Tranilast and hypertensive heart disease

The pharmacological characteristics of tranilast (N(3,4dimethoxycinnamoyl)-anthranilic acid) were first described as early as the 1970s [1]. The drug was initially introduced for the treatment of dermatological disorders such as hypertrophic scar formation and keloid [2,3]. Because of its ability to suppress mast cell degranulation and histamine release, tranilast is also used in allergic diseases such as allergic rhinitis, atopic dermatitis and asthma bronchiale [4–6]. In recent years, the effects of tranilast have been investigated in different models both in vitro and in vivo. There are a number of results indicating beneficial effects of tranilast in disorders involving matrix deposition and fibrosis. Those effects appear to be caused by several mechanisms: (i) inhibition of matrix protein synthesis (e.g. collagen); (ii) reduction of cell proliferation (e.g. fibroblasts and smooth muscle cells); and (iii) inhibition of inflammatory responses that result in fibrosis Tumour growth factor (TGF)β1 is known to play an important role for the synthesis of matrix proteins and cardiac fibrosis [7]. In cell culture experiments, tranilast inhibits the TGFβ1-mediated synthesis of collagen [2,8]. The expression of TGFβ1 mRNA decreased after tranilast treatment in vitro and in vivo [8–10]. Additionally, mRNA for the type 2 TGFβ1-receptor is reduced by tranilast [11]. Another in vivo model of fibrosis with high clinical relevance is diabetic nephropathy. Mifsud et al. [12] treated diabetic rats in the phase of established kidney disease and found a strong reduction of tubulointerstitial fibrosis, tubular atrophy and a near normalization of albuminuria. Anti-proliferative effects of tranilast have been demonstrated in cardiac fibroblasts, vascular smooth muscle cells [13–16] and even in mouse lung cancer cells [17]. Noguchi et al. [18] described reduction of tumour growth in a model of squamous carcinoma in vivo. The anti-proliferative effect in vivo may be at least partly attributed to inhibition of tumour angiogenesis [19]. On the other hand, tranilast also induces the tumour suppressor gene p53 [20]. The pathways of the anti-fibrotic and anti-proliferative effects are incompletely understood. A microfibril- associated glycoprotein has been identified as a possible target of tranilast [21]. Growth inhibition of vascular smooth muscle cells appears to be mediated by inhibition of platelet-derived growth factor homodimer-BB receptor binding [22]. There is also evidence that tranilast suspends the cell cycle of fibroblasts at the G0/G1 phase. Tranilast induces the cyclin-dependent kinase (CDK) inhibitor p21 which suppresses CDK2 and CDK4 activity and thus inhibits proliferation [15,20]. Several clinical studies involving humans have been performed. In a recent pilot study in humans with advanced diabetic nephropathy, tranilast was reported to slow the decline in glomerular filtration rate over a 12-month period [23]. Because tranilast combines anti-fibrotic and growth-inhibiting properties, it appeared provocative to evaluate its possible effects on the prevention of neointima formation and restenosis after coronary artery dilatation. After promising results were obtained from animal experiments [10,24,25] and preliminary clinical studies (TREAT 1 and 2) [26], the large PRESTO trial (involving more than 10 000 patients) had to be cancelled ahead of schedule because no beneficial effect of tranilast could be found [27]. At a high price, we have learned once more that hypotheses arising from cell culture and animal experiments do not necessarily correspond to large-scale clinical settings in humans. In addition to anti-proliferative and anti-fibrotic effects, an anti-inflammatory potency of tranilast (apart from mast-cell stabilization) is discussed but the relevant mechanisms remain poorly understood. Capper et al. [28] demonstrated an inhibition of pro-inflammatory activity of monocytes after incubation with tranilast. Inhibition of interleukin (IL)-1β-induced expression of the monocyte chemo-attractant peptide (MCP)-1 was reported [29]. Tranilast also prevents leukocyte accumulation around porcine coronary arteries after stenting [10]. Recent cell culture experiments investigated the molecular mechanisms of anti-inflammatory tranilast action. In human umbilical vein endothelial cells the effects of tranilast on the pro-inflammatory transcription factor nuclear factor-κB (NF-κB) were investigated [30]. NF-κB is essential for the induction of numerous pro-inflammatory genes such as endothelial cell adhesion molecules, namely intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), E-selectin and cytokines (e.g. TNFα and -β, interleukin-2, -6 and -8) [31,32]. Expression of ICAM-1, VCAM-1 and E-selectin mediate the adhesion of monocytes, lymphocytes and granulocytes to the vascular endothelium and thus play an important role for inflammatory processes [33]. TNFα-induced expression of endothelial VCAM-1, ICAM-1 and E-selectin, and the secretion of IL-6 is dose-dependently inhibited by tranilast whereas NF-κB independent processes remain unchanged. Moreover, the expression of the transcriptional coactivator CBP (cAMP response element binding protein binding protein) is inhibited by tranilast, resulting in a loss of interaction between NF-κB and CBP. It is therefore suggested that tranilast inhibits the expression of adhesion molecules via inhibition of the interaction between NF-κB and its transcriptional coactivator CBP [30]. Hypertensive heart disease entails structural remodelling of cardiomyocyte and non-myocyte cells. This remodelling involves an inflammatory component mediated by adhesion molecules [34,35]. Transcriptional activation of the genes of those molecules is tightly regulated by NF-κB. Consecutive monocyte/macrophage infiltration and activation results in activation of the renin–angiotensin–aldosterone system (RAS) and synthesis of TGFβ1 [36]. Inhibitors of the RAS are well known to attenuate cardiac remodelling [16,37]. Angiotensin II itself, which is primarily activated in many hypertensive mouse models, is known to induce TGFβ1 in different cardiac cells [38,39]. TGFβ1 directly stimulates