Loss of Cardiomyocyte Integrin Linked Kinase Produces …circep.ahajournals.org/content/circae/early/2015/06/12/CIRCEP.115... · Loss of Cardiomyocyte Integrin Linked Kinase Produces

DOI: 10.1161/CIRCEP.115.001668

1

Loss of Cardiomyocyte Integrin Linked Kinase Produces an Arrhythmogenic

Cardiomyopathy in Mice

Running title: Le Quang et al.; ILK-deletion arrhythmogenic cardiomyopathy

Khai Le Quang, MD, PhD1*; Ange Maguy, PhD1*; Xiao-Yan Qi, PhD1*; Patrice Naud, PhD1;

Feng Xiong, PhD1,2; Artavazd Tadevosyan, MSc1; Yan-Fen Shi, MD1; Denis Chartier, MSc1;

Jean-Claude Tardif, MD1; Dobromir Dobrev, MD3; Stanley Nattel, MD1,2

1Department of Medicine, Montreal Heart Institute & Université de Montréal, Montreal; 2Department of Pharmacology and Therapeutics, McGill University, Montreal, Quebec, Canada; 3Institute of

Pharmacology, West German Heart and Vascular Center, University Duisburg-Essen, Essen, Germany *shared 1st authorship

Correspondence:

Stanley Nattel, MD

Department of Medicine

Montreal Heart Institute &

Université de Montréal

5000 Belanger Street E

Montreal, Quebec, H1T 1C8

Canada

Tel.: 514-376-3330 ext. 3990

Fax: 514-593-2493

E-mail: [email protected]

Journal Subject Codes: [132] Arrhythmias - basic studies, [106] Electrophysiology, [5] Arrhythmias, clinical electrophysiology, drugs

Jean Claude Tardif, MDff ; Dobromir Dobrev, MD ; Stanley Nattel, MD

1DeDeepapapartrtrtmemementntnt ooof MeMeMedicine, Montreal Heart Institute &&& UnUnU iversité de Montréall,,, MMMontreal; 2Department of Pharmamaacccologygyg and Therapeutics, McGill Univerrrsitty, Montreal, QuQuQuebec, Canada;; 3Institute of

Phhharmacologyyy, WWWesttt GGGererermamamannn HeHeHeararttt anannd d VaVaVascuululararr Cennteeer, UUUnniniveveersrsrsity DDuiuuisbsbsburururg-g--y Essesesenn,n, EEEsssss enenen, GeGeGermrmrmanana yyy **s*shhareeed 1st aautututhhhort shshhipipip

Correspopopondndndenenencecece:::

Stanleyyy NNatt llel,, MDMDD

f di i

by guest on June 16, 2018http://circep.ahajournals.org/

Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from

http://circep.ahajournals.org/



























DOI: 10.1161/CIRCEP.115.001668

2

Abstract:

Background - Integrin-linked kinase (ILK), a serine/threonine protein-kinase, has roles in cell-

signaling and molecular scaffolding. ILK-mutation/deletion causes cardiomyopathic

phenotypes, but the functional and electrophysiological features have not been characterized.

This study investigated the structural, functional, ion-channel and electrophysiological changes

associated with cardiomyocyte-directed ILK-deletion in mice.

Methods and Results - Adult mice with cardiomyocyte-directed ILK-knockout (KO) were

compared to littermate controls. KO-mice showed markedly-increased mortality, with sudden-

death beginning after 5 weeks and 100%-mortality at 18 weeks. Spontaneous and inducible

ventricular tachyarrhythmias were common in 10-week KO-mice, occurring in 60% and 86%

respectively, and absent in controls (P<0.001, P<0.05 vs KO-mice). Ventricular refractoriness

was prolonged, along with both QRS and QT interval. Action-potentials were prolonged and

displayed triggered activity. A wide range of ion-currents were downregulated, including total,

fast and slow components of transient-outward K+-current and inward-rectifier K+-current, along

with corresponding ion-channel subunit genes, providing a plausible explanation of AP-

prolongation. At 5 weeks, only voltage-dependent K+-currents were reduced, possibly related to

direct ILK-Kv4.2 subunit interactions. Action-potentials were prolonged, but no arrhythmias or

cardiac dysfunction were noted. Structural remodeling was prominent at 10 weeks: connexin-43

was downregulated and redistributed to lateral cell-margins, and left-ventricular fibrosis

occurred, with a strong regional distribution (predominating in the basal left ventricle).

Conduction was slowed. High-throughput quantitative polymerase-reaction gene-expression

studies in 10-week ILK-KO showed upregulation of structural, remodeling and fibrosis-related

genes and downregulation of a wide range of ion channel and transporter subunits.

Conclusions - Cardiomyocyte ILK-deletion produces a lethal arrhythmogenic cardiomyopathy

associated with important ion-channel and structural remodeling.

Key words: cardiomyopathy, ventricular tachycardia, sudden cardiac death, potassium channels, remodeling, integrin linked kinase

p y, ( , )

was prolonged, along with both QRS and QT interval. Action-potentials were prooololoongnggededed anananddd

displayed triggered activity. A wide range of ion-currents were downregulated, inclul ding total,

faststst aaandndnd slololow ww coompmpm onents of transient-outward KKK+++ c-ccuru rent and inwwwaraa d-reeectctctifier K+-current, along

wwwithhh corresponnndididinngg iiononon-c-cchahahannnnnnelelel suububuuuniitit gggenenees,s,, proooviidiingngng a pppllalausibii lelle expxpxplalaanaatititioonon ooof f f APAPAP---

prprrolollonoo gation.. AAtt 555 weeeekkks, onlnlnlyyy voltttagaa eee-dddepeeenndenttt KKK+--cucucurreeenttts wwwerrre reddduccceddd, pppossibblblyyy rrrellateddd tto

directctt IIILKLKLK K-Kv4v44 22.2 ssububb nunititit interacacctititiono s. AcAcActtitionon-pp tototenentitit alallsss wewererre proollolongng dded, bbubut nonoo aaarrrhyththth imimiasas oorr

cardiac dydydysfsfsfunununctctctioioion nn wewewererere nnnotototededed.. StStStruuuctctctuuurararal ll rereremomoodededelililingngng wwwasasas ppprororomimiminenenentntnt aaat t t 101010 wwweeeeeeksksks:: cococonnnnnnexin-43

wawas dddownregulllatteddd anddd redididi tst iiribbbutteddd tto lll tater llal c llellll-margiiins, anddd lllefftft-ve tnt iiriculllar fififibbbrosiiis


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

3

Introduction

Integrin-Linked Kinase (ILK) is a widely-expressed protein that contributes to cytoskeletal

anchoring and macromolecular organization.1 In the heart, ILK promotes cardiomyocyte actin-

cytoskeleton linkage to the plasma membrane at costameres.2 By binding to the cytoplasmic tail

of -integrins and scaffolding/recruiting multiprotein complexes, ILK contributes to the

adaptation of cellular processes (growth, proliferation, survival and differentiation) in response

to mechanical forces applied to the extracellular matrix.1,3

ILK upregulates in human pressure-load cardiac hypertrophy, and appears central to the

adaptive hypertrophic response.4 Adenovirus-mediated ILK gene-transfer attenuates adverse

left-ventricular post-myocardial infarction5 and enhances functional preservation/survival in rats

with doxorubicin-induced dilated cardiomyopathy.6 A focal-adhesion complex including ILK,

PINCH-1 and -parvin promotes cell- -4 treatment

post-MI.7 We previously demonstrated that cardiac-directed knockout of ILK causes a severe

cardiomyopathy and premature death in mice.8 Subsequently, ILK-mutation was demonstrated

to cause dilated cardiomyopathy in man.9 Most of the deaths in ILK-deleted mice are sudden,

presumably due to an acute arrhythmic cause.8 However, the functional, electrophysiological

and arrhythmic phenotype associated with ILK-deletion has not been evaluated. Accordingly,

the present study was designed to evaluate the effects of ILK-deletion on: 1) in vivo

electrophysiology and cardiac rhythm; 2) associated cellular electrophysiology; 3) underlying

ionic current changes and 4) changes in cardiac structure and function.

Methods

A detailed description of Methods is provided in the online Data Supplement. The most

important methods are briefly summarized here.

eft-ventricular post-myocardial infarction5 and enhances functional preservation/n/n/suuurvvviivavaval l l ininin rats

with ddoxorubbicinn-i- nduced dilated cardiomyopathy.y.y 6 A focal-adhesion coompm lex including ILK,

PIPIPINNCNCH-1 and dd --papaparvvvininin ppprororomomom tetetes s s cececelllll --- --4 4 trtrtreaeaeatmtmtmenenentt

popoosttt--MI.7 Wee prrrevvviouuusllly y demomomonstratatatededd tthat caaardiaaac---dirererectccteddd kkknockckkout ooof IILKLKK ccauseees aa sssevvvereee

cardiomyyyopopopathyhyhy aand ppprereremam tureee dddeath in miceee..8 SuSuSubsbb eqquentnn lyyy, , ILK-m-m-mutation n n waaas ss demonstrated

ooo cccauauausesese dddililatatatededed cccararardididiomomomyoyoyopapapaththt yy y inin mmmananan...999 MoMoMoststst oooff ththt ee e dededeatatathshss iinn n ILILILK-KK dededeleleetetetedd d mimicecece aaarerere sssudududdededen,n,n,


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

4

Animal Model

Age-matched mice expressing the cardiac mckCRE transgene and 2 loxP1- flanked ILK alleles

(mckCRE ILKfl/fl-KO mice; ILK-KO) were compared to littermate controls (mckCRE) as

described previously.8 All procedures followed National Institutes of Health guidelines and were

approved by the Animal Research Ethics Committee of the Montreal Heart Institute. Studies

were performed in 5- and 10-week old mice, as well as age-matched controls, as indicated. Male

and female mice were used, in equal proportions.

Echocardiography

Transthoracic echocardiography was performed under 2.5%-isofluorane anesthesia. The person

performing and analyzing echocardiographic imagery was blinded to mouse-group assignment.

The average of 3 consecutive cardiac cycles was used for each measurement. Special care was

taken to get similar imaging planes at follow-up studies.

Histology and Fibrosis Quantification

Ventricular mass was assessed using total ventricular weight-to-body-weight ratio (TVW/BW).

For histological studies, hearts were perfused with 10% neutral-buffered formalin, dehydrated

and embedded in paraffin. Longitudinal sections (6- were stained with Masson’s trichrome.

Connective tissue content was quantified as percentage of surface area, with perivascular

collagen excluded from measurement.

Connexin-43 Immunostaining and Quantification of Lateralization

LV-tissues were embedded in Optimal Cutting-Temperature solution (Sakura) and snap-frozen in

liquid-N2. Cryosections (12- Primary antibody

(rabbit anti-Cx43, AB1727, Millipore) was diluted in a solution containing Triton X-100 for

overnight incubation with cryosections. Alexa Fluor-conjugated donkey anti-rabbit IgG (488-

performing and analyzing echocardiographic imagery was blinded to mouse-grouuupp asassisisigngngnmememe tnt.

The avveraggge of 33 cconsecutive cardiac cycles was usede for each measuremment. Special care was

aaakekeen to get sssimimimilililaraa iimamamagigigingngng plplplanananeseses atata fffololollolowww-u-u-up stttudddiesss..

HHiHiststtolo ogy anndd FFFibbbrosssiss Quaaantntntificatattioi nnn

Ventricuulalalar r massssss wass aaasssss essed d usuu ing total veeentnn ririicucuculalalar weigigight-to-bodddy-y-y-weiggghththt ratttioioi (((TVW/BW).

FoFoForr r hihistststololo ogogogiciccalala ssstututudididieseses,, , heheeararartststs wwwerereree e pepeperfrfusususededed wwwititthh 101010%% % neneneutututrararal-l bububufffferererededed ffororormamamalilinnn,,, dededehyhyhydrdrdratatatededed


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

5

nm) was used as a secondary antibody. Alexa Fluor-conjugated phalloidin (647-nm) was used as

an actin-filament marker. Slides were mounted examined with a Zeiss Axiovert 100-M

microscope coupled to a Zeiss LSM-710 laser-scanning confocal system. Identical settings were

used to image samples from control and ILK-KO animals. Images were deconvolved with the

Maximum Likelihood Estimation algorithm. Connexin-43 lateralization was determined as

previously described in detail,10 using Image Pro Plus 6.0 software (Media Cybernetics). Total

pixel-intensity associated with Connexin-43 staining (x) was first quantified, then lateral

Connexin-43 positive staining was selected and the corresponding pixel-intensity sum (b) was

quantified. The extent of lateralization (y) was expressed as a percentage, as follows:

y=(100b)/x.

Protein-extraction and Immunoblotting

LV-samples from control and ILK-KO hearts were excised, snap-frozen in liquid-N2 and

homogenized in TNE buffer. Homogenized samples were centrifuged at 1000g for 10 minutes,

supernatant collected and ultracentrifuged at 100,000g for 1 hour. The resulting membrane-

fraction pellet was suspended and incubated in TNE buffer containing 1% Triton-X100. The

protein concentration was determined by Bradford assay (BioRad). Protein samples (20 μg) were

separated on 8% poly-acrylamide SDS-PAGE and transferred electrophoretically onto PVDF

membranes. The PVDF membranes were blocked in a PBS-solution containing 0.05% (v/v)

Tween-20 and 5% (w/v) non-fat dried milk (NDM) and incubated overnight at 4°C with primary

antibodies. After washing, membranes were hybridized with HRP-conjugated secondary

antibody. Immunoreactive bands were detected by electrochemoluminescence. Quantification

was performed with Quantity-One software (BioRad). All expression-data are provided relative

to GAPDH staining for the same samples on the same gels. Primary antibodies (1/2000)

y=(100b)/x.

Proteiinn-eextx raracttioioion and Immunoblotting

LVLVLV---samples frffromomom cononntrtrtrololol aaandndn IIILKLKLK---KOKOKO heheh ararartststs wwwereee eeexciiisesesed,d,d, sssnanappp-f-frororozezezenn n ininin liqqquiuiuid-d-d-NNN222 aandndnd

hohoommomogenized d in TNNE bubuuffer... HHHomogogogennnizzzed saaamplless wwwererere cccennntriffufuggged aatt 11100000g0gg for 11000 mmminnnuteees,,,

upernatananant colllecece ted ananand dd ultracaccenee trifuggged at t 10101 0,0,0,0000000g0gg forrr 1 hour. ThTT e resuuultltlting g g memm mbrane-

frfracacactitit ononon pepepelllletetet wawawass s susususpspspenenendedededd d ananandd d inincucucubababatetetedd d inin TTTNENENE bbbufufu fefeerr r cococontntntaiaiaininingngng 111%%% TrTrTritittononon-X-XX1010100.0.0. ThThee e


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

6

included polyclonal rabbit anti-connexin43 (AB1727; Millipore) and monoclonal anti-GAPDH

(10R-G109a) from Fitzgerald. Peroxidase-conjugated AffiniPure donkey anti-rabbit IgG (111-

035-152) and Affinipure donkey anti-mouse IgG (715-035-151) from Jackson ImmunoResearch

were used as secondary antibodies (1/10000). Co-immunoprecipitation was performed as

detailed in Supplemental Methods.

Quantitative Polymerase Chain-reaction (qPCR) Analysis

High-throughput qPCR was performed on samples from the superior LV free-wall of 10-week

ILK-KO and control mice were homogenized, RNA was isolated using Trizol, purified and

quantified via Nanodrop. Absence of genomic DNA-contamination was confirmed by PCR.

The integrity of total RNA was assessed with an Agilent bioanalyser. cDNA was synthesized

from 220 ng total RNA using the High Capacity cDNA Reverse Transcription Kit (Applied

Biosystems) with random hexamer primers. TaqMan low-density arrays (Applied Biosystems)

were used in two-step RT-PCR as previously reported.10 Real-time PCR was performed on the

7900HT Fast Real-Time PCR System. TLDA plates (96-gene capacity) were designed for the

study of genes related to cardiac remodeling, electrophysiology, contractility and fibrosis

(Supplemental Table 1). GAPDH was the internal standard. Results were subjected to non-

directed hierarchical clustering to obtain a global expression-profile of the genes studied

(Supplemental Figure 1). Conventional qPCR was performed with TaqMan probes on additional

5- and 10-week mice samples to compare directly their expression of ion-channel, collagen and

hypertrophic-marker subunits.

