Abstract
After a myocardial infarction (MI), an increase in the cardiac ratio of matrix metalloproteinases (MMPs) relative to their inhibitors (TIMPs) causes extracellular matrix modulation that leads to ventricular dilatation and congestive heart failure. Cell therapy can mitigate these effects. In this study, we tested whether increasing MMP inhibition via cell-based gene transfer of Timp-3 further preserved ventricular morphometry and cardiac function in a rat model of MI. We also measured the effect of treatment timing. We generated MI (coronary artery ligation) in adult rats. Three or 14 days later, we implanted medium (control) or vascular smooth muscle cells transfected with empty vector (VSMCs) or Timp-3 (C-TIMP-3) into the peri-infarct region (n = 15—24/group). We assessed MMP-2 and −9 expression and activity, TIMP-3, and TNF-α expression, cell apoptosis, infarct size and thickness, ventricular morphometry, and cardiac function (by echocardiography). Relative to medium, VSMCs delivered at either time point significantly reduced cardiac expression and activity of MMP-2 and −9, reduced expression of TNF-α, and increased expression of TIMP-3. Cell therapy also reduced apoptosis and scar area, increased infarct thickness, preserved ventricular structure, and reduced functional loss. All these effects were augmented by C-TIMP-3 treatment. Survival and cardiac function were significantly greater when VSMCs or C-TIMP-3 were delivered at 3 (vs. 14) days after MI. Upregulating post-MI cardiac TIMP-3 expression via cell-based gene therapy contributed additional regulation of MMP, TIMP, and TNF-α levels, thereby boosting the structural and functional effects of VSMCs transplanted at 3 or 14 days after an MI in rats. Early treatment may be superior to late, though both are effective.
Keywords
Introduction
A main contributor to the adverse changes in ventricular dimensions and function that follow a myocardial infarction (MI) is a shift in the expression or activity of the degradative matrix metalloproteinase (MMP) enzymes relative to that of their natural tissue inhibitors (TIMPs). The TIMP family of protease inhibitors comprises four proteins (TIMP-1, −2, −3, and −4) involved in the maintenance of tissue structure. The loss of any one of these proteins can initiate the MMP/TIMP imbalance that leads to maladaptive remodeling after an MI. Specifically, a relative increase in MMP levels triggers remodeling of the extracellular matrix—a dynamic scaffold that supports cardiomyocytes and facilitates contraction (9). These changes promote ventricular dilatation and dysfunction (24) and are not reversed by current medical therapies (21). Reestablishing the balance by increasing TIMP expression has been shown to mitigate matrix remodeling and treat ischemic cardiomyopathy caused by an MI (5), but establishing clinically relevant methods has been a challenge (30). For example, chronic administration of MMP inhibitors was not useful (12,22,25), and gene therapies are nonspecific (25).
We recently demonstrated that the beneficial paracrine effects of cell transplantation (8) are enhanced by genetically modifying the implanted cells for increased cytokine production (11,14,32). Potentially more clinically effective than traditional gene therapies, so-called cell-based gene therapies permit a temporally and spatially regulated release of the gene product. Using Timp-3 knockout (Timp-3-/-) mice, we confirmed the unique role of TIMP-3 in maintaining ventricular dimensions and preventing matrix disruption after an MI (28) and showed that reinstating TIMP activity via cell-based Timp-3 gene therapy enhanced the functional effects of cell therapy by transiently inhibiting MMP activity in the mutant mice (2). These findings have yet to be confirmed in wild-type animals.
In another study, we directly compared early versus late post-MI implantation of unmodified skeletal myoblasts into the infarcted myocardium in rats and found that cardiac function was improved regardless of the timing of cell delivery (6). In that study, myocardial MMP activities were suppressed by cells delivered during, but not after, the first week following an MI. We hypothesized that, when treatment is given later after an MI, an enhanced therapy (rather than cells alone) might be required to restore the MMP/TIMP ratio and prevent matrix disruption and ventricular dilatation.
