1-s2.0-s0163725810000793-main

Upload: dawn-kent

Post on 02-Jun-2018

212 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/11/2019 1-s2.0-S0163725810000793-main

    1/37

  • 8/11/2019 1-s2.0-S0163725810000793-main

    2/37

    Contents

    1. Cardiac hypertrophy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192

    2. Experimental and genetic mouse models utilized in the identication of

    signaling pathways that mediate cardiac hypertrophy . . . . . . . . . . . . . . . . . . . . . . . . . . 192

    3. Overview of signaling cascades/proteins implicated

    in mediating physiological and pathological cardiac growth . . . . . . . . . . . . . . . . . . . . . . . 192

    4. Molecular mechanisms associated with structural features of pathological and physiological hypertrophy . . . 193

    5. Molecular mechanisms associated with differences in

    energy metabolism in pathological and physiological hypertrophy . . . . . . . . . . . . . . . . . . . . 1936. Characteristic gene expression changes associated with pathological and physiological hypertrophy . . . . 194

    7. Gender differences in cardiac hypertrophy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194

    8. Therapeutic strategies for the treatment of heart failure . . . . . . . . . . . . . . . . . . . . . . . . . 194

    9. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195

    Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

    References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

    1. Cardiac hypertrophy

    1.1. Introduction and overview

    Cardiac hypertrophy can broadly be dened as an increase in heartmass. Growth of the postnatal heart is closely matched to its

    functional load (Zak, 1984). In response to an increase in load (e.g.

    pressure overload in a setting of hypertension), the heart must work

    harder than under normal conditions. To counterbalance the chronic

    increase in wall stress the muscle cells within the heart enlarge

    leading to an increasein size and mass (Cooper, 1987; Sugden & Clerk,

    1998; Hunter & Chien, 1999). The increase in heart mass is largely due

    to an increase in ventricular weight. In the subsequent sections we

    have described cardiac hypertrophy at the cellular level, different

    types of cardiac hypertrophy (pathological and physiological), the

    molecular mechanisms responsible for different forms of cardiac

    hypertrophy, gender differences, and possible treatment strategies

    based on the distinct molecular mechanisms associated withphysiological and pathological cardiac hypertrophy.

    1.2. Cardiac hypertrophy at the cellular level

    The heart is composed of cardiac myocytes (muscle cells), non-

    myocytes (e.g.broblasts, endothelial cells, mast cells, vascular smooth

    musclecells), and the surrounding extracellularmatrix (Nag, 1980; Zak,

    1984). Ventricular cardiacmyocytesmake up only one-third of the total

    Fig. 1. Cellular processes involved in the development of cardiac hypertrophy. ECM: extracellular matrix, FAO: fatty acid oxidation, GPCR: G protein-coupled receptor, MAPK:

    mitogen-activated protein kinase, PI3K: phosphoinositide 3-kinase, ROS: reactive oxygen species, SERCA: sarcoplasmic reticulum Ca2+

    ATPase.

    196

    197

    207

    209210

    212

    214

    217

    217

    217

    192 B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    3/37

    cell number, butaccount for7080% ofthe heart'smass (Nag, 1980; Zak,

    1984; Popescu et al., 2006). In mammals, at birth or soon after, the

    majority of cardiac myocytes lose the ability to proliferate, thus heart

    growth occurs primarily via an increase in myocyte size (Soonpaa et al.,

    1996). The inability of adult cardiac myocytes to divide has come under

    some debate (Anversa & Nadal-Ginard, 2002; Anversa et al., 2002;

    Pasumarthi & Field, 2002). However, estimates of DNA labeling indicate

    that DNA synthesis is taking place in a very small fraction of the total

    adultcardiac myocyte population (Nakagawa et al.,1988;Soonpaa et al.,1996; MacLellan & Schneider, 2000; Anversa et al., 2002; Pasumarthi &

    Field, 2002), indicating thatthe postnatal heart enlarges primarily by an

    increase in myocyte size.

    Myocytes are composed of bundles of myobrils. Myobrils contain

    myolaments which consist of sarcomeres, the basic contractile unit of

    the heart. Myocytes are arranged in a circumferential and spiral

    orientation around the left ventricle, and need to contract simulta-

    neously to ensure the heart pumps with a normal rhythm. Intercalated

    discs,located at thebipolar ends of cardiacmyocytes, areresponsiblefor

    maintaining cellcell adhesion while allowing contractile force to be

    transmitted between adjacent cardiac myocytes (Estigoy et al., 2009).

    Growth of cardiac myocytes is dependent on the initiation of several

    events in response to an increase in functional load, including activation

    of signaling pathways, changes in gene expression, increases in the rate

    of protein synthesis, and the organization of contractile proteins into

    sarcomeric units (Fig. 1). Cardiac myocytes appear to have an intrinsic

    mechano-sensing mechanism. Stretch sensitive ion channels present in

    the plasmamembraneof cardiac myocytes and structural proteins (such

    as integrins) play a rolein linking the extracellular matrix, cytoskeleton,

    sarcomere, calcium handling proteins and nucleus (Knoll et al., 2003;

    Hoshijima, 2006). Thus, there is an interactive continuum from integrins

    at the cell surface to the contractile apparatus and nucleus ( Fig. 1).

    1.3. Association between cardiac hypertrophy and heart failure

    Understanding the molecular mechanisms responsible for the

    induction of cardiac hypertrophy has been of great interest due to the

    known association between cardiac hypertrophy and nearlyall forms of

    heart failure (Levy et al., 1990). Cardiac hypertrophy is also an

    independent risk factor for myocardial infarction, arrhythmia and

    sudden death (Levy et al., 1990). In response to a chronic increase in

    load, there is an initial increase in heart mass to normalize wall stress

    and permit normal cardiovascular function at rest i.e. compensated

    growth. However, if the chronic increase in wall stress is not relieved,

    the hypertrophied heart can dilate, contractile function falls and the

    heart can fail.Heart failure affects approximately 13% of people in Western

    society. The incidence of heart failure increases withage, affecting 34%

    of those over45 years old, 5% of those aged between 60 and 69 years of

    age, and 10% of people over the age of 70 (Davies et al., 2001; AIHW,

    2004; Thom et al., 2006; Lloyd-Jones et al., 2009). Symptoms of heart

    failure patients include fatigue, insomnia, anxiety, depression,shortness

    of breath, edema, dizziness, and nausea, all of which contribute to a

    reduced qualityof life for these patients (Blinderman et al., 2008). With

    an aging population, rising rates of obesity and diabetes, as well as the

    availability of interventions that prolong survival following cardiac

    insults, the incidence of heart failure is likely to rise over the coming

    decades. Thecostsassociated with an expandingnumber of patients and

    specialized treatment strategies are expected to contribute signicantly

    to the economic burden caused by heart failure (Blinderman, et al.,

    2008). Currentlythere is no cure forheart failure, andlong term survival

    following heart failure remains poor, with one third of patients dying

    within a year of diagnosis (Zannad et al., 1999; Cowie et al., 2000;

    Bleumink et al., 2004; McMurray & Pfeffer, 2005). Thus, a number of

    studies have focused on identifying the molecular mechanisms

    associated with cardiac hypertrophy and the transition to heart failure,

    to identify new therapeutic targets to prevent or reverse cardiac

    hypertrophy and heart failure.

    1.4. Cardiac hypertrophy and the athlete's heart

    Theathlete'shearthas generally been dened asa benignincreasein

    heart mass, associated with morphological alterations, thatrepresents a

    Fig. 2.Cardiac hypertrophy can be classied as physiological, which occurs during pregnancy or in response to chronic exercise training, is reversible and characterized by normal

    cardiac morphology and function. In contrast, hypertrophy that occurs in settings of disease is detrimental for cardiac structure and function and can lead to heart failure.

    Developmental hypertrophyis associatedwith the normal growth of the heart after birthuntil adulthood. RV: right ventricle, LV: leftventricle. Normal/ physiological heart growth is

    shown in green, pathological heart growth is shown in red.

    193B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    4/37

    physiological adaptation to chronic training. Though, with attention

    from media reports of sudden death in young athletes, it has been

    questioned whether highly trained athletes develop pathological

    conditions. Notably, there is currently no evidence in the healthy

    population (excluding persons with underlying cardiovascular disease

    or genetic disorders) showing that remodeling due to exercise training

    leads to long-termcardiac disease progression,cardiovascular disability,

    or sudden cardiac death (Maron & Pelliccia, 2006). The overall risk of

    sudden death in athletes is not well de

    ned but considered low. A 12-year surveyin high-school athletes participatingin organized sportsin a

    US state (Maron et al., 1998) reported a frequency of sudden death of

    1:200,000 per year (based on only 3 deathsamong 1.4 million students;

    including27 sports). Suddendeathin young trainedathletesin response

    to physical exertion has largely been causally linked to congenital but

    clinically unsuspected cardiovascular disease (see Maron, 2003). In

    large autopsy-based surveys of athletes in the US, hypertrophic

    cardiomyopathy is the most common cause of sudden death (account-

    ing for about one-third of events), followed by congenital coronary-

    artery anomalies. A wide range of other, largely congenital malforma-

    tions account for the remaining sudden deaths from cardiovascular

    disease among athletes (Maron, 2003). Thus, it is generally accepted

    that cardiac hypertrophy in response to exercise is protective, in some

    instances improves cardiac function, and does not progress to heart

    failure. A comprehensive understanding of why cardiac hypertrophy

    progresses to heart failure in a setting of disease, but does not in

    response to exercise, is considered important for identifying and

    targeting the critical molecular mechanisms responsible for the

    transition from hypertrophy to heart failure.

    1.5. Distinct forms of cardiac growth and hypertrophy

    1.5.1. Pathological and physiological cardiac growth and hypertrophy

    Cardiac growth or hypertrophy can broadly be classied as either

    physiological (normal) or pathological (detrimental). Physiolog-

    ical heart growth includes normal postnatal growth, pregnancy-

    induced growth, and exercise-induced cardiac hypertrophy. In

    contrast, pathological growth occurs in response to chronic pressure

    or volume overload in a disease setting (e.g. hypertension, valvular

    heart disease), myocardial infarction or ischemia associated with

    coronary artery disease, or abnormalities/conditions that lead to

    cardiomyopathy (e.g. inherited genetic mutations, diabetes) (Fig. 2).

