Journal of Molecular and Cellular Cardiology 82 (2015) 167–173

Contents lists available at ScienceDirect

Journal of Molecular and Cellular Cardiology journal homepage: www.elsevier.com/locate/yjmcc

Review article

TNNI3K in cardiovascular disease and prospects for therapy Annalisa Milano, Elisabeth.M. Lodder, Connie.R. Bezzina ⁎ Department of Clinical and Experimental Cardiology, Academic Medical Centre, Amsterdam, The Netherlands

a r t i c l e

i n f o

Article history: Received 9 October 2014 Received in revised form 23 February 2015 Accepted 9 March 2015 Available online 16 March 2015 Keywords: Tnni3k Heart failure Ischemia/reperfusion Arrhythmia Kinase

a b s t r a c t Cardiovascular diseases are an important cause of morbidity and mortality worldwide and the global burden of these diseases continues to grow. Therefore new therapies are urgently needed. The role of protein kinases in disease, including cardiac disease, is long recognized, making kinases important therapeutic targets. We here review the knowledge gathered in the last decade about troponin I-interacting kinase (TNNI3K), a kinase with cardiac-restricted expression that has been implicated in various cardiac phenotypes and diseases including heart failure, cardiomyopathy, ischemia/reperfusion injury and conduction of the cardiac electrical impulse. © 2015 Elsevier Ltd. All rights reserved.

Contents 1. 2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Background on protein kinases . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. TNNI3K: gene and transcript . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. TNNI3K protein: domain structure, kinase activity, binding partners and localization 2.3. TNNI3K in cardiac function and disease . . . . . . . . . . . . . . . . . . . . 2.3.1. Quantitative trait locus mapping studies in mice . . . . . . . . . . . . 2.3.2. TNNI3K in heart failure and hypertrophy . . . . . . . . . . . . . . . 2.3.3. TNNI3K in ischemia/reperfusion injury . . . . . . . . . . . . . . . . 2.3.4. TNNI3K in cardiac conduction . . . . . . . . . . . . . . . . . . . . 2.4. TNNI3K in human cardiac disease . . . . . . . . . . . . . . . . . . . . . . . 3. Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Disclosure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Cardiovascular diseases are an important cause of morbidity and mortality world-wide and the global burden of these diseases continues to grow [1–3]. Thus, new therapeutic approaches for prevention and treatment of these diseases are urgently needed. The role of protein kinases in disease, including cardiovascular disease, is long recognized, making kinases important targets for therapy [4]. We here review ⁎ Corresponding author at: Academic Medical Center, Department of Experimental Cardiology, Room L2-108, Meibergdreef 9, 1105AZ Amsterdam, The Netherlands. Tel.: +31 20 5665403. E-mail address: [email protected] (C.R. Bezzina).

http://dx.doi.org/10.1016/j.yjmcc.2015.03.008 0022-2828/© 2015 Elsevier Ltd. All rights reserved.

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

167 167 168 168 169 169 169 171 171 171 172 172 172 172

knowledge about troponin I-interacting kinase (TNNI3K), a kinase with cardiac-restricted expression that has in the last ten years been implicated in various cardiac phenotypes. These include progression of cardiomyopathy [5], hypertrophy [6,7], conduction of the cardiac electrical impulse [8] and ischemia/reperfusion injury, making TNNI3K an interesting target for the development of specific antagonists for these diseases. 2. Background on protein kinases Protein kinases catalyze the phosphorylation of proteins as a means of signal transduction. Catalysis by protein kinases brings about the

168

A. Milano et al. / Journal of Molecular and Cellular Cardiology 82 (2015) 167–173

transfer of a phosphate group from ATP onto the hydroxyl group of serine or threonine (in the case of serine/threonine protein kinases), or tyrosine residues (in the case of tyrosine kinases) in target proteins to modulate their function. The human kinome is vast and protein kinases constitute approximately 2% of human genes (N500 genes) [9]; they can be grouped into families based on their molecular target [10]. The function of protein kinases can be inhibited by numerous strategies including competition for ATP binding, prevention of binding to anchoring proteins, and binding to substrate interaction sites [4,10]. Although the generation of highly selective kinase inhibitors is challenging (for example because of the similarity in the ATP-binding pocket across different kinases), targeting protein kinases with cardiac-restricted expression is an attractive goal for therapy of cardiac diseases as targeting such kinases limits unwanted off-target effects in other organs.

2.1. TNNI3K: gene and transcript The cDNA for TNNI3K was first isolated from a human adult heart cDNA library [11]. In humans the TNNI3K gene is located on chromosome 1 (1p31.1) and at the moment of writing eight TNNI3K transcripts were annotated in Ensembl, of which four are protein coding [12,13]. In humans, the longest TNNI3K transcript encodes a 92 kD protein of 835 amino acids. Some studies have indicated the presence of conjoined transcripts in human heart [14], resulting from the transcription of TNNI3K and the neighboring gene FPGT; 6 such transcripts (of which 5 are protein coding) are annotated in Ensembl. The longest FPGTTNNI3K transcript encodes a 105 kD protein of 949 amino acids. Experimental evidence of the function of such fusion transcripts is thus far lacking [14]. No such fusion transcripts have been reported or detected in the murine heart (Ensembl GRCm38 and unpublished data from our group). Analysis of TNNI3K mRNA expression in multiple human tissues revealed that expression is restricted to the heart, being highly expressed in both adult and fetal heart, and expressed in left and right atria and ventricles [11]. A more recent study reported similar findings except for a slight expression in the brain and testis [15]. The latter study additionally demonstrated that within the heart TNNI3K expression is limited to cardiomyocytes [15]. The TNNI3K gene is highly conserved and there are only a few differences between humans and other species including mice [16]. The mouse Tnni3k cDNA was cloned and the basal promoter was characterized [17]. Here a series of reporter studies uncovered Mef2c binding sites as critical cis-acting elements in the regulation of Tnni3k expression. Mef2c is a member of the MADS (MCM1-agamous-deficiensserum response factor) superfamily of transcription factors [18] and is involved among others in myogenesis [19] and in the development of structures derived from the second heart field [20]. The role of Mef2c binding in control of Tnni3k expression was confirmed in the same study through experiments in primary rat cardiomyocytes. In these cells transfection with Mef2c antisense morpholinos resulted in reduced Tnni3k expression [17]. Tnni3k may be a direct target of Mef2c transcriptional regulation, as shown by electrophoretic mobility shift assay (EMSA) in primary rat neonatal cardiomyocytes nuclear extracts [17].

