Determinants of the Cardiomyopathic Phenotype in Chimeric Mice Overexpressing Cardiac Gsα
Abstract—Mice with overexpressed cardiac Gsα develop cardiomyopathy, characterized by myocyte hypertrophy and extensive myocardial fibrosis. The cardiomyopathy likely involves chronically enhanced β-adrenergic signaling, because it can be blocked with long-term propranolol treatment. It remains unknown whether the genotype of the myocyte is solely responsible for the progressive pathological changes. A chimeric population in the heart should answer this question. Accordingly, we developed a chimeric animal, which combined cells from a transgenic overexpressed Gsα parent and a Rosa mouse containing the LacZ reporter gene, facilitating identification of the non–Gsα cells, which express a blue color with exposure to β-galactosidase. We studied these animals at 14 to 17 months of age (when cardiomyopathy should have been present), with the proportion of Gsα cells in the myocardium ranging from 5% to 88%. β-Galactosidase staining of the hearts demonstrated Gsα and Rosa cells, exhibiting a mosaic pattern. The fibrosis and hypertrophy, characteristic of the cardiomyopathy, were not distributed randomly. There was a direct correlation (r=0.85) between the extent of myocyte hypertrophy (determined by computer imaging) and the quantity of Gsα cells. The fibrosis, determined by picric acid Sirius red, was also more prominent in areas with the greatest Gsα cell density, with a correlation of r=0.88. Thus, the overexpressed Gsα can exert its action over the life of the animal, resulting in a local picture of cardiomyopathic damage in discrete regions of the heart, where clusters of the overexpressed Gsα cells reside, sparing the clusters of normal cells derived from the normal Rosa parent.
With the advent of molecular biology and genetic transformation of animal models and with the potential for gene therapy in patients, an important question arises, ie, does the resulting phenotype reflect precisely the altered genotype, or have other compensatory changes modified the original phenotype? For example, the mouse with overexpressed cardiac Gsα responds to sympathetic stimulation with enhanced β-adrenergic signal transduction as a young adult.1 2 3 However, over the life of this animal, β-adrenergic receptor desensitization mechanisms are ineffective in the presence of overexpressed Gsα,4 and the prolonged exposure to enhanced β-adrenergic receptor signaling exerts a toll, resulting in a picture of cardiomyopathy.2 5 This cardiomyopathy is reflected by depressed left ventricular (LV) function with a dilated heart, and the architecture of the heart is characterized by extensive interstitial fibrosis and hypertrophy of myocytes. The question is posed again more specifically: Is the cardiomyopathy observed in older mice with overexpressed Gsα due to chronically enhanced β-adrenergic signaling, resulting in reduced subendocardial coronary reserve, an imbalance between myocardial oxygen demand and supply, with consequent myocardial ischemia and consequent myocyte necrosis and apoptosis? Or, does the overexpressed Gsα exert an effect locally or in the microenvironment that leads to hypertrophy, myocyte necrosis, and fibrosis?
One approach to address this question is to study a chimeric animal with a heart composed of both normal myocytes and myocytes overexpressing cardiac Gsα. Our goal was to study these chimeras, when the cardiomyopathy should have been fully manifest (14 to 17 months of age). By using a Rosa parent, whose cells are tagged with the LacZ reporter gene, the origin of the cells, ie, from the normal Rosa or Gsα parent, could be identified by exposing the cells to β-galactosidase, which results in the Rosa cells expressing a blue color. The major goal was to quantitate the extent of fibrosis and the extent of myocyte hypertrophy in cells of overexpressed Gsα origin and to compare this with normal Rosa cells, to determine whether the characteristic features of cardiomyopathy, fibrosis and hypertrophy, exhibit a predilection for the overexpressed Gsα cells or are expressed in areas of normal Rosa cells as well.
