Involvement of the nicotinamide adenosine dinucleotide phosphateoxidase isoform Nox2 in cardiac contractile dysfunction occurring inresponse to pressure overload.Grieve, D., Shah, A. M., Byrne, J. A., Siva, A., Layland, J., Johar, S., & Cave, A. C. (2006). Involvement of thenicotinamide adenosine dinucleotide phosphate oxidase isoform Nox2 in cardiac contractile dysfunctionoccurring in response to pressure overload. Journal of the American College of Cardiology, 47 (4)(4), 817-826.https://doi.org/10.1016/j.jacc.2005.09.051

Published in:Journal of the American College of Cardiology

Document Version:Publisher's PDF, also known as Version of record

Queen's University Belfast - Research Portal:Link to publication record in Queen's University Belfast Research Portal

General rightsCopyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or othercopyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associatedwith these rights.

Take down policyThe Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made toensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in theResearch Portal that you believe breaches copyright or violates any law, please contact openaccess@qub.ac.uk.

Download date:29. Jun. 2020

P

IAODDSL

ApcbadLpRoattct

TFCt

2

Journal of the American College of Cardiology Vol. 47, No. 4, 2006© 2006 by the American College of Cardiology Foundation ISSN 0735-1097/06/$32.00P

RECLINICAL STUDIES

nvolvement of the Nicotinamidedenosine Dinucleotide Phosphatexidase Isoform Nox2 in Cardiac Contractileysfunction Occurring in Response to Pressure Overload

avid J. Grieve, PHD, Jonathan A. Byrne, PHD, MRCP, Anjana Siva, PHD, MRCP, Joanne Layland, PHD,ofian Johar, MRCP, Alison C. Cave, PHD, Ajay M. Shah, MD, FAHAondon, England

OBJECTIVES This study sought to examine the role of Nox2 in the contractile dysfunction associated withpressure-overload left ventricular hypertrophy (LVH).

BACKGROUND Reactive oxygen species (ROS) production is implicated in the pathophysiology of LVH. Thenicotinamide adenosine dinucleotide phosphate oxidase isoform, Nox2, is pivotally involvedin angiotensin II-induced hypertrophy but is not essential for development of pressure-overload LVH. Its possible impact on contractile function is unknown.

METHODS The effects of aortic banding or sham surgery on cardiac contractile function and interstitialfibrosis were compared in adult Nox2�/� and matched wild-type (WT) mice.

RESULTS Banding induced similar increases in left ventricular (LV) mass in both groups. BandedNox2�/� mice had better LV function than WT by echocardiography (e.g., fractionalshortening 33.6 � 2.5% vs. 21.4 � 2.2%, p � 0.05). Comprehensive LV pressure-volumeanalyses also showed significant contractile dysfunction in banded WT compared with sham,whereas banded Nox2�/� mice had preserved function (e.g., maximum rate of rise of LV pressure:banded WT, 4,879 � 213; vs. banded Nox2�/�, 5,913 � 259 mm Hg/s; p � 0.05). Similarpreservation of function was observed in isolated cardiomyocytes. The 24-h to 36-h treatmentof banded WT mice with N-acetylcysteine resulted in recovery of contractile function.Cardiac interstitial fibrosis was significantly increased in banded WT but not Nox2�/� mice,together with greater increases in procollagen I and III mRNA expression.

CONCLUSIONS The Nox2 oxidase contributes to the development of cardiac contractile dysfunction andinterstitial fibrosis during pressure overload, although it is not essential for development ofmorphologic hypertrophy per se. These data suggest divergent downstream effects of Nox2 ondifferent components of the overall response to pressure overload. (J Am Coll Cardiol 2006;

ublished by Elsevier Inc. doi:10.1016/j.jacc.2005.09.051

47:817–26) © 2006 by the American College of Cardiology Foundation

ctcitnc

rotoaizvapcg

n increase in oxidative stress is implicated in the patho-hysiology of left ventricular hypertrophy (LVH) (1). In-reased reactive oxygen species (ROS) production occursoth in experimental models of cardiomyocyte hypertrophynd with in vivo pressure-overload LVH, whereas antioxi-ants may exert beneficial effects in experimental in vivoVH and heart failure (2–6). The ROS influence LVHathophysiology through several mechanisms. IncreasedOS production contributes to development of cardiomy-cyte hypertrophy induced by G-protein–coupled receptorgonists, cytokines, and mechanical stretch, at least in parthrough activation of redox-sensitive protein kinases andranscription factors (3–6). Oxidative stress promotes myo-yte apoptosis and necrosis, especially in advanced hyper-rophy and heart failure (1,3,7). The ROS may also alter

From the King’s College London, Cardiovascular Division, Guy’s, King’s, and St.homas’ School of Medicine, London, England. Supported by British Heartoundation (BHF) Program Grant RG/03/008. Dr. Shah holds the BHF Chair ofardiology at King’s College London. Drs. Grieve and Byrne contributed equally to

his work.

rManuscript received January 1, 2005; revised manuscript received September 14,

005, accepted September 19, 2005.

ardiac contractile properties, for example by modulatinghe function of sarcolemmal ion channels/exchangers, sar-oplasmic reticular proteins, myofilaments, and/or enzymesnvolved in energy metabolism (8–10). Finally, ROS con-ribute to endothelial dysfunction through inactivation ofitric oxide, and also promote interstitial fibrosis and extra-ellular matrix remodeling (1,11).