2.2. TNNI3K protein: domain structure, kinase activity, binding partners and localization A homology search and phylogenetic analysis showed that the overall protein domain structure of TNNI3K is similar to that of integrinlinked kinase (ILK) that regulates signaling pathways controlling cardiac growth, contractility and repair [11]. TNNI3K contains four main recognizable domains (Fig. 1): a coiled coil domain at the N-terminal part of the protein, followed by a domain containing ankyrin (ANK) repeats, an active protein kinase domain and a C-terminal serine-rich domain [11]. Most likely TNNI3K interacts with its partners via the ANK domain as this domain is an established protein-binding target of proteins in other pathways [21] TNNI3K has also been shown to be able to form homo-dimers or oligomers [22] and an in silico analysis has uncovered putative interaction interfaces at which TNNI3K monomers interact during dimerization or oligomerization [16]. In spite of intense research on TNNI3K, the identification of downstream phosphorylation targets of TNNI3K remains scarce. Similarly knowledge on interacting partners that may be involved in TNNI3Kmediated signaling as well as data on the subcellular localization of TNNI3K, which might provide hints concerning its function is largely lacking, limiting insight into the exact molecular function of this protein in myocardial biology. The fact that TNNI3K is a functional kinase has been principally established through studies investigating the autophosphorylation of the protein [7,22]. The domain structure of TNNI3K suggested that it is a dual-function kinase with both tyrosine and serine/threonine kinase activity. This was confirmed in in vitro studies that identified multiple auto-phosphorylation sites involving tyrosine as well as serine and threonine residues [7,22]. These sites are mainly located in the ANK-repeat and serine-rich domains of the protein. Studies conducted in mice over-expressing human TNNI3K (DBA/2J-hTNNI3Ktg mice) and mice over-expressing a kinase-dead version of human TNNI3K (DBA/2J-TNKD mice; due to residue 490 LysNArg substitution at the canonical ATP binding site of the kinase domain) demonstrated that auto-phosphorylation of TNNI3K also occurs in vivo [7]. In an early study, candidate proteins that interact with TNNI3K have been identified by means of a yeast two-hybrid screen using the Cterminal serine-rich domain of the protein as bait [11]. Among the candidate proteins identified in this screen the most frequently observed was cardiac troponin I (cTnI), hence the name troponin-I-interacting kinase. [11] Another yeast two-hybrid screen, this time using the Nterminal ankyrin-containing domain of TNNI3K as bait, led to the identification of the mitochondrial protein anti-oxidant protein 1 (AOP-1, also called peroxiredoxin 3) as another possible interaction partner of TNNI3K [23]. AOP-1 functions as a thioredoxin-dependent peroxidase that scavenges ROS, such as H2O2 [24]. Yet, for both cTnI and AOP-1, confirmatory studies of the interaction with TNNI3K have thus far only been conducted by means of co-immunoprecipitation of overexpressed proteins [11,25]. Evidence for an in vivo interaction in the setting of physiologic levels of TNNI3K and the putative interacting proteins is thus far lacking. In a series of preliminary studies also conducted in an overexpression system it has been proposed that TNNI3K can phosphorylate

Fig. 1. Schematic representation of the domain structure of TNNI3k. Protein domains (with amino acid positions in brackets) of TNNI3K drawn to scale based on uniprot_ID Q59H18 (http://www.uniprot.org/uniprot/Q59H18).

A. Milano et al. / Journal of Molecular and Cellular Cardiology 82 (2015) 167–173

cTnI [6,25]. The authors of these studies additionally implicated this phosphorylation in the regulation of cardiomyocyte contractility, where phosphorylation promotes contractility [25]. It should be noted however that the implicated cTnI phosphorylation sites involve residues that have been established as protein kinase C or protein kinase A phosphorylation sites [26], and it could therefore be still possible that the high level of TNNI3K expression in the experimental systems used leads to promiscuous phosphorylation at illegitimate sites. Studies that have investigated the subcellular localization of TNNI3K have provided conflicting findings, that likely stem from the different anti-TNNI3K antibodies used and variability in tissue sample processing. Immunohistochemical studies in mice over-expressing hTNNI3K (DBA/ 2J-hTNNI3K mice, see Table 1) and C57BL/6J mice using an antibody raised against the C-terminus of the protein demonstrated localization of TNNI3K to the sarcomeric Z-disk, [7] an observation that could be in line with the putative interaction with cTnI as suggested by the yeast two hybrid and co-immunoprecipitation data [11]. This localization was also observed in DBA/2J-TNKD mice, suggesting that the kinase activity of TNNI3K is not required for its Z-disk localization [7]. Besides its structural role in the sarcomere, the Z-disk is recognized to serve as a nodal point for signaling in general and mechanosensation and mechanotransduction in particular [27], and also links to the t-tubular system and the sarcoplasmic reticulum [28]. In contrast however, Vagnozzi et al. reported the enrichment of TNNI3K in the nuclear fraction in subcellular fractionation studies, and, localization around nuclei in immunofluorescence studies in rat cardiomyocytes [15]. Crucially, these investigators did not detect co-localization with α-actinin, an observation that does not support an interaction of TNNI3K with cTnI, at least at the sarcomere. Studies conducted in a limited set of human heart tissue have shown both nuclear and cytoplasmic localization, although the relative intensity of cytoplasmic staining differed between the two studies that demonstrated this [11,16]. All in all, the exact sub-cellular localization remains to be conclusively demonstrated. 2.3. TNNI3K in cardiac function and disease 2.3.1 . Quantitative trait locus mapping studies in mice Robust relations between Tnni3k and distinct cardiac phenotypes have been found through quantitative trait locus (QTL) genetic mapping studies conducted in different inbred strains of mice. These studies have linked the Tnni3K locus on mouse chromosome 3 to cardiac function and survival [29,30] and the electrocardiographic PR-interval [29]. QTL mapping studies in mice have also identified the Tnni3k locus as a susceptibility locus for viral myocarditis caused by Coxsackievirus B3 infection; however while Tnni3k is considered a candidate for this effect, evidence in support of its role as the causal gene within the locus is as yet lacking [30,31]. QTL mapping is based on the idea that one can identify a genetic locus that controls a phenotype of interest through the identification of genetic markers linked to the locus in a segregating population with known pedigree [32]. Although not known at the outset, the studies identifying the Tnni3k locus as a modulator of cardiac function [33,34]

