Reduction of Infarct Size and Prevention of Cardiac Rupture in Transgenic Mice Overexpressing FrzA
Background— FrzA/sFRP-1, a secreted, frizzled-related protein and antagonist for the wnt/frizzled pathway, is expressed in the heart and vessels during mouse embryogenesis and adulthood. FrzA is involved in cell cycle control of vascular cells and angiogenesis. We assessed the hypothesis that FrzA could control the healing process after myocardial infarction (MI).
Methods and Results— We demonstrated an upregulation of sFRP-1 and distinct wnt and fz member expression after MI. We established transgenic (Tg) mice that overexpress FrzA under a cytomegalovirus promoter and developed a model of MI by coronary artery ligation. FrzA reduced cardiac rupture after MI in Tg (6.5% versus 26.4% in controls; n=165, P<0.01). MI was smaller in Tg at each time point (18±10.8% of left ventricular circumference versus 30±14.2% in controls at day 30; P<0.001). Similar results were found in cryolesion-induced MI. Cardiac function was improved in Tg mice (3800±370 mm Hg/s dP/dtmax versus 2800±840 in controls; −2800±440 dP/dtmin versus −1800±211 in controls at day 15; P<0.001). Early leukocyte infiltration had decreased in Tg mice during the first week. Apoptotic index was decreased by 50% in Tg mice at day 7. Matrix metalloproteinase-2 and −9 activity was reduced in Tg mice at day 4, and collagen deposition in the scar was increased in Tg mice. Capillary density in the scar was higher in Tg mice (290±103 vessels/mm2 versus 104±43 in controls at day 15; P<0.001). Vessels were more muscularized, and mean lumen area was 3-fold higher in Tg animals.
Conclusions— Overexpression of FrzA, through direct or indirect interaction with different phases of infarct healing, reduced infarct size and improved cardiac function.
Received January 28, 2003; de novo received April 21, 2003; revision received June 27, 2003; accepted June 30, 2003.
The wnt proteins are secreted glycoproteins that bind to the cell surface or the extracellular matrix and thus, probably act locally in an autocrine or paracrine fashion.1 The frizzled (fz) proteins are 7-pass transmembrane proteins characterized by an extracellular N-terminal, cysteine-rich domain that might constitute part or all of the ligand-binding domain of the wnt proteins.2 Besides being essential for cell fate determination during embryonic development, various wnt and fz family members also influence cell proliferation and the response to activating stimuli. Recently, a family of secreted proteins was described containing a cysteine-rich domain similar to that of fz but lacking the transmembrane domain and was termed “secreted frizzled-related protein” (sFRP).3 sFRPs could compete for wnt binding and antagonize wnt function by interacting with wnt or their corresponding receptors.2,3 A member of this family, FrzA, was originally isolated from bovine aortic endothelial cells and characterized in our laboratory.4 FrzA is related to mouse sFRP-1.3
A growing body of evidence points to the involvement of wnt or fz in cardiac physiology, although their overall role is not known. Interventions in wnt or fz are thought to affect cardiac morphogenesis,5,6 and several members of the wnt/fz signal-transduction pathway have been found to be expressed during cardiac development in vertebrates.7,8 It has been shown that wnt-1 is a potent inducer of connexin43 expression in cardiomyocytes, enhancing the formation of functional gap-junction channels.9 fz2 has enhanced reexpression in overloaded myocardium and is also highly expressed in myofibroblasts during their migration to the area of infarction.10 Moreover, members of the wnt/fz signal-transduction pathway have been involved in cardiac hypertrophy and myocardial repair after myocardial infarction (MI).11,12 sFRP-3 and −4 levels were found to be elevated in the hearts of patients with dilated cardiomyopathy and coronary heart disease.13 Our laboratory previously detected high FrzA levels during cardiovascular maturation and in the cardiovascular system during adulthood.4,8 FrzA proved capable of regulating vascular cell proliferation4 and induced angiogenic response.14
Because of evidence of the potential involvement of wnt/fz in cardiovascular physiology described, we hypothesized that enhanced expression of FrzA might play a role in infarct healing after MI. The aim of the study was, therefore, to examine the disruption of the wnt/fz receptor system in transgenic mice (Tg) overexpressing FrzA as a regulator of the healing process after MI.