Given these effects of oxidative stress, elucidation of theoles of different ROS sources (e.g., mitochondria, xanthinexidase, uncoupled nitric oxide synthases) is important. Inhis regard, recent studies from several laboratories, includingur own, have addressed the potential involvement of nicotin-mide adenosine dinucleotide phosphate (NADPH) oxidasesn LVH pathophysiology. These superoxide-generating en-ymes are major cardiovascular ROS sources, particularly in theasculature, where they play important roles in hypertension,therosclerosis, angiogenesis, and other conditions (12). Therototypic NADPH oxidase (first characterized in neutrophils)omprises a membrane-bound heterodimer consisting of onep91phox (or Nox2) and one p22phox subunit, and several

egulatory subunits (p47phox, p67phox, p40phox, and Rac) that

actismNfi

hoiheoos

a(rh(sAcpNn((carea

M

EWai(NHmk

F(c

818 Grieve et al. JACC Vol. 47, No. 4, 2006Nox2 and Contractile Dysfunction in LVH February 21, 2006:817–26

ssociate with the heterodimer in the activated enzymeomplex (12,13). The Nox subunit is the catalytic core ofhe oxidase and, recently, several isoforms (Nox1 to 5) weredentified that have specific tissue distributions and mayubserve distinct (patho)physiological functions (13). Theajor isoforms expressed in heart seem to be Nox2 andox4, with the Nox2 oxidase found in cardiomyocytes,broblasts, and endothelial cells (12,14–16).The NADPH oxidases are implicated in cardiomyocyte

ypertrophic signaling in response to angiotensin II (AngII)r alpha-adrenergic agonists (16,17), whereas Rac1 (whichs involved in oxidase activation) induces cardiomyocyteypertrophy (18,19). Increased cardiac NADPH oxidasexpression and activity was shown in experimental pressure-verload LVH, with evidence that ROS derived from thexidase may contribute to contractile dysfunction in thisetting (14,20,21). Similarly, increased NADPH oxidase

igure 1. Effect of pressure overload on cardiac fibrosis. (A) Interstitial co

Abbreviations and AcronymsANF � atrial natriuretic factorAngII � angiotensin IIEDPVR � end-diastolic pressure volume relationESPVR � end-systolic pressure volume relationIVSD � interventricular septal thickness in diastoleLVEDD � left ventricular end-diastolic dimensionLVESD � left ventricular end-systolic dimensionLVH � left ventricular hypertrophyNADPH � nicotinamide adenosine dinucleotide

phosphate, reducedROS � reactive oxygen speciesWT � wild-type

B to D) messenger ribonuclerc acid expression of procollagen I, procollagen III,orresponding sham; †p � 0.05 versus banded Nox2�/�. WT � wild type.

ctivity was also documented in failing human hearts22,23). Recently, we began to specifically investigate theole of the Nox2 oxidase in the development of cardiacypertrophy (15,24). In Nox2-deficient mice (Nox2�/�)25), LVH induced by short-term subpressor AngII infu-ion was substantially inhibited together with abolition ofngII-induced NADPH oxidase activation (24). In marked

ontrast, increases in LV mass and myocyte size duringressure overload induced by aortic banding were similar inox2�/� and wild-type (WT) mice, indicating that Nox2 is

ot essential for development of pressure-overload LVH15). The latter findings were confirmed by Maytin et al.26). Because ROS potentially affect several aspects ofardiac structure and function, however, these studies do notddress the role of Nox2 in other components of the overallesponse to pressure overload. In the present study, wexamined the role of Nox2 in the contractile dysfunctionnd fibrosis associated with pressure-overload LVH.

ETHODS

xperimental LVH. Adult male Nox2�/� and matchedT littermate controls underwent suprarenal abdominal

ortic banding or sham surgery (15). Banding caused similarncreases in invasively measured systolic blood pressure in WTsham, 90 � 3 mm Hg; banded, 135 � 5 mm Hg) andox2�/� mice (sham, 86 � 3 mm Hg; banded, 130 � 8 mmg). Some animals were pretreated with N-acetylcysteine (500g/kg/day in drinking water) for 24 to 36 h before being

illed. All procedures were performed in accordance with

n assessed in left ventricular (LV) sections stained with Picrosirius red.

llage and fibronectin. Data are mean � SEM from six animals. *p � 0.05 versus

tPHstUcfcKaCmREatS(gngGGCCCCACGGrarE1IMdpDtvv

(tLwlmHsapaHrtptwCcqSfmds

R

D(im0Aba(NuIcsoamp

819JACC Vol. 47, No. 4, 2006 Grieve et al.February 21, 2006:817–26 Nox2 and Contractile Dysfunction in LVH

he Guidance on the Operation of the Animals (Scientificrocedures) Act, 1986 (United Kingdom).istology. Interstitial fibrosis was assessed in frozen LV

ections (6 �m) by Picrosirius red staining. Blinded quan-itative image analysis (Openlab, Improvision, Coventry,nited Kingdom), excluding coronary vessels and perivas-

ular regions, was undertaken (27). Immunohistochemistryor 3-nitrotyrosine (28) was performed with a rabbit poly-lonal antibody (Upstate Biotechnology, Dundee, Unitedingdom; 1:1,000), using diaminobenzidine for final visu-

