Arrhythmogenic right ventricular dysplasia type 1 and mutations in transforming growth factor ?3 gene regulatory regions: a breakthrough?

See article by Beffagna et al.(pages 366–373) in this issue. Arrhythmogenic right ventricular dysplasia (ARVD) is the second most common cause of sudden unexpected death in young adults [1], responsible for ∼20% of sudden deaths in this population [2]. Subsequent to the description of the first case in 1961 [3], great progress has been made in the diagnosis, evaluation and management of the condition. The hallmark of ARVD is right ventricular atrophy with fibro-fatty infiltration. There is approximately threefold male predominance, and typical abnormalities include signs of right ventricular thinning and dysfunction and fragmented, low-amplitude late-QRS potentials in the right precordial leads identified on standard (“epsilon waves”) or signal-averaged electrocardiograms [2]. Familial occurrence is common, and in ∼30% of cases evidence has been found for autosomal-dominant inheritance with variable penetrance [4]. A variety of chromosomal locations have been identified in various kindreds with the condition (for review, see Ref. [2]). Eight subtypes identified as ARVD1–8 have been recognized. The specific gene underlying ARVD has been identified for a small number of subtypes. ARVD2, with features perhaps more typical of a catecholamine-dependent ventricular tachyarrhythmia syndrome than of ARVD, is associated with a mutation in ryanodine receptor-2 [5]. All other mutations described to date have been in genes encoding cytoskeletal proteins involved in desmosomal cell–cell interaction: plakoglobin (γ-catenin), which interacts with α- and β-catenin to mediate cell–cell adhesion in cardiac and dermal–epidermal junctions and for which homozygous deficiency causes Naxos disease [6]; desmoplakin, which attaches intermediate filaments to desmosomes and which causes the autosomal-dominant ARVD8 [7,8]; and plakophilin-2, an armadillo repeat protein strongly expressed in desmosomes [9]. Presumably because of the particularly important roles of desmosomes in heart and skin, desmosomal protein mutations can produce specific cardio/cutaneous phenotypes. The first form of ARVD connected to a chromosomal abnormality was ARVD1, linked in 1994 to a locus on 14q23–q24 [10]. Five potential candidate genes, including transforming growth factor β3 (TGFβ3), were identified in the region of interest, but detailed screening failed to identify mutations in the coding regions of any of these [11,12]. In the present issue of Cardiovascular Research, Beffagna et al. [13] report the results of further investigations searching for mutations in the non-coding region of the TGFβ3 gene in AVRD1 kindreds. In one family, a mutation in the 5′-untranslated region (UTR) was identified. The mutation was found in all nine members of the kindred with proven ARVD, and was also present in three of over 40 asymptomatic members. An unrelated proband was found to have a mutation in the 3′-UTR. Extensive sequencing failed to identify mutations in the TGFβ3 gene in two other kindreds with the same ARVD1 linkage pattern. Both the 5′- and 3′-UTR mutations were found to increase luciferase reporter activity by about 2.5-fold, and neither mutation was found in 300 control patients (600 chromosomes). The results of this study are intriguing. They clearly suggest that TGFβ3 overexpression resulting from a regulatory site mutation produces ARVD1. This is an extremely interesting finding, with potentially major significance for our understanding of the pathophysiology of ARVD and of the function of TGFβ3. However, a number of issues need to be resolved before TGFβ3 overexpression can be accepted as the molecular basis of ARVD1: (1) The lack of mutations identified in the TGFβ3 gene in two kindreds with typical ARVD1 linkage is of concern. It may be that the mutation lies in a linked site outside the TGFβ3 locus that nevertheless regulates TGFβ3 expression. Another possibility is that there is another gene closely linked to TGFβ3 that is mutated in these kindreds without identified TGFβ3 mutations and causes ARVD. Still another possibility is that the TGFβ3 mutations identified by Beffagna et al. are not causal for AVRD, but are closely linked to another mutated gene that causes ARVD. (2) Confirmation of in vivo overexpression of TGFβ3 in the mutation carriers is not available. Demonstration of TGFβ3 overexpression in cardiac biopsy samples from affected patients would be helpful in confirming the functional relevance of the in vitro promoter studies. (3) The biological connection between TGFβ3 overexpression and ARVD remains to be established. TGFβ3 is a member of the TGFβ family, which consists of multifunctional cytokines with widespread distribution. Targeted deletion of TGFβ1 results in extensive, eventually lethal inflammation ∼3 weeks after birth, whereas TGFβ2 and TGFβ3 mutations cause perinatal death [14]. TGFβ3 ablation impairs palatogenesis and delays pulmonary development, with the Alk-5/Smad2 pathway playing a key role in TGFβ3-driven palatal fusion [15]. TGFβ3 is also involved in trophoblastic [16] and cardiac valve [17] development. Beffagna et al. argue for a role of TGFβ3-induced fibrosis in AVRD. TGFβ3 overexpression has been detected in the skin of patients with scleroderma and implicated in increased collagen production [18]. However, TGFβ3 antagonizes TGFβ1-induced extracellular matrix (ECM) production by fibroblasts [19] and suppresses TGFβ1-induced scar formation in cutaneous wounds [20]. Interestingly, TGFβ3 appears to reduce the production of a variety of tight-junction integral membrane proteins in Sertoli cells, including occludin, zona occludens-1, and claudin-11 [21], by a p38 mitogen-activated protein kinase pathway [22]. It is conceivable that ECM or tight-junction modulation by TGFβ3 is pathogenic in ARVD1, but this possibility remains speculative for the moment. Given the widespread tissue expression of TGFβ3, the apparently quite selective cardiac phenotype of AVRD1 requires explanation. Further work, particularly in genetically engineered models of TGFβ3 overexpression, would be of great interest. Despite the many questions that remain to be answered, the results reported by Beffagna et al. have the potential to lead to many important new discoveries about the molecular basis of ARVD1 and the biological function of TGFβ3. We look forward to the further unfolding of this fascinating story with great interest. The continuing development of insights into genetic causes of sudden cardiac death promises to provide not only new insights into cardiac physiology and pathophysiology, but potential new approaches to the diagnosis, prevention, and therapy of the conditions underlying this important public health problem [23,24]. Supported by the Canadian Institutes of Health Research and the Quebec Heart and Stroke Foundation.