INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

MYOCARDIAL FIBRILLAR COLLAGEN serves as a supportive scaffolding to the myocardium, thereby maintaining ventricular shape and size and governing tissue stiffness. Accordingly, an activation of matrix metalloproteinases (MMP), which in turn would cause a disruption of myocardial collagen fibers, would be expected to precede the ventricular remodeling associated with a chronic biventricular volume overload (14). Indeed, morphological evidence of myocardial MMP activation and collagen fiber disruption/degradation has been found in the dilated, thin-walled ventricle in human heart failure patients (11, 37), in the cardiomyopathic Syrian hamster (8, 13), and in the chronically rapid-paced animal (33, 34). However, a biochemical, histological, or morphological assessment of interstitial collagen and MMP activity using tissue obtained from explanted human hearts, from endomyocardial biopsies, from hearts at autopsy, or from diseased hearts subjected to several months of an experimentally induced increase in ventricular preload provides a description of the collagen network "after the fact" with little insight into the remodeling events that preceded the development of heart failure. The infrarenal abdominal aortocaval (AV) fistula model of sustained ventricular volume overload in rats has been shown to induce progressive ventricular dilatation, increased myocardial compliance, and inappropriate hypertrophy (2). After an extended period of compensated volume overload lasting ~8 wk, there is an accelerated transition to overt congestive heart failure associated with a marked increase in left ventricular (LV) dilatation and compliance (3). Thus it was the purpose of this study to examine the temporal response of MMP activity and collagen matrix degradation in this AV fistula model over the first 8 wk of chronic ventricular volume overload. In addition, because mast cells are known to contain the metalloproteinase-activating enzymes chymase and tryptase, the possible regulatory role of cardiac mast cells in determining MMP activity was investigated.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

All experiments were performed using adult male Sprague-Dawley (Hsd:SD) rats housed under standard environmental conditions and maintained on commercial rat chow and tap water ad libitum. All studies conformed with the principles of the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and the protocol was approved by our institution's Animal Care and Use Committee. Anesthesia for surgical procedures and subsequent euthanasia at the experimental end point was effected by pentobarbital sodium (50 mg/kg) injected into the peritoneal cavity. Postoperative analgesia was provided by buprenorphine HCl (0.025 mg/kg sq) administered to the rats at the time of surgery. After all surgical procedures, the rats were alert and resumed normal activity within 24 h.

Surgical preparation. Infrarenal abdominal AV fistula was created in rats as previously described (2). Briefly, a ventral abdominal laparotomy was performed to expose the aorta and caudal vena cava ~1.5 cm below the renal arteries. Both vessels were then occluded, and an 18-gauge needle was inserted into the aorta and advanced through the medial wall into the vena cava to create the fistula. The needle was then withdrawn, and the aortic puncture site was sealed with cyanoacrylate. Creation of a successful fistula was evident by the pulsatile flow of oxygenated blood into the vena cava. The musculature and skin incisions were closed by standard techniques with absorbable sutures and autoclips, respectively.

Experimental design. AV fistulas were created in a total of 77 rats, which were then randomly divided into nine groups corresponding to 0.5 (n = 6), 1 (n = 6), 2 (n = 6), 3 (n = 9), 5 (n = 6), 14 (n = 9), 21 (n = 6), 35 (n = 10), and 56 (n = 19) days after creation of the fistula. We previously identified increased variability in ventricular function and remodeling at 5 wk and beyond in this animal model (2, 3); therefore, the number of animals in the groups for these time points was increased. In addition, the ability of the mast cell-stabilizing compound cromolyn sodium to prevent the increases in mast cell number and MMP activation during the first 3 days after creation of an AV fistula was assessed in three groups at 1 (n = 13), 2 (n = 12), and 3 (n = 11) days postfistula. Five days before the creation of the fistula, cromolyn sodium (24 mg · kg¹ · day¹) administration via an intraperitoneal osmotic minipump (model 2ML2, Durect; Cupertino, CA) was initiated and continued for the duration of the study. Age-matched control groups for each study period consisted of sham-operated rats (n = 34). At the end of the study periods, the rats were anesthetized, fistula patency was visually confirmed, and the hearts were rapidly removed and rinsed in ice-cold saline. The atria and great vessels were then dissected free, the LV (plus septum) and right ventricle (RV) were separated and weighed (except for groups from the first week postfistula), and a complete cross section taken from the midportion of each ventricle was placed in buffered formalin for fixation. The remainder of the cardiac tissue was frozen at 80°C for subsequent zymographic determination of MMP activity. Lung wet weight was obtained (except for groups from the first week postfistula) after the esophagus and trachea were trimmed away, and the pleural surface was blotted dry. Heart and lung weights were not recorded for the groups during the first week postfistula because significant increases for these values were not seen before 2 wk in a previous study (2).