lization and counterstaining with nuclear fast red (Dako-ytomation, Glostrup, Denmark). Non-immune rabbit im-unoglobulin G was used as a negative control.eal-time reverse transcriptase-polymerase chain reaction.xpression of atrial natriuretic factor (ANF), procollagen I/III,

nd fibronectin mRNA were analyzed by real-time reverseranscriptase-polymerase chain reaction with fluorescentYBR Green technology on a Prism 7000 HT systemApplied Biosystems, Warrington, United Kingdom), usinglyceraldehyde-3-phosphate dehydrogenase mRNA forormalization (15). The primers were (5= to 3=):lyceraldehyde-3-phosphate dehydrogenase forward CGT-CCGCCTGGAGAA, reverse CCCTCAGATGCCT-CTTCAC; ANF forward CGTGCCCCGACCCA-GCCAGCATGGGCTCC, reverse GGCTCCGAGGG-CAGCGAGCAGAGCCCTCA; procollagen I�1 forwardCTCAGGGTATTGCTGGACAAC, reverse TTGAT-CAGAAGGACCTTGTTTG; procollagen III�1 forwardGGAGCCAGTGGCCATAATG, reverse TGACCAT-TGATCCAGGGTTTC; fibronectin forward CCGGTG-CTGTCAGTCAGA, reverse CCGTTCCCACTGCT-ATTTATC. The comparative Ct method was used for

elative quantification, with validation experiments showingpproximately equal efficiencies of primers. All samples wereun in triplicate, and data were expressed in arbitrary units.chocardiography. Mice were anesthetized with 1% to.5% isoflurane/oxygen and placed on a warming pad.mages were acquired using a Sonos 5500 system with a 15

Hz linear probe (Philips, Bothell, Washington). Two-imensional LV short-axis views were acquired at theapillary muscle level, and M-mode recordings were made.epth settings were adjusted to maximize frame rate (220

o 260 Hz) and optimize temporal resolution. The inter-entricular septal thickness in diastole (IVSD) and the leftentricular end-diastolic and end-systolic dimensions

Table 1. Echocardiographic Parameters in Sha

Sham WT Ban

Heart rate (beats/min) 431 � 18 451IVSD (mm) 0.77 � 0.02 1.13LVEDD (mm) 3.92 � 0.05 3.82LVESD (mm) 2.48 � 0.10 3.01FS (%) 36.9 � 2.4 21.4

Data are mean � SEM from 8 to 10 animals. *p � 0.05 ve

FS � fractional shortening; IVSD � interventricular septal thi

dimension; LVESD � left ventricular end-systolic dimension; WT

LVEDD and LVESD, respectively) were measured usinghree consecutive cycles.V pressure-volume analyses. Left ventricular functionas analyzed in isolated ejecting hearts under controlled

oading and heart rate of 450 bpm (29,30). A custom 1.4-Ficroconductance catheter (SPR-853; Millar Instruments,ouston, Texas) was inserted into the LV apex for mea-

urement of volume and high-fidelity pressure. Atrial inflownd aortic outflow were measured by ultrasonic transitrobes (1N; Transonic Systems, Ithaca, New York). Lefttrial filling pressure (preload) was varied from 10 to 25 cm

2O to generate Starling curves. The LV pressure-volumeelationships were also studied during transient occlusion ofhe aortic outflow tract, and end-systolic and end-diastolicressure-volume relations (ESPVR and EDPVR, respec-ively) were determined. Absolute volume measurementsere corrected for alpha and parallel conductance (30,31).ardiomyocytes. Ventricular myocytes were isolated by

ollagenase digestion, and single myocyte shortening wasuantified (29).tatistics. Between 5 and 10 animals per group were studied

or all protocols. Data were analyzed by two-way repeatedeasures ANOVA, one-way ANOVA, or an unpaired Stu-

ent’s t test, as appropriate. p � 0.05 was consideredignificant.

ESULTS

evelopment of LVH. Consistent with our previous study15), two weeks of pressure overload caused similar increasesn the LV:body weight ratio in banded WT (4.57 � 0.29

g/g vs. 3.21 � 0.07 mg/g, p � 0.05) and Nox2�/� (4.87 �.18 mg/g vs. 3.40 � 0.07 mg/g, p � 0.05) versus shams.trial, right ventricular, lung, and body weights were similaretween groups. The ANF mRNA expression increased tosimilar extent in banded WT and Nox2�/� versus sham

WT, 16.7 � 2.1 arbitrary units vs. 2.3 � 0.1 arbitrary units;ox2�/�, 19.3 � 3.6 arbitrary units vs. 2.9 � 1.5 arbitrary

nits, p � 0.05).nterstitial fibrosis. After 2 weeks, interstitial fibrosis in-reased significantly in banded WT mice compared withham (63 � 11%, p � 0.05) but no significant changeccurred in banded Nox2�/� mice (Fig. 1A). Procollagen Ind III mRNA were also significantly increased in banded WTice (Figs. 1B and 1C). In contrast, no significant increase in

rocollagen I mRNA occurred in banded Nox2�/� mice, whereas

perated and Banded WT and Nox2�/� Mice

T Sham Nox2�/� Banded Nox2�/�

411 � 11 437 � 1704* 0.75 � 0.04 1.06 � 0.02*08 3.94 � 0.15 3.93 � 0.0812* 2.64 � 0.19 2.61 � 0.102* 33.6 � 3.1 33.6 � 2.5

propriate sham control.

m-O

ded W

� 18� 0.� 0.� 0.� 2.

