Genetic Etiologies and Molecular Mechanisms of Hypertrophic Cardiomyopathy: Strategies to Predict and Prevent Disease
Summary: Christine Seidman is interested in understanding the genetic basis of human cardiovascular disorders such as cardiomyopathy (hypertrophic and dilated), heart failure, and congenital heart malformations. Using experimental models that are engineered to carry human mutations, her lab examines the consequences of mutations on cardiac biology that lead to clinical manifestations of disease. She hopes to combine knowledge of genetic etiologies and molecular mechanisms to improve therapeutic opportunities for patients.
Hypertrophic cardiomyopathy (HCM) is a primary disease of the myocardium that is caused by a dominant mutation in genes encoding sarcomere proteins. Many HCM mutations are missense, encoding a single amino acid substitution in an otherwise normal sarcomere protein. Early in life, HCM mutations are clinically quiescent, but during adolescence there is insidious emergence of ventricular hypertrophy along with changes in cardiac histology, including myocyte enlargement, disorganization of myocyte alignment, and increased amounts of cardiac fibrosis. Over time these changes progressively impact heart function, produce shortness of breath and diminished exercise capacity, and may cause life-threatening arrhythmias. Relentless progression of HCM can also result in heart failure. How a subtle change—a single amino acid substitution in one sarcomere protein gene—produces the multiple anatomic and functional manifestations of HCM remains incompletely understood.
We have tackled this question by producing mouse lines that carry human HCM missense mutations (Arg403Gln and Arg719Trp) in the cardiac myosin heavy chain. Mice carrying either of these mutations exhibit protean manifestations of HCM: a period of clinical latency during which there is no hypertrophy and normal cardiac histology (denoted as prehypertrophic), followed by the emergence of prototypic disease. Our earlier studies of prehypertrophic mice showed that functional abnormalities, including enhanced cardiac contraction and diminished cardiac relaxation, precede the development of cardiac hypertrophy. In collaboration with David Warshaw (University of Vermont), we also showed that HCM missense mutations profoundly alter myosin biophysical properties. Myosins with HCM mutations had increased actin-activated ATPase activity, increased actin sliding velocity, and increased myofilament and myocardial force generation, data that indicate that HCM mutations activate, rather than diminish, the motor properties of myosin.
To elucidate how HCM-activating mutations in myosin triggered the changes in myocardial architecture and perturbed heart function, we have harnessed next-generation sequencing technologies to define global transcriptional responses that emerge in HCM mice over time. These analyses longitudinally examined all RNAs expressed in the heart in prehypertrophic mice and in mice with overt HCM. Furthermore, because HCM histopathology implicates changes in both the myocyte and nonmyocyte populations of the heart, we perform transcriptional analyses in isolated cells from each of these populations.
We have observed that HCM mutations induce robust transcriptional responses in myocytes, as these cells express abundant amounts of the mutant and wild-type sarcomere proteins. But unexpectedly we also observed early and substantial changes in RNAs expressed in nonmyocytes, cells in the heart that do not express sarcomere proteins. We used the functional annotation tools DAVID and Gene Ontology to categorize the processes associated with differentially expressed RNAs in nonmyocytes from prehypertrophic and HCM mice compared to wild-type mice. These analyses suggested that nonmyocytes had enriched expression of molecules involved in nucleic acid synthesis, cell cycling, and proliferation. To ask directly if HCM mutations activate proliferation of nonmyocyte cells, we performed BrdU-labeling and Ki67 immunohistochemical studies in mice carrying HCM mutations. Cardiac sections showed that 4-fold more nonmyocyte cells were labeled by these markers of proliferation in HCM hearts than hearts from wild-type mice, especially around regions where the interstitium was expanded. There was no evidence for proliferation in myocytes.
Another class of RNAs with significantly increased expression in nonmyocyte cells encoded molecules involved in extracellular matrix biology, including periostin, connective tissue growth factor, and collagens. Immunohistochemical analyses confirmed robust expression of these molecules in cardiac sections from mice with overt HCM, as would be expected based on cardiac histopathology of increased amounts of fibrosis. But remarkably, RNA analyses also indicated increased expression of these profibrotic genes in nonmyocyte populations from prehypertrophic hearts, in which histology is normal. As many of these molecules are known to be responsive to transforming growth factor β (Tgfβ), we examined and found significantly increased Tgfβ in nonmyocyte cells from both prehypertrophic and HCM mice. We also observed increased Smad2 phosphorylation in HCM hearts, implicating activation of the canonical Tgfβ-signaling pathway in nonmyocyte cells.