Microscopy. The formalin-fixed tissue was processed for routine histopathology, and 5-µm-thick paraffin-embedded sections were stained with hematoxylin and eosin for evaluating myocardial morphology and determining the presence of cellular infiltrates or necrosis, toluidine blue for determining mast cell number, or sirius red F3BA (PSR) for evaluating myocardial interstitial collagen (36). With the use of an overall magnification of ×200, the number of mast cells in the entire ventricular cross sections was determined in a blinded fashion. Mast cell density was then determined by dividing the total number of mast cells per ventricular cross section by the area determined from scanning the respective cross section. Similarly, LV interstitial collagen volume fraction (CVF) was determined from the PSR-stained sections using a Quantimet 520 Image Analysis System (Leica; Deerfield, IL) and established methods (41). Perivascular collagen was excluded from the analysis of CVF, and the Quantimet operator was blinded as to the source of the tissue.

Mast cell-mediated MMP activation. To demonstrate a direct linkage between cardiac mast cells and subsequent MMP activation, in situ degranulation of cardiac mast cells was induced with the mast cell secretagogue compound 48/80 (Sigma; St. Louis, MO). Briefly, using an in vitro, blood-perfused isolated heart preparation as previously described (2, 3), the aorta was cannulated for retrograde perfusion, and the heart was extirpated and attached to a modified Langendorff preparation. The apparatus consisted of a pressurized perfusion reservoir and a collection reservoir connected in circuit with a support rat. The heart was perfused with oxygenated blood obtained from the carotid artery of the support rat. The coronary venous effluent, with the exception of the portion containing compound 48/80, was collected and returned to the support rat through a jugular vein catheter to filter and oxygenate the blood supply to the isolated heart. We then introduced a bolus of compound 48/80 (7.2 mg/ml) or diluent into the perfusion reservoir for delivery to the heart. Perfusion of the heart was maintained for an additional 30 min, after which the heart was processed for histology and zymography as described above.

MMP activity. MMP activity in cardiac tissue extracts was determined by gelatin zymography performed by a standard procedure using a SDS-PAGE matrix containing gelatin (1 mg/ml) (40). Cardiac tissue extracts were prepared by washing 40-50 mg of frozen LV in cold saline. The tissue was then minced into 1-mm fragments, which were agitated for 48 h at 4°C in 1 ml of ice-cold extraction buffer containing cacodylic acid (10 mmol/l), NaCl (150 mmol/l), ZnCl₂ (1 µmol/l), CaCl₂ (20 mmol/l), NaN₃ (3.0 mmol/l), and 0.01% Triton X-100. Upon completion of the extraction incubation, the supernatant was decanted and added to 200 µl of 0.1 mol/l Tris buffer, and the samples were centrifuged (4°C, 15 min, 13,000 g). The extracted samples were then aliquoted and stored at 80°C. During subsequent analysis, each sample was mixed with Laemmli SDS sample buffer in the absence of a reducing agent and electrophoresed on a 10% polyacrylamide gel at 20°C. After electrophoresis, the gels were washed four times for 20 min in 2.5% Triton X-100 to remove the SDS and permit enzyme renaturation. The gel was then placed in substrate buffer [40 mM Tris · HCl, 200 mM NaCl, 5 mM CaCl₂, 0.02% (wt/vol) Brij-35, and 0.02% sodium azide; pH 7.4] and incubated on a shaker for 18 h at 37°C. The gels were then stained with 0.1% Coomassie brilliant blue and destained in distilled deionized H₂O, and the activity of the MMP bands was quantified by densitometry (ImageQuant, Molecular Dynamics). All of the zymograms had two lytic bands corresponding to standards for gelatinase A (MMP-2) proenzyme (68 kDa) and activated (62 kDa) forms (Chemicon International; Temecula, CA). Less consistent was the appearance of a lytic band at 92 kDa consistent with gelatinase B (MMP-9). To combine results from different gels, a single extract from the same control heart was used as a standard on all gels. The activity of the lytic bands in the other lanes of a gel were expressed as a percentage of the activity of this standard. Once normalized in this fashion, the percent activities from hearts belonging to the same group (i.e., control, 1 day postfistula, etc.) were averaged.