Construction of Tg Overexpressing FrzA Mice and Characterization
The FrzA transgene, under a cytomegalovirus (CMV) promoter and fused into the C-terminus with a myc:his epitope in the pcDNA3 expression vector (Invitrogen), was excised from the plasmid backbone.4 Microinjection into fertilized (C57BL/6J×CBA) F2 oocytes and other surgical procedures were performed as described earlier15 within the transgenic core facility at Bordeaux 2 University. All experiments were performed with FrzA transgenic (Tg) mice hemizygous for the transgene and backcrossed onto the C57BL/6J strain for >6 generations (Charles River, Lyon, France). C57BL/6J mice were used as controls.
Analysis of wnt, fz, and sFRP-1 RNA by Reverse Transcription–Polymerase Chain Reaction
Total RNAs were prepared from uninjured and infarcted hearts in guanidinium thiocyanate buffer, and reverse transcription–polymerase chain reactions (RT-PCRs) were carried out as previously described.4 Negative controls without RT were prepared in parallel for each RNA sample. Semiquantitative PCR was performed as previously described by our group,14 and the results were standardized with respect to those obtained with 28S rRNA, which was coamplified.
Male Tg and control C57BL/6J mice, 10 to 12 weeks old, were used for experimental protocols. All experiments were performed in accordance with the Principles of Laboratory Animal Care. The mouse model of MI was based on that previously described.16 Mice were anesthetized with a mixture of ketamine (2.5 mL/kg) and xylazine (0.8 mL/kg) by intraperitoneal injection. After intubation, a suture was placed around the left anterior descending coronary artery near the origin of the pulmonary artery.
To demonstrate the independence of the surgical process or mouse strain, a second model was established by a freeze-thaw injury procedure.17 On the free wall of the left ventricle (LV), cryoinjury was performed with the use of a metallic probe cooled to −190°C by immersion in LN2. It was applied twice for 20 seconds to ensure transmural necrosis. These mice were used to determine infarct size only. Control and Tg mice underwent surgery at different time points throughout the year. During the same experimental day, mice from the 2 groups were operated on simultaneously by the same investigator (L.B.), who was blinded to group assignment, and were randomly assigned to 1-day euthanization.
Mice were anesthetized as described earlier on day 15 (D15). The right carotid artery was cannulated with a 1.4F microtipped transducer (Millar). Aortic blood pressure was recorded, and the catheter was pushed into the LV; hemodynamic parameters with or without dobutamine infusion (60 μg IP) were recorded in a closed-chest preparation.
For morphometric studies, the heart was arrested in diastole by KCl injection and was fixed for 20 minutes under pressure (diastolic arterial pressure) in 4% paraformaldehyde. The heart was excised, weighed, and stored for 3 hours in 4% paraformaldehyde at 4°C. For immunochemistry procedures, hearts were fixed in 100% methanol before paraffin embedding. One day before the mice were humanely killed, they received bromodeoxyuridine (30 mg/kg body weight IP). Infarct size was determined by planimetry on Masson’s trichrome–stained slides, as previously described.18 Percentage of collagen was determined in the scar with picrosirius red stain, and the ratio of positive Sirius red–stained area to LV area was calculated.
Immunohistochemistry was performed as previously described.19 To detect FrzA transgene expression, sections were stained with a tag polyclonal c-myc antibody (Euromedex). To detect endogenous sFRP1 in control hearts, sections were stained with a polyclonal antibody against FrzA, as previously described.8 The following antibodies were used to detect the various components of interest: myeloperoxidases (Myeloperoxydases, Dako), T lymphocytes (CD-3, Serotec), macrophages (F4/80, Caltag), endothelial cells (CD-31, Pharmingen), myofibroblast cells and smooth muscle cell (α-actin, Sigma), bromodeoxyuridine (BrdU, Harlan), and β-catenin (C-2206, Sigma). To detect apoptosis in scar tissue, terminal deoxynucleotidyl transferase–mediated dUTP nick end-labeling staining was used as previously described.19
After MI, C57BL/6J and Tg mouse hearts were harvested, and the scar and septum were separated and frozen to study matrix metalloproteinase (MMP)-2 and −9 activity. Tissue samples were homogenized in 500 μL buffer (50 mmol/L Tris, 0.2% Triton X-100, 10 mmol/L CaCl2, and 2 mol/L guanidine hydrochloride, pH 7.5). Samples were applied to nondenaturing 10% polyacrylamide gels containing 1 mg/mL gelatin. After electrophoresis, the gels were washed with 2.5% Triton X-100, incubated overnight at 37°C in zymography buffer (50 mmol/L Tris and 10 mmol/L CaCl2), and stained with Coomassie Brilliant Blue. Gelatinolytic activity was visualized as clear areas of lysis in the gel.
