Abstract
This study evaluated the effects of angiotensin-converting enzyme (ACE) inhibition after myocardial infarction (MI) on cardiac remodeling and gene expression and localization of components (ligands, receptors, and binding proteins) of the cardiac insulinlike growth factor (IGF) system. After ligation of the coronary artery, rats were randomized to vehicle or ACE inhibitor (Captopril, 50 mg/kg/day) for 4 weeks. Blood pressure, cardiac remodeling, and components of the IGF system were localized in the heart using in situ hybridization (ISH) and immunohistochemistry (IHC). The average infarct size was 42%. There were regional differences in the expression of the IGF system after MI, with increased IGF-I mRNA abundance in the border (24-fold) and infarct (12-fold) and increased IGF-binding protein (IGFBP)-3 mRNA in all areas of the failing left ventricle (threefold). Captopril reduced blood pressure, attenuated cardiac remodeling, and caused a threefold increase in IGF-I receptor mRNA and protein in infarct, border zone, and viable myocardium (p<0.01). Captopril had no effect on IGF-I mRNA but further increased IGFBP-3 mRNA and protein in the border zone, (p<0.05). The changes in the cardiac IGF system following MI are highly localized, persist for at least 4 weeks, and can be modulated by ACE inhibition. These data suggest that the benefits of ACE inhibitors in attenuation of cardiac remodeling may be mediated in part through increased expression of the IGF-I receptor sensitizing the myocardium to the positive effects of endogenous IGF-I.
Keywords
M
A number of studies have assessed the role of the cytokine insulin-like growth factor (IGF)-I and growth hormone (GH) on myocyte injury and myocardial function after MI, and both IGF-I and GH have emerged as potential agents in the therapy of CHF (Buerke et al. 1995; Duerr et al. 1995; Jin et al. 1995; Yang et al. 1995; Fazio et al. 1996; Genth-Zotz et al. 1999). GH mediates its effects through IGF-I, which binds to the IGF-I receptor to influence tissue growth and differentiation (Bach and Rechler 1995). In the heart, GH stimulates myofibril development in rat myocytes (Donath et al. 1994) and limits the effect of myocyte stretch to form angiotensin II and cause apoptosis (Leri et al. 1999), whereas high-dose angiotensin II infusion increases ventricular IGF-I expression (Brink and Delafontaine 1995).
Both the renin–angiotensin system (RAS) and the IGF system are activated after MI, with increases in IGF-I mRNA (Dean et al. 1999; Loennechen et al. 2001) coinciding with activation of ACE and angiotensin II (Yamada et al. 1991; Passier et al. 1996; Duncan et al. 1997). Although the interactions between the RAS and the IGF system in the infarcted heart have not been fully elucidated, the actions of IGF-I suggest that it acts as a counterregulatory mechanism to minimize the adverse effects of the RAS. In human and experimental CHF, GH and IGF-I have cardioprotective effects to improve cardiac remodeling and function (Duerr et al. 1995; Jin et al. 1995; Yang et al. 1995; Fazio et al. 1996; Genth-Zotz et al. 1999). In addition, although a lack of increase in circulating levels of IGF-I after acute MI in humans is associated with worsened left ventricular (LV) dysfunction and poorer outcomes (Lee et al. 1999; Conti et al. 2001), cardiac-specific IGF-I overexpression attenuates dilated cardiomyopathy in a transgenic mouse model of CHF (Welch et al. 2002) and prevents myocardial cell death and remodeling in an infarct model of CHF (Li et al. 1997).
Although the data suggest a potential therapeutic role for IGF-I/GH after MI, the effects of standard treatment with ACE inhibition after MI on the cardiac IGF system have not been assessed. Such information is useful in determining the mechanism of benefit of GH/IGF-I treatment, particularly because if it is to enter the therapeutic arena, it will be most likely used as an adjunct to ACE inhibition. The aim of this work was therefore to assess the effect of ACE inhibition with captopril on the cardiac IGF system after coronary artery ligation in the rat. The study included evaluation of cardiac remodeling and gene expression and localization of all components (ligands, receptors, and binding proteins) of the IGF system in the heart. Although most physiological actions of IGF-I and IGF-II are mediated via the IGF-I receptor, this interaction is modulated by at least six structurally related IGF binding proteins (IGFBP-1 to 6) that bind to IGF-I to inhibit or potentiate its activity (Bach and Rechler 1995), which provides a flexible means of regulation of IGF-I and IGF-II activity. Although we have previously demonstrated low-level cardiac expression of IGF-II mRNA (Dean et al. 1999), because IGF-II has positive effects on myocardial function (Battler et al. 1995) we also studied its expression.