Two potential explanations might account for why sarcomere gene mutations trigger Tgfβ activation in nonmyocyte cells. First, because our previous studies have identified abnormal calcium cycling in HCM myocytes, we hypothesize that calcium-mediated signals are involved in transcriptional activation in nonmyocyte cells. Work is ongoing to better define these signals. Second, because HCM mutations increase the biophysical properties of myosins (increasing force, actin sliding velocity, and ATPase activity), we speculate that enhanced biomechanical forces on the surrounding nonmyocytes directly activate Tgfβ signaling. In support of this mechanical signal transduction mechanism, we note that other researchers have demonstrated that cultured fibroblasts subjected to stretch increase Tgfβ expression above levels found in stationary cultures.
We posited that Tgfβ signals are essential for nonmyocyte proliferation and increased expression of profibrotic molecules and that these signals contribute to the emergence of an expanded extracellular matrix and focal fibrosis that is observed in HCM histopathology. To test this, we assessed the impact of silencing Tgfβ signals by treating young, prehypertrophic mice with Tgfβ-neutralizing antibodies. This strategy resulted in marked reduction in nonmyocyte proliferation, suppression of extracellular matrix proteins, and reduced amounts of developed myocardial fibrosis. We extended this study by administering losartan, an angiotensin II type 1 receptor inhibitor that can also inhibit Tgfβ activation and reduce circulating Tgfβ, to prehypertrophic mice. Longitudinal assessment of losartan-treated prehypertrophic mice showed less development of myocardial fibrosis and hypertrophy compared to untreated mice. In contrast, losartan treatment of mice with established HCM was not beneficial.
We have begun to assess the relevance of transcriptional studies in HCM mice in human HCM, by using cardiac imaging studies and biomarkers to compare asymptomatic young carriers of HCM mutations without cardiac hypertrophy and age-matched, mutation-negative control subjects. Although both groups had normal cardiac dimensions, carriers of HCM mutations had both increased contractile performance (supra-normal ejection fraction) and reduced diastolic function (lower global E' velocity, which assesses passive filling of the ventricle and provides an estimate of ventricular compliance). These findings may be the clinical equivalencies of enhanced biophysical properties in myosins with an HCM mutation and early relaxation deficits found in prehypertrophic mice. To consider if young mutation carriers also had evidence for the transcriptional changes identified in mice, particularly Tgfβ activation, we assessed serum levels of biomarkers of fibrosis. In comparison to age-matched mutation-negative controls, young HCM mutation carriers had significantly higher serum levels of carboxy-terminal propeptide of procollagen type I (PICP), a marker of collagen synthesis and degradation. Similar analyses in HCM patients with overt hypertrophy showed hyper-dynamic contraction, more severe diastolic dysfunction, and still higher serum PICP levels.
These studies suggest that an early consequence of HCM mutations is to activate nonmyocyte cells of the heart to proliferate and to expand the extracellular matrix that surrounds myocytes—changes that may contribute to impaired relaxation capacity even before hypertrophy emerges—in mice and humans. Based on the transcriptional signals activated in the hearts of prehypertrophic mice, we hypothesize that Tgfβ signaling is central to this process and promotes profibrotic gene expression in the heart. A corollary to these studies is that early inhibition of Tgfβ signals in young patients with HCM mutations might reduce the remodeling of the extracellular matrix and attenuate the emergence of overt HCM, as we observed in prehypertrophic mice in which Tgf signaling was suppressed.
With the availability of genetic testing for HCM in at-risk individuals and biomarkers of early disease, we are now poised to translate these research studies into the clinics. In collaboration with our colleagues, we are developing protocols to assess the clinical impact in early phases of HCM of suppression of Tgf signals via FDA-approved drugs. We plan to assess whether treatment impacts the development and/or progression of relaxation abnormalities and elevated PICP and ultimately retards or diminishes the emergence of hypertrophy. We hope that continued mechanistic insights from basic investigations may advance the real opportunity presented by discovering the genetic basis of HCM—accurate prediction and prevention of disease development.
Grants from the National Institutes of Health provide partial support for these projects.
As of December 29, 2011