169

and PR-interval [29] all exploited the fact that the endogenous levels of Tnni3k mRNA and protein vary widely among different strains of inbred mice, ranging from robust levels in some strains (e.g. the C57BL/ 6J, WSB/Eij, AKR/J and 129P2 strains) to a total lack/undetectable expression in others (e.g. the FVB/N and DBA/2J strains) [5,8,35]. Although the latter mice effectively represent Tnni3k-null mice, of note, they do not show baseline cardiac abnormalities as determined by echocardiographic and electrocardiographic (ECG) measurements [5,8]. Thus, while the high conservation of Tnni3k [16] points to evolutionary constraints for retention of the gene, clearly the fact that some mouse strains have lost Tnni3k expression indicates that it is not necessary for survival in the protected environment of the laboratory. 2.3.2 . TNNI3K in heart failure and hypertrophy While initial in vitro work suggested a cardioprotective role for increased levels of TNNI3K [36] more recent work has provided strong evidence that increased levels of TNNI3K have detrimental effects on cardiac structure and function. The Tnni3k locus was first linked to cardiac function and survival by Marchuk, Rockman and co-workers in two independent genetic mapping studies, each utilizing a different mouse N2 cross (Fig. 2A) [33,34]. Both of these studies were conducted in cardiomyopathy-sensitized mice due to the cardiac-specific overexpression of the Ca2 + binding protein calsequestrin (Csq, Csqtg mice) [37]. Additionally for genetic mapping, they both used inbred mouse lines that displayed marked differences in cardiac function and survival as a consequence of Csq over-expression, namely DBA/2J, with a less severe phenotype, and C57Bl6 [33] or AKR/J [34], with a more severe phenotype. This variability in disease severity in the setting of diseasesensitization mediated by Csq overexpression, alongside the genetic differences between the mouse strains used, enhanced the possibility of genetic locus discovery in these studies. Following the initial reports that identified the modulatory chromosomal interval, a series of elegant studies led to the unequivocal identification of Tnni3k as the causal gene within the interval [5]. One crucial observation these investigators made was the high Tnni3k expression in the susceptible strains (AKR/J, C57Bl6) and the low/undetectable levels in the protected strain (DBA/2J). This they attributed to defective splicing of Tnni3k leading to a frameshift, that was associated with decreased Tnni3k mRNA by a mechanism of nonsense-mediated decay in mice with undetectable Tnni3k expression. The in vivo role of TNNI3K was subsequently investigated by cardiacspecific over-expression of human TNNI3K in DBA/2J mice (DBA/2JhTNNI3Ktg) (Table 1). Crossing these DBA/2J-hTNNI3Ktg mice with DBA/2J-Csqtg ‘sensitizer mice’ led to a marked acceleration in the development of heart failure, with severe systolic dysfunction, chamber dilatation and a profound reduction in survival in double transgenic mice (overexpressing hTNNI3K and Csq) compared to single transgenic DBA/2J-Csqtg mice [5]. These investigators went on to generate a DBA/ 2J.AKR/J congenic mouse strain by repeated backcrossing of DBA/ 2J.AKR/JN2 mice into DBA/2J while selecting for the AKR/J genotype at the Tnni3k locus (Fig. 2A). This congenic line essentially harbored a 20 Mb chromosomal interval containing Tnni3k of the AKR/J strain in

Table 1 Glossary of mice generated and investigated in the various studies. Mouse

Details

Reference

DBA/2J-hTNNI3Ktg C57BL/6J-hTNNI3Ktg FVB/N-hTNNI3Ktg DBA/2J-TKND DBA/2J-Csqtg DBA/2J-hTNNI3K-Csq (double transgenic) DBA/2J.AKR/J congenic mouse

Overexpression of hTNNI3K in DBA/2J mice; wt DBA/2J mice do not have detectable murine Tnni3k Overexpression of hTNNI3K in C57Bl/6J mice; wt C57Bl/6J mice express robust levels of murine Tnni3k Overexpression of hTNNI3K in FVB/N mice; wt FVB/N mice do not have detectable murine Tnni3k Overexpression of kinase-dead hTNNI3K in DBA/2J mice; wt DBA/2J mice do not have detectable murine Tnni3k Overexpression of murine calsequestrin in DBA/2J mice Overexpression of murine calsequestrin and hTNNI3K in DBA/2J mice

