMP-1, -2, -8, -9, -13, and -14, and TIMP-1, -2, -3, and TIMP-4 were determined. With AF, collagen content was greater within the atrial myocardium but less in the ventricular myocardium. There were chamber-specific differences in MMPs and TIMPs with AF. For example, MMP-1 in the right atrium and MMP-9 in the left atrium were greater with AF. TIMP-3 levels were greater in the right ventricle, left atrium, and left ventricle. Although total LA collagen was positively correlated with AF duration (r ⴝ 0.49, p <0.03), there was an inverse relation between soluble collagen I and AF duration (n ⴝ 6, r ⴝ ⴚ0.84, p <0.04). In conclusion, AF is associated with chamber-specific alterations in myocardial collagen content and MMP and TIMP levels, indicative of differential remodeling and altered collagen metabolism. Differences in MMP and TIMP profiles may provide diagnostic and mechanistic insights into the pathogenesis of AF with CHF. © 2006 Elsevier Inc. All rights reserved. (Am J Cardiol 2006;97: 532–537)
Despite the presence of atrial fibrillation (AF) in both atria, the predominant structural changes, including dilation, occur in the left atrium,1– 4 suggesting that there may be differential activation of remodeling pathways with AF.4,5 The ventricular remodeling response to stress stimuli, such as chronic pressure or volume overload, may also occur in a differential manner.6,7 Taken together, these findings suggest that cardiac remodeling may occur in a chamber-specific manner. Enzymes implicated in myocardial remodeling include the matrix metalloproteinases (MMPs).8 –14 Changes in myocardial levels of many MMP
a
Division of Cardiothoracic Surgery, Department of Surgery, and Division of Cardiology, Department of Medicine, Medical University of South Carolina, Charleston, South Carolina. Manuscript received June 7, 2005; revised manuscript received and accepted August 24, 2005. This study was supported by Grants HL-66029, HL-45024, HL97012, and PO1-48788 from the National Heart, Lung, and Blood Institute, Bethesda, Maryland. * Corresponding author: Tel: 843-876-5186; fax: 843-876-5187. E-mail address: [email protected] (R. Mukherjee).
b
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2005.08.073
types and the endogenous tissue inhibitors of the metalloproteinases (TIMPs) occur in a number of cardiac disease states.3,6,8 –13,15 However, it remains unclear whether the superimposition of AF differentially modulates MMP and/or TIMP levels in the atria and ventricles of patients with end-stage heart failure. Accordingly, this study tested the hypothesis that changes in the abundance of MMPs and TIMPs would show chamber specificity and that the presence of AF would differentially modulate myocardial MMP and TIMP levels and extracellular matrix composition. Methods Myocardial samples: Explanted hearts were obtained from patients who underwent cardiac transplantation for end-stage New York Heart Association class III or IV congestive heart failure (CHF; n ⫽ 43, mean age 54 ⫾ 10 years). Informed consent for this study, which was approved by the institutional review board, was obtained from all patients. The samples used in this study were obtained from www.AJConline.org
Arrhythmias and Conduction Disturbances/MMPs and TIMPs With AF
533
Table 1 Antibodies used to determine levels of matrix metalloproteinases and tissue inhibitors of the metalloproteinases Antigen
Catalog No.
From
Concentration
Secondary
Standard
MMP-1 MMP-8 MMP-13 MMP-14 TIMP-1 TIMP-2 TIMP-3 TIMP-4
IM35L PC493 AB8114 AB815 AB8116 IM11L CL2T3 AB816
Oncogene Oncogene Chemicon Chemicon Chemicon Oncogene Cedarlane Chemicon
0.2 g/ml 0.2 g/ml 0.2 g/ml 0.2 g/ml 0.2 g/ml 0.2 g/ml 0.4 g/ml 0.1 g/ml
Vector mouse Vector rabbit Vector rabbit Vector rabbit Vector rabbit Vector mouse Vector rabbit Vector rabbit
CC1031 CC067 CC068 CC1043 CC1062 PF021 CC1065 H-TIMP-4
From Chemicon Chemicon Chemicon Chemicon Chemicon Oncogene Chemicon Cedarlane
Standard: recombinant standard included with immunoblot.
consecutive explanted hearts at our institution from 1997 to October 2003, for which myocardium from all 4 cardiac chambers was available. Tissue samples from right atrial (RA), right ventricular (RV), left atrial (LA), and left ventricular (LV) walls of the explanted hearts were obtained. Because the atrial appendages of the recipient hearts were variably retained during transplantation, RA and LA samples were obtained from the lateral atrial walls. Normal LV myocardial samples (n ⫽ 6, age range 33 to 45 years) without a history of cardiac disease were obtained from donor hearts not matched for transplantation or used for valve harvest (Cryolife, Inc., Kennesaw, Georgia) and used as reference controls.10,14 All myocardial samples were stored at ⫺80°C until analyses. Clinical data: A review of the medical records of these patients revealed that 23 patients had a pretransplantation history of persistent AF, with AF being documented in the last pretransplantation electrocardiogram. For the patients with AF, the date at which AF was first diagnosed was recorded to calculate AF duration. The remaining 20 patients served as non-AF controls. Echocardiography was performed on all patients as part of routine clinical care (5.4 ⫾ 3.8 months) before transplantation. LA size, LV dimensions, LV fractional shortening, and the ejection fraction were measured by standard techniques. Biochemistry: Biochemical assays for the abundance of MMP-1, -2, -8, -9, -13, and -14, and TIMP-1, -2, -3, and TIMP-4, and collagen content were performed using previously described techniques.10 –14 Briefly, myocardial samples were weighed, homogenized, and standardized to a protein concentration of 1 g/l. For immunoblot analyses, samples were loaded into a gel, separated electrophoretically, and transferred onto nitrocellulose membranes. The membranes were incubated with primary antisera (Table 1) at room temperature for 1 hour. After washing (1⫻ trisbuffered saline with 0.1% Tween-20, USB Corporation, Cleveland, Ohio), the membranes were incubated for 1 hour in peroxidase-conjugated secondary antibody (1:5,000 dilution, either Vector antimouse or Vector antirabbit, Vector Laboratories Inc., Burlingame, California; Table 1) and subjected to chemiluminescent activation (PerkinElmer, Inc., Boston, Massachusetts). The luminescent signal was
detected by exposure to x-ray film (Hyperfilm MP, Amersham Biosciences UK Ltd., Little Chalfont, United Kingdom). Appropriate recombinant standards were included in every immunoblot (Table 1). Relative levels of MMP-2 and -9 were determined using substrate-specific zymography as previously described.11,12 A commercially available purified MMP-2 and -9 standard (CC073, Chemicon, Temecula, California) was included in every zymogram, as were the internal reference control samples. Immunoblots and zymograms were digitized (Expression 1680, Epson, Nagano, Japan), and the banding patterns corresponding to the molecular weight of the positive signal from the recombinant standards were quantified (Gel Pro Analyzer, Media Cybernetics, Silver Spring, Maryland) and expressed in terms of an integrated optical density (IOD). The linearity of densitometric analysis was confirmed by ensuring that the largest and smallest IOD values for each analyte were recorded within a linear response range. Total collagen content was determined through a colorimetric reaction against picrosirius red.16,17 Briefly, 100 l of sample homogenates were dehydrated and stained with 0.1% picrosirius red in saturated picric acid (w/v) in a 96-well plate. The dye was solubilized and absorbance read at 540 nm. Readings were converted to protein units using a linear calibration curve generated from standards (Vitrogen 100, Angiotech Biomaterials, Palo Alto, California) and normalized to the weight of each myocardial sample. In a subset of patients with AF (n ⫽ 6), LA collagen I levels were determined by immunoblotting (anticollagen I, CL50151AP, Cedarlane Laboratories, Ltd., Hornby, Ontario, Canada). Statistical analysis: Echocardiographic parameters were compared between the non-AF and AF groups using the unpaired Student’s t test. For biochemical analyses, duplicate values were averaged and treated as a complete observation for that sample. Comparisons between the reference normal control group and the CHF group were performed by averaging IOD values for the normal samples and using the resultant value to normalize the IOD values for samples from the CHF group. Similarly, to examine the effects of AF on any of these biochemical determinants, IOD values for the non-AF samples from each zymogram or immuno-
534
The American Journal of Cardiology (www.AJConline.org) Table 2 Clinical parameters recorded before transplantation in patients without and with atrial fibrillation Variable
Figure 1. Representative immunoblots and zymograms for MMP-1, -2, -8, -9, -13, and -14, and TIMP-1, -2, -3, and TIMP-4. For the immunoblots, the areas of protein staining are depicted as dark bands on a lighter background. For the zymograms (MMP-2 and -9), the areas of substrate lysis caused by these proteases are depicted as lighter bands on a dark background. Myocardial samples from the right atrium (RA), right ventricle (RV), left atrium (LA), and left ventricle (LV) of patients with a documented history of AF (A) or no known AF history (N) were homogenized and electrophoretically separated. There were AF-associated chamberspecific differences in the levels of certain MMP and TIMP types. For example, MMP-14 levels were lower in the RA with AF, and there was a differential increase in left atrial MMP-9 levels with AF.
