The p38 Mitogen-Activated Protein Kinase Pathway-A Potential Target for Intervention in Infarction, Hypertrophy and Heart Failure

Abstract

The p38 mitogen-activated protein kinases (p38s) are stress activated ser/thr kinases. Their activation has been associated with various pathological stressors in the heart. Activated p38 is implicated in a wide spectrum of cardiac pathologies, including hypertrophy, myocardial infarction, as well as systolic and diastolic heart failure. In this review, the specific contribution of different isoforms of p38 kinases to cardiac diseases as well as TAB-1 mediated non-canonical activation pathway are discussed as a rationale for inhibiting p38 activity to treat cardiac hypertrophy, ischemic injury and heart failure. Finally, a summary of current clinical trials targeting p38 kinases in cardiovascular diseases is provided to highlight the potential promise as well as existing challenges of this therapeutic approach.

The p38 mitogen-activated protein kinases (p38s) are stress-activated kinases that together with the extracellular signal-regulated kinases 1/2 (ERK1/2), the c-Jun N-terminal kinases (JNKs) and extracellular signal-regulated kinase 5 (ERK5) form the 4 principal limbs of the mitogen-activated protein kinase (MAPKs) system. The MAPKs are serine/threonine protein kinases that can be activated by ligands of G-protein-coupled, and growth factor, receptors, as well as other stressors such as reactive oxidative species.

Isoforms of p38 and their mechanism of activation

Four p38 isoforms (α, β, γ and δ) exist which have preserved structure but variable sensitivity to pharmacological inhibition. p38α and β have high sequence homology and share sensitivity to pharmacological inhibition by pyridinyl imidazole molecules (such as SB203580) but have only 60% homology with p38γ and δ which are resistant to inhibition by SB203580[1].

All investigators agree that p38α exists in the myocardium of rodents, large animals and humans[2–9]. Similarly, amongst the few investigators that have performed a systematic examination, there is general agreement that p38γ is preferentially expressed in cardiac muscle and is present at all stages of development[9–11]. Although several earlier studies also claimed p38 expression is rather restricted to skeletal muscle[12, 13]. In contrast, it is generally accepted that p38β and p38δ are less abundant in the myocardium[2, 10, 14].

All four isoforms have a Thr-Gly-Tyr (TGY) dual phosphorylation motif/epitope which is used by investigators to infer activation. The traditional view is that this dual phosphorylation event is achieved by upstream, dual specificity, MAPK kinases (MAPKKs) or MKKs. The major activators of p38 in vivo are MKK3 and MKK6[15, 16] with perhaps some contribution from MKK4[17, 18]. When p38-MAPK is in the non-dual phosphorylated, inactive conformation the loop is thought to reside in the peptide-binding channel that lies in the cleft between amino- and carboxy-terminal lobes of the kinase. In addition, there is a misalignment of these lobes, that prevents the co-operation between Lys⁵³, in the N-terminal lobe, and Asp¹⁶⁸, in the C-terminal lobe, imperative to the binding and stabilisation of ATP[19, 20]. The prediction therefore is that the non-dual phospho-form of p38-MAPK is inactive as a result of steric obstruction of the peptide-binding channel and low-ATP affinity. The MKK-induced, dual phosphorylation event is thought to cause the activation loop to refold[21] and move out of the peptide-binding channel. This movement is then thought to exert a “crank-handle” effect on the overall tertiary structure of the kinase reorientating the N-, and C-, terminal lobes so that Lys⁵³ and Asp¹⁶⁸ move towards one another by 2.5–5Å enabling the cooperation necessary for ATP binding[19] and restoring the active conformation of the “Catalytic Spine” of the kinase[22].

Thus based on structure and function, ATP binding occurs after, and not before, dual phosphorylation of the activation loop. Nonetheless, and of particular relevance to the heart, a number of lines of enquiry have suggested that p38s can also auto-activate through at least 2 distinct mechanisms, one involving a scaffold protein known as TGF-β-activated protein kinase 1 binding protein 1 (TAB1)[20, 23] and the other a priming phosphorylation of Tyr323 by a tyrosine kinase of the SYK-related family[24, 25].