Materials and Methods
Chimeric mice were generated by the investigators by the fusion of morulae from which the zona pellucida had been removed by pronase treatment. The morulae from a line of transgenic mice expressing the LacZ gene controlled by a promoter, which directs expression to essentially all tissues (Rosa 26, Jackson Laboratories), on a Balb C background with an albino phenotype were fused with the morulae from the Gsα transgenic mice on a C57/SJL background with an agouti phenotype. Of 386 fusions, 338 fusions were successful. The fused embryos were transferred to 42 pseudopregnant female recipients. Of these, 7 chimeric offspring originating from different litters survived 14 to 17 months so that they could be studied at a time when the cardiomyopathy should have been manifest. The numbers of animals studied included 7 chimeric, 3 Rosa controls, and 6 Gsα wild-type controls. Animals used in this study were maintained in accordance with the Guide for the Care and Use of Laboratory Animals (DHHS publication No. [NIH] 83-23, revised 1996).
Heart rate was measured using telemetry techniques in the conscious, unrestrained state6 in 7 chimeric mice and 6 age-matched Gsα wild-type controls. Arterial pressure was measured in 6 chimeric mice using a 1.4F micromonometer catheter (Millar Instruments). Echocardiographic assessment was possible in 6 of 7 chimeric mice. One animal did not tolerate the anesthesia. After determination of body weight, mice were anesthetized with ketamine (0.065 mg/g), acepromazine (0.002 mg/g), and xylazine (0.013 mg/g), which was injected intraperitoneally. The procedure for echocardiography (Apogee X-200; Interspec, Inc) has been described in previous reports from our laboratory.2 5
Tissue Preparation and Histochemical Examination
After the echocardiography, the atria and great vessels were removed, and the combined left and right ventricles were weighed. Cryosections cut at 10 μm were incubated overnight at 37°C in β-galactosidase, which results in expression of a blue color in the cells from the Rosa parent but does not affect the color of Gsα cells. One entire cross section from each mouse was examined to determine the fraction of Gsα versus Rosa cells. Cryosections stained with β-galactosidase were subsequently stained with picric acid Sirius red for collagen. Picric acid Sirius red–stained collagen, highlighted by the Metamorph system, was expressed as volume percent collagen for each region, which was classified according to the percentage of Gsα cells. On average, 7 areas were analyzed from each mouse. Staining was adequate in 6 of 7 chimeric animals. Myocyte cross-sectional area was measured on 10-μm-thick cryosections stained in β-galactosidase solution overnight followed by laminin immunostaining (Sigma) to outline the basement membrane of cardiac myocytes. This procedure used a biotinylated secondary antibody and ExtrAvidin peroxidase for colorization. Individual regions within the ventricular wall were analyzed by parental origin based on β-galactosidase staining and classified according to the percentage of Gsα cells. On average, 7 (range 5 to 9) areas were analyzed from each mouse. Staining was adequate in 6 of 7 animals.
The myocyte cross-sectional area and the volume percent of collagen were correlated using linear regression analysis versus the percentage of Gsα cells in the chimeric animals. All data are expressed as mean±SEM. All statistical data were analyzed with ANOVA using a computer with appropriate software (StatView).
In conscious, unrestrained mice, we observed no differences in heart rates between the chimeric mice (624±24 bpm, n=7) and wild-type mice (611±18 bpm, n=6) (Figure 1⇓).
Echocardiography (n=6) demonstrated that fractional shortening was significantly less in the chimeric mice (23±2%, n=7) compared with the control Gsα wild-type littermates (32±1%; n=6) (Figure 1⇑). One control Rosa mouse was studied and found to have normal LV fractional shortening (38.1%). Other than fractional shortening, which was reduced, and arterial pressure, which was elevated, there were no differences in hemodynamics or LV mass between the chimerics and controls (Table⇓).