blot were averaged and used to normalize the IOD values for the samples from the AF group. Therefore, changes in the levels of collagen, the MMPs, and the TIMPs in each of the cardiac chambers were expressed as percentages of non-AF values of 100% and compared using a 2-sided Student’s t test. Correlations between MMP levels, TIMP levels, collagen content, and AF duration and the ejection fraction were examined using least-squares linear regression analyses. When appropriate, comparisons of categorical variables between the non-AF and AF groups were performed using Pearson’s chi-square analysis. All statistical analyses were performed using the BMDP software package (BMDP Statistical Software, Los Angeles, California). Results are presented as mean ⫾ SD. Values of p ⬍0.05 were considered to be statistically significant.
Non-AF (n ⫽ 20)
Age (yrs) LV size (cm) LV fractional shortening (%) LV ejection fraction (%) LA size (cm) Ischemic/nonischemic CHF Pretransplantation medications ACE inhibitors or ARB Aldosterone inhibitors Amiodarone Aspirin Beta-adrenergic receptor blockade Calcium channel blockade Coumadin Digoxin Diuretics Inotropes Nitrates Statins
AF (n ⫽ 23)
All (n ⫽ 46)
56 ⫾ 9 54 ⫾ 10 54 ⫾ 10 7.2 ⫾ 0.9 7.4 ⫾ 1.0 7.3 ⫾ 0.9 12.2 ⫾ 6.0 13.8 ⫾ 5.1 13.0 ⫾ 5.5 25.3 ⫾ 10.8 23.9 ⫾ 7.5 24.6 ⫾ 9.2 4.9 ⫾ 0.7 5.5 ⫾ 0.8* 5.2 ⫾ 0.8 8/12 11/12 19/24 18 (90%) 9 (45%) 3 (15%) 7 (35%) 10 (50%)
16 (70%) 7 (30%) 11 (48%)* 9 (39%) 5 (22%)
34 (79%) 16 (37%) 14 (33%) 16 (37%) 15 (35%)
1 (5%) 11 (55%) 14 (70%) 17 (85%) 2 (10%) 10 (50%) 13 (65%)
3 (13%) 19 (83%)* 18 (78%) 22 (96%) 7 (30%) 11 (48%) 10 (43%)
4 (9%) 30 (70%) 32 (74%) 39 (91%) 9 (21%) 21 (49%) 23 (53%)
Values are presented as mean ⫾ SD. * p ⬍0.05 versus non-AF. ACE ⫽ angiotensin-converting enzyme; ARB ⫽ angiotensin receptor blockade. Table 3 Matrix metalloproteinase levels, tissue inhibitor of metalloproteinase levels, and collagen content in right atrial, right ventricular, left atrial, and left ventricular myocardium of patients with atrial fibrillation and normalized to values obtained from non–atrial fibrillation hearts Variable
RA
RV
LA
LV
MMP-1 MMP-2 MMP-8 MMP-9 MMP-13 MMP-14 TIMP-1 TIMP-2 TIMP-3 TIMP-4 Collagen
139 ⫾ 49* 111 ⫾ 48 70 ⫾ 36* 130 ⫾ 99 107 ⫾ 45 56 ⫾ 26* 154 ⫾ 141 116 ⫾ 82 122 ⫾ 89 96 ⫾ 13 171 ⫾ 68*
118 ⫾ 45 96 ⫾ 35 121 ⫾ 96 112 ⫾ 96 105 ⫾ 51 122 ⫾ 50 161 ⫾ 125* 110 ⫾ 59 141 ⫾ 71* 100 ⫾ 16 79 ⫾ 41*
121 ⫾ 61 97 ⫾ 37 99 ⫾ 65 166 ⫾ 123* 103 ⫾ 41 112 ⫾ 126 122 ⫾ 94 102 ⫾ 90 202 ⫾ 194* 106 ⫾ 15 155 ⫾ 71*
123 ⫾ 68 106 ⫾ 33 120 ⫾ 48 113 ⫾ 112 112 ⫾ 39 108 ⫾ 33 100 ⫾ 47 138 ⫾ 79* 134 ⫾ 150* 99 ⫾ 17 90 ⫾ 13*
Values are presented as mean ⫾ SD. * p ⬍0.05 versus non-AF values of 100%.
Results Representative immunoblots for MMP-1, -8, -13, -14, and TIMP-1, -2, -3, and TIMP-4 and zymograms for MMP-2 and -9 are shown in Figure 1. To assess the effect of advanced CHF on these measures, the pooled results from patients with CHF (AF and non-AF cohorts) were compared with LV samples from subjects with no structural heart disease. Irrespective of AF status, LV myocardial levels of MMP-8 (256 ⫾ 69%), MMP-13 (309 ⫾ 125%), MMP-14 (295 ⫾ 80%), and TIMP-2 (210 ⫾ 28%) were increased and
TIMP-1 levels (30 ⫾ 31%) were decreased with CHF compared with reference normal samples (p ⬍0.05 vs the normal value of 100%). Clinical characteristics of the patients with CHF are listed in Table 2. The mean age, LV end-diastolic dimension, and LV ejection fraction were not different between the non-AF and AF groups. Moreover, the percentages of subjects with CHF with ischemic causes versus nonischemic causes did not differ between groups (Pearson’s chi-square 0.61). As expected, LA size was larger in patients with AF, as
Arrhythmias and Conduction Disturbances/MMPs and TIMPs With AF
Figure 2. There was a positive relation between LA collagen content and the duration of documented AF (y ⫽ 0.300 ⫹ 0.004x, r ⫽ 0.49, p ⫽ 0.023, SEE 0.148).
