Skip to main content
Download PDF
  • Original Articles
  • Published:

Coordinate Changes in Myosin Heavy Chain Isoform Gene Expression Are Selectively Associated With Alterations in Dilated Cardiomyopathy Phenotype

Molecular Medicine volume 8, pages 750–760 (2002)Cite this article

Abstract

Background

The most common cause of chronic heart failure in the US is secondary or primary dilated cardiomyopathy (DCM). The DCM phenotype exhibits changes in the expression of genes that regulate contractile function and pathologic hypertrophy. However, it is unclear if any of these alterations in gene expression are disease producing or modifying.

Materials and Methods

One approach to providing evidence for cause-effect of a disease-influencing gene is to quantitatively compare changes in phenotype to changes in gene expression by employing serial measurements in a longitudinal experimental design. We investigated the quantitative relationships between changes in gene expression and phenotype n 47 patients with idiopathic DCM. In endomyocardial biopsies at baseline and 6 months later, we measured mRNA expression of genes regulating contractile function (β-adrenergic receptors, sarcoplasmic reticulum Ca2+ ATPase, and α- and β-myosin heavy chain isoforms) or associated with pathologic hypertrophy (β-myosin heavy chain and atrial natriuretic peptide), plus β-adrenergic receptor protein expression. Left ventricular phenotype was assessed by radionuclide ejection fraction.

Results

Improvement in DCM phenotype was directly related to a coordinate increase in α- and a decrease in β-myosin heavy chain mRNA expression. In contrast, modification of phenotype was unrelated to changes in the expression of β1- or β2-adrenergic receptor mRNA or protein, or to the mRNA expression of sarcoplasmic reticulum Ca2+ ATPase and atrial natriuretic peptide.

Conclusion

We conclude that in human DCM, phenotypic modification is selectively associated with myosin heavy chain isoform changes. These data support the hypothesis that myosin heavy chain isoform changes contribute to disease progression in human DCM.

Introduction

Disease phenotypes are ultimately the result of changes in gene expression. In the chronically failing hypertrophied human heart, selective changes in the expression of genes that could potentially modify phenotype have been reported (1–8), typically in explanted human hearts at the end stage of the disease process (1–6). Reported alterations in gene expression in this setting include changes in the mRNA or protein expression of components of β-adrenergic signal transduction (2–4), calcium handling proteins (1,5), and contractile proteins (6). However, in these studies it has been difficult to determine which alterations in gene expression are fundamentally related to contractile dysfunction as opposed to those which are epiphenomena associated with advanced disease. Another interpretative challenge in these studies has been the difficulty of distinguishing between adaptive versus maladaptive processes, because many of the described changes are products of compensatory mechanisms that can produce either benefit or harm.

Additionally, because end-stage failing hearts or the organ donor-procured controls used in these studies may be affected by a multitude of ancillary factors, it is important to ultimately investigate gene expression in the intact heart. In the intact heart, starting material can be obtained from less advanced myocardial failure and controls are not subjected to the profound homeostatic disruption of brain death (9). Also, studies in the intact heart may be conducted longitudinally employing serial measurements (8,10–12), which theoretically allows for detection of gene expression changes more directly associated with phenotypic modification. Importantly, because gene regulation is combinatorial (13), a strong case can be made for investigating gene expression in intact tissues where the net effect of multiple regulatory influences can be assessed.

Using reverse transcription quantitative PCR (RT-QPCR), it is possible to measure the steady-state mRNA abundance of multiple genes in small amounts of starting material (14), such as endomyocardial biopsy (1,7,8,12,15). Treatment of chronic heart failure subjects with β-adrenergic blocking agents is associated with improvement in systolic function and a reversal of remodeling in the majority of subjects with a dilated phenotype (16), within a time frame of 4 to 12 months (16–21) that is amenable to serial investigation. This contrasts with the time course of the natural history of dilated cardiomyopathies, which is that of slow (time course of years) progression in phenotype and clinical sequelae (22,23). The current investigation is an analysis of the quantitative relationships between changes in phenotype and the expression of contractility-regulating or pathologic hypertrophy-associated genes in a subject population of idiopathic dilated cardiomyopathy (IDC) patients treated for 6 months with placebo or β-adrenergic blockade.

Methods

Clinical Protocol

This study was designed to investigate changes in gene expression associated with phenotypic modulation in IDC, irrespective of treatment. Fifty-three subjects with symptomatic IDC and left ventricular ejection fraction (LVEF) were randomized to 6 months treatment with placebo, metoprolol tartrate, or carvedilol. Details of this treatment protocol are published elsewhere (8). Forty-nine subjects completed this protocol, and the current report concerns the 47 individuals who had complete (baseline and after 6 months of treatment) mRNA (n = 45) or β-adrenergic receptor protein (n = 36) measurements.

Right Ventricular Endomyocardial Biopsy and Right Heart Catheterization

Right heart catheterization and endomyocardial biopsy with allocation of tissue for mRNA and β-receptor protein measurements were performed from the right internal jugular vein as previously described (7,8).

Messenger RNA Expression Measurements in Endomyocardial Biopsy Material

Messenger RNA abundance for six different genes (β1- and β2-adrenergic receptors, α- and β-myosin heavy chain [MyHC], atrial natriuretic peptide [ANP], and SR Ca2+ ATPase [SERCA-2a or SRCA]) was quantified in extracted total RNA, by RT-PCR using previously described methods (7,8).