Electrocardiography and Telemetry Recording

ECGs were recorded in 5 and 10-week old mice under 2.5%-isoflurane anesthesia. Recordings

were filtered between 0.5 and 500 Hz. Measurements were based on averages of 10 consecutive

The integrity of total RNA was assessed with an Agilent bioanalyser. cDNA waass sysysyntntnthehehesisisizezezeddd

from 2220 ng g totaal l RNA using the High Capacity cDDNAN Reverse Transcrripipption Kit (Applied

BBiBiosssystems)) wwwititithhh rarr ndndndomomom hhhexexamamamererer pppririimememersss.. TaTT qMMMaaan lowowow---dededensn ititityy y arararrarar ysysys (A(A( ppppplililiededed BBBioiosysysystststememems)s)s)

wwewereree used in ttwooo-ssstep RRRTT-PCRCRCR as prprrevevviooouslyyy rrrepoorteeed..10100 ReReeaaal-timememe PCRRR wwwasasas ppperformrmrmededd ooon thhheeR

7900HT FFFasa t Reeealaa -Timememe PCR SSSysysystem. TLDADAA ppplalalatetetess (9(( 6-gegegene capppacacacity)y)y) weeererer desesesigggned for the

tttudududyy y ofofo gggeneneneseses rrrelele atatatededed tttooo cacacardrddiaiaacc c rereremomomodededelilingngng,,, elele ececectrtrtropopophyhyysisis ololo ogogogy,y,y, cccononontrtrtracacactitit lilitytyty aaandndd ffibibbrororosisisisss


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

7

complexes. Standard criteria were used to measure RR, PR, QRS, and QT intervals.11 The

QT-interval (QTc) was corrected for heart rate (HR) as follows: QTc=QT/(RR/100)1/2.12 To

obtain 24-hr ECG recordings in free-moving mice, mice were anaesthetized at 9 weeks of age

with 2.5%-isoflurane, telemetry-transmitters implanted and paired wire electrodes placed under

the skin (to obtain a bipolar chest ECG lead). One week post-implantation, ECG-signals were

computer-recorded on free-moving animals for 24 hours. Heart-rate values were determined

from RR-interval averages over 10 seconds.

Intracardiac Recording, Programmed Stimulation and Optical Mapping

Mice were anaesthetized with 2.5%-isoflurane. An octapolar electrophysiology catheter was

positioned in the RV. Surface (ECG lead-I) and intra-cardiac electrograms (filtered between

0.5 and 500 Hz) were recorded. Refractory periods were determined with a nine-stimulus drive

train (S1) at a cycle-length of 100 ms followed by a 1.5×threshold-current premature stimulus

(S2) decremented in 2-ms intervals. Effective refractory periods (ERPs) were defined as the

longest S1-S2 coupling-interval that failed to capture. Ventricular tachycardia (VT) was defined

as 10 or more successive spontaneous ventricular complexes. Optical mapping studies were

performed to analyze conduction changes in 10-week mouse hearts as detailed in Supplemental

Methods.

Action-potential (AP) Recordings

Fine-tipped microelectrode recordings were obtained from the hearts of 10 week-old mice.

Isolated hearts were perfused with Krebs-Henseleit solution containing (mmol/L): 120 NaCl, 4

KCl, 1.2 NaH2PO4, 1.2 MgSO4, 25 NaHCO3, 1.25 CaCl2, and 5.5 glucose (95%-O2/5%-CO2, pH

7.4) at 36°C. Epicardial AP-duration (APD) was recorded from perfused ventricles with fine-

tipped borosilicate-glass microelectrodes. Preparations were stimulated with 2-ms 1.5xdiastolic-

positioned in the RV. Surface (ECG lead-I) and intra-cardiac electrograms (filteerrred d d bebebetwtwtweeeeeenn n

0.5 annd 500 HzH )) wew re recorded. Refractory periodsd were determined witith h a nine-stimulus drive

rrraainnn (S1) at a aa cycycyclclcle-lelelengngngththth ooof f 1010100 0 0 msmsm fffolololloll wewewed dd by a 1.55×t×t×thrhrhreseseshoooldldld-cucucurrrrrenenent t t prememmatatatururure ee ststtimimimulululususus

SSS2)2)2) decremenenteed in 22-mmms inteteterrrvals. EEffffeeectivveve refrrraccctorrry y y perrriooods (EEERPs))) wwewerrere dddefinnededd ass ttthe

ongest S1S1S1-S2 coooupupplingngng--inii tervalalal that failed tooo cacaaptptptururure. Veeentnn ricular tatatachycyy arrrdidd a (VVVT))) was defined

asasas 11100 0 ororor mmmorororee e sususuccccccesesessisis veveve ssspopopontntntanananeoeoeoususus vvvenenentrtrtriciccululu ararar cococompmpmpleleexexexesss... OpOpOptititicacacall mamamappppppiningg g stststudududieieies s s wewewererere


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

8

threshold current square-wave pulses. After 5 minutes of pacing at 0.5 Hz for equilibration,

preparations were paced successively at increasing frequencies from 0.5 to 10 Hz, with 200 beats

at each frequency to ensure steady state prior to recording. APD values were based on the

average of five consecutive APs at three different sites in the LV free-wall. To obtain data about

the time-course of AP-changes, APs were recorded under current-clamp with perforated-patch

methods from cells isolated from hearts of 5 and 10 week-old mice. For perforated-patch

recording, nystatin-free intracellular solution was placed in the tip of the pipette, and then

pipettes were backfilled with nystatin-

Ventricular-cardiomyocyte Isolation

Adult male mice (10-week) were heparinized (100 IU, intraperitoneal), anaesthetized with

2.5%-isoflurane and sacrificed by cervical dislocation. The hearts were retrogradely perfused

through the aorta on a modified Langendorff apparatus and cells isolated with collagenase type 2

(for details, see Data Supplement). After digestion, single LV-cardiomyocytes obtained by

trituration, then placed into storage-solution at 4°C.

Ion-current Recordings and Analysis

IK1, IK,Total and ICa,L were recorded at 37°C with tight-seal patch-clamp in voltage-clamp mode.

Cell-capacitances were obtained with 5-mV, 10-ms hyperpolarizing steps from-60 mV. For

recording conditions of specific currents, see Data Supplement. Peak transient-outward current

(Ipeak) was defined as the maximal outward K+-current. The amplitude of Ito,f, IK,slow and Iss were

determined from exponential fits to the decay-phases of the total outward K+-current as

previously described.13,14 Currents are expressed as densities (pA/pF).

Data Analysis

Normality of distribution was assessed via Shapiro-Wilk test. Normally-distributed data are

Adult male mice (10-week) were heparinized (100 IU, intraperitoneal), anaesthetttizizededed wiwiwiththth

2.5%-iisoflurana e anand sacrificed by cervical dislocatit ono . The hearts were reretrogradely perfused

hhhrooough the aoaoaortrtrtaaa onoo aaa mmmodododififi ieeed d d LaLaLangngngenenendodd rfrfrff f f apaa parararattus ananand d d cecec lls ss iiisolololataa ededed wiww thh cccololollalalagegeg nananasesese tttypypype ee 2

fofofor r ded tails, seee DDDaaata SSuSuppplememement). AfAfA ttterrr digggr eesstionnn, singngnglelele LVLVLV-caaardddiomyyyocccytyttesess obtaaiainneneddd bbby

rituratioonnn, , , then plplplacededd iiintnn o stoorararagegg -solution atata 444°CCC.

ooon-nn cucucurrrrrrenenenttt ReReRecococordrdrdininingsgsgs aaandndnd AAAnananalylylysisisisss


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

9

expressed as mean±SEM and analyzed by Student’s nonpaired t-test (for comparisons involving

only CTL vs ILK-KO group-comparisons without other factors or repeated measures) or by

ANOVA; data not satisfying normal-distribution criteria are shown as median and 25%-75%

percentile and were analyzed by Mann-Whitney U-test. For work involving multiple sample-

measures from thesame animal (e.g. multiple cell-impalements, multiple cells from the same

mouse available for patch-clamp work; numbers are provided in figures as n/N for cells/mice),

data were analyzed using a multi-level mixed-effects model to take into account correlation

between multiple levels of within-mouse measurements. In this model, the mouse effect and the

interaction “cell×mouse” effect were considered as random factors. According to the design of

the experiments, fixed factors for repeated measures (e.g. voltages, frequencies, weeks) were

added to the model to produce multi-level repeated measures mixed-effects models. Two-way

ANOVA on log-transformed data was used to compare control versus ILK-KO mice across time

(5 and 10 weeks) for PCR parameters. Bonferroni correction was used to take into account

multiple statistical tests to identify specific mean-differences (P-value from t-test multiplied by

nc=number of comparisons). Fisher’s exact test was used to compare differences in occurrence-

rates. A two-tailed P-value<0.05 denoted statistically-significant differences.

Results

Premature Death and DCM Phenotype in ILK-KO Mice

ILK-KO mice began to die prematurely at 5 weeks, and by 18 weeks all 30 ILK mice (vs no

controls) had died, with a median age at death of 10 weeks (Figure 1A). Detailed

characterization was obtained at the median time for death (10 weeks). In addition, selected

experiments were repeated in 5 week-old mice, to assess the changes that preceded the onset of

manifest cardiomyopathic changes. After 5 weeks, ILK-KO mice began to exhibit signs of

he experiments, fixed factors for repeated measures (e.g. voltages, frequencies, weweweekekeks)s)) wwwererere ee

added to the moddele to produce multi-level repeatedd mmeasures mixed-effeectc s models. Two-way

ANANANOOOVA on lllogogog-t-t-trarr nsnssfofoformrmrmedede dddatatata a a wawaw sss usususeddd tttoo o compmpmpara e e cococontntntroror l veveversrssususu IIILKLKLK-KOOO mmmicicice e e acaccrororossssss tititimemm

555 aandnn 10 weeke s)) fffor PPPCCCR paaaraamametererrs.s BBBonffferrrroniii cccorrererectctctioioon wasss uuused ttto tatakkeke into acacaccocoununnt

multiple ssstatt tisticccalaa tessstststs ttto idenntititifyyy spep cific memem ananan-dididifff erenncecc s (((P-valululue frommm ttt-teststst multiplpp ied by

ncncnc=nnnumumumbebeberr r ofofo cccomomompapaparirisososonsnsns)))... FFFisissheheer’rr sss exexexacacactt t teteteststst wwwasasas uuusesesedd d tototo cccomomompapaparerere dddifififfefeerererencncnceseses iinn n ocococcucucurrrrrrenenencecece-


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

10

cardiac failure, including dyspnea, weakness, disorientation and withdrawn behavior. However,

mortality was apparently sudden, with no prior progressive deterioration. Cardiac chambers

were dilated in 10-week ILK-KO mice, with thinning and fibrosis of the superior septal and free-

wall regions of the LV, along with severe dilation and intracavitary thrombus in the left atrium

(Figure 1B). The ventricular-weight/tibia-length ratio was significantly increased (Figure 1C,

Table 1). The hypertrophic biomarkers ANP and BNP were substantially increased (Figure 1D).

Structural and Hemodynamic Remodeling

ILK-KO produced substantial changes in cardiac structure and function by 10 weeks (Table 1).

Ventricular and atrial weights were increased and heart rate was significantly increased. A wide

range of indices indicated LV systolic/diastolic dysfunction, along with RV dilation and reduced

global cardiac performance, in 10-week ILK-KO mice (Table 1). Interestingly, at 5 weeks there

were virtually no differences from control to ILK-KO mice in cardiac dimensions or functional

indices.

Figure 2A illustrates regional changes in connective-tissue composition and distribution

on Masson’s-trichrome staining in 10 week-ILK mice. Quantitative analysis indicates

substantial, regionally-selective LV-fibrosis, with the superior LV free-wall being particularly

affected (Figure 2B). The mRNA-expression of fibrosis-related genes in the LV free-wall was

significantly increased (Figure 2C), with increases in vimentin (3.3-fold) and -SMA (2.6-fold)

and a dramatic 70-fold increase in pro-collagen -1 type-1. Immunostaining showed marked

lateralization of connexin43 (Figure 2D). Connexin-43 expression was quantified by

immunoblot, which confirmed significant downregulation in ILK-KO (Figure 2E).

In vivo Electrophysiological Abnormalities

Representative ECG-recordings at 10 weeks are shown in Figure 3A. QRS-durations and

ange of indices indicated LV systolic/diastolic dysfunction, along with RV dilatititionnn aandndnd rrredededucuu ed

globall cardiaca ppererformance, in 10-week ILK-KO mmici e (Table 1). Intereststini gly, at 5 weeks there

wwwererere virtuallylyy nnnooo didd ffffferererenenencececes frfrfromomom cccononntrtrtroloo ttto o o ILILILK--KKOKO mmmiciciceee iiin nn caaarrdrdiaiaiac cc dididimemem nssioioionsnsns ooor r r fuuuncncnctititionononalalal

nnndidiicecc s.

FiFiigugugure 22A AA illuuststs rarar tes regegegioii nal changegees s s innn ccconononnectivvvee-tissue cococompppositttioii n ananand distribution

ononon MMMasasassososon’nn s-ss trtrtricicchrhrhromomomee e stststaiaia niningngng iinn n 101010 wwweeeeeek-k-ILILILKK K mimicecece... QuQuQuananantitit tatatatititiveveve aaanananalylyysisis ss s inindidid cacacatetetesss


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

11

corrected QT intervals were significantly prolonged in ILK-KO mice, and RR-interval was

decreased (Table 2A). P-wave duration and PR-interval were unchanged. AH and HV intervals

were not significantly altered (Table 2B), suggesting that QRS-prolongation was due to slowing

in intraventricular, as opposed to conduction-system, impulse-propagation. Sinus-node and

AV-node function were also unaltered. While atrial effective refractory period (ERP) was

unchanged, ventricular ERPs were significantly greater, by -KO. Consistent with

their lack of structural/functional remodeling, 5 week ILK-mice showed no statistically-

significant changes in ECG-intervals or in vivo electrophysiological parameters (Table 2).

To further assess the conduction changes suggested by the QRS-interval prolongation in

10-week ILK-KO mice, we used optical mapping to analyze ventricular impulse-propagation.

As illustrated in Figure 4, ILK-KO appreciably slowed conduction, with statistically-significant

decreases in both longitudinal and transverse conduction velocity of similar magnitude (25% and

21% respectively).

Programmed stimulation induced VTs (Figure 3B), with a mean duration of 626±28

milliseconds, in 10-week ILK-KO mice, but not controls. Twenty-four hour telemetry at 10

weeks of age revealed frequent VPCs in ILK-KO mice (1362±669/24 hr) and none in controls.

No arrhythmias were seen in 5 week-old mice. Spontaneous VTs were common in ILK-KO

mice (Figure 3C), with an overall occurrence-rate of 60% (Figure 3D) and a mean duration of

2.4±0.2 seconds. No spontaneous ventricular arrhythmias were seen in control-mice. A

substrate for VT was revealed by inducible arrhythmias in 6/7 ILK-KO mice (85.7%), versus 0/8

controls (Figure 3E).

Cellular Basis for Electrophysiological Abnormalities

Representative standard floating-microelectrode ventricular AP-recordings are shown in Figures

10-week ILK-KO mice, we used optical mapping to analyze ventricular impulse---prrropopopagagagatatatioioion.nn

As illusu tratedd in FiF gure 4, ILK-KO appreciably slowowed conduction, withh sstatistically-significant

dededecrrreases in bbbototothhh loll ngngngitititudududinininall aaandndnd tttrarar nsnsnsvevv rsrsse e e cocc nddduccctiononn vvvelelelocoo itttyyy ofofof sssimmmilililaraa mmmagagagnininitututudedee (((252525%% % ananand

21211%% % respectivev lyyy).