Materials and Methods
Experimental procedures were performed according to the National Health and Medical Research Council of Australia Guidelines for Animal Experimentation. Rats were housed at 23–25C in a 12:12 light:dark cycle, with ad libitum food containing 0.4–0.6% NaCl and water.
Experimental Design
LV free-wall MI was induced in female Sprague–Dawley rats (200–250 g) by ligation of the proximal anterior descending artery as described previously (Pfeffer et al. 1985; Burrell et al. 1998,2000). Sham-operated (Control) rats underwent an identical operation but the suture was not tied. Rats that survived for 24 hr postoperatively were randomized to vehicle (once daily gavage of 5% arabic gum) or captopril (25 mg/kg twice daily gavage) for 4 weeks. Body weight and systolic blood pressure (SBP) were measured weekly. SBP was measured by the indirect tail-cuff technique (38L flatbed recorder, model 229 Amplifier; ITCH Life Science, Woodland Hills, CA) in conscious, lightly restrained rats. At 4 weeks, rats were weighed, sacrificed and blood collected for the measurement of atrial natriuretic peptide (ANP) and plasma renin activity (PRA). Lungs were weighed and the LV interventricular septum was dissected, weighed, and fixed in 10% buffered formalin. The LV was sectioned at four levels from the base to the apex, paraffin-embedded, and sections cut and stained with Masson's trichome for assessment of infarct size, and with hematoxylin and eosin. The mean epicardial and endocardial scar circumference was compared to total LV circumference to calculate total infarct size (Pfeffer et al. 1985).
In situ hybridization (ISH) was used to localize mRNA for the IGF system (IGF-I, IGF-II, IGF-I receptor and IGFBPs 1–6) in Control and MI LV and to examine changes in this system after captopril (Dean et al. 1999). Quantitative IHC was used to determine whether changes at the mRNA level were detectable at the protein level.
ISH Histochemistry
ISH was performed on 4-μm sections from four levels of LV from MI (vehicle n=5, captopril n=5) and Control rats (vehicle n=5, captopril n=5) to localize the IGF system mRNA (IGF-I, IGF-II, IGF-I receptor, and IGFBPs 1–6).
Labeling of RNA Probes and Hybridization. Complementary (antisense) or noncomplementary (sense) RNA probes were synthesized for rat IGFBP-1–6 (kindly provided by Dr. S. Shimasaki; Whittier Institute, La Jolla, CA), IGF-II (provided by Dr. P.K. Lund; University of North Carolina, Chapel Hill, NC), and IGF-I receptor and IGF-I (provided by Dr. C. J. Roberts, Jr and Dr. D. Le Roith, NIH, Bethesda, MD) and hybridized as previously described (Dean et al. 1999). After the final posthybridization washes, slides were air-dried and autoradiographed with Kodak XAR (Eastman Kodak; Rochester, NY) film for 1–3 days at room temperature (RT). Slides were dipped in photographic emulsion (Amersham; Poole, UK), stored with desiccant at 4C for 14–21 days, developed in Kodak D19, fixed in Ilford Hypam, and stained with hematoxylin for cellular localization.
Quantitation of Macroscopic Autoradiographs. Quantitation was performed using a microcomputer imaging device (Imaging Research; St. Catharines, Ontario, Canada) run by an IBM PC. Sections from four levels of the LV from each animal were used for quantitation. In MI hearts, the viable myocardium, infarct, and border zone were quantitated separately. The border zone is the area of high cellular infiltrate at the border zone of the fibrotic scar tissue of the infarct. The optical densities of the autoradiographs were calibrated in terms of radioactivity density as dpm/mm2 by reference to radioactive standards (Amersham) carried through the procedures (Le Moine et al. 1994). Results are expressed as specific labeling (using antisense probes) minus nonspecific labeling (using sense probes).