β-Adrenergic Receptor Measurements in Endomyocardial Biopsy Material, Catecholamine Measurements

Total β-receptor density and β1- and β2-receptor subtypes were measured in biopsy material as previously described (7,8). Norepinephrine and epinephrine were determined in plasma from coronary sinus blood by a previously described radioenzymatic method (11).

Left and Right Ventricular Ejection Fraction Measurements by Radionuclide Ventriculography

LVEF and right ventricular ejection fraction (RVEF) were measured by in vitro red blood cell labeling and imaging techniques as previously described (18,20), and the results were expressed in EF units ([stroke volume/end diastolic volume] × 100). For RVEF a first-pass technique was employed, and for LVEF an equilibrium technique was used.

Statistical Analysis

The 45 subjects who had baseline and end-of-study mRNA measurements were rank ordered according to change in LVEF into three groups without regard to treatment type. For the 45 subjects who had baseline and end-of-study mRNA measurements, 15 subjects with an LVEF change from −11 and +2 (average −2.2 ± 1.1[standard error of the mean; SEM]) EF units comprised the Declined/No Change function group, 15 subjects with an LVEF change from +5 and +17 (average 12.2 ± 0.9) EF units comprised the Improved group, and 15 subjects with changes from + 18 and +44 (average 25.5 ± 1.8) EF units comprised the Marked Improvement group. The 36 subjects who had baseline and end-of-study receptor protein measurements were also rank ordered by LVEF change. Eleven subjects comprised the Declined/No Change function group (average change −3.5 ± 1.2 EF units, range −11 to + 1 EF units), 13 subjects were in the Improved group (average change 8.5 ± 1.1 EF units, range 2–13 EF units), and 12 subjects were in the Marked Improvement group (average change 23.1 ± 1.6 EF units, range 16–34 EF units). We chose to rank order ventricular function change by LVEF rather than RVEF because (1) gene expression changes on the right side of the septum are closely related to those that occur in both the LV and RV free walls (6), presumably because the septum is a wall shared by both chambers, (2) with radionuclide techniques LVEF is more accurately measured than is RVEF (24), (3) 14 of the 45 study subjects had normal (≥40 EF units) RVEF values at baseline, and (4) as shown in Results, the ordered classification of LVEF changes also resulted in incrementally increasing RVEF changes.

Changes in receptor protein, mRNA, and other variables of interest among the three LV function change groups were analyzed by three methods. In the first method, in which the LV function change groups were treated as ordinal variables, the relationships between various interval-level outcome variables and the ordinal-level change in LVEF were assessed with an ordinal trend test. This test is analogous to a test of bivariate regression used to assess a dose-response relationship between two interval-level variables. When the predictor variable (X) is ordinal level then the interpretation of a dose response is that changes in the outcome variable (Y) between successive values of X be uniform. In the general linear model, X is submitted as a nominallevel variable and specific contrasts are used to ask if the deltas are uniform. When X has three levels, with corresponding Y values Y1, Y2, and Y3, a single contrast, with one degree of freedom, is used to ask if Y2 equals the mean of Y1 and Y3. A low p value for this contrast indicates substantial deviation from uniformity of the deltas. This analysis assumes only that each ascending class of LV function change is greater than the preceding one, and it is useful for detecting the presence of a systematic relationship between dependent and independent variables while making no assumptions about the quantitative nature of such a relationship.

In the second method, the relationship of changes in LV function to changes in the variables of interest was assessed by linear regression analysis, treating LVEF change as a continuous interval variable, with and without statistical adjustment for effect of treatment group. In the third method, the three LV function change groups were analyzed independently by ANOVA and the Bonferroni multiple comparison test, with the model adjusted to statistically control for any effect of treatment group.

Changes in measured parameters among the three treatment groups were assessed by ANOVA/Bonferroni, and changes within treatment group by paired t test. LVEF and RVEF class differences among treatment groups were also assessed by contingency table analysis of the rank-ordered EF changes into Decline/No Change, Improved, and Marked Improvement groups. Changes in New York Heart Association (NYHA) class among LV function change groups were assessed by the Spearman rank correlation coefficient. Potential differences within functional change and treatment groups were assessed by Student’s paired t test. Differences between baseline values in IDC subjects versus nonfailing controls were assessed by Student’s unpaired t test. Baseline data are given with variance estimated as standard deviation, and changes are given with variance estimated as SEM. A two-sided p value < .05 was considered to be statistically significant. p values for the ordinal trend test, ANOVA, and linear regression were adjusted for the effect of treatment group.

Results

Subject Demographics

Table 1 gives the demographics and other clinical descriptors of the 47 subjects with heart failure from IDC that had complete (baseline and after 6 months of treatment) ventricular function and mRNA (n = 45) or β-adrenergic receptor protein (n = 36) measurements. The subject population is a relatively young (mean age 53 years) heart failure group with moderate symptoms (majority NYHA Class III), severe LV dysfunction (mean LVEF 21.6 EF units), preserved resting hemodynamics (mean cardiac index 2.44 mL/min/m2), mild volume overload (mean pulmonary wedge pressure 13.6 mm Hg, mean right atrial mean pressure 5.9 mm Hg), moderate exercise intolerance (mean peak VO2 16.2 ml/kg/min), and marked cardiac adrenergic activation (mean coronary sinus norepinephrine 982 pg/ml). There were no differences in any baseline demographic measurement between subjects assigned to the three treatment groups.

