Atrial fibrillation and fibrosis: role of connective tissue growth factor

This editorial refers to ‘Angiotensin II-mediated up-regulation of connective tissue growth factor promotes atrial tissue fibrosis in the canine atrial fibrillation model’ by M. Kiryu et al., on page 1206 Atrial fibrillation (AF) induces significant electrophysiological and structural changes in the myocardium and substantially compromises the microvascular blood flow.1 Atrial fibrosis comprises a hallmark of AF-dependent structural remodelling and, to a large extent, depends on increased activity of the renin–angiotensin system (RAS).2 In addition to experimental data, histological studies have also shown that the amount of fibrotic tissue is increased in fibrillating human atria.1,3 Fibrosis isolates groups of atrial myocytes as well as individual myocytes and thus impairs cell-to-cell coupling, leading to inhomogeneities in intra- and interatrial conduction and slowed conduction velocity. Both mechanisms favour the inducibility of prolonged episodes of AF.1 Atrial fibrillation itself is promoted by atrial fibrosis.1 Atrial fibrillation can develop secondary to underlying disease as observed, e.g. in heart failure models, and appears to be associated with increased atrial angiotensin II (AngII) levels.1 Atrial fibrillation has also been associated with increased plasma and tissue levels of AngII,2,4 which were shown to result from increased expression/activity of angiotensin-converting enzyme.3 Substantial alterations in atrial and ventricular expression levels of other components of the RAS have been observed during AF and include changes in the AngII receptor subtypes, AT1R and AT2R. Interestingly, the expression patterns differ between paroxysmal and persistent AF, which points to a profound impact which the duration of AF exerts on structural remodelling.1 In this issue of the journal, Kiryu et al.5 report on a sustained (6 weeks) canine in vivo rapid pacing model to verify the hypothesis that AF-dependent increase in AngII levels leads to connective tissue growth factor (CTGF) induction and promotes subsequent fibrosis. Six weeks of AF led to significant interstitial fibrosis. In accordance with this finding, the authors observed increased synthesis of extracellular matrix (ECM) proteins such as collagen 3 and fibronectin. Increased atrial tissue fibrosis and ECM accumulation were associated with a strong induction of CTGF expression. The induction of ECM synthesis (collagen production) occurred in parallel with the onset of fibrosis, and both could be significantly reduced by the administration of the AT1R-blocker, olmesartan. The results clearly demonstrate that, in the absence of a concomitant disease, there is an AF-dependent induction of pro-fibrotic gene expression and fibrosis, which is associated with an induction of CTGF expression. The most interesting findings of Kiryu et al.5 are, however, (i) that the left and right atria showed a very similar response to both AF and olmesartan treatment during AF, (ii) that even during a 6-week period of sustained AF, the development of fibrosis could be prevented to a significant extent, and (iii) that ECM production and tissue fibrosis occur in the absence of any increase in atrial transforming growth factor-β1 (TGF-β1) expression. Previous studies have proposed CTGF as a potent pro-fibrotic factor in various pathophysiological conditions, including AF.6 Increased amounts of CTGF in the left atrial tissue from patients with AF were associated with increased fibrosis and, notably, with increased tissue levels of AngII.7 Angiotensin II has been shown to induce CTGF via activation of Rac1 and NADPH oxidase, leading to increased expression of target genes.7 Angiotensin II, however, can induce CTGF also via TGF-β1 and Smad-signalling. In a porcine AF model, both the tissue levels of AngII and CTGF were increased in the left and right atria.6 Similarly, AngII induced CTGF expression in perfused rat hearts.6 This demonstrates that activation of the atrial AngII system by lone AF, and/or by existing concomitant cardiovascular disease, can provide a ‘morphologic, arrhythmogenic substrate’ for re-entry arrhythmia and allow AF to persist. Recent data suggest, however, that CTGF alone is not sufficient to promote the development of fibrosis in all pathophysiological settings. Despite the fact that enhanced CTGF expression is regularly associated with fibrosis, in vivo evidence proving any causality between elevated CTGF expression and induction of fibrosis is still lacking. Notably, over-expression of CTGF in cardiomyocytes failed to induce fibrosis.8 Likewise, in a rapid pacing-induced model of heart failure, there was no evidence of fibrosis, in spite of substantially increased levels of CTGF. In contrast, Shinagawa et al.,9 using a very similar model of pacing-induced heart failure, observed the occurrence of typical atrial fibrosis. Although the reasons for these discrepancies are yet to be fully established, it is reasonable to assume that the simultaneous (synergistic) activity of several factors is required to turn CTGF into a pro-fibrotic and ‘pro-arrhythmogenic’ factor. These concomitant factors may include pre-existing structural alterations of the myocardium (underlying disease), systemic activation of the RAS, and, of course, AF duration. Diabetes mellitus is one such established risk factor of AF and was shown to promote fibrosis, activation of AGE/RAGE (advanced glycation endproduct–receptor for AGE) signalling, and CTGF expression, all of which are sensitive to sartan treatment.10 The interesting questions that arise from the present study and may stimulate future work are (i) whether CTGF requires additional synergistic factors to become pro-fibrotic and what, in fact, the identity of these factors might be, and (ii) what is the maximum duration, if there is any, that keeps fibrosis subject to reverse remodelling (e.g. by AT1R-blockade)? Conflict of interest: none declared.