    Both physiological and pathological heart growth are associated with

    an increase in heart size, however pathological hypertrophy is also

    typically associated with loss of myocytes and brotic replacement,

    cardiac dysfunction, and increased risk of heart failure and sudden

    death (Levy et al., 1990; Weber et al., 1993; Cohn et al., 1997). In

    contrast, physiological growth is associated with normal cardiacstructure, normal or improved cardiac function, and is reversible in

    the instance of exercise- or pregnancy-induced hypertrophy (Ferrans,

    1984; Schaible & Scheuer, 1984; Fagard, 1997)(Fig. 2).

    1.5.2. Concentric and eccentric hypertrophy

    Pathological and physiological hypertrophy has classically been

    subdivided as concentricor eccentric. Theseclassications arebasedon

    changes in shape, which is dependent on the initiating stimulus

    (Grossman et al., 1975; Pluim et al., 2000) (Fig. 3). Concentric

    hypertrophy refers to an increase in relative wall thickness and

    cardiac mass, with a small reduction or no change in chamber volume.

    Concentric hypertrophy is characterized by a parallel pattern of

    sarcomere addition leading to an increase in myocyte cell width

    (Fig. 3). Eccentric hypertrophy refers to an increase in cardiac mass

    with increased chamber volume, i.e. dilated chambers. Relative wall

    thickness may be normal, decreased, or increased. In eccentric

    hypertrophy, addition of sarcomeres in series leads to an increase in

    myocyte cell length (Fig. 3) (Grossmanet al., 1975).

    A pathological stimulus causing pressure overload (e.g. hyperten-

    sion, aorticstenosis) produces an increase in systolic wall stresswhich

    results in concentric hypertrophy (Grossman et al., 1975). In contrast,

    a stimulus causing volume overload (e.g. aortic regurgitation,

    arteriovenous stulas) produces an increase in diastolic wall stress

    and results in eccentric hypertrophy (Grossman et al., 1975; Pluim et

    al., 2000). Clinical studies suggest that eccentric cardiac hypertrophy

    induced by pathological stimuli poses a greater risk than concentric

    cardiac hypertrophy (Berenji et al., 2005).

    Physiological stimuli can also produce concentric or eccentric

    hypertrophy. Aerobic exercise (also referred to as endurance training,

    Fig. 3.Different stimuli induce different forms of cardiac hypertrophy. Pressure overload causes thickening of the left ventricle wall due to the addition of sarcomeres in parallel and

    results in concentric hypertrophy. Volume overload induces an increase in muscle mass via the addition of sarcomeres in series and results in eccentric hypertrophy.

    194 B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    5/37

    isotonic or dynamic exercise e.g. long-distance running, swimming)

    and pregnancy increase venous return to the heart resulting in

    volume overload and eccentric hypertrophy (Zak, 1984; Pluim et al.,

    2000; Eghbali et al., 2005). This type of eccentric hypertrophy is

    usually characterized by chamber enlargement and a proportional

    change in wall thickness, whereas eccentrichypertrophy in settings of

    disease is generally associated with thinning of the ventricular walls.

    Strength training (also referred to as isometric or static exercise, e.g.

    weight lifting, wrestling, throwing heavy objects) results in a pressure

    load on theheart ratherthan volumeload andconcentric hypertrophy

    (Zak 1984; Pluim et al., 2000)(Fig. 3).

    1.6. Distinct features of pathological and physiological hypertrophy

    Despite comparable increases in heart size, pathological and

    physiological hypertrophy are associated with distinct 1) structural

    and functional, 2) metabolic, and 3) biochemical and molecular

    features (Fig. 4,Table 1).

    Fig. 4. Fourdistinctfeatures of cardiac hypertrophy includeheart size,cardiac function, cardiac brosis and gene expression. The upper leftquadrant showsrepresentative images of

    mouse hearts that have been subjected to a pathological (aortic banding; band) stimulus for one week, compared with sham (sham) operated controls, or mice subjected to a

    physiological stimulus for four weeks (chronic swim training; exercise) compared with sedentary controls. An increase in heart s ize is observed in mice that have undergone aortic

    banding (band) or chronic swim training (exercise) compared to sham and sedentary controls. The upper right quadrant depicts cardiac function as shown by M-mode

    echocardiography. Cardiac function is depressed in a mouse model of pathological (decompensated) hypertrophy but is conserved in a mouse model of physiological hypertrophy

    (exercise). The lower left quadrant depicts histological analysis of heart sections stained with Masson's trichrome; increased brosis is shown in blue and is only present in

    pathological hypertrophy. Representative sections from the left ventricular wall of untrained mice (control), aortic banded mice (pressure overload, band) and swim trained mice

    (exercise). Bar = 10 m. Sections from sham operated mice were similar to those of untrained (control). The lower right quadrant shows gene expression changes associated with

    cardiac hypertrophy. Representative Northern blot showing total RNA from ventricles of sham, aortic banded (band), untrained (sedentary) and trained (exercise) mice. Expression

    of GAPDH was determined to verify equal loading of RNA. There is increased expression of atrial natriuretic peptide (ANP), B-type natriuretic peptide (BNP),-myosin heavy chain

    (

    -MHC) and

    -skeletal actin, and decreased expression of sarcoplasmic reticulum Ca

    2+

    ATPase (SERCA) and

    -MHC in mice subjected to aortic banding (band) compared tocontrols, while gene expression remains relatively unchanged in exercised trained mice compared to sedentary controls.

    Table 1

    Features of pathological and physiological hypertrophy.

    Feature Pathological cardiac hypertrophy Physiological cardiac hypertrophy

    Stimuli Disease Aerobic exercise training

    Pressure or volume overload Postnatal growth

    Cardiomyopathy (familial, viral, diabetes, metabolic, alcoholic/toxic) Pregnancy

    Cardiac morphology Increased myocyte volume Increased myocyte volume

    Formation of new sarcomeres Formation of new sarcomeres

    Increase in heart size Increase in heart size

    Cardiacbrosis Yes No

    Apoptosis Yes No

    Fetal gene expression Upregulation of ANP, BNP,-MHC, and-skeletal actin Relatively unchanged

    Expression of genes associated with contractile function Downregulation of SERCA2a,-MHC Normal or increased

    Cardiac function Depressed Normal or enhanced

    Metabolism Decreased fatty acid oxidation Enhanced fatty acid oxidation

    Increased glucose utilization Enhanced glucose utilization

    Reversible No Yes

    Association with heart failure and mortality Yes No

    195B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    6/37

    1.6.1. Structural and functional features

    Cardiac hypertrophy is associated with structural remodeling of

    components of the ventricular walls to accommodate increases in

    myocyte size, including changes in the brillar collagen network

    and angiogenesis. Under basal conditions or a setting of physiolog-

    ical hypertrophy, the brillar collagen network provides structural

    integrity of adjoining myocytes, facilitating myocyte shortening

    which translates into efcient cardiac pump function (Gunasinghe

    & Spinale, 2004). Pathological hypertrophy is associated with celldeath (apoptosis, necrosis) and the loss of myocytes is replaced

    with excessive collagen (known as brosis). The main brillar

    collagen present in cardiac brosis is type 1 collagen. Excessive

    accumulation of collagen stiffens the ventricles, which impairs

    contraction and relaxation, impairs the electrical coupling of cardiac

    myocytes with extracellular matrix proteins, and reduces capillary

    density. Fibrosis and reduced capillary density increases oxygen

    diffusion distances, leading to myocardial ischemia, and is likely to

    contribute to the transition from hypertrophy to failure (Guna-

    singhe & Spinale, 2004).

    1.6.2. Cardiac metabolism

    In the normal healthy heart, fatty acid oxidation is the main

    metabolic pathway responsible for generating energy, accounting for

    6070% of ATP production (van der Vusse et al., 1992); glucose and

    lactate metabolism account for approximately 30% of ATP synthesis.

    The heart is capable of switching energy substrates depending on

    workload and the relative concentrations of fuel molecules in the

    bloodstream (seevan der Vusse et al., 1992). This is considered an

    adaptive mechanism which allows the heart to produce a continuous

    supply of ATP under various physiological conditions (e.g. fasting,

    during exercise, etc.).

    Pathological cardiac hypertrophy is associated with decreases in

    fatty acid oxidation and increases in glucose metabolism (Allard et al.,

    1994; Christe & Rodgers 1994; Davila-Roman et al., 2002). This switch

    in substrate utilization may be a protective mechanism, allowing the

    heart to produce more ATP per molecule of oxygen consumed (see

    van Bilsen et al., 2009). This is reminiscent of what occurs during fetal

    cardiac development, when oxygen supply is limited and fatty acidtransport and metabolism are impaired (due to carnitine deciency

    and delayed maturation of enzymes involved in fatty acid oxidation).

    Thus, glucose is the primary substrate used by the fetal heart to

    generate ATP (Ostadal, et al., 1999). In contrast, physiological cardiac

    hypertrophy induced by exercise training is characterized by

    enhanced fatty acid and glucose oxidation (Gertz et al., 1988). Of

    note, in advanced pathological hypertrophy and failure, glucose

    metabolism decreases as the heart becomes resistant to insulin,

    reducing the overall ability of the heart to generate sufcient ATP (see

    Neubauer, 2007).

    1.6.3. Biochemical and molecular features

    In the 1970s and early 1980s it was recognized that physiological

    hypertrophy (induced by exercise training/ thyroid hormone) wasassociated with elevations in myosin ATPase activity and enhance-

    ment of contractility, whereas pathological hypertrophy (induced by

    renal hypertension, aortic banding) was associated with decreased

    myosin ATPase activity and depressed contractile function (Wikman-

    Coffelt et al., 1979; Rupp, 1981). Since then, there have been a number

    of studies demonstrating that physiological and pathological cardiac

    hypertrophy are associated with some distinct biochemical and

    molecular signatures.Iemitsu et al. (2001)compared mRNA expres-

    sion in a rat model of pathological cardiac hypertrophy (spontane-

    ously hypertensive rat) and physiological hypertrophy (chronic swim

    training). The investigators reported a distinct pattern of gene

    expression in the two models (Iemitsu et al., 2001). It is now well

    recognized that pathological cardiac hypertrophy is associated with

    distinctalterations in cardiac contractile proteins (- and -myosin

    heavy chain (MHC)), increased expression of fetal genes (e.g. atrial

    natriuretic peptide (ANP), B-type natriuretic peptide (BNP), -

    skeletal actin) but down-regulation of calcium-handling proteins

    (e.g. sarcoplasmic reticulum Ca2+-ATPase 2a (SERCA2a)). Since this

    time, generation of numerous transgenic and knockout mouse models

    in combination with models of physiological and pathological

    hypertrophy have allowed investigators to delineate signaling

    proteins that appear to play distinct roles in regulating physiological

    and pathological cardiac hypertrophy (discussed in detail inSection 3).

