increased lysyl oxidase expression and collagen cross-linking during atrial fibrillation
TRANSCRIPT
Journal of Molecular and Cellular Cardiology 50 (2011) 678–685
Contents lists available at ScienceDirect
Journal of Molecular and Cellular Cardiology
j ourna l homepage: www.e lsev ie r.com/ locate /y jmcc
Original article
Increased lysyl oxidase expression and collagen cross-linking during atrial fibrillation
Oliver Adam a,⁎,1, Katharina Theobald a,1, Daniel Lavall a, Markus Grube b, Heyo K. Kroemer b,Sabine Ameling c, Hans-Joachim Schäfers d, Michael Böhm a, Ulrich Laufs a
a Klinik für Innere Medizin III, Kardiologie, Angiologie und Internistische Intensivmedizin, Universitätsklinikum des Saarlandes, Homburg/Saar, Germanyb Department of Pharmacology, Ernst Moritz Arndt University, Greifswald, Germanyc Department of Functional Genomics, Ernst Moritz Arndt University, Greifswald, Germanyd Klinik für Thorax- und Herz-Gefäßchirurgie; Universitätsklinikum des Saarlandes, Homburg/Saar, Germany
⁎ Corresponding author. Klinik für Innere Medizin IInternistische Intensivmedizin, Universitätsklinikum deSaar, Germany. Tel.: +49 6841 1623000.
E-mail address: [email protected] (O. Adam).1 Both authors contributed equally.
0022-2828/$ – see front matter © 2011 Elsevier Ltd. Aldoi:10.1016/j.yjmcc.2010.12.019
a b s t r a c t
a r t i c l e i n f oArticle history:Received 28 September 2010Received in revised form 20 December 2010Accepted 30 December 2010Available online 6 January 2011
Keywords:Atrial fibrillationRac1CTGFLysyl oxidaseFibrosis
The aim of the study is to characterize the signal transduction leading to interstitial fibrosis in the patho-genesis of atrial fibrillation (AF) and atrial remodeling.Samples of the left atrial appendage (LA) from patients with AF showed higher collagen content (73±5 vs.38±2 μg/mg protein) and 2.5-fold increased collagen crosslinking compared to patients with sinus rhythm(SR). Affymetrix-assays, RT-PCR and western Blot analysis revealed that LA of AF patients are characterized byincreased lysyl oxidase (LOX) mRNA (218±42%) and protein (253±11%) expression. This was associatedwith increased expression of connective tissue growth factor (CTGF), fibronectin and Rac1 activity comparedto SR. In neonatal cardiac fibroblasts, the Rac1 specific small molecule inhibitor NSC23766 preventedangiotensin II (AngII) induced upregulation of LOX (214±16%) expression. Inhibition of CTGF by siRNAtransfections completely inhibited AngII induced LOX expression. The LOX specific small molecule inhibitorBAPN prevented AngII and CTGF induced fibronectin expression. Left atria of transgenic mice with cardiacoverexpression of Rac1 (RacET) that develop AF at high age exhibited upregulation of CTGF as well as LOX(187±7%) and fibronectin (627±146%) expression. Atria of RacET showed increased collagen content(28±2 μg/mg protein) and crosslinking (10±0.7) compared to wildtypes (20±0.4 μg/mg protein; 5±0.9).Left atrial myocardium of patients with atrial fibrillation is characterized by increased lysyl oxidase andfibronectin expression as well as collagen cross-linking. In cardiac fibroblasts, Rac1 GTPase mediatesupregulation of fibronectin via LOX and CTGF. Inhibition of this signaling pathway may therefore represent atarget for the prevention of fibrotic atrial remodeling.
II, Kardiologie, Angiologie unds Saarlandes, 66424 Homburg/
l rights reserved.
© 2011 Elsevier Ltd. All rights reserved.