The TAB1-faciliated autoactivation of myocardial p38 has now been described as a non-canonical mode of p38 activation by a number of independent groups and always seems to lead to adverse consequences[26–31]. This mechanism of activation, which is likely confined to p38α, is worthy of further discussion since it may allow circumstance-selective inhibition of p38 activity, which cannot be achieved with the current agents in early phase clinical trial (see below).

TAB1 is a scaffold protein that promotes the autoactivation of another kinase, TGF-β-activated protein kinase 1 (TAK1), which is upstream of p38 and is capable of directly activating MKK3 and MKK6[32]. Furthermore TAB1 is a p38 substrate, suggesting it may also be a component of closed feedback loop regulating p38 activity[32, 33]. This complexity obfuscates data interpretation. Nonetheless, TAB1 and p38 associate during myocardial ischaemia[26–28] and this association alone is capable of causing p38 activation[29, 30, 34]. Furthermore, a natural splice variant of TAB1 exists that can perform this function without interacting with TAK1[35]. Finally, the characteristic readout of autoactivation, a reduction in p38 dual phosphorylation when the ATP-binding site is occupied by a catalytic site inhibitor, is commonly observed when p38 is activated by cardiac stresses such as ischaemia (for some examples see figure 2 of Maulik et al[36], figure 4 of Clanachan et al[37], figure 2 of Yada et al[38], figure 3 of Aleshin et al[39], figure 10 of Capano et al[40], figure 4 of Gorog et al[41], figure 3 of Wang et al[42], figure 5 of House et al[43], figure 5A of Lau et al[44] and figure 4 of Sy et al[45]). The reason why this is a readout of autoactivation is that p38 inhibitors do not interfere with the catalytic activity of MKK3/6 and thus would not interfere with transphosphorylation by an activating upstream kinase. Interestingly, Ota et al recently reported that Hsp90/Cdc37 also plays an inhibitory role in p38 autophosphorylation and selectively modulates TAB1 mediated p38 activation[30].

Isoforms of p38 and their pathophysiological role

Most of the studies to date have used pharmacological inhibition of p38 to infer its role in myocardial biology. The vast majority of these compounds only inhibit p38α and p38β and do so at very similar IC₅₀s. The use of these pharmacological agents has spawned a large literature that suggests the sustained activation of p38 contributes to myocyte death during lethal myocardial ischaemia and to cardiac dysfunction and fibrosis during post-infarction remodeling, hypertrophy and with transgenic expression of upstream activating kinases (see[33, 46, 47] for reviews). These studies have been reviewed elsewhere and will not be discussed further here, instead we will concentrate on possible isoform-specific functions. Using adenoviral-mediated expression of p38α and β in rat neonatal cardiomyocytes, Saurin et al have previously shown that during simulated ischaemia p38 is activated, whereas p38β is inhibited[4]. Inhibition during lethal ischaemia of p38α, but not β, resulted in increased cell viability[4]. This strongly supports Wang et al who previously suggested that p38α activation in cardiac myocytes causes apoptosis whereas activation of the isoform leads to protection and hypertrophy[3]. The possible protective role of p38β has subsequently been highlighted by a number of investigators[5, 8, 47–49]. However, none of these studies involve manipulation of endogenous p38β. Nonetheless, it has been suggested that ischaemic preconditioning is the result of selective activation of p38β[4, 5]. This perhaps explains why pharmacological inhibition of p38s α and β during preconditioning blocks protection, (since β is perhaps the dominant isoform activated), whilst during lethal ischaemia the same inhibitors, at the same concentrations, performed by the same investigators, often within the same study, cause protection, (since α is perhaps the dominant isoform activated)[26, 50–53]. The detrimental effect of p38α activation during ischaemia is also supported by studies that do not rely on pharmacological inhibition or ectopic overexpression of a wild-type isoform. For example, we have shown that the protective effect of SB203580 in isolated cardiomyocytes exposed to simulated ischaemia is lost in myocytes expressing a form of p38α resistant to pharmacological inhibition[54]. Similarly, mice heterozygous for a p38α null allele, with consequently reduced levels of myocardial p38α protein, are resistant to infarction[55] as are mice transgenic for a kinase dead form of p38α[56]. Thus there is strong direct evidence that the activation of p38α during lethal ischaemia increases injury, whilst the activation of p38β, most usually before ischaemia, maybe protective. Very recently, this paradigm has been challenged using knock-in mice where either the endogenous p38α or p38β alleles have been replaced with a gene encoding an identical isoform with a single residue substitution that renders it resistant to pharmacological inhibition. These studies confirm that sustained p38α activation during lethal ischaemia does indeed aggravate injury[57] but that surprisingly brief activation of the same isoform during ischaemic preconditioning contributes to protection[58]. Furthermore, p38β null mice could be preconditioned by transient ischaemia and there was no change in sensitivity to lethal ischaemia[58]. Similar results also suggest p38α activity is required to adapt to hypertrophic stress. Global gene targeting to disrupt both p38α alleles results in embryonic lethality[59, 60] however myocyte-specific knockouts are viable but respond poorly to haemodynamic stress[61] again suggesting that under some circumstances even the activation of p38α has beneficial consequences. We (Rose et al, unpublished results) have recently observed that cardiomyocyte specific knockout of both p38α and p38β in mice led to spontaneous right ventricular hypertrophy, suggesting that both isoforms may contribute to maintain normal development and function of the heart. In contrast to the chronic absence of p38, acute activation results in hypertrophy and heart failure, in part mediated by the downstream kinase mitogen-activated protein kinase activated protein kinase 2 (MAPKAPK2 or MK2)[62]. In summary, the physiological role, if any, of myocardial p38β remains uncertain. In contrast there is compelling evidence that the activation of p38α can aggravate cardiac injury, but tantalizingly there are also data to suggest it is involved in chronic adaptation to ischaemia[57] and to abrupt increases in after load[61].