PPrororogrgg ammmmemm d stststimimmulatioon nn induced VTTTsss (FFFigigigururu e 3B))),, , with a mmmeaee n durararatit onnn of 626±28

mimillllisissecececononondsdsds,, , inin 1110-0-0-weweweekeke ILILILKKK-KOKOKO mimicecece,, , bububutt t nononott t cococontntntrororolslss... TwTwTwenenentytyty-ffououour r r hohooururur tttelele emememetetetryryry aaat t t 101010


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

12

4A and B. Control recordings showed no spontaneous arrhythmia, whereas 10-week ILK-KO

mouse-APs showed periods of rapid firing compatible with triggered activity (Figure 5A). ILK-

KO APs had slower early and terminal repolarization than controls (Figure 5B). Durations of all

phases of the AP were significantly prolonged in ILK-KO (Figure 5C). Because of the profound

changes observed at 10 weeks, we performed perforated-patch AP-recording in a series of mice

at 5 weeks of age as well as 10-week ILK-mice and age-matched controls. Examples of such

recordings are shown in Supplemental Figure 2A-B. Single-cell APD was prolonged at both

5 weeks and 10 weeks (Figure 5E). For some single-cell recordings, phase 0 overlapped with the

stimulus artefact and obscured the upstroke. However, this was not the case for many cells, in

which phase-0 upstroke velocity (Vmax) could be analyzed. In the single-cell recordings, resting

membrane potential, AP-amplitude and Vmax were significantly reduced in 10-week mice but

not at 5 weeks (Supplemental Figures 2C-E). These results suggest that phase-0 Na+-current

may be reduced in 10-week ILK-KO mice. Since SCN5a mRNA-expression wasn’t altered in

ILK-KO mice (Supplemental Table 1), this finding suggests that post-transcriptional

mechanisms are involved.

To evaluate mechanisms underlying AP-prolongation in ILK-KO mice, we performed

whole-cell voltage-clamp recordings of ionic currents governing repolarization. Supplemental

Figure 3 shows original K+- and L-type Ca2+- (ICaL) current recordings. Voltage-gated K+-

currents were reduced in ILK-KO, with changes already present at 5 weeks (Figure 6A) and

continuing to 10 weeks (Figure 6B). Kinetic analyses revealed that in addition to total (peak)

current, 2 of the 3 kinetically-dissociable components reported in WT-mice13 are downregulated

by ILK-KO: Ito,f and IK,slow. The sustained component Iss was not altered. Quantitative-PCR

confirmed significant changes in subunit-expression corresponding to the ionic-current changes

which phase-0 upstroke velocity (Vmax) could be analyzed. In the single-cell reececororordididingngngs,s,s, rerere tsting

membbrane ppoto enntitial, AP-amplitude and Vmax werere significantly reduceded in 10-week mice but

nononot aaat 5 weekkksss (((SuSuSupppplelelememementntntalal FFFigigigururureseses 222C-CC E)E)E).. Theeseee resssululultststs sssuguu gegegeststt ttthahahattt phphphase-e-e-0 0 0 NaNaNa+-cucucurrrrrrenenentt t

mmamay y y beb reduccede innn 10---wweweek IIILKLKLK-KOOO mmmiccce. SSSinnnce SCSCSCN5N5N5aaa mRmRRNNANA---exxxpressssiononn wwwasa n’t t t alallteterered innn

LK-KO mimm ce (((SuSuS ppppppleeememm ntal TTTabaa le 1),),), this fififindndndinining g g sus ggggggessststt that popopostss -transcccrirr ptpp ioioional

mememechchc anananisissmsmsms aaarerere iinvnvnvolololveveved.d.d.


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

13

at 10 weeks: - -subunits corresponding to Ito,f,13,14 were

significantly reduced, by 55 and 66% respectively (Supplemental Figure 4A). Kv4.3-expression

was not altered. Kv1.5, the subunit believed to underlie IK,slow,13 was also downregulated, by

30%. Ito,s, a component with kinetics intermediate between Ito,f and IK,slow, is difficult to detect in

native myocytes but is likely encoded by Kv1.4.13 We found that Kv1.4 was expressed at very

low levels that were significantly upregulated by ILK-KO. Two other subunits, Kv1.2 and

Kv2.1, were unaltered. Interestingly, despite the significant reductions in Ito,f and IK,slow at 5

weeks, the corresponding channel-subunit expression was unchanged (Supplemental Figure 4A).

Changes in inward-rectifier background current (IK1) are shown in Figure 6: ILK-KO did

not affect the current at 5 weeks (Figure 7A), but significantly reduced it at 10 weeks

(Figure 7B). All 3 subunits implicated in cardiac IK1 (Kir2.1-3) were downregulated at 10 weeks

in ILK-KO but unaffected at 5 weeks (Supplemental Figure 4B). Similarly, ICa,L was unaffected

at 5 weeks (Figure 7C) but significantly reduced at 10 weeks in ILK-KO (Figure 7B).

-subunit in the ventricles, Cav1.2, was downregulated at 10 weeks

and unaffected at 5 weeks (Supplemental Figure 4C). The atrial-predominant subunit Cav1.3

was expressed at very low levels and upregulated in 10-week ILK-KO.

Gene-expression Changes

Non-directed hierarchical clustering showed discrete patterns of gene-expression between 10-

week ILK-KO and age-matched control-mice (Supplemental Figure 1). ILK-KO was associated

with upregulation of multiple structural/adaptor (e.g. -parvin, integrin, laminin, -SMA,

vimentin, procollagen, E-cadherin, vinculin) and remodeling (ANP, BNP) genes, with

downregulation of ion-channel subunit-genes (Kir2.1, Kir2.2, Kir3.1, Kir 3.4, Kir6.1, Kir6.2,

Kv1.5, Kv4.2, Cav3.1, Cx43, KChIP2, SUR2). Complete details of gene-expression analyses are

not affect the current at 5 weeks (Figure 7A), but significantly reduced it at 10 wwweeeeksksk

Figurre 7B).) All l 33 subunits implicated in cardiac IIK1K1K (Kir2.1-3) were doownw regulated at 10 weeks

nnn ILLLK-KO bbbuutut uuunananafffffececectetetedd d atata 55 wwweeeeeeksksk (S(S(Supuu plplplememementttalll Fiigugugurerere 4B44 ).).). SiSiSimimim lalalarlrlrly,y ICaCaCa,L,LL wawawas s unununafafaffefefectctctedede

atatt 5 weww eks (FFigi uuureee 7C)) bbbut siiigngnnificananntlyyy rrredududuccceddd aaat t 1000 wwweeksss innn d ILILLKKK-KOOO (F(F(Figgguuure 777BBB).

-s-ssubu unit in the e e veveventntntriririclcc es,, CaCaCav1.2,, waaasss downrereregugg laaatet d at 10 weeks

ananandd d unununafafa fefeectctctededed aaatt t 555 weweweekeke sss (S(S(Supupupplplp emememenenentatatall FiFigugugurerere 444C)C)C).. . TTThehee aaatrtrtriaiaall-p-p-prereredododomimiminananantntnt sssubububunununititit CCCavavav1.1.1.333


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

14

provided in Supplementary Table 1. Pathognomonic changes typical for heart-failure paradigms

were seen: highly-significant increases in ANP and BNP, and reciprocal

- -MHC. Profibrotic fibroblast-differentiation was

indicated by 3- -SMA, along with a 17-fold increase in

procollagen-1 expression. In addition to the ion-channel expression changes described above,

the Ca2+-dependent K+-channel subunits SK1 and SK2, the acetylcholine-dependent K+-channel

subunits Kir3.1 and 3.4 and the ATP-dependent K+-channel subunits Kir6.1 and 6.2 were

downregulated. Finally, the Ca2+-handling subunits phospholamban, SERCA2a and RyR2, were

significantly downregulated.

Conventional qPCR was used to compare collagen (Col1a1), ANP and BNP mRNA

expression directly in 5 versus 10 week-old mice. The results (Supplemental Figure 5) show

statistically-significant changes in the expression of all 3 genes at 10 weeks, and of BNP and

collagen at 5 weeks. However, for BNP and collagen the magnitudes of increases were

significantly larger at 10 weeks than at 5 weeks: for Col1a1, 221% increase at 5 versus 1,448%

increase at 10 weeks (group×time interaction P<0.0001); for BNP, 118% increase at 5 versus

317% increase at 10 weeks (group×time interaction P=0.010).

ILK co-immunoprecipitation

We noted changes in voltage-dependent K+-currents before any overt cardiomyopathy. We

therefore examined whether ILK might interact physically with relevant cardiac ion-channel

subunits. Co-immunoprecipitation demonstrated selective pulldown of ILK, Kv4.1 and Kv4.2-

subunits with antibodies directed to ILK (Supplemental Figure 6), suggesting physical

interactions between ILK and specific voltage-dependent K+-channel subunits. No interaction

was demonstrable in ILK-KO ventricles.

Conventional qPCR was used to compare collagen (Col1a1), ANP and BNNNP PP mRmRmRNANANA

expressssisisiononon diririrecctltltly yy in 5 versus 10 week-old mice. ThTT e results (Supplemmmenenental Figure 5) show

tttatiiistically-ssigigignininififificaaantntnt ccchahahanngngeseses iiinn n thththe e exexe prprp esesessisision ooof f all 333 gegegenenenes atatat 1110 00 weweweekekeks, aaandndnd ooof f f BNBNBNP P P ananand d d

cocoollllagagagen at 5 weww ekekeks. HHHooowevvvererr,,, for BNBNB P ana ddd ccoollaggeeen thththeee mmmagggnitudududes offf innncrrreaeaases wewewerrre

ignificananntltlt y y y lalargrgrgerer aat t 101010 wweeeeksksks ttthahan atat 55 wweeeeeeksksk ::: fofofor r r CoCol1l1a1a1a , , , 22221%1%% iiincncn rereasasee atatat 55 vvverersususs 1,1,,444 8%

nnncrcrcreaeaeasesese aaattt 101010 wwweeeeeeksksks (((grgrgrouououp×p×p×tititimememe iiintntnterereracacactititiononon PPP<0<0<0 0.0.00000001)1)1);;; fofoforrr BNBNBNPPP,, 111111888% %% ininincrcrcreaeaeasesese aaattt 5 55 veveversrsrsususus


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

15

Discussion

In this study, we investigated for the first time the arrhythmogenic cardiac phenotype produced

by cardiomyocyte-targeted ablation of ILK in the mouse. The arrhythmogenic substrate included

both repolarization-deficiency due to K+-channel dysfunction and conduction abnormalities in

the presence of structural remodeling (cardiac fibrosis and connexin remodeling).

ILK-dysfunction Cardiomyopathy

Two prior studies have described dilated cardiomyopathies with genetic deficits in ILK, one in

humans9 and one in the same mouse model used here.8 Phenotypic characterization has been

very limited: increased mortality, cardiac hypertrophy, ventricular dilation and reduced LVEF

were described in mice with targeted ILK-deletion,8 whereas reduced integrin-binding capacity

and loss of endothelial cells were reported with clinical ILK-mutation associated dilated

cardiomyopathy.9

The present study constitutes the first detailed phenotypic characterization of the

cardiomyopathy caused by ILK-deletion. A schematic representation of our findings is provided

in Figure 8. We provide detailed information about cardiac function abnormalities, structural

remodeling, and in particular arrhythmic/electrophysiological disturbances. Two important sets

of candidate-mechanisms contributing to ILK-KO mouse arrhythmias were identified:

1) structural remodeling, including downregulation/lateralization of Cx43 and regionally-

distributed tissue-fibrosis and 2) electrophysiological remodeling causing APD-prolongation

through K+-current downregulation. The high sudden-death mortality-rate was associated with

frequent ventricular tachyarrhythmias, likely attributable to profound abnormalities in ion-

channel subunit expression and ion-current function that caused important APD-prolongation, as

well as to tissue-fibrosis, connexin-dysregulation and possibly Na+-current downregulation that

were described in mice with targeted ILK-deletion,8 whereas reduced integrin-biindndndinininggg cacacapapapacicicityt

and looss of enndoththelial cells were reported with clinnici al ILK-mutation assssociated dilated

cacaardddioi myopatatathyhyhy.9

The ppprer seeennnt stuuuddydy connnstststitutesss thhhe firsttt dddetaiiileeed phphpheeenoootyyypiccc cchharacccteererizzzatattioii n offf tththeee

cardiomyyyopopopathyhyhy ccauseeed d d bybyby ILKKK---deletion. A ssschhhemememataa ic repepeprer sentatiooon n of oururr findidid ngggs is ppprovided

nn FFFigiggurururee e 8.8.8. WeWeWe ppprororoviviv dedede dddetetetaiaia leleedd d ininfofoormrmrmatatatioioonn n abababouououtt t cacacardrddiaiaacc c fufuuncncnctititiononon aaabnbnbnororormamamalilititit eseses,,, stststruruructctcturururalala


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

16

were associated with significant conduction-slowing.

Functional Importance of ILK

ILK is a crucial component of a macromolecular complex, including ILK, melusin, PINCH-1

-parvin, which links the extracellular matrix to the contractile machinery at

costameres.7,15 ILK-functions include both mechanosensitive properties16 and action as a

kinase.8 Cardioprotective protein-kinase B (PKB) is an important target of ILK-phosphorylation:

the cardiomyopathy resulting from loss of ILK-phosphorylation in PINCH-inactivated zebrafish

is reversed by overexpression of constitutively-activated PKB.17 ILK is upstream to a wide

range of protein-kinases involved in cell-growth and cardiac hypertrophy, including PKB,

glycogen-synthase kinase-3- -related kinase (ERK) and

mammalian-target of rapamycin (mTOR),2 and ILK-deletion causes severe cardiomyocyte-

apoptosis.18 ILK is upregulated in human cardiac hypertrophy caused by left-ventricular outflow

obstruction, along with a wide range of hypertrophic genes.4 ILK gene-transfer improves cardiac

function in post-MI and doxorubicin-induced cardiomyopathies.5,6 Thus, ILK is a central protein

in cardiac function and adaptive responses to a range of stressors, particularly mechanical.

Are there Specific Electrophysiological Features of the Arrhythmogenic Cardiomyopathy

Caused by ILK-depletion?

We observed downregulation of a wide range of ion-channel subunits and currents in ILK-KO

mice. The downregulation of K+-currents and corresponding subunits (Kv1.5, Kv4.2, KChIP2,

Kir2.1-2.3) likely explains the substantial APD-prolongation, associated with triggered activity,

that we saw at 10 weeks. One important issue is whether any of the ion-channel changes we

observed specifically result from ILK-depletion or whether they are all secondary to the

cardiomyopathic state. Ion-current remodeling typical of a wide variety of heart-failure

glycogen-synthase kinase-3- -related kinase (ERRRK)K)K) aaandndnd

mammmalian-tatargggetete of rapamycin (mTOR),2 and ILKK-deletion causes sevverere cardiomyocyte-

apapapopopptosis.18 ILILILKK K isii uuuprprpregegegulululatatededed iiin n n huhuhumamaman nn cacacardrdrdiac hyyypeertrtrtrororophphphy y cacacausssededed bbbyy y lell ft-v-vvenenentrtrtriciciculllararar oooutututflflflowoo

obobbstttrurr ction, aalol nnng withhh aaa widdde e rar nge e e ofoff hhhypy eeertrtrrophhhiccc gegeenennes.444 ILKKK gggene--ttraananssfsfeerer imppprorooveees cardddiiaac

function iiin nn popp st---MIMM andndnd doxorubububicin-induceddd cararardididiomoo yoy papapathies.5,6 ThTT us, , ILLLK KK isss aa central pprotein

nn cccararardidid acacac ffunununctctctioioonn n ananandd d adadadapapaptitit veveve rrresesespopoponsnsnseseses tttoo o aa a rararangngngee e ofofo ssstrtrtresesessososorsrsrs,,, papapartrtrticicicululu arararlylyy mmmececechahahanininicacacal.l..