[34] [6,15] [15] [7] [34] [34]

Congenic mice generated by back-crossing DBA/2J mice (recipient strain) with AKR/J mice (donor strain) for N10 generations; resultant congenic mice contain a 20 Mb region from AKR/J chromosome 3 including Tnni3k in the setting of a DBA/2J genetic background

[34]

170

A. Milano et al. / Journal of Molecular and Cellular Cardiology 82 (2015) 167–173

Fig. 2. Mouse breeding strategies: (A) Backcross strategy for generation of N2 mice for QTL mapping (as used by Suzuki et al. [33], Wheeler et al. 2005 [34] and Wiltshire et al. [31]) and for generation of a congenic mouse strain for the locus of interest (marked by X) such as the hrtfm2.DBA/2J-AKR/J congenic strain produced by Wheeler et al. 2009 [5]. (B) Production of second filial (F2) generation mice for quantitative trait locus (QTL) mapping as used by Scicluna et al. [29] and Aly et al. [30].

the setting of the DBA/2J genome. Crucially, by crossing this DBA/ 2J.AKR/J congenic line with DBA/2J-Csqtg mice, they could demonstrate that even physiological levels of Tnni3k (as offered by the AKR/Jderived Tnni3k alleles), are associated with decreased survival and acceleration of cardiomyopathy in the setting of Csq-overexpression [5]. Studies in three different hTNNI3k overexpression mouse models have produced conflicting data concerning the role of TNNI3K in the development of hypertrophy. At two months of age DBA/2J-hTNNI3Ktg mice (i.e. without Csq sensitization) showed increased heart weight to body weight ratio and enlarged cardiomyocytes compared to wildtype DBA/2J mice [7]. Consistent with the observed hypertrophy,

mRNA levels of Nppa (which encodes the atrial natriuretic peptide, ANF) and Nppb (which encodes the brain natriuretic peptide, BNP) were elevated. However, the phosphorylation status of p38, ERK1/2 and AKT, established effectors transducing cardiac remodeling signals, was unchanged, suggesting that TNNI3K may not regulate these effectors in this setting. Examination of the ultrastructure of cardiomyocytes in these mice uncovered a decreased resting sarcomere length. DBA/2J mice overexpressing a kinase-dead version of TNNI3K (TNKD, with kinase activity ablated by the 490Lys → Arg substitution; here referred to as DBA/2J-TKND mice, Table 1), provided insight as to whether the hypertrophy relied on kinase activity [7]. Although not as severe as in

A. Milano et al. / Journal of Molecular and Cellular Cardiology 82 (2015) 167–173

DBA/2J-hTNNI3Ktg mice, cardiac mass and cardiomyocyte size was also increased in DBA/2J-TKND mice, implying that while the kinase activity of TNNI3K is involved in mediating cardiac remodeling, effects of TNNI3K on cardiac remodeling that are independent of its kinase activity may also exist. After pressure overload (induced by transverse aortic constriction) DBA/2J-hTNNI3Ktg transgenic mice showed greater diastolic and systolic dysfunction, an effect that appeared to depend on the kinase activity of TNNI3K [5,7]. A hypertrophic phenotype was also reported by another group in a mouse model with cardiac overexpression of hTNNI3K, in this case in a C57BL/6J genetic background (C57BL/6J-hTNNI3Ktg, Table 1) [6]. Of note, these investigators uncovered hypertrophy in mice with a high level of overexpression of the transgene but not in mice with intermediate and low expression. At variance with the DBA/2J-hTNNI3Ktg, enhanced contractility and diastolic function was observed. In contrast, in a third overexpression model generated by Vagnozzi et al., this time in the FVB/N strain (FVB/N-hTNNI3Ktg), echocardiographic analysis did not reveal hypertrophy at 3 months of age [15]. Yet, as no histological analysis of cardiomyocyte size has been reported in this model, one cannot exclude the presence of less overt hypertrophy. The observed differences in hypertrophic phenotype between the three overexpression models [6,7,15] may stem from a number of factors, the most important of which is likely the absolute level of transgene expression achieved in the three models. The fact that the hypertrophy was not observed in the low and intermediate overexpression models described by Wang et al. calls for caution as it suggests that the hypertrophic response is evident only in the setting of considerable overexpression of TNNI3K. Another aspect is the different mouse genetic background used in the three studies; while the C57BL6/6J strain used by Wang and co-workers [6] expressed murine Tnni3k, this is lacking in the DBA/2J strain used by Tang et al. [7] and in the FVB/N strain used by Vagnozzi et al. [15]. Clearly further studies are necessary to elucidate whether physiologic levels of TNNI3K play a role in the development of hypertrophy in the face of stressors, and if so, through which signaling pathways it occurs. 2.3.3 . TNNI3K in ischemia/reperfusion injury A study by Vagnozzi et al. [15] provided evidence for a role of Tnni3k in mediation of ischemic-reperfusion injury (I/R), a pathology that is associated with increased oxidative stress, mitochondrial dysfunction, and cardiomyocyte death [38]. In a series of elegant ischemia/reperfusion studies conducted in mice over-expressing hTNNI3K (FVB/N strain, which do not naturally express murine Tnni3k), in mice with inducible cardiomyocyte-specific knock-out of Tnni3k (C57BL6 strain, which naturally express Tnni3k), and in a neonatal rat ventricular cardiomyocyte model of ischemic injury, these investigators provided convincing evidence that TNNI3K worsens I/R injury. They demonstrated that this effect is mediated, at least in part, through the overproduction of mitochondrial reactive oxygen species (mROS) with ensuing mitochondrial and bioenergetic impairment, and is dependent on activation of the mitogen-activated protein kinase p38 MAPK. No activation of p38 MAPK was observed in FVB/N-hTNNI3Ktg animals without I/R injury. This is similar to previous findings in DBA/2J-hTNNI3Ktg animals in spite of the observed hypertrophy in baseline conditions [7]. Clearly, the lack of apparent involvement of p38 MAPK activation in mediating hypertrophy necessitates further studies. By administering active-site binding inhibitors of TNNI3K that they developed (GSK854 and GSK329) to wild-type C57BL6 mice at reperfusion following ischemia (mimicking clinical intervention in patients), they went on to show reduction of infarct size following I/R. They also demonstrated long-term beneficial effects of Tnni3k inhibition during reperfusion coupled to sustained inhibition thereafter, in C57Bl6 mice after a severe ischemic insult. In this model of human acute coronary syndrome, GSK854 inhibition of Tnni3k reduced left ventricular dysfunction and limited progressive remodeling, with improvement in cardiac function, dimensions and a reduction in cardiac fibrosis and