was the pretransplantation use of amiodarone and warfarin (Table 2). All other pretransplantation medications were similar between the 2 groups. For the AF group, this arrhythmia was present for 23 ⫾ 25 months (range 1 to 84). AF differentially affected selected MMPs and TIMPs. Cardiac chamber–specific differences in the levels of MMP-1, -9, and -14, and TIMP-3 were observed with AF compared with reference non-AF values (Table 3). Specifically, MMP-1 levels in the right atrium and MMP-9 levels in the left atrium were higher with AF. MMP-8 and -14 levels were lower in the right atrium with AF. TIMP-1 levels in the right ventricle and TIMP-3 levels in the right ventricle, left atrium, and left ventricle were higher with AF. Collagen content in the right atrium and the left atrium was greater with AF, whereas collagen content was less in the right ventricle and left ventricle in the AF group compared with non-AF values (Table 3). There was a linear correlation between AF duration and total LA collagen content (Figure 2). In a subset of patients with AF, collagen I levels were determined by immunoblotting. Immunopositive, or soluble, collagen I levels in the left atrium tended to be higher in the AF group than the non-AF group (214 ⫾ 101 vs 123 ⫾ 117 pixels). However, this difference did not achieve statistical significance (p ⫽ 0.12). Nevertheless, when expressed as a percentage of non-AF values, soluble collagen I content in the left atrium was greater with AF (175 ⫾ 83%, p ⫽ 0.011). In a subset of the AF samples for which soluble collagen content in the left atrium was determined, there was an inverse relation between soluble collagen I levels and AF duration (y ⫽ 296.15 ⫺ 2.84x, r ⫽ ⫺0.84, p ⫽ 0.036, SEE 75.20). Discussion AF is associated with a number of changes in the extracellular matrix (ECM) of the atrial myocardium, including
535
increases in intracellular glycogen accumulation, interstitial fibrosis, and a loss of intermyocyte connections.2,18,19 Because the MMPs and the TIMPs constitute a key system that regulates ECM composition and structure, the present study examined levels of collagen and MMP and TIMP species in samples obtained at the time of transplantation from all 4 cardiac chambers of the end-stage hearts of patients with and without persistent AF. There are several unique and important findings of the present study. First, there were chamber-specific differences in collagen content and levels of MMP and TIMP types with AF. For example, in patients with AF, collagen content was greater in the right atrium and left atrium and less in the right ventricle and left ventricle. AF was also associated with differential changes in MMP and TIMP levels in the atria and the ventricles. Second, in the left atrium, collagen content was linearly correlated with AF duration. Finally, there was an inverse correlation between AF duration and soluble collagen I levels. Taken together, these findings suggest that several post-translational events with respect to collagen metabolism occurred in the left atrium as a function of AF duration. It is important that the observed changes of myocardial levels of collagen, MMPs, and TIMPs occurred over and above those associated with other myocardial remodeling processes associated with LV dilation and systolic dysfunction. Although AF is primarily considered to be an electrical disease,20 it can result in changes in atrial structure, which make the atrial myocardium more susceptible to the maintenance of the arrhythmia (“AF begets AF”).21 Structural abnormalities documented to occur with AF include myocyte loss, the hypertrophy of existing myocytes, and altered composition of the ECM.2,4,5 The ECM is a dynamic structure that provides supportive scaffolding for myocytes and is critical for the maintenance of the structural integrity and the overall geometry of the cardiac chambers. In a canine model of AF induced by rapid ventricular pacing, Hanna et al5 demonstrated that LA collagen content increased earlier and achieved higher levels than in the LV myocardium. Changes in myocardial structure and/or geometry, collectively termed remodeling, can occur because of changes in the conformation and composition of the ECM.10 –12 Consistent with the findings of past studies,1– 4,18,19,22,23 an increase of the collagen content in the right atrium and left atrium was associated with the presence of AF in the present study. Moreover, the present study demonstrated an inverse correlation in the amount of soluble collagen I and the duration of AF, suggesting an increase of collagen cross linking as AF perpetuates. It is intriguing to speculate that this is 1 mechanism leading to the arrhythmia becoming permanent and irreversible. Using transgenic mice constitutively overexpressing transforming growth factor-1, Verheule et al24 recently demonstrated that increased atrial interstitial fibrosis was associated with greater AF inducibility, in the absence of abnormalities in action potentials. In a canine model of pacing-induced LV dysfunction, Hanna et al5 reported that there was a correlation between
536
The American Journal of Cardiology (www.AJConline.org)
P-wave duration and the amount of atrial fibrosis. Taken together, these findings raise the possibility that atrial fibrosis may, in and of itself, create a substrate conducive to the perpetuation of AF, and modification of this process may alter AF incidence and duration. The study by Hanna et al5 also provided a potential explanation for the differences with respect to atrial versus ventricular collagen content with AF. Specifically, in this past study, the progressive prolongation of the P-wave duration was associated with a rapid increase in angiotensin II content in the LA myocardium concomitant with an increase in the profibrotic cytokine transforming growth factor-.5 Given that recent clinical studies have reported a lower AF incidence in patients with CHF and hypertension who were prescribed angiotensin-converting enzyme inhibitors,25,26 it is likely that changes in atrial angiotensin II content contribute, at least in part, to atrial fibrosis with AF. Although a causative role of the angiotensin system in the development of fibrosis with AF remains to be elucidated, the findings of the present study taken together with those of past studies raise the intriguing possibility of targeting atrial fibrosis as a potential means to limit the incidence of AF. Previously, indirect evidence that changes in MMP or TIMP levels occur with AF was provided by measurements of plasma levels.27 In the present study, MMPs and TIMPs were measured directly from the myocardium of all 4 cardiac chambers, confirming the origin of these proteins. It is important that the findings of the present study demonstrated that AF was associated with alterations in the levels of certain key MMP and TIMP species in the atrial myocardium. Specifically, levels of MMP-9 and TIMP-3 were differentially increased in the left atrium and MMP-14 levels lower in the right atrium with AF. Two recent studies have examined a limited number of MMP and/or TIMP species in atrial samples from patients with AF.3,15 Xu et al3 reported that atrial MMP-2 levels were increased and TIMP-2 levels reduced with AF. Nakano et al15 demonstrated that although MMP-2 levels were unchanged with AF, RA MMP-9 messenger ribonucleic acid and protein levels were increased with AF. Consistent with the findings of these past studies, LA MMP-9 levels were observed to be differentially increased with AF in the present study. In addition, the findings of the present study provide evidence that regional changes of collagen content occurred as a function of changes in the proteolytic capacity in the myocardium of each chamber. For example, the lower MMP-14 levels in the right atrium with AF may have contributed to a reduction of collagen proteolysis. Similarly, the increase in LA collagen content may have been due to increased TIMP-3 levels with AF. These observations provide evidence that a likely mechanistic underpinning of interstitial atrial fibrosis with AF is changes in MMP and TIMP abundance and/or MMP and TIMP stoichiometry. The findings of the present study should be interpreted in light of certain limitations. First, all patients evaluated had a history of advanced CHF, which may have affected the