Cytoplasmic β-Catenin Western Blotting
Cytoplasmic extracts were isolated by a slight modification of a previous method.20 In brief, fresh tissues from noninjured heart or from infarcted areas were homogenized in buffer (250 mmol/L sucrose, 1 mmol/L EGTA, and 20 mmol/L HEPES, pH 7.4, plus a protease inhibitor) at 4°C. After a low-speed spin (700g for 10 minutes at 4°C), the supernatant was collected and then centrifuged at 100 000g for 30 minutes at 4°C. The clear supernatant was saved as the cytoplasmic protein fraction and run for Western blotting with the anti–β-catenin antibody (Sigma), as previously described,14 and normalized to α-tubulin.
Percentage of infarcted LV was calculated as the ratio of infarcted area to infarcted plus noninfarcted LV area. For capillary density, vessel muscularization, proliferation, and apoptosis, a minimum of 30 randomized pictures were recorded and then counted at 40× magnification for each animal (n≥5) at each time point in the scar area with a CCD camera connected to a personal computer (IBM). Capillary density was reported as the number of capillaries per square millimeter. The percentage of muscularized arteries was calculated as the ratio of α-actin–positive vessels to CD-31–positive vessels (total vessel number). Apoptosis index was calculated as the ratio of positive apoptotic cells to the total cell number. Proliferation index was calculated as the ratio of bromodeoxyuridine-positive cells to total cell number.
Results were expressed as mean±SD. Comparisons of continuous variables between 2 groups were performed by a 1-way ANOVA and subsequently, when statistical significance was observed, by a 2-sided, paired t test (StatView 5-1, Abacus). A value of P<0.05 was considered significant.
Tg Mice Characterization
To further examine the in vivo potential of FrzA to control myocardial repair after MI, we established 3 transgenic lines, 11, 73, and 83, with the pCMV-FrzA-myc vector portion containing the CMV-directed, full-length COOH-terminus, Myc-tagged FrzA cDNA. In the selected founder 73 (Tg), abundant FrzA mRNA expression was detected in the heart and skeletal muscle, as described. Hearts were immunostained with anti-myc antibody, revealing FrzA protein expression in myocytes from Tg mouse hearts from line 73 but not from controls C57Bl/6J (not shown). This line was viable and fertile and did not present any differences in heart morphology by light microscopy.
sFRP-1, wnt, and fz Expression After MI
We investigated the expression of FrzA, wnt, and fz in a mouse model of MI. We first delineated sFRP-1 expression (endogenous expression) after coronary artery ligation in control C57BL/6J mice. By RT-PCR analysis, sFRP-1 mRNA was detectable at low levels in nonischemic myocardium; by D7, abundant sFRP-1 mRNA was detected in the infarcted area (Figure 1A). By D15, expression had returned to near-baseline levels (not shown). Endogenous expression of sFRP-1 was not modified or downregulated by overexpression of the FrzA transgene (not shown). Immunohistochemistry showed sFRP-1 protein expression to be predominant at the border ischemic zone but was less expressed in the scar itself (Figure 1B). Comparable expression was obtained in samples of human MI, with increased expression of SARP2 (the sFRP-1 human homologue) in the ischemic areas (Figure 1B).
Semiquantitative RT-PCR analysis was performed on myocardial extracts to examine the expression patterns of wnt and fz genes before (D0) and after (D7) infarction in control and Tg mice (Figure 1C). Among the different wnt members, 1, 3, 3A, 4, 8B, 10A, and 11 were not detected. wnt10B was upregulated after infarction, although wnt7B was downregulated to undetectable levels in ischemic versus nonischemic tissues in both control and Tg mice. Interestingly, some of them were differentially expressed in Tg compared with control mouse tissues. We detected little or no expression of wnt7A and 7B in the heart at D0 in Tg tissues; wnt8A was upregulated in Tg mice. Among the 9 fz members examined, some showed stable expression, whereas fz1, 2, 5, and 10 were induced during ischemia in both control and Tg mice. Again, we observed that some of the fz genes were differentially regulated in control versus Tg mice. fz6 and 7 expression levels were increased in nonischemic and ischemic conditions in Tg compared with control mice, whereas fz8 was downregulated in ischemic conditions in control mice compared with Tg mice (Figure 1C). Because MI involved the upregulation of sFRP-1/FrzA and distinct wnt and fz members, it was tempting to consider that sFRP1/FrzA antagonism of the wnt/fz cascade was involved in infarct healing.