Table 1 Patient demographics ( n = 47)
Full size table

Baseline Gene Expression Data

Baseline mRNA and β-receptor protein data in the 47 IDC subjects are reported elsewhere (8) and consisted of decreases in the mRNA and protein expression of β1-adrenergic receptors, decreases in the mRNA expression of SRCA and α-MyHC genes, and a respective increase or trends for an increase in the mRNA expression of ANP and β-MyHC.

Effect of Treatment

Between-Group Changes By treatment group comparison the only statistically significant changes among groups were in LVEF, where by ANOVA the metoprolol and carvedilol groups exhibited greater numerical increases than the placebo group (that were not quite significant after Bonferroni multiple group correction), and peak exercise heart rate, where both β-blocker groups exhibited greater degrees of reduction (Table 2).

Table 2 Changes in measured parameters by treatment group (±SEM), subjects who had complete mRNA or β -adrenergic receptor measurements ( n = 47)
Full size table

Within-Group Changes There were multiple significant (p < 0.05) changes in the measured parameters within treatment groups, as presented in Table 2. The placebo group exhibited decreases in ANP mRNA abundance and coronary sinus or arterial norepinephrine, and an increase in β1-adrenergic receptors percent. The metoprolol group had increases in total β- and β1-adrenergic receptor densities, LVEF, RVEF, stroke volume, and LV stroke work indices, and decreases in peak exercise heart rate, NYHA class, and resting heart rate. In the carvedilol group, there were increases in total β- and β1-adrenergic receptor density, LVEF, RVEF, stroke volume, and LV stroke work indices, and decreases in peak exercise heart rate, NYHA class, ANP mRNA abundance, pulmonary wedge mean pressure, pulmonary artery mean pressure, and resting heart rate.

Changes in Hemodynamics, Cardiac Adrenergic Drive, Functional Capacity, NYHA Class, and RVEF by LV Function Change Group or by Linear Regression Analysis of LVEF Change

Between-Group Changes Table 3 gives the changes in LVEF, RVEF, hemodynamics, and other measured responses in the three LVEF change groups for subjects who had complete mRNA measurements. As shown in Table 3, LVEF by definition and RVEF data exhibit sequential increases among the three LV function change groups. The only other hemodynamic variable listed in Table 3 that had a significant change by ordinal trend test was pulmonary artery mean pressure. NYHA functional class did not differ among groups by ordinal trend test (p = 0.42) or Spearman rank correlation coefficient (p = 0.16). By ANOVA, pulmonary wedge pressure and pulmonary artery mean pressure exhibit decreases in the Marked Improvement group as compared to the Improved or Decline/No Change groups.

Table 3 Changes in LVEF, RVEF, hemodynamics, coronary sinus norepinephrine, and peak Vo 2 in three LV function change groups, subjects with complete mRNA measurements ( n = 15/group)
Full size table

With linear regression analysis using change in LVEF as a continuous variable, RVEF change was positively related (r = 0.51, p = 0.001), and pulmonary artery mean pressure (r =−0.37, p = 0.029) was negatively related. No other hemodynamic changes were significantly related to change in LVEF.

Within-Group Changes The Decline/No Change group exhibited a decrease in heart rate and an increase in stroke volume index. The Improved group had increases in LVEF, RVEF, and stroke volume index, and a decrease in heart rate. The Marked Improvement group exhibited an increase in LVEF, RVEF, stroke volume index, and left ventricular stroke work index, and a decrease in pulmonary wedge mean pressure, pulmonary arterial mean pressure, and heart rate.

Changes in mRNA Expression by LV Function Change Group or Linear Regression Analysis of LVEF Change

Between-Group Changes Figures 1 through 4 give the changes in mRNA expression in the LV function change groups. As shown in Figure 1, there were no changes in β-adrenergic receptor mRNA expression associated with changes in left ventricular function. There were also no changes in β1- or β2-adrenergic mRNA abundance within any of the function change groups. Similarly, Figure 2 demonstrates that SRCA and ANP gene expression is not different among the different functional groups.

Fig. 1
figure 1

Changes in β1- and β2-adrenergic receptor (AR) mRNA abundances by three LV function change group. p values refer to expression-response analysis. ANOVA was also nonsignificant for both mRNAs.

Full size image
Fig. 2
figure 2

Changes in ANP and SRCA mRNA abundances by three LV function change group. Note y axis is logarithmic. p values refer to expression-response analysis. ANOVA was also nonsignificant for both mRNAs.

Full size image
Fig. 3
figure 3

Changes in α- and β-MyHC mRNA abundances by three LV function change groups. Note y axis is logarithmic. p values refer to ordinal trend test. ANOVA p values were 0.027 for α-MyHC and 0.022 for β-MyHC.

Full size image
Fig. 4
figure 4

Changes in total (α- plus β-) MyHC mRNA abundance and in percent β-MyHC/total MyHC by three LV function change groups. Note y axis is logarithmic for total MyHC. p values refer to expression-response analysis. ANOVA p values are 0.034 for total MyHC and 0.076 for percent α-MyHC.

