Long-Term Localized High-Frequency Electric Stimulation Within the Myocardial Infarct
Effects on Matrix Metalloproteinases and Regional Remodeling
Background— Disruption of the balance between matrix metalloproteinases (MMP) and MMP inhibitors (TIMPs) within a myocardial infarct (MI) contributes to left ventricular wall thinning and changes in regional stiffness at the MI region. This study tested the hypothesis that a targeted regional approach through localized high-frequency stimulation (LHFS) using low-amplitude electric pulses instituted within a formed MI scar would alter MMP/TIMP levels and prevent MI thinning.
Methods and Results— At 3 weeks after MI, pigs were randomized for LHFS (n=7; 240 bpm, 0.8 V, 0.05-ms pulses) or were left unstimulated (UNSTIM; n=10). At 4 weeks after MI, left ventricular wall thickness (echocardiography; 0.89±0.07 versus 0.67±0.08 cm; P<0.05) and regional stiffness (piezoelectric crystals; 14.70±2.08 versus 9.11±1.24; P<0.05) were higher with LHFS than in UNSTIM. In vivo interstitial MMP activity (fluorescent substrate cleavage; 943±59 versus 1210±72 U; P<0.05) in the MI region was lower with LHFS than in UNSTIM. In the MI region, MMP-2 levels were lower and TIMP-1 and collagen levels were higher with LHFS than in UNSTIM (all P<0.05). Transforming growth factor-β receptor 1 and phosphorylated SMAD-2/3 levels within the MI region were higher with LHFS than in UNSTIM. Electric stimulation (4 Hz) of isolated fibroblasts resulted in reduced MMP-2 and MT1-MMP levels but increased TIMP-1 levels compared with unstimulated fibroblasts.
Conclusions— These unique findings demonstrate that LHFS of the MI region altered left ventricular wall thickness and material properties, likely as a result of reduced regional MMP activity. Thus, LHFS may provide a novel means to favorably modify left ventricular remodeling after MI.
Received March 31, 2009; accepted May 10, 2010.
Events after a myocardial infarction (MI) include left ventricular (LV) remodeling in terms of progressive chamber dilation and heterogeneous changes in the cellular and extracellular constituents of the LV myocardium.1–3 In addition, the formation of the fibrotic MI scar alters myocardial material properties such as regional myocardial stiffness.4–6 Moreover, these changes in myocardial structure at the MI region are speculated to result in a progressive thinning of the MI region, which is called infarct expansion.4 Increased matrix metalloproteinase (MMP) abundance and discordant alterations in the balance between MMPs and endogenous tissue inhibitors of the metalloproteinases (TIMPs) have been implicated in structural sequelae to MI.2,4,5,7 Although systemic pharmacological approaches can attenuate global LV dilation after MI, a targeted modulation that results in reversing the adverse remodeling of a “mature” MI scar in terms of normalizing the imbalance between MMP/TIMP levels and the material characteristics remains problematic.
Clinical Perspective on p 32
In vivo and in vitro studies have demonstrated that electric stimulation can modulate a number of factors relevant to tissue remodeling.8–15 For example, electric stimulation of dermal wounds accelerates the wound healing response in terms of collagen deposition and wound contraction (see elsewhere11 for a review). Electrically stimulating cartilage explanted from patients with osteoarthritis resulted in increased collagen deposition and reduced messenger RNA expression of certain MMP types.12 In fibroblast cultures, cellular viability, migration, and rate of protein synthesis, including that of matrix proteins, have been shown to be increased with electric stimulation.11 However, whether and to what degree in vivo electric stimulation of the MI region, which contains a high population of fibroblasts, would alter remodeling within the MI region remained unknown. Accordingly, this study tested the hypothesis that localized high-frequency stimulation (LHFS) instituted within a formed MI scar with low-amplitude electric pulses would reduce MMP activity and prevent progressive MI thinning.