abundance of the MMPs and TIMPs. Comparisons were made with LV measurements from a small group of normal controls, demonstrating important effects of CHF on MMPs and TIMPs. Therefore, a direct comparison between chamber-specific MMP and TIMP levels obtained in the present study and those obtained from the other chambers of the nonfailing hearts could not be performed. Moreover, myocardial samples of the CHF hearts were obtained from available portions of the free wall of each of the cardiac chambers, precluding the analysis of within-chamber variability for any of the analytes. Nevertheless, these findings demonstrate that the presence of AF resulted in a differential regulation of MMPs and TIMPs in failing hearts and that these changes were over and above those that may have been caused by the underlying cardiac disease state. Second, MMPs and TIMPs were assayed using semiquantitative immunoblotting or zymographic techniques. Therefore, direct comparisons of changes in the levels of these proteins among the cardiac chambers could not be performed. Finally, myocardial but not plasma levels of the MMP and TIMP types were measured. Further studies are needed to determine the relation between myocardial and plasma levels of MMPs and TIMPs to evaluate clinical opportunities for potential diagnostic and prognostic tests of these findings for patients with AF. 1. Boldt A, Wetzel U, Lauschke J, Weigl J, Gummert J, Hindricks G, Kottkamp H, Dhein S. Fibrosis in left atrial tissue of patients with atrial fibrillation with and without underlying mitral valve disease. Heart 2004;90:400 – 405. 2. Kostin S, Klein G, Szalay Z, Hein S, Bauer EP, Schaper J. Structural correlate of atrial fibrillation in human patients. Cardiovasc Res 2002; 54:361–379. 3. Xu J, Cui G, Esmailian F, Plunkett M, Marelli D, Ardehali A, Odim J, Laks H, Sen L. Atrial extracellular matrix remodeling and the maintenance of atrial fibrillation. Circulation 2004;109:363–368. 4. Li D, Shinagawa K, Pang L, Leung TK, Cardin S, Wang Z, Nattel S. Effects of angiotensin-converting enzyme inhibition on the development of the atrial fibrillation substrate in dogs with ventricular tachypacing-induced congestive heart failure. Circulation 2001;104:2608 – 2614. 5. Hanna N, Cardin S, Leung TK, Nattel S. Differences in atrial versus ventricular remodeling in dogs with ventricular tachypacing-induced congestive heart failure. Cardiovasc Res 2004;63:236 –244. 6. Nagatomo Y, Carabello BA, Coker ML, McDermott PJ, Nemoto S, Hamawaki M, Spinale FG. Differential effects of pressure or volume overload on myocardial MMP levels and inhibitory control. Am J Physiol Heart Circ Physiol 2000;278:H151–H161. 7. Woodiwiss AJ, Tsotetsi OJ, Sprott S, Lancaster EJ, Mela T, Chung ES, Meyer TE, Norton GR. Reduction in myocardial collagen cross-linking parallels left ventricular dilatation in rat models of systolic chamber dysfunction. Circulation 2001;103:155–160. 8. Boixel C, Fontaine V, Rucker-Martin C, Milliez P, Louedec L, Michel JB, Jacob MP, Hatem SN. Fibrosis of the left atria during progression of heart failure is associated with increased matrix metalloproteinases in the rat. J Am Coll Cardiol 2003;42:336 –344. 9. Hoit BD, Takeishi Y, Cox MJ, Gabel M, Kirkpatrick D, Walsh RA, Tyagi SC. Remodeling of the left atrium in pacing-induced atrial cardiomyopathy. Mol Cell Biochem 2002;238:145–150. 10. Thomas CV, Coker ML, Zellner JL, Handy JR, Crumbley AJ III, Spinale FG. Increased matrix metalloproteinase activity and selective
Arrhythmias and Conduction Disturbances/MMPs and TIMPs With AF
11.
12.
13.
14.
15.
16. 17.
18.
upregulation in LV myocardium from patients with end-stage dilated cardiomyopathy. Circulation 1998;97:1708 –1715. Mukherjee R, Brinsa TA, Dowdy KB, Scott AA, Baskin JM, Deschamps AM, Lowry AS, Escobar GP, Lucas DG, Yarbrough WM, et al. Myocardial infarct expansion and matrix metalloproteinase inhibition. Circulation 2003;107:618 – 625. Spinale FG, Coker ML, Krombach SR, Mukherjee R, Hallak H, Houck WV, Clair MJ, Kribbs SB, Johnson LL, Peterson JT, et al. Matrix metalloproteinase inhibition during the development of congestive heart failure: effects on left ventricular dimensions and function. Circ Res 1999;85:364 –376. Wilson EM, Moainie SL, Baskin JM, Lowry AS, Deschamps AM, Mukherjee R, Guy TS, John-Sutton MG, Gorman JH III, Edmunds LH Jr, et al. Region- and type-specific induction of matrix metalloproteinases in post-myocardial infarction remodeling. Circulation 2003;107: 2857–2863. Spinale FG, Coker ML, Heung LJ, Bond BR, Gunasinghe HR, Etoh T, Goldberg AT, Zellner JL, Crumbley AJ. A matrix metalloproteinase induction/activation system exists in the human left ventricular myocardium and is upregulated in heart failure. Circulation 2000;102: 1944 –1949. Nakano Y, Niida S, Dote K, Takenaka S, Hirao H, Miura F, Ishida M, Shingu T, Sueda T, Yoshizumi M, et al. Matrix metalloproteinase-9 contributes to human atrial remodeling during atrial fibrillation. J Am Coll Cardiol 2004;43:818 – 825. Marotta M, Martino G. Sensitive spectrophotometric method for the quantitative estimation of collagen. Anal Biochem 1985;150:86 –90. Tullberg-Reinert H, Jundt G. In situ measurement of collagen synthesis by human bone cells with a sirius red-based colorimetric microassay: effects of transforming growth factor beta2 and ascorbic acid 2-phosphate. Histochem Cell Biol 1999;112:271–276. Frustaci A, Chimenti C, Bellocci F, Morgante E, Russo MA, Maseri A. Histological substrate of atrial biopsies in patients with lone atrial fibrillation. Circulation 1997;96:1180 –1184.
537
19. Ausma J, van der Velden HM, Lenders MH, van Ankeren EP, Jongsma HJ, Ramaekers FC, Borgers M, Allessie MA. Reverse structural and gap-junctional remodeling after prolonged atrial fibrillation in the goat. Circulation 2003;107:2051–2058. 20. Moe GK, Abildskov JA. Atrial fibrillation as a self-sustaining arrhythmia independent of focal discharge. Am Heart J 1959;58:59 –70. 21. Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA. Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation 1995;92:1954 –1968. 22. Aime-Sempe C, Folliguet T, Rucker-Martin C, Krajewska M, Krajewski S, Heimburger M, Aubier M, Mercadier JJ, Reed JC, Hatem SN. Myocardial cell death in fibrillating and dilated human right atria. J Am Coll Cardiol 1999;34:1577–1586. 23. Schotten U, Ausma J, Stellbrink C, Sabatschus I, Vogel M, Frechen D, Schoendube F, Hanrath P, Allessie MA. Cellular mechanisms of depressed atrial contractility in patients with chronic atrial fibrillation. Circulation 2001;103:691– 698. 24. Verheule S, Sato T, Everett T, Engle SK, Otten D, Rubart-von der LM, Nakajima HO, Nakajima H, Field LJ, Olgin JE. Increased vulnerability to atrial fibrillation in transgenic mice with selective atrial fibrosis caused by overexpression of TGF-1. Circ Res 2004;94:1458 –1465. 25. L’Allier PL, Ducharme A, Keller PF, Yu H, Guertin MC, Tardif JC. Angiotensin-converting enzyme inhibition in hypertensive patients is associated with a reduction in the occurrence of atrial fibrillation. J Am Coll Cardiol 2004;44:159 –164. 26. Vermes E, Tardif JC, Bourassa MG, Racine N, Levesque S, White M, Guerra PG, Ducharme A. Enalapril decreases the incidence of atrial fibrillation in patients with left ventricular dysfunction: insight from the Studies of Left Ventricular Dysfunction (SOLVD) trials. Circulation 2003;107:2926 –2931. 27. Marin F, Roldan V, Climent V, Garcia A, Marco P, Lip GYH. Is thrombogenesis in atrial fibrillation related to matrix metalloproteinase-1 and its inhibitor, TIMP-1? Stroke 2003;34:1181–1186.