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

17

paradigms includes downregulation of Ito, IK1 and ICa,L.19 Changes in voltage-gated K+-channels

occurred at 5 weeks, preceding most changes in cardiac structure or function and suggesting that

these alterations are not secondary to the cardiomyopathic phenotype. The co-

immunoprecipitation of Kv4.2-subunits with ILK suggests that ILK may contribute to the

stabilization or membrane insertion of K+-channels, a potentially-interesting subject for future

investigation. There is evidence that Ito may contribute to electromechanical coupling, and that

its dysfunction may cause contractile abnormalities,20 so the early Ito-changes we observed may

in fact have contributed to the development of cardiomyopathy. Interestingly, there were

virtually no transcriptional changes in ion-channel subunit expression at 5 weeks, in contrast to

the marked changes at 10 weeks. This observation suggests that the widespread current-changes

at 10 weeks are secondary to cardiomyopathy, whereas the changes in specific voltage-

dependent K+-currents at 5 weeks may be caused by loss of direct interactions with ILK due to

ILK-ablation in KO-mice, as illustrated by our co-immunoprecipitation data (Supplemental

Figure 6).

In addition to APD-prolongation, we noted connexin-43 downregulation and

lateralization, extensive regionally-variable fibrosis and QRS-prolongation indicating ventricular

conduction-slowing. It is likely that the conduction-changes caused by fibrosis and connexin-

dysfunction contributed to the vulnerability to VT-induction, likely due to intracardiac reentry,

that we observed. Connexin-43 downregulation and lateralization, along with increased fibrous-

tissue content, are typical of end-stage heart failure in man and cause significant conduction-

abnormalities.21 One interesting aspect of the phenotype we noticed was the regional distribution

of tissue-fibrosis, which was restricted to the basal regions of the heart, where it was quite

marked (Figure 2B). This basal localization of fibrosis has not, to our knowledge, been

he marked changes at 10 weeks. This observation suggests that the widespread cucuurrrr enenenttt---chchchanananges

at 10 wew eks ara e sesecondary to cardiomyopathy, wherereae s the changes in spepep cific voltage-

dededepepeendent KK+++-cucucurrrrrrenntststs aaat t t 555 weweweekekeks s s mamam y y y bebebe cccauauausess d bybyby lossssss ofofof dddiri ececctt ininintetet rararactctctioi nss wiwiwiththth IIILKLKK dududuee e tototo

LLLKKK-a- blation in KKKOOO-mmmiccce, asss illlll ustratatatededd bbby oouurrr co---immmmmmunununopppreeecipipiitaaation ddaaata a (S(SSupplemememennntaaal

Figure 6).).).

IIInn n adadaddidid titit ononon ttto oo APAPAPDDD-ppprororololoongngngatatatioioon,n,n, wwwee e nononotetetedd d cococonnnnnnexexexinin-444333 dododownwnwnrereregugugulalaatitit ononon aaandndd


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

18

previously reported in other animal models of cardiomyopathy. However, it is characteristic of

patients demonstrating monomorphic VTs associated with nonischemic cardiomyopathies.22,23 A

similar distribution of fibrosis has recently been described with the use of late gadolinium-

enhancement magnetic-resonance imaging in patients with lamin A/C (LMNA)-mutation

cardiomyopathy.24 This regional localization may be related to the effects of local mechanical

stresses, which might have particularly important effects when part of the mechanosensor

apparatus, like ILK, is deleted.

Potential Limitations

The mckCRE-lox system that we used for ILK-deletion creates a myocyte-specific knockout.

ILK is important for fibroblast function, and fibroblast-selective ILK-deletion prevents

myofibroblast differentiation.25 Our studies therefore apply specifically to the result of ILK-

deficiency in cardiomyocytes. Complete loss of ILK-function in all cell-types might produce a

different phenotype, particularly with respect to fibrotic changes.

We recorded ECGs and performed electrophysiological studies under isoflurane

anesthesia. Isoflurane has potential electrophysiological actions and could have affected the

results obtained. This limitation does not apply to the in vitro data, such as the AP and ionic

current recording results. We used Bazett’s formula to correct QT intervals for heart-rate

changes, as previously reported.12 This formula may not provide optimal correction, which may

explain our ability to identify statistically-significant in vivo QT/QTc-changes in 5-week ILK-

mice, despite statistically-significant APD-increases (albeit less than at 10 weeks). Care must be

taken in extrapolating results from animal models to man. While mouse-models are recognized

to be of value in studying the molecular basis of clinical arrhythmia-syndromes, some of their

ion-current systems may differ substantially from what is seen in humans.26

LK is important for fibroblast function, and fibroblast-selective ILK-deletion prreevevenene tststs

myofibbroblasst diifffferentiation.25 Our studies therefforore apply specificallyyy tto o the result of ILK-

dededefiiicciency innn ccararardididiommmyoyoyocycycytetetes.. CCComomomplplpletetete ee looossssss ooof ILLLKKK-fuuuncncnctititiononon in nn alalll ll cececellllll-t-t-typyy esss mmmigigighththt pprororodududucecece aaa

didiifffffeererent phennotypypype, ppararrtticulaaarlrlrly y withthth rressspep cctt ttto fibrbrbroto icicic ccchahahanggges.

WeWeWe recororrded d ECECECGsG and d d pepeperformed elllecee trtrropopophyhyhysiologogogical studididies undererer isoooflff urane

anananesesesththt esesesiaiaa... IIIsososoflflurururanananee e hahaasss popopotetetentntntiaiaall elele ececectrtrtropopophyhyysisis ololo ogogogiciccalala aaactctctioioonsnsns aaandndd cccouououldldd hhavavavee e afafa fefeectctctededed ttthehee


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

19

We did not observe in vivo arrhythmias or mortality in 5-week ILK-KO mice and no

EADs were noted during single-cell AP recording. However, we did not perform recordings on

multicellular preparations from 5-week ILK-KO mice and therefore cannot completely exclude

EAD-susceptibility at 5 weeks.

Acknowledgments: The authors thank Nathalie L'Heureux, Chantal St. Cyr, Marc-Antoine

Gillis, Louis Villeneuve and Marie-Elaine Clavet for technical support, Annik Fortier for

biostatistical expertise and analysis and France Thériault for secretarial help with the manuscript.

Funding Sources: Supported by Canadian Institutes of Health Research (MOP68929), the

Quebec Heart and Stroke Foundation, and Fondation Leducq (ENAFRA, 07CVD03).

Conflict of Interest Disclosures: None.

References

1. Hannigan GE, Leung-Hagesteijn C, Fitz-Gibbon L, Coppolino MG, Radeva G, Filmus J, Bell JC, Dedhar S. Regulation of cell adhesion and anchorage-dependent growth by a new beta 1-integrin-linked protein kinase. Nature. 1996;379:91-96.

2. Bendig G, Grimmler M, Huttner IG, Wessels G, Dahme T, Just S, Trano N, Katus HA, Fishman MC, Rottbauer W. Integrin-linked kinase, a novel component of the cardiac mechanical stretch sensor, controls contractility in the zebrafish heart. Genes Dev. 2006;20:2361-2372.

3. Hannigan G, Troussard AA, Dedhar S. Integrin-linked kinase: A cancer therapeutic target unique among its ilk. Nat Rev Cancer. 2005;5:51-63.

4. Lu H, Fedak PW, Dai X, Du C, Zhou YQ, Henkelman M, Mongroo PS, Lau A, Yamabi H, Hinek A, Husain M, Hannigan G, Coles JG. Integrin-linked kinase expression is elevated in human cardiac hypertrophy and induces hypertrophy in transgenic mice. Circulation.2006;114:2271-2279.

5. Ding L, Dong L, Chen X, Zhang L, Xu X, Ferro A, Xu B. Increased expression of integrin-linked kinase attenuates left ventricular remodeling and improves cardiac function after myocardial infarction. Circulation. 2009;120:764-773.

Quebec Heart and Stroke Foundation, and Fondation Leducq (ENAFRA, 07CVD003)3)3).

Confliictct oof InI teerereest Disclosures: None.

RReR ffeferences

1. HHHananannininigagagan n n GEGEGE,,, LeLeLeununung-g-g-HaHH geeeststs eiieijnjnjn CC, FiFiFitztztz-G-G-Gibibibbobobonnn L,L,L, CCCopopoppopopolililinonon MMMG,GG RRRadddeveve a aa G,G,G, FiFiFilmlmlmuuus JJJ,,, BBBellllll JC, Dedhhharara S. ReReRegugg laatititionono of ceellllll adhesion annnd dd anananchchchoro aggge---dedd pepp ndennnt t t grgg owthhh bby yy a a a new beta 1-ntegrinn-l-l-linininkekeked d d prprprotototeieiein n n kikikinananasesese. NaNaNatututureree.. 191919969696;3;3;37999:9:9:91-1--969696. .

22 Bendig G Grimmler M Huttner IG Wessels G Dahme T Just S Trano N Katus HA


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

20

6. Gu R, Bai J, Ling L, Ding L, Zhang N, Ye J, Ferro A, Xu B. Increased expression of integrin-linked kinase improves cardiac function and decreases mortality in dilated cardiomyopathy model of rats. PLoS One. 2012;7:e31279.

7. Sopko N, Qin Y, Finan A, Dadabayev A, Chigurupati S, Qin J, Penn MS, Gupta S. Significance of thymosin beta4 and implication of pinch-1-ilk-alpha-parvin (pip) complex in human dilated cardiomyopathy. PLoS One. 2011;6:e20184.

8. White DE, Coutu P, Shi YF, Tardif JC, Nattel S, St Arnaud R, Dedhar S, Muller WJ. Targeted ablation of ilk from the murine heart results in dilated cardiomyopathy and spontaneous heart failure. Genes Dev. 2006;20:2355-2360.

9. Knoll R, Postel R, Wang J, Kratzner R, Hennecke G, Vacaru AM, Vakeel P, Schubert C, Murthy K, Rana BK, Kube D, Knoll G, Schafer K, Hayashi T, Holm T, Kimura A, Schork N, Toliat MR, Nurnberg P, Schultheiss HP, Schaper W, Schaper J, Bos E, Den Hertog J, van Eeden FJ, Peters PJ, Hasenfuss G, Chien KR, Bakkers J. Laminin-alpha4 and integrin-linked kinase mutations cause human cardiomyopathy via simultaneous defects in cardiomyocytes and endothelial cells. Circulation. 2007;116:515-525.

10. Burstein B, Comtois P, Michael G, Nishida K, Villeneuve L, Yeh YH, Nattel S. Changes in connexin expression and the atrial fibrillation substrate in congestive heart failure. Circ Res.2009;105:1213-1222.

11. Lande G, Kyndt F, Baro I, Chabannes D, Boisseau P, Pony JC, Escande D, Le Marec H. Dynamic analysis of the qt interval in long qt1 syndrome patients with a normal phenotype. Eur Heart J. 2001;22:410-422.

12. Mitchell GF, Jeron A, Koren G. Measurement of heart rate and Q-T interval in the conscious mouse. Am J Physiol. 1998;274:H747-751.

13. Xu H, Guo W, Nerbonne JM. Four kinetically distinct depolarization-activated K+ currents in adult mouse ventricular myocytes. J Gen Physiol. 1999;113:661-678.

14. Kuo HC, Cheng CF, Clark RB, Lin JJ, Lin JL, Hoshijima M, Nguyêñ-Trân VT, Gu Y, Ikeda Y, Chu PH, Ross J, Giles WR, Chien KR. A defect in the Kv channel-interacting protein 2 (KChIP2) gene leads to a complete loss of I(to) and confers susceptibility to ventricular tachycardia. Cell. 2001;107:801-813.

15. Hannigan GE, Coles JG, Dedhar S. Integrin-linked kinase at the heart of cardiac contractility, repair, and disease. Circ Res. 2007;100:1408-1414.

16. Bendig G, Grimmler M, Huttner IG, Wessels G, Dahme T, Just S, Trano N, Katus HA, Fishman MC, Rottbauer W. Integrin-linked kinase, a novel component of the cardiac mechanical stretch sensor, controls contractility in the zebrafish heart. Genes Dev. 2006;20:2361-2372.

endothelial cells. Circulation. 2007;116:515-525.

10. Burstein B, Comtois P, Michael G, Nishida K, Villeneuve L, Yeh YH, Nattel SSS. CChahahangngngeseses iiinnn connexxin expxppresssioi n and the atrial fibrillation subsstrtrate in congestive hearart failure. Circ Res.2000090909;1;1;10505:1:112121213-12121222.

111 . LaL nde G, Kyyyndddt F,,, BBaBaro III, CChabababaaannnenes D,,, BBBoissseeaau PP,P, PPPononony JCCC, Escaandndnde D,D,D, Le MaMaMarerecc HHH. DyDyDynananamic analalysy iiis of thhheee qt innnteeerval ininin lonnng qqqt11 syndndndrrromememe paaatieentsss wwwith aaa nnnorormamamal pheenenoootyypype. EEEurrr Heararart t t JJJ. 202020010101;2;22222:4141410-0-0-424222.22 JJJJ

12. Mitccchehehellllll GGGF,FF JJJerererononon AAA,, KoKoKorereren n n G.GG MeMeMeasasasurururemememenenentt t ofofof hhheaeaeartrtrt rrratatate ee ananand d d QQQ-T-T-T iiintntntererervavavall l ininin tthehehe ccconononsciousmomoousususe.e.e. AmAmAm JJJ PPPhyhyhysisisiololol... 191919989898;2;2;2747474:H:H:H7474747-77 7575751.1.1.


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

21

17. Meder B, Huttner IG, Sedaghat-Hamedani F, Just S, Dahme T, Frese KS, Vogel B, Köhler D, Kloos W, Rudloff J, Marquart S, Katus HA, Rottbauer W. PINCH proteins regulate cardiac contractility by modulating integrin-linked kinase-protein kinase B signaling. Mol Cell Biol.2011;31:3424-3435.

18. Chen H, Huang XN, Yan W, Chen K, Guo L, Tummalapali L, Dedhar S, St-Arnaud R, Wu C, Sepulveda JL. Role of the integrin-linked kinase/PINCH1/alpha-parvin complex in cardiac myocyte hypertrophy. Lab Invest. 2005;85:1342-1356.

19. Nattel S, Maguy A, Le Bouter S, Yeh YH. Arrhythmogenic ion-channel remodeling in the heart: Heart failure, myocardial infarction, and atrial fibrillation. Physiol Rev. 2007;87:425-456.

20. Sah R, Ramirez RJ, Oudit GY, Gidrewicz D, Trivieri MG, Zobel C, Backx PH. Regulation of cardiac excitation-contraction coupling by action potential repolarization: role of the transientoutward potassium current (Ito). J Physiol. 2003;546(Pt 1):5-18.

21. Glukhov AV, Fedorov VV, Kalish PW, Ravikumar VK, Lou Q, Janks D, Schuessler RB, Moazami N, Efimov IR. Conduction remodeling in human end-stage nonischemic leftventricular cardiomyopathy. Circulation. 2012;125:1835-1847.

22. Hsia HH, Callans DJ, Marchlinski FE. Characterization of endocardial electrophysiological substrate in patients with nonischemic cardiomyopathy and monomorphic ventricular tachycardia. Circulation. 2003;108:704-710.

23. Marchlinski FE. Perivalvular fibrosis and monomorphic ventricular tachycardia: toward a unifying hypothesis in nonischemic cardiomyopathy. Circulation. 2007;116:1998-2001.

24. Holmström M, Kivistö S, Heliö T, Jurkko R, Kaartinen M, Antila M, Reissell E, Kuusisto J, Kärkkäinen S, Peuhkurinen K, Koikkalainen J, Lötjönen J, Lauerma K. Late gadolinium enhanced cardiovascular magnetic resonance of lamin A/C gene mutation related dilated cardiomyopathy. J Cardiovasc Magn Reson. 2011;13:30.