Immunohistochemistry
Immunohistochemistry was carried out on 4-μm sections from four levels of LV from MI (vehicle n=10, captopril n=10) and Control rats (vehicle n=10, captopril n=5). A rabbit anti-mouse IGFBP-3 polyclonal antiserum (GroPep; Adelaide, Australia), which crossreacts with rat IGFBP-3, at a dilution of 1:150 was employed. A biotinylated goat anti-rabbit secondary antibody at a dilution of 1:400 (Vector Laboratories; Burlingame, CA) and the Elite Vectastain ABC kit (Vector Laboratories), followed by diaminobenzidine (Sigma; St Louis, MO) were used to visualize antibody binding. Negative control slides were incubated with normal goat serum; the primary antibody was excluded. For detection of IGF-I receptor protein, a mouse MAb to the IGF-I receptor (β-subunit) (Neomarkers; Union City, CA), which crossreacts with the rat IGF-I receptor, was used at a dilution of 1:100. This antibody was used in conjunction with a Catalyzed Signal Amplification System (DAKO, Carpinteria, CA) and the peroxidase method of labeling according to the manufacturer's instructions.
Quantitation of IHC Staining. Immunohistochemical staining for IGFBP-3 and IGF-I receptor protein was quantitated (n=5-10/group) using computerized image analysis (AIS Imaging; St. Catharines, Ontario, Canada). All sections used for quantitation were fixed, processed, sectioned, and immunolabeled at the same time and under the same conditions to limit variability. Sections from four levels of the LV from each animal were used for quantitation. In MI, the viable myocardium, infarct and border zone were quantitated separately. Twelve fields (X20 objective) from each region of the heart (viable myocardium, infarct, and border) were selected for assessment according to a predefined grid pattern. Images were imported into the AIS imaging program using a color video camera and a standard light microscope. The detection level threshold for positively stained areas (brown for DAB staining) was set so that the processed image accurately reflected the positively stained areas as visualized by light microscopy and on the unprocessed digital image. An average intensity for the selected area was then calculated. The percentage area of chromogen staining was determined by calculating the number of selected pixels (positively stained areas) in a given area and was expressed as a percentage of the entire image. The average intensity and area of staining were then multiplied to give the final figure (arbitrary units) (James and Hauer-Jensen 1999; Rimsza et al. 1999).
Statistics
Results are presented as mean ± SEM. Before analysis of ISH histochemistry quantitation, results were log-transformed to stabilize variance where appropriate. Differences between values for both ISH and IHC were assessed using two-factor ANOVA, one factor being the presence of MI and the other being treatment (vehicle or captopril), followed by post-hoc analysis using the Fisher LSD test.
Results
All sham-operated rats (Control, n=32) survived to 24 hr and were randomized to vehicle (n=17) or captopril (n=15). Of the rats operated on to produce an MI, ∼80% (n=20) were alive at 24 hr and were randomized to vehicle (n=10) or captopril (n=10).
Infarct Size, Body Weight, Blood Pressure, Cardiac Mass, and Hormones
The average infarct size was 42% and was similar in vehicle and captopril-treated MI (Table 1). No Control animal had evidence of cardiac damage. There was no pretreatment difference in body weight or systolic blood pressure between Control and MI rats (not shown). Control rats gained weight throughout the duration of the study with no treatment effect, whereas MI rats treated with captopril gained less weight than vehicle-treated rats (p<0.01). Captopril reduced SBP in both Control and MI (p<0.01). Results after 4 weeks of treatment are shown in Table 1. MI rats had CHF with increased LV and lung mass (p<0.05) compared to Control rats (Table 1). Captopril reduced LV mass in both MI and Control, whereas lung mass was reduced by captopril in MI only (p<0.05). Captopril increased PRA in all rats (p<0.01). Plasma ANP concentrations were elevated in MI rats compared to Control (p<0.01) and were reduced by captopril (p<0.05) (Table 1).
Expression of the Cardiac IGF System After Infarction and Effects of ACE Inhibition
The cellular distribution of components of the IGF system 4 weeks after MI was similar to that previously published 24 weeks after MI (Dean et al. 1999). The expression of most components of the IGF system (i.e., IGF-I, IGF-II, BPs) were similar in Control and viable myocardium of MI rats. In Control rats, ACE inhibition had no effect on the IGF system, whereas there were significant effects on the cardiac IGF system after MI.