Full size image

In contrast, there were changes in both α- and β-MyHC that incrementally increased with the function change groups (Fig. 3). Across the three LV function change groups, by adjusted ordinal trend test, there were ascending increases in α-MyHC (p = 0.012) and ascending decreases in β-MyHC (p = 0.007). A similar pattern of ascending decrease was shown for total MyHC (Fig. 4, p = 0.014). Because of the progressive increases in α-MyHC and decreases in β-MyHC gene expression with increasing LV and RV function, there was also a progressive increase in the percent of β-MyHC/total MyHC as function improved (Fig. 4, ordinal trend test p = 0.026).

For linear regression analysis using change in LVEF as a continuous variable, LVEF change was positively related to change in α-MyHC steady-state abundance (r = 0.34, p = 0.023) and to percent α-MyHC (r = 0.29, p = 0.045), and negatively related to change in β-MyHC (r =−0.43, p = 0.002), and total MyHC (r =−0.42, p = 0.003). No other changes in mRNA expression were significantly related to change in LVEF. values refer to expression-response analysis. ANOVA was also nonsignificant for both mRNAs.

Within-Group Changes Within function change groups, the Declined/No Change group exhibited an increase (p < 0.05 by paired t-test) in β-MyHC and total MyHC, and a trend (p < 0.10) for a decrease in ANP mRNA expression. The Improved group had an increase in α-MyHC and percent α-MyHC plus a trend for an increase in β2-receptor, and a decrease in ANP mRNA expression. The Marked Improvement group exhibited trends in increases in α-MyHC and percent β-MyHC, and trends for decreases in β-MyHC, total MyHC, and ANP mRNA expression.

Changes in β-Adrenergic Receptor Protein Expression by LV Function Change Group or Linear Regression Analysis of LVEF Change

Between-Group Changes As shown in Figure 5, by expression-response analysis, there were no differences in the expression of total β-, β1-, or β2-adrenergic receptor protein expression among the three function change groups constructed for the 36 subjects with complete β-adrenergic receptor protein measurements. For β2-receptors, by ANOVA the Improved group had an increased level of expression compared to the Decline/No Change or Marked Improvement group. There were no significant differences between LV function change and receptor protein expression using LVEF change as an ordinal or continuous variable for any receptor type.

Fig. 5
figure 5

Changes in β1- and β2-AR protein in three LV function change groups. N = 11 for Decline/No Change, N = 13 for Improved, and N = 12 for Marked Improvement. p values refer to expression-response analysis. Corrected ANOVA p values are nonsignificant for total β- and β1-receptor densities, and 0.025 for β2-receptor density.

Full size image

Within-Group Changes As shown in Figure 5, the Decline/No Change and Marked Improvement groups exhibited similar within-group increases in β1- and total β-adrenergic receptor densities, and the Improved group exhibited an increase in β2-adrenergic receptor protein expression that was not present in the Decline/No Change or Marked Improvement groups.

Coordinate Changes in α-, β-MyHC mRNA Expression

Figure 6 illustrates the changes in β-MyHC versus α-MyHC, SRCA, ANP, and β1-adrenergic receptor mRNA abundance across all subjects and all EF changes. As can be seen, changes in β-MyHC and α-MyHC are coordinate, with a decreases in β-MyHC related to increases in α-MyHC. In contrast, changes in β-MyHC are unrelated to two other genes ordinarily considered to be part of the fetal gene program, ANP and SRCA. Changes in β-MyHC are also unrelated to changes in β1- orβ2-adrenergic receptor mRNA (data not shown).

Fig. 6
figure 6

(A) Changes between baseline and 6 months in α-MyHC expression versus β-MyHC mRNA abundance for all subjects. (B) Changes in SRCA versus β-MyHC mRNA abundance. (C) Changes in ANP versus β-MyHC mRNA abundance. (D) Changes in β1-adrenergic receptor versus β-MyHC mRNA abundance.

Full size image

Discussion

The most common cause of heart failure in the United States is secondary or primary dilated cardiomyopathy, phenotypically characterized by hypertrophy, ventricular chamber dilatation, and decreased contractile function (25). The alterations in gene expression responsible for dilated cardiomyopathy phenotypes remain unidentified. In terms of contractile dysfunction, leading contenders include changes in components of β-adrenergic signal transduction, calcium-handling mechanisms, and contractile proteins (2). For ventricular pathologic hypertrophy, a net increase in contractile protein and sarcomeres through the activation of myocardial growth fetal genes that include β-MyHC has been described (6,7,26–29). The induction of pathologic hypertrophy and the development of contractile dysfunction are interrelated, because the increased expression of β-MyHC and associated decreased expression of α-MyHC and other genes such as SRCA (28,29) decrease systolic and diastolic function (2).