All animals were treated and cared for in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (National Research Council; Washington, DC; 1996), and the protocol was approved by the Institutional Animal Care and Use Committee. Permanent coronary ligation was performed in mature pigs (Yorkshire; n=17; weight, 25 kg; Hambone Farms, Orangeburg, SC) as previously described.4 Briefly, the LV was accessed through a left-sided thoracotomy, and a pericardiectomy was performed. MI was induced by direct ligation of the first 2 obtuse marginal arteries (OM1, OM2) at the origin from the circumflex coronary artery (4.0 proline).4,5
Determination of LHFS Parameters
The main concern in determining the LHFS stimulation parameters pertained to the safety of being able to confine electric activation to within the MI region but not cause escape of the triggered impulses that could result in ventricular tachycardia. Accordingly, the pacemaker parameters used in the present study were determined in preliminary studies performed in 3 post-MI pigs. At the terminal study on post-MI day 28, a short-term pacing protocol was instituted using a pacemaker lead sutured to the center of the MI region on the epicardial surface and connected to an external pacemaker (Medtronic, Minneapolis, Minn). The pacemaker was activated using pulses with amplitudes ranging from 0.5 to 1.5 V (in increments of 0.1 V) and pulse durations of 0.01, 0.05, and 0.10 ms. The resultant propagation of the introduced impulses was mapped by measuring pulse amplitude (PONEMAH, St. Paul, Minn; sensitivity, 5 mV) on the epicardial surface with a template in the polar coordinate system (radial intervals in increments of 0.5 cm up to 2.5 cm and angle increments of 45°). Pulses with an amplitude of 0.8 V and duration of 0.05 ms represented the largest combination of amplitude/duration parameters that was completely confined to the MI region (at distances of 2.0 cm from the stimulating electrode; Figure 1A) in all 3 pigs. Therefore, these pulse parameters were used for long-term LHFS for the MI pigs in the present study. The same pacing parameters resulted in the capture of the ventricular rhythm in 2 non-MI pigs. Accordingly, referent controls used in this study were non-MI pigs not subjected to long-term LHFS.
Long-Term LHFS Model
MI was induced as described above, and a shielded pacemaker lead was sutured onto the epicardium between OM1 and OM2, 2 cm below the circumflex artery (n=17). A pacemaker (EnPulse, Medtronic) was buried in a subcutaneous pocket and connected to the pacemaker lead, which was tunneled through a purse-string incision on the thoracic wall. Pacemaker capture of the LV was confirmed by transiently (<30 seconds) activating the pacemaker at a rate 20% higher than the intrinsic heart rate of the pig.
LV echocardiography measurements were performed (S4 probe; 3.5 MHz; spatial resolution, 800 μm; Agilent Sonos 5500, Agilent Technologies Inc, Santa Clara, Calif) to determine LV wall thickness and volumes with the method-of-disks variant of the Simpson algorithm.16 Interobserver and intraobserver coefficients of variance while using this system were 9% and 4%, respectively. In these studies, all echocardiographic measurements were performed by a single observer (W.T.R.) who was blinded to the group assignment of the pigs. After echocardiography on post-MI day 21, pacemakers in the pigs randomized to the LHFS group (n=7) were activated at 240 bpm with low-amplitude and short-duration pulses (VOO mode; 0.8 V, 0.05 ms). ECG recordings were obtained to confirm that pacemaker activation did not cause ventricular tachycardia through capture of the ventricular rate (Figure 1B). Pacemakers in the other MI pigs (n=10) were left deactivated, and these pigs made up the unstimulated (UNSTIM) group.
Myocardial Function and Microdialysis Measurements
At 28 days after MI, pacemaker stimulation of the LV in the LHFS group was reconfirmed in the ECG tracing, after which the pacemakers were deactivated. Repeat echocardiographic measurements were obtained in all pigs. The pigs were anesthetized and instrumented for hemodynamic measurements of arterial pressures, pulmonary artery pressures, and cardiac output. A sternotomy was performed, and a vascular ligature was placed around the inferior vena cava to perform transient caval occlusion. A calibrated microtipped transducer (7.5F, Millar Instruments Inc, Houston, Tex) was placed in the LV through a small apical stab wound. Piezoelectric crystals (2 mm; Sonometrics Corp, London, Ontario, Canada) were positioned in the central portion of the MI region to record regional LV dimensions and wall thickness at a sampling frequency of 1000 Hz (Pentium-Sonolab, Sonometrics).
Steady-state hemodynamics measurements included systemic and pulmonary artery pressures, cardiac output, and LV pressures. After steady-state measurements, LV preload was altered by sequential occlusion and release of the inferior vena cava, and isochronal measurements of LV pressure and dimensions were recorded. From the digitized pressure-dimension data, regional myocardial stiffness of the MI region was computed.5
For interstitial MMP activity measurements, microdialysis probes with a molecular weight cutoff of 20 kDa and an outer diameter of 0.5 mm were placed in the remote and MI regions (2 probes per pig). A previously validated fluorogenic substrate specific for MMP-1, -2, -3, -7, and -9 (Calbiochem, La Jolla, Calif)17 at a concentration of 60 μmol/L was infused at a rate of 5 μL/min and allowed to equilibrate for 30 minutes. Dialysate returning from both probes was collected into amber microcentrifuge tubes at 30-minute intervals. Fluorescence from dialysate samples (100 μL, FLUOstar Galaxy, BMG Laboratory Technologies, Durham, NC) was read at an excitation wavelength of 280 nm and an emission wavelength of 360 nm. For the purposes of obtaining reference control values, 5 age- and weight-matched pigs were instrumented to measure LV myocardial function and interstitial MMP activity.