25. Blumbach K, Zweers MC, Brunner G, Peters AS, Schmitz M, Schulz JN, Schild A, Denton CP, Sakai T, Fassler R, Krieg T, Eckes B. Defective granulation tissue formation in mice with specific ablation of integrin-linked kinase in fibroblasts - role of tgfbeta1 levels and rhoa activity. J Cell Sci. 2010;123:3872-3883.

26. Berul CI. Electrophysiological phenotyping in genetically engineered mice. Physiol Genomics. 2003;13:207-216.

Moazami N, Efimov IR. Conduction remodeling in human end-stage nonischemic leleleftventricular cardiomyopathy. Circulation. 2012;125:1835-1847.

22. Hssia HH,,, Caallllans DJ, Marchlinski FE. Characteterir zation of endocarddiaial electrophysiological uubsbsbstrtrtratate ininin pppaata ieieenntnts with nonischemic cardiomyopppattthy and monomomm rphihihiccc ventricular aaachhhycardia. CiCiCircrcrculuu atattioioionnn... 2020200333;1;1;1080808:7:7:7040404-7-7-71000...

23233. MMaM rchlinsks i FEEE. Peeerriivalvvvulullara fibrbrrosossisss anddd mmmonnnommmorororphphphicicc vvventrrricccular tttaccchyhyycacaardia::: tototowwawarrrd aa unnifififyiyiyingngng hhhypypypototothehehesisiiss s ininin nnnonononischchchememicicic carrrdididiomomomyoyoyopapapathththy.y.y. CCiCircrcrculululatatatioioionn... 202020070707;1;1;1161616:1119999998-8-8-20202001011.

24. Holmmmststströröröm m m M,M,M, KKKivivvisisistötötö SSS,,, HeHeHelililiööö T,T,T, JJJurururkkkkkkooo RR,R, KKKaaaaaartrtrtinininenenen MMM, , , AnAnAntititilaaa MMM,, ReReReisisissesesellllll EEE,,, KuKuKuusususisto J, KäKäKärkrkkäkääininenenen SSS,,, PePePeuhuhuhkukukuririnenenenn n K,K,K, KKKoioio kkkkalala aiaia nenenenn n J,J,J, LLLötötötjöjöjönenenenn n J,J,J, LLLauauauererermamama KKK... LaLaLatetete gggadadadololo inininiuiuium mm eenhanced cardiovascular magnetic resonance of lamin A/C gene mutation related dilated


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

22

Table 1: General and Echocardiographic Variables

5 weeks 10 weeks

Weight Parameters Control ILK P-value Control ILK P-valuen=7 n=7 n=7 n=7BW (g) 21.50 (19.90, 22.20) 19.40 (17.60, 21.00) 0.11 24.0 (21.80, 26.10) 22.30 (17.50, 23.20) 0.17RAW (mg) 3.00 (2.75, 3.25) 3.20 (2.95, 3.50) 0.24 4.00 (3.50, 4.30) 4.70 (4.10, 6.10) 0.03LAW (mg) 3.10 (3.00, 3.45) 3.70 (3.30, 3.85) 0.05 4.30 (3.80, 4.60) 14.30 (9.00, 28.60) 0.003RVW (mg) 15.10 (14.40, 16.45) 15.60 (14.85, 16.30) 0.72 20.10 (18.70, 22.20) 26.80 (24.90, 29.70) 0.01RVW/TL (mg/mm) 0.95 (0.91, 0.98) 0.95 (0.91, 1.00) 0.78 1.16 (1.07, 1.25) 1.55 (1.42, 1.70) 0.008LVW (mg) 58.50 (56.75, 62.60) 61.90 (59.75, 65.85) 0.20 70.20 (59.70, 81.60) 85.70 (78.00, 95.40) 0.03LVW/TL (mg/mm) 3.70 (3.53, 3.79) 3.85 (3.67, 4.00) 0.14 3.90 (3.41, 4.80) 4.91 (4.46, 5.45) 0.03TVW/TL (mg/mm) 4.61 (4.49, 4.74) 4.81 (4.66, 4.92) 0.09 5.13 (4.66, 5.92) 6.19 (6.15, 7.00) 0.007EchocardiographyCO (ml/min.) 36.57 (34.30, 39.92) 35.56(31.04, 39.69) 0.45 40.00 (37.00, 42.00) 31.50 (30.00, 36.00) 0.001LVDd (mm) 3.72 (3.35, 3.88) 3.72 (3.50, 3.74) 0.75 3.63 (3.36, 3.88) 4.75 (4.31, 4.94) 0.0003LVDs (mm) 2.23 (2.08, 2.31) 2.26 (2.00, 2.50) 0.56 2.29 (2.14, 2.33) 3.79 (3.67, 4.42) 0.0002LVFS (%) 40.37 (34.53, 41.27) 37.16 (35.38, 39.29) 0.99 38.77 (34.06, 41.24) 19.78 (11.03, 22.57) 0.0003LVEF (%) 77.54 (70.50, 78.60) 73.96 (71.66, 76.37) 0.98 75.79 (69.23, 78.59) 46.68 (28.12, 51.88) 0.0006IVRTc 0.69 (0.66, 1.30) 1.07 (0.84, 1.27) 0.37 0.91 (0.77, 1.37) 1.70 (1.49, 1.83) 0.01Mitral E/Em 28.71 (22.70, 34.30) 29.40 (27.00, 32.40) 0.70 26.43 (22.09, 32.76) 50.48 (36.93, 60.52) 0.002LV MPI (%) 14.30 (11.50, 22.60) 17.73 (11.90, 28.25) 0.35 23.60 (14.29, 37.55) 80.17 (54.91, 140.26) 0.01RVAWd/RVDd 0.22 (0.21, 0.25) 0.28 (0.25, 0.32) 0.04 0.23 (0.21, 0.27) 0.28 (0.25, 0.32) 0.02TAPSE (mm) 1.14 (1.01, 1.23) 1.06 (0.93, 1.15) 0.41 1.16 (1.03, 1.32) 0.72 (0.53, 1.90) <0.0001Tricuspid E/Em 10.40 (7.70, 10.90) 9.60 (9.40, 12.00) 0.76 10.34 (8.06, 10.77) 33.02 (11.28, 44.79) 0.04RV MPI (%) 16.50 (14.65, 26.45) 22.60 (15.73, 34.73) 0.52 24.85 (14.05, 32.64) 59.47 (35.45, 68.91) 0.001

Results are median (25th percentile, 75th percentile). BW: body-weight, RAW: RA-weight, LAW: LA-weight, TVW: total ventricular weight, bpm: beats/minute, CO: cardiac output, LVDd: LV end-diastolic dimension, LVDs: left ventricular end-systolic dimension, LVFS: LV fractional-shortening, LVEF: LV ejection-fraction, IVRTc: heart rate-corrected isovolumic relaxation time, Mitral E/Em: early-diastolic mitral-inflow velocity ratio, LV MPI: LV myocardial-performance index, RVAWd: RV anterior-wall thickness at end-diastole, RVDd: RV end-diastolic dimension, TAPSE: tricuspid annular-plane systolic excursion, TL: tibia length, Tricuspid E/Em: filling-velocity to tricuspid-annular early-diastolic flow velocity, RV MPI: RV myocardial-performance index. P-values based on Mann-Whitney U-test.

0.0.0.282828 (((0.0.0.252525,, , 0.0.0.323232))) 0.0.0.040404 0.0.0.232323 (((0.0.0.212121,, , 0.0.0.272727))) 0.0.0.282828 (((0.0.0.252525,,, 0.0.0.323232))) 0.0.0.000EE (mm) 1 14 (1 01, 1 23) 1 06 (0 93, 1 15) 0 41 1 16 (1 03, 1 32) 0 72 (0 53, 1 90)

( g) ( , ) ( , ) ( , ) ( 9090, ) 0.0/TL (mg/mm) 0.95 (0.91, 0.98) 0.95 (0.91, 1.00) 0.78 1.16 (1.07, 1.25) 1.55 (((11.1.424242,, 1.1.707070))) 0.00(mg) 58.50 (56.75, 62.60) 61.90 (59.75, 65.85) 0.20 70.20 (59.70, 81.60) 85.70 (7(7(78.8.0000,, , 959595 4.4.40)0)0) 0.0 0TL (mg/mm) 3.70 (3.53, 3.79) 3.85 (3.67, 4.00) 0.14 3.90 (3.41, 4.80) 4.91 (4.44 464646, 5.454545))) 0.00 000TL (mg/mm) 4.611 (4.49, 4.74) 4.81 (4.66, 4.92) 0.09 5.13 (4.66, 5.92) 6.19 (6.15, 7.00) 0.00

cardrddioioiogrgrgraapaphhyhymlll//m/mininin.) 36.57 (3(( 4.30, , 39.92))) 35.56((31.04,, 39.69)) 0.4445 40.00 (37.7.7.000000, , 42.00) 31.50 (3(( 0.00, 36.00))) 0.00

(mmmmm) 3.33 7277 (3.353535,, , 3.888888))) 3.727272 (((3.33 5000, 33.74) 0.7775 3.3 6366 (3..336,, , 3.888))) 4.4.4.77575 (4.4.4 313131,, 4.4.949494) 0.0.000((mmmm) 2.2223 (2.08,, 2.331) 2.266 ((2.000, 22.50) 0.5556 2.2 2229 (2..14,, , 2.33) 33.79 (3.676767, 444.4222) 0..0000(((%)%)% 404 .337 (((34.533, 411.27) 373 .16 (3(335.5 388, 3339.29))) 0.9999 3888.7777 (344.0666, 41.24))) 1191 .7.778 (11.0003, , 222.557))) 0..0000(%)%)%) 777.55544 (7(7(70.0 50500, , 7888.6.60) 73733 99.9666 (7(7(71.6666, ,, 767676 3.3.37)7)7) 0.99898 757575.7999 (6(( 9.9.232323, , 787878.5.59)9)9) 46466.6.6. 888 (2(2( 8.8.8.121212, 511..8.88))) 0.00000

c 0.69 (0.66, , 1.30) 1.07 (0.84, 1.27) 0.37 0.91 (0.77, 1.37) 1.70 ((1.49, 1.83) 0.0E/Em 282828.7.7.7111 (2(2(22.2.2.707070,, 343434.3.3.30)0)0) 29292 .4.4.4000 (27.7.7.000000,, 323232.4.4.40)0)0) 000.70700 262626.4.4.4333 (2(2(22.2.2.090909,, 323232.7.7.76)6)6) 505050.4.4.4888 (3(3(36.6.6.939393, , 606060.5.5.52)2)2) 0.00

PI (%) 141414.333000 ((1(11.1.1 505050, ,, 222222.6.6.60)0)0) 17177.7.77333 (111.1.909090, , 282828.2.2.25)5)5) 000.353535 232323.6.6.6000 (1(1(144.4.292929, , , 373737.5.5.55)5)5) 808080.1.11777 (5(5(54.4.4.919191,, 114140.0.26266) 0.0WWWd/d/d/RVRVRVDdDdDd 0.0.0.222222 (((0.0.0.212121,, , 0.0.0.252525)))

<0.00


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

23

Table 2: Electrophysiological Differences Based on ECG and Electrophysiological Study (EPS)

Results are median (25th percentile, 75th percentile). SNRTc: Corrected sinus-node recovery time, SACT: sino-atrial conduction time, WBCL: Wenckebach cycle-length, AH: Atrio-His interval, HV: His-Ventricular interval, AERP: Atrial ERP, VERP: Ventricular ERP. All EPS-variables except SNRTc and WBCL were measured at a cycle-length of 100 ms. P-values based on Mann-Whitney U-test..

5 weeks 10 weeks

Control ILK P-value Control ILK P-valuen=12 n=12 n=12 n=12

A. ECG

P-wave (ms) 16.00 (14.25, 17.75) 16.00 (15.00, 17.00) 1.00 17.00 (14.50, 18.00) 17.00 (16.00, 18.00) 0.08

PR-interval (ms) 29.00 (28.00, 30.80) 29.50 (28.00, 30.80) 0.72 29.50 (28.00, 30.00) 29.00 (28.00, 30.00) 0.35

QRS-duration (ms) 12.00 (10.00, 12.80) 13.00 (11.00, 14.80) 0.09 12.00 (10.50, 14.00) 16.00 (12.00, 18.00) 0.01

QT-interval (ms) 49.00 (46.50, 53.50) 53.50 (47.50, 63.00) 0.34 50.50 (46.00, 54.80) 70.50 (61.00, 77.30) <0.0001

QTc (ms) 44.00 (41.80, 47.30) 48.00 (43.00, 54.30) 0.21 44.00 (44.00, 47.30) 65.50 (56.30, 71.00) <0.0001

RR interval (ms) 127.50 (117.00, 140.00) 120.00 (114.30, 135.80) 0.48 130.50 (126.50, 141.00) 119.50 (112.50, 124.30) 0.03

B. EPS

SNRTc (ms) 34.00 (19.50, 38.50) 28.00 (21.50, 34.00) 0.70 17.50 (23.50, 40.30) 32.00 (23.80, 40.80) 0.24

SACT (ms) 19.00 (17.00, 24.00) 24.00 (19.00, 25.50) 0.31 14.00 (9.30, 16.80) 17.50 (14.30, 25.50) 0.08

WBCL (ms) 68.00 (63.00, 74.00) 67.00 (62.00, 74.00) 0.84 68.00 (64.30, 69.50) 72.00 (65.00, 75.50) 0.07

AH (ms) 20.00 (17.50, 21.50) 20.00 (18.00, 21.50) 0.87 20.00 (19.00, 21.80) 21.00 (19.30, 22.00) 0.23

HV (ms) 9.00 (8.50, 10.00) 10.00 (8.50, 11.50) 0.24 8.50 (8.00, 9.00) 8.80 (7.00, 11.50) 0.32

AERP (ms) 24.00 (22.00, 27.00) 24.00 (22.00, 29.00) 0.77 26.00 (24.00, 28.00) 26.00 (24.00, 30.50) 0.61

VERP (ms) 30.00 (29.00, 33.00) 30.00 (29.00, 35.00) 0.73 27.00 (23.80, 30.80) 39.00 (28.50, 45.50) 0.04

terval (ms) 29.00 (28.00, 30.80) 29.50 (28.00, 30.80) 0.72 29.50 (28.00, 30.00) 29.00 (2(2(28.88 00,,, 303030.0.0.00)0)0)

duration (ms) 12.00 (10.00, 12.80) 13.00 (11.00, 14.80) 0.09 12.00 (10.50, 14.00) 16.0000 (1(112.000000,,, 181818.00.00)0)0)

terval (ms) 49.00 (46.50, 53.50) 53.50 (47.50, 63.00) 0.34 50.50 (46.00, 54.80) 70.50 (6(61.00, 77.30) <

ms) 4444 .000000 (4(4(41.80, 47.30) 48.00 (43.00, 54.30) 0.0.0.212 44.00 (44.00, 47.30)0)0) 65.50 (56.30, 71.00) <

tererervav ll l ((ms) 1212127.50 (117.00, 140.00) 120.00 (114.30, 135.80) 0..484 130.50 (12266.6.5550, 141.00) 119.50 (112.50, 124.30)

PPPSS

c (m(m(ms)s)) 33434.0000 (1(119.50,,, 3888.50) 282828.0.0.000 0 ((2(21.5505 , 34343 .0.00000) 0...707 1117.5550 0 0 (22333.50500,, 40404 .3000) 32322.00 (2(2(23.3..8880, 4440.880)

(ms) 19.00 (17.00,, 24.00) 24.00 (19.00, 25.50))) 0.31 14.00 (9.30,,, 16.80) 17.50 (14.30, 25.50)

L (ms) 686868.0.0.00 0 0 (6(663.3.3.000000, 747474.0.0.00)00) 676767.0.0.000 0 (6662.2.2 000000, 747474.0.0.00)0)0) 0.0.0.848484 686868.0.0.00 0 0 (6(6(64.4.4.30300, 696969.5.550)0)0) 727272.0.00000 (6(6(65.5.5.000000,,, 757575..550)

ms) 20.00 (1117.7.7 550, 21.50) 20.00 (18.00, 21.550) 0.877 20.00 (111999.00, 2111.8.88000) 21.00 (1( 9.9.303030, 222222.00)

mms) 9.00 (8.50, 10.00) 10.00 (8.50, 11.50) 0.24 8.50 (8.00, 9.00) 8.80 (7.00, 11.50)


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

24

Figure Legends:

Figure 1: Premature death and dilated-cardiomyopathy phenotype in ILK-KO mice. A:

Cumulative survival curves obtained from 30 animals/group. B: Representative Masson’s

trichrome-stained longitudinal heart-sections from control and ILK-KO mice at 10 weeks of age.