1. Introduction
Atrial fibrillation (AF) remains the most common arrhythmia inclinical practice and a major cause of morbidity and mortality [1].Fibrosis is a hallmark of atrial arrhythmogenic structural remodeling[2–4]. Tissue fibrosis results from an accumulation of fibrillar collagendeposits, occurring most commonly as a reparative process to replacedegenerating myocardial parenchyma with concomitant reactivefibrosis, which causes interstitial expansion [5]. Increased collagendeposition has been documented in lone-AF patients compared withindividuals in sinus rhythm [6], as well as in patients with secondaryAF [7].
Collagen is synthesized and secreted by fibroblasts and myofibro-blasts as a precursor, procollagen. Its amino-terminal and carbox-yterminal propeptides are cleaved to yield the triple helicalmonomers of collagen by specific procollagen proteinases [8]. Afterthese proteolytic reactions, the collagen molecules are rapidly andspontaneously assembled into collagen fibrils. Fibrillar collagen typesI and III must be cross-linked to form the final collagen types I and IIIfibers that are highly resistant against proteolytic enzymes and possesthe physical properties of tensile strength [9]. The copper (Cu)-dependent amine oxidase lysyl oxidase (LOX) catalyzes the oxidationof the ε-amino groups in lysine or hydroxylysine residues, resulting inthe formation of corresponding aldehydes. Two such aldehydes canthen spontaneously react with each other. Alternatively, one aldehydecan bind to another ε-amino group; both mechanisms produce cross-links connecting two polypeptide chains [10]. LOX therefore plays animportant role for collagen cross-linking.
Recent studies form our lab have shown that the fibrosis of left atria(LA) of patients with AF is associated with increased expression andactivity of the small Rho-GTPase Rac1 and increased expression of the
Table 1Patient characteristics. ACE-I = angiotensin-converting enzyme inhibitor; AT antag-onist = angiotensin receptor I antagonist; FS = fractional shortening; IVSd =interventricular septum; LA = left atrium; LVEDd = left ventricular end-diastolicdiameter; LVEF = left ventricular ejection fraction; LVESd = left ventricular end-systolic diameter.
SR Sem AF Sem P
No. 10 10Male(female) 5(5) 5(5)Age [ y] 66 6.6 72 2.5 nsHeight [cm] 163 4.6 174 2.7 nsWeight [kg] 82 3.2 85.0 4.2 nsLA [mm] 53.2 2.4 54.2 5.2 nsLVEF [%] 54.6 9.2 58.7 3.4 nsLVEDd [mm] 56.7 4.7 54.7 3.2 nsLVESd [mm] 41.5 5.2 40.6 8.3 nsIVSd [mm] 12.0 1.0 12.1 0.9 nsLVPWd [mm] 12.5 1.2 11.8 1.5 nsBeta-blocker 7 7ACE-I or AT antagonist 8 8Aldosterone antagonist 1 1Calcium antagonist 1 1Diuretic 7 10Nitrate 0 1Digitalis 0 2Amiodaron 0 0Sotalol 0 0Statin 0 0
679O. Adam et al. / Journal of Molecular and Cellular Cardiology 50 (2011) 678–685
pro-fibrotic cytokine connective tissue growth factor (CTGF) [11,12].Our previously reported Affymetrix expression analysis of human LAhas suggested that LOX gene-expression is increased in AF patients [12].The present study was undertaken to follow up on these observationsbecause the role of LOX in atrial fibrillation was completely unknown—
despite its importance for collagen cross-linking. Therefore, the aim ofthis study was to characterize the lysyl oxidase in left atrialmyocardium from patients with AF and sinus rhythm (SR), in culturedprimary cardiac cells and in Rac1 transgenic mice.
2. Materials and methods
2.1. Cell isolation and culture
Cardiac fibroblasts were isolated from 5 days old neonatal Sprague–Dawley rat hearts [13]. Purity of fibroblasts was confirmed by vimentinstaining (data not shown). Cells were used for experiments after3–6 days of culture.