In this study, we utilized the substantial changes in the dilated cardiomyopathy phenotype that occur variably in chronic heart failure patients treated with β-blockade or placebo to create a dynamic phenotypic model suitable for investigation of the molecular basis of myocardial failure. The advantage of such a model is considerable, because candidate genes whose expression is not associated with phenotypic modification can be distinguished and eliminated from consideration. Improvement in dilated cardiomyopathy phenotype in response to blocker therapy involves an increase in velocity of ventricular pressure development (16), an upward and leftward shift in depressed left (16) and right (20) ventricular function curves indicative of improved intrinsic systolic function, and a decrease in ventricular volume (16–21) and mass (16,17,21). Primary dilated cardiomyopathy—IDC—was chosen for investigation because IDC is relatively common and its gene expression profile is similar to more common secondary cardiomyopathies (6,25,30). The high response rate of IDC to β-blockade plus spontaneous improvement in some subjects in the placebo group led to substantial numbers of subjects exhibiting improvement in LVEF, including 15 subjects with a change of more than 17 EF units, a degree of improvement that amounts to normalization or near normalization of left ventricular function. LVEF was used to measure systolic contractile function because the test is readily available and is standardized, and because LVEF is less load-dependent than standard hemodynamic measurements (31). Moreover, across the three LV function change groups analyzed by ordinal trend test, there were no changes in LV loading conditions that might have confounded the analysis. That is, mean arterial pressure did not change significantly across the three groups, and the trend toward a lower mean pulmonary wedge pressure in the Marked Improvement group would tend to lower, not increase, ejection fraction. Thus, in this study LVEF change should accurately reflect changes in LV systolic function. Because the ejection fraction calculation is heavily influenced by end-diastolic volume, in dilated cardiomyopathies LVEF is also an estimate of the degree of chamber dilatation or remodeling (32).

The IDC population investigated in this study exhibited the gene expression abnormalities previously reported by our (1,4,6-8,30) and other (3,5,15) laboratories. These changes include components of fetal gene program induction (increased mRNA expression of ANP and β-MyHC, decreased expression of α-MyHC and SR Ca2+ ATPase), and the consequences of chronic adrenergic activation (down-regulation in β1-receptor mRNA and protein expression). The question addressed by this study is which of these gene expression abnormalities most closely relates to modifications in dilated cardiomyopathy phenotype measured over a 6-month period in all subjects, regardless of treatment. Because the nature of the relationship between gene expression and any related LVEF change is uncertain and because EF change is unlikely to be linearly related to changes in intrinsic function or degree of pathologic hypertrophy, several types of statistical models were employed.

The results of this study indicate that changes in the mRNA expression of two genes whose regulation is coordinately linked (33), α- and β-MyHC, are selectively associated with modifications in ventricular systolic function and remodeling. The expression of α-MyHC exhibited a statistically significant progressive increase across ascending LV function change groups, from a negative value in the Decline/No Change group to increasingly positive values in the Improved and Marked Improvement groups. Conversely, β-MyHC gene expression exhibited a decrease from a positive change value in the Decline/No Change group to progressively more negative values in the Improved and Marked Improvement groups. If translated into proportional changes in protein expression (34), such changes in MyHC isogene expression could potentially improve intrinsic systolic and diastolic function (35) because the α-MyHC isoform has a much higher myosin ATPase activity than the β-isoform, resulting in a faster velocity of shortening and rate of cross-bridge cycling (26,27,35). In rodent (35–37) and rabbit (38) models of pathologic hypertrophy and myocardial failure, coordinate decreases in α-MyHC and increases in β-MyHC mRNA and protein expression are associated with a reduction in velocity of shortening and other measures of systolic function.

Although in the current study the small amounts of starting material precluded measurement of α-and β-MyHC protein, in gram quantities of failing explanted human left ventricular myocardium changes in gel electrophoresis-measured MyHC iso-forms are directionally similar to changes in mRNA (34). However, in human ventricular myocardium the absolute expression of α-MyHC protein is much less than its mRNA expression (32,39). For example, in nonfailing left ventricles α-MyHC protein expression is 5–10% of total MyHC, versus 25–30% for mRNA (34). Although the protein expression of α-MyHC decreases to near zero in failing hearts (34,39), it is unclear if such a small absolute change can account for a decrease in intrinsic contractile function. In favor of such small changes being functionally significant are the observations that (1) in model systems small amounts (∼10%) (40) or changes (37) of α-MyHC have relatively large effects on force development and (2) recent data from our laboratory demonstrating that human cardiac α-MyHC has greater (relative to β-MyHC) specific myosin and myofibrillar ATPase activities than has been in reported in rat or rabbit hearts (K. Nunley et al., submitted). Thus an increase in percent αMyHC isoform expression is a molecular change that could causally relate to improved function. Alternatively, changes toward normal in MyHC isoforms could simply be a molecular marker of de-induction of the fetal gene program, with other components accounting for the actual improvement in function.

Similarly, the progressive decrease in total MyHC (α-MyHC + β-MyHC) gene expression with increasing LVEF (and decreasing end diastolic volume) (32) is a molecular measure of reverse remodeling; MyHC is a major structural protein of the myocardial contractile element (41). As a molecular index of reverse remodeling, in the current study changes in total MyHC or β-MyHC appeared to be superior to the indirect hypertrophic marker ANP, because total or β-MyHC but not ANP expression exhibited an incremental decline as remodeling progressively decreased. ANP gene expression decreased or tended to decrease in all LV function/diastolic volume change groups, indicating that this molecular phenotypic marker was nonspecifically responding to the general treatment of chronic heart failure. This was confirmed by significant reductions in ANP gene expression within the placebo and carvedilol treatment groups.