Here, we evaluate the effects of MMP inhibition by cell-based gene transfer of Timp-3 on ventricular morphometry and cardiac function in a normal rat model of MI and compare the results of early versus late treatment (3 vs. 14 days after MI).
Materials and Methods
Animals
All experiments were performed in accordance with the principles of laboratory animal care formulated by the Guide for the Care and Use of Laboratory Animals by the Institute of Laboratory Animal Resources (Commission on Life Sciences, National Research Council). All animal procedures were approved by the Harbin Medical University Animal Care Committee. Female Wistar rats (aged 10—12 weeks) were obtained from the Harbin Medical University Animal Center (Harbin, China).
Isolation, Transfection, and Characterization of Vascular Smooth Muscle Cells
Vascular smooth muscle cells (VSMCs) were isolated from the adult female rat aorta as previously described (19). The aorta was minced into small pieces after it was washed with PBS and scraped to remove the endothelium. Pieces of tissue that adhered to the culture dish were mixed with 5 ml Dulbecco's modified Eagle medium (DMEM) containing 20% fetal bovine serum to obtain primary VSMCs. The primary cells were subsequently cultured in DMEM containing 10% fetal bovine serum, 100 U/ml penicillin, and 100 μg/ml streptomycin for 30 min at 37°C and 5% CO2. The purity of the cultures was evaluated via immunohistochemical staining with monoclonal antibodies against α-smooth muscle actin. VSMCs were transfected using liposomal lipofectamine 2000 (Invitrogen), either with a purified plasmid comprising murine Timp-3 cDNA cloned into pcDNA3.1+ (Invitrogen) as previously described (13), or with an empty vector (control). Cellular Timp-3 mRNA levels were assessed using RT-PCR with specific primers as previously described (4). Timp-3 transfection was confirmed by the clear presence of a transfection band at the 464 bp location.
Surgical Procedures
With the rat under general anesthesia, a thoracotomy was performed and the left coronary artery was permanently ligated distal to the left atrial appendage. Ischemia was confirmed by the appearance of hypokinesis and pallor distal to the occlusion. The animals were randomly separated into six groups (n = 28/group) and allowed to recover. Three or 14 days later, a second thoracotomy was performed, and VSMCs transfected 3 days earlier with Timp-3 plasmid (C-TIMP-3 groups) or vector (VSMC groups) were delivered by three direct, intramyocardial injections (1 × 106 cells in 100 μl of DMEM per injection) into three separate points in the peri-infarct area. Control animals (medium groups) received three injections of culture medium (100 μl/ injection) at 3 or 14 days after MI. A group of sham animals underwent both operations without coronary artery ligation or injections. Each animal was treated postoperatively with antibiotics (Duplocillin LA 30,000 IU/kg) and analgesics (buprenorphine 0.05 mg/kg) delivered subcutaneously.
Functional Cardiac Assessment
Left ventricular (LV) dilatation and function were evaluated by echocardiography (Sequoia C256 System, Siemens Medical; 15-MHz linear array transducer) in all groups (n = 15—24 rats per group). M-mode and 2D images were obtained in the parasternal short axis view at the level of the papillary muscles. For each measurement, three consecutive cardiac cycles were recorded and averaged by a single, blinded examiner. LV internal diastolic dimension (LVIDd) and internal systolic dimension (LVIDs) were determined in M-mode imaging. LV end-diastolic volume (EDV) and end-systolic volume (ESV) were calculated using the following formulae: EDV = 1.047 x LVIDd3; ESV = 1.047 x LVIDs3. Percent ejection fraction (%EF) and fractional shortening (%FS) of the LV were calculated as follows: %EF = [(EDV − ESV)/EDV] x 100; %FS = [(LVIDd − LVIDs)/LVIDd] x 100 (3,18).
