Intramural clefts and structural discontinuities in Brugada syndrome: the missing gap?

This editorial refers to ‘Normal interventricular differences in tissue architecture underlie right ventricular susceptibility to conduction abnormalities in a mouse model of Brugada syndrome’ by A. Kelly et al., pp. 724–736. Brugada syndrome (BrS) is an inherited syndrome characterized by an increased risk for ventricular arrhythmias and sudden death occurring mostly during rest or sleep, in particular in young and otherwise healthy male individuals. The typical Brugada ECG pattern of ST-segment (J point) elevation in the right-precordial leads may be variably present, and can be unmasked or increased after administration of Class 1A or 1C anti-arrhythmic sodium channel blocking drugs (ajmaline or flecainide), or during exercise.1 Cardiac conduction disease is often also observed in BrS patients, evidenced by increased PQ- and QRS-duration on the ECG.2 In ∼20% of BrS patients, SCN5A mutations leading to reduced cardiac sodium channel function are identified. The consequent decreased action potential upstroke velocity underlies the observed prolonged ECG conduction indices, but the right-precordial ST-segment elevation and its relation to arrhythmogenesis is less well understood. There has been considerable debate on the mechanism underlying BrS, with abnormalities in both repolarization and depolarization suggested.1 It has been proposed that the action potential duration in the subepicardium of the right ventricular outflow tract (RVOT) is extremely short due to fast phase-1 repolarization. The short subepicardial action potential would result in systolic current sink leading to ST-segment elevation on the body surface ECG, explaining the BrS sign. Moreover, the plateau phase of the action potential of subendocardial myocytes would provide current to activate the already repolarized subepicardial myocytes causing focal activity (Phase-2 re-entry).3 Accordingly, mice have been suggested to be a good model for BrS since their ECG resembles that of BrS patients, namely J point elevation and a negative T wave.4 In wild-type mice, however, the morphology of the ECG can be fully explained based on the shape of the action potential and RVOT myocardium fully activates before the onset of the J wave.5 Moreover, follow-up clinical studies have shown that the subepicardial action potential is not shorter in BrS patients when compared with controls.6 In addition, late activation of the RVOT myocardium after the end of the QRS complex in addition to highly fractionated local electrograms have been demonstrated in BrS patients.7,8 Both are indicative of abnormal discontinued ventricular conduction and form the basis for the hypothesis that BrS is a conduction disease with structural discontinuities in the RVOT myocardium as a prerequisite.9 Here, cardiac fibrosis in combination with reduced sodium current secondary to SCN5A mutations would facilitate discontinued conduction in BrS patients. In this issue of Cardiovascular Research, Kelly et al.10 explore an alternative mechanism underlying right ventricular conduction slowing in BrS, by employing heterozygous Scn5a-deficient (Scn5a+/−) mice. Based on their findings, they propose that structural heterogeneities intrinsic to the right ventricle, instead of or next to fibrosis, predispose to conduction slowing in the setting of reduced sodium channel function. Histological investigations revealed a higher number of intramural clefts in the right ventricle as compared with the left ventricle whereas fibrosis levels were not different. These clefts run parallel to the subendo- and subepicardium in between cardiomyocyte layers, thereby separating myocardial bundles within the transmural wall and potentially creating multiple activation waves in the myocardium.11 If this separation causes two myocardial bundles to be connected via a very small bundle of cardiomyocytes this may cause activation delay (see Figure 1) . Moreover, conduction block due to current to-load-mismatch may occur12 when sodium current is reduced to such an extent that activation of neighbouring cardiomyocytes fails.13,14 Thus, these clefts may explain the delayed activation of the subepicardial myocardium observed by Kelly et al. in Scn5a+/− mice. Using a start-of-the-art approach combining wide field epifluorescence and two-photon microscopy, they observed slowed transmural conduction particularly in the right ventricle of Scn5a+/− mouse hearts, while the rise time of the optical action potential (a surrogate for upstroke velocity) was unchanged. In line with this, clefts were predominantly observed within the myocardial wall of the right ventricle explaining why reduced sodium current slows down conduction more severely in the right than left ventricle as is observed in the majority of Scn5a-deficient mice.15 Subsequent in silico testing supported this hypothesis and showed that these intramural clefts could explain the observed delayed transmural activation in the right