C: Total ventricular-weight/body-weight (TLV/BW) ratio. Mean±SEM. n=7 mice/group.

*P<0.01, Student t-test. D: Mean±SEM ANP and BNP mRNA expression in left-ventricular

(LV) tissues of control and ILK-KO mice, n=8 in each group. ***P<0.001, Mann-Whitney

U-test. RA=right atrium, RV=right ventricle, LA=left atrium.

Figure 2: Interstitial fibrosis and Cx-43 lateralization in 10-week ILK-KO mice. A:

Representative images of interstitial fibrosis within the superior part of the left ventricle (LV-

Sup), the inferior part of the left ventricle (LV-Inf), the interventricular septum (IVS) and the

right ventricle (RV) in both control (Upper panels) and ILK-KO (Lower panels) myocardium.

B: Fibrosis (n=8 /group), mean±SEM. *P<0.05, ***P<0.001, Mann-Whitney U-test. C:

Mean±SEM vimentin, -SMA and procollagen-1 1 mRNA-expression in left-ventricular (LV)

tissues of control and ILK-KO mice, n=8 for each group, except n=7 for -SMA in control mice.

***P<0.001, Student t-test. D: Representative connexin-43 (green) and actin (red) staining in

LV-cryosections from control and ILK-KO mice (Top). Percentage of connexin-43

lateralization, mean±SEM (Bottom). **P<0.01, Mann-Whitney U-test. E: Representative

immunoblots for connexin-43 and GAPDH (Top). Mean ± SEM connexin-43 normalized to

GAPDH (bottom). *P<0.05, Mann-Whitney U-test.

Figure 3: ECGs and arrhythmias in 10-week ILK-KO mice. A: Representative surface

Figure 2: Interstitial fibrosis and Cx-43 lateralization in 10-week ILK-KO mice. A:A:A:

Reeeprprpreesesentataatititivvve iiimamam ges of interstitial fibrosis withihihinnn ttthe superior pararart of ttthehehe left ventricle (LV-

SSuS pp)p), the inferiorrr pppart offf theh lllefefeft veeentntntricccleee (LV-V-V Inf)f)f), ttheee inininteteervrvrventrrricccular ssseppptuuummm (IVSVSS) )) anannd tht e

ighghhttt veveventntntriririclclcle e (R(R(RV)V)V) iiinnn bobobothtt ccononontrtrt ololol (((UpUpppepeperr r papapanenenelslsls) ) ) ananand d d ILILLK-K-K-KOKOKO (((LoLoLowewewerrr papapanenen lslsls)) ) mymymyooocararardiiiumumum.

B: Fibrosososisisis (((n=8=8=8 ///grgrgrouououp)p)p),,, mememeananan±S±S±SEMEMEM.. ***PPP<0<0<0.0.0.055,, *********PPP<0<0<0.0.0.0010101,,, MaMaMannnnnn--WhWhWhitititneneney y y UUU--teteteststst.. C:C:C:

MMean±SEM vimentin SMA and procollagen 1 1 mRNA expression in left ventricular (LV)


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

25

electrograms (lead 1, L1) in control (left) and ILK-KO (right) mice. B: Representative L1 ECG

and right-ventricular (Endo) electrograms obtained during intracardiac programmed stimulation

in an ILK-KO mouse. C: Representative examples of telemetry recordings in control and ILK-

KO mice. Spontaneous ventricular tachycardia was observed exclusively in ILK-KO mice. D:

The proportion of mice with induced ventricular tachycardia during programmed electrical

stimulation. E: The proportion of mice with spontaneous ventricular tachycardia on 24-hour

telemetry. *P<0.01, ***P<0.001, Fisher’s exact-test.

Figure 4: Optical mapping of conduction in 10-week mice. A. Activation maps in representative

control (CTL) and ILK-KO mice. Arrows show the longitudinal and transverse conduction axes.

B. Mean±SEM conduction velocity (CV). CV was significantly decreased by ILK-KO in both

longitudinal and transverse directions. **P<0.01, ***P<0.001; two-way ANOVA for repeated-

measures with fixed factors group and direction; followed by t-test with Bonferroni-correction

(nc=2 for comparisons).

Figure 5: Action-potential (AP)-recordings. A: Representative AP-recordings from 10-week

control and ILK-KO mouse multicellular preparations at a basic cycle-length of 150 ms.

Triggered activity was observed in (3/6) ILK-KO preparations, and no control preparations (0/6).

B: Typical APs recorded within a left-ventricular tissue preparation of control (grey) and ILK-

KO (black) mice. C: Mean±SEM AP-duration (APD) values measured in multicellular

preparations at 25%, 50% and 90% repolarization (APD 25, 50 and 90 respectively). n=6/group.

***P<0.001, One-way ANOVA followed by Bonferroni-corrected t-tests (nc=3 for

comparisons). D: Mean±SEM action-potential duration to 90%-repolarization (APD90) from

single-cell recordings at frequencies shown in 5- and 10-week mice (n/N=cells/mice).

control (CTL) and ILK-KO mice. Arrows show the longitudinal and transverse cconononduduductctctioioion n n axaxaxesee .

B. Mean±SEM conduction velocity (CV). CV was significantly decreased by ILK-KKO in both

ooongngngiiitudinalll aaand transverse directions. **P<0.01, *******P<0.001; twowowo-way ANOVA for repep ated-

mmmeaaasures with fiixeeed faaactttors grgrgrooup and didirrectttiooon; fffooollllowewewed d d bybby t-teesst with BBononnfefeerroni---cooorrrreccctionnn

nc=2=22 fffororor ccomompapa iririsosonsns)).).

FFigure 5: Action potential (AP) recordings A: Representative AP recordings from 10 week


Dow

nloaded from


DOI: 10.1161/CIRCEP.115.001668

26

Comparison by multi-level mixed-effects model followed by Bonferroni-corrected t-tests (nc=3

for comparisons). **P<0.01, ***P<0.001. Results in A-C were obtained with floating standard

microelectrodes on Langendorff-perfused hearts; results in D were obtained by perforated-patch

recording.

Figure 6: Changes in K+-currents. A, B: Mean±SEM peak (Ipeak), fast-inactivating (Ito,f), slow-

inactivating (IK,slow) and steady-state (Iss) transient-outward current component current-densities

(pA/pF) plotted as a function of test-pulse potential (mV) in control (CTL) vs ILK-KO

(n/N=cells/mice) at 5 weeks (A) and 10 weeks (B). *P<0.05, **P<0.01, ***P<0.001 versus

CTL, multi-level mixed-effects model ANOVA followed by Bonferroni-corrected t-test (nc=13

for comparisons).

Figure 7: IK1 and L-type Ca2+-current changes. A, B: Mean±SEM IK1 current-density as a

function of test-potential (mV) in control (CTL) vs ILK-KO cardiomyocytes at 5 (A) and 10 (B)

weeks (n/N=cells/mice). **P<0.01, ***P<0.001 versus CTL, multi-level mixed-effects model

ANOVA followed by Bonferroni-corrected t-test (nc=12 for comparisons). C, D: Mean±SEM

ICaL current-density as a function of test-potential (mV) in CTL vs ILK-KO cardiomyocytes at 5

(C) and 10 (D) weeks (n/N=cells/mice). *P<0.05, ***P<0.001, multi-level mixed-effects model

ANOVA followed by Bonferroni-corrected t-test (nc=12 for comparisons).

Figure 8: Schematic of the changes we observed in 10-week ILK-KO mice and their possible

relation to ventricular arrhythmias.

CTL, multi-level mixed-effects model ANOVA followed by Bonferroni-correcteeedd d ttt--teteteststst (((ncncnc=1=1=13

for comparisons). r

FFiF guguure 7: IK1 andndd LL-tyyypeee CCa2+2++---currrrenenent chchchanggesss. A,, BBB: MeMeMeannn±S±S±SEMMM IIIK1I cuuurrrrenennttt-dddensititity y y asass aaa

funcncctititiononon ooof f f teteteststt--popopotetetentntntiaiaalll (m(( V)V)V) iiin n cococontrorooll (C(C(CTLTLTL) )) vsvsvs IIILKLKLK-KOKOKO cccararardidiiomomomyoyoyocycycytetetes s atatat 555 (((AAA))) ananandd 101010 (((BBB)))

weeks ((n/n/n/NNN=c=c=ceelellslsls/m/m/micicice)e)e)... ******PPP<0<0<0.0.0.0111, *********PPP<0<0<0.00010101 vvvererersusususss CTCTCTL,L,L, mumumultltltiii---lelelevevevel l l mimimixexexeddd---efefeffefefectctctsss mmom del

AANOVA followed by Bonferroni corrected t test (nc=12 for comparisons)r C D: Mean±SEM


Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from



Dow

nloaded from


Yan-Fen Shi, Denis Chartier, Jean-Claude Tardif, Dobromir Dobrev and Stanley NattelKhai Le Quang, Ange Maguy, Xiao-Yan Qi, Patrice Naud, Feng Xiong, Artavazd Tadevosyan,

in MiceLoss of Cardiomyocyte Integrin Linked Kinase Produces an Arrhythmogenic Cardiomyopathy

Print ISSN: 1941-3149. Online ISSN: 1941-3084 Copyright © 2015 American Heart Association, Inc. All rights reserved.

Dallas, TX 75231is published by the American Heart Association, 7272 Greenville Avenue,Circulation: Arrhythmia and Electrophysiology

published online June 12, 2015;Circ Arrhythm Electrophysiol.

http://circep.ahajournals.org/content/early/2015/06/12/CIRCEP.115.001668World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circep.ahajournals.org/content/suppl/2015/06/12/CIRCEP.115.001668.DC1Data Supplement (unedited) at:

http://circep.ahajournals.org//subscriptions/

is online at: Circulation: Arrhythmia and Electrophysiology Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answerinformation about this process is available in the

requested is located, click Request Permissions in the middle column of the Web page under Services. FurtherCenter, not the Editorial Office. Once the online version of the published article for which permission is being

can be obtained via RightsLink, a service of the Copyright ClearanceCirculation: Arrhythmia and Electrophysiology Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:


Dow

nloaded from

http://circep.ahajournals.org/content/early/2015/06/12/CIRCEP.115.001668

http://circep.ahajournals.org/content/suppl/2015/06/12/CIRCEP.115.001668.DC1

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circep.ahajournals.org//subscriptions/


1

Supplemental Material

Supplemental Methods

Animal model

Age-matched mice expressing the cardiac mckCRE transgene and 2 loxP1- flanked ILK

alleles (mckCRE ILKfl/fl-KO mice; ILK-KO) were compared to littermate controls

(mckCRE) as described previously.1 All procedures followed National Institutes of

Health guidelines (http://oacu.od.nih.gov/training/index.htm) and were approved by the

Animal Research Ethics Committee of the Montreal Heart Institute. All studies were

performed in 10-week old mice, unless otherwise indicated.

Echocardiography

Transthoracic echocardiography was performed under 2.5%-isofluorane anesthesia, with

an i13L (10-14 Megaherz) probe and Vivid 7 Dimension system (GE Healthcare

Ultrasound, Horten, Norway). The person performing and analyzing echocardiographic

imagery was blinded to mouse-group assignment. Left-ventricular (LV) M-mode

imaging was obtained in parasternal short-axis view at the level of the papillary muscle.

LV dimensions at end-diastole (LVDd) and systole (LVDs) were measured, LV fractional

shortening (LVFS) was calculated as (LVDd – LVDs)/LVDd X 100%. LV volumes and

ejection fraction (LVEF) were calculated by the formula packed in GE Healthcare

Ultrasound as recommended by the American Society of Echocardiography. Dimension

(D) of the LV outflow-tract (LVOT) was measured in a zoomed parasternal long-axis

view. LVOT cross sectional area (CSALVOT) was calculated as π(D/2). LVOT flow

was obtained by pulsed-wave Doppler (PW) in apical 5-chamber view. LVOT

2

velocity-time integral (VTILVOT) was traced, stroke volume (SV) was calculated as

VTILVOT X CSALVOT and cardiac output (CO) as HR X SV. The right-ventricular

(RV) M-mode image was recorded in parasternal long-axis view, the thickness of RV

anterior wall (RVAW) and RV dimension at end-diastole were measured, and the

RVAWd/RVDd ratio was calculated. Tricuspid annulus plane systolic excursion

(TAPSE) was measured with RV lateral tricuspid annulus M-mode in apical 4-chamber

view. PW was used to record both transmitral and transtricuspid flow in apical

4-chamber view, peak velocity in early filling E-wave, and time interval from

mitral/tricuspid closing to opening (MVco, TVco) were measured. Mitral and tricuspid

annulus movements were recorded with tissue Doppler imaging, velocity during early

filling Em was measured, and E/Em ratios were calculated for both mitral and tricuspid

studies. LV isovolumetric relaxation time (IVRT) was measured using continuous wave

Doppler at the conjunction of LV inflow and outflow in apical 5-chamber view and was

corrected (IVRTc) by the square root of the R-R interval. PW was also used to record

aortic and pulmonic flow in apical 5-chamber and parasternal short-axis view

respectively. LV/RV ejection time (ET) was measured from the beginning to the ending

of aortic/pulmonic flow. LV global myocardial performance index (MPI) was calculated

as (MVco-LVET)/LVET X 100%, RV MPI as (TVc-o–RVET)/RVET X 100%. The

average of 3 consecutive cardiac cycles was used for each measurement. Special care

was taken to get similar imaging planes at follow-up studies.

Histology and fibrosis quantification

Ventricular mass was assessed using total ventricular weight-to-body-weight ratio

(TVW/BW). For histological studies, the hearts were perfused with 10% neutral-

3

buffered formalin, immersed in the fixative, dehydrated and embedded in paraffin.

Longitudinal sections (6-μm) were stained with Masson’s trichrome to identify collagen-

deposition. Microscopic images at 200×magnification were digitized and analyzed with

ImagePro Plus 7.0 software (MediaCybernetics). Connective tissue content was

quantified as percentage of surface area, with perivascular collagen excluded from

measurement.

Connexin-43 immunostaining and quantification of lateralization

LV tissues were embedded in Optimal Cutting-Temperature solution (Sakura) and

snap-frozen in liquid nitrogen. Cryosections (12-μm thickness) were fixed with a

1×Phosphate-Buffer Saline (PBS) solution containing 4% paraformaldehyde (PFA, pH

7.3), blocked and permeabilized with 1×PBS solution containing 2%-normal donkey

serum (NDS) and 0.5%-Triton X-100. Primary antibody (rabbit anti-Cx43 AB1727

Millipore) was diluted 1/200 in 1×PBS solution containing 2%-NDS and 0.1%-Triton X-

100 for overnight incubation with cryosections. Alexa Fluor-conjugated donkey

anti-rabbit IgG (488-nm, Invitrogen) was used as a secondary antibody (1/600 dilution).

Alexa Fluor-conjugated phalloidin (647-nm) was used as an actin-filament marker (1/600

dilution). Slides were mounted in DABCO/Glycerol (25%/75%) and examined with a

Zeiss Axiovert 100-M microscope coupled to a Zeiss LSM-710 laser-scanning confocal

system. Identical settings were used to image samples from control and ILK-KO

animals. A 40x objective with a zoom-out of 0.6 was used. Images were deconvolved

with the Maximum Likelihood Estimation algorithm (Huygens software, Scientific

Volume Imaging). Connexin-43 lateralization was determined using Image Pro Plus 6.0

software (Media Cybernetics). Briefly, the total pixel-intensity associated with

4

Connexin-43 staining (x) was first quantified. In a second step, the lateral Connexin-43

positive staining was selected and the corresponding pixel-intensity sum (b) was

quantified. The extent of lateralization (y) was expressed in percentage as follows:

y = (b x 100)/x.