Infarct size, blood pressure, tissue weights, and blood hormones
a p<0.05.
b p<0.01 Control vehicle vs MI vehicle.
c p<0.05.
d p<0.01 captopril vs corresponding vehicle group.
Values are expressed as mean ±SEM. MI, myocardial infarction; PRA, plasma renin activity; ANP, atrial natriuretic peptide. LV, left ventricle.
Expression of IGF-I and IGF-II mRNA in Control and MI rats; LV is divided into infarct, border and viable myocardium. White bars are vehicle and black bars captopril-treated. ∗p<0.05, ∗∗p<0.01 MI viable vs. infarct/border; ‡p<0.05; ‡‡p<0.01, border vs infarct.
IGF-I and IGF-II mRNA. As we have shown previously, after MI IGF-I mRNA levels are elevated with 24- and 12-fold increases in the border and infarct zones, respectively, compared to the viable myocardium (p<0.05) and twofold increases in the border compared to infarct (p<0.05) (Figure 1). ACE inhibition had no effect on IGF-I mRNA or on IGF-II mRNA.
IGF-I Receptor mRNA and Protein. IGF-I receptor mRNA levels were similar in the viable myocardium of MI compared to Control rats, and were increased three- to fourfold in all areas of the infarcted heart (viable myocardium, border, infarct) by ACE inhibition (p<0.01; Figure 2A). ACE inhibition also increased IGF-I receptor protein levels (Figure 2B) in viable myocardium (p<0.05), border, and infarct (p<0.01). IGF-I receptor protein was localized to the cell membrane of myocytes in the Control (data not shown) and MI rats (Figures 3A and 3B). An increase in IGF-I receptor protein levels was also seen in the infarct and border zone (Figure 2B; p<0.01), mainly in the endothelium of vessels within the infarct (Figures 3C and 3D) and border (Figures 3E and 3F).
Expression of IGF-I receptor mRNA (A) and protein (B) in Control and MI rats; LV is divided into infarct, border, and viable myocardium. White bars are vehicle and black bars captopril-treated. ∗∗p<0.01 MI viable vs. infarct/border; †p<0.05, ‡p<0.01, captopril vs vehicle.
IGF Binding Proteins. IGFBP-3 mRNA levels were increased 2.4-fold in the viable myocardium of MI compared to Control rats (p<0.05) and were further increased in the border (3.5-fold) and infarct (threefold) (p<0.01) (Figure 4A). ACE inhibition increased IGFBP-3 mRNA by 50% in the infarct and border (p<0.05; Figure 4A) but had no effect in the viable myocardium. This was associated with changes at the protein level, with ACE inhibition increasing IGFBP-3 protein significantly in the border (p<0.05; Figure 4B) and approaching significance in the infarct area (p=0.07; Figure 4B).
IGFBP-3 protein was localized to myocardial cells (Figures 5A and 5B) and to infiltrating cells and vessels of the infarct (Figures 5C and 5D) and border (Figures 5E and 5F), correlating with the distribution of mRNA in these areas. The distribution of IGFBP-3 protein was more widespread in the myocardial cells than that of the mRNA, which may reflect the fact that IGFBP-3 is a secreted protein and is therefore not confined to sites of synthesis (Bach and Rechler 1995).
IHC staining of IGF-I receptor protein (brown) in viable myocardium (
Expression of IGFBP-1 and IGFBP-2 mRNA was minimal in normal and infarcted hearts and was unchanged with ACE inhibition (data not shown). IGFBP-4 mRNA levels were 2.8-fold higher in viable myocardium of MI than Control (p<0.05) (Table 2) and were further increased in the border (6.6-fold) and infarct (4.6-fold) (p<0.01). IGFBP-6 mRNA expression was increased 5.5-fold in the border (p<0.01) (Table 2). Expression of IGFBP-5 mRNA was not significantly different between Control and CHF. There was no effect of ACE inhibition on IGFBP-4, −5, or −6 mRNA levels (Table 2).