The chimeric mice exhibited a variegated coat color, reflecting the expression of the different parental genotypes. The myocytes in the hearts also reflected the origin of both parents (Figure 2⇓), with the blue-stained cells from the normal Rosa parent and the unstained cells reflecting the overexpressed Gsα parent. Note that the two different cell types were not distributed randomly but rather were clustered (Figure 2⇓), as might be predicted from the techniques used to produce the chimeras. Histopathological evaluations at higher power revealed a pattern of larger cells and enhanced interstitial fibrosis clustered around the Gsα cells (Figure 2⇓). When this was quantitated for the 6 chimeric animals, there was a direct correlation between the extent of myocyte hypertrophy and the percentage of Gsα cells for the cells overexpressing Gsα (r=0.85) (Figure 3⇓). There was no correlation between Rosa myocyte size and percentage of Gsα cells (r=0.32). The y-intercept for the cell size in the Rosa chimerics (152 μm2) was not different from the average Gsα wild-type control (154±2 μm2) or the Rosa control (155±7 μm2). The extent of fibrosis in the 6 chimerics was correlated to the density of Gsα cells (r=0.88) (Figure 4⇓). The y-intercept for the volume percent collagen in the chimerics (2.6%) was not different from the Gsα wild-type controls (2.4±0.2%) or the Rosa controls (2.3±0.4%). Body weights and LV/body weights were not different between groups.
To determine the extent of overexpression of Gsα in the chimeric heart, we performed immunoblotting using total protein extracts from the chimeric, Gsα, Rosa, and wild-type control mice. As expected from the histology, there was a graded increase in cardiac Gsα from wild-type controls to the chimeras to the pure, overexpressed Gsα hearts. The Gsα hearts showed a 3- to 5-fold overexpression, as demonstrated previously,4 whereas the chimeric demonstrated an intermediate amount of cardiac Gsα.
In the present study, the key features of cardiomyopathy, interstitial fibrosis and myocyte hypertrophy, were evident in these unique chimeric transgenic animals. We asked the question: Is the resultant phenotype of the altered genotype solely the consequence of the altered gene, or are there other effects in vivo that alter the phenotype secondarily?
It is possible that the enhanced β-adrenergic signaling with increased heart rate, LV wall stress, and contraction in the face of limited coronary reserve is, in part, the cause of hypertrophy and cardiomyopathy in the overexpressed Gsα mouse.2 5 If this is the mechanism, then all myocytes in the chimeric heart would be hypertrophied. In contrast, the fibrosis and hypertrophy were limited to the Gsα cells. This suggests that the adverse effects of overexpressed Gsα are expressed locally and are not due to the secondary changes of enhanced global LV contractility, wall stress, tachycardia, and reduced coronary reserve in the subendocardium. Both the extent of hypertrophy and fibrosis increased proportionally to the concentration of Gsα cells. This indicates that simple expression of the gene was not sufficient to induce the full cardiomyopathic phenotype, because the hypertrophy and fibrosis were not evenly distributed among the Gsα cells, regardless of their location. In areas of low concentration of Gsα cells, little fibrosis and hypertrophy were observed. Rather, it was in areas of larger concentrations of clustered Gsα cells where hypertrophy and/or fibrosis were most prominent, suggesting that local areas of enhanced contraction or neurohormonal signaling were responsible for the cardiomyopathic phenotype. These observations, in combination with a recent study from our laboratory demonstrating that chronic propranolol rescued the cardiomyopathic phenotype, implicate β-adrenergic signaling.7 Thus, the effects of locally amplified β-adrenergic signaling are sufficient to induce phenotypic characteristics of cardiomyopathy. Moreover, these changes are due to the chronically enhanced contraction or to the altered microenvironment created in the Gsα clusters, and this effect is proportional to the numbers of cells that exhibit enhanced β-adrenergic signaling.
Another feature of the overexpressed Gsα model is chronic tachycardia.6 Interestingly, the tachycardia in the chimeric transgenic mice was attenuated, ie, the heart rates were not different, in the chimeras compared with the wild-type controls. Why the heart rates were not different is not clear. Given that the Gsα and Rosa cells were not distributed homogeneously, it is possible that the sinoatrial nodal environment is not influenced in a linear fashion by the overall percentage of Gsα cells in the entire heart. However, the lack of relationship between heart rate and fibrosis or hypertrophy suggests that the mechanism of hypertrophy and interstitial fibrosis in this model is not exclusively due to the more rapid rate of contraction that is characteristic of Gsα mice. Even if heart rate had been significantly elevated, a heart rate mechanism would not be responsible for the hypertrophy and fibrosis, because all cells in the chimeric heart experience the same frequency of contraction. It is conceivable that the enhanced β-adrenergic signaling increases the local force of contraction in isolated parts of the heart containing the overexpressed Gsα cells. However, it is equally plausible that other signaling pathways that play a key role in hypertrophy and cell death are responsible for the pathological changes, independent from global or local hemodynamics. In either event, it is important that the altered genotype is responsible for the altered phenotype on a local level. It is interesting to speculate that the genotypic alteration may be the reason that this Gsα model does not desensitize and consequently permits the development of myocyte hypertrophy and cardiomyopathy.4 These results also have implications for understanding the effects of β-blockade therapy in heart failure where chronically enhanced sympathetic signaling to the heart occurs. In view of these findings, β-adrenergic receptor blockade clearly has effects that reach beyond simple reduction in heart rate.