Protein extraction and immunoblotting

LV-samples from control and ILK-KO hearts were excised, snap-frozen in liquid

nitrogen and mechanically homogenized in a TNE buffer containing: Tris 25-mmol/L,

EDTA 5-mmol/L, EGTA 5-mmol/L, NaCl 150-mmol/L, NaF 20-mmol/L, Na3VO4

0.2-mmol/L, β-glycerophosphate 20-mmol/L, AEBSF 0.1-mmol/L, leupeptin 25-µg/mL,

aprotinin 10-µg/mL, pepstatin 1-µg/mL, microcystin-LR 1-µmol/L, pH 7.34, HCl.

Homogenized samples were centrifuged at 1000×g for 10 minutes, supernatant collected

and ultracentrifuged at 100,000×g for 1 hour. The resulting pellet corresponding to the

membrane fraction was then suspended and incubated in TNE buffer containing 1%

Triton-X100. The protein concentration was determined by Bradford assay (BioRad).

All steps were carried out on ice at 4-5°C. Protein samples (20 µg) were separated on 8%

poly-acrylamide SDS-PAGE and transferred electrophoretically onto PVDF membranes.

The PVDF membranes were blocked in a PBS-solution containing 0.05% (v/v) Tween-20

and 5% (w/v) non-fat dried milk (NDM) and incubated overnight at 4°C with primary

antibodies diluted in PBS containing 0.05% Tween-20 and 1%-NDM. After washing

with PBS-Tween solution/1%-NDM, membranes were hybridized with HRP-conjugated

secondary antibody. Immunoreactive bands were detected by electrochemoluminescence

with BioMax MS/MR films. Protein quantification was performed with Quantity-One

software (BioRad). All expression data are provided relative to GAPDH staining for the

5

same samples on the same gels. Primary antibodies (1/2000) included polyclonal rabbit

anti-connexin43 (AB1727; Millipore) and monoclonal anti-GAPDH (10R-G109a) from

Fitzgerald. Peroxidase-conjugated AffiniPure donkey anti-rabbit IgG (111-035-152) and

Affinipure donkey anti-mouse IgG (715-035-151) from Jackson ImmunoResearch were

used as secondary antibodies (1/10000).

Co-immunoprecipitation Sheep anti-mouse IgG M-280 coated Dynabeads were used for co-immunoprecipitation

(co-IP). Five μg of anti-ILK antibody were incubated overnight at 4ºC in PBS containing

1.0×107 magnetic beads and 0.1%-bovine serum albumin with gentle rotation. The

immobilised antibody attached to the beads was cross-linked with BS3

(Bis[sulfosuccinimidyl] suberate). Following multiple washings, membrane-protein

extracted from left ventricular tissue samples (300 μg) either from control or ILK-KO

mice were incubated with beads and rotated overnight at 4ºC. Control incubations were

made with non-immune IgG. Samples were eluted, and the proteins were separated by

SDS-PAGE prior to immunoblotting. The primary antibodies used for immunoblots on

co-IPs are provided below:

6

Antibody Species Raised In Dilution Source

ILK Mouse 1:500 Sigma

Kv4.1 Rabbit 1:600 Alomone





KChIP2 Rabbit 1:1000 Alomone

Kir2.1 Rabbit 1:1000 Alomone

Cav1.2 Rabbit 1:200 Alomone

mRNA isolation and TLDA analysis

Samples from the superior LV free-wall were homogenized, total RNA was isolated

using Trizol, purified and further quantified via Nanodrop. DNAse I (Invitrogen)

treatment was performed. Absence of genomic DNA-contamination was confirmed by

PCR. The integrity of total RNA was assessed with an Agilent bioanalyser. RNA with a

RIN number above 7,5 was used for gene expression analysis. cDNA was synthesized

from 220 ng total RNA using the High Capacity cDNA Reverse Transcription Kit

(Applied Biosystems) with random hexamer primers. TaqMan low-density arrays

(TLDA, Applied Biosystems) were used in two-step RT-PCR as previously reported.2

Real-time PCR was performed on the 7900HT Fast Real-Time PCR System. Data were

collected with SDS2.3 software and grouped with RQ Manager software. The TLDA

plates were designed for the study of 96 selected genes related to cardiac remodeling,

electrophysiology, contractility and fibrosis (Supplemental Table 1). Taqman assays

were used for quantification. The thermal profile for the PCR reactions was: 2 min 50°C;

7

10 min 95°C; then 40 cycles with 30 sec 95°C; 1 min 60°C. GAPDH was selected as the

housekeeping-gene internal standard.

Electrocardiography and telemetry recording

All mice were anaesthetized for ECG-recording under 2.5%-isoflurane anesthesia. Body

temperature was maintained at 37°C with a heating pad (Harvard Apparatus, USA). The

surface ECG (lead I) was obtained with three 25-gauge subcutaneous electrodes and

transmitted to a computer via an analogue-digital converter (IOX v1.585, EMKA

Technologies) for monitoring and off-line analysis with ECG-Auto 1.5.12.10 software

(EMKA Technologies). Recordings were filtered between 0.5 and 500 Hz.

Measurements were based on averages of 10 consecutive complexes. Standard criteria

were used to measure RR, PR, QRS, and QT intervals.3 The QT interval was corrected

for heart rate (HR) as follows: corrected QT interval (QTc)=QT/(RR/100)1/2, with QT and

RR, expressed in milliseconds (ms).4

At 9 weeks of age, telemetry devices were implanted in mckCRE ILKfl/fl and

control mice for 24-hour ECG recordings. Briefly, mice were anesthetized with 2.5%

isoflurane and an abdominal incision was made 1.5 cm lateral to the midline to insert a

telemetry transmitter (TA10EA-F20, Data Sciences International) into a subcutaneous

pocket with paired wire electrodes placed under the skin over the thorax (to obtain a

bipolar chest ECG lead). Mice were housed in individual cages with free access to food

and water and exposed to 12-hr light/dark cycles (light, 6:00 AM to 6:00 PM) in a

thermostatically controlled room. One week following implantation (10 weeks of age),

telemetry devices were turned on and ECG-signals were computer-recorded on free-

moving animals for 24 hours with a telemetry receiver (RPC1, DSI International). IOX

8

2.3 and ECG-Auto 2.4.0.30 (EMKA) systems were used for display and analysis. Heart-

rate values were determined from RR-interval averages over 10 seconds.

Intracardiac recording and programmed stimulation

Both mckCRE-ILKfl/fl and control mice were anaesthetized with 2.5% isoflurane. An

octapolar electrophysiology catheter (1.2F) designed for mouse electrophysiology

(Biosense Inc.) was positioned in the RV via internal jugular vein. Intra-cardiac

electrograms were used to guide catheter positioning. Surface ECG (lead I) and

intra-cardiac electrograms were recorded on a computer through an analogue-digital

converter (IOX 1.11, EMKA) for monitoring and later analysis (ECG-Auto 2.1.4.15).

Intra-cardiac electrograms were filtered between 0.5 and 500 Hz. Pacing was performed

with a custom-built computer-based stimulator triggering a stimulus isolator

(509 Stimulator, Grass Telefactor). Standard pacing-protocols were used to determine

electrophysiological parameters. Refractory periods were determined with a

nine-stimulus drive train (S1) at a cycle-length of 100 ms followed by a 1.5×threshold-

current premature stimulus (S2) progressively decremented in 2-ms intervals. Atrial,

atrioventricular nodal, and ventricular effective refractory periods (AERP, AVERP, and

VERP, respectively) were defined as the longest S1-S2 coupling interval that failed to

generate a propagated beat. Programmed single, double, and triple extrastimuli were

applied at a 100-ms drive cycle length. Ventricular tachycardia (VT) was defined as 10

or more successive spontaneous ventricular complexes.

9

Optical mapping

A total of 16 mice were studied: 6 ILK-KO mice and 10 wild type mice. Mouse hearts

were excised and placed in oxygenated (95% O2, 5% CO2) Krebs solution (mmol/L:

120 NaCl, 4 KCl, 1.2 MgSO4 0.7, 1.2 KH2PO4, 25 NaHCO3, 5.5 glucose, 1.25 CaCl2),

and then perfused in Langedorff mode with blebbistatin-containing (15 µmol/L) 37°C

Krebs solution at 1.5 mL/min. The heart was paced with bipolar silver wires and 2 ms,

1.5×threshold current stimuli at cycle lengths of 100, 150 and 200 ms. After

stabilization, di-4-ANEPPS (Biotium, CA, dissolved in DMSO) was injected. Optical

mapping was obtained with a Redshirt Imaging CCD (charge coupled device) camera

with an 80 by 80 photodiode array. Fluorescence images were recorded at 2 kHz while

pacing.

The recorded data were processed with software written in Matlab (version 7.11,

MathWorks). Biorthogonal symlet wavelet filtering was used to filter fluorescence

intensity signals at each pixel. The thresholding value was calculated by the minimax

method, and soft thresholding used for high frequency components. The activation time

was defined as the time of maximum dF/dtmax of the upstroke of the filtered signal. A

five point Gaussian filter was applied to filter the calculated isochronal activation map.

Conduction velocity was calculated from the gradient of the scalar field of the isochronal

activation map, based on the average of local velocity vectors in the principal propagation

direction. Longitudinal and transverse propagation were determined by calculating mean

conduction velocities for individual vector-axes (by 2-degree increments) and selecting

the vector-axes with greatest and least mean conduction-velocity respectively. The

vectors selected were perpendicular to each other (±5 degrees) in all cases.

10

Action-potential (AP) recordings

Isolated hearts from 10-week control and ILK-mice were perfused with Krebs-Henseleit

solution containing (mM): 120 NaCl, 4 KCl, 1.2 NaH2PO4, 1.2 MgSO4, 25 NaHCO3,

1.25 CaCl2, and 5.5 glucose (95% O2-5% CO2, pH 7.4) at 36°C. Epicardial AP-duration

(APD) was recorded from perfused ventricles with fine-tipped borosilicate-glass

microelectrodes (resistances 10-20 MΩ when filled with 3 M KCl) coupled to an

Axoclamp-2B amplifier (Axon Instruments, Foster City, CA), digitized with a Digidata-

1200 series A/D-converter, and displayed with Axotape 2.0 software (10-kHz sampling).

Preparations were stimulated with 2-ms 1.5xdiastolic-threshold current square-wave

pulses through bipolar Teflon-coated silver electrodes. After 5 minutes of pacing at 0.5

Hz for equilibration, preparations were paced successively at increasing frequencies from

0.5 to 10 Hz, with 200 beats at each frequency to ensure steady state prior to recording.

APD was analyzed with a custom-made Matlab algorithm, with each value based on the

average of five consecutive APs at three different regions of the LV. The data were

recorded using pClamp 8.0 software and analyzed with Clampfit 8.0 (Axon Instruments,

Foster City, USA). In addition, APs were recorded in single ventricular myocytes from

5-week and 10-week mice (ILK-KO and age-matched littermates; for isolation method,

see below) with nystatin perforated-patch methods, in current-clamp mode. For

perforated-patch recording, nystatin-free intracellular solution was placed in the tip of the

pipette, and then pipettes were backfilled with nystatin-containing (600 µg/mL) solution.

Ventricular-cardiomyocyte isolation

Adult male mice (10-week) were heparinized (100 IU, intraperitoneal), anaesthetized

with 2.5%-isoflurane and sacrificed by cervical dislocation. The hearts were rapidly

11

removed, and retrogradely perfused through the aorta on a modified Langendorff

apparatus with the following solutions (at 37±1°C; 100% O2): (i) 5 minutes with HEPES-

buffered Tyrode solution containing (mM): 130 NaCl, 5.4 KCl, 1 CaCl2, 1 MgCl2, 0.33

Na2HPO4, 10 HEPES, 5.5 glucose (pH adjusted to 7.4 with NaOH); (ii) 10 min with

nominally Ca2+-free Tyrode solution; (iii) 20 min with Ca2+-free Tyrode solution

containing 73.7 U/ml collagenase type 2 (Worthington Co. Ltd, Freehold, NJ), 0.1%

bovine serum albumin (BSA; Fraction V, Sigma Chemicals Co., St. Louis, Mo), 20 mM

taurine and 30 μM CaCl2; and (iv) 5 min with a storage-solution containing (in mM):

100 K-glutamate; 10 K-aspartate, 25 KCl, 10 KH2PO4, 2 MgS04, 20 taurine, 5 creatine,

0.5 EGTA, 5 HEPES, 0.1%-BSA, 20 glucose (pH adjusted to 7.3 with KOH). After

perfusion, the hearts were removed and single LV-cardiomyocytes obtained by

trituration, then placed into storage-solution at 4°C.

Ion-current recordings and analysis

IK1, IK,Total and ICa,L were recorded at 37°C with tight-seal patch-clamp in voltage-clamp

mode. Borosilicate glass electrodes with a tip resistance of 3 to 5 MΩ (Sutter Instrument)

were filled with pipette solution and connected to a patch-clamp amplifier (Axopatch

200A, Axon). Cell-capacitances were obtained with 5-mV, 10-ms hyperpolarizing steps

from -60 mV. For both IK1 and IK,Total recordings, the extracellular solution (Tyrode)

contained (in mmol/L): NaCl 136, KCl 5.4, MgCl2 1, CaCl2 1, NaH2PO4 0.33, HEPES 5,

and dextrose 10 (pH 7.35, NaOH). CdCl2 (200µmol/L) was used to inhibit L-type Ca2+

current. The internal solution contained (mmol/L): potassium-aspartate 110, KCl 20,

MgCl2 1, Mg-ATP 5, GTP (lithium salt) 0.1, HEPES 10, sodium-phosphocreatinine 5,

and EGTA 10.0 (pH 7.3, KOH). The extracellular solution for ICaL-recording contained

12

(mmol/L) tetraethylammonium-chloride 136, CsCl 5.4, MgCl2 0.8, CaCl2 2, NaH2PO4

0.33, dextrose 10, and HEPES 10, pH 7.4 (CsOH). Niflumic acid (50-µmol/L) was

added to inhibit Ca2+-dependent Cl- current, and 4-aminopyridine (2-mmol/L) to suppress

Ito. The pipette-solution for ICaL-recording contained (mmol/L) CsCl 120,

tetraethylammonium chloride 20, MgCl2 1,EGTA 20, ATP-Mg 5, HEPES 10, and GTP

(lithium salt) 0.1, pH 7.4 (CsOH).

IK1 was recorded by applying 300-ms test-pulses from a holding potential

of -40 mV. Total voltage-gated outward K+ current (IK,Total) was recorded from a holding

potential of -70 mV. ICa,L was recorded with a holding potential of -50 mV. IK1 was

measured at the end of the 300-ms test-pulses. Peak current (Ipeak) was defined as the

maximal outward K+-current. The amplitude of Ito,f, IK,slow and Iss were determined from

exponential fits to the decay-phases of the total outward K+-current as previously

described.5,6 Currents are expressed as densities (pA/pF).

Data analysis

Normality of distribution was assessed via Shapiro-Wilk test. Normally-distributed data

are expressed as mean±SEM and analyzed by Student’s nonpaired t-test (for comparisons

involving only CTL vs ILK-KO group-comparisons without other factors or repeated

measures) or by ANOVA; data not satisfying normal-distribution criteria are shown as

median and 25%-75% percentile and were analyzed by Mann-Whitney U-test. For work

involving multiple sample-measures from thesame animal (e.g. multiple cell-

impalements, multiple cells from the same mouse available for patch-clamp work;

numbers are provided in figures as n/N for cells/mice), data were analyzed using a multi-

level mixed-effects model to take into account correlation between multiple levels of

13

within-mouse measurements. In this model, the mouse effect and the interaction

“cell×mouse” effect were considered as random factors. According to the design of the

experiments, fixed factors for repeated measures (e.g. voltages, frequencies, weeks) were

added to the model to produce multi-level repeated measures mixed-effects models.