Discussion
The main findings of this study are as follows: (a) the changes in the cardiac IGF system after MI are highly localized and persist for at least 4 weeks; (b) ACE inhibition attenuates cardiac remodeling and pulmonary congestion; and (c) ACE inhibition enhances expression of IGF-I receptor gene and protein in the failing heart and causes further increases in IGFBP-3 gene and protein in the border zone. These data suggest a significant interaction between the cardiac IGF system and the RAS in the remodeling heart after MI. The benefits of ACE inhibitors in attenuation cardiac remodeling after MI may be mediated in part through increased expression of the IGF-I receptor sensitizing the myocardium to the positive effects of endogenous IGF-I.
Expression of IGFBP-3 mRNA (
This model of MI-induced CHF causes hemodynamic alterations and neurohormonal changes (Burrell et al. 1998; Pfeffer et al. 1985) similar to those seen in patients with anterior MI and results with this model have clinical implications. All infarcted rats had histological verification of infarct sizes and all showed signs of pulmonary congestion. Our previous studies have shown that CHF in this model is characterized by cardiac remodeling with LV cavity dilatation, hypertrophy of surviving myocardium, and impaired systolic function (Burrell et al. 1996,2000). The success of ACE inhibitors in slowing the deterioration of the failing heart is at least in part due to inhibition of the action of angiotensin II in the heart and attenuation of remodeling (Pfeffer et al. 1985). Both cardiac ACE (Passier et al. 1996; Yamamura et al. 2001) and angiotensin II (Duncan et al. 1997) increase after MI, and treatment with ACE inhibitors attenuates hypertrophy (Burrell et al. 2000) and reduces the cardiac angiotensin II/angiotensin I ratio, indicating effective inhibition of cardiac ACE (Duncan et al. 1996).
The results of this study agree with our previous findings, in that MI increased expression of IGF-I and IGFBP-3, −4, and −6 mRNA in the border and infarct and increased IGFBP-3 and −4 mRNA in the viable myocardium, with no change in the gene expression of IGF-II or the IGF-I receptor (Dean et al. 1999). Over-expression of the IGF system occurs in the same areas of the heart in which there is upregulation of the RAS after MI, i.e., those areas of the myocardium subjected to high mechanical stress. This suggests that the IGF system may be activated as a response to injury by viable myocytes and infiltrating cells adjacent to the infarct to repair potentially viable myocytes at these sites, or to prevent apoptosis induced by an activated RAS (Anversa and Kajstura 1998).
Many studies support a protective and anti-apoptotic role for GH and/or IGF-I in myocardial ischemia and MI (Buerke et al. 1995; Li et al. 1997; Omerovic et al. 2000), and in cardiac myocytes (Wang et al. 1998; Yamamura et al. 2001). IGF-I interferes with the myocyte RAS and reduces secretion of angiotensin II (Leri et al. 1999). Activation of cardiac IGF-I occurs as early as 12–24 hr after MI (Reiss et al. 1994) in a region- and time-specific manner (Dean et al. 1999; Loennechen et al. 2001). In this work, we found that upregulation of IGF-I after MI was not associated with any change in IGF-I receptor levels, which is in accordance with a report that IGF-I receptor mRNA increased for 2–3 days after coronary ligation but had returned to basal values at 7 days (Reiss et al. 1994).
This study assessed whether the cardiac benefits of ACE inhibition may be mediated in part through changes in the cardiac IGF system. As we have shown previously (Burrell et al. 2000), ACE inhibition had the expected effects of lowering blood pressure, increasing PRA, and attenuating cardiac remodeling and pulmonary congestion. A major finding of this study was that ACE inhibition caused a threefold increase in expression of the IGF-I receptor in the failing heart. Because decreased density of IGF-I receptors is associated with increased apoptosis (Resnicoff et al. 1995), enhanced expression of IGF-I receptor mRNA and protein in all areas of the infarcted LV with ACE inhibition may result in the same biological outcome as IGF-I overexpression, i.e., inhibition of cell death. In support of this, ACE inhibition attenuates cardiomyocyte apoptosis in dogs with CHF (Goussev et al. 1998), suggesting that ACE inhibitors may attenuate remodeling through upregulation of the cardiac IGF-I receptor and consequent inhibition of cardiomyocyte apoptosis.