    2. Experimental and genetic mouse

    models utilized in the identication of

    signaling pathways that mediate cardiac hypertrophy

    The denition of cardiac hypertrophy as either physiological or

    pathological has not been without contention (Dorn et al., 2003).

    However, there is now substantial evidence from animal studies that

    the different phenotypes associated with pathological and physiolog-

    ical hypertrophy can be due to distinct stimuli and activation of some

    distinct signaling pathways, at least under certain conditions.

    Studies utilizing genetic mouse models (transgenic and knockout)

    alone or in combination with morphologically distinct models of

    hypertrophy (e.g. pathological, physiological, concentric and eccen-

    tric) have become powerful tools for understanding molecular

    pathways responsible for different forms of heart growth in vivo.

    Genetic mouse models have typically utilized the-MHC promoter to

    achieve cardiac myocyte specic expression (Subramaniam et al.,

    1991), or the Cre-loxP systemto generate cardiac-specic or inducible

    knockout mice (Chien, 2001). Inducible transgenic mouse models

    have also been valuable as they allow investigators to switch on the

    activity of the protein of interest in myocytes by injection of a drug

    (e.g. tamoxifen) at a specic time point, i.e. after completion of

    developmental growth (Fan et al., 2005b;Hoesl et al., 2008; Lu et al.,

    2009; Ruan et al., 2009). Commonly used experimental mouse models

    of pathological hypertrophy include pressure overload (constriction/

    banding of the renal, abdominal, ascending or transverse aorta),

    volume overload (aortocaval shunt), and minipump infusions ofvasoactive substances (e.g. isoproterenol, angiotensin II (Ang II)).

    Physiological models include treadmill running, freewheel running

    and chronic swim training. Below, we have described hypertrophic

    triggers/stimuli, signaling proteins and cascades which appear to play

    an important role for the development of pathological or physiological

    hypertrophy.

    2.1. Hypertrophic triggers/stimuli

    In response to hemodynamic overload, cardiac myocytes are

    subjected to mechanical stretch, and autocrine and paracrine humoral

    factors including Ang II, endothelin 1 (ET-1), insulin-like growth

    factor 1 (IGF1), transforming growth factor- (TGF-) and cardio-

    trophin 1 (CT-1) are released. These factors bind to receptors oncardiac cells, in turn activating intracellular signaling pathways that

    leads to cell growth. Signaling cascades and proteins responsible for

    cardiac growth and hypertrophy are complex and extensive crosstalk

    has been identied (Fig. 5). The subsequent sections of this review

    focus on signaling cascades and proteins that have been reported to

    play distinct roles in regulating pathological and physiological

    hypertrophy.

    2.1.1. Physiological and pathological triggers/stimuli

    Human and animal studies have demonstrated that certain factors

    are preferentially released in response to pathological and physiolog-

    ical stimuli. It is well recognized that IGF1 is released during postnatal

    development and in response to exercise training (Yeh et al., 1994;

    Conlon and Raff 1999; Koziris et al., 1999; Neri Serneri et al., 2001b;

    196 B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    7/37

    Perrino et al., 2006), and IGF1 levels in the heart were increased in

    swim-trained rats (Scheinowitz et al., 2003). Furthermore, production

    of cardiac IGF1 (but not Ang II or ET-1) was increased in professional

    athletes compared with control subjects (Neri Serneri et al., 2001b).In

    contrast,pressure overload is associated with elevated levelsof AngII,

    catecholamines and ET-1 (Schunkert et al., 1990; Arai et al., 1995;

    Yamazaki et al., 1999; Rapacciuolo et al., 2001; Yayama et al., 2004),

    and cardiac formation of Ang II was increased in heart failure patients

    with hypertrophied hearts (Neri Serneri et al., 2001a).

    3. Overview of signaling cascades/proteins implicatedin mediating physiological and pathological cardiac growth

    The best characterized signaling cascades responsible for mediat-

    ing physiological and pathological cardiac hypertrophy are the IGF1-

    phosphoinositide 3-kinase [PI3K, (p110)]-Akt pathway and Gq

    signaling (downstream of G protein-coupled receptors (GPCR)

    activated by Ang II, ET-1 and catecholamines), respectively (Fig. 6).

    Other signaling pathways associated with physiological cardiac

    hypertrophy and/or protection include the gp130/JAK/STAT pathway,

    thyroid hormone signaling, and heat shock transcription factor 1

    (HSF1). In contrast, pathological hypertrophy has also been associated

    with abnormalities leading to enhanced PI3K(p110), mitogen

    activated protein kinases (MAPKs), protein kinase C (PKC) and D

    (PKD), and calcineurin.

    3.1. Signaling proteins/pathways

    implicated in mediating physiological hypertrophy

    3.1.1. IGF1-PI3K(p110)-Akt pathway

    Substantial evidence from genetic mouse models has demonstrat-

    ed the critical role of the IGF1-PI3K(p110)-Akt pathway in

    regulating physiological cardiac growth.

    3.1.1.1. Insulin-like growth factor (IGF1) receptor signaling.IGF1 is best

    known for being produced by the liver in response to growth

    hormone stimulation and is essential for normal fetal and postnatal

    growth and development (Adams et al., 2000). IGF1 is also produced

    by theheart(see reviews: Renet al., 1999; McMullen,2008) and binds

    to a cell surface receptor, insulin-like growth factor 1 receptor

    (IGF1R), a receptor tyrosine kinase that activates downstream

    signaling proteins. A number of studies have examined the role of

    IGF1 in the heart using gene targeted mice.

    a) Mice with increased cardiac myocyte specic expression of IGF1

    (human IGF1B transgene expression driven by the -MHC

    promoter) had enlarged hearts with normal cardiac function

    (Reiss et al., 1996). A confounding factor of this study was that

    transgene expression increased IGF1 secretion from cardiac

    myocytes which resulted in a signicant rise in systemic plasma

    levels of IGF1 (approximately 80%) and an increase in other organ

    weights. The increase in heart size was attributed to an increase in

    Fig. 5.A schematic of the major signaling pathways involved in cardiac hypertrophy, showing cross-talk and integration of various pathways. N.B. due to the complex nature of

    signaling cascades and on-going discoveries it was not possible to illustrate all interactions.

    197B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    8/37

    cardiac myocyte number rather than myocyte size. This result was

    unexpected given that the majority of mammalian cardiac

    myocytes are thought to lose their ability to proliferate at birth

    or within the rst few weeks of postnatal life (Soonpaa et al.,

    1996).

    b) In a second study, IGF1 transgenic mice were generated using the

    -skeletal actin promoter (transgene expression in heart and

    skeletal muscle) (Delaughter et al., 1999). Serum IGF1 levels were

    not elevated but persistent transgenic expression was associatedwith increases in gut, liver, and spleen weight (Fiorotto et al.,

    2003). Up to 10 weeks of age, IGF1 transgenic mice displayed

    cardiac hypertrophy, which was associated with enhanced cardiac

    systolic function i.e. physiological hypertrophy (Delaughter et al.,

    1999). However, the investigators concluded that the physiolog-

    ical cardiac phenotype ultimately progressed to pathological

    hypertrophy, because mice had depressed cardiac function by

    12 months of age (Delaughter et al., 1999).

    c) We examined the role of IGF1 specically in cardiac myocytes by

    over-expressing the IGF1R rather than IGF1 (transgenic expression

    of human IGF1R using the -MHC promoter) (McMullen et al.,

    2004b). Expression of the IGF1R was considered an advantage

    because it allowed examination of IGF1 signaling in the absence of

    effects of secreted IGF1 on other tissues or non-myocytes. At

    3 months of age, IGF1R transgenic mice had enlarged hearts

    (approximately 40% increase in weight, with a proportional

    increase of all chambers and ventricular wall thickness), increased

    myocyte size, no evidence of histopathology (e.g. necrosis, brosis,

    myocyte disarray) and enhanced systolic function (McMullen et

    al., 2004b). These characteristic features of physiological hyper-

    trophywere maintained at 1216 months of age (IGF1R transgenic

    mice had enhanced systolic function in comparison to non-

    transgenic mice) (McMullen et al., 2004b). Thus, IGF1R transgenic

    mice did not progress to pathological hypertrophy with aging,

    which was observed in the earlier study (model b) ( Delaughter et

    al., 1999). Consistent with the hypothesis that IGF1 activates PI3K

    (p110) and Akt (a known downstream target of PI3K) to induce

    physiological cardiac hypertrophy (Fig. 6), activation of PI3K and

    phosphorylationof Aktwere elevated in heartsof IGF1R transgenic

    mice (McMullen et al., 2004b). In contrast, signaling proteinsdownstream of Gq (implicated in pathological hypertrophy)

    including MAPKs and calcineurin were not activated in hearts of

    IGF1R transgenic mice (McMullen et al., 2004b).

    d) In corroboration with the idea that IGF1 signaling is critical for

    physiological heart growth, cardiac myocyte-specic ablation of

    the IGF1R gene in mice attenuated the hypertrophic response to

    swim exercise training compared to non-transgenic mice (Kim et

    al., 2008b). A basal cardiac phenotype was not observed.

    3.1.1.2. Phosphoinositide 3-kinase (PI3K, p110) signaling. PI3Ks are a

    family of enzymes and have been linked to a diverse group of cellular

    functions, particularly cell growth, survival, differentiation, and

    proliferation (Cantley, 2002). PI3K is a lipid kinase that releases

    inositol lipid products from the plasma membrane which in turn

    mediate intracellular signaling (Toker & Cantley, 1997; Vanhaeseb-

    roeck et al., 1997). Activation of PI3Ks is coupled to both receptor

    tyrosine kinases (e.g. insulin receptor and IGF1R) and GPCRs. There

    are three major classes of PI3Ks (classes I, II and III), which are

    classied based on sequence homology in the catalytic domain,

    structure and substrate specicity (Vanhaesebroeck et al., 2001; Kok

    et al., 2009). Class I PI3Ks are heterodimers and further divided into

    Fig. 6. A schematicoverview of pathological andphysiologicalhypertrophy outliningkey differences in initiatingstimuli,signaling pathways, cellular responsesand cardiac function.