Myocardial Histological and Biochemical Measurements
After the final set of measurements, the heart was removed and the LV was divided into MI and remote regions. These myocardial sections were flash-frozen for biochemistry or placed in formalin for histological staining.
LV sections (5 μm thick) were stained with picrosirius red, and the relative collagen percent areas in the MI and remote regions were determined with computer-assisted morphometric methods as described previously.18 Sections were imaged on an inverted microscope (Axioskop-2, Zeiss Corp, Peabody, Mass), and the images were digitized (AxioCam MRc, Zeiss). Collagen content was determined from the digitized images as a percentage of total tissue area in a minimum of 5 random high-power fields from each myocardial region of each pig.
Immunostaining was used to determine whether TIMPs colocalized with cells that stained positive for α-sarcomeric actin (marker for myocytes) and α-smooth muscle actin (marker for myofibroblasts) in sections from the MI region of UNSTIM pigs (n=2) and after LHFS (n=3).18 Sections from non-MI pigs (n=2) were stained identically and used as referent controls. Briefly, these myocardial sections were subjected to antigen unmasking (0.1 mg/mL proteinase K) and blocked for 1 hour at room temperature (3% normal goat serum and 1% BSA). After an overnight incubation at 4°C with primary antibodies (TIMP-1: Chemicon AB770, 1:100, Chemicon International, Inc, Temecula, Calif; α-sarcomeric actin: Sigma A2172, 1:400; α-smooth muscle actin: Sigma A5228, 1:400; α-sarcomeric actin: Sigma A2172, 1:400, Sigma, St Louis, Mo), slides were washed and incubated at room temperature with fluorochrome-conjugated secondary antibodies (FITC-conjugated goat anti-rabbit IgG [for α-smooth muscle actin staining], Cy3-conjugated donkey anti-mouse IgM [for α-sarcomeric actin staining], and Cy5-conjugated goat anti-mouse IgM [for TIMP-1 staining], Jackson ImmunoResearch, West Grove, Pa). Negative controls included the use of the secondary antibody only with preimmune serum. Fluorescent images of the sections were captured with a laser confocal microscope (Leica TCS SP2, Exton Pa). A minimum of 3 high-powered fields (63× objective) were imaged from each myocardial section, and immunopositive staining for each antibody was quantified as a percentage of the total area of each image.
Relative MMP-2 and MMP-9 levels were determined by gelatin zymography, and levels of MMP-1, MMP-8, MMP-13, MT1-MMP, TIMP-1, TIMP-2, and TIMP-4 were determined by immunoblotting.4 Abundance of α-smooth muscle actin, vimentin, prolyl-4-hydroxylase, hyaluronan-binding protein, transforming growth factor (TGF)-β, TGF-β receptors R1 and R2, SMAD-2/3, and phosphorylated SMAD-2/3 was determined by immunoblotting.19 Positive controls were included in each gel as appropriate. The zymograms and immunoblots were digitized, and levels of all analytes were quantified (Gel Pro Analyzer, Media Cybernetics, Bethesda, Md) by 2-dimensional integrated optical density. Myeloperoxidase activity was determined with ELISA (R&D Systems, Minneapolis, Minn).
In Vitro Electric Stimulation of Fibroblasts
Fibroblasts were isolated from the LV free wall of non-MI pigs with the outgrowth technique.20,21 Briefly, the LV myocardial samples were minced, transferred to cell culture flasks (75 cm2, Falcon, Franklin Lakes, NJ), and allowed to adhere. Sterile growth medium was added to the flasks, and the cells were incubated under standard cell culture conditions (37°C; 21% O2, 5% CO2) with culture media consisting of fibroblast growth medium (C23010, Promocell, Heidelberg, Germany), 20% FBS, and Promocell Supplement Mixture (C39315). After a 2-week incubation period, myocardial fibroblasts were scraped and transferred to 0.2% gelatin–coated (Sigma-Aldrich) tissue culture flasks and grown to confluence.
For the electric stimulation studies, confluent cultures from passage 2 were used. Fibroblasts were plated onto a 4-well chamber (plastic base) at a density of 6×105 cells per well. Once the fibroblasts were 80% confluent, the medium was changed to a serum-free medium, in which the cells incubated for 24 hours. The culture medium was replaced with fresh serum-free medium, and carbon (graphite) electrodes were placed at the ends of each chamber.22 The fibroblasts were electrically stimulated with 5-ms, 2-mA, 4-V/cm pulses of alternating polarity in each chamber. The cells either were stimulated at 4 Hz or were left unstimulated (0 Hz). After 24 hours of stimulation, the cell media was collected. The cells were trypsinized, and the cell count from each well (hemocytometer) was recorded. The cells were then centrifuged, and the cell pellets were resuspended in an extraction buffer.