rsus ap

ckness in diastole; LVEDD � left ventricular end-diastolic

� wild-type.

ti(Edisc

IsaLossa

F(c

T

CLLLLLTLLSESEE

S� end

L ic volu

820 Grieve et al. JACC Vol. 47, No. 4, 2006Nox2 and Contractile Dysfunction in LVH February 21, 2006:817–26

he increase in procollagen III mRNA was significantly lower thann banded WT. Fibronectin mRNA did not increase in any groupFig. 1D).

chocardiography. Consistent with the morphometricata, banded WT and Nox2�/� mice showed similar

ncreases in IVSD after two weeks (Table 1). Fractionalhortening was significantly reduced in banded WT miceompared with sham, with a significant increase in LVESD.

igure 2. Left ventricular (LV) function in isolated ejecting hearts from w

able 2. Hemodynamic Parameters in Sham-Operated and Band

Sham WT

F (ml min�1 g�1) 44.8 � 6.2VPmax (mm Hg) 84.3 � 1.1VESP (mm Hg) 52.2 � 0.7VEDP (mm Hg) 4.5 � 0.4VdP/dtmax (mm Hg s�1) 5,757 � 178VdP/dtmin (mm Hg s�1) �3,576 � 135 �au (ms) 8.0 � 0.3VESV (�l) 20.2 � 1.6VEDV (�l) 47.7 � 2.4troke volume (�l) 27.5 � 1.1jection fraction (%) 58.0 � 1.5W (mm Hg·�l g�1) 18,925 � 1,263 1SPVR 4.81 � 0.36DPVR 0.065 � 0.004

teady-state cardiac function was assessed at 20-cm pre-load. Data are mean � SEMCF � coronary flow; EDPVR � end-diastolic pressure-volume relation; ESPVR

V end-diastolic volume; LVESP � LV end-systolic pressure; LVESV � LV end-systol

EDV). Changes in A and D: maximum rate of rise of LV pressure (LVdP/dtmaxhanges in C and F: stroke work (SW). Data are mean � SEM from eight an

n contrast, neither LVESD nor fractional shortening wereignificantly altered in banded Nox2�/� mice. Heart ratend LVEDD were similar among groups.V pressure-volume analyses. Table 2 shows parametersf steady-state cardiac function in WT and Nox2�/� mice,tudied two weeks after surgery. Banded WT mice showedignificant reductions in LVdP/dtmax, end-systolic pressure,nd ejection fraction, and an increase in end-systolic volume

pe (WT) and Nox2�/� mice across a range of LV end-diastolic volumes

T and Nox2�/� Mice

ed WT Sham Nox2�/� Banded Nox2�/�

� 3.3 56.8 � 5.4 49.5 � 5.0� 1.2 86.1 � 1.5 87.2 � 2.0� 1.3* 51.3 � 1.3 51.0 � 0.7� 1.0 7.6 � 1.1 7.8 � 0.5� 213* 5,733 � 169 5,913 � 259� 179* �3,474 � 116 �3,273 � 169� 0.5* 8.8 � 0.5 9.2 � 0.7� 3.1* 21.9 � 1.4 23.4 � 1.6� 4.1 49.5 � 1.9 53.8 � 2.2� 1.4 27.7 � 0.8 30.4 � 1.4� 2.0* 56.1 � 1.3 56.7 � 2.1� 977* 19,057 � 1,029 17,741 � 1,520� 0.36* 4.74 � 0.30 4.67 � 0.15� 0.006* 0.080 � 0.005 0.087 � 0.005

eight animals. *p � 0.05 versus appropriate sham control.-systolic pressure-volume relation; LVEDP � LV end-diastolic pressure; LVEDV �me; LVPmax � maximum LV systolic pressure; SW � stroke work; WT � wild-type.

ild-ty

ed W

Band

40.381.448.65.9

4,8793,041

9.628.853.825.147.3

2,8443.68

0.087

from

); changes in B and E: maximum rate of fall of LV pressure (LVdP/dtmin);imals. *p � 0.05 banded (solid squares) versus sham (open squares).

cLrme

wsmTmpw

dwrmcs3Ccpcr

(f(EdwebLi5iiib�0�Naft

a3

F( . (B ap

821JACC Vol. 47, No. 4, 2006 Grieve et al.February 21, 2006:817–26 Nox2 and Contractile Dysfunction in LVH

ompared with sham, indicating impaired systolic function.VdP/dtmin was also significantly reduced, and isovolumic

elaxation time-constant (Tau) increased in banded WTice, indicating impaired relaxation. None of these param-

ters were significantly altered in banded Nox2�/� mice.Figure 2 shows LVdP/dtmax, LVdP/dtmin, and stroke

ork measured across a range of LV volumes and confirmsignificant reductions in these parameters in banded WTice but preservation of function in banded Nox2�/� mice.he differences in stroke work between WT and Nox2�/�

ice are further appreciated from the representative LVressure-volume loops shown in Figures 3A and 3D, inhich stroke work is the area of the loop.Figure 3 shows representative examples of data obtained

uring transient aortic occlusion. The slope of the ESPVRas significantly decreased in banded WT mice, indicating

educed contractility, but was unaltered in banded Nox2�/�

ice (Figs. 3B and 3E). The slope of the EDPVR was signifi-antly increased in banded WT mice, indicating increased dia-tolic stiffness, but was unaltered in banded Nox2�/� mice (Figs.C and 3E). Mean data are shown in Table 2.ardiomyocyte function. To assess whether differences in

ontractile function found at the whole-heart level were alsoresent at the cellular level, we studied isolated cardiomyo-ytes. Myocytes from banded WT mice had significantly

igure 3. Representative examples of left ventricular (LV) pressure-volumeWT) and Nox2�/� mice. (A and D) Steady-state pressure-volume loopsressure volume relation (EDPVR).