Morphometric Cardiac Assessment
After the rats were sacrificed at 28 days post-MI, hearts (n = 15—24 per group) were fixed with buffered 10% formalin at 20 mmHg ventricular pressure for 5 days. The fixed hearts were cut into 1-mm-thick transverse slices from apex to base and photographed. Images were analyzed by a blinded observer using ImageJ software. The border between infarcted and noninfarcted myocardium was clearly identifiable in each image (infarcted myocardium appeared visibly thinner and paler in color than noninfarcted myocardium). Scar area and thickness were calculated using computed planimetry as we have previously described (27). Briefly, both apical and basal sections were digitally photographed and quantified. We calculated the epicardial surface areas of the LV free wall (LVFW) and any scar tissue in the LVFW as the sum of the epicardial length times the section thickness (1 mm). We next calculated the scar surface area as a percentage of the LVFW area as follows: (epicardial scar area)/(epicardial LVFW area) x 100. Scar thickness was presented as an average of wall thickness measurements taken at the middle and at each edge of the scar area.
Ultrastructural Cardiac Analysis
For transmission electron microscopy (TEM), we sampled cardiac tissue from the peri-infarct region. Small blocks (0.5—1 mm3) of cardiac tissue from the peri-infarct (border) region were fixed in 1% glutaraldehyde in phosphate buffer and then sent to the Department of Cardiovascular Pathology at the Harbin Medical University (Harbin, China) for electron microscopic analysis. The samples were postfixed with 1% osmium tetroxide, embedded, sliced, and photographed as we have previously reported (17).
Myocardial MMP and TIMP-3 Expression Measurement
Hearts [n = 9 per treated group, n = 4 per sham group (normal controls)] were snap frozen. The heart homogenate was treated with an appropriate extraction buffer and then protein expression levels of MMP-2, MMP-9, and TIMP-3 were measured by Western blot (9). Each sample contained tissue from the infarct area and the remote area. We used housekeeping protein (β-actin) as the internal standard for the MMP and TIMP-3 calculation. The ratio of protein/actin was used to quantify MMP and TIMP-3 protein levels in each group. All antibodies were from Chemicon.
Myocardial MMP Activity Measurement
Hearts (n = 5 per treated group, n = 3 per sham group) were snap frozen at 3 days after cell transplantation. The activities of myocardial MMP-2 and MMP-9 were quantified using gelatin zymography as described previously (28). The values of gelatinase activity are expressed as intensity unit.
Myocardial Tumor Necrosis Factor-α (TNF-α) Expression Measurement
Myocardial tissue (n = 9 hearts per treated group, n = 4 per sham group) was homogenized and the supernatant was collected. Tumor necrosis factor-α (TNF-α) levels were quantified using an enzyme-linked immunosorbent assay (commercial TNF-α ELISA kit, Invitrogen Canada Inc., Burlington, Canada, Catalog # KRC3011, used according to manufacturer's instructions). TNF-α concentrations were calculated based on a standard curve. Each sample was analyzed in triplicate and then averaged.
Myocardial Inflammation Measurement
The number of infiltrating inflammatory cells was measured by a semiquantitative method (assessment of cell nuclear density) from H&E-stained slides 3 days following cell implantation (n = 5 per treated group, n = 3 per sham group). The degree of infiltration was classified on the scale of + (indicating the lowest infiltration) to +++ (indicating the highest infiltration).
Myocardial Apoptosis Measurement
For this analysis, terminal dUTP nick-end labeling (TUNEL) assay was performed on tissue sections 3 days following implantation (n = 5 per treated group, n = 3 per sham group) with an in situ Cell Death Detection kit (Roche Inc.) as previously described (28). Apoptotic cell death was determined by counting the number of TUNEL-positive nuclei per microscopic field (200x equals 0.4 mm2) in five fields per slide and then averaged.
Statistical Methods
Results are expressed as mean ± standard deviation of the mean. Analyses were performed using GraphPad Prism software (v.4) with the critical α-level set at p < 0.05. Mortality rates were compared using a chi-square test. Survival was analyzed with Kaplan-Meier survival curves and log-rank statistics. Two-way analyses of variance (with Bonferroni multiple comparison posttests to specify differences between the groups) compared the effects of treatment (sham, medium, VSMC, C-TIMP-3) and treatment timing (3 vs. 14 days after MI) on cardiac morphometry, cardiac function, levels of MMP, TIMP-3, and TNF-α, and cell apoptosis.