the formation of extracellular matrix including collagen (type I and III) and also blocks extracellular matrix degradation by stimulating protease inhibitors, including platelet activator inhibitor-1 (PAI-1) and tissue inhibitor of metalloproteinase (TIMP) [40]. The excess of ventricular collagen partly results from insufficient collagen degradation by interstitial collagenases and metalloproteinases (MMPs), whose activity is regulated by specific tissue inhibitors of metalloproteinases (TIMPs). Finally, the fibrosis results in progressive abnormalities of diastolic ventricular filling and relaxation, systolic dysfunction and conduction disturbances, thus contributing to the clinical risk associated with left ventricular hypertrophy. Cardiac fibrosis is recognized as a key process in the progression from compensated cardiac hypertrophy to cardiac dysfunction and heart failure [41]. Of course, this suggested line of events is oversimplified and many other factors are involved in the pathogenesis of hypertensive heart disease [42]. In deoxycorticosterone acetate (DOCA) salt-sensitive hypertensive rats, the cardiac inflammatory response was recently analysed [43]. It was demonstrated that NF-κB, VCAM-1, the platelet-endothelial cell adhesion molecule-1 and MMPs are upregulated in the heart of DOCA-salt rats. However, NF-κB decoy oligodeoxynucleotides were unable to reduce gene expression of TGFβ1 and fibrosis in a model of nitro-l-arginine methyl ester-induced cardiac fibrosis [44], suggesting the existence of an additional NF-κB-independent pathway for cardiac fibrosis. This is supported by the finding that collagen is already upregulated after 1 week of DOCA-salt treatment whereas NF-κB binding activity rose only after the second week [43]. In another model of mineralocorticoid-induced hypertension, the time course of cardiac fibrosis development was described. Interestingly, the first change to be noted was a significantly elevated immunohistochemical level of collagen III [45]. The fact that blood pressure is not yet significantly elevated at that time supports previous findings that the origin of cardiac fibrosis is humoral rather than haemodynamic, whereas cardiomyocyte hypertrophy is clearly associated to hypertension [42]. There are two studies performed in vivo showing that tranilast is capable of suppressing myocardial fibrosis in hypertensive animals. The study by Pinto et al. [9] compared tranilast to losartan in hypertensive TGR (mRen2)27 rats, a transgenic rat model characterized by increased cardiac angiotensin II concentrations [46]. Both drugs prevented the increase in left ventricular TGFβ1 mRNA, collagen accumulation and left ventricular hypertrophy without lowering blood pressure. Moreover, even a significant positive effect on survival was seen after treatment with tranilast. The second study was designed to investigate whether tranilast would also inhibit collagen accumulation in a rat model of renovascular hypertension (Goldblatt two-kidney, one-clip model). In accordance with the previous study, a preventive effect on cardiac fibrosis was shown [16]. In this issue of the journal, Kagitani et al. [47] evaluated tranilast effects in DOCA-salt hypertensive rats. In its approach, the study goes beyond previous studies, focusing on anti-inflammatory aspects of tranilast action in the course of cardiac fibrosis development. In comparison to previous studies, only one-quarter of the tranilast dose was used (100 mg/kg per day). The study confirms that inflammatory processes are involved in fibrosis development in this model. Apart from increased expression of TGFβ1, PAI-1 and collagen accumulation the presence of the inflammation-related peptides MCP-1 and IL-6 is detected. For the first time, a suppression of inflammatory processes by tranilast could be shown in vivo. MCP-1, IL-6 and monocyte/macrophage infiltration could be reduced by tranilast in a blood pressure-independent manner. Kagitani et al. [47] suggest that reduction of collagen synthesis and fibrosis might be a consequence of the attenuation of the inflammatory response. This concept certainly merits further attention. It would be particularly interesting to define the effects of tranilast on the pro-inflammatory transcription factor NF-κB and related adhesion molecules that are known to be involved in the DOCA-salt model of cardiac fibrosis, and also known to be influencable by tranilast. Kagitani et al. [47] propose an anti-MCP-1 or MCP-1 knockout strategy to further establish the line from macrophages and inflammation to fibrosis. However, the study by Fujisawa et al. [45] demonstrated that part of the collagen deposition (type III) precedes inflammation, thus suggesting a second pathway independent of inflammation, possibly via direct action of angiotensin II or inflammation-independent release of TGFβ1. This finding could not be confirmed by Kagitani et al. [47]. Pinto et al. [9] and Hocher et al. [16] found angiotensin-converting enzyme (ACE) inhibitors (i.e. angiotensin II receptor antagonists) to be about as effective as tranilast. Apart from many other effects, blocking of the RAS also prevents the initial inflammatory cell response in models of hypertensive heart disease [48]. A possible additive effect of tranilast combined with ACE inhibitors, angiotensin II receptor antagonists or antagonists of aldosterone remains to be evaluated, although an antagonizing effect on angiotensin II has also been described for tranilast on its own [49]. This angiotensin II antagonizing effect of tranilast is not strong enough to cause lowering of blood pressure even at a four-fold higher dose compared to the present study [9,16]. The study by Kagitani et al. [47] adds another piece to the puzzle concerning pathways leading from hypertension and the activation of RAS to cardiac fibrosis. Tranilast, most likely in addition to its other pharmacological approaches, is a promising compound against fibrotic remodelling of the heart, and possibly other organs such as the kidney in diabetic nephropathy. Negative results from the first clinical studies should not discourage further research, especially because no major side-effects of tranilast have been reported to date.