In distinction to p38s α and β, there is little information in the literature regarding the roles of either or isoforms during myocardial ischaemia or hypertrophy. Since these isoforms lack a crucial “gatekeeper” residue (Thr106) within the ATP-binding pocket they are resistant to classic rationally designed ATP-competitive inhibitors. However, conserved cardiac expression of p38 and perhaps δ amongst several different species suggests that these isoforms play an important role in the heart and are unlikely to be functionally redundant[6, 9, 11, 63, 64]. This notion is further reinforced by the findings that their in vitro substrate preferences[64–68] and scaffold partners[69] differ from those of α and β. Although p38 was reported to be highly expressed in heart and translocated from cytoplasm to nuclei upon pressure-overload, its specific role in the heart is still unknown[10]. In short, although the relative contribution of individual p38 isoforms to cardiac pathologies still needs to be further defined, based on current evidence p38 alpha appears to be the major player.

Despite an extensive body of literature describing the phenotypic consequences of p38 manipulation, surprisingly little is known about their underlying molecular basis. Myocyte death, pro-inflammatory gene induction, myofilament modulation and extracellular matrix remodeling are among the main factors contributing to the deleterious effects of p38 activation. The beneficial effects of activation include an anti-hypertrophic action and enhanced myocardial adaptation to stress perhaps through myocardial angiogenesis. However, the p38 substrates and interacting partners that mediate these effects remain unknown and in urgent need of study.