Results
Mortality
Mortality rates were calculated for the 28-day duration of the experiment. Whether implantation was at 3 or 14 days after MI, mortality rates were significantly higher (p < 0.05) in the medium groups than in the VSMC groups and were the lowest in the C-TIMP-3 groups (p < 0.05 vs. VSMC) (Fig. 1a). Comparing early and later treatment, both VSMC and C-TIMP-3 groups experienced significantly less mortality when the cells were delivered at 3 days rather than 14 days after MI (p < 0.05 for both groups). Log-rank statistics confirmed that percent survival was significantly higher (p < 0.05) in animals that received cells at the 3-day time point (Figs. 1b, c). Treatment timing did not affect mortality in the control groups.
Mortality and survival rates after myocardial infarction (MI). (a) Whether treatment was delivered at 3 or 14 days after MI (D3, D14, respectively), the mortality rates (%) were highest in the medium groups and lowest in the C-TIMP-3 groups. Mortality rates were also significantly reduced when VSMC or C-TIMP-3 was injected at D3 rather than D14. VSMC, vascular smooth muscle cells transfected with vector; C-TIMP-3, VSMCs transfected with tissue inhibitor of metalloproteinases protein 3 (Timp-3). *p < 0.05 versus medium, #p < 0.05 versus VSMC; n = 15—24/ group. Kaplan-Meier survival curves illustrate percent survival (at different time points from 14 to 28 days after MI) in animals that were treated on D3 or D14 with VSMC (b) or C-TIMP-3 (c). Survival rates were higher (p < 0.05) when the cells were delivered at 3 versus 14 days after the MI.
Left Ventricular Morphometry and Ultrastructure
Measures of LV dilatation (LVIDd and LVIDs) and LV volumes (EDV and ESV) assessed using echocardiography at 28 days post-MI (Fig. 2), as well as LV volume indexed by animal body weight (Table 1), indicated that LV structure was significantly preserved (vs. medium) after cell treatment (VSMC), with a further beneficial effect after treatment with gene-enhanced cells (C-TIMP-3) (p < 0.05 for all comparisons). Similarly, myocardial scars were the smallest and thickest in the C-TIMP-3 groups (p < 0.05 vs. VSMC) and the largest and thinnest in the medium groups (p < 0.01 vs. VSMC, C-TIMP-3) (Fig. 3a—c).
Ventricular morphometry by echocardiography at 28 days after MI. Whether treatment was delivered at 3 or 14 days after MI (D3, D14, respectively), parameters of left ventricular (LV) dilatation and LV volumes were largest in the medium groups and smallest in the C-TIMP-3 groups. (a) Left ventricular internal diastolic dimension (LVIDd), (b) left ventricular internal systolic dimension (LVIDs), (c) end-diastolic volume, (d) end-systolic volume. LVIDd was significantly reduced when VSMC or C-TIMP-3 was injected at D3 rather than D14. Sham, sham-operated controls; VSMC, vascular smooth muscle cells transfected with vector; C-TIMP-3, VSMCs transfected with Timp-3. *p < 0.05 versus medium, #p < 0.05 versus VSMC; n = 15—24/group.
Scar morphometry and ultrastructure at 28 days after MI. (a) Photographs of representative heart sections from each group. D3, treatment delivered at 3 days after MI; D14, treatment delivered at 14 days after MI; VSMC, vascular smooth muscle cells transfected with vector; C-TIMP-3, VSMCs transfected with Timp-3. Arrows indicate the location of the infarct in sample sections. Whether treatment was delivered at D3 or D14, scar areas (b) were reduced and scar thicknesses (c) were preserved in the VSMC and C-TIMP-3 groups compared with the medium groups, and scars were smallest and thickest in the C-TIMP-3 groups. There were no differences due to treatment timing (D3 or D14). **p < 0.01 versus medium, #p < 0.05 versus VSMC; n = 15—24/ group. (d) Representative transmission electron micrographs illustrating ultrastructure of infarct area tissue from each group. S, sarcomere; N, nucleus. Arrows indicate mitochondria.