educed twitch shortening (by �20%) compared with sham s

Figs. 4A and 4C). In contrast, the shortening of myocytesrom banded Nox2�/� mice was non-significantly reducedFigs. 4B and 4D).ffect of N-acetylcysteine. Treatment with the antioxi-ant N-acetylcysteine for 36 h before killing at the two-eek post-surgical time point resulted in normalization of

chocardiographic fractional shortening and LVESD inanded WT mice (Figs. 5A and 5C); heart rate, IVSD, andVEDD (Fig. 5B) were unaffected. N-acetylcysteine also

ncreased LVdP/dtmax, stroke work, ESPVR (Figs. 5D toF), ejection fraction, and LVdP/dtmin (data not shown) insolated ejecting hearts of banded WT mice. Furthermore,n vivo N-acetylcysteine treatment for 24 h significantlymproved contraction of cardiomyocytes isolated fromanded WT hearts (e.g., shortening velocity, untreated,2.0 � 0.07 �ms�1; treated, �2.9 � 0.16 �ms�1; p �

.05), but had minimal effects in the sham group (untreated,2.4 � 0.07 �ms�1, treated, �2.6 � 0.11 �ms�1; p �S). A similar improvement of contractility was found with

ddition of N-acetylcysteine to isolated hearts or myocytesor 15 to 20 min, but this effect did not stabilize over thisime period (i.e., continued to increase).

To assess whether the effects of N-acetylcysteine werettributable to reduced oxidative stress, we assessed-nitrotyrosine staining in LV sections. Banded WT hearts

ses in banded (solid squares) and sham-operated (open squares) wild-typend E) End-systolic pressure relation (ESPVR). (C and F) End-diastolic

analy

howed increased 3-nitrotyrosine staining, but this was

swCTbdw

esatmLr

Fo ur heav

Fa

822 Grieve et al. JACC Vol. 47, No. 4, 2006Nox2 and Contractile Dysfunction in LVH February 21, 2006:817–26

ignificantly reduced in hearts from banded animals treatedith N-acetylcysteine for 36 h (Fig. 6).ontractile function after eight weeks of pressure overload.o assess whether the differences in contractile functionetween WT and Nox2�/� mice persisted after longerurations of pressure overload, animals were studied at eighteeks after surgery. Figure 7 shows cardiac functional param-

igure 4. Effect of pressure overload on cell shortening (CS) in cardiomyocf myocyte contraction. (C and D) Mean � SEM from �60 cells from foersus sham.

igure 5. Effect of N-acetylcysteine on cardiac function assessed (A to C) by echocanimals; darker columns � N-acetylcysteine-treated animals. Data are mean �

ters measured in isolated ejecting hearts. Banded WT micehowed substantial reductions in LVdP/dtmax, LVdP/dtmin,nd cardiac work compared with shams, which were greaterhan those observed after two weeks (Fig. 3). Banded Nox2�/�

ice now also showed significant reductions in LVdP/dtmax,VdP/dtmin, and cardiac work, but the degree of impairment

emained much less than in banded WT mice.

rom wild-type (WT) and Nox2�/� mice. (A and B) Representative tracesrts. Dark columns � banded mice; lighter columns � shams. *p � 0.05

ytes f

rdiography and (D to F) in isolated ejecting hearts. Light columns � controlSEM from eight animals. *p � 0.05 versus corresponding control.

D

TNcpoippb

psofiwoio

mrdfotsimiAiwNatAs

F ) mic(

823JACC Vol. 47, No. 4, 2006 Grieve et al.February 21, 2006:817–26 Nox2 and Contractile Dysfunction in LVH

ISCUSSION

he major novel finding of this study is that the Nox2ADPH oxidase contributes to development of cardiac

ontractile dysfunction and interstitial fibrosis in response toressure overload, even though it is not essential for devel-pment of LVH per se. The Nox2 oxidase thereforenfluences specific components of the overall response toressure overload, indicating that related but distinct com-onents of the overall cardiac hypertrophic phenotype maye subject to independent regulation.Oxidative stress is recognized to be involved in LVH

athophysiology, but the precise roles of different ROSources remains unclear (1–6). Furthermore, the possibilityf divergent effects on contractile function or interstitialbrosis as opposed to increases in LV mass has not beenidely addressed. Recent work suggests a role for NADPHxidases in the development of cardiac hypertrophy andnterstitial fibrosis (14,21,24). A relatively unique attribute

igure 6. Representative left ventricular (LV) sections from wild-type (WTC) N-acetylcysteine-treated sham; (D) N-acetylcysteine-treated banded.