2.2. Animal studies
Micewith cardiac overexpression of constitutively active (V12) Rac1under the control of the α-myosin heavy chain (MHC) promoter(RacET) and wild type controls (WT) promoter [14] were fed withnormal chow (ssniff, Germany) or normal chow supplemented with0.4 mg/d of commercially available rosuvastatin (Crestor, AstraZeneca)for 10 months. The study was not funded by the pharmaceuticalindustry. At the age of 10 months 44% of RacET showed AF compared to20% of RacET+statin treatment (pb0.05). None of the wildtype miceexhibited atrial arrhythmias [11,12,15]. The study was approved bythe animal ethics committee of the Universität des Saarlandes and isin accord with the Guide for the Care and Use of Laboratory Animalspublished by the US National Institutes of Health (NIH Pub. No. 85-23,revised 1996).
2.3. Human left atrial tissue
Tissue samples of the left atrial appendage of patients undergoingmitral valve surgery were analyzed in 10 patients with sinus rhythm(SR) and 10 patients with permanent atrial fibrillation (AF; docu-mented by ECG for N3 months). There was no difference in atrialdiameter, left ventricular function and medication between bothgroups (Table 1). The patients did not receive drugs at least 12 hbefore surgery, no patient was treated with a statin. The analysis wasapproved by the ethics committee of the Ärztekammer des Saarlandes(No.131/00). The investigation is conformwith the principles outlinedin the Declaration of Helsinki.
2.4. Affymetrix expression analysis
As described previously, transcriptional profiling of atrial tissue ofthe AF and SR patients was performed using GeneChip-HumanGenome-HGU 133-Plus 2.0-arrays (Affymetrix, Santa Clara, USA)[12]. Affymetrix HGU133-Plus 2.0 arrays query the expression of54,675 probe sets. After analysis with Gene Chip Operating software(GCOS), expression raw data were transferred to GeneSpring GXversion 7.3.1. (Agilent Technologies) to perform per-chip normaliza-tion for signals that were flagged in GCOS as present or marginal(P/M). Cyber-T was used for the identification of genes differentiallyexpressed between controls and AF patients. Application of Cyber-Tresults in an average ratio and a p-value for each probe set.
2.5. Angiotensin II concentration
Angiotensin II concentration in left atrial myocardium was deter-mined by ELISA using Assay Max from Assay Pro; USA.
2.6. Western analysis
Total protein lysates were prepared as described [13]. Immuno-blotting was performed using Anti-Rac1 (Rac1, Upstate, clone 23A8),LOX (Abcam), fibronectin (Sigma) and tubulin (Santa Cruz). Immu-nodetection was accomplished using goat anti-rabbit or goat anti-mouse secondary antibody (1:4000 dilution, Sigma) and an enhancedchemiluminescence kit (Amersham) followed by densitometry.Expression was calculated using a standard curve generated withrecombinant Rac1 protein (Upstate).
2.7. siRNA-transfection
Transfection of rat neonatalfibroblastswith CTGF siRNA (sc-39330,Santa Cruz; USA) was performed according to the manufacturesinstructions, fluorescin conjugated scrambled siRNA (sc-36869) wasused as control.
2.8. Reverse-transcriptase polymerase-chain reaction (RT-PCR)
Total RNA isolation, reverse transcription, and competitive PCRwas performed according to standard techniques. The sense, (5′-TTACCCAGCCGACCAAGATA-3′) and anti-sense (5′-CTGAGGCATACG-CATGATGT-3′) primers were used to amplify a 399-bp LOX cDNAfragment. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH)was amplified as an external standard. Each PCR cycle consistedof denaturing at 94 ° C for 30 s, annealing at 53 ° C for 30 s, andelongation at 72 ° C for 60 s. The linear exponential phases for LOXand GAPDH PCR were 40 and 22 cycles, respectively. Equal amountsof corresponding LOX and GAPDH RT-PCR products were loaded on1.5% agarose gels and optical densities of ethidium-bromide stainedDNA bands were quantitated.
2.9. Histological analysis
10 μm cryosections were stained with 0.1% Sirius Red F3BA(Polysciences). Lucia Measurement Version 4.6 software was usedfor quantification of interstitial fibrosis.
Fig. 1. Increased content and cross-linking of collagen in left atria of patients with AF.(A) Quantification of soluble, insoluble and total collagen as well as (B) the degree ofcollagen cross-linking in left atrial myocardium of patients