Left Ventricular Volume Index (mm3/g Body Weight)
Left ventricular volume index was measured at 28 days after myocardial infarction (MI). Sham, sham-operated controls; medium, medium-injected controls; VSMC, vascular smooth muscle cells transfected with vector; C-TIMP-3, VSMCs transfected with tissue inhibitor of metalloproteinase protein 3 (Timp-3). p < 0.01 for all medium, VSMC, C-TIMP-3 compared to sham.
p < 0.01 for C-TIMP-3 and VSMC compared to medium.
p < 0.01 for C-TIMP-3 compared to VSMC.
p <0.05 for transplantation on day 14 compared to day 3.
An ultrastuructural analysis of myocardial tissue at the border zone collected at 28 days after MI revealed clear group differences in histology (Fig. 3d). In the medium group, cardiomyocytes in the border area had condensed nuclei, fractured and dissolved mitochondrial cristae, and low matrix density. In the VSMC group, the sarcomeres were clear with visible myocyte contraction bands, and the mitochondria were only mildly swollen. In the C-TIMP-3 group, the cardiomyocytes were in good condition, with rich cytoplasmic organelles (including visible nucleoli), intact mitochondrial cristae, and balanced matrix density.
Most of these effects were not dependent on timing of treatment, with two exceptions: although treatment timing did not affect ventricular dimensions in the medium groups, LVIDd and LV volume indices were significantly smaller following the delivery of VSMC or C-TIMP-3 at 3 days, rather than 14 days, after MI (p < 0.05 for all comparisons) (Fig. 2a, Table 1). Therefore, treatment with VSMCs or VSMCs overexpressing TIMP-3 reduced the magnitude of ventricular dilatation, scar area, and wall thinning recorded in control animals. Overall LV geometry was further preserved by the gene-enhanced cells, and there was some evidence that cell (and cell-based gene) treatment may be most effective if delivered early, rather than later, after an MI.
Left Ventricular Function
By echocardiography, %EF and %FS were significantly increased at 28 days after MI in the VSMC groups (p < 0.05 vs. medium), with further improvements in the C-TIMP-3 groups (p < 0.05 vs. VSMC). Also, %EF and %FS were greatest in VSMC and C-TIMP-3 groups that received cells at 3 days (vs. 14 days) after MI (p < 0.05 for all comparisons), indicating that there was a benefit associated with early cell transplantation (Fig. 4).
Cardiac function by echocardiography at 28 days after MI. Whether treatment was delivered at 3 or 14 days after MI (D3, D14, respectively), percent ejection fraction (EF, a) and fractional shortening (FS, b) were improved (vs. medium groups) in the VSMC groups, and further improved in the C-TIMP-3 groups. EF and FS were significantly greater when VSMC or C-TIMP-3 were injected on D3 rather than D14. Sham, sham-operated controls; VSMC, vascular smooth muscle cells transfected with vector; C-TIMP-3, VSMCs transfected with Timp-3. *p < 0.05 versus medium, #p < 0.05 versus VSMC; n = 15—24/group.
MMP and TIMP-3 Protein Expression
Myocardial protein levels of MMP-2, MMP-9, and TIMP-3 were evaluated at 28 days after MI. The analysis revealed stronger MMP and weaker TIMP-3 protein bands in the medium control group compared with the VSMC and C-TIMP-3 groups (Fig. 5a). Inversely, compared with levels recorded in control animals (medium groups), there were significant decreases in MMP-2 and MMP-9 protein levels in those that received VSMCs (p < 0.01 vs. medium) (Fig. 5b, c) with further reductions in animals that received the gene-enhanced cells (p < 0.05 vs. VSMC) (Fig. 5b, c). Meanwhile, TIMP-3 protein levels were significantly increased in the VSMC groups (p < 0.01 vs. medium), and most increased in the C-TIMP-3 groups (p < 0.05 vs. VSMC) (Fig. 5d).