f NADPH oxidases in contrast to other ROS sources (e.g., N

itochondria, xanthine oxidase) is that ROS generation inesponse to specific stimuli and subsequent modulation ofownstream signal transduction seems to be a primaryunction of these enzymes (12,13). In addition, the presencef at least two cardiac Nox isoforms (Nox2 and Nox4) raiseshe possibility of not only stimulus-specific but also isoform-pecific roles (12,13,15). Our previous findings that LVHnduced by subpressor AngII was inhibited in Nox2�/�

ice (24) indicated that Nox2-derived ROS are specificallynvolved in cardiac AngII signaling. Consistent with this,ngII-induced cardiac hypertrophy and fibrosis were inhib-

ted in mice lacking apoptosis signal-regulating kinase 1,hich is known to be redox-sensitive (32). However,ox2�/� mice developed a similar degree of LVH to WT

nimals in response to pressure overload (15,26), indicatinghat even though Nox2 mediates hypertrophy in response tongII, other pathways may be more important or may

ubstitute for Nox2 when the stimulus is pressure overload.

e stained for 3-nitrotyrosine (20� magnification). (A) Sham, (B) banded;

evertheless, the current study shows that Nox2 plays a

mcINNapdNo

ccrheasosNrgNsN

dotpWwcrdaiaEwtdriWto(gt

FLv

824 Grieve et al. JACC Vol. 47, No. 4, 2006Nox2 and Contractile Dysfunction in LVH February 21, 2006:817–26

ajor role in the development of interstitial fibrosis andontractile dysfunction in response to pressure overload.nterestingly, pressure overload was found to increaseADPH oxidase activity and ROS production even inox2�/� mice, secondary to upregulation of Nox4 in these

nimals (15). This finding, together with the results of theresent study, indicates that the downstream effects of Nox2uring pressure overload are highly specific; i.e., neitherox4 nor other ROS sources apparently promotes fibrosis

r contractile dysfunction in Nox2�/� mice.Previous seminal studies have established that although

ardiomyocyte hypertrophy and interstitial fibrosis ofteno-exist, they are to a significant extent independentlyegulated (33). Oxidative stress is pro-fibrotic both in theeart and in other organs, by stimulating fibroblast prolif-ration, collagen deposition, and activation of matrix met-lloproteinases (1,33–35). The novel finding of the currenttudy is the demonstration of a specific role for the Nox2xidase in promoting interstitial cardiac fibrosis during pres-ure overload. Inhibition of interstitial fibrosis in bandedox2�/� mice was shown both by histology and real-time

everse transcriptase-polymerase chain reaction. Taken to-ether with previous data indicating an involvement ofox2 in AngII-induced cardiac, vascular, and hepatic fibro-

is (24,28,36), these results may suggest a broader role of

igure 7. Left ventricular (LV) function in isolated ejecting hearts from wildVdP/dtmax; (B and E) LVdP/dtmin; and (C and F) cardiac work (CW). Dersus sham (open squares).

ox2 in mediating pathological fibrosis. w

The other major effect of Nox2 was found to be in theevelopment of contractile dysfunction during pressureverload. We undertook detailed characterization of con-ractile function by both echocardiography in vivo and LVressure-volume analyses in isolated ejecting hearts. The

T mice had significant systolic dysfunction after twoeeks of pressure overload as evidenced by reduced echo-

ardiographic fractional shortening, increased LVESD, andeduction in pressure (LVdP/dtmax) and volume-based in-ices (LV end-systolic volume and ejection fraction) as wells the slope of the ESPVR. Banded WT mice also hadmpaired relaxation (reduced LVdP/dtmin, prolonged tau)nd evidence of diastolic dysfunction (increased slope ofDPVR). In marked contrast, none of these parametersere significantly altered in banded Nox2�/� mice. Fur-

hermore, brief (24-h to 36-h) treatment with the antioxi-ant N-acetylcysteine improved contractile function andeduced LV 3-nitrotyrosine staining in WT bands, suggest-ng that the dysfunction was indeed mediated via ROS.

hereas we previously showed that chronic (two-week)reatment with N-acetylcysteine inhibited the developmentf pressure-overload hypertrophy (15), the current brief24-h to 36-h) protocol was specifically chosen to investi-ate effects on contractile function independent of hyper-rophy. With more prolonged pressure overload (eight

(WT) and Nox2�/� mice eight weeks after surgery. Changes in (A and D):e mean � SEM from five to six animals. *p � 0.05 banded (solid squares)

-typeata ar

eeks), we found that contractile dysfunction did develop in

NlmtoipOwttpb�sbot

ficsfesfchHboarno(

ocpohoai(gsepTrt

RpR

R

1

1

1

1

1

1

1

1

1

1

2

825JACC Vol. 47, No. 4, 2006 Grieve et al.February 21, 2006:817–26 Nox2 and Contractile Dysfunction in LVH

ox2�/� mice but the magnitude of dysfunction remainedess than with banded WT. These data suggest that Nox2

ay be particularly important for contractile dysfunction inhe early stages of pressure overload. With more prolongedverload, it is likely that other mechanisms, such as ROS-ndependent abnormalities of excitation-contraction cou-ling and energetic deficit, may also become relevant (37).ur results contrast with a recently published study inhich no differences in cardiac function were found be-

ween Nox2�/� and WT mice subjected to aortic constric-ion (26). However, in that study, the pressure overload wasrobably much more severe than in the present studyecause it resulted in �50% LVH after just one week and50% mortality, compared with �20% mortality in our

tudy. These data leave open the possibility that the contri-ution of Nox2 to contractile dysfunction during pressureverload may depend on the degree of overload, a questionhat warrants further investigation.