Relative MMP-2 and MMP-9 levels were examined using gelatin zymography from the media samples.4 The relative abundance of MT1-MMP (AB38971) in the cell pellets and TIMP-1 (AB8116) in the media samples was determined by immunoblotting.4 The zymograms and immunoblots were quantified as above.
Data were collected in a blinded fashion and remained coded until the end of the study. Normality of data distribution for each variable was checked with the Shapiro-Wilk test. Nonparametric tests, as detailed below, were used for variables in which a normal distribution could not be assumed. Echocardiographic measurements of LV geometry and function were compared between the 2 MI groups with a 2-way ANOVA model, with time (21 and 28 days after MI) and treatment group (UNSTIM and LHFS) as the factors for the model. Single point measurements were compared between the control and MI groups with a 1-way ANOVA. After the ANOVA, posthoc pair-wise comparisons were performed with t tests corrected for number of comparisons by the Tukey method (module prcompw, STATA Corp, College Station, Tex). Differences in LV end-diastolic volume between post-MI days 21 and 28 were computed as a percentage change and compared between the Winsorized (robust) means of the 2 MI groups with the Mann–Whitney test.23 For the interstitial MMP activity measurements and biochemical and morphometric studies, comparisons to reference control values were performed with a 1-way ANOVA. Comparisons between the MI groups, in which the treatment effects were group and region, were performed with 2-way ANOVA. The relationships between collagen content, LV regional myocardial stiffness, and relative levels of MMP-1, MMP-8, and MMP-13 were examined through the use of least-squares linear regression analysis. Areas of positive immunostaining were compared between groups with the Kruskal-Wallis test. For the fibroblast studies, the integrated optical density values recorded from the zymographic and immunoblot assays were normalized to the number of cells in each well. The change in MMP and TIMP levels from unstimulated values was determined as the ratio of values recorded at each frequency and that of the average value for the unstimulated wells. This normalization procedure resulted in the MMP and TIMP levels in the unstimulated (0 Hz) group being assigned a value of 100%. MMP and TIMP levels between unstimulated and 4-Hz values were compared with a 1-way ANOVA, and comparisons to the unstimulated group were performed with a 2-tailed, 1-sample mean comparison test against the value of 100%. All statistical analyses were performed with the STATA statistical software package (version 8.0). Results are presented as mean±SEM. Two-tailed values of P<0.05 were considered statistically significant.
All 17 pigs entered in the study survived the initial instrumentation and MI induction. LV echocardiographic measurements recorded for the control pigs and at post-MI days 21 and 28 for the 2 MI groups are presented in Table 1. LV posterior wall thickness was reduced and septal wall thickness was increased at 21 days after MI in the UNSTIM group or the MI group with LHFS, with no difference between groups. At 28 days after MI, LV posterior wall thickness was reduced further in the UNSTIM group but remained similar to 21-day values in the LHFS group. The change in LV end-diastolic volume from 21 to 28 days after MI was smaller with LHFS compared with the UNSTIM group (3.2±2.6% versus 12.9±5.3%, respectively; P=0.03). LV ejection fraction in the post-MI period was similar in both groups.
Post-MI Day 28 Terminal Studies
At 28 days after MI, measurements of hemodynamics and regional LV geometry and function were performed with sonomicrometry. Hemodynamics and steady-state sonomicrometry recordings for the control and both MI groups are summarized in Table 2. Mean arterial pressures in both MI groups were similar to control values. Cardiac index and steady-state regional segmental shortening (Table 2) were reduced from reference control values in both MI groups. Within the MI region, LV regional myocardial stiffness was increased in both post-MI groups but was higher with LHFS compared with the UNSTIM group (Figure 2A). Interstitial MMP activity (Figure 2B) was higher than control values in the MI region of the UNSTIM group but was similar to control values with LHFS.
LV Myocardial Collagen Content
Morphometrically determined percent collagen (Figure 3) was increased within the MI regions in both MI groups compared with reference control levels. Compared with the UNSTIM group, however, relative collagen volume fraction within the MI region was increased with LHFS. There was a significant correlation between LV collagen content and regional myocardial stiffness in the MI region (Figure 3B).
LV Myocardial Biochemistry
LV profiles for the MMPs and TIMPs are shown in Figure 4. Compared with the UNSTIM group, MMP-2 levels were lower and there was a trend for lower MMP-1 levels (P=0.08) within the MI region of the MI+LHFS group. TIMP-1 levels were higher within the MI region with LHFS compared with the UNSTIM group. There was a significant inverse relationship between MMP-1 levels and collagen content (P<0.05; Figure 5A).