Pharmacological inhibition of p38

As mentioned above, and reviewed elsewhere, there is compelling evidence that pharmacological inhibition of p38 enhances contractile performance, reduces ischaemic injury and consequently heart failure in animal models. Furthermore, it seems likely similar mechanisms could operate in the human heart since myocardial p38 is identically activated in patients with ischaemia[70, 71] and/or heart failure[72, 73]. It would therefore seem reasonable to attempt to inhibit p38 as a therapeutic strategy. Fairly specific inhibitors of p38 have existed for almost 2 decades and in fact their anti-inflammatory activity led to the discovery of p38α as a proinflammatory kinase[74]. As a consequence, the structure-activity relationship of p38 with an archetypal inhibitor was described more than a decade ago[20]. Given the necessary conservation of key aspects of the ATP-binding pocket amongst the kinome it is inevitable ATP competitive inhibitors lack selectivity. Nonetheless, it is acknowledged that such inhibitors targeting p38 are often more selective than many ATP-competitive inhibitors developed for other kinases[75]. Over the years a number of other structural features of p38 have been discovered that can be exploited by medicinal chemists to create ever more potent and selective inhibitors. Some of these features, such as the Asp-Phe-Gly (DFG)-motif sequence at the base of the activation loop, are common to all 4 isoforms enabling pharmacological inhibition of p38γ and p38δ[76, 77]. As summarized in a recent excellent review,[78] the 3 principal features that have been successfully exploited in designing inhibitors are (i) DFG binding, largely to increase potency since this motif is highly conserved amongst kinases; (ii) the small gatekeeper Thr residue (106 in p38α and p38β, equivalent residue in p38γ and p38δ is Met), which also confers selectivity since it is found in less than 20% of kinases and (iii) a mobile Gly110 adjacent to the linker Met109, which also confers selectivity occurring in less than 10% of kinases.

A large number of scaffolds that inhibit p38 have been patented however only a small number have proceeded to clinical trial and of these none have progressed beyond phase 2. The principal reasons for this lack of progress are toxicity which has limited the doses at which inhibitors can be used. In the most recent trials there has also been limited anti-inflammatory efficacy perhaps caused by the use of relatively low doses, p38 escape (see below) and/or potent anti-inflammatory background therapy.

Clinical trials with p38 inhibitors

To date the majority of the experience with p38 inhibition has been for inflammatory, non-cardiac, conditions. Information from such trials has recently entered the public domain. The conditions that have been reported on most extensively are rheumatoid arthritis and Crohn’s disease. Since these are not directly relevant to the cardiovascular system interested readers are directed to recent reviews[79, 80]. The “take home” message from these trials is that despite some favourable trends inhibition of p38 has not shown efficacy when added to traditional anti-inflammatory disease modifying drugs and there have been signs of toxicity most notably with an increase in intercurrent infections, raised liver enzymes and skin rashes. One very interesting observation common to a number of trials is escape from p38 inhibition with chronic therapy that is discussed in more detail below under unresolved issues requiring future research.

Of more interest to the readership of this journal are trials of p38 inhibition for cardiovascular disease. Thus far these have been limited to small detailed pathophysiological studies only published as preliminary abstracts. Vertex’s compound VX-702 was used in a dose-escalation study in 45 patients undergoing planned coronary artery angioplasty (PCI)[81]. The salient finding was a suppression of the elevation in C reactive protein (CRP) and other inflammatory markers (monocytes, neutrophils and total white blood cell count) for 3–4 days post procedure. Oral dosing with the inhibitor (5–40 mg, once per day) was for 5 days starting the day before the procedure and was well tolerated[81]. In a study of a similar design SB681323 at a dose of 7.5 mg per day was compared to placebo in 92 patients with 1:1 randomisation (ClincalTrials.gov identifier NCT00291902). Dosing in this double blind study was for 28 days starting 3 days before planned PCI. SB681323 was associated with a reduction in CRP at day 5 and day 28 post procedure[82, 83]. There were also reductions in other inflammatory markers such as myelopreoxidase and IL8. The study was not powered to look at clinical endpoints but there were a significant reductions in episodes of angina and the need for further revascularisation, albeit on a post-hoc analysis[82, 83].

Based on ClincalTrails.gov both GlaxoSmithKline (with GW85655 also known as Losmapimod, identifier NCT00633022) and Bristol-Myers Squibb (with BMS582949, identifier NCT00570752) have active programs examining the effects of p38 inhibition on inflammation within atherosclerotic plaques as assessed by ¹⁸F-Flurodeoxyglucose (FDG) uptake on PET/CT. The GSK study with over 100 patients has completed recently but has not reported. In this study 2 dosing regimens of Losmapimod at 7.5 mg o.d. or 7.5 mg b.i.d were compared with placebo. The BMS study is ongoing with a target of 70 patients dosed with BMS582949 at 100 mg o.d. against placebo. Completion is expected in Jan 2011. In both studies dosing is for 3 months with a presumed primary endpoint of change in FDG uptake at baseline compared with 12 weeks. The study design and the endpoints are based on preclinical translational studies demonstrating that p38 activity within macrophages contributes to atherosclerosis progression[84].