Increased ROS production may lead to contractile dys-unction through several mechanisms, including alterationsn cellular energetics, changes in excitation-contractionoupling, and/or alterations in myofilament calcium respon-iveness (1,8–10,38). At the whole-heart level, contractileunction may also be influenced by alterations in thextracellular matrix (11). In the current study, we foundignificant differences in function of cardiomyocytes isolatedrom banded Nox2�/� and WT, suggesting that a signifi-ant proportion of the contractile dysfunction in wholeearts may be attributable to intrinsic myocyte dysfunction.owever, the differences in diastolic properties observed

etween banded Nox2�/� and WT may be related to thebserved changes in interstitial fibrosis. The precise mech-nisms underlying Nox2-dependent contractile dysfunctionequire further investigation, but our preliminary studies doot show significant differences in expression of SERCA2ar phospholamban between banded Nox2�/� and WTdata not shown).

In summary, this study shows that the Nox2 NADPHxidase specifically contributes to the development of bothontractile dysfunction and interstitial fibrosis in response toressure overload, despite not being essential for the devel-pment of LVH per se. This divergent regulation of LVypertrophy and LV function is analogous to a number ofther recent reports of experimental models in which thesespects of cardiac hypertrophy have been found to bendependently regulated, e.g., by the PI3K-PTEN pathway39,40) and the MEKK1-JNK pathway (41). Also analo-ous to these reports, the Nox2 oxidase in the heart not onlyeems to be activated in an isoform-specific manner, but alsoxerts isoform-specific downstream effects on distinct com-onents of the overall cardiac hypertrophic phenotype.hese data underscore the likely importance of independent

edox regulation of different aspects of the cardiac response

o pressure overload.

eprint requests and correspondence: Dr. Ajay M. Shah, De-artment of Cardiology, GKT School of Medicine, Bessemeroad, London SE5 9PJ England. E-mail: ajay.shah@kcl.ac.uk.

EFERENCES

1. Dhalla NS, Temsah RM, Netticadan T. Role of oxidative stress incardiovascular diseases. J Hypertens 2000;18:655–73.

2. Nakamura K, Fushimi K, Kouchi H, et al. Inhibitory effects ofantioxidants on neonatal rat cardiac myocyte hypertrophy induced bytumor necrosis factor-� and angiotensin II. Circulation 1998;98:794–9.

3. Pimentel DR, Amin JK, Xiao L, et al. Reactive oxygen species mediateamplitude-dependent hypertrophic and apoptotic responses to me-chanical stretch in cardiac myocytes. Circ Res 2001;89:453–60.

4. Hirotani S, Otsu K, Nishida K, et al. Involvement of nuclear factor-�Band apoptosis signal-regulating kinase 1 in G-protein-coupled recep-tor agonist-induced cardiomyocyte hypertrophy. Circulation 2002;105:509–15.

5. Yamamoto M, Yang G, Hong C, et al. Inhibition of endogenousthioredoxin in the heart increases oxidative stress and cardiac hyper-trophy. J Clin Invest 2003;112:1395–406.

6. Takemoto M, Node K, Nakagami H, et al. Statins as antioxidanttherapy for preventing cardiac myocyte hypertrophy. J Clin Invest2001;108:1429–37.

7. Cesselli D, Jakoniuk I, Barlucchi L, et al. Oxidative stress-mediatedcardiac cell death is a major determinant of ventricular dysfunction andfailure in dog dilated cardiomyopathy. Circ Res 2001;89:279–86.

8. Josephson RA, Silverman HS, Lakatta EG, Stern MD, Zweier JL.Study of the mechanisms of hydrogen peroxide and hydroxyl freeradical-induced cellular injury and calcium overload in cardiac myo-cytes. J Biol Chem 1991;266:2354–61.

9. Gao WD, Liu Y, Marban E. Selective effects of oxygen free radicals onexcitation-contraction coupling in ventricular muscle. Implications forthe mechanism of stunned myocardium. Circulation 1996;94:2597–604.

0. Xie YW, Kaminski PM, Wolin MS. Inhibition of rat cardiac musclecontraction and mitochondrial respiration by endogenous peroxynitriteformation during posthypoxic reoxygenation. Circ Res 1998;82:891–7.

1. Kinugawa S, Tsutsui H, Hayashidani S, et al. Treatment withdimethylthiourea prevents left ventricular remodeling and failure afterexperimental myocardial infarction in mice. Role of oxidative stress.Circ Res 2000;87:392–8.

2. Griendling KK, Sorescu D, Ushio-Fukai M. NAD(P)H oxidase. Rolein cardiovascular biology and disease. Circ Res 2000;86:494–501.

3. Lambeth JD. Nox enzymes and the biology of reactive oxygen. NatRev Immunol 2004;4:181–9.

4. Li JM, Gall NP, Grieve DJ, Chen M, Shah AM. Activation ofNADPH oxidase during progression of cardiac hypertrophy to failure.Hypertension 2002;40:477–84.

5. Byrne JA, Grieve DJ, Bendall JK, et al. Contrasting roles of NADPHoxidase isoforms in pressure overload versus angiotensin II-inducedcardiac hypertrophy. Circ Res 2003;93:802–5.

6. Xiao L, Pimentel DR, Wang J, Singh K, Colucci WS, Sawyer DB.Role of reactive oxygen species and NAD(P)H oxidase in �1-adrenoceptor signaling in adult rat cardiac myocytes. Am J Physiol CellPhysiol 2002;282:C926–34.