MMP and TIMP-3 protein expression at 28 days after MI. The hearts were evaluated by Western blot for matrix metalloproteinase-2 (MMP-2), MMP-9, and TIMP-3 proteins. (a) Blots shown are representative of three independent experiments and normalized to β-actin. Whether treatment was delivered at 3 or 14 days after MI (D3, D14, respectively), MMP-2 protein levels (b) and MMP-9 protein levels (c) were reduced (vs. medium groups) in the VSMC groups, and further reduced in the C-TIMP-3 groups. (d) Whether treatment was delivered at D3 or D14, TIMP-3 protein levels were increased (vs. medium groups) in the VSMC groups, and further increased in the C-TIMP-3 groups. Sham, sham-operated controls; VSMC, vascular smooth muscle cells transfected with vector; C-TIMP-3, VSMCs transfected with Timp-3. **p < 0.01 versus medium, #p < 0.05 versus VSMC; n = 9/group (medium, VSMC, C-TIMP-3), n = 4/group (sham).
MMP Activity
MMP activities were measured by gel zymography at 3 days after cell transplantation (Fig. 6a). The MMP-2 and −9 activities were highest in the day 3 and day 14 medium groups (Fig. 6a—c), followed by the day 3 and day 14 VSMC groups (p < 0.05 vs. medium group) (Fig. 6a—c), and were lowest in the day 3 C-TIMP-3 groups (p < 0.05 vs. VSMC group) (Fig. 6a—c). There was no difference between day 14 VSMC and C-TIMP-3 groups.
MMP activity at 3 days after implantation. The hearts were evaluated by gelatin zymography for MMP-2 and MMP-9 activity. (a) Representative zymography of D3 and D14 groups. Densitometry analyses illustrate that MMP-2 (b) and MMP-9 (c) gelatinolytic activities were highest in the medium groups for both D3 and D14, followed by the VSMC group for D3. The enzyme activities in the C-TIMP-3 group were the lowest for D3, but similar to VSMC for D14. Sham, sham-operated controls; VSMC, vascular smooth muscle cells transfected with vector; C-TIMP-3, VSMCs transfected with Timp-3. *p < 0.05 versus medium, #p < 0.05 versus VSMC; n = 5/group (medium, VSMC, C-TIMP-3), n = 3/group (sham).
TNF-α Expression, Inflammation, and Apoptosis
By a semiquantitative method (assessment of cell nuclear density), the number of infiltrating cells (Fig. 7a, arrows) was highest in the medium group (+++), followed by the VSMC group (++), and was lowest in the C-TIMP-3 group (+) at 3 days after cell transplantation (Fig. 7a).
Inflammation at 3 days after implantation and TNF-α expression at 28 days after MI. (a) Representative light micrographs (400x) illustrating infiltrating inflammatory cells (arrows) in the infarct border zone of rat hearts after implantation of medium, VSMC, or C-TIMP-3. Normal myocardium (sham) is presented as a control. Using a semiquantitative method (assessment of cell nuclear density), the number of infiltrating cells was highest in the medium group (+++), followed by the VSMC group (++), and the C-TIMP-3 group (+). MI, myocardial infarction; M, cardiomyocytes. (b) Whether treatment was delivered at 3 or 14 days after MI (D3, D14, respectively), tumor necrosis factor-α (TNF-α) expression was reduced (vs. medium groups) in the VSMC groups, and further reduced in the C-TIMP-3 groups. There were no differences due to treatment timing (D3 or D14). Sham, sham-operated controls; VSMC, vascular smooth muscle cells transfected with vector; C-TIMP-3, VSMCs transfected with Timp-3. **p < 0.01 versus medium, #p < 0.01 versus VSMC; n = 5—9/group (medium, VSMC, C-TIMP-3), n = 3—4/group (sham).
At 28 days after the MI, TNF-α levels were elevated in control animals (in medium vs. sham groups). In comparison to levels in the medium group, expression of this cytokine was significantly reduced in animals that received VSMCs (p < 0.01 vs. medium) (Fig. 7a) and was the lowest (much closer to sham levels) after treatment with the gene-enhanced cells (p < 0.01 vs. VSMC) (Fig. 7b). TNF-α levels did not vary with the timing of treatment.