Cell Type Markers
Representative images (Figure 6) from immunohistological analysis showed that low amounts of TIMP-1, which was localized to interstitial space between myocytes (α-sarcomeric actin–positive cells stained in blue), could be detected within the viable myocardium from the non-MI control animals. The MI region in the UNSTIM and LHFS groups was devoid of α-sarcomeric actin staining, indicating an absence of myocytes. Nevertheless, TIMP-1 staining was clearly present within the MI region of both groups and was colocalized with α-smooth muscle actin–positive cells. However, the number of α-smooth muscle actin–positive cells and TIMP-1 colocalization appeared to be higher with LHFS than in UNSTIM. Immunoblotting of the myocardium from the remote and MI regions revealed that levels of α-smooth muscle actin, vimentin, prolyl-4-hydroxylase, and hyaluronan-binding protein were differentially higher with LHFS in the UNSTIM group.
Levels of TGF-β and TGF-β R1 within the MI region were increased over control values and that of the remote regions for both the UNSTIM and LHFS groups (Figure 7). However, TGF-β R1 levels were differentially higher in the LHFS group compared with the UNSTIM group. Levels of SMAD-2/3 and phosphorylated SMAD-2/3 were higher than control levels or levels in the remote myocardium within the MI region of both groups. However, phosphorylated SMAD-2/3 levels were higher with LHFS compared with UNSTIM. Consequently, the ratio of phosphorylated SMAD-2/3 to total SMAD-2/3 within the MI region was higher with LHFS than in the UNSTIM group.
Electric Stimulation of Isolated Fibroblasts
Two independent cultures of myocardial fibroblasts from each pig either were stimulated at 4 Hz or left unstimulated for a total of 6 experiments for the UNSTIM and 4-Hz groups. After 24 hours of stimulation, the cells retained spindle-shaped morphology (Figure 8). The number of fibroblasts increased with stimulation frequency and was significantly higher in the 4-Hz stimulated group compared with the UNSTIM group (Figure 8). Accordingly, levels of the MMPs and TIMP-1 were normalized to the number of cells in each well. MMP-2, MMP-9, and MT1-MMP levels at 4 Hz were lower than values obtained from fibroblasts that were left unstimulated (Figure 8). TIMP-1 levels normalized to cell number were higher at 4 Hz compared with that recorded in fibroblasts in the UNSTIM group (Figure 8).
A proteolytic event that contributes to adverse LV remodeling after an MI is changes in the abundance and balance between the MMPs and the endogenous TIMPs. Accordingly, the present study determined whether electric stimulation with LHFS pulses within a formed MI scar would alter the course of regional MI remodeling. Specifically, in the present study, LHFS was initiated in the MI scar at 21 days after MI and maintained for the final 7 days of the study period. In addition, the effects of electrically stimulating isolated fibroblasts with high-frequency (4-Hz) pulses on MMP/TIMP release were examined. The main findings of this study were 3-fold. First, LHFS attenuated MI thinning and increased regional stiffness of the MI region. Second, LHFS reduced the levels of certain MMP subtypes and increased TIMP-1 levels, which was accompanied by increased collagen content in the MI region. Finally, electric stimulation of myocardial fibroblasts resulted in a decrease in the release of specific MMP types and an increase in TIMP-1 levels, suggesting that the in vivo effects with respect to the decrease in interstitial MMP activity may have been due to direct effects of electric stimulation on fibroblasts/myofibroblasts that populate the MI region. Taken together, these findings provide evidence that a targeted nonpharmacological approach through subthreshold electric stimulation of the MI region may represent a novel means to attenuate and/or even prevent adverse LV remodeling after MI.
LV remodeling after MI is characterized by changes in the cellular and extracellular constituents at the infarcted region.2,3 The present study used a porcine model of MI that has been characterized in terms of infarct expansion, regional changes in MI geometry, and MMP and TIMP levels.4,5 Specifically, in this clinically relevant MI model, permanent occlusion of the obtuse marginals of the circumflex artery results in an MI size of 21% and is associated with progressive LV dilation and thinning of the MI region (called infarct expansion).4 The rationale for selecting 21 days after MI as the time point at which LHFS was initiated in the present study was 2-fold. First, in rodent MI models, interference with the early post-MI response in terms of modifying the cellular and/or extracellular characteristics of the MI region has been associated with deleterious effects such as an increased incidence of LV rupture.24 Second, during the later stages of MI remodeling, cellular and extracellular events that occur within the MI region culminate in the formation of a fibrotic scar (see elsewhere3 for a review). The presence of a fibrotic scar, which is generally considered to be nonconductive, was used to advantage in the present study with respect to confining the electric propagation of LHFS to within the MI region, as evidenced in ECG recordings. Therefore, initiating LHFS at 21 days after MI in this proof-of-concept study avoided the confounding influences with respect to interrupting the short-term MI “healing” response and provided sufficient time for the development of a nonconductive substrate. Nevertheless, it must be recognized that sufficient remodeling of the MI region with respect to fibrotic deposition could occur earlier than 21 days and thus may provide a suitable substrate to initiate LHFS earlier in the course of the post-MI remodeling process.