ventricle of Scn5a+/− mice. Using this approach, Kelly et al. have provided a potential novel mechanism underlying preferential right ventricular conduction slowing in the setting of Scn5a deficiency. Their findings moreover reinstate the notion that mice may in fact mimic the substrate for BrS, especially when a mutation in Scn5a is present. Effect of clefts on transmural conduction. Panels (B–D) have been redrawn from the original figure from Kelly et al.10 depicting a section of the right ventricle (panel A). The presence of clefts (in white) may separate activation waves leading to delayed activation of the subepicardium (B). When sodium current (INa) is reduced, activation waves may block leaving part of the subepicardium inactivated (C). This causes a flow of depolarizing current from the activated myocardium to the inactivated myocardium leading to ST-segment elevation in the extracellular space as seen in BrS patients. If the inactivated subepicardium becomes activated by a wave front from a different direction this may set the stage for reentry (D). Although the elegant study by Kelly et al. provides an answer to why conduction slows down specifically in the right ventricle when sodium current is decreased, it also raises several questions. Are these clefts present in vivo or are they merely the result of the tissue sectioning process? Why is the number of clefts higher in the right ventricle than in the left? Are these clefts also more pronounced in the RVOT? To exclude the possibility of clefts being an artefact, Kelly et al. quantified clefts in the intact heart using two-photon microscopic imaging before tissue section. These results supported the histological observations thus making it unlikely that the clefts are merely an artefact. The reason why clefts are predominantly observed in the right ventricular wall is unclear. During development, the compact wall of the right and left ventricle forms from early chamber myocardium upon epicardial signalling.16 In addition, part of the embryonic trabecular myocardium will become part of the compact wall via a process called compaction. Whether compaction is involved in cleft formation is unclear. However, compaction differs between species and may vary between ventricles as well, possibly contributing to the different numbers of clefts observed between the right and left ventricle. Another explanation could be mechanical strain differences between the left and right ventricle as a result of early and late activation, respectively. The RVOT myocardium is the last to activate and, therefore, it would be interesting to see whether the RVOT contains more intramural clefts than the left ventricle as well. Intriguingly, evidence is accumulating that sodium channels may play additional roles in cell adhesion and extracellular matrix regulation through their interaction with other proteins within the macromolecular complex.17 Although speculative, these non-canonical actions may play a (additional) mechanistic role in BrS. It remains to be seen whether the intraventricular differences in tissue architecture observed in mice are also present in human right ventricular myocardium. Furthermore, the question arises to what extent Scn5a+/− mice constitute a relevant model of BrS. SCN5A mutations typically lead to reduced sodium channel availability, either through decreased trafficking and membrane surface channel expression, or through altered channel gating properties including disruption of activation, accelerated inactivation, and impaired recovery from inactivation.2 BrS-associated SCN5A mutations may not always lead to a reduced SCN5A expression in the heart, as is the case in Scn5a+/− mice. In the latter, less Nav1.5-based channels are present on the cell membrane, whereas this is not necessarily the case in the setting of SCN5A mutations. Whether this limits the transferability of the findings by Kelly et al. to BrS-associated mutations remains to be elucidated. However, the fact that TTX resulted in similar results as in Scn5a+/− mice suggests a more generally applicable mechanism. In conclusion, the observation of intramural clefts in the right ventricle comprises a potential gap in our knowledge on why conduction slows down predominantly in the right ventricle in Scn5a deficient mice. Such clefts in the human right ventricle would challenge cardiac conduction and contribute to the substrate for arrhythmias in BrS patients. It will be interesting to see the potential clinical implications, including the presence of similar clefts in human heart, possible modulatory effects of age and co-morbidities, and potential relevance for other arrhythmia syndromes with right ventricular involvement, i.e. arrhythmogenic right ventricular cardiomyopathy. Conflict of interest: none declared. This work was supported by an Innovational Research Incentives Scheme Vidi grant from The Netherlands Organization for Health Research and Development (ZonMw) (91714371 to C.A.R.) and the Dutch Heart Foundation (2016T047 to B.J.B.).