Despite the presence of atrial fibrillation (AF) in both atria, the predominant structural changes, including dilation, occur in the left atrium,1– 4 suggesting that there may be differential activation of remodeling pathways with AF.4,5 The ventricular remodeling response to stress stimuli, such as chronic pressure or volume overload, may also occur in a differential manner.6,7 Taken together, these findings suggest that cardiac remodeling may occur in a chamber-specific manner. Enzymes implicated in myocardial remodeling include the matrix metalloproteinases (MMPs).8 –14 Changes in myocardial levels of many MMP
a
Division of Cardiothoracic Surgery, Department of Surgery, and Division of Cardiology, Department of Medicine, Medical University of South Carolina, Charleston, South Carolina. Manuscript received June 7, 2005; revised manuscript received and accepted August 24, 2005. This study was supported by Grants HL-66029, HL-45024, HL97012, and PO1-48788 from the National Heart, Lung, and Blood Institute, Bethesda, Maryland. * Corresponding author: Tel: 843-876-5186; fax: 843-876-5187. E-mail address: [email protected] (R. Mukherjee).
b
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2005.08.073
types and the endogenous tissue inhibitors of the metalloproteinases (TIMPs) occur in a number of cardiac disease states.3,6,8 –13,15 However, it remains unclear whether the superimposition of AF differentially modulates MMP and/or TIMP levels in the atria and ventricles of patients with end-stage heart failure. Accordingly, this study tested the hypothesis that changes in the abundance of MMPs and TIMPs would show chamber specificity and that the presence of AF would differentially modulate myocardial MMP and TIMP levels and extracellular matrix composition. Methods Myocardial samples: Explanted hearts were obtained from patients who underwent cardiac transplantation for end-stage New York Heart Association class III or IV congestive heart failure (CHF; n ⫽ 43, mean age 54 ⫾ 10 years). Informed consent for this study, which was approved by the institutional review board, was obtained from all patients. The samples used in this study were obtained from www.AJConline.org
Arrhythmias and Conduction Disturbances/MMPs and TIMPs With AF
533
Table 1 Antibodies used to determine levels of matrix metalloproteinases and tissue inhibitors of the metalloproteinases Antigen
Catalog No.
From
Concentration
Secondary
Standard
MMP-1 MMP-8 MMP-13 MMP-14 TIMP-1 TIMP-2 TIMP-3 TIMP-4
IM35L PC493 AB8114 AB815 AB8116 IM11L CL2T3 AB816
Oncogene Oncogene Chemicon Chemicon Chemicon Oncogene Cedarlane Chemicon
0.2 g/ml 0.2 g/ml 0.2 g/ml 0.2 g/ml 0.2 g/ml 0.2 g/ml 0.4 g/ml 0.1 g/ml
Vector mouse Vector rabbit Vector rabbit Vector rabbit Vector rabbit Vector mouse Vector rabbit Vector rabbit
CC1031 CC067 CC068 CC1043 CC1062 PF021 CC1065 H-TIMP-4
From Chemicon Chemicon Chemicon Chemicon Chemicon Oncogene Chemicon Cedarlane
Standard: recombinant standard included with immunoblot.
consecutive explanted hearts at our institution from 1997 to October 2003, for which myocardium from all 4 cardiac chambers was available. Tissue samples from right atrial (RA), right ventricular (RV), left atrial (LA), and left ventricular (LV) walls of the explanted hearts were obtained. Because the atrial appendages of the recipient hearts were variably retained during transplantation, RA and LA samples were obtained from the lateral atrial walls. Normal LV myocardial samples (n ⫽ 6, age range 33 to 45 years) without a history of cardiac disease were obtained from donor hearts not matched for transplantation or used for valve harvest (Cryolife, Inc., Kennesaw, Georgia) and used as reference controls.10,14 All myocardial samples were stored at ⫺80°C until analyses. Clinical data: A review of the medical records of these patients revealed that 23 patients had a pretransplantation history of persistent AF, with AF being documented in the last pretransplantation electrocardiogram. For the patients with AF, the date at which AF was first diagnosed was recorded to calculate AF duration. The remaining 20 patients served as non-AF controls. Echocardiography was performed on all patients as part of routine clinical care (5.4 ⫾ 3.8 months) before transplantation. LA size, LV dimensions, LV fractional shortening, and the ejection fraction were measured by standard techniques. Biochemistry: Biochemical assays for the abundance of MMP-1, -2, -8, -9, -13, and -14, and TIMP-1, -2, -3, and TIMP-4, and collagen content were performed using previously described techniques.10 –14 Briefly, myocardial samples were weighed, homogenized, and standardized to a protein concentration of 1 g/l. For immunoblot analyses, samples were loaded into a gel, separated electrophoretically, and transferred onto nitrocellulose membranes. The membranes were incubated with primary antisera (Table 1) at room temperature for 1 hour. After washing (1⫻ trisbuffered saline with 0.1% Tween-20, USB Corporation, Cleveland, Ohio), the membranes were incubated for 1 hour in peroxidase-conjugated secondary antibody (1:5,000 dilution, either Vector antimouse or Vector antirabbit, Vector Laboratories Inc., Burlingame, California; Table 1) and subjected to chemiluminescent activation (PerkinElmer, Inc., Boston, Massachusetts). The luminescent signal was
detected by exposure to x-ray film (Hyperfilm MP, Amersham Biosciences UK Ltd., Little Chalfont, United Kingdom). Appropriate recombinant standards were included in every immunoblot (Table 1). Relative levels of MMP-2 and -9 were determined using substrate-specific zymography as previously described.11,12 A commercially available purified MMP-2 and -9 standard (CC073, Chemicon, Temecula, California) was included in every zymogram, as were the internal reference control samples. Immunoblots and zymograms were digitized (Expression 1680, Epson, Nagano, Japan), and the banding patterns corresponding to the molecular weight of the positive signal from the recombinant standards were quantified (Gel Pro Analyzer, Media Cybernetics, Silver Spring, Maryland) and expressed in terms of an integrated optical density (IOD). The linearity of densitometric analysis was confirmed by ensuring that the largest and smallest IOD values for each analyte were recorded within a linear response range. Total collagen content was determined through a colorimetric reaction against picrosirius red.16,17 Briefly, 100 l of sample homogenates were dehydrated and stained with 0.1% picrosirius red in saturated picric acid (w/v) in a 96-well plate. The dye was solubilized and absorbance read at 540 nm. Readings were converted to protein units using a linear calibration curve generated from standards (Vitrogen 100, Angiotech Biomaterials, Palo Alto, California) and normalized to the weight of each myocardial sample. In a subset of patients with AF (n ⫽ 6), LA collagen I levels were determined by immunoblotting (anticollagen I, CL50151AP, Cedarlane Laboratories, Ltd., Hornby, Ontario, Canada). Statistical analysis: Echocardiographic parameters were compared between the non-AF and AF groups using the unpaired Student’s t test. For biochemical analyses, duplicate values were averaged and treated as a complete observation for that sample. Comparisons between the reference normal control group and the CHF group were performed by averaging IOD values for the normal samples and using the resultant value to normalize the IOD values for samples from the CHF group. Similarly, to examine the effects of AF on any of these biochemical determinants, IOD values for the non-AF samples from each zymogram or immuno-
534
The American Journal of Cardiology (www.AJConline.org) Table 2 Clinical parameters recorded before transplantation in patients without and with atrial fibrillation Variable
Figure 1. Representative immunoblots and zymograms for MMP-1, -2, -8, -9, -13, and -14, and TIMP-1, -2, -3, and TIMP-4. For the immunoblots, the areas of protein staining are depicted as dark bands on a lighter background. For the zymograms (MMP-2 and -9), the areas of substrate lysis caused by these proteases are depicted as lighter bands on a dark background. Myocardial samples from the right atrium (RA), right ventricle (RV), left atrium (LA), and left ventricle (LV) of patients with a documented history of AF (A) or no known AF history (N) were homogenized and electrophoretically separated. There were AF-associated chamberspecific differences in the levels of certain MMP and TIMP types. For example, MMP-14 levels were lower in the RA with AF, and there was a differential increase in left atrial MMP-9 levels with AF.