In the present study, LHFS over the final 7 days of the study period prevented the progressive thinning of the MI region. The presence of a thinned MI region results in a heterogeneous distribution of LV stress and strain patterns,25 which may place the viable, remote myocardium at a mechanical disadvantage and lead to a downward spiral of further MI expansion.3 Using mathematical stimulation, Pilla et al6 reported that increased stiffness in the MI region was predictive of an attenuation of LV dilation after MI. In the present study and consistent with past reports, myocardial stiffness at the MI region was differentially increased with LHFS and associated with the increase in collagen content in the MI region. Moreover, LHFS instituted during the final week of the post-MI study period attenuated the progression of LV dilation. However, despite this attenuation, the LV remodeling process was not “reversed” in that LV end-diastolic volumes were similar between the 2 MI groups. A potential explanation for this finding is that the duration of LHFS was too short for the effects on regional LV geometry to be translated into effects on global LV dilation. Future studies in which LHFS is instituted earlier and/or maintained for a longer duration after MI are required to further characterize the temporal effects of LHFS with respect to post-MI LV remodeling.
The MMPs contribute to proteolysis and remodeling of the extracellular matrix, and a causal role for several MMP types in LV remodeling after MI has been described.2,24,26 Consistent with past findings, MMP-2 and MMP-9 levels were increased within the MI region, but MMP-2 levels were reduced and there was a trend toward lower MMP-1 levels with LHFS. Concomitantly, there was an inverse relationship between MMP-1 levels and collagen content. TIMP-1 levels within the MI region, however, were higher with LHFS. Although changes in the relative ex vivo abundance of MMP types are an important consideration with respect to in vivo proteolytic activity, it must be recognized that processing the samples for assay can result in dissociation of MMP complexes with TIMPs and other interstitial proteins, as well as linearization of the latent and active forms of the MMPs. To avoid these inherent limitations of ex vivo approaches, the present study used a previously validated small-peptide substrate with a quenched fluorescent moiety, which, when cleaved by MMPs, would yield a detectable fluorescent signal.17 Results from this analysis showed that in vivo interstitial MMP activity within the MI region was reduced with LHFS. However, it must be recognized that the fluorogenic substrate used in the present study provided an “integrative” index of contributions from several MMP types. This laboratory has recently demonstrated that MMP substrates with a more specific amino acid sequence, which therefore impart greater MMP specificity, can be successfully used with this microdialysis system.17,27 Thus, future studies that use different MMP substrates would allow identification of the effects of LHFS on the activity of specific MMP types within the MI region. Nevertheless, findings from the present study demonstrated that the ex vivo reductions in the levels of certain MMPs with a concomitant increase in TIMP-1 levels likely contributed to the in vivo reduction of interstitial MMP activity, which in turn suggests that there was a net in vivo reduction in extracellular matrix proteolysis that contributed to the increase in collagen content of the MI region with LHFS.
MMPs and TIMPs are synthesized and released by a number of myocardial cell types, including myocytes and fibroblasts.18,28,29 In addition, inflammatory cell types, such as macrophages and neutrophils, and myofibroblasts, which are differentiated fibroblasts, release MMPs.18 Although myocytes make up a majority of the cellular volume of the viable myocardium, there are relatively few, if any, myocytes within the MI region.1,2 Indeed, in the present study, there was an absence of immunostaining for α-sarcomeric actin within the MI region. Therefore, it is unlikely that myocytes were a source for the LHFS-related changes in MMP/TIMP levels within the MI region. Furthermore, myeloperoxidase activity levels were similar to non-MI control values in the remote and MI regions of pigs with or without LHFS. This finding suggests that by 28 days after MI, the effects of LHFS on MMP/TIMP levels were unlikely to be driven by modulating inflammation. Nevertheless, immunohistochemical staining of the MI region revealed that TIMP-1 colocalized with cells that stained for α-smooth muscle actin, a well-established marker for myofibroblasts.18 However, the colocalization of TIMP-1 to cells staining for α-smooth muscle actin appeared to be greater with LHFS. Moreover, the levels of cell markers for mesenchymal cell types such as α-smooth muscle actin and vimentin within the MI region were differentially higher with LHFS. Finally, in vitro electric stimulation of fibroblasts reduced the abundance of MMP-2, MMP-9, and MT1-MMP compared with fibroblasts that were left unstimulated. Concomitantly, TIMP-1 levels were increased in the fibroblasts subjected to in vitro electric stimulation. Therefore, the stoichiometric balance between MMPs and TIMPs released from fibroblasts in response to electric stimulation was shifted to one that would likely favor collagen accumulation. Taken together, these findings of the present study suggest that the cellular source for the differentially higher TIMP-1 levels and collagen content with LHFS was likely fibroblasts and/or myofibroblasts.