171

cardiomyocyte hypertrophy. Although the mechanism by which TNNI3K activates p38 MAPK in I/R remains unclear, taken together these data provide an approach for treating ischemia–perfusion injury in a clinical setting [39]. In this respect, the exclusive localization of TNNI3K in the heart makes it a particularly interesting target as opposed to targeting other effectors, such as p38 MAPK, which are more broadly expressed and which regulate numerous biological processes. 2.3.4 . TNNI3K in cardiac conduction In 2012 our group linked Tnni3k to conduction of the cardiac electrical impulse, particularly in control of the electrocardiographic PRinterval which represents the time between the beginning of electrical activation of the atria and the beginning of activation of the ventricles [8]. Our identification of Tnni3k as a gene influencing cardiac conduction started with a QTL mapping approach [29] similar to that used by Marchuk, Rockman and co-workers [33,34], but in our case conducted in an F2 mouse population (Fig. 2B). Our study employed conduction disease-sensitized F2 mice; this sensitization was achieved by virtue of the Scn5a-1798insD mutation (insertion of an aspartic acid (D) at codon 1798 within the murine Scn5a gene) [40]. This mutation is homologous to the SCN5A-1795insD mutation we had identified in a large Dutch family presenting with conduction disease and Long QT Syndrome [41]. SCN5A encodes the major sodium channel isoform in heart. It mediates the depolarization phase of the cardiac action potential and thereby plays a major role in cardiac conduction. F2 mice were generated by crossing Scn5a-1798insD mutant mice of the 129P2 strain with Scn5a-1798insD mutant mice of the FVB/NJ strain, where these two inbred mouse strains vary in their susceptibility to cardiac conduction disease [42]. By QTL mapping we mapped a locus controlling the PR-interval on mouse chromosome 3, overlapping with the locus previously linked to cardiac function and survival by Marchuk, Rockman and co-workers [33,34]. As genetic variation within a locus may affect a given phenotype through effects on expression levels of neighboring genes, we next used cardiac tissue from the same F2 mice to map genetic loci controlling cardiac gene expression (commonly called expression QTLs, eQTLs), including Tnni3k. This uncovered an eQTL controlling Tnni3k transcript levels overlapping the PR-interval QTL on chromosome 3. Furthermore, Tnni3k transcript abundance showed a strong positive correlation with the PR-interval in the F2 mice. Taken together this made Tnni3k a prime candidate for the effect observed on the PR-interval at the locus [8]. To confirm the hypothesis that the PR interval is modulated by TNNI3K we compared the PR-interval between DBA/2J-hTNNI3K transgenic mice and wild-type DBA/2J mice (naturally lacking Tnni3k expression), showing that hTNNI3K over-expression resulted in a marked increase in the PR interval. Furthermore, the PR interval in DBA/ 2J.AKR/J congenic mice (with natural levels of Tnni3k as offered by the AKR/J-derived Tnni3k allele) was longer compared to DBA/2J mice (which lack Tnni3k); congenic mice in fact showed PR-intervals that were comparable to those of the AKR/J donor strain. This demonstrated that natural levels of Tnni3k are associated with a longer PR-interval in comparison with the absence of Tnni3k [8]. While altogether this data provided a robust link between Tnni3k and cardiac conduction, as for Tnni3k-mediated mechanisms of cardiac dysfunction and hypertrophy, the molecular mechanism whereby increased levels of Tnni3k delay cardiac conduction await elucidation. Cardiac conduction is affected by multiple processes involving cardiomyocyte depolarization, gapjunctional coupling, and fibrosis, and it is likely that TNNI3K affects atrio-ventricular conduction by impacting on one or more of these. 2.4. TNNI3K in human cardiac disease Recently, using an approach combining linkage analysis and exome sequencing, Theis and co-workers [16] reported on a TNNI3K mutation (TNNI3K-G526D) co-segregating with a cardiac phenotype in a small family. This family presented with a phenotype consisting of variably