FrzA Prevented Cardiac Rupture After MI
After MI, there was no difference in early mortality (within 24 hours after surgery) between the 2 groups. Between D 4 and D5 after MI, 26.4% of control mice (n=87) suffered fatal cardiac rupture of the LV free wall, as confirmed by necropsy. In contrast, only 6.5% of Tg mice (n=61) died by this mechanism (P<0.01). This difference in mortality was not due at this time point to differences in LV pressures between the 2 groups (P=NS; data not shown).
FrzA Reduced Infarct Size and Improved Cardiac Function After MI
Before and after MI, control and Tg hearts displayed no difference in weight, septal and free wall thickness, and LV dimensions (Table 1). At D2, D15, and D30 after MI, the percentage of infarcted LV was statistically decreased and the thickness of the scar was significantly increased in Tg mice compared with controls (n≥5, P<0.001; Figure 2A and 2B). In the infarction model induced by cryolesion, differences in the percentage of infarcted LV and thickness of the scar between Tg and controls were confirmed by morphometric analysis at all time points (Figure 2A and 2B).
This reduction in infarct size in Tg mice improved cardiac function. Before MI, hemodynamic parameters were comparable in the 2 groups of mice (Table 2). As shown in Figure 3, on D15 after MI, LV end-diastolic pressure was significantly decreased, dP/dtmax was significantly increased, and dP/dtmin was significantly decreased in Tg compared with control mice (despite the use of cardiodepressant drugs). Differences in dP/dtmax and dP/dtmin were confirmed and amplified after dobutamine infusion.
FrzA Reduced Early Leukocyte Infiltration and Apoptosis in the Scar and Improved Cellularity
At D2 and D7 after MI, Tg mice displayed significantly fewer myeloperoxidase-positive cells in the scar compared with control mice (Table 3). Macrophages and lymphocytes counts in the scar were not different between the 2 groups. At baseline, bone marrow and serum counts for neutrophils were comparable between Tg and control mice (not shown).
At D7, the apoptotic index was significantly decreased by 50% in Tg mice compared with control mice (P<0.001; Table 3). The index of proliferation was lower in Tg mice at D7. Cell density was significantly higher in the scar of Tg mice on D15 after MI compared with control mice (P<0.001). At D7 and D15 in Tg mice, the cellular population was composed essentially of myofibroblasts, whichdisplayed a more concentric organization compared with myofibroblasts in control mouse scars (Figure 4 and Table 3).
FrzA Increased Collagen Deposition and Decreased Metalloproteinase Activity
Collagen density was significantly increased in the scars of Tg mice compared with controls at D15 (53±2.3% of the scar area was Sirius red–positive in Tg versus 26±6% in controls; P<0.001; Figure 4). No difference in collagen density or myofibroblast organization was reported at remote areas at D15 or D30.
The MMP study showed a large decrease in MMP-9 activity and a moderate decrease in MMP-2 activity in the scar at D4 in Tg mice compared with controls. At D7, MMP-2 activity was similar in both groups, although we constantly found a moderate increase in MMP-9 activity in Tg animals. Immunostaining showed a decrease in MMP-9–positive cells in Tg mice compared with controls at D4, overlapping myeloperoxidase staining (not shown).
FrzA Controlled the Accumulation of β-Catenin
MI was accompanied by an accumulation of cytosolic β-catenin. This accumulation was significantly decreased by FrzA overexpression (Figure 5A). In control and Tg mouse uninfarcted hearts, β-catenin was expressed in the intercalated disks between the uninjured cardiomyocytes but not in the endothelial cells of capillaries or large vessels (not shown). After MI, many endothelial and smooth muscle cells of the infarcted area stained positively for β-catenin in control mice, but few stained positively in Tg mice (Figure 5A and 5B).