Increased levels of TGF-β and activation of the TGF-β signaling pathway, which includes the TGF-β R1 and R2 and the SMAD second messenger system, are associated with fibrosis in MI.30–33 Consistent with these past findings, the present study demonstrated that levels of TGF-β, TGF-β R1, SMAD-2/3, and phosphorylated SMAD-2/3 were higher within the MI region than control levels. Moreover, the levels of TGF-β R1 and the ratio of phosphorylated SMAD-2/3 to total SMAD-2/3 were differentially higher with LHFS. SMAD-2/3 requires phosphorylation to be chaperoned across the nuclear membrane to effect transcription of a number of genes.19,32,33 Fibroblasts and myofibroblasts have been previously shown to respond to TGF-β stimulation through increased synthesis and release of fibrillar collagens.1,28,31,33,34 Therefore, the findings of the present study suggest that a fundamental mechanism, at least in part, for the differentially higher collagen content within the MI region with LHFS was activation of the TGF-β signaling pathway and release of fibrillar collagen from fibroblasts/myofibroblasts.
A major consideration for selecting the LHFS parameters used in the present study was to prevent “escape” of the introduced stimuli into viable myocardium, which may have resulted in ventricular tachycardia. This is an important consideration because the presence of a fibrotic post-MI scar can form an area of conduction block35 and origination sites for reentrant rhythms as a result of changes in the refractory properties and/or changes in the length of conduction pathways.36 Moreover, variability in MI size, the presence of alternate high-resistance conduction (eg, through islets of remnant viable myocardium in the MI region), and case-to-case variation in fibrillation thresholds are some factors that must be considered in future studies that implement LHFS in the setting of MI.
The present study provided proof of concept that LHFS could be designed as a “therapy” to provide a targeted, nonpharmacological approach to attenuate infarct expansion after the formation of a fibrotic scar. Potential advantages of the LHFS approach include regional and temporal specificity in terms of directing treatment. For example, regional specificity would be achieved through the placement of the electrode in the area of targeted intervention, which may hold therapeutic benefit over systemic pharmacological approaches. With respect to temporal specificity, LHFS may afford the opportunity to activate and deactivate the pacemaker based on achieving a potential remodeling response. Although the prospects of implementing LHFS in the clinical setting are provocative, it must be recognized that the findings reported here, albeit determined in a clinically relevant animal model, represent just the short-term effects of LHFS. For instance, it is possible that long-term LHFS could result in late arrhythmias, which would not be apparent in the present study design. Therefore, extrapolation to the clinical context must be undertaken with caution and only after further characterization of the later effects of LHFS on LV geometry and associated changes in myocardial material properties.
Sources of Funding
This study was supported in part by National Heart, Lung, and Blood Institute grants HL-45024, HL-97012, and PO1-48788 and a Veterans Administration Merit Award (Dr Spinale).
Dr Spinale has received grants from the National Institutes of Health and the Veterans Administration. The other authors report no conflicts.
Creemers EE, Cleutjens JP, Smits JF, Daemen MJ. Matrix metalloproteinase inhibition after myocardial infarction: a new approach to prevent heart failure? Circ Res. 2001; 89: 201–210.
Sutton MG, Sharpe N. Left ventricular remodeling after myocardial infarction: pathophysiology and therapy. Circulation. 2000; 101: 2981–2988.
Mukherjee R, Brinsa TA, Dowdy KB, Scott AA, Baskin JM, Deschamps AM, Lowry AS, Escobar GP, Lucas DG, Yarbrough WM, Zile MR, Spinale FG. Myocardial infarct expansion and matrix metalloproteinase inhibition. Circulation. 2003; 107: 618–625.
Wilson EM, Moainie SL, Baskin JM, Lowry AS, Deschamps AM, Mukherjee R, Guy TS, St John-Sutton MG, Gorman JH III, Edmunds LH Jr, Gorman RC, Spinale FG. Region- and type-specific induction of matrix metalloproteinases in post-myocardial infarction remodeling. Circulation. 2003; 107: 2857–2863.
Kloth LC. How to use electrical stimulation for wound healing. Nursing. 2002; 32: 17.
Kloth LC. Electrical stimulation for wound healing: a review of evidence from in vitro studies, animal experiments, and clinical trials. Int J Low Extrem Wounds. 2005; 4: 23–44.
Kamrin BB. Induced collagenolytic activity by electrical stimulation of embryonic fibroblasts in tissue culture. J Dent Res. 1974; 53: 1475–1483.
Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum H, Gutgesell H, Reichek N, Sahn D, Schnittger I, Silverman NH, Tajik AJ. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography: American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. J Am Soc Echocardiogr. 1989; 2: 358–367.