172

A. Milano et al. / Journal of Molecular and Cellular Cardiology 82 (2015) 167–173

expressed atrial tachyarrhythmia, conduction system disease and dilated cardiomyopathy. The G526D mutation identified in this family concerns a conserved residue in TNNI3K, which although located in the kinase domain of the protein is not predicted to directly impact functional sites within the ATP-binding catalytic domain. However a peptide consisting of the mutated kinase domain (with 526D) aggregated in vitro, as opposed to a peptide consisting of the wild-type kinase domain (with 526G) that was soluble, suggesting an effect on protein assembly. In support of such a scenario, in silico protein modeling showed that while the G N D substitution is expected to have a minimal effect on the monomeric protein structure, it alters the relative affinity of docking configurations during oligomerization. The energetically favorable interactions reported for the mutant protein, also with wildtype TNNI3K, would allow for a dominant-negative effect of the mutant, which would be in line with the autosomal dominant inheritance in the family. Immunohistochemical analysis on a right ventricular endomyocardial biopsy from the proband using two separate antibodies targeting respectively the C-terminal and N-terminal domains of the protein showed reduced TNNI3K staining in the cytoplasm. Ultrastructural analysis of this biopsy by electron microscopy uncovered features of degenerative myopathy with abnormal mitochondria, amorphous inclusions in the cytoplasm and nuclei, and myofilament loss. The aggregation observed in vitro for the mutant and the reduced staining for TNNI3K in myocardium from the proband could point to decreased bioavailability or a loss-of-function defect for the TNNI3K-G526D mutation. Such a mechanism is however difficult to reconcile with findings from rodent studies that have largely identified protective effects for loss of TNNI3K function in the heart, including faster cardiac conduction and less severe cardiomyopathy. Clearly, more research on this mutation is warranted to understand the mechanism whereby it leads to the phenotype observed in patients. Furthermore, analysis of TNNI3K in additional patient sets may uncover more mutations that could increase our insight into the role of this gene in human cardiac disease. TNNI3K mRNA expression was found to be higher in cardiac tissue samples from patients with end-stage idiopathic dilated cardiomyopathy compared to healthy controls [43]. In another study, TNNI3K protein levels were shown to be increased in a set of left ventricular samples from failing hearts of patients with ischemic cardiomyopathy undergoing transplantation compared to non-failing control hearts [15]. The fact that both of these studies were conducted in a very small sample set and in end-stage tissue precludes any inference concerning the role of TNNI3K in cardiac disease and whether elevated levels may represent causal or adaptive changes. Recent human genetic association studies, including genome-wide association studies, have identified robust association between genetic variation at the chromosomal locus harboring TNNI3K with variation in body mass index [44,45]. Considering the cardiac-specific expression of TNNI3K and the fact that another gene exists at this locus which appears to be a more likely candidate, namely FPGT (encoding fucose-1phosphate guanylyltransferase, involved in the turnover of L-fucose in the liver) [46], TNNI3K may not be a likely causal gene mediating this effect. On the other hand, the role of the TNNI3K-FPGT fusion transcript in relation to phenotypes mapping to this locus may need to be taken into account. Clearly more studies are needed to understand the genetic mechanism at this locus and any possible involvement of TNNI3K. 3. Concluding remarks In the last decade, studies largely conducted in mice have uncovered TNNI3K as a cardiac-specific kinase that plays an important role in a broad spectrum of cardiac pathologies. However, in spite of the robust links that have been laid between TNNI3K and distinct cardiac phenotypes, insight into the molecular mechanisms and signaling pathways whereby TNNI3K mediates its effects is very scarce. In particular information concerning upstream regulators and downstream effectors of TNNI3K is lacking and their identification is likely to be an important

goal of future research on TNNI3K. Such in-depth knowledge on TNNI3K is necessary as its cardiac-specific expression makes it an attractive target for therapy in human cardiac disease [47]. Furthermore, future research in humans, including genetic studies, may provide an increased insight into the role of TNNI3K in human cardiac disease, which has only just started to be uncovered and awaits confirmatory studies.

Disclosure None.

Acknowledgments Work by the authors is funded by the Netherlands CardioVascular Research Initiative (CVON2012-10) (Dutch Heart Foundation, Dutch Federation of University Medical Centres, the Netherlands Organisation for Health Research and Development and the Royal Netherlands Academy of Sciences) PREDICT project and by the Dutch Heart Foundation (project 2009B66).