Two-way ANOVA on log-transformed data was used to compare control versus ILK-KO

mice across time (5 and 10 weeks) for PCR parameters. Bonferroni correction was used

to take into account multiple statistical tests to identify specific mean-differences (P-

value from t-test multiplied by nc=number of comparisons). Fisher’s exact test was used

to compare differences in occurrence-rates. A two-tailed P-value<0.05 denoted

statistically-significant differences.

14

References

1. White DE, Coutu P, Shi YF, Tardif JC, Nattel S, St Arnaud R, Dedhar S,

Muller WJ. Targeted ablation of ilk from the murine heart results in dilated

cardiomyopathy and spontaneous heart failure. Genes Dev. 2006;20:2355-2360.

2. Gaborit N, Le Bouter S, Szuts V, Varro A, Escande D, Nattel S, Demolombe S.

Regional and tissue specific transcript signatures of ion channel genes in the non-

diseased human heart. J Physiol. 2007;582:675-693.

3. Lande G, Kyndt F, Baro I, Chabannes D, Boisseau P, Pony JC, Escande D,

Le Marec H. Dynamic analysis of the qt interval in long qt1 syndrome patients

with a normal phenotype. Eur Heart J. 2001;22:410-422.

4. Mitchell GF, Jeron A, Koren G. Measurement of heart rate and Q-T interval in the

conscious mouse. Am J Physiol. 1998;274:H747-751.

5. Xu H, Guo W, Nerbonne JM. Four kinetically distinct depolarization-activated K+

currents in adult mouse ventricular myocytes. J Gen Physiol. 1999;113:661-678.

6. Kuo HC, Cheng CF, Clark RB, Lin JJ, Lin JL, Hoshijima M, Nguyêñ-Trân VT,

Gu Y, Ikeda Y, Chu PH, Ross J, Giles WR, Chien KR. A defect in the Kv channel-

interacting protein 2 (KChIP2) gene leads to a complete loss of Ito and confers

susceptibility to ventricular tachycardia. Cell. 2001;107:801-813.

15

Supplemental Table 1. Full Results of Changes in All Genes Assayed

Gene Protein Control (n) ILK-KO (n) P-value

ILK Pathway

Parva α-parvin 47.7 ± 1.9 (8) 79.4 ± 5.5 (8) <0.01

Itga7 integrin α7 188.7 ± 10.7 (8) 207.7 ± 12.4 (8) NS

Itgb1 integrin β1 641.4 ± 21.1 (8) 1266.2 ± 52.2 (8) <0.001

Lama2 Laminin α2 91.8 ± 3.7 (8) 118.8 ± 7.5 (8) <0.01

Lama4 Laminin α4 124.7 ± 5.0 (8) 159.8 ± 18.7 (8) NS

Lims1 PINCH1 135.5 ± 9.4 (8) 110.2 ± 7.9 (8) NS

Lims2 PINCH2 228.7 ± 13.6 (8) 190.7 ± 15.1 (8) <0.05

Ctnnb1 Beta 1 Catenin 383.6 ± 22.7 (8) 395.9 ± 29.9 (8) NS

Jup Gamma Catenin (Junction

plakoglobin) 159.6 ± 6.0 (7) 202.7 ± 13.3 (8) NS

Cav1 Caveolin 1 441.5 ± 20.2 (8) 288.5 ± 35.6 (8) <0.01

Cav3 Caveolin 3 79.5 ± 4.1 (8) 89.1 ± 13.2 (8) NS

Cdh1 E-cadherin 0.1 ± 0.0 (6) 1.5 ± 0.7 (8) <0.05

Cdh2 N-cadherin 597.8 ± 25.7 (8) 565.5 ± 32.7 (8) NS

Dmd Dystrophin 106.3 ± 6.2 (8) 107.6 ± 7.2 (8) NS

Snta1 Syntrophin alpha 1 83.4 ± 4.6 (8) 71.0 ± 2.6 (7) <0.05

Snai1 Snail 1 0.2 ± 0.1 (6) 0.2 ± 0.1 (7) NS

Csk c-src tyrosine kinase 38.9 ± 2.2 (8) 76.9 ± 4.8 (7) <0.001

Ptk2 focal adhesion kinase 42.7 ± 1.9 (8) 45.8 ± 3.4 (8) NS

16


Cytoskeleton

Nppa ANP 403.7 ± 52.2 (8) 13432.6 ± 1767.6 (8) <0.001

Nppb BNP 1296.5 ± 273.1 (8) 4175.7 ± 436.3 (8) <0.001

Acta2 α-SMA 288.8 ± 11.6 (7) 753.3 ± 65.6 (8) <0.001

Acta1 α actin, s 2448.4 ± 521.8 (8) 1660.4 ± 363.6 (8) NS

Actc1 α actin, c 13921.0 ± 840.9 (8) 9273.0 ± 1003.2 (8) <0.01

Ank2 Ankyrin B 142.9 ± 5.7 (8) 89.5 ± 8.0 (8) <0.001

Cnn1 Calponin 1 2.9 ± 0.3 (8) 15.3 ± 2.1 (8) <0.001

Col1a1 Procollagen, 1, α1 422.2 ± 55.6 (8) 7050.6 ± 1161.3 (8) <0.001

Ttn Titin 2440.5 ± 131.8 (8) 1357.2 ± 76.6 (7) <0.001

Tnni3 TnIc 10576.9 ± 514.8 (8) 3630.1 ± 388.4 (8) <0.001

Tnni1 TnIs 0.3 ± 0.0 (8) 0.5 ± 0.1 (8) <0.05

Myh6 α-MHC 25531.4 ± 1640.5 (8) 9182.8 ± 1391.3 (8) <0.001

Myh7 β-MHC 296.6 ± 47.1 (8) 2521.9 ± 812.3 (8) <0.01

Myl4 MLC1A 46.2 ± 12.7 (8) 79.3 ± 17.0 (8) NS

Myl2 MLC2V 26739.9 ± 2064.4 (8) 8783.2 ± 953.7 (8) <0.001

Vim vimentin 554.3 ± 40.1 (8) 1843.1 ± 135.7 (8) <0.001

Vcl vinculin 129.6 ± 7.8 (8) 177.4 ± 9.0 (8) <0.01

17


Connexins

Gja5 Cx40 6.6 ± 0.5 (8) 8.1 ± 1.1 (8) NS

Gja1 Cx43 706.4 ± 33.2 (8) 603.3 ± 54.3 (8) NS

Gja7 Cx45 2.4 ± 0.4 (7) 4.3 ± 0.7 (8) <0.05

Sodium channels

Scn5a Nav1.5 309.2 ± 17.5 (8) 367.8 ± 28.2 (8) NS

Scn1b Navβ1 64.2 ± 2.0 (8) 130.5 ± 10.6 (8) <0.001

Potassium channels

Kcna2 Kv1.2 1.0 ± 0.1 (8) 0.9 ± 0.1 (7) NS

Kcna4 Kv1.4 0.9 ± 0.1 (8) 2.0 ± 0.3 (8) <0.01

Kcna5 Kv1.5 78.8 ± 5.5 (8) 42.5 ± 5.1 (8) <0.01

Kcnb1 Kv2.1 53.7 ± 4.0 (8) 45.5 ± 2.5 (8) NS

Kcnd2 Kv4.2 37.3 ± 1.9 (8) 16.9 ± 2.3 (8) <0.001

Kcnd3 Kv4.3 7.6 ± 0.5 (8) 6.7 ± 0.6 (8) NS

Kcnq1 KvLQT1 54.9 ± 2.6 (8) 44.8 ± 4.1 (8) NS

Kcnh2 m-erg 80.7 ± 8.8 (8) 43.2 ± 2.7 (8) <0.001

Kcnn1 SK1 7.8 ± 0.6 (8) 5.3 ± 0.5 (8) <0.01

Kcnn2 SK2 6.8 ± 0.7 (8) 2.8 ± 0.5 (8) <0.001

Kcnn3 SK3 2.9 ± 0.6 (8) 4.8 ± 1.2 (8) NS

Kcnma1 BigK 0.5 ± 0.1 (8) 1.8 ± 0.5 (8) <0.05

Kcnj2 Kir2.1 92.4 ± 3.4 (8) 37.2 ± 5.2 (8) <0.001

Kcnj12 Kir2.2 58.7 ± 1.3 (7) 45.2 ± 2.5 (8) <0.001

Kcnj4 Kir2.3 3.1 ± 0.4 (8) 1.0 ± 0.1 (8) <0.001

18


Potassium channels

Kcnj3 Kir3.1 52.1 ± 3.5 (8) 10.1 ± 2.0 (8) <0.001

Kcnj5 Kir3.4 34.5 ± 1.4 (7) 23.9 ± 1.1 (8) <0.01

Kcnj8 Kir6.1 55.2 ± 2.7 (8) 37.0 ± 5.3 (8) <0.05

Kcnj11 Kir6.2 102.9 ± 4.5 (7) 58.8 ± 4.8 (8) <0.001

Potassium channels auxiliary subunits

Kcne1 MinK 0.3 ± 0.1 (4) 1.0 ± 0.3 (8) NS

Kcne2 MIRP1 No Value No Value NS

Kcnip2 KChIP2 121.6 ± 5.3 (8) 41.6 ± 8.0 (8) <0.001

Abcc8 SUR1 42.5 ± 2.8 (8) 29.7 ± 2.2 (7) <0.05

Abcc9 SUR2 594.7 ± 24.7 (8) 247.4 ± 22.0 (8) <0.001

Calcium channels

Cacna1c Cav1.2 328.2 ± 20.7 (8) 188.2 ± 12.6 (8) <0.001

Cacna1d Cav1.3 0.5 ± 0.1 (8) 1.6 ± 0.2 (8) <0.001

Cacna1g Cav3.1 32.5 ± 2.5 (8) 15.3 ± 2.7 (8) <0.001

Cacna1h Cav3.2 0.4 ± 0.0 (7) 0.5 ± 0.1 (8) NS

Cacna2d1 Cavα2δ1 177.6 ± 8.7 (8) 179.1 ± 6.2 (8) NS

Cacna2d2 Cavα2δ2 2.4 ± 0.3 (8) 2.3 ± 0.5 (8) NS

Cacna2d3 Cavα2δ3 0.4 ± 0.1 (8) 0.4 ± 0.1 (8) NS

Cacnb1 Cavβ1 18.1 ± 1.2 (8) 41.8 ± 3.9 (8) <0.001

Cacnb2 Cavβ2 81.4 ± 6.1 (8) 61.2 ± 3.6 (7) NS

Cacnb3 Cavβ3 4.7 ± 0.2 (8) 16.6 ± 1.4 (8) <0.001

Cacng7 Cavγ7 8.5 ± 0.5 (8) 15.7 ± 0.7 (7) <0.001

19


Calcium handling and signaling

Calm1 Calmodulin1 372.8 ± 19.6 (8) 598.2 ± 55.0 (8) <0.001

Calm3 Calmodulin3 319.1 ± 7.4 (8) 421.2 ± 11.7 (7) <0.001

Casq2 Calsequestrin 957.0 ± 29.7 (8) 728.9 ± 37.7 (8) <0.001

Camk2d CAMK IIδ 160.2 ± 8.6 (8) 151.0 ± 5.3 (8) NS

Pln PLB 7937.0 ± 478.1 (8) 3351.4 ± 332.5 (8) <0.001

Itpr2 IP3R-2 46.0 ± 1.8 (7) 40.4 ± 3.7 (8) NS

Ryr2 RYR2 1201.6 ± 61.8 (8) 810.4 ± 85.5 (8) <0.01

Pumps and exchangers

Atp1a1 Na/K-ATPase, α1 843.4 ± 47.4 (8) 721.5 ± 17.3 (7) <0.05

Atp1a2 Na/K-ATPase, α2 514.6 ± 30.9 (8) 208.8 ± 38.7 (8) <0.001

Atp1b1 Na/K-ATPase, β1 643.9 ± 21.0 (8) 860.9 ± 28.3 (8) <0.001

Atp2a2 SERCA2 11917.0 ± 565.3 (8) 3925.1 ± 284.7 (7) <0.001

Slc8a1 NCX1 326.6 ± 11.2 (8) 379.4 ± 27.1 (8) NS

TRP channels

Trpc1 TRPC1 5.9 ± 0.3 (7) 5.6 ± 0.4 (8) NS

Trpc3 TRPC3 14.0 ± 0.8 (8) 7.8 ± 1.5 (8) <0.01

Trpc6 TRPC6 0.5 ± 0.1 (8) 1.0 ± 0.2 (8) <0.05

All data are expressed mean±S.E.M. of 2−∆Ct versus GAPDH (×100). Statistical comparison was by

Student t-test.

20

Supplemental Figure 1. Hierarchical clustering. Individual sample results are arranged vertically, genes horizontally. Green indicates lower-level expression, red higher-level. Samples are organized by the program in terms of similarity in gene-expression. All the control samples group together at the left; all the ILK-KO samples at the right. Results are from Taqman TLDA high throughput qPCR.

21

Supplemental Figure 2. A, B: Examples of single-cell action potentials recorded with perforated-patch methods for one cell from each group. C-E: Mean (horizontal bar) and individual-value phase-0 upstroke velocity (Vm, panel C), action potential amplitude (APA, panel D) and resting membrane potential (RP, panel E) from control (Ctl) and ILK-KO cells. Results for APA and RP were available from all cells; results for Vmax were available for those cells in which most of the phase-0 upstroke could be distinguished from the stimulus artifact. **P<0.01, ***P<0.001, multi-level mixed-effects model.

22

Supplemental Figure 3. A: Representative-whole cell voltage-dependent K+-currents evoked in isolated cardiomyocytes from control (left) and ILK-KO (right) mice during 4.5-second 10-mV incremental voltage-clamp steps to potentials between -60 mV and +60 mV from a holding potential of -70 mV. B: Representative recordings of IK1 obtained in isolated cardiomyocytes from control (left) and ILK-KO (right) mice by applying 300-ms voltage steps at 0.1 Hz from -120 mV to -10 mV with 10-mV increments and a holding potential of -40 mV. C: Representative L-type Ca2+-current (ICaL) recorded in isolated cardiomyocytes with 200-ms voltage steps at 0.1 Hz to voltages from -40 mV to +70 mV with 10-mV increments and a holding potential of -50 mV.

23

Supplemental Figure 4. Conventional qPCR analysis of ion-channel subunit expression. A: Mean (horizontal bar) and individual-heart mRNA-expression of Kv1.2, Kv1.4, Kv1.5, Kv2.1, Kv4.2, Kv4.3 and KChIP2 subunits in left-ventricular (LV) tissues. B: Mean (horizontal bar) and individual-heart Kir2.1, Kir2.2 and Kir2.3 mRNA-expression in LV-tissues. C: Mean (horizontal bar) and individual-heart Cav1.2 and Cav1.3 mRNA-expression in left-ventricular (LV) tissues. *P<0.05, **P<0.01, ***P<0.001, two-factor ANOVA for non-repeated measures with Bonferroni-corrected t-tests (nc=6 for comparisons).

24

Supplemental Figure 5. Mean (horizontal bar) and individual-heart mRNA-expression. *P<0.05, **P<0.01, P<0.001 by Bonferroni-adjusted t-tests (nc=6 for comparisons). The group*time interaction was significant for all 3 effects (for Col1a1 P<0.0001; for ANP, P=0.001; for BNP P=0.01).

25

Supplemental Figure 6. Co-immunoprecipitation between ILK and the various ion-channel subunits shown. This result is representative of 3 similar experiments performed with 3 washes. Because of the faint band seen in the ILK-KO and IgG lanes for Kv4.2, we repeated that experiment with 5 washed (n=4 experiments). As shown, the potentially unspecific bands disappeared and the Kv4.2 band remained.

Documents

Loss of Cardiomyocyte Integrin Linked Kinase Produces …circep.ahajournals.org/content/circae/early/2015/06/12/CIRCEP.115... · Loss of Cardiomyocyte Integrin Linked Kinase Produces