blot were averaged and used to normalize the IOD values for the samples from the AF group. Therefore, changes in the levels of collagen, the MMPs, and the TIMPs in each of the cardiac chambers were expressed as percentages of non-AF values of 100% and compared using a 2-sided Student’s t test. Correlations between MMP levels, TIMP levels, collagen content, and AF duration and the ejection fraction were examined using least-squares linear regression analyses. When appropriate, comparisons of categorical variables between the non-AF and AF groups were performed using Pearson’s chi-square analysis. All statistical analyses were performed using the BMDP software package (BMDP Statistical Software, Los Angeles, California). Results are presented as mean ⫾ SD. Values of p ⬍0.05 were considered to be statistically significant.
Non-AF (n ⫽ 20)
Age (yrs) LV size (cm) LV fractional shortening (%) LV ejection fraction (%) LA size (cm) Ischemic/nonischemic CHF Pretransplantation medications ACE inhibitors or ARB Aldosterone inhibitors Amiodarone Aspirin Beta-adrenergic receptor blockade Calcium channel blockade Coumadin Digoxin Diuretics Inotropes Nitrates Statins
AF (n ⫽ 23)
All (n ⫽ 46)
56 ⫾ 9 54 ⫾ 10 54 ⫾ 10 7.2 ⫾ 0.9 7.4 ⫾ 1.0 7.3 ⫾ 0.9 12.2 ⫾ 6.0 13.8 ⫾ 5.1 13.0 ⫾ 5.5 25.3 ⫾ 10.8 23.9 ⫾ 7.5 24.6 ⫾ 9.2 4.9 ⫾ 0.7 5.5 ⫾ 0.8* 5.2 ⫾ 0.8 8/12 11/12 19/24 18 (90%) 9 (45%) 3 (15%) 7 (35%) 10 (50%)
16 (70%) 7 (30%) 11 (48%)* 9 (39%) 5 (22%)
34 (79%) 16 (37%) 14 (33%) 16 (37%) 15 (35%)
1 (5%) 11 (55%) 14 (70%) 17 (85%) 2 (10%) 10 (50%) 13 (65%)
3 (13%) 19 (83%)* 18 (78%) 22 (96%) 7 (30%) 11 (48%) 10 (43%)
4 (9%) 30 (70%) 32 (74%) 39 (91%) 9 (21%) 21 (49%) 23 (53%)
Values are presented as mean ⫾ SD. * p ⬍0.05 versus non-AF. ACE ⫽ angiotensin-converting enzyme; ARB ⫽ angiotensin receptor blockade. Table 3 Matrix metalloproteinase levels, tissue inhibitor of metalloproteinase levels, and collagen content in right atrial, right ventricular, left atrial, and left ventricular myocardium of patients with atrial fibrillation and normalized to values obtained from non–atrial fibrillation hearts Variable
RA
RV
LA
LV
MMP-1 MMP-2 MMP-8 MMP-9 MMP-13 MMP-14 TIMP-1 TIMP-2 TIMP-3 TIMP-4 Collagen
139 ⫾ 49* 111 ⫾ 48 70 ⫾ 36* 130 ⫾ 99 107 ⫾ 45 56 ⫾ 26* 154 ⫾ 141 116 ⫾ 82 122 ⫾ 89 96 ⫾ 13 171 ⫾ 68*
118 ⫾ 45 96 ⫾ 35 121 ⫾ 96 112 ⫾ 96 105 ⫾ 51 122 ⫾ 50 161 ⫾ 125* 110 ⫾ 59 141 ⫾ 71* 100 ⫾ 16 79 ⫾ 41*
121 ⫾ 61 97 ⫾ 37 99 ⫾ 65 166 ⫾ 123* 103 ⫾ 41 112 ⫾ 126 122 ⫾ 94 102 ⫾ 90 202 ⫾ 194* 106 ⫾ 15 155 ⫾ 71*
123 ⫾ 68 106 ⫾ 33 120 ⫾ 48 113 ⫾ 112 112 ⫾ 39 108 ⫾ 33 100 ⫾ 47 138 ⫾ 79* 134 ⫾ 150* 99 ⫾ 17 90 ⫾ 13*
Values are presented as mean ⫾ SD. * p ⬍0.05 versus non-AF values of 100%.
Results Representative immunoblots for MMP-1, -8, -13, -14, and TIMP-1, -2, -3, and TIMP-4 and zymograms for MMP-2 and -9 are shown in Figure 1. To assess the effect of advanced CHF on these measures, the pooled results from patients with CHF (AF and non-AF cohorts) were compared with LV samples from subjects with no structural heart disease. Irrespective of AF status, LV myocardial levels of MMP-8 (256 ⫾ 69%), MMP-13 (309 ⫾ 125%), MMP-14 (295 ⫾ 80%), and TIMP-2 (210 ⫾ 28%) were increased and
TIMP-1 levels (30 ⫾ 31%) were decreased with CHF compared with reference normal samples (p ⬍0.05 vs the normal value of 100%). Clinical characteristics of the patients with CHF are listed in Table 2. The mean age, LV end-diastolic dimension, and LV ejection fraction were not different between the non-AF and AF groups. Moreover, the percentages of subjects with CHF with ischemic causes versus nonischemic causes did not differ between groups (Pearson’s chi-square 0.61). As expected, LA size was larger in patients with AF, as
Arrhythmias and Conduction Disturbances/MMPs and TIMPs With AF
Figure 2. There was a positive relation between LA collagen content and the duration of documented AF (y ⫽ 0.300 ⫹ 0.004x, r ⫽ 0.49, p ⫽ 0.023, SEE 0.148).