There were also no changes in β1-adrenergic receptor gene or protein expression or in SRCA gene expression that were systematically associated with changes in ventricular function. In subjects treated with β-blocking agents SRCA increases selectively in subjects who exhibit an improvement in LVEF, but spontaneous LVEF responders treated with placebo do not exhibit increases in SRCA (8). Therefore, in the current analysis that included subjects treated with or without β-blockade, there was no relationship between change in SRCA gene expression and change in phenotype. Total β- and β1-adrenergic receptor protein density increased or tended to increase in all three LV function change groups. This indicates that changes in these molecular myocardial dysfunction candidates can occur in response to heart failure medical treatment, unrelated to modification of ventricular function or remodeling. This was confirmed by significant increases in total β- and β1-adrenergic receptor protein density within the metoprolol and carvedilol treatment groups.

In this study, α-MyHC and β-MyHC were coordinately and reciprocally regulated across all changes in LV function. In contrast, two other genes ordinarily considered to be part of the fetal gene program, ANP and SRCA, were not coordinately regulated with β-MyHC. Coordinate, reciprocal regulation of cardiac MyHC isoforms in a variety of settings, including pathologic hypertrophy and failure, is well known in small animal models (42) and our data extend these findings to the human heart. The dissociation of regulation of β-MyHC from SRCA and ANP indicates that fetal gene regulation is complex, and may be individualized for MyHC isoforms versus other members of this program.

In summary, the data presented in this study indicate that in chronic heart failure patients with a dilated cardiomyopathy phenotype changes in MyHC isogene expression measured in septal endomyocardial biopsy material are closely associated with modification in left ventricular function and chamber remodeling. These associations are specific for MyHC isoform expression, and are not observed for several other molecular changes previously shown to be present in the failing human heart. These observations suggest that modulation of MyHC isoforms may be a therapeutic target for drug development in chronic heart failure, particularly because such treatment should result in cardiac specific effects.

Disclosures

Drs Gilbert, Abraham, and Bristow have served as paid consultants to Glaxo SmithKline, and Dr Bristow has been a paid consultant for AstraZeneca. Dr Bristow is a founder, officer, and stockholder in Myogen, which owns a license to aspects of the current study that are covered in a patent application.

References

  1. Feldman AM, Ray PA, Silan CM, Mercer JA, Minobe W, Bristow MR. (1991) Selective gene expression in failing human heart. Quantification of steady-state levels of messenger RNA in endomyocardial biopsies using the polymerase chain reaction. Circulation 83: 1866–1872.

    Article  CAS  PubMed  Google Scholar 

  2. Bristow MR. (1998) Why does the myocardium fail? New insights from basic science. Lancet 352(suppl I): 8–14.

    Article  CAS  Google Scholar 

  3. Ungerer M, Böhm M, Elce JS, Erdmann E, Lohse MJ. (1993) Altered expression of β-adrenergic receptor kinase and β1-adrenergic receptors in the failing human heart. Circulation 87: 454–463.

    Article  CAS  PubMed  Google Scholar 

  4. Bristow MR, Minobe WA, Raynolds MV, et al. (1993) Reduced β1 receptor messenger RNA abundance in the failing human heart. J. Clin. Invest. 92: 2737–2745.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Mercadier JJ, Lompre AM, Duc P, et al. (1990) Altered sarcoplasmic reticulum Ca-ATPase gene expression in the human ventricle during end-stage heart failure. J. Clin. Invest. 85: 305–309.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Nakao K, Minobe WA, Roden RL, Bristow MR, Leinwand LA. (1997) Myosin heavy chain gene expression in human heart failure. J. Clin. Invest. 100: 2362–2370.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Lowes BD, Minobe WA, Abraham WT, et al. (1997) Changes in gene expression in the intact human heart: down-regulation of α-myosin heavy chain in hypertrophied, failing ventricular myocardium. J. Clin. Invest. 100: 2315–2324.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Lowes BD, Gilbert EM, Abraham WT, et al. (2002) Myocardial gene expression in dilated cardiomyopathy treated with beta-blocking agents. N. Engl. J. Med. 346: 1357–1365.

    Article  CAS  PubMed  Google Scholar 

  9. Lopau K, Mark J, Schramm L, Heidbreder E, Wanner C. (2000) Hormonal changes in brain death and immune activation in the donor. Transplant. Int. 13: S282–S285.

    Article  Google Scholar 

  10. Heilbrunn SM, Shah P, Bristow MR, Valantine HA, Ginsburg R, Fowler MB. (1989) Increased β-receptor density and improved hemodynamic response to catecholamine stimulation during long-term metoprolol therapy in heart failure from dilated cardiomyopathy. Circulation 79: 483–490.

    Article  CAS  PubMed  Google Scholar 

  11. Gilbert EM, Abraham WT, Olsen S, et al. (1996) Comparative hemodynamic, left ventricular functional, and antiadrenergic effects of chronic treatment with metoprolol versus carvedilol in the failing heart. Circulation 94: 2817–2825.