    For simplicity we have focussed on the best characterized signaling pathways implicated in mediating pathological (shaded red) and physiological (shaded green) cardiac

    hypertrophy. Other important mediators are described in detail in Section 3. Ang II: angiotensin II, ET-1: endothelin-1, GPCR: G protein-coupled receptor, IGF-1: insulin-like growth

    factor 1, MAPK: mitogen-activated protein kinase, NE: norepinephrine, PI3K(p110): phosphoinositide 3-kinase p110, RTK: receptor tyrosine kinase.

    198 B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    9/37

    the subclasses IA and IB. Class IA PI3Ks consist of a p110 catalytic

    subunit (, or ) and a p85 or p55 regulatory subunit. The only Class

    IPI3K is p110, which is regulated by p101 (Vanhaesebroeck et al.,

    1997). Of the Class I PI3Ks, p110 and p110 are abundantly

    expressed in the heart. p110is also expressed in the heart but at a

    lower level (Crackower et al., 2002). The p110 isoform of PI3K is

    exclusively expressed in leukocytes (Vanhaesebroeck & Watereld,

    1999).

    PI3Ks were

    rst shown to regulate organ size in Drosophila. Over-expression of the Drosophila PI3K homolog, Dp110, resulted in

    formation of larger wings and eyes (Leevers et al., 1996). In contrast,

    expression of a catalytically inactive Dp110 caused the opposite

    phenotype, i.e. smaller wings and eyes (Leevers et al., 1996). Null

    homozygous mice for p110 were embryonically lethal due to

    proliferation defects in the embryo (Bi et al., 1999). Evidence that

    PI3K activity regulates heart size was obtained from studies that

    expressed a cardiac-specic constitutively active (ca) form of PI3K

    (p110) or dominant negative (dn) form of PI3K(p110) in

    transgenic mice utilizing the -MHC promoter (Shioi et al., 2000;

    McMullen et al., 2003). PI3K activity was increased by 6.5-fold in

    hearts of caPI3K transgenic mice, and hearts were 20% larger than

    non-transgenicmice. Theincreasein heart size of thecaPI3K mice was

    proportional, resembling physiological hypertrophy. In contrast, PI3K

    activity was 77% lower in hearts of dnPI3K transgenic mice, which

    resulted in a 17% decrease in heart size compared to non-transgenic

    mice (Shioi et al., 2000). Changes in heart size were due to changes in

    myocyte size ratherthan number(Shioi et al., 2000). Cardiac function,

    structure and life span were normal in caPI3K and dnPI3K transgenic

    mice under basal conditions (Shioi et al., 2000; McMullenet al., 2003).

    These studies demonstrated that PI3K(p110) is critical for physio-

    logical postnatal growth of the heart.

    It was later demonstrated that PI3K(p110) is also critical for

    physiological exercise-induced growth of the heart but not patholog-

    ical hypertrophy. Adult dnPI3K transgenic mice were subjected to a

    physiological stimulus (exercise; chronic swim training) and a

    pathological stimulus (pressure overload; ascending aortic banding).

    dnPI3K mice showed signicant hypertrophy in response to pressure

    overload, but an attenuated hypertrophic response to swim training,compared with non-transgenic mice (McMullen et al., 2003; McMul-

    len et al., 2007). These studies were also conrmed using a muscle-

    specic knockout approach of the p85/p55/p50 and p85

    (global) regulatory subunits in mice (Luo et al., 2005), as well as

    cardiac-specic ablation ofp110(Lu etal.,2009). These mice showed

    a decrease in heart weight to body weight ratio of approximately 20%

    and 16%, respectively (Luo et al., 2005; Lu et al., 2009), similar to that

    reported in dnPI3K mice (Shioi et al., 2000; McMullen et al., 2003).

    The small heart phenotype of p85/p85 knockout mice was

    accompanied by a reduction in mean myocyte cell area, and mice

    exhibited an attenuated hypertrophic response to exercise training

    (Luo et al., 2005). Together, these studies support the critical role of

    class IAPI3Ks in the regulation of physiological cardiac hypertrophy.

    Finally, to assess whether PI3K(p110) was the critical mediatorresponsible for physiological growth in the IGF1R transgenic mice, we

    crossed transgenic mice over-expressing IGF1R with dnPI3K trans-

    genic mice and examined heart size. Hearts of double transgenic mice

    (i.e. expressing both the IGF1R and dnPI3K transgenes) were not

    signicantly different in size to that in dnPI3K mice alone (McMullen

    et al., 2004b). This result demonstrated that the physiological heart

    growth in the IGF1R transgenic mice was dependent on PI3K(p110)

    signaling (McMullen et al., 2004b).

    3.1.1.3. Akt. Akt, a serine/threonine kinase (also known as protein

    kinase B), is a well characterized target of PI3K. The Akt family is

    involved in a number of cellular processes, including cell survival, cell

    cycle, metabolism, and protein synthesis. There are three isoforms of

    Akt (Akt1, Akt2 and Akt3), each is encoded by distinct genetic loci, in

    which the genes code for enzymes that are members of the serine/

    threonine-specic protein kinase family (Matsui & Rosenzweig,

    2005). The Akt homologue Dakt regulates cell and organ growth in

    Drosophila, in the same manner as PI3K; activation of Dakt increased

    cell size whereas loss-of-function caused a reduction in cell size, but

    had no impact on cell proliferation (Verdu et al., 1999; Rintelen et al.,

    2001). In mammals, Akt1 null mice had a 20% reduction in body

    weight (Cho et al., 2001). Although all three isoforms are broadly

    expressed, only Akt1 and Akt2 are highly expressed in the heart(Matsui & Rosenzweig 2005).

    Initial characterization of the cardiac phenotypes of Akt transgenic

    mice led to confounding results. Phenotypes ranged from absence of

    hypertrophy associated with protection from ischemiareperfusion

    injury to substantial hypertrophy associated with a pathological

    phenotype and premature death (Condorelli et al., 2002; Matsui et al.,

    2002; Shioi et al., 2002; Shiraishi et al., 2004; Shiojima et al., 2005). The

    varying phenotypes have been attributed to different degrees of Akt

    activation, angiogenesis,and subcellular localization. Of note, Aktcan be

    activated by both receptor tyrosine kinases (e.g. IGF1R) and GPCR

    (Fig. 5), and appears to be differently regulated depending on the

    initiating stimulus. Myostatin, an inhibitor of cardiac growth, reduced

    GPCR-induced Akt phosphorylation but not receptor tyrosine kinase

    (IGF1R)-induced phosphorylation in neonatal cardiac myocytes (Mor-

    issette et al., 2006). The biological signicance of this differential

    activation is currently unclear.

    More recent studies in Akt knockout mice suggest Akt1 is required

    for physiological rather than pathological heart growth. Akt1

    knockout mice (normal cardiac phenotype under basal conditions)

    showed a blunted hypertrophic response to swim training but not to

    pressure overload (DeBosch et al., 2006b). These ndings are

    reminiscent of those in mice with reduced PI3K activity ( McMullen

    et al., 2003; Luo et al., 2005). It is now generally accepted that Akt1

    mediates cardiac cell growth whereas Akt2 is important for cardiac

    metabolism (DeBosch et al., 2006a,b).

    Glycogen synthase kinase 3 (GSK3), a cellular substrate of Akt, is

    an important regulatory kinase with a number of cellular targets,

    including cytoskeletal proteins and transcription factors. Both GSK3

    isoforms (GSK3 and GSK3) are expressed in the heart (Ferkey &Kimelman, 2000; Harwood, 2001; Hardt & Sadoshima, 2002). Initial

    reports demonstrated that GSK3negativelyregulated heart growth

    and that inhibition of GSK3 by hypertrophic stimuli was an

    important mechanism for stimulating growth (Haq et al., 2000;

    Morisco et al., 2000; Morisco et al., 2001; Antos et al., 2002; Badorff et

    al., 2002). More recent studies have shown that GSK3 inhibits

    postnatal cardiac growth and reduces pressure overload-induced

    hypertrophy (Zhai et al., 2007). However, transgene expression was

    associated with increased brosis and apoptosis both under basal

    conditions and during pressure overload. Furthermore, the reduced

    hypertrophic phenotype in response to pressure overload was

    associated with severe cardiac dysfunction and heart failure (Zhai et

    al., 2007). Interestingly, the GSK3 and GSK3 isoforms appear to

    have distinct roles in a setting of pressure overload. Phosphorylationof GSK3 was essential for the development of pathological

    hypertrophy whereas phosphorylation of GSK3 played a compen-

    satory role (Matsuda et al., 2008). Thus, selective modulation of the

    phosphorylation status of the two isoforms may be required to

    maximize the therapeutic potential of modulating this kinase.

    3.1.2. Gp130/JAK/STAT pathway

    Leukemia inhibitory factor (LIF), CT-1, and other members of the

    interleukin-6 cytokine family activate the gp130 receptor associated

    with the LIF receptor (Fig 5). Once activated, this cytokine receptor

    interactswith janus kinase1 (JAK1), leading to phosphorylationof the

    signal transducer and activator of transcription (STAT) class of

    transcription factors (Kodama et al., 1997; Pellegrini & Dusanter-

    Fourt 1997; Aoki & Izumo, 2001; Molkentin & Dorn 2001 ). Cardiac-

    199B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    10/37

    specic transgenic mice over-expressing STAT3 displayed cardiac

    hypertrophy that was protective against doxorubicin-induced car-

    diomyopathy (Kunisada et al., 2000). In contrast, mice with

    ventricular deletion of gp130 had normal cardiac structure and

    function under basal conditions, but displayed a rapid onset of dilated

    cardiomyopathy in response to pressure overload (Hirota et al., 1999).