was the pretransplantation use of amiodarone and warfarin (Table 2). All other pretransplantation medications were similar between the 2 groups. For the AF group, this arrhythmia was present for 23 ⫾ 25 months (range 1 to 84). AF differentially affected selected MMPs and TIMPs. Cardiac chamber–specific differences in the levels of MMP-1, -9, and -14, and TIMP-3 were observed with AF compared with reference non-AF values (Table 3). Specifically, MMP-1 levels in the right atrium and MMP-9 levels in the left atrium were higher with AF. MMP-8 and -14 levels were lower in the right atrium with AF. TIMP-1 levels in the right ventricle and TIMP-3 levels in the right ventricle, left atrium, and left ventricle were higher with AF. Collagen content in the right atrium and the left atrium was greater with AF, whereas collagen content was less in the right ventricle and left ventricle in the AF group compared with non-AF values (Table 3). There was a linear correlation between AF duration and total LA collagen content (Figure 2). In a subset of patients with AF, collagen I levels were determined by immunoblotting. Immunopositive, or soluble, collagen I levels in the left atrium tended to be higher in the AF group than the non-AF group (214 ⫾ 101 vs 123 ⫾ 117 pixels). However, this difference did not achieve statistical significance (p ⫽ 0.12). Nevertheless, when expressed as a percentage of non-AF values, soluble collagen I content in the left atrium was greater with AF (175 ⫾ 83%, p ⫽ 0.011). In a subset of the AF samples for which soluble collagen content in the left atrium was determined, there was an inverse relation between soluble collagen I levels and AF duration (y ⫽ 296.15 ⫺ 2.84x, r ⫽ ⫺0.84, p ⫽ 0.036, SEE 75.20). Discussion AF is associated with a number of changes in the extracellular matrix (ECM) of the atrial myocardium, including
535
increases in intracellular glycogen accumulation, interstitial fibrosis, and a loss of intermyocyte connections.2,18,19 Because the MMPs and the TIMPs constitute a key system that regulates ECM composition and structure, the present study examined levels of collagen and MMP and TIMP species in samples obtained at the time of transplantation from all 4 cardiac chambers of the end-stage hearts of patients with and without persistent AF. There are several unique and important findings of the present study. First, there were chamber-specific differences in collagen content and levels of MMP and TIMP types with AF. For example, in patients with AF, collagen content was greater in the right atrium and left atrium and less in the right ventricle and left ventricle. AF was also associated with differential changes in MMP and TIMP levels in the atria and the ventricles. Second, in the left atrium, collagen content was linearly correlated with AF duration. Finally, there was an inverse correlation between AF duration and soluble collagen I levels. Taken together, these findings suggest that several post-translational events with respect to collagen metabolism occurred in the left atrium as a function of AF duration. It is important that the observed changes of myocardial levels of collagen, MMPs, and TIMPs occurred over and above those associated with other myocardial remodeling processes associated with LV dilation and systolic dysfunction. Although AF is primarily considered to be an electrical disease,20 it can result in changes in atrial structure, which make the atrial myocardium more susceptible to the maintenance of the arrhythmia (“AF begets AF”).21 Structural abnormalities documented to occur with AF include myocyte loss, the hypertrophy of existing myocytes, and altered composition of the ECM.2,4,5 The ECM is a dynamic structure that provides supportive scaffolding for myocytes and is critical for the maintenance of the structural integrity and the overall geometry of the cardiac chambers. In a canine model of AF induced by rapid ventricular pacing, Hanna et al5 demonstrated that LA collagen content increased earlier and achieved higher levels than in the LV myocardium. Changes in myocardial structure and/or geometry, collectively termed remodeling, can occur because of changes in the conformation and composition of the ECM.10 –12 Consistent with the findings of past studies,1– 4,18,19,22,23 an increase of the collagen content in the right atrium and left atrium was associated with the presence of AF in the present study. Moreover, the present study demonstrated an inverse correlation in the amount of soluble collagen I and the duration of AF, suggesting an increase of collagen cross linking as AF perpetuates. It is intriguing to speculate that this is 1 mechanism leading to the arrhythmia becoming permanent and irreversible. Using transgenic mice constitutively overexpressing transforming growth factor-1, Verheule et al24 recently demonstrated that increased atrial interstitial fibrosis was associated with greater AF inducibility, in the absence of abnormalities in action potentials. In a canine model of pacing-induced LV dysfunction, Hanna et al5 reported that there was a correlation between
536
The American Journal of Cardiology (www.AJConline.org)
P-wave duration and the amount of atrial fibrosis. Taken together, these findings raise the possibility that atrial fibrosis may, in and of itself, create a substrate conducive to the perpetuation of AF, and modification of this process may alter AF incidence and duration. The study by Hanna et al5 also provided a potential explanation for the differences with respect to atrial versus ventricular collagen content with AF. Specifically, in this past study, the progressive prolongation of the P-wave duration was associated with a rapid increase in angiotensin II content in the LA myocardium concomitant with an increase in the profibrotic cytokine transforming growth factor-.5 Given that recent clinical studies have reported a lower AF incidence in patients with CHF and hypertension who were prescribed angiotensin-converting enzyme inhibitors,25,26 it is likely that changes in atrial angiotensin II content contribute, at least in part, to atrial fibrosis with AF. Although a causative role of the angiotensin system in the development of fibrosis with AF remains to be elucidated, the findings of the present study taken together with those of past studies raise the intriguing possibility of targeting atrial fibrosis as a potential means to limit the incidence of AF. Previously, indirect evidence that changes in MMP or TIMP levels occur with AF was provided by measurements of plasma levels.27 In the present study, MMPs and TIMPs were measured directly from the myocardium of all 4 cardiac chambers, confirming the origin of these proteins. It is important that the findings of the present study demonstrated that AF was associated with alterations in the levels of certain key MMP and TIMP species in the atrial myocardium. Specifically, levels of MMP-9 and TIMP-3 were differentially increased in the left atrium and MMP-14 levels lower in the right atrium with AF. Two recent studies have examined a limited number of MMP and/or TIMP species in atrial samples from patients with AF.3,15 Xu et al3 reported that atrial MMP-2 levels were increased and TIMP-2 levels reduced with AF. Nakano et al15 demonstrated that although MMP-2 levels were unchanged with AF, RA MMP-9 messenger ribonucleic acid and protein levels were increased with AF. Consistent with the findings of these past studies, LA MMP-9 levels were observed to be differentially increased with AF in the present study. In addition, the findings of the present study provide evidence that regional changes of collagen content occurred as a function of changes in the proteolytic capacity in the myocardium of each chamber. For example, the lower MMP-14 levels in the right atrium with AF may have contributed to a reduction of collagen proteolysis. Similarly, the increase in LA collagen content may have been due to increased TIMP-3 levels with AF. These observations provide evidence that a likely mechanistic underpinning of interstitial atrial fibrosis with AF is changes in MMP and TIMP abundance and/or MMP and TIMP stoichiometry. The findings of the present study should be interpreted in light of certain limitations. First, all patients evaluated had a history of advanced CHF, which may have affected the