7. Wenzel S, Taimor G, Piper HM, Schluter KD. Redox-sensitiveintermediates mediate angiotensin II-induced p38 MAP kinase acti-vation, AP-1 binding activity, and TGF-� expression in adult ven-tricular cardiomyocytes. FASEB J 2001;15:2291–3.

8. Pracyk JB, Tanaka K, Hegland DD, et al. A requirement for the rac1GTPase in the signal transduction pathway leading to cardiac myocytehypertrophy. J Clin Invest 1998;102:929–37.

9. Higuchi Y, Otsu K, Nishida K, et al. The small GTPase bindingprotein Rac1 induces cardiac myocyte hypertrophy through the acti-vation of apoptosis signal-regulating kinase 1 and nuclear factor-�B.J Biol Chem 2003;278:20770–7.

0. MacCarthy PA, Grieve DJ, Li JM, Dunster C, Kelly FJ, Shah AM.

Impaired endothelial regulation of ventricular relaxation in cardiac

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

826 Grieve et al. JACC Vol. 47, No. 4, 2006Nox2 and Contractile Dysfunction in LVH February 21, 2006:817–26

hypertrophy: role of reactive oxygen species and NADPH oxidase.Circulation 2001;104:2967–74.

1. Nadruz W Jr., Lagosta VJ, Moreno H Jr., Coelho OR, Franchini KG.Simvastatin prevents load-induced protein nitration in overloadedhearts. Hypertension 2004;43:1060–6.

2. Heymes C, Bendall JK, Ratajczak P, et al. Increased myocardialNADPH oxidase activity in human heart failure. J Am Coll Cardiol2003;41:2164–71.

3. Maack C, Kartes T, Kilter H, et al. Oxygen free radical release inhuman failing myocardium is associated with increased activity ofRac1-GTPase and represents a target for statin therapy. Circulation2003;108:1567–74.

4. Bendall JK, Cave AC, Heymes C, Gall N, Shah AM. Pivotal role ofgp91phox-containing NADPH oxidase in angiotensin II-induced car-diac hypertrophy. Circulation 2002;105:293–6.

5. Pollock JD, Williams DA, Gifford MA, et al. Mouse model ofX-linked chronic granulomatous disease, an inherited defect in phago-cyte superoxide production. Nat Genet 1995;9:202–9.

6. Maytin M, Siwik DA, Ito M, et al. Pressure overload-inducedmyocardial hypertrophy in mice does not require gp91phox. Circulation2004;109:1168–71.

7. Brilla CG, Janicki JS, Weber KT. Impaired diastolic function andcoronary reserve in genetic hypertension: role of interstitial fibrosis andmedial thickening of intramyocardial coronary arteries. Circ Res1991;69:107–15.

8. Wang HD, Xu S, Johns DG, et al. Role of NADPH oxidase in thevascular hypertrophic and oxidative stress response to angiotensin II inmice. Circ Res 2001;88:947–53.

9. Layland J, Grieve DJ, Cave AC, Sparks E, Solaro RJ, Shah AM.Essential role of troponin I in the positive inotropic response to�-adrenergic stimulation in the ejecting mouse heart. J Physiol

2004;556:835–47.

0. Grieve DJ, Cave AC, Byrne JA, Layland J, Shah AM. Analysis of exvivo left ventricular pressure-volume relations in the isolated murineejecting heart. Exp Physiol 2004;89:573–82.

1. Georgakopoulos D, Mitzner WA, Chen CH, et al. In vivo murine leftventricular pressure-volume relations by miniaturized conductancemicromanometry. Am J Physiol 1998;274:H1416–22.

2. Izumiya Y, Kim S, Izumi Y, et al. Apoptosis signal-regulating kinase1 plays a pivotal role in angiotensin II-induced cardiac hypertrophyand remodeling. Circ Res 2003;93:874–83.

3. Weber KT. Extracellular matrix remodeling in heart failure. A role forde novo angiotensin II generation. Circulation 1997;96:4065–82.

4. Irani K, Xia Y, Zweier JL, et al. Mitogenic signaling mediated byoxidants in Ras-transformed fibroblasts. Science 1997;275:1649–52.

5. Poli G, Parola M. Oxidative damage and fibrogenesis. Free Radic BiolMed 1997;22:287–305.

6. Bataller R, Schwabe RF, Choi YH, et al. NADPH oxidase signaltransduces angiotensin II in hepatic stellate cells and is critical inhepatic fibrosis. J Clin Invest 2003;112:1383–94.

7. Hunter JJ, Chien KR. Signaling pathways for cardiac hypertrophy andfailure. N Engl J Med 1999;341:1276–83.

8. Ide T, Tsutsui H, Hayashidani S, et al. Mitochondrial DNA damageand dysfunction associated with oxidative stress in failing hearts aftermyocardial infarction. Circ Res 2001;88:529–35.

9. Crackower MA, Oudit GY, Kozieradzki I, et al. Regulation ofmyocardial contractility and cell size by distinct PI3K-PTEN signal-ling pathways. Cell 2002;110:737–49.

0. Patrucco E, Notte A, Barberis L, et al. PI3K� modulates the cardiacresponse to chronic pressure overload by distinct kinase-dependentand -independent effects. Cell 2004;118:375–87.

1. Sadoshima J, Montagne O, Wang Q, et al. The MEKK1-JNKpathway plays a protective role in pressure overload but does not

mediate cardiac hypertrophy. J Clin Invest 2002;110:271–9.