    However, expression of a dominant negative mutant of gp130 (to

    decrease activation of this pathway) appeared to protect transgenic

    mice against pressure overload-induced hypertrophy (Uozumi et al.,2001). Despite this discrepancy, the majority of data in the literature

    suggests that thegp130/JAK/STAT pathway hasa protective role in the

    heart (see Fischer & Hilker-Kleiner, 2007; Boengler et al., 2008;

    Fischer & Hilker-Kleiner, 2008).

    3.1.3. Thyroid hormone receptor signaling

    Thyroid hormone is a classic hormonal mediator of normal

    postnatal heart growth. The thyroid gland secretes two biologically

    active hormones: thyroxine (T4, prohormone) and triiodothyronine

    (T3). T4 and T3 diffuse across the plasma membrane and T4 is

    converted to T3(Danzi & Klein 2002; Dillmann, 2002). Postnatal heart

    growth was reduced in a setting of depressed thyroid gland activity,

    whereas administration of excess thyroid hormone to animals led to

    an increase in heart weight (Bedotto et al., 1989; Hudlicka & Brown

    1996). Furthermore, administration of T4 in humans was associated

    with increased heart mass but no fall in systolic function ( Ching et al.,

    1996), and patients with chronic hyperthyroidism have increased

    cardiac contractility which is often associated with cardiac hypertro-

    phy (Forfar et al., 1982; Feldman et al., 1986). Thus, it has been

    suggested that thyroid hormone induces physiological heart growth.

    The biological effects of thyroid hormone have largely been

    attributed to nuclear transcriptional mechanisms. T3passes through

    the nuclear membrane to bind to nuclear thyroid hormone receptors

    (TRs), which act as transcription factors to directly repress or activate

    cardiac genes (Lazar & Chin 1990; Lazar 1993; Mangelsdorf et al.,

    1995; Danzi & Klein,2002; Dillmann, 2002; Harvey & Williams, 2002).

    Thyroid hormone has been shown to regulate - and-MHC, cardiac

    troponin, SERCA2a, and voltage gated potassium channels (Izumo et

    al., 1986; Rohrer & Dillmann, 1988; Nishiyama et al., 1998; Danzi &Klein, 2002).In mammals TRs are encoded bytwo genes:TR and TR

    (Lazar, 1993; Harvey & Williams, 2002). Thyroid hormone treatment

    increased heart mass by approximately 64% in wildtype mice, 44% in

    TR knockout micebut only6% in TR knockout mice, suggesting that

    TR plays a predominant role in regulating heart growth (Weiss et al.,

    2002).

    More recently, cytosolic and membrane-initiated effects of thyroid

    hormones have been reported (Bassett et al., 2003; Farach-Carson &

    Davis,2003). Kenessey andOjamaademonstrated a directinteraction of

    cytosol-localized TR1 with the p85 regulatory subunit of PI3K in

    neonatal rat ventricular myocytes (Kenessey & Ojamaa, 2006). This

    interaction was shown to be critical for T3-induced protein synthesis.

    The authors concluded that rapid T3 mediated activation of PI3K by

    TR1 may underlie the mechanisms via which thyroid hormoneinduces physiological heart growth (Kenessey & Ojamaa, 2006) (Fig 5).

    3.1.4. Heat shock transcription factor 1

    Sakamoto and colleagues identied HSF1 as a promising critical

    mediator of physiologicalcardiac hypertrophy from a genetic proling

    study that compared gene expression in hearts from rats subjected to

    pressure overload with exercise trained rats (voluntary running

    wheel). Gene expression of HSF1, which regulates heat shock proteins

    (Hsp) including Hsp70 and Hsp27, was upregulated in hearts from

    exercise trained rats but not in hearts subjected to pressure overload

    (Sakamoto et al., 2006). To examine whether HSF1 had a protective

    role in exercise-induced cardiac hypertrophy, HSF1-decient hetero-

    zygote mice (HSF1+/) were subjected to voluntary wheel running

    for 4 weeks. Interestingly, exercise-induced hypertrophy was not

    blunted in HSF1+/ but cardiac function was signicantly reduced

    (Sakamoto et al., 2006).

    3.2. Signaling proteins/pathways

    implicated in mediating pathological hypertrophy

    3.2.1. G proteins

    G proteins canbe divided into two main subgroups: heterotrimeric

    G proteins and small-molecular-weight monomeric G proteins (smallG proteins).

    3.2.1.1. Heterotrimeric G proteins.Heterotrimeric G proteins consist of

    three subunits (,and) and couple to GPCR. Binding of an agonist

    to the GPCR leads to dissociation of the G and G subunits,

    followed by activation of downstream signaling pathways (Gutkind,

    1998a,b; Rockman et al., 2002). Isoforms of the heterotrimeric G

    proteins are largely determined by the isoform of the subunits,

    which fall into four subfamilies: Gs, Gi, G12, and Gq (e.g. Gq, G11)

    (Simon et al., 1991; Neer, 1995).

    In response to a pathological stimulus (e.g. pressure overload),

    hormones/vasoactive factors such as Ang II, ET-1 and noradrenaline

    (norepinephrine, NE) are released and induce cardiac growth

    (Schunkert et al., 1990; Arai et al., 1995; Yamazaki et al., 1999;

    Rapacciuolo et al., 2001; Yayama et al., 2004). These ligands bind to

    GPCR: Ang II receptortype1 (AT1 receptor), endothelin receptors (ETAand ETB) and1-adrenergic receptors (ARs), respectively. This causes

    activation of Gq/11 and downstream signaling proteins, including

    phospholipase C (PLC), MAPKs, PKC and protein kinase A (PKA).

    Transgenic mouse studies have highlighted the critical role of Gq/11in mediating pathological cardiac hypertrophy (Fig. 6). Cardiac-

    specic transgenic mice over-expressing Gq developed cardiac

    hypertrophy that was associated with cardiac dysfunction and

    premature death (D'Angelo et al., 1997; Mende et al., 1998). In

    contrast, mice lacking G proteins (Gq/11) in cardiac myocytes and

    cardiac-specic transgenic mice expressing a peptide specic for

    inhibiting Gq-coupled receptor signaling displayed no hypertrophy or

    a signicantly blunted response to pressure overload (Akhter et al.,

    1998; Wettschureck et al., 2001). Taken together, these studiessuggest that the Gq/11 pathway is important for the induction of

    pathological hypertrophy.

    3.2.1.1.1. Angiotensin II receptors.Ang II is the principal vasoactive

    substance of the reninangiotensin system with a variety of

    pathophysiological actions in the cardiovascular system via systemic

    and local effects including vasoconstriction, aldosterone release, and

    cell growth (Zimmerman & Dunham, 1997; de Gasparo et al., 2000).

    Two pharmacologically distinct Ang II receptors have been cloned

    (AT1 and AT2). Rodents have two AT1 receptor isoforms (AT1A and

    AT1B) (de Gasparo et al., 1995; Lorell 1999). The heart contains a local

    reninangiotensin system which is activated in response to hemody-

    namic stress (e.g. pressure overload) (Yamazaki & Yazaki, 1997;

    Lijnen & Petrov, 1999). It is well established that blocking Ang II

    formation with angiotensin converting enzyme inhibitors (ACEI)attenuates pressure overload-induced hypertrophy in animal models

    and humans (Sadoshima et al., 1996; Zhu et al., 1997; Lijnen and

    Petrov 1999; Yamazaki et al., 1999; Devereux 2000; Modesti et al.,

    2000).

    Hypertrophy dueto activation of theAT1 receptor in rodent cardiac

    neonatal myocytes or mouse models has been associated with

    activation of MAPKs, increased intracellular calcium, PKC, and

    transactivation of the epidermal growth factor receptor (EGFR)

    (Sadoshima & Izumo, 1993; Miyata & Haneda, 1994; Sadoshima et

    al., 1995; Kagiyama et al., 2002; Thomaset al., 2002; Chan et al., 2006).

    Under basal conditions, single global Ang II receptor knockout mice

    (AT1A, AT1B, AT2) have been reported to have no cardiac phenotype or

    a small decrease in heart mass (Hein et al., 1995; Ichiki et al., 1995;

    Hamawaki et al., 1998; Harada et al., 1998; Oliverio et al., 1998;

    200 B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    11/37

    Gembardt et al., 2008). Interestingly, double knockouts lacking AT1Aand AT1Bhad smaller hearts than AT1A alone, whereas the reduction

    was less pronounced in triple knockout mice lacking AT1A, AT1Band

    AT2 (Gembardt et al., 2008). Cardiac-specic AT1 transgenic mice

    developed pathological hypertrophy and died prematurely of heart

    failure (Paradis et al., 2000). Unexpectedly, pressure overload-

    induced hypertrophy was not blunted in AT1A knockout mice

    (Hamawaki et al., 1998; Harada et al., 1998), possibly suggesting

    that AT1Bis able to compensate for loss of AT1A. Mice with a gain offunction mutated AT1Areceptor did not display cardiac hypertrophy

    but developed progressive cardiac brosis and diastolic function

    abnormalities (Billet et al., 2007). Though, this study is confounded by

    a modest increase in blood pressure in the mutant mice. Cardiac-

    specic AT2transgenic mice had no cardiac phenotype (Masaki et al.,

    1998), but pressure overload-induced hypertrophy was inhibited in

    AT2knockout mice (Senbonmatsu et al., 2000). While it is reasonably

    clear that Ang II is critical for mediating cardiac hypertrophy, the

    precise role of Ang II receptor subtypes requires further examination

    (see Billet et al., 2008)). In addition, recent studies have demonstrated

    the cardiovascular effects of a number of breakdown products of Ang

    II including Ang 17, Ang III and Ang IV; as well as new mechanisms

    concerning the functional regulation of Ang II receptors such as

    receptor dimerization, ligand-independent activation, receptor-inter-

    acting proteins, and the existence of an agonistic antibody against the

    AT1receptor (Jones et al., 2008; Mogi et al., 2009). The role of these

    peptides and new regulation of Ang II receptors in relation to cardiac

    hypertrophy also requires further investigation.

    3.2.1.1.2. Endothelin-1 receptors. ET-1 is the predominant endothe-

    lin in the heart and is a potent hypertrophic stimulus in neonatal

    cardiac myocytes (Shubeita et al., 1990). ET-1binds to two GPCRs: ETAand ETB. ETA receptors account for 90% of endothelin receptors on

    cardiac myocytes (Kedzierski & Yanagisawa, 2001). Ang II increased

    ET-1 levels in rat cultured cardiac myocytes, and an ETA antagonist

    (BQ123) reduced Ang II-induced myocyte hypertrophy (Ito et al.,

    1993). ETA receptor antagonists (alone or in combination with ETBreceptor antagonists) have been shown to attenuate pathological

    hypertrophic responses in animal models (seeBrunner et al., 2006).