Deschamps AM, Yarbrough WM, Squires CE, Allen RA, McClister DM, Dowdy KB, McLean JE, Mingoia JT, Sample JA, Mukherjee R, Spinale FG. Trafficking of the membrane type-1 matrix metalloproteinase in ischemia and reperfusion: relation to interstitial membrane type-1 matrix metalloproteinase activity. Circulation. 2005; 111: 1166–1174.
Mukherjee R, Mingoia JT, Bruce JA, Austin JS, Stroud RE, Escobar GP, McClister DM Jr, Allen CM, Alfonso-Jaume MA, Fini ME, Lovett DH, Spinale FG. Selective spatiotemporal induction of matrix metalloproteinase-2 and matrix metalloproteinase-9 transcription after myocardial infarction. Am J Physiol Heart Circ Physiol. 2006; 291: H2216–H2228.
Rivest L-P. Statistical properties of Winsorized means for skewed distributions. Biometrika. 1994; 81: 373–383.
Heymans S, Luttun A, Nuyens D, Theilmeier G, Creemers E, Moons L, Dyspersin GD, Cleutjens JP, Shipley M, Angellilo A, Levi M, Nube O, Baker A, Keshet E, Lupu F, Herbert JM, Smits JF, Shapiro SD, Baes M, Borgers M, Collen D, Daemen MJ, Carmeliet P. Inhibition of plasminogen activators or matrix metalloproteinases prevents cardiac rupture but impairs therapeutic angiogenesis and causes cardiac failure. Nat Med. 1999; 5: 1135–1142.
Deschamps AM, Zavadzkas J, Murphy RL, Koval CN, McLean JE, Jeffords L, Saunders SM, Sheats NJ, Stroud RE, Spinale FG. Interruption of endothelin signaling modifies membrane type 1 matrix metalloproteinase activity during ischemia and reperfusion. Am J Physiol Heart Circ Physiol. 2008; 294: H875–H883.
Spinale FG. Myocardial matrix remodeling and the matrix metalloproteinases: influence on cardiac form and function. Physiol Rev. 2007; 87: 1285–1342.
Ellmers LJ, Scott NJ, Medicherla S, Pilbrow AP, Bridgman PG, Yandle TG, Richards AM, Protter AA, Cameron VA. Transforming growth factor-beta blockade down-regulates the renin-angiotensin system and modifies cardiac remodeling after myocardial infarction. Endocrinology. 2008; 149: 5828–5834.
Jugdutt BI. Extracellular matrix and cardiac remodeling. In: Villarreal FJ, ed. Interstitial Fibrosis in Heart Failure. New York, NY: Springer; 2000: 23–56.
Bujak M, Ren G, Kweon HJ, Dobaczewski M, Reddy A, Taffet G, Wang XF, Frangogiannis NG. Essential role of Smad3 in infarct healing and in the pathogenesis of cardiac remodeling. Circulation. 2007; 116: 2127–2138.
Bujak M, Frangogiannis NG. The role of TGF-beta signaling in myocardial infarction and cardiac remodeling. Cardiovasc Res. 2007; 74: 184–195.
Sun Y, Weber KT. Infarct scar: a dynamic tissue. Cardiovasc Res. 2000; 46: 250–256.
Soejima K, Stevenson WG, Maisel WH, Sapp JL, Epstein LM. Electrically unexcitable scar mapping based on pacing threshold for identification of the reentry circuit isthmus: feasibility for guiding ventricular tachycardia ablation. Circulation. 2002; 106: 1678–1683.
Stevenson WG, Khan H, Sager P, Saxon LA, Middlekauff HR, Natterson PD, Wiener I. Identification of reentry circuit sites during catheter mapping and radiofrequency ablation of ventricular tachycardia late after myocardial infarction. Circulation. 1993; 88 (pt 1): 1647–1670.
Adverse left ventricular remodeling after myocardial infarction (MI) remains an important cause of morbidity and mortality. Current pharmacological strategies fail to interrupt or reverse the inexorable progress of post-MI remodeling. Recent translational or early clinical research has focused on the development of new pharmacological modalities or the feasibility of delivery of exogenous cell types to prevent post-MI remodeling. However, whether endogenous cell types can be recruited and/or targeted to prevent, or even reverse, left ventricular remodeling after MI has not been explored. In the present proof-of-concept study, low-amplitude, high-frequency electric stimulation instituted within a formed MI scar arrested the progressive thinning of the MI region and attenuated left ventricular dilation. A likely mechanism for these findings was the direct effects of electric stimulation of fibroblasts/myofibroblasts resident within the MI region. These findings suggest that targeted electric stimulation of the MI region can alter the endogenous substrate for post-MI remodeling. In light of the fact that multisite myocardial pacing is a commonly used clinical tool, translational studies to extend the basic observations of the present study are warranted.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.110.936872/DC1.