Perhaps the most interesting and ambitious study is in the setting of non-ST Elevation Myocardial Infarction (NSTEMI) by GlaxoSmithKline (with Losmapimod at 7.5 mg b.i.d. with or without a 15 mg loading dose for 12 weeks compared to placebo, see ClincalTrials.gov NCT00910962). In this study 500 pts with NSTEMI will be randomized. The primary endpoints relate principally to safety in this high risk population. However, there are some measures of efficacy such as alterations in CRP and measures of infarct size based on biomarkers. In addition there is a substudy in which changes in infarction and cardiac function will be analyzed by MRI. The study is due to complete in early 2011.

Alternative Strategies to Target the p38 Pathway for Heart Disease

The lessons learnt from current p38 inhibitors for inflammatory arthritis and inflammatory bowel disease suggest that preclinical data cannot necessarily predict clinical response. There appear to be two main challenges facing the current p38 kinase targeted therapies, one is escape of efficacy from chronic inhibition and another is potential side-effects in unintended organs.

With chronic studies in man it is clear that p38 escapes from inhibition since the initial suppression of systemic cytokine concentrations is transient. The reasons why p38 inhibition only results in transient suppression of inflammation are complex but they likely relate to activation of the feedback loops known to control activity. A variety of mechanisms that allow p38 to escape inhibition have been reviewed recenty[85] and include the feedback mechanism mentioned above involving the protein TAB1. In addition, p38 kinases may function as a major signaling conduit of compensatory stress response, thus systemic inhibition may lead to undesired consequences. Therefore, finding an alternative strategy to target the p38 pathway would be valuable. Based on our current understanding of p38 regulation, targetting the TAB1-mediated non-canonical activation of p38 has a number of attractions. This non-canonical mode of p38 activation appears to occur relatively specifically in certain circumstances, such as ischemic hearts and differs from the canonical activation of p38 by upstream kinases that leads to inflammation[29]. Furthermore, it appears that the TAB1-mediated non-canonical p38 pathway can be regulated by a different set of molecular partners from the canonical pathway, such as Hsp90/Cdc37. It therefore has the possible attraction of circumstance specificity and potentially would avoid the escape mechanisms occurring through pathways that are amplified with inhibitors of p38 catalytic activity. It would also potentially avoid the toxicity seen with classic inhibitors that likely relates to the key homeostatic function of p38α[59, 86]. To achieve these goals, we would need to have a better understanding at the molecular level of the different regulatory mechanisms controlling p38 under specific pathological conditions. In this case, uncovering the structural basis for the p38 and TAB-1 interaction and its regulation by other molecular partners would be an essential step leading to more effective p38 targeted therapies for ischemic heart diseases. As an example we have recently described a mechanism responsible for regulating the interaction between TAB1 and p38α in the heart that if potentiated could prevent activation during myocardial ischaemia[30]. What is certain is that a more complete understanding of the mechanisms of p38 activation and of the relevant downstream substrates will significantly advance the field and enable more refined therapies targeting this crucial kinase.

Table 1.

Summary of cardiovascular trials of p38 inhibition in the public domain

Compound	Sponsor	Registration No (clincaltrails.gov)	Clinical Circumstance	Endpoint/Results
VX-702	Vertex	Unknown	PCI/Short term	Reduced Inflammatory Markers
SB681323	GSK	00291902	PCI/Medium term	Reduced inflammatory markers
GW85655 (Losmapimod)	GSK	00633022	Active atherosclerosis	FDG uptake in great vessels. Due to report at AHA 2010
BMS582949	BMS	00570752	Active atherosclerosis	FDG uptake in great vessels. Due to complete early 2011
GW85655 (Losmapimod	GSK	00910962	Non-ST elevation MI	Safety/Inflammation. Due to complete 2011