    Though, cardiac myocyte specic ETA knockout mice displayed anormal hypertrophic response to Ang II (Kedzierski et al., 2003).

    A number of clinical trials have assessed the potential of ET-1

    antagonists to treat heart failure patients; results have largely been

    disappointing, and/or remain unpublished (reviewed by (Seed et al.,

    2001; Rich & McLaughlin, 2003; Kelland & Webb, 2006; Kirkby et al.,

    2008)). In the majority of clinical trials using ET-1 antagonists [either

    non-selectivefor ETA orETB (e.g. bosentan,tezosentan) or selectivefor

    ETA (e.g. darusentan)] heart failure patients developed adverse side

    effects without improvement in cardiac remodeling or clinical

    symptoms (Louis et al., 2001; Kalra et al., 2002; Luscher et al., 2002;

    Rich & McLaughlin, 2003; Anand et al., 2004; Packer et al., 2005;

    Kaluski et al., 2008).

    3.2.1.1.3. Adrenergic receptors (ARs). Catecholamines activate ARs

    (members of the GPCR superfamily) (Scheuer 1999; Lomasney &Allen, 2001; Rockman et al., 2002). There are three major AR

    subfamilies: 1-AR, 2-AR, and -AR. At least 6 types of ARs are

    present in the mammalian heart (three 1-ARs: 1A, 1B, 1D and

    three-ARs:1, 2, 3), with 1-ARs predominating, accounting for

    approximately 80% of the-ARs in the healthy heart (Xiang & Kobilka,

    2003; Barki-Harrington et al., 2004; Salazar et al., 2007; Woodcock et

    al., 2008). ARs are coupled to Gq, Gs, a n d Gi, leading to

    modulation of adenylate cyclase, PLC and ion channels (Rockman et

    al., 2002). Specically,1A, 1B, and 1D-ARs activate Gqsignaling,

    1-ARs couple to Gs, and 2-ARs coupleto Gs andGi (Exton, 1985;

    Garcia-Sainz et al., 1999; Rockman et al., 2002). Species-dependent

    differences exist between the functions of1- and2-AR subtypes in

    the heart which may be attributed to differential coupling to Gsand

    Gi(seeKaumann et al., 1999). In the human heart,2-ARs appear to

    largely couple to the Gs/cAMP pathway (Kaumann et al., 1999). The

    role of3-ARs in the heart remains unclear (seeBarki-Harrington et

    al., 2004; Kaumann & Molenaar, 2008).

    Heart failure patients have elevated circulating catecholamines

    and increased adrenergic drive, which initially increases contractility

    and may be benecial. However, prolonged adrenergic drive is

    detrimental and associated with desensitization and downregulation

    of -ARs (Bristow, 2000). This is consistent with ndings from

    cardiac-speci

    c transgenic mice over-expressing 1-ARs. Before15 weeks of age, transgenic mice displayed increased cardiac

    contractility compared to controls. However, cardiac function pro-

    gressively fell in 1-AR transgenic mice after 16 weeks and the mice

    rapidly developed cardiac dysfunction and heart failure (Engelhardt

    et al., 1999). Progressive deterioration in cardiac function with

    chronic transgenic expression of 1-AR was later conrmed by

    another report (Bisognano et al., 2000).

    Cardiac-specic transgenic mice expressing a dominant negative

    -AR receptor kinase 1 (-ARK1/GRK2; dominant negative mutant

    restores-AR signaling;-ARK1 phosphorylates-ARsleadingto their

    desensitization) were protected against pathological hypertrophy and

    heart failure (Koch et al., 1995), and targeted deletion of GRK2 in

    cardiac myocytes of mice preventedand rescued heart failure induced

    by myocardial infarction(Raakeet al., 2008).In this respect, it hasbeen

    difcult to explain why -AR agonists are poorly tolerated in heart

    failure patients but -blockers have a protective role (Bristow, 2003;

    Molenaar& Parsonage, 2005, discussedin further detailin Section 8.1).

    Though it has been suggested that at the molecular level, inhibition of

    -ARK1/GRK2 shares a number of properties with -blockade as

    opposed to-AR agonism (Rockman et al., 2002). For instance,-AR

    agonists promote desensitization and receptor downregulation due to

    constant activation of the -AR system. In contrast, inhibition of-

    ARK1 may allow-ARs to return to a more normal state of signaling

    because desensitization will be inhibited (Rockman et al., 2002).

    While it is generally accepted that chronic stimulation of the -AR

    system has an adverse effect on the heart that contributes to the

    pathogenesis of heart failure, preservation of normal -AR-G protein

    coupling is critical during times of need, such as periods of stress and

    during exercise (Christensen & Galbo, 1983; Lefkowitz et al., 2000;Rockman et al., 2002). In this instance, acute versus chronic activation

    of -ARs may explain differences in phenotype observed with

    pathological and physiological hypertrophy. Plasma resting levels of

    catecholamines were signicantly higher in a mouse model of

    pathological hypertrophy induced by chronic pressure overload in

    comparison to a mouse model of physiological hypertrophy induced

    by swim training (Perrino et al., 2006).

    Therole of AR subtypes in mediatingcardiac hypertrophy based on

    genetically modied mouse models has been described in detail (see

    Du, 2008). Global knockout mice decient of1and2-ARs displayed

    an attenuated hypertrophic response to pressure overload, with

    reducedbrosis(Kiriazis et al., 2008). Based on various transgenic and

    knockout models, 1B-AR and 2-AR appear to contribute to

    maladaptation and the onset of heart failure in a setting of pressureoverload, whereas activation of1A-AR may be benecial (Du, 2008).

    Of note, in rat neonatal cardiac myocytes, it was shown that 2-ARs

    that couple with G i proteins mediate cardiac protection due to

    activation of the PI3K-Akt pathway (Chesley et al., 2000). Further-

    more,1-ARs were critical for normal postnatal heart growth in male

    but not female mice (O'Connell et al., 2003), and protected the heart

    against pressure overload-induced maladaptive hypertrophy (O'Con-

    nell et al., 2006).

    3.2.1.2. Small G proteins (also called GTPases). The family of small G

    proteins can be divided into 5 subfamilies (Ras, Rho, ADP ribosylation

    factors, Rab, and Ran). Small G proteins act as molecular switches,

    which link receptors to downstream signaling cascades. Ras and Rho

    can be activated in myocytes in response to Ang II, ET-1,

    201B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    12/37

    phenylephrine (PE) and mechanical stress and have been implicated

    in the development of cardiac hypertrophy (Ramirez et al., 1997a;

    Aoki et al., 1998; Aikawa et al., 1999; Chiloeches et al., 1999; Clerk and

    Sugden, 2000; Clerk et al., 2001). Cardiac-specic transgenic mice

    expressing a constitutively active form of Ras or over-expressing

    Rab1a developed pathological cardiac hypertrophy (Hunter et al.,

    1995; Wu et al., 2001). Furthermore, using a Rho-kinase inhibitor

    (Fasudil), it was shown that Rho-kinase was critical for pressure

    overload-induced pathological hypertrophy in rats but not swim-ming-induced physiological cardiac hypertrophy (Balakumar & Singh,

    2006).

    3.2.2. PI3K(p110) signaling

    In contrast to the p110 isoform of PI3K (coupled to RTKs e.g.

    IGF1R), PI3K(p110) is coupled to GPCRs (specic GPCR not all

    denitively dened but thought to include Gs/Gi/Gq e.g. adren-

    ergic receptors, Ang II receptors and endothelin receptors) and

    appears to have a detrimental effect in the heart (Oudit et al., 2004).

    PI3K(p110) does not affect heart size under basal conditions but is a

    negative regulator of cardiac contractility, as PI3K(p110) knockout

    mice displayed enhanced contractile function (Crackower et al.,

    2002). PI3K(p110) may have an impact on heart growth in settings

    of pathological stress (Naga Prasad et al., 2000; Oudit et al., 2003),

    although therole of PI3K(p110) in the diseased heart is complex and

    appears to differ depending on the nature of the pathological

    stimulus. PI3K(p110) knockout mice were protected from heart

    failure induced by chronic activation of -ARs, displaying less

    hypertrophy and brosis, and better cardiac function than controls

    (Oudit et al., 2003). However, PI3K(p110) knockout mice displayed

    an accelerated progression to dilated cardiomyopathy in response to

    pressure overload (Patrucco et al., 2004; Oudit & Kassiri, 2007). PI3K

    (p110) might contribute to cardiac dysfunction via its effects on -

    AR internalization and regulation of phosphodiesterases (Fig. 5)

    (Oudit & Kassiri, 2007; Pretorius et al., 2009b). Downregulation and

    desensitization of-ARs is detrimental for heart function (Bristow et

    al., 1982; Perrino et al., 2007), and is dependent on the binding of

    p110 to -ARK1 (Naga Prasad et al., 2001). Expression of a

    catalytically inactive p110 mutant or disruption of the interactionbetween-ARK1 and p110restored -AR signaling and contractile

    function in transgenic mice subjected to chronic -AR stimulation

    (Nienaber et al., 2003; Perrino et al., 2005; Perrino et al., 2007 ). PI3K

    (p110) may also reduce cardiomyocyte contractility by regulating

    the activity of phosphodiesterases (PDEs) (Patrucco et al., 2004;

    Kerfant et al., 2007). PDEs hydrolysecAMP, a secondmessenger which

    plays a critical role in mediating Ca2+ release from the sarcoplasmic

    reticulum to induce contraction. PDE inhibitors improve contractile

    function by increasing intracellular cAMP levels, however the safety of

    PDE inhibitors as therapeutic agents in patients with heart failure is

    still being investigated (seeOsadchii, 2007for review).