References [1] Fuster V. Global burden of cardiovascular disease: time to implement feasible strategies and to monitor results. J Am Coll Cardiol 2014;64:520–2. http://dx.doi.org/10. 1016/j.jacc.2014.06.1151. [2] Levy D, Kenchaiah S, Larson MG, Benjamin EJ, Kupka MJ, Ho KKL, et al. Long-term trends in the incidence of and survival with heart failure. N Engl J Med 2002;347: 1397–402. http://dx.doi.org/10.1056/NEJMoa020265. [3] Heeringa J, Van Der Kuip DAM, Hofman A, Kors JA, Van Herpen G, Stricker BHC, et al. Prevalence, incidence and lifetime risk of atrial fibrillation: the Rotterdam study. Eur Heart J 2006;27:949–53. http://dx.doi.org/10.1093/eurheartj/ehi825. [4] Anderson ME, Higgins LS, Schulman H. Disease mechanisms and emerging therapies: protein kinases and their inhibitors in myocardial disease. Nat Clin Pract Cardiovasc Med 2006;3:437–45. http://dx.doi.org/10.1038/ncpcardio0585. [5] Wheeler FC, Tang H, Marks O a, Hadnott TN, Chu P-L, Mao L, et al. Tnni3k modifies disease progression in murine models of cardiomyopathy. PLoS Genet 2009;5: e1000647. http://dx.doi.org/10.1371/journal.pgen.1000647. [6] Wang X, Wang J, Su M, Wang C, Chen J, Wang H, et al. TNNI3K, a cardiac-specific kinase, promotes physiological cardiac hypertrophy in transgenic Mice. PLoS One 2013;8:e58570. http://dx.doi.org/10.1371/journal.pone.0058570. [7] Tang H, Xiao K, Mao L, Rockman HA, Marchuk Da. Overexpression of TNNI3K, a cardiac-specific MAPKKK, promotes cardiac dysfunction. J Mol Cell Cardiol 2013; 54:101–11. http://dx.doi.org/10.1016/j.yjmcc.2012.10.004. [8] Lodder* EM, Scicluna* BP, Milano A, Sun AY, Tang H, Remme CA, et al. Dissection of a quantitative trait locus for PR interval duration identifies Tnni3k as a novel modulator of cardiac conduction. PLoS Genet 2012;8:e1003113. http://dx.doi.org/10.1371/ journal.pgen.1003113. [9] Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S. The protein kinase complement of the human genome. Science 2002;298(80-):1912–34. http://dx.doi.org/ 10.1126/science.1075762. [10] Anastassiadis T, Deacon SW, Devarajan K, Ma H, Peterson JR. Comprehensive assay of kinase catalytic activity reveals features of kinase inhibitor selectivity. Nat Biotechnol 2011;29:1039–45. http://dx.doi.org/10.1038/nbt.2017. [11] Zhao Y, Meng X-M, Wei Y-J, Zhao X-W, Liu D-Q, Cao H-Q, et al. Cloning and characterization of a novel cardiac-specific kinase that interacts specifically with cardiac troponin I. J Mol Med (Berl) 2003;81:297–304. http://dx.doi.org/10.1007/s00109003-0427-x. [12] Farrell CM, O’Leary NA, Harte RA, Loveland JE, Wilming LG, Wallin C, et al. Current status and new features of the Consensus Coding Sequence database. Nucleic Acids Res 2014;42:D865–72. http://dx.doi.org/10.1093/nar/gkt1059. [13] Flicek P, Ridwan Amode M, Barrell D, Beal K, Billis K, Brent S, et al. Ensembl 2014. Nucleic Acids Res 2014;42:D749–55. http://dx.doi.org/10.1093/nar/gkt1196. [14] Prakash T, Sharma VK, Adati N, Ozawa R, Kumar N, Nishida Y, et al. Expression of conjoined genes: another mechanism for gene regulation in eukaryotes. PLoS One 2010;5:e13284. http://dx.doi.org/10.1371/journal.pone.0013284. [15] Vagnozzi RJ, Gatto GJ, Kallander LS, Hoffman NE, Mallilankaraman K, Ballard VLT, et al. Inhibition of the cardiomyocyte-specific kinase TNNI3K limits oxidative stress, injury, and adverse remodeling in the ischemic heart. Sci Transl Med 2013;5: 207ra141. http://dx.doi.org/10.1126/scitranslmed.3006479. [16] Theis JL, Zimmermann MT, Larsen BT, Rybakova IN, Long PA, Evans JM, et al. TNNI3K mutation in familial syndrome of conduction system disease, atrial tachyarrhythmia and dilated cardiomyopathy. Hum Mol Genet 2014. http://dx.doi.org/10.1093/hmg/ ddu297. [17] Wang H, Chen C, Song X, Chen J, Zhen Y, Sun K, et al. Mef2c is an essential regulatory element required for unique expression of the cardiac-specific CARK gene. J Cell Mol Med 2008;12:304–15. http://dx.doi.org/10.1111/j.1582-4934.2007.00155.x.

A. Milano et al. / Journal of Molecular and Cellular Cardiology 82 (2015) 167–173 [18] Martin JF, Schwarz JJ, Olson EN. Myocyte enhancer factor (MEF) 2C: a tissuerestricted member of the MEF-2 family of transcription factors. Proc Natl Acad Sci U S A 1993;90:5282–6. http://dx.doi.org/10.1073/pnas.90.11.5282. [19] Sartorelli V, Huang J, Hamamori Y, Kedes L. Molecular mechanisms of myogenic coactivation by p300: direct interaction with the activation domain of MyoD and with the MADS box of MEF2C. Mol Cell Biol 1997;17:1010–26. [20] Rosenthal N, Harvey RP. Heart development and regeneration volume II; 2010 1–496. [21] Mosavi LK, Cammett TJ, Desrosiers DC, Peng Z. The ankyrin repeat as molecular architecture for protein recognition. Protein Sci 2004;13:1435–48. http://dx.doi.org/ 10.1110/ps.03554604.ity. [22] Feng Y, Cao HQ, Liu Z, Ding JF, Meng XM. Identification of the dual specificity and the functional domains of the cardiac-specific protein kinase TNNI3K. Gen Physiol Biophys 2007;26:104–9. [23] Feng Y, Liu D-Q, Wang Z, Liu Z, Cao H-Q, Wang L-Y, et al. AOP-1 interacts with cardiac-specific protein kinase TNNI3K and down-regulates its kinase activity. Biochemistry 2007;72:1199–204. http://dx.doi.org/10.1134/S0006297907110053. [24] Cardoso TC, Ferrari HF, Garcia AF, Bregano LC, Andrade AL, Nogueira AH. Immunohistochemical approach to the pathogenesis of clinical cases of bovine herpes virus type 5 infections. Diagn Pathol 2010;5:57. http://dx.doi.org/10.1186/1746-1596-5-57. [25] Wang H, Wang L, Song L, Zhang YW, Ye J, Xu RX, et al. TNNI3K is a novel mediator of myofilament function and phosphorylates cardiac troponin I. Br J Med Biol Res 2013; 46:128–37. http://dx.doi.org/10.1590/1414-431X20122515. [26] Layland J, Solaro RJ, Shah AM. Regulation of cardiac contractile function by troponin I phosphorylation. Cardiovasc Res 2005;66:12–21. http://dx.doi.org/10.1016/j. cardiores.2004.12.022. [27] Frank D, Frey N. Cardiac Z-disc signaling network. J Biol Chem 2011;286:9897–904. http://dx.doi.org/10.1074/jbc.R110.174268. [28] Knöll R, Buyandelger B, Lab M. The sarcomeric Z-disc and Z-discopathies. J Biomed Biotechnol 2011;2011:569628. http://dx.doi.org/10.1155/2011/569628. [29] Scicluna BP, Tanck MWT, Remme CA, Beekman L, Coronel R, Wilde AAM, et al. Quantitative trait loci for electrocardiographic parameters and arrhythmia in the mouse. J Mol Cell Cardiol 2011;50:380–9. http://dx.doi.org/10.1016/j. yjmcc.2010.09.009. [30] Aly M, Wiltshire S, Chahrour G, Osti J-CL, Vidal SM. Complex genetic control of host susceptibility to coxsackievirus B3-induced myocarditis. Genes Immun 2007;8: 193–204. http://dx.doi.org/10.1038/sj.gene.6364374. [31] Wiltshire S a, Leiva-Torres GA, Vidal SM. Quantitative trait locus analysis, pathway analysis, and consomic mapping show genetic variants of Tnni3k, Fpgt, or H28 control susceptibility to viral myocarditis. J Immunol 2011;186:6398–405. http://dx.doi. org/10.4049/jimmunol.1100159. [32] Flint J, Valdar W, Shifman S, Mott R. Strategies for mapping and cloning quantitative trait genes in rodents. Nat Rev Genet 2005;6:271–86. http://dx.doi.org/10.1038/ nrg1576. [33] Suzuki M, Carlson KM, Marchuk DA, R. H. A. Genetic modifier loci affecting survival and cardiac function in murine dilated cardiomyopathy. Circulation 2002;105: 1824–9. http://dx.doi.org/10.1161/01.CIR.0000014926.32463.89. [34] Wheeler FC, Fernandez L, Carlson KM, Wolf MJ, Rockman H a, Marchuk D a. QTL mapping in a mouse model of cardiomyopathy reveals an ancestral modifier allele