At 3 days after cell transplantation, the number of apoptotic cells (Fig. 8a, arrows) was highest in the medium group, followed by the VSMC group (p < 0.05 vs. medium) (Fig. 8b), and was lowest in the C-TIMP-3 group (p < 0.05 vs. VSMC) (Fig. 8b), indicating that the medium control group experienced increased cell loss.
Cell apoptosis at 3 days after implantation. (a) Representative light micrographs (200x) illustrating terminal dUTP nick-end labeling (TUNEL) staining of apoptotic cells (arrows) in the border zone of rat hearts after implantation of medium, VSMC, or C-TIMP-3. Normal myocardium (sham) is presented as a control. (b) The number of apoptotic cells was highest in the medium group, followed by the VSMC group (*p < 0.05 vs. medium group), and was lowest in the C-TIMP-3 group (#p < 0.05 vs. VSMC group). Sham, sham-operated controls; VSMC, vascular smooth muscle cells transfected with vector; C-TIMP-3, VSMCs transfected with Timp-3. *p < 0.05 versus medium, #p < 0.05 versus VSMC; n = 5/group (medium, VSMC, C-TIMP-3), n = 3/ group (sham).
Discussion
In a rat model of MI, we demonstrated that inhibiting expression and activity of matrix proteases (MMP-2 and −9) in the infarcted myocardium by cell-based overexpression of TIMP-3 significantly inhibited ventricular remodeling, which is a main contributor to post-MI heart failure, and also reduced functional loss. Although the effect on functional outcome was slightly, but significantly, weaker and mortality was increased when the cells (modified or not) were delivered later after the injury (at 14 days compared to 3 days post-MI), we found that cell-based Timp-3 gene therapy was superior to cell therapy alone even at the 14-day time point.
MMPs contribute to tissue development, morphogenesis, and inflammation (23,29) and are involved in the pathological progression of tissue and organ abnormalities. Thus, after an MI, MMPs have a role in the progressive ventricular dilatation and dysfunction that leads to heart failure. For example, animals with chronic myocardial MMP-1 overexpression experience marked cardiac dysfunction (15), and mice lacking a natural MMP inhibitor (Timp-3-/- mice) have increased end-diastolic volumes and pressures and decreased contractility after an MI (28). LV dilatation and rupture are decreased after myocardial injury in mice that lack MMP-9 (Mmp-9-/-mice) (5). We previously reported that increased MMP activity and decreased TIMP-3 expression are associated with human heart failure (9). Here, we found evidence of enhanced myocardial MMP-2 and −9 protein expression by Western blot after an MI in rats (medium vs. sham groups) and increased MMP-2 and −9 activity by gel zymography in the medium group compared with the VSMC and C-TIMP-3 groups. Despite the fact that in situ zymography was not used for the analysis of MMP activity, which is a limitation of our study, we demonstrated that implantation of smooth muscle cells or cells with the Timp-3 gene significantly decreased MMP-2 and −9 activity and improved the recovery of myocardial function following MI. Indeed, we found that enhanced myocardial MMP activity in untreated animals was associated with LV dilatation and functional loss. But cell therapy or cell-based Timp-3 gene therapy administered within 14 days of the MI rebalanced the MMP/TIMP ratio and mitigated the harmful cardiac effects, providing support for the theory that post-MI ventricular remodeling and cardiac dysfunction are triggered by a myocardial MMP/TIMP imbalance.
In addition, TNF-α, an inflammatory cytokine associated with cardiac remodeling and cell apoptosis, is upregulated in the infarcted myocardium (10) and is known to increase MMP gene expression (26). We quantified the apoptotic cells in the border zone of the infarct heart by TUNEL staining and found that, based on cell shape, most apoptotic cells were cardiomyocytes. These data are in agreement with our previously published data (28). We also quantified TUNEL-stained inflammatory cells and found no significant difference in the number of apoptotic inflammatory cells among the medium, VSMC, and C-TIMP-3 groups (data not shown). These data suggest that one of the mechanisms of the beneficial effect of cell transplantation or cell-based TIMP-3 overexpression was likely the limitation of cardiomyocyte loss. Thus, the enhanced benefits of TIMP-3 overexpression compared to unmodified VSMCs may have involved a TIMP-3-mediated inhibition of TNF-α converting enzyme (TACE) (1,31) and a resultant drop in cardiac TNF-α production, thereby limiting cardiomyocyte loss.