Immunohistochemical staining of IGFBP-3 protein (brown) in viable myocardium (
We found that the increase in IGF-I receptor expression was not associated with any change in expression of the endogenous ligand IGF-I. Gene expression profiling has produced similar results in the same model; captopril increased IGF-I receptor expression in the absence of changes in IGF-I expression (Jin and Wang 2001). Therefore, some elements involved in the pathophysiology of MI-induced heart failure may be unaffected by ACE inhibition and are potentially targets for new therapies.
The IGFBPs, which are potent modulators of the biological actions of IGF-I, were also assessed to more fully understand the role of the local IGF system in CHF. IGFBP-3 mRNA is present in the human heart and is more abundant in ventricles from patients with ischemic heart disease and hypertrophic cardiomyopathy than in controls (Granata et al. 2000). This study demonstrated increased IGFBP-3 mRNA in all areas of the infarcted LV in untreated rats, with increased expression at the gene and protein levels with ACE inhibition in the infarct and border areas. However, the relationship between IGFBP-3 mRNA and protein levels was not as strong as that observed for IGF-I and IGF-IR. In particular, protein levels appeared disproportionately high relative to mRNA levels in control myocardium. Regulation of IGFBP-3 protein levels is not purely transcriptional, so that other mechanisms may explain this apparent discrepancy. For example, IGFBP-3 is subject to specific proteolysis in a number of situations, resulting in decreased affinity for IGFs (Bach and Rechler 1995). It is possible that differences in IGFBP protein levels without changes in mRNA levels are due to differing degrees of proteolysis. Further studies are necessary to address this issue. In particular, the specificity of the antibody used in the present study for intact IGFBP-3 would need to be characterized, followed by specific measurement of IGFBP-3 proteolysis in the heart under different conditions. Therefore, the precise role of cardiac IGFBP-3 is complex, and may enhance or inhibit IGF-I actions. Although the physiological significance of our data remains to be determined, it is possible that increased IGFBP-3 may target IGF-I to viable myocytes.
Effect of ACE inhibition on the expression of IGF binding proteins −4, −5, and −6 in control and MI rats (density of mRNA labeling in dpm/mm2)
aValues are expressed as mean ± SEM. ACEi, angiotensin, converting enzyme inhibition. p<0.05 MI vehicle/viable vs Control Vehicle and p<0.01 MI infarct/border vehicle vs MI vehicle/viable.
Experimental studies have clearly demonstrated the benefits of GH and/or IGF-I after myocardial ischemia or infarction. However, because randomized clinical trials have also established ACE inhibitors as standard therapy following MI (Pfeffer et al. 1992), any new treatment strategy must be considered as an adjunct to ACE inhibition. At present there is a paucity of data comparing and contrasting the effects of ACE inhibition alone and in combination with GH/IGF-I after MI. Data available in experimental CHF show that the combination of GH/IGF-I and an ACE inhibitor improved cardiac function over and above that observed with ACE inhibition alone, suggesting that improvement occurred via independent pathways (Jin et al. 1995). In clinical trials in dilated cardiomyopathy, when most patients will be on an ACE inhibitor, GH supplementation was also associated with a myocardial growth response (Fazio et al. 1996; Osterziel et al. 1998; Genth-Zotz et al. 1999).
Our present study uses a histological approach to allow better localization of mRNA and protein, an approach that is not as sensitive as Northern and Western blotting, but that clearly demonstrates significant, highly localized changes in expression of the cardiac IGF system with MI-induced CHF in the rat that can be modulated by ACE inhibition. The benefits of ACE inhibition in attenuation of cardiac remodeling may be mediated in part through the upregulation of the cardiac IGF-I receptor, sensitizing the myocardium to the positive effects of endogenous IGF-I. These results are of functional significance if a therapeutic strategy is directed at the potentiation of a specific IGF-I/IGF-I receptor interaction. Because we found no change in expression of endogenous cardiac IGF-I with ACE inhibition, this study provides a rationale for the addition of GH and/or IGF-I to ACE inhibition in the management of ischemic heart failure.
Footnotes
Acknowledgements
Supported by the National Heart Foundation of Australia, the Austin Hospital Medical Research Foundation, and the Sir Edward Dunlop Medical Research Foundation.