Statistics. Statistical analyses were performed using SPSS 10.0 software (SPSS; Chicago, IL). Grouped data are expressed as means ± SD. Grouped data comparisons were made by one-way ANOVA. When a significant F-test (P  0.05) was obtained, intergroup comparisons were analyzed using Fisher's protected least-significant-difference post hoc testing. Statistical significance was taken to be P  0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

A patent AV fistula was confirmed to be present in all rats from the AV fistula groups.

MMP activity. The temporal response of LV MMP activity measured after creation of the AV fistula is presented in Fig. 1. Marked increases in MMP activity occurred rapidly, being increased above control at 12 h postfistula (44% greater than control, P < 0.05) and remaining significantly elevated for the first 5 days postfistula. It then returned toward normal until days 35 and 56, when a moderate but still significant increase in MMP activity was observed (24% and 22% greater than control, respectively, P < 0.05).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   The temporal response of left ventricular gelatinase A [matrix metalloproteinase (MMP)-2] activity after creation of an infrarenal abdominal aortocaval fistula. Bars represent means ± SD. * P < 0.05 vs. control;  P < 0.01 vs. control.

Mast cell profile. The temporal responses in the number of myocardial mast cells for the LV and RV are shown in Table 1, whereas the percent changes in cardiac mast cell density relative to control are depicted in Fig. 2. As might be expected in this model of biventricular volume overload, the temporal response in RV mast cells postfistula was similar to that seen in the LV. The average number of mast cells per ventricular cross section increased rapidly after opening the AV fistula, peaking within 24 h and remaining elevated for the first 5 days postfistula before returning to normal by day 14. Furthermore, the increase in the number of LV mast cells closely paralleled the changes in MMP activity seen after opening the AV fistula. This close relationship between total mast cell number and MMP activity was maintained during the first 2 wk; however, the subsequent significant increases in mast cell number occurring at 35 and 56 days were associated with a relatively modest increase in MMP activity. However, it should be noted that by this time, substantial myocardial hypertrophy had developed, and when this is taken into account by instead comparing MMP activity to mast cell density (i.e., the number of mast cells per area of tissue, Fig. 3), the 35 and 56 day points cluster relatively near that of the control.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   The temporal response of myocardial mast cell density after creation of an infrarenal abdominal aortocaval fistula. Bars represent mean percent differences from control in the number of mast cells/mm2 ± SD for the left ventricle (top) or right ventricle (bottom). * P < 0.05 vs. control;  P < 0.01 vs. control.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Relationship between left ventricular mast cell density postfistula and the percent increase in left ventricular MMP-2 activity relative to controls. Data points represent group mean values ± SE.

Effects of inhibiting mast cell degranulation. Treatment with the mast cell-stabilizing compound cromolyn sodium was able to prevent the subsequent increase in both the number of LV mast cells and MMP activity after 1-3 days of volume overload. As shown in Figs. 4 and 5, MMP activity and mast cell number in the cromolyn-treated fistula rats were similar to the sham-operated control group values.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   The ability of mast cell stabilization after treatment with cromolyn sodium to prevent the increase in MMP-2 activity at 1, 2, and 3 days after creation of an infrarenal abdominal aortocaval fistula. Solid bars, untreated aortocaval fistula; hatched bars, cromolyn-treated aortocaval fistula. Values are means ± SD.  P < 0.01 vs. control.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   The ability of mast cell stabilization after treatment with cromolyn sodium to prevent the increase in mast cell number at 1, 2, and 3 days after creation of an infrarenal abdominal aortocaval fistula. Solid bars, untreated aortocaval fistula; hatched bars, cromolyn-treated aortocaval fistula. Values are means ± SD. * P < 0.05 vs. control;  P < 0.01 vs. control.

Myocardial interstitial collagen. The temporal response in LV CVF is presented in Fig. 6. A notable decrease in CVF was apparent by days 3 (48%) and 5 (61%). CVF then rebounded and actually exceeded control on days 14 (55% increase) and 35 (43% increase), whereas it was statistically similar to the control value at days 1, 2, 21, and 56.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   The temporal response of left ventricular collagen volume fraction after creation of an infrarenal abdominal aortocaval fistula. Bars represent means ± SD. * P < 0.05 vs. control;  P < 0.01 vs. control.

Mast cell-mediated MMP activation. The results of degranulating cardiac mast cells with compound 48/80 on MMP activation are presented in Fig. 7. As can be clearly seen, mast cell degranulation produced a substantial reduction in latent MMP-2 and a corresponding marked increase in active MMP-2. Microscopic examination of the compound 48/80-treated hearts revealed extensive to complete degranulation in the majority of cardiac mast cells.