FrzA Improved Angiogenesis in the Scar
Coronary distribution, capillary density, and area of vessel muscularization were similar in normal uninjured hearts in both groups. At D7 after MI, capillary density in the scar was higher in control mice compared with Tg mice. However, by D15, the capillary density had decreased in control mice but remained constant in Tg mice between D7 and D15 (Figure 5C). The percentage of muscularized vessels was significantly higher in Tg than in control mice at D7 and D15 after MI (Figure 5C). Moreover, the lumen area of capillaries was 3-fold greater in Tg mice at D7 and D15 compared with controls (Figure 5C).
The aim of our study was to characterize the role of FrzA in myocardial healing after MI in the light of knowledge regarding FrzA expression and function. FrzA/sFRP-1 is detected at high levels during mouse cardiovascular embryogenesis and in adult heart and vessels,4,8 and we have demonstrated its role in vascular cell proliferation and angiogenesis.4,14
In this study of a mouse model of MI, we report that sFRP-1 expression increased after MI at the border zone. We therefore identified the expression and regulation of numerous wnt and fz members in the normal and/or ischemic heart. sFRPs, which are secreted proteins, are thought to compete for wnt binding and to antagonize wnt function by interacting with wnt or their corresponding fz receptors.21 The exact target of FrzA is not known, but it might antagonize different wnt. Besides the presence of FrzA potential targets in ischemic/infarcted myocardium, we showed a cytosolic β-catenin accumulation after MI (wnt pathway activation), corresponding to β-catenin modulation in endothelial cells, as previously reported.12 However, FrzA overexpression decreased the cytosolic accumulation of β-catenin, and the Tg mouse heart displayed fewer β-catenin–positive vascular cells than did control mouse hearts. Therefore, the wnt/fz pathway is activated after MI, so blocking this pathway by overexpressing FrzA was physiologically relevant. The cardiac phenotype of the Tg mouse has been found to be normal, but after MI, infarct size and cardiac rupture were reduced in comparison with control mice, and cardiac function was preserved.
The reduced apoptosis seen in Tg mice probably played a role in limiting the scar area and attenuating hemodynamic degradation.22 The role of the wnt/fz pathway in programmed cell death is a matter of debate. sFRPs were first related to apoptosis, thus making cells more prone to apoptosis under apoptotic conditions.21 However, in recent reports, it was shown that different components of the fz cascade could induce apoptosis and that this effect was dependent on cell type.23,24 The protective effect of FrzA against cell death in transgenic mice could also be indirect, through limitation of inflammatory cell infiltration. In transgenic mice overexpressing FrzA, myeloperoxidase-positive cell infiltration in infarcted areas (mainly polymorphonuclear cells) was significantly reduced in the first week, but T-lymphocyte and macrophage infiltrations were not different between Tg mice and controls. It is now admitted that granulocytes adhere and migrate in response to release of several cytokines, like interleukin (IL)-6 and IL-8.25 Sen et al26 have recently shown that transfection of normal synovial fibroblasts with wnt5A increases IL-6, IL-8, and IL-15 mRNA expression. Therefore, it is conceivable that some members of the wnt/fz family modulate the inflammatory process, and future experiments will aim to investigate whether FrzA could influence the production of cytokines by cardiac cells through impairment of the wnt/fz pathway.
FrzA overexpression was found to lead to a decrease in MMP-9 activity in the infarcted area. This decrease was mainly related to the decrease in granulocyte infiltration (not shown) and was accompanied by a decrease in cardiac rupture, as suggested previously,27 and a reduction in infarct wall thinning.28 In parallel, abundant myofibroblasts and collagen deposition were found in organized arrays in the Tg heart, in alignment with the epicardium and endocardium (see Figure 4). Wnt/fz interaction has been reported to be essential in architectural and polarity control in Drosophila29 through stabilization of cytosolic β-catenin and interaction with cadherin linking the cytoskeleton to the extracellular environment. In summary, FrzA could contribute to preserving the contractile properties of the heart after MI through its action on myofibroblast recruitment and organization and on collagen content and deposition.