Two experimental models commonly used to study the pathophysiological processes associated with MI are ischemia/reperfusion and permanent coronary artery ligation. While ischemia/reperfusion is the preferred model for assessing therapies and outcomes following reperfusion in an acute MI setting, the permanent ligation model used here is useful for studying cardiac remodeling and outcomes later after an MI (in this case, at 28 days). The pattern of ventricular dilatation and dysfunction manifested in our model (28) is comparable to that seen in patients who progress rapidly to heart failure after an MI, perhaps due to an inadequate TIMP injury response. Cell therapy alone has repeatedly been demonstrated to significantly improve ventricular function after MI in animals (7,11,19,20,33), primarily due to paracrine effects that stimulate angiogenesis, reduce apoptosis, increase bone marrow cell recruitment, and decrease matrix modulation (8). Unfortunately, data from clinical trials indicate that the approach is less effective in patients who have experienced an MI, perhaps because the paracrine effects of implanted cells are diminished in aged patients. We have previously shown, in a mutant mouse model, that modifying the donor cells to overexpress TIMP-3 before implantation contributes transient MMP inhibition that enhances the cells' paracrine effects and boosts the functional benefits of post-MI cell therapy (2). Our current results confirm this pattern in a normal rat model of MI, suggesting that cell-based gene therapy could improve treatment outcomes for aging patients whose responses to implanted cells are diminished.
Another factor affecting clinical outcomes for cell therapy is time of implantation. Cell therapy is normally less effective when it is delivered during the late stages of congestive heart failure rather than soon after a myocardial injury (16). Here, we found that functional restoration was less pronounced when the cell or cell-based gene treatment was delayed (to 14 vs. 3 days after the MI), perhaps because the cells are more effective if delivered during the earlier stages of ventricular dilatation, which increases with time after MI. Also, the first week after an MI is a period of intense inflammatory activity. Since significantly increased MMP-2 and −9 activity at 3 days after cell transplantation has been observed in the myocardium, these proteases could be contributed by myocardial cells as well as by infiltrating cells following injury. In our current study, we found that the number of infiltrating cells was highest in the medium group, followed by the VSMC group, then the C-TIMP-3 group. These data corroborate with the myocardial TNF-α production in the different groups. Therefore, a large part of the MMP activity in the medium, VSMC, and C-TIMP-3 groups could most likely arise from infiltrating inflammatory cells. Thus, gene-transfected cells may have not only directly inhibited MMP-2 and −9 activity, but also suppressed early myocardial inflammation. We expected that overexpressing TIMP-3 in the implanted cells would boost their effects. Indeed, the gene-enhanced cells showed greater preservation of ventricular volumes, wall thickness, and cardiac function than did unmodified VSMCs at both time points. Also, functional outcomes after late implantation of gene-enhanced cells were better than those after early implantation of unmodified cells, and morphometric outcomes were similar in response to cell-based gene therapy delivered at either time point.
We conclude that short-term amplification of TIMP-3 expression via cell-based gene therapy improves healing after an MI in rats; this approach may provide a clinically relevant option for patients with ischemic cardiomyopathy. However, this study did not evaluate angiogenesis, marrow-derived progenitor cell recruitment, or myofibroblast activity in the heart—all of which are believed to be affected by cell or cell-based gene therapy. A more detailed understanding of the numerous mechanisms responsible for the benefits of TIMP-3 overexpression may eventually permit more directed therapeutic interventions.
Footnotes
Acknowledgments
R.K.L. holds a Canada Research Chair in cardiac regeneration and is a Career Investigator of the Heart and Stroke Foundation of Canada. This work was supported by a grant from the Heart and Stroke Foundation of Ontario (T6604) to R.K.L. The authors declare no conflicts of interest.