View larger version (45K): [in this window] [in a new window]   Fig. 7.   Zymogram depicting the ability of the mast cell secretagogue compound 48/80 to induce MMP-2 activation. Mast cell degranulation produced a substantial reduction in latent MMP-2 and a corresponding marked increase in active MMP-2.

Morphometric data. Average LV, RV, and lung weights for the control and volume-overloaded groups are reported in Table 1. The progressive myocardial hypertrophy and development of pulmonary edema induced over the temporal period postfistula is consistent with previously reported findings in this model (2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The objectives of the present study were as follows: 1) to determine the temporal response of LV MMP activity and collagen matrix degradation over the first 8 wk of chronic ventricular volume overload and 2) to ascertain the possible role of cardiac mast cells in regulation of MMP activity. The principal findings of this study were as follows. Marked increases in myocardial MMP activity occurred within hours and remained significantly elevated for the first 5 days after creation of the AV fistula before returning toward normal. Concomitant with this initial response in MMP activity, significant fibrillar collagen degradation had occurred within 3 days postfistula. However, beyond this initial phase of LV remodeling, the CVF subsequently rebounded to normal or above normal values for the remainder of the study period. There was also a rapid increase in the number of cardiac mast cells, and the significant increases in MMP activity closely paralleled the changes in LV mast cell density throughout the study. This close association between cardiac mast cell density and MMP activity throughout the 56 days of ventricular volume overload is further strengthened by the observation that the initial increase in both the number of mast cells and MMP activity postfistula were preventable in hearts treated with the mast cell membrane-stabilizing drug cromolyn. In addition, this is the first study to demonstrate that in vivo activation of gelatinase A can be mediated by cardiac mast cell degranulation. From these results, it is concluded that mast cells play a major role in regulating myocardial MMP activity and the initial extracellular matrix remodeling induced by a chronic volume overload.

The view that cardiac mast cells may be involved in the pathological course of myocardial disease is a recent one. Initial observations were focused on the relationship of mast cell number to the vasculature (1, 24, 31), and most of the subsequent studies were concentrated on the role of mast cells in the development of atherosclerosis (5, 16, 17, 20). More recently, increased numbers of cardiac mast cells have been reported in human hearts with end-stage cardiomyopathy (20, 27, 28) and in animal models of hypertension (25), mitral regurgitation (7), and myocardial infarction (9). However, none of these studies have identified or really even speculated on how this influx of mast cells in the heart might be associated with the pathogenesis of these various conditions. Whereas the source of this mast cell influx is not known, the observations herein clearly demonstrate that infiltration of mature mast cells can occur within hours, not days, as evidenced by more than a 40% increase in the number of myocardial mast cells at 12 h postfistula. The relatively short period required for this significant increase to occur indicates that these additional mast cells are recruited either from the circulation or from a resident population of mast cell progenitors in the heart (10). Although the role of mast cells in extracellular matrix remodeling has not been clearly defined, they appear to modulate extracellular matrix synthesis and degradation through paracrine release of cytokines, which regulate fibroblast phenotype and MMP activity (6, 18, 30, 39). For example, mast cells contain substantial stores of tumor necrosis factor-, a cytokine capable of inducing MMP synthesis. The few studies that have characterized human cardiac mast cells have described them as containing tryptase, chymase, carboxypeptidase, and cathepsin G (26, 29). In addition, mast cells may themselves contain MMPs, as evidenced by the demonstration of stromelysin in normal murine skin and lung mast cells and collagenase in peritoneal mast cells (4, 38). Although mast cell enzymes have been implicated as being involved in the MMP activation cascade by several in vitro studies (19, 22, 35), this study is the first to demonstrate that mast cells are capable of mediating MMP activation in the heart. That significant MMP activation can be induced by cardiac mast cell degranulation represents an important new concept. These findings are also novel in that they identify mast cells as potential regulators of a mechanism thought by many to mediate the adverse cardiac remodeling associated with heart failure. The direct relationship between MMP activity and mast cell density, together with the fact that we were able to prevent the acute increase in MMP activity and mast cell number with the mast cell-stabilizing compound cromolyn sodium, provide additional evidence that cardiac mast cells are in vivo regulators of MMP activity. More than likely, inhibiting mast cell degranulation prevented the release of MMP activating substances such as chymase, cathepsin G, tryptase, and stromelysin. Cromolyn also apparently prevented the release of substances responsible for the chemoattraction and subsequent maturation of mast cell progenitors in the heart.