    Article  CAS  PubMed  Google Scholar 

  12. Ladenson PW, Sherman SI, Baughman KL, Ray PE, Feldman AM. (1992) Reversible alterations in myocardial gene expression in a young man with dilated cardiomyopathy and hypothyroidism. Proc. Natl. Acad. Sci. U.S.A. 89: 5251–5255.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Adolph EA, Subramanian A, Cserjesi P, Olson EN, Robbins J. (1993) Role of myocyte-specific enhancer-binding factor (Mef-2) in transcriptional regulation of the α-cardiac myosin heavy chain gene. J. Biol. Chem. 268: 5349–5352.

    PubMed  CAS  Google Scholar 

  14. Wang AM, Doyle MV, Mark DF. (1989) Quantitation of mRNA by the polymerase chain reaction. Proc. Natl. Acad. Sci. U.S.A. 86: 9717–9721.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Engelhardt S, Bohm M, Erdmann E, Lohse MJ. (1996) Analysis of beta-receptor mRNA levels in human ventricular biopsy specimens by quantitative polymerase chain reactions: progressive reduction of beta 1-adrenergic receptor mRNA in heart failure. J. Am. Coll. Cardiol. 27: 146–154.

    Article  CAS  PubMed  Google Scholar 

  16. Eichhorn EJ, Bristow MR. (1996) Medical therapy can improve the biologic properties of the chronically failing heart: a new era in the treatment of heart failure. Circulation 94: 2285–2296.

    Article  CAS  PubMed  Google Scholar 

  17. Hall SA, Cigarroa CG, Marcoux L, Risser RC, Grayburn PA, Eichhorn EJ. (1995) Time course of improvement in left ventricular function, mass, and geometry in patients with congestive heart failure treated with β-adrenergic blockade. J. Am. Coll. Cardiol. 5: 1154–1161.

    Article  Google Scholar 

  18. Quaife RA, Gilbert EM, Christian PE, et al. (1996) Effects of carvedilol on systolic and diastolic left ventricular performance in idiopathic dilated cardiomyopathy. Am. J. Cardiol. 78: 779–784.

    Article  CAS  PubMed  Google Scholar 

  19. Doughty RN, Whalley GA, Gamble G, MacMahon S, Sharpe N. (1996) Left ventricular remodeling with carvedilol in patients with congestive heart failure due to ischemic heart disease. J. Am. Coll. Cardiol. 29: 1060–1066.

    Article  Google Scholar 

  20. Quaife RA, Christian PE, Gilbert EM, Datz FL, Volkman K, Bristow MR. (1998) Effects of carvedilol on right ventricular function in chronic heart failure. Am. J. Cardiol. 81: 247–250.

    Article  CAS  PubMed  Google Scholar 

  21. Lowes BD, Gill WT, Larrain JR, Robertson AD, Bristow MR, Gilbert EM. (1999) Effects of carvedilol on left ventricular mass, chamber geometry and mitral regurgitation in chronic heart failure. Am. J. Cardiol. 83: 1201–1205.

    Article  CAS  PubMed  Google Scholar 

  22. Cintron G, Johnson G, Francis G, Cobb F, Cohn JN. (1993) Prognostic significance of serial changes in left ventricular ejection fraction in patients with congestive heart failure. Circulation 87(suppl VI): VI–17–VI–23.

    Google Scholar 

  23. Cetta F, Michels VV. (1995) The natural history and spectrum of idiopathic dilated cardiomyopathy, including HIV and peripartum cardiomyopathy. Curr. Opin. Cardiol. 10: 332–338.

    Article  CAS  PubMed  Google Scholar 

  24. Zaret BL, Wackers FJ. (1991) Measurement of right ventricular function. In Gerson MC, ed. Cardiac Nuclear Medicine, 2nd ed. New York: McGraw-Hill.

    Google Scholar 

  25. Bristow MR, Mestroni L, Bohlmeyer T, Gilbert EM. (2001) Dilated cardiomyopathies. In Fuster V, Alexander RW, O’Rourke RA, eds. Hurst’s the Heart, Volume 2, 10th ed. New York: McGraw-Hill; pp. 1947–1966.

    Google Scholar 

  26. Swynghedauw B. (1986) Developmental and functional adaptation of contractile proteins in cardiac and skeletal muscles. Physiol. Rev. 66: 710–771.

    Article  CAS  PubMed  Google Scholar 

  27. Nadal-Ginard B, Mahdavi V. (1989) Molecular basis of cardiac performance. Plasticity of the myocardium generated through protein isoform switches. J. Clin. Invest. 4: 1693–1700.

    Article  Google Scholar 

  28. Schwartz K, Chassagne C, Boheler KR. (1993) The molecular biology of heart failure. J. Am. Coll. Cardiol. 22: 30A–33A.

    Article  CAS  PubMed  Google Scholar 

  29. Colucci WS. (1997) Molecular and cellular mechanisms of myocardial failure. Am. J. Cardiol. 80: 15L–25L.

    Article  CAS  PubMed  Google Scholar 

  30. Bristow MR, Anderson FL, Port JD, et al. (1991) Differences in β-adrenergic neuroeffector mechanisms in ischemic versus idiopathic dilated cardiomyopathy. Circulation 84: 1024–1039.

    Article  CAS  PubMed  Google Scholar 

  31. Rahko PS. (1994) Comparative efficacy of three indexes of left ventricular performance derived from pressure-volume loops in heart failure induced by tachypacing. J. Am. Coll. Cardiol. 23: 209–218.