    3.2.3. Mitogen activated protein kinase (MAPK) pathways

    MAPKs are divided into 3 subfamilies based on the terminal kinasein the pathway: the extracellular signal-regulated kinases (ERKs), the

    c-Jun amino-terminal kinase (JNKs), and the p38-MAPKs (Clerk &

    Sugden, 1999; Widmann et al., 1999; Pearson et al., 2001). All three

    types of MAPKs areactivated in cultured cardiac myocytes in response

    to GPCR agonists (couple to Gq: AT1receptors, endothelin receptors

    and1-ARs) and mechanical stress, as well as in pressure overloaded

    hearts and failing human hearts, but the exact role of MAPKs has

    remained unclear (Yamazaki et al., 1993; Sadoshima et al., 1995;

    Komuro et al., 1996; Sugden & Clerk, 1998; Cook et al., 1999; Esposito

    et al., 2001; Pearson et al., 2001; Takeishi et al., 2001; Purcell et al.,

    2007).

    3.2.3.1. ERK1/2.ERKs are protein kinases that phosphorylate a range of

    cytosolic and nuclear substrates (see Chen et al., 2001bfor review).

    ERK1/2 are ubiquitously expressed (Boulton et al., 1991) and

    activation has been reported in numerous settings of cardiac

    hypertrophy and failure (seeMuslin, 2008) however it is still unclear

    whether ERK1/2 is a critical mediator of hypertrophic responses.

    ERK1/2 was activated in response to agonists that induce pathological

    heart growth, such as Ang II, ET-1 and NE, but not in response to the

    physiological hypertrophic agonist IGF1 (Clerk et al., 2006). In isolated

    cardiac myocytes, activation of ERK1/2 was essential for protein

    synthesis (a requirement for cell growth) following stimulation withhypertrophic agonists that signal via Gq protein coupled receptors

    (Wang & Proud, 2002). Consistent with this nding, expression of a

    dominant negative mutant of Raf-1 (a MAPK kinase kinase down-

    stream of Gq;Fig. 5) blunted cardiac hypertrophy in mice subjected

    to pressure overload, implicating ERK1/2 in the development of

    pathological cardiac hypertrophy (Harris et al., 2004). However,

    transgenic mice expressing cardiac-specic constitutively active

    MAPK kinase 1 (MEK1; a MAPK kinase immediately upstream of

    ERK1/2; does not activate JNK or p38-MAPK) developed a physiolog-

    ical ratherthan a pathological phenotype, which was characterized by

    concentric cardiac hypertrophy, enhanced systolic cardiac function

    and no interstitial brosis (Bueno et al., 2000). Furthermore, loss of

    ERK1 (global knockout mice) had no effect on heart size in mice

    subjected to pressure overload or swim training, indicating that ERK1

    is not a critical mediator of pathological or physiological hypertrophy,

    or that the remaining ERK2 activity was sufcient to drive the

    hypertrophic response (Purcell et al., 2007). Reduced expression of

    ERK2 (ERK2+/ mice; ERK2/ mice are embryonically lethal) alone

    or in mice decient for ERK1 (ERK1/ERK2+/ mice) also failed to

    block hypertrophy induced by pressure overload or swim training

    (Purcell et al., 2007). Together, these studies suggest that activation of

    ERK1/2 is sufcient, but not critical, for inducing cardiac hypertrophy,

    although the results of the latter study were confounded by the fact

    that the ERK1/ and ERK2+/ mice were not cardiac-specic.

    Fig. 7. ERK1/2 appears to contribute to hypertrophic responses via two distinct

    mechanisms. Stimulation of GPCRs leads to dissociation of Gq proteins. A) The G

    subunit of Gq activates traditional MAP kinase signaling cascades, resulting in

    phosphorylation and activation of ERK1/2 by MEK1/2. This in turn leads to protein

    synthesis and cell growth. B) Association of Gsubunits with the Raf1/MEK/ERK1/2

    complex is necessary for autophosphorylation of ERK1/2 at residue Thr188 and

    localization of ERK1/2 in the nucleus. This leads to phosphorylation of nuclear targets

    (such as Elk1, MSK1 and c-Myc) and transcription of hypertrophic genes.

    202 B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    13/37

    Induction of ERK1/2 kinase activity requires phosphorylation of

    the threonine and tyrosine residues within the TEY motif of the

    activation loop by MEK1/2 (Anderson et al., 1990; Payne et al., 1991;

    Robbins et al., 1993). A recent study reported that autophosphoryla-

    tion of ERK2 at residue Thr188 promoted nuclear localization and

    subsequent phosphorylation of hypertrophic factors, including Elk1,

    MSK1 and c-Myc (Lorenz et al., 2009). This occurred following

    association of the Raf/MEK/ERK complex with the subunits of Gq

    proteins (Lorenz et al., 2009). Thr188 phosphorylation appears to be akey event in the development of ERK1/2-mediated cardiac hypertro-

    phy, as transgenic mice with suppressed ERK2 Thr188 phosphoryla-

    tion were resistant to cardiac hypertrophy induced by pressure

    overload, while mice with enhanced ERK2 Thr188 phosphorylation

    displayed more pronounced hypertrophy compared to wildtype mice

    (Lorenz et al., 2009). Phosphorylation of Thr188 was also evident in

    biopsies from human failing hearts, suggesting that ERK2 autopho-

    sphorylation of Thr188 is clinically relevant (Lorenzet al., 2009). Thus,

    ERK1/2 appears to contribute to cardiac hypertrophic responses via

    two distinct mechanisms (Fig. 7). Activation of the traditional MAPK

    signaling cascade (Raf/MEK/ERK) following binding of Gq proteins to

    GPCRs results in phosphorylation of ERK1/2 within the TEY motif and

    induction of ERK1/2 kinase activity. Subsequent phosphorylation of

    substrates (such as p90 ribosomalS6 kinase (Lorenzet al., 2009))may

    contribute to cell hypertrophy by increasing protein synthesis. This

    mechanism may be responsible for driving the physiological hyper-

    trophy observed in the MEK1 transgenic mice. Secondly, interaction of

    the Raf/MEK/ERK complex with Gproteins causes autophosphor-

    ylation of ERK2 at Thr188. This results in nuclear localization, allowing

    ERK1/2 to phosphorylate nuclear targets, which in turn promotes

    transcription of hypertrophic genes. This event may be critical for

    inducing the maladaptive phenotype associated with pathological

    hypertrophic responses (Fig 7.)

    3.2.3.2. ERK5.ERK5 (also known as big MAP kinase 1, BMK1) may play

    a role in mediating pathological eccentric cardiac hypertrophy, as

    transgenic mice expressing constitutively active MEK5 developed

    eccentric cardiac hypertrophy which progressed to dilated cardiomy-opathy and death (Nicol et al., 2001).

    3.2.3.3. JNKs. The JNK family consists of at least ten isoforms, derived

    from three genes:JNK1,JNK2and JNK3(Waetzig & Herdegen, 2005).

    JNK1and JNK2are ubiquitously expressed, whereas JNK3has a more

    restricted expression prole in the heart, brain and testis (Waetzig &

    Herdegen, 2005). JNK was activated in hearts from heart failure

    patients (Cook et al., 1999) and in the remote myocardium of

    infarcted rat hearts (Li et al., 1998). A number of in vitro studies have

    suggested that JNKs may be important regulators of pathological

    hypertrophy (Bogoyevitch et al., 1996; Ramirez et al., 1997a;Choukroun et al., 1998; Wang et al., 1998b; Choukroun et al., 1999),

    although in vivo studies have been more difcult to interpret

    (described below and summarized inTable 2).

    JNK is phosphorylated and activated by MAPK kinase 4 (MEK4) and

    MEK7 (Fig. 5), and preferentially upregulated by MAPK kinase kinase 1

    (MEKK1) (Fig5). Transgenic mice withcardiac-specic activationof JNK

    (via constitutive activation of MEK7) did not develop cardiac hypertro-

    phy but died prematurely from congestive heart failure (Petrich et al.,

    2003). Several loss of function approaches have also been utilized to

    determinethe role of JNKsin pathologicalcardiac hypertrophy. Pressure

    overload-induced hypertrophy was attenuated in dnMEK4 transgenic

    mice(Choukrounet al.,1999), andMEKK1was essentialfor pathological

    cardiac hypertrophy and dysfunction induced by cardiac-specic

    transgenic expression of Gq (Minamino et al., 2002). These data

    suggest JNKs may be necessary regulators of pathological cardiac

    hypertrophy. However, dnJNK1/2 transgenic mice and JNK1/2 gene-

    targeted micedisplayed an enhanced hypertrophicresponseto pressure

    overload (Liang et al., 2003), suggesting JNKs antagonize cardiac

    growth. Furthermore, cardiac-specic MEK4 knockout mice with

    reduced JNK activity displayed normal cardiac growth and function

    under basal conditions (Liu et al., 2009). However, following aortic-

    banding and chronic -adrenergic stimulation, unlike the dnMEK4

    transgenic mice, cardiac specic MEK4 knockout mice had enhanced

    cardiac growth, increased hypertrophic gene transcription and ventric-

    ular brosis compared to wildtype aortic-banded controls. Following

    swim training, the hypertrophic response was unchanged compared to

    wildtype controls in this model (Liu et al., 2009). Null MEKK1 transgenic

    mice had comparable heart weights to wildtype mice under basal

    conditions.In response to aortic-banding, MEKK1/mice displayedanenhanced hypertrophic response rather than a blunted response after

    Table 2

    JNK mouse models and their phenotype under basal conditions and in response to pressure overload as assessed by heart weight/body weight or left ventricular/body weight ratios.

    Mouse model Basal phenotype Pressure overload Reference

    Versus Ntg/WT Hypertrophic response versus Ntg/WT

    caMEK7

    (cardiac-specic)

    n/a Petrich et al., 2004

    dnMEK4

    (cardiac-specic by adenovirus-mediated gene transfer)

    n/a Choukroun et al., 1999

    dnJNK1/2

    (cardiac-speci

    c)

    Liang et al., 2003

    Jnk1+//Jnk2/

    (targeted deletion, global)

    Liang et al., 2003

    MEK4/

    (cardiac-specic)

    Liu et al., 2009

    MEKK1/

    (targeted deletion, global)

    Sadoshima et al., 2002

    Jnk1/

    (targeted deletion, global)

    Tachibana et al., 2006

    Jnk2/

    (targeted deletion, global)

    Tachibana et al., 2006

    Jnk3/

    (targeted deletion, global)

    Tachibana et al., 2006

    Key:

    = JNK has no effect on heart growth, possibly due to redundancy of isoforms.