abundance of the MMPs and TIMPs. Comparisons were made with LV measurements from a small group of normal controls, demonstrating important effects of CHF on MMPs and TIMPs. Therefore, a direct comparison between chamber-specific MMP and TIMP levels obtained in the present study and those obtained from the other chambers of the nonfailing hearts could not be performed. Moreover, myocardial samples of the CHF hearts were obtained from available portions of the free wall of each of the cardiac chambers, precluding the analysis of within-chamber variability for any of the analytes. Nevertheless, these findings demonstrate that the presence of AF resulted in a differential regulation of MMPs and TIMPs in failing hearts and that these changes were over and above those that may have been caused by the underlying cardiac disease state. Second, MMPs and TIMPs were assayed using semiquantitative immunoblotting or zymographic techniques. Therefore, direct comparisons of changes in the levels of these proteins among the cardiac chambers could not be performed. Finally, myocardial but not plasma levels of the MMP and TIMP types were measured. Further studies are needed to determine the relation between myocardial and plasma levels of MMPs and TIMPs to evaluate clinical opportunities for potential diagnostic and prognostic tests of these findings for patients with AF. 1. Boldt A, Wetzel U, Lauschke J, Weigl J, Gummert J, Hindricks G, Kottkamp H, Dhein S. Fibrosis in left atrial tissue of patients with atrial fibrillation with and without underlying mitral valve disease. Heart 2004;90:400 – 405. 2. Kostin S, Klein G, Szalay Z, Hein S, Bauer EP, Schaper J. Structural correlate of atrial fibrillation in human patients. Cardiovasc Res 2002; 54:361–379. 3. Xu J, Cui G, Esmailian F, Plunkett M, Marelli D, Ardehali A, Odim J, Laks H, Sen L. Atrial extracellular matrix remodeling and the maintenance of atrial fibrillation. Circulation 2004;109:363–368. 4. Li D, Shinagawa K, Pang L, Leung TK, Cardin S, Wang Z, Nattel S. Effects of angiotensin-converting enzyme inhibition on the development of the atrial fibrillation substrate in dogs with ventricular tachypacing-induced congestive heart failure. Circulation 2001;104:2608 – 2614. 5. Hanna N, Cardin S, Leung TK, Nattel S. Differences in atrial versus ventricular remodeling in dogs with ventricular tachypacing-induced congestive heart failure. Cardiovasc Res 2004;63:236 –244. 6. Nagatomo Y, Carabello BA, Coker ML, McDermott PJ, Nemoto S, Hamawaki M, Spinale FG. Differential effects of pressure or volume overload on myocardial MMP levels and inhibitory control. Am J Physiol Heart Circ Physiol 2000;278:H151–H161. 7. Woodiwiss AJ, Tsotetsi OJ, Sprott S, Lancaster EJ, Mela T, Chung ES, Meyer TE, Norton GR. Reduction in myocardial collagen cross-linking parallels left ventricular dilatation in rat models of systolic chamber dysfunction. Circulation 2001;103:155–160. 8. Boixel C, Fontaine V, Rucker-Martin C, Milliez P, Louedec L, Michel JB, Jacob MP, Hatem SN. Fibrosis of the left atria during progression of heart failure is associated with increased matrix metalloproteinases in the rat. J Am Coll Cardiol 2003;42:336 –344. 9. Hoit BD, Takeishi Y, Cox MJ, Gabel M, Kirkpatrick D, Walsh RA, Tyagi SC. Remodeling of the left atrium in pacing-induced atrial cardiomyopathy. Mol Cell Biochem 2002;238:145–150. 10. Thomas CV, Coker ML, Zellner JL, Handy JR, Crumbley AJ III, Spinale FG. Increased matrix metalloproteinase activity and selective
Arrhythmias and Conduction Disturbances/MMPs and TIMPs With AF
11.
12.
13.
14.
15.
16. 17.
18.
upregulation in LV myocardium from patients with end-stage dilated cardiomyopathy. Circulation 1998;97:1708 –1715. Mukherjee R, Brinsa TA, Dowdy KB, Scott AA, Baskin JM, Deschamps AM, Lowry AS, Escobar GP, Lucas DG, Yarbrough WM, et al. Myocardial infarct expansion and matrix metalloproteinase inhibition. Circulation 2003;107:618 – 625. Spinale FG, Coker ML, Krombach SR, Mukherjee R, Hallak H, Houck WV, Clair MJ, Kribbs SB, Johnson LL, Peterson JT, et al. Matrix metalloproteinase inhibition during the development of congestive heart failure: effects on left ventricular dimensions and function. Circ Res 1999;85:364 –376. Wilson EM, Moainie SL, Baskin JM, Lowry AS, Deschamps AM, Mukherjee R, Guy TS, John-Sutton MG, Gorman JH III, Edmunds LH Jr, et al. Region- and type-specific induction of matrix metalloproteinases in post-myocardial infarction remodeling. Circulation 2003;107: 2857–2863. Spinale FG, Coker ML, Heung LJ, Bond BR, Gunasinghe HR, Etoh T, Goldberg AT, Zellner JL, Crumbley AJ. A matrix metalloproteinase induction/activation system exists in the human left ventricular myocardium and is upregulated in heart failure. Circulation 2000;102: 1944 –1949. Nakano Y, Niida S, Dote K, Takenaka S, Hirao H, Miura F, Ishida M, Shingu T, Sueda T, Yoshizumi M, et al. Matrix metalloproteinase-9 contributes to human atrial remodeling during atrial fibrillation. J Am Coll Cardiol 2004;43:818 – 825. Marotta M, Martino G. Sensitive spectrophotometric method for the quantitative estimation of collagen. Anal Biochem 1985;150:86 –90. Tullberg-Reinert H, Jundt G. In situ measurement of collagen synthesis by human bone cells with a sirius red-based colorimetric microassay: effects of transforming growth factor beta2 and ascorbic acid 2-phosphate. Histochem Cell Biol 1999;112:271–276. Frustaci A, Chimenti C, Bellocci F, Morgante E, Russo MA, Maseri A. Histological substrate of atrial biopsies in patients with lone atrial fibrillation. Circulation 1997;96:1180 –1184.
537
19. Ausma J, van der Velden HM, Lenders MH, van Ankeren EP, Jongsma HJ, Ramaekers FC, Borgers M, Allessie MA. Reverse structural and gap-junctional remodeling after prolonged atrial fibrillation in the goat. Circulation 2003;107:2051–2058. 20. Moe GK, Abildskov JA. Atrial fibrillation as a self-sustaining arrhythmia independent of focal discharge. Am Heart J 1959;58:59 –70. 21. Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA. Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation 1995;92:1954 –1968. 22. Aime-Sempe C, Folliguet T, Rucker-Martin C, Krajewska M, Krajewski S, Heimburger M, Aubier M, Mercadier JJ, Reed JC, Hatem SN. Myocardial cell death in fibrillating and dilated human right atria. J Am Coll Cardiol 1999;34:1577–1586. 23. Schotten U, Ausma J, Stellbrink C, Sabatschus I, Vogel M, Frechen D, Schoendube F, Hanrath P, Allessie MA. Cellular mechanisms of depressed atrial contractility in patients with chronic atrial fibrillation. Circulation 2001;103:691– 698. 24. Verheule S, Sato T, Everett T, Engle SK, Otten D, Rubart-von der LM, Nakajima HO, Nakajima H, Field LJ, Olgin JE. Increased vulnerability to atrial fibrillation in transgenic mice with selective atrial fibrosis caused by overexpression of TGF-1. Circ Res 2004;94:1458 –1465. 25. L’Allier PL, Ducharme A, Keller PF, Yu H, Guertin MC, Tardif JC. Angiotensin-converting enzyme inhibition in hypertensive patients is associated with a reduction in the occurrence of atrial fibrillation. J Am Coll Cardiol 2004;44:159 –164. 26. Vermes E, Tardif JC, Bourassa MG, Racine N, Levesque S, White M, Guerra PG, Ducharme A. Enalapril decreases the incidence of atrial fibrillation in patients with left ventricular dysfunction: insight from the Studies of Left Ventricular Dysfunction (SOLVD) trials. Circulation 2003;107:2926 –2931. 27. Marin F, Roldan V, Climent V, Garcia A, Marco P, Lip GYH. Is thrombogenesis in atrial fibrillation related to matrix metalloproteinase-1 and its inhibitor, TIMP-1? Stroke 2003;34:1181–1186.