[35]

[36]

[37]

[38]

[39] [40]

[41]

[42]

[43]

[44]

[45]

[46] [47]

173

affecting heart function and survival. Mamm Genome 2005;16:414–23. http://dx. doi.org/10.1007/s00335-005-2468-7. Auerbach SS, Thomas R, Shah R, Xu H, Vallant MK, Nyska A, et al. Comparative phenotypic assessment of cardiac pathology, physiology, and gene expression in C3H/ HeJ, C57BL/6J, and B6C3F1/J mice. Toxicol Pathol 2010;38:923–42. http://dx.doi. org/10.1177/0192623310382864. Lai Z, Chen Y, Feng L, Meng X, Feng Ding J, Wang L, et al. Overexpression of TNNI3K, a cardiac-specific MAP kinase, promotes P19CL6-derived cardiac myogenesis and prevents myocardial infarction-induced injury. Am J Physiol Heart Circ Physiol 2008; 295:H708–16. http://dx.doi.org/10.1152/ajpheart.00252.2008. Cho M-C, Rapacciuolo a, Koch WJ, Kobayashi Y, Jones LR, Rockman HA. Defective-adrenergic receptor signaling precedes the development of dilated cardiomyopathy in transgenic mice with calsequestrin overexpression. J Biol Chem 1999;274:22251–6. http://dx.doi.org/10.1074/jbc.274.32.22251. Hausenloy DJ, Yellon DM. Myocardial ischemia–reperfusion injury : a neglected therapeutic target. J Clin Invest 2013;123:92–100. http://dx.doi.org/10.1172/ JCI62874.92. Abraham DM, Marchuk DA. Commentary on cutting edge science. Circ Res 2014; 114:938–40. http://dx.doi.org/10.1126/scitranslmed.3006479. Remme CA, Verkerk AO, Nuyens D, Ginneken ACG van, Brunschot S van, Belterman CNW, et al. Overlap syndrome of cardiac sodium channel disease in mice carrying the equivalent mutation of human SCN5A-1795insD. Circulation 2006;114: 2584–94. http://dx.doi.org/10.1161/CIRCULATIONAHA.106.653949 [CIRCULATIONAHA.106.653949 [pii]]. Bezzina CR, Veldkamp MW, Berg MP van den, Postma AV, Rook MB, Viersma J, et al. A single Na(+) channel mutation causing both long-QT and Brugada syndromes. Circ Res 1999;85:1206–13. http://dx.doi.org/10.1161/01.RES.85.12.1206. Remme CA, Scicluna BP, Verkerk AO, Amin AS, Brunschot S van, Beekman L, et al. Genetically determined differences in sodium current characteristics modulate conduction disease severity in mice with cardiac sodium channelopathy. Circ Res 2009;104:1283–92. http://dx.doi.org/10.1161/CIRCRESAHA.109.194423 [CIRCRESAHA.109.194423 [pii]]. Colak D, Kaya N, Al-Zahrani J, Al Bakheet A, Muiya P, Andres E, et al. Left ventricular global transcriptional profiling in human end-stage dilated cardiomyopathy. Genomics 2009;94:20–31. http://dx.doi.org/10.1016/j.ygeno.2009.03.003. Speliotes EK, Willer CJ, Berndt SI, Monda KL, Thorleifsson G, Jackson AU, et al. Association analyses of 249,796 individuals reveal 18 new loci associated with body mass index. Nat Genet 2010;42:937–48. http://dx.doi.org/10.1038/ng.686. Graff M, Ngwa JS, Workalemahu T, Homuth G, Schipf S, Teumer A, et al. Genomewide analysis of BMI in adolescents and young adults reveals additional insight into the effects of genetic loci over the life course. Hum Mol Genet 2013;22: 3597–607. http://dx.doi.org/10.1093/hmg/ddt205. Pastuszak I. GDP-L-fucose pyrophosphorylase. J Biol Chem 1998;273:30165–74. http://dx.doi.org/10.1074/jbc.273.46.30165. Lal H, Ahmad F, Parikh S, Force T. Troponin I-interacting protein kinase: a novel cardiac-specific kinase, emerging as a molecular target for the treatment of cardiac disease. Circ J 2014;78:1514–9.