A potential shortcoming of this study is that collagenase activity cannot be determined by this technique because collagenase is a relatively poor degrader of gelatin. However, a collagen-substrate zymogram is not possible due to the insolubility of native collagen. Furthermore, based on the work of Nagase (22) and others, one would expect the activities of the various MMPs in this degradation cascade to be tightly coupled. In fact, a comparison of zymography-determined gelatinase A and B activity with collagenase activity determined by degradation of tritium-labeled telopeptide-free collagen was reported by Gunja-Smith et al. (11) to be closely correlated in both nondiseased and cardiomyopathic human hearts having little and markedly elevated activities, respectively. Thus the increase in MMP-2 activity observed in this study was in all likelihood also reflective of increased collagenase activity. This, together with the fact that fibrillar collagen was significantly reduced within 3 days postfistula represents strong evidence that myocardial collagenase activity (which for the rat would be MMP-13) was similarly elevated.

The temporal responses in LV size and stiffness over the first 8 wk of AV fistula were recently reported (2). In the initial 3 wk postfistula, there is an essentially parallel shift of the LV end-diastolic pressure-volume relationship to the right, indicating primarily LV dilatation and a relatively small change in LV compliance. The initial decrease in extracellular collagen reported herein is most likely responsible for the ventricular dilatation observed after the first week of volume overload. Although the subsequent response in CVF and its relationship to the marked LV enlargement and increasing compliance occurring postfistula are difficult to interpret, we recently found both MMP activity and interstitial CVF to be significantly increased in decompensated hearts from rats with symptomatic congestive heart failure (3). Clearly, the up and down fluctuations in CVF and MMP activity beyond the first week reflect a dynamic imbalance between collagen synthesis and degradation as LV remodeling progresses. Moreover, continued remodeling of the extracellular matrix would be expected to have an impact on the size and passive mechanical behavior of the ventricle (15). The composition of new collagen deposition would be expected to contain a greater proportion of the more compliant type III collagen. This, together with less extensive collagen cross-linking, could account for much of the progressive increase in compliance in this model of chronic volume overload. This hypothesis is supported by the findings of Gunja-Smith et al. (11) related to remodeling of myocardial collagen in human idiopathic dilated cardiomyopathy. They concluded that despite marked elevations in total myocardial collagen, collagen critical to the mechanical stability of the heart is degraded by increased MMP activity and is replaced by collagen that is poorly cross-linked, resulting in a more compliant, dilated ventricle. Additional factors, such as a decrease in integrin-mediated cardiomyocyte adhesion to the extracellular matrix, could also lead to ventricular dilatation and increased compliance.

To date, no other investigators have assessed MMP activity and the extracellular matrix in the AV fistula model at a point earlier than 1 wk. With the exception of the report by Ruzicka et al. (32), all other studies (including ours) that have investigated cardiac extracellular matrix remodeling secondary to 1 mo or longer of AV fistula found myocardial collagen concentration to be normal or elevated (3, 12, 21, 23, 41). Interestingly, although Namba et al. (23) found no significant differences in myocardial collagen content postfistula, they did report a persistent upregulation in collagen gene expression; however, because MMP activity was not increased at 28 days postfistula, they concluded that collagen degradation was unlikely to play a significant role in this disparity between collagen synthesis and accumulation. The nuances of ventricular remodeling are further mystified by the findings of Ruzicka and co-workers (32), who reported total collagen to be normal at 1 wk and significantly decreased at 4 and 10 wk postfistula. These findings are in contrast to a recent study in which we found that both MMP activity and interstitial collagen content were significantly increased in decompensated hearts from rats with symptomatic congestive heart failure, whereas the collagen content and MMP activity in comparable compensated hearts were similar to controls. While the reason for these discrepancies in the literature is not clear, it seems likely from the temporal changes observed in our study that continual shifts in the balance of compensatory ventricular remodeling may be responsible for wide fluctuations in the parameters of interest. Additional inconsistencies may be related in part to differences in animal strain, technique, or the degree of volume overload produced.

In summary, the results of this study indicate that cardiac mast cells may play a major role in regulation of the myocardial extracellular matrix and ventricular remodeling that occurs in response to a chronic ventricular volume overload. This is also the first study to identify a functional role for the mast cell in myocardial remodeling, demonstrating that rapid MMP activation mediated by cardiac mast cell degranulation produces the adverse ventricular remodeling contributing to the development of congestive heart failure. Furthermore, the fact that the initial increase in MMP activity and mast cell number in response to volume overload could be prevented with a mast cell membrane-stabilizing drug suggests a potential role for such drugs in the prevention of ventricular dilatation associated with heart failure. However, long-term studies will be required to determine whether these initial beneficial effects result in the prevention of adverse ventricular remodeling and ultimately translate into a reduction in the incidence of heart failure and its associated mortality.