However, MMP inhibition induced by FrzA did not impair beneficial neovascularization after MI, as suggested in a recent demonstration in rabbit MI.28 Indeed, overexpression of FrzA led to an increase in capillary density, lumen area, and muscularization in the scar at D15 compared with controls. In fact, FrzA appeared to preserve already formed capillaries at D7 more than to increase vessel formation. In addition to studies on FrzA,14 there is increasing evidence from the literature that the wnt/fz pathway is involved in the formation of new blood vessels.6,12
Overexpression of FrzA interacted directly or indirectly with the different phases of infarct healing, apoptosis, granulocyte infiltration, blood vessel formation, myofibroblast invasion, and granulation tissue formation and destruction. The exact mechanism of FrzA in some of these steps needs further investigation. Because of diffuse expression of the transgene, FrzA might have extracardiac effects. However, the present findings suggest that infarct healing and scar structure, the major determinants of heart function, are improved when FrzA is overexpressed.
This work was supported by a grant from the Association pour la Recherche contre le Cancer and from the Fédération Française de Cardiologie, Paris, France. We thank Dr Dubosq-Marcheney (hematologist) for helpful mouse blood and bone marrow analyses.
Hsieh J-C, Rattner A, Smallwood PM, et al. Biochemical characterization of Wnt-Frizzled interactions using a soluble, biologically active vertebrate Wnt protein. Proc Natl Acad Sci U S A. 1999; 96: 3546–3551.
Rattner A, Hsieh JC, Smallwood PM, et al. A family of secreted proteins contains homology to the cysteine-rich ligand binding domain of frizzled receptors. Proc Natl Acad Sci U S A. 1997; 94: 2859–2863.
Duplàa C, Jaspard B, Moreau C, et al. Identification and cloning of a secreted protein related to the cysteine-rich domain of frizzled: evidence for a role in endothelial cell growth control. Circ Res. 1999; 84: 1433–1445.
Ishikawa T, Tamai Y, Zorn AM, et al. Mouse Wnt receptor gene Fzd5 is essential for yolk sac and placental angiogenesis. Development. 2001; 128: 25–33.
Antos CL, McKinsey TA, Frey N, et al. Activated glycogen synthase-3β suppresses cardiac hypertrophy in vivo. Proc Natl Acad Sci U S A. 2002; 99: 907–912.
Blankesteijn WM, van Gijn ME, Essers-Janssen YP, et al. β-Catenin, an inducer of uncontrolled cell proliferation and migration in malignancies, is localized in the cytoplasm of vascular endothelium during neovascularization after myocardial infarction. Am J Pathol. 2000; 157: 877–883.
Schumann H, Holtz J, Zerkowski HR, et al. Expression of secreted frizzled related proteins 3 and 4 in human ventricular myocardium correlates with apoptosis related gene expression. Cardiovasc Res. 2000; 45: 720–728.
Dufourcq P, Couffinhal T, Ezan J, et al. FrzA, a secreted frizzled related protein, induced angiogenic response. Circulation. 2002; 106: 3097–3103.
Hofmann SL, Russell DW, Brown MS, et al. Overexpression of low density lipoprotein (LDL) receptor eliminates LDL from plasma in transgenic mice. Science. 1988; 239: 1277–1281.
Pfeffer MA, Pfeffer JM, Steinberg C, et al. Survival after an experimental myocardial infarction: beneficial effects of long-term therapy with captopril. Circulation. 1985; 72: 406–412.
Sellin JH, Umar S, Xiao J, et al. Increase β-catenin expression and nuclear translocation accompany cellular hyperproliferation in vivo. Cancer Res. 2001; 61: 2899–2906.
Melkonyan HS, Chang WC, Shapiro JP, et al. SARPs: a family of secreted apoptosis-related proteins. Proc Natl Acad Sci U S A. 1997; 94: 13636–13641.
Frangogiannis NG, Smith CW, Entman ML. The inflammatory response in myocardial infarction. Cardiovasc Res. 2002; 53: 31–47.
Sen M, Lauterbach K, El-Gabalawy H, et al. Expression and function of wingless and frizzled homologs in rheumatoid arthritis. Proc Natl Acad Sci U S A. 2000; 97: 2791–2796.
Lindsey ML, Gannon J, Aikawa M, et al. Selective matrix metalloproteinase inhibition reduces left ventricular remodeling but does not inhibit angiogenesis after myocardial infarction. Circulation. 2002; 105: 753–758.