    Article  CAS  PubMed  Google Scholar 

  32. Cohn JN, Ferrari R, Sharpe N, on behalf of the International Forum on Cardiac Remodeling. (2000) Cardiac remodeling—concepts and clinical implications: A consensus paper from an international forum on cardiac remodeling. J. Am. Coll. Cardiol. 35: 569–582.

    Article  CAS  PubMed  Google Scholar 

  33. Weiss A, Leinwand LA. (1996) The mammalian myosin heavy chain gene family. Ann. Rev. Cell. Dev. Biol. 12: 417–439.

    Article  CAS  Google Scholar 

  34. Miyata S, Minobe WA, Bristow MR, Leinwand LA. (2000) Myosin heavy chain isoform expression in the failing and non-failing human heart. Circ. Res. 86: 386–390.

    Article  CAS  PubMed  Google Scholar 

  35. Vanburen P, Harris DE, Alpert NR, Warshaw DM. (1995) Cardiac V1 and V3 myosins differ in their hydrolytic and mechanical activities in vitro. Circ. Res. 77: 439–444.

    Article  CAS  PubMed  Google Scholar 

  36. Imamura SI, Matsuoka R, Hiratsuka E, et al. (1991) Adaptational changes of MyHC gene expression and isoenzyme transition in cardiac overloading. Am. J. Physiol. 260: H73–H79.

    PubMed  CAS  Google Scholar 

  37. Herron TJ, McDonald KS. (2002) Small amounts of α-myosin heavy chain isoform expression significantly increase power output of rat cardiac myocyte fragments. Circ. Res. 90: 1–4.

    Article  CAS  Google Scholar 

  38. Litten RZ 3rd, Martin BJ, Low RB, Alpert NR. (1982) Altered myosin isozyme patterns from pressure-overloaded and thyrotoxic hypertrophied rabbit hearts. Circ. Res. 50: 856–864.

    Article  CAS  PubMed  Google Scholar 

  39. Reiser PJ, Portman MA, Ning XH, Schomisch Moravee C. (2001) Human cardiac myosin heavy chain isoforms in fetal and failing adult atria and ventricles. Am. J. Physiol. 280: H1814–H1820.

    CAS  Google Scholar 

  40. Razumova MV, de Tombe PP, Moss RL. (2001) Simulations predict that expression of small amounts of αMHC in mammalian ventricle significantly accelerates the rate of rise of force. Biophys. J. 80: 261a.

    Google Scholar 

  41. Jones WK, Grupp IL, Doetschman T, et al. (1996) Ablation of the murine alpha myosin heavy chain gene leads to dosage effects and functional deficits in the heart. J. Clin. Invest. 98: 1906–1917.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Morkin E. (2000) Control of myosin heavy chain gene expression. Microsc. Res. Tech. 50: 522–531.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgments

This investigation was sponsored by NIH Grants 1R01 HL48013 from the NHLBI, M01-RR00051 and M01-RR00064 from the National Center for Research Resources, and by GlaxoSmithKline Company. The authors are grateful to Laurel Hunter for her assistance in manuscript and figure preparation.

Author information

Authors and Affiliations

  1. Division of Cardiology, University of Colorado Health Sciences Center, 4200 East 9th Avenue, Campus, Box B139, Denver, Colorado, 80262, USA

    W. T. Abraham, B. D. Lowes, W. A. Minobe, R. L. Roden, D. Dutcher, J. Sederberg, J. A. Lindenfeld, E. E. Wolfel, S. F. Shakar, D. Ferguson, J. V. Linseman, R. A. Quaife, A. D. Robertson & Michael R. Bristow (Head)

  2. Division of Cardiology, University of Utah Health Sciences Center, Salt Lake City, Utah, USA

    P. Larrabee & K. Volkman

  3. Division of Cardiovascular Medicine, University of Kentucky College of Medicine, Lexington, Kentucky, USA

    E. M. Gilbert

Authors
  1. W. T. Abraham
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. E. M. Gilbert
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. B. D. Lowes
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. W. A. Minobe
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. P. Larrabee
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. R. L. Roden
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. D. Dutcher
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. J. Sederberg
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. J. A. Lindenfeld
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. E. E. Wolfel
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. S. F. Shakar
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. D. Ferguson
    View author publications

    You can also search for this author in PubMed Google Scholar

  13. K. Volkman
    View author publications

    You can also search for this author in PubMed Google Scholar

  14. J. V. Linseman
    View author publications

    You can also search for this author in PubMed Google Scholar

  15. R. A. Quaife
    View author publications

    You can also search for this author in PubMed Google Scholar

  16. A. D. Robertson
    View author publications

    You can also search for this author in PubMed Google Scholar

  17. Michael R. Bristow
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Michael R. Bristow.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Abraham, W.T., Gilbert, E.M., Lowes, B.D. et al. Coordinate Changes in Myosin Heavy Chain Isoform Gene Expression Are Selectively Associated With Alterations in Dilated Cardiomyopathy Phenotype. Mol Med 8, 750–760 (2002). https://doi.org/10.1007/BF03402039

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/BF03402039

Keywords

Molecular Medicine

ISSN: 1528-3658