    = Indicates JNK antagonizes pathological growth.

    = JNK is essential for pathological growth.

    n/a = not examined/assessed.

    203B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    14/37

    14 days compared to wildtype controls (Sadoshima et al., 2002).

    Furthermore, aortic-banded MEKK1/ mice had a higher mortality

    rate and congestive heart failure compared to wildtype banded mice.

    This study implies that cardiac hypertrophy induced by pressure

    overload occurs in the absence of JNK activation. In further support of

    the latter, mice with selective deletion of the three JNK genes (Jnk1/,

    Jnk2/ and Jnk3/) subjected to aortic-banding developed cardiac

    hypertrophy that was comparable to wildtype mice (Tachibana et al.,

    2006). Thus, it appears individual members of the JNK family are not

    required to induce cardiac growth, or that the JNK isoforms are

    functionally redundant.

    3.2.3.4. p38-MAPK.p38-MAPK is an important mediator of numerousbiological functions including cell growth, cell proliferation, cell cycle

    and cell death, and is considered a critical component of stress

    response pathways (Wilson et al., 1996; Bassi et al., 2008). In the

    heart, p38-MAPK is known to be activated during ischemia, and p38-

    MAPK activity was increased in the myocardium from patients with

    ischemic heart disease (Cook et al., 1999). p38-MAPK has been

    implicated in the regulation of cardiac gene expression, cardiac

    myocyte apoptosis, myocyte hypertrophy, contractility, remodeling

    and metabolism (Liao et al., 2001; Petrich & Wang, 2004; Baines &

    Molkentin, 2005; Wang, 2007).

    The p38-MAPK family consists of four isoforms (, , and ),

    though it appears onlyandare expressed in the heart (Jiang et al.,

    1997; Clark et al., 2007). The p38 isoform is predominately

    expressed in the human and rodent myocardium (Lemke et al.,2001; Braz et al., 2003). The role of p38and p38in the heart has

    been examined extensively using both transgenic and knockout

    mouse models. As with JNK, many of the studies appear to be

    contradictory (summarized in Table 3). Cardiac-specic transgenic

    mice expressing a dominant negative form of p38 were generated

    and analyzed by two independent groups (Braz et al., 2003; Zhang et

    al., 2003a).Zhang et al. (2003a) reported no basal phenotype in dn-

    p38transgenics. In contrast, Braz et al. (2003) found that dn-p38

    transgenics developed concentric hypertrophy associated with a fall

    in cardiac function and elevated fetal gene expression (ANP and BNP),

    with the majority of mice dying prematurely from cardiomyopathy by

    8 months of age. Both groups subjected the dn-p38 transgenics to

    pressure overload and reported a signicant increase in heart size.

    Zhang et al. reported a hypertrophic response (to pressure overload)

    in the dn-p38 transgenics similar to control animals but, interest-

    ingly, the transgenics had signicantly less brosis (Zhang et al.,

    2003a). In contrast, Braz et al. (2003) reported an exaggerated

    hypertrophic response in dn-p38 transgenics compared with

    controls in response to pressure overload or 14 day minipump

    infusions of phenylephrine, Ang II and isoproterenol. Finally,

    cardiac-specic p38 knockout mice displayed normal cardiac

    structure and function under basal conditions (Nishida et al., 2004).

    In response to pressure overload, knockout mice developed a similar

    degree of cardiac hypertrophy to controls, but displayed greater

    cardiac dysfunction, morebrosisand apoptosis(Nishida et al., 2004).

    The authors concluded that p38 plays a critical role in protecting the

    heart in a setting of pressure overload.Cardiac-specic dn-p38transgenic mice had no cardiac hypertro-

    phy under basal conditions but seemed to have reduced systolic

    function (Zhang et al., 2003a). The hypertrophic response to pressure

    overload was not different from that observed in non-transgenic mice,

    though the dn-p38transgenics were reported to display less brosis

    (Zhang et al., 2003a).

    MEK3 and MEK6 are regulators of p38-MAPK (Fig 5). Under basal

    conditions dnMEK3 and dnMEK6 transgenic mice developed patho-

    logical hypertrophy by 2 and 8 months of age, respectively. Both

    models displayed cardiac dysfunction, and brosis was also reported

    in dnMEK3 transgenics at 4 months of age. The majority of dnMEK3

    mice died by 8 months of age due to cardiomyopathy (Braz et al.,

    2003). In response to pressure overload, dnMEK3 and dnMEK6

    transgenic mice displayed an exacerbated cardiac hypertrophicresponse, increased brosis and depressed cardiac function, similar

    to the observations in dn-p38transgenic mice (Braz et al., 2003).

    Activation of p38-MAPK in either MEK3 or MEK6 transgenic hearts

    under baseline conditions led to the increased expression of the fetal

    gene program, substantial induction of interstitial brosis, and loss of

    contractility (Liao et al., 2001). Bothtransgenic mousemodels developed

    heart failure, although this was not associated with hypertrophy of

    cardiac myocytes (Liao et al., 2001). In this setof studies, p38-MAPKdoes

    not appear to promote hypertrophy, but seems to contribute to brosis,

    loss of contractility and the development of dilated cardiomyopathy.

    Explanations for the possible discrepancies between many of these

    models (seeTable 3) may be attributed to the generation of transgenic

    mice on different genetic backgrounds (Braz et al., 2003; Zhang et al.,

    2003a), gender differences (Liu et al., 2006) (discussed in detail in

    Table 3

    p38 mouse models and their phenotype under basal conditions and in response to pressure overload as assessed by heart weight/body weight or left ventricular/body weight ratios.

    Mouse model Basal phenotype Pressure overload Reference

    Versus Ntg/WT Hypertrophic response versus Ntg/WT

    dn-p38

    (cardiac-specic, Black Swiss background)

    Zhang et al., 2003a

    dn-p38

    (cardiac-specic, FVB/N background)

    Braz et al., 2003

    dn-p38

    (cardiac-specic females compared to males)

    Liu et al., 2006

    p38CKO

    (cardiac-specic KO)

    Nishida et al., 2004

    dn-p38

    (cardiac-specic)

    Zhang et al., 2003a

    dnMEK3

    (cardiac-specic)

    Braz et al., 2003

    dnMEK6

    (cardiac-specic)

    Braz et al., 2003

    MEK3

    (cardiac-specic)

    n/a Liao et al., 2001

    MEK6

    (cardiac-specic)

    n/a Liao et al., 2001

    Key:

    = p38 has no effect on heart growth.

    = p38 is essential for pathological hypertrophy.

    n/a = not examined/assessed.

    204 B.C. Bernardo et al. / Pharmacology & Therapeutics 128 (2010) 191227

  • 8/11/2019 1-s2.0-S0163725810000793-main

    15/37

    Section 7), p38-MAPK having an anti-apoptotic function that is

    independent of kinase activity or a biphasic dose response curve to

    p38-MAPK (Muslin, 2008), as well as distinct roles of p38 and that

    maybe differentially regulated by MEK3 andMEK6 (Wang et al.,1998a).

    3.2.3.5. Protein kinases. Extracellular stimuli such as pressure overload

    activate PKC and PKD via GPCRs to trigger hypertrophic responses

    (Dorn & Force, 2005; Harrison et al., 2006)(Fig 5).

    3.2.3.5.1. PKC. PKC is considered a critical signal transducerdownstream of Gq. There are at least 12 isoforms of PKC and at least

    4 (, , , and ) have been implicated in the induction of cardiac

    hypertrophy (see Dorn & Force, 2005). PKC and PKC are

    conventional Ca2+-dependent isoforms whereas PKC and PKC are

    novel Ca2+-independent isoforms (Mackay & Mochly-Rosen, 2001;

    Sabri & Steinberg, 2003). Mice null for PKC,,, or had no obvious

    cardiac phenotype under basal conditions (seeDorn & Force, 2005),

    however it has been suggested that the role of specic PKC isoforms

    may be masked by compensatory signaling by other PKC isoforms

    (Dorn & Force, 2005). Cardiac-specic PKC transgenic mice devel-

    oped cardiac hypertrophy associated with cardiac dysfunction,

    brosis and premature death (Bowman et al., 1997; Wakasaki et al.,

    1997; Chen et al., 2001a), but knockout mice displayed a typical

    hypertrophic response to a GPCR agonist (PE) or aortic-banding

    (Roman et al., 2001). Thus, it appears that PKC is not required for the

    pathological hypertrophic response. Cardiac-specic transgenic mice

    over-expressing PKCor PKCdisplayed mild concentric hypertrophy

    with a physiological phenotype (no evidence ofbrosis, and normal

    cardiac function) (Takeishi et al., 2000; Chen et al., 2001a). However,

    in response to a cardiac insult (ischemia-induced damage), PKChad

    a protective role, whereas activation of PKCexacerbated the damage

    (Chen et al., 2001a). PKCappears to be critical for regulating cardiac

    contractility but not cardiac hypertrophy (Braz et al., 2002; Hahn et

    al., 2003; Braz et al., 2004). Transgenic mice with over-expression of

    PKChad diminished cardiac contractility while PKC/ mice had

    improved cardiac contractility (Braz et al., 2004). Furthermore,

    inhibition of PKC activity in a model of pathological hypertrophy

    (Gq transgenic mice) improved cardiac contractility, whereas

    activation of PKCresulted in a lethal cardiomyopathy (Hahn et al.,2003). There were no apparent effects on heart size under basal

    conditions or in response to cardiac stress (e.g. pressure overload)

    with the PKCmouse models.

    3.2.3.5.2. PKD. Cardiac-specic transgenic mice expressing a

    constitutively active form of PKD1 developed pathological hypertro-

    phy and died prematurely (Harrison et al., 2006). In contrast, mice

    with conditional cardiac-specic deletion of PKD1 had no phenotype

    under basal conditions but displayed a blunted hypertrophic response

    to various pathological models (pressure overload, Ang II-dependent

    hypertrophy and isoproterenol-dependent hypertrophy) which was

    associated with better cardiac function, less brosis, and less fetal

    gene activation compared to control mice (Fielitz et al., 2008).

    3.2.4. Calcium signalingEach heartbeat is associated with entry of calcium into cardiac

    myocytes. Calcium is central to the control of contractile function and

    cardiac growth. Calcium/calmodulin is an important second messen-