Communication among cells in the heart (myocytes, fibroblasts, endothelial cells) through electrical or paracrine pathways is important in determining the overall progression of cardiac architecture and function.8 With communication between Gsα and Rosa cells, there would have been less of a clear-cut distinction between the normal and Gsα cells in terms of hypertrophy and fibrosis. The data from the present study support the position that the effects of the overexpressed Gsα are predominantly local, because adjacent clusters of Gsα or Rosa cells demonstrated markedly different phenotypes. The present data support the position that the genotype in combination with the microenvironment is responsible for the hypertrophy and fibrosis characteristic of the cardiomyopathic Gsα phenotype.
An important limitation to the interpretation of the data from the present study is derived from the nature of the chimeric model. As noted above, the cells from the two lines were not distributed randomly or homogeneously but were clustered. Therefore, clusters of Gsα cells developing hypertrophy and fibrosis in one part of the left ventricle could affect LV function out of proportion to the total number of cells present. However, this limitation in experimental design does not have an impact on the major conclusion of this study, demonstrating hypertrophy and fibrosis in one cell type (Gsα) to the exclusion of the other. One final point needs to be mentioned. This chimeric approach is useful not only to address questions raised in this study relevant to this transgenic animal model but also for other genetically altered models. For example, some altered genotypes will be toxic (ie, embryonically lethal) when expressed confluently in the heart and lead to premature death, particularly in homozygous models. However, if these altered cells could be expressed in a mosaic fashion, the remainder of the normal cells could permit the animal to survive to adulthood and allow examination of the local action of either gene overexpression or gene deletion in the heart.9
This study was supported in part by United States Public Health Service Grants HL59139, R01 HL33107, R01 AG14121, R01 HL33065, R01 HL62716, and R01 HL59874.
↵1 Both authors contributed equally to this study.
- Received January 3, 2000.
- Accepted January 13, 2000.
- © 2000 American Heart Association, Inc.
Gaudin C, Ishikawa Y, Wight DC, Vatner DE, Mahdavi V, Nadal-Ginard B, Wagner TE, Homcy CJ. Overexpression of Gsα protein in the hearts of transgenic mice. J Clin Invest. 1995;95:1676–1686.
Iwase M, Bishop SP, Uechi M, Vatner DE, Shannon RP, Kudej R, Wight DC, Wagner TE, Ishikawa Y, Homcy CJ, Vatner SF. Adverse effects of chronic endogenous sympathetic drive induced by cardiac Gsα overexpression. Circ Res. 1996;78:517–524.
Iwase M, Uechi M, Vatner DE, Asai K, Shannon RP, Kudej RK, Wagner TE, Wight DC, Patrick TA, Ishikawa Y, Homcy CJ, Vatner SF. Cardiomyopathy induced by cardiac Gsα overexpression. Am J Physiol. 1997;272:H585–H589.
Uechi M, Asai K, Osaka M, Smith A, Sato N, Wagner TE, Ishikawa Y, Vatner DE, Shannon RP, Homcy CJ, Vatner SF. Depressed heart rate variability and arterial baroreflex in conscious transgenic mice with overexpression of cardiac Gsα. Circ Res. 1998;82:416–423.
Cichowski K, Shih TS, Schmitt E, Santiago S, Reilly K, McLaughlin ME, Bronson RT, Jacks T. Mouse models of tumor development in neurofibromatosis type 1. Science. 1999;286:2172–2176.