Received Date: March 23, 2017; Accepted Date: March 28, 2017; Published Date: April 05, 2017
Citation: Ilaria S, Valentina C, Michela C, Claudio T, Giulio P, et al. (2017) The Guilty Party in Arrhythmogenic Cardiomyopathy Adipogenesis. J Stem Cell Res Ther 7:383. doi: 10.4172/2157-7633.1000383
Copyright: © 2017 Ilaria S, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of Stem Cell Research & Therapy
Arrhythmogenic Cardiomyopathy (ACM) is a genetic disease characterized by fibro-fatty substitution of the ventricular myocardium, arrhythmias and sudden death. Although adipogenesis is today well recognized in ACM as an aberrant reparative response following myocardial loss, it is still unknown which cell type(s) is (are) accountable for the adipose replacement. This review provides a comprehensive overview of the different cells that have been, over time, considered as the main players in adipose tissue deposition. Understanding the source of fibro-fatty substitution in ACM will both help to deepen our current knowledge on disease pathogenesis, and to identify a potential significant therapeutic target.
Arrhythmogenic cardiomyopathy; Adipogenesis; Progenitors; Cardiac mesenchymal cells; Cardiomyocytes
Arrhythmogenic Cardiomyopathy (ACM) is a genetic severe cardiac condition predominantly affecting the right ventricle with unfavorable prognosis due to malignant ventricular arrhythmias and heart failure. The pathological hallmark of the disease is a progressive loss of contractile myocardium, that is replaced by fibrous and adipose tissue . This fibro-fatty substitution process extends transmurally with an epi-endocardial gradient, causing ventricular free walls thinning and aneurysmal dilation, especially in the so-called “triangle of dysplasia” (inflow tract, outflow tract and apex)  of the right ventricle. Fibro-fatty substitution predominantly drives a worsening of the arrhythmogenic phenotype, characterized by typical electrocardiographic features (ε waves, localized right precordial QRS prolongation, T-wave inversion) and life-threatening ventricular arrhythmias, usually with a left bundle branch block . Moreover, fibro-fatty substitution provokes progressive regional and global bi-ventricular dysfunction  and ultimately congestive heart failure.
PKP2 and the other desmosomal genes are frequently mutated in ACM , causing alterations of intermediate filament-mediated anchorage of cardiac cells to each other, impairing tissue structural integrity, electrical conduction, mechanical contraction and cellular functionality . To date, enough evidence has been provided indicating that electrical instability in ACM is caused by cardiomyocyte defects . On the contrary, adipose replacement, which is likely to be attributed to a defective reparative response following myocardial loss , lacks an indisputable proof regarding its origin.
Cardiomyocyte loss has been attributed to different pathogenic mechanisms, among which stretch-induced cell damage  facilitated by desmosomal mutations. Moreover, since inflammatory infiltrates have been found in ACM hearts, the co-occurrence of myocarditis has been postulated to further produce cardiomyocytes injury and death . Finally, the myocyte apoptotic theory has been formulated due to the presence of high levels of CPP-32, a cysteine protease required for apoptosis, detected in ACM patients’ hearts . Whatever the reason, cardiomyocytes death is now a well-recognized trigger of fibro-fatty substitution.
The “infiltrative theory” of adipocytes and fibroblasts from the epicardial cell layer  has been advanced as a plausible hypothesis for fibro-fatty substitution, according to evidences at both tissue and clinical levels that fibro-fatty infiltration progresses from the epicardium towards the endocardium .
On the contrary, The “differentiation theory”, proposes that resident cardiac cells differentiate into adipocytes, through Wnt pathway alterations. Indeed, it was hypothesized that an altered desmosomal structure provokes Plakoglobin (PG) translocation to the nucleus, where it induces cell transcriptional activity changes. This provokes the increase of adipogenic and fibrogenic gene expression, thus contributing to differentiation . Recently the Hippo pathway, activated by mechanical stress, G-protein-coupled receptor signaling, and oxidative stress , has been found involved, with the effect of dysregulating cell proliferation/apoptosis and further suppressing Wnt pathway .
Although both theories are supported by strong rationales, so far a single theory has not been exclusively proven by clear-cut evidence. Which cell type(s) is (are) responsible for the aberrant adipogenic differentiation process is therefore still an open and burning question. This review aims to provide a systematic overview about possible cell effectors that have been proposed as players of adipose replacement in ACM, and is meant to provide a new vision of the mechanisms of fibrofatty development, in an attempt to harmonize the “infiltrative” and “differentiation” theories.
The comprehension of molecular mechanisms underpinning adipose tissue development is crucial to dissect and characterize the adipogenic differentiation process in ACM.
In white adipose tissue, adipocyte formation is a multifaceted and stepwise differentiation process involving many mediators. Adipocytes derive from Adipose Tissue-Mesenchymal Stromal Cells (A-MSC), differentiating at first into lipoblasts, then to preadipocytes and finally into mature adipocytes . The acquisition of the adipocyte phenotype is characterized by sequential changes in the expression of many genes : Peroxisome Proliferator-Activated Receptor Gamma (PPARγ), CCAAT/Enhancer-Binding Proteins (C/EBPs) and sterol regulatory element binding protein (SREBP) transcription factors are the major determinants of adipocyte fate . PPARγ and C/EBP-α are expressed early and induce cell growth arrest; levels of PPARγ progressively increase to allow the adipogenic switch . At later phases, when preadipocytes have committed to the adipogenesis program, activation of metabolic genes and adipokines occurs, such as Fatty Acid-Binding Protein 4 (FABP4), Glucose Transporter 4 (GLUT4), leptin, and adiponectin (ADIPO Q) . Increased levels of total adrenergic receptors are also reported . In addition, adipocytes synthesize other adipose tissue-specific products, such as the fatty acid transporter CD36  and perilipin , a lipid droplet-surrounding protein. This process provokes changes in cell morphology, with conversion from fibroblastic to spherical shape, along with modifications of cytoskeletal and extracellular matrix components .
Wnt-mediated signaling is one of the major pathways involved in A-MSC differentiation. In particular, the presence of Wnt ligands normally leads to intracellular molecular events that sustain adipogenesis inhibition . In addition to transcription changes induced by transcription factors, epigenetic mechanisms are involved in the adipogenic process: the histone methylation dependent on the activity of the H3K9-methyltransferase-G9a, is enriched on the PPARγ locus in preadipocytes . During adipogenesis the increase of PPARγ activity inversely correlates with methylation and G9a levels. Indeed, G9a has been proven to inhibit adipogenesis through a direct role on PPARγ and by targeting Wnt signal . Moreover, microRNAs, fine regulators of gene expression, are also implicated in adipogenic commitment .
Strong analogies can be found in the differentiation process occurring in adipose tissue and in ACM hearts. As described, the involvement of certain adipogenic pathways is clearly established in ACM [14,23,24], while other mechanisms are still unexplored, even if potentially exploitable.
The hypothesis that adipocytes in ACM hearts derive from cardiomyocyte trans-differentiation has gained in the past a widespread consensus. Interestingly, a single report, published more than 15 years ago, described that the histological, immunochemical and ultrastructural analysis of an ACM heart were suggestive of cardiomyocytes adipogenic transdifferentiation . The authors found, in both ventricles, that cardiomyocytes contiguous to the adipose tissue showed, instead of the myofibril component, multiple sarcoplasmic vacuoles, that made these cells indistinguishable from preadipocytes. By means of immunohistochemistry staining for desmin to mark muscle tissue, and vimentin, a protein expressed in adipocytes and absent in adult cardiomyocytes , they observed that a few myocytes showing cytoplasmic vacuoles were positive for both these markers. They interpreted these cells as transitional elements between cardiomyocytes and fat cells. However, vimentin is not exclusively expressed in adipocytes, being a major cytoskeletal component of mesenchymal cells, therefore this evidence is not per se demonstrative of myocyte trans-differentiation. This evidence has never been confirmed to date. In a recent work, we could not detect in human ACM hearts cardiomyocyte positivity for the adipocyte specific marker PLIN1 as well as pre-adipocyte positivity for α-sarcomeric actin . Thus, this evidence excludes a direct contribution of adult cardiomyocytes transdifferentiation into adipocytes.
Interestingly, in 2008, Fujita et al.  observed in an ACM patient’s heart biopsy a large amount of adipose tissue and a group of isolated myocardial cells with an island-like appearance between the adipocytes. Analyzing this group of cardiomyocytes, they detected polymorphic nuclear changes in shape, perinuclear vacuolization and accumulation of small granules, which visually appeared similar to lipid droplets. By means of electron microscopy, they speculated on the presence of lipid droplets of various size in the cardiomyocytes. This phenomenon was accompanied by degeneration of intracellular organelles and occasional disruption of the plasma membrane with discharge of intracellular content, including mitochondria, into the interstitial space , all being characteristics of cellular autophagy . Although the microscopic analysis is not conclusive for cardiomyocyte lipid accumulation, a limited lipogenesis (fat droplets accumulation) may occur in myocytes . It is then conceivable that the contractile heart compartment undergoes lipogenesis rather than adipogenesis in ACM. Of note, intracellular fat droplet accumulation has the potential to impact on myocardial function .
Taking advantage of genetic fate-mapping experiments in mice, second heart field-derived progenitor cells, characterized by Isl-1, an early marker of multipotent mesodermal precursors, have been shown to take part to adipogenesis. This finding was corroborated in ACM human hearts by the evidence of coexpression of second heart field markers and adipogenic transcription factors . Notably, Isl-1 cell expansion has been related to Wnt signaling . On this basis, it has been suggested that the adipogenic switch of these precursors may occur before myocyte commitment, depending on the suppression of canonical Wnt pathway signaling by nuclear PG translocation . However, the small number of Isl-1+ pre-adipocytes found in ACM hearts  is a strong limitation challenging this hypothesis. Moreover, the embryologic origin of cardiac progenitors from the second heart field cannot provide an exhaustive explanation for the left-predominant or biventricular forms of ACM .
A further hypothesis involving cardiac precursors is based on the common developmental origin of second heart field RV progenitors and epicardial cells . Epicardial cells, giving rise to non-myocyte stromal components through epithelial-to-mesenchymal transition (EMT) , have been proposed as source of adipocytes in ACM . Indeed, Matthes et al. demonstrated that epicardium-derived cells obtained from neonatal hearts express desmosomes and, if silenced for PKP2, they have increased migration velocity, higher proliferative rate and can be driven to adipogenesis. These cells have been in fact suggested to cause an excess of fibroblasts, adipocytes, or their progenitors, in the myocardial interstitium . As an E-Cadherin negative and α-smooth muscle actin (α-sma) positive subpopulation was identified, the authors hypothesized that this epicardial cell subpopulation may correspond to resident myofibroblasts, or alternatively, cells in which EMT occurred . After PKP2 silencing, an increase in α-sma positive cells was reported, and it was not clear whether this depended on cell proliferation or on a boost of EMT. Indeed, in the EMT process both E-Cadherin and PKP2 are known to be downregulated , raising the intriguing hypothesis that PKP2 loss could be associated to EMT. The epicardium-endocardium gradient of the ACM adipose substitution is a key element supporting this theory .
Alternatively, c-kit+/Sca1+ progenitor cells have also been proposed as adipocyte precursors . Lombardi et al. in 2011 generated transgenic mice overexpressing wild-type and truncated PG. Histological examination showed a moderate increased number of adipocytes, predominantly in the epicardium, and fibrosis in the heart of these transgenic models. Moreover, they exhibited cardiac dysfunction and premature death. The authors isolated c-kit+/Sca1+ cells from PG transgenic mice and they obtained adipogenic differentiation of these cells through the contribution of Wnt signaling . Since c-kit+/Sca1+ cells from PG homozygous knock-out embryos have been shown to be resistant to adipogenesis, and showed normal activation of canonical Wnt pathway, they demonstrated the role of PG in the induction of lipid accumulation in these cells . However, a quantitative evaluation providing evidence on the effective contribution of these cardiac progenitor cells on adipogenesis is currently lacking . Our recent findings revealed that the percentage of c-kit+ preadipocytes in the human ACM myocardium is too low to account for the massive adipocyte presence in ACM hearts, suggesting that only a small proportion of adipocytes in ACM may ultimately originate from c-kit+ progenitors .
Differentiation of pluripotent cells
In 2015, we introduced for the first time a novel hypothesis on the origin of ectopic adipocytes in ACM , identifying Cardiac Mesenchymal Stromal Cells (C-MSC) as an adult, non-contractile cell compartment involved in lipid accumulation and adipogenic differentiation . C-MSC are a large population of supportive cells of epicardial origin, characterized by multipotency. They play a critical role in maintaining healthy heart structure and function. Importantly, C-MSC are involved in cardiac remodeling in pathological conditions . Noteworthy, as for A-MSC, C-MSC can differentiate into adipocytes under appropriate stimuli.
First of all, to understand the cellular origin of adipocytes in ACM hearts, we performed immunohistochemical staining on ACM cardiac tissue, founding cells actively differentiating into adipocytes expressing the typical mesenchymal markers CD29 and CD105 .
Moreover, we isolated C-MSC from ACM and control heart biopsy specimens and demonstrated that C-MSC express desmosomal genes, and therefore suffer the consequence of desmosomal mutations. Importantly, C-MSC from ACM patients’ hearts have been shown to differentiate into adipocytes in culture, by both phenotypic analysis and expression analysis of specific adipogenic genes and relative proteins. We also tested the involvement of the Wnt pathway, increasing its activity by inhibiting Glycogen Synthase Kinase 3 beta (GSK3β), and obtained a reduction in the accumulation of lipid droplets. The adipogenic differentiation of C-MSC occurs with a PKP2-dependent mechanism: the overexpression of PKP2 led to a decrease of lipid accumulation. In contrast, silencing PKP2, we obtained a significant increase in adipogenesis. All together, these findings provide convincing evidence that C-MSC are a source of adipocytes in ACM.
More recently, cardiac Fibro-Adipose Progenitors (FAP) characterized by the expression of PDGFRα have been proposed as the source of both adipocytes and fibrosis , as previously suggested by Paylor and colleagues , and as demonstrated by the expression of either adipogenic or fibrogenic markers. FAP progenitors contribute to the vascular and mesenchymal compartments of the human hearts and descend from epicardial derivatives, PDGFRα being necessary for epithelial-mesenchymal transition . Lombardi et al. demonstrated that the subset of FAP that express the adipogenic marker C/EBP-α also express desmosomal genes, and, if mutated, can differentiate into adipocytes through the suppression of canonical Wnt signaling in ACM caused by Dsp haploinsufficiency. Cardiac FAP give rise to approximately 40% of adipocytes in the heart of an ACM mouse model, indicating the possibility of the co-existence of another cellular origin of lipid accumulation in ACM. Further experiments will be needed to ascertain communalities and differences of C-MSC  with FAP subpopulation  in ACM pathogenesis.
Research in ACM has recently progressed both from clinical, genetic and mechanistic standpoints. Pathogenic information is however still missing for a complete understanding of ACM etiology, even if scientific consensus exist on key points as follows:
2) Aberrant repair after cardiomyocyte death takes place through signaling pathways [14,16] specifically driven by desmosomal gene mutations, resulting in over-deposition of extracellular matrix and desmosome-expressing-cell differentiation into adipocytes.
Relevant steps forward have been undertaken over time in the understanding of fibro-fatty pathogenesis and cellular effectors. As outlined in this manuscript, cardiomyocyte trans-differentiation is unlikely to happen. Furthermore, cardiac progenitors are present at a too low rate in an adult heart to account to a wide effect contribution. Most likely, epicardial cells, highly expressing desmosomal genes, as all epithelia, undergo EMT, of which Wnt represents a contributing pathway, resulting in a centripetal invasion of mesenchymal cells from the epicardium to the endocardium, both in ACM and in physiological conditions. C-MSC, are characterized by a lower expression of desmosomal genes with respect to epithelial cells; particularly, the further reduction of desmosomal proteins in ACM C-MSC may be the culprit for adipogenic switch. Indeed, C-MSC or a C-MSC sub-population are well known to possess a differentiation ability into adipocytes (Figure 1).
Figure 1: Hypothesized developmental steps of cardiac adipocytes in ACM. The epithelial component of epicardium undergoes epithelial-to-mesenchymal transition (EMT), giving rise to a mesenchymal stromal cell (C-SC) population. C-MSC differentiate firstly into pre-adipocytes, accumulating lipid droplets, and, subsequently, into mature adipocytes.
Whether adipogenesis in ACM hearts merely represents distorted wound healing or has a distinct functional role in disease pathogenesis, is still not known. Electrically inert tissue may induce compulsory current routes though remaining cardiomyocytes, exacerbating ACM arrhythmic phenotype. Moreover, adipose tissue is known to exert a strong paracrine activity, influencing both inflammation and cardiomyocyte contractility performances by attenuating intracellular calcium ion levels .
Overall, a comprehensive understanding of the fibro-fatty substitution process in ACM is today a matter of active scientific debate with potential relevant repercussion at a clinical level, since the cellular component giving rise to substrate defects may represent both an essential tool for mechanistic studies of ACM pathogenesis and a possible novel therapeutic target. Indeed, the in vitro study of the adipogenesis process in the cells responsible for the adipose tissue deposition in ACM may give scientists the tools to counteract it by directly revert involved pathways. Another interesting option may be to assess a potential therapeutic effect by means of high-throughput screening of FDA-approved drugs or new druggable compounds. Perspectively, targeted cell-specific administration could be envisaged, in order to avoid systemic effects.
Thanks to Federico Zoofito for technical support with the figure, and to Dr. Aoife Gowran for English revisions. This work was supported by the National Institutes of Health [Ricerca corrente 2016] and by the Telethon Grant GGP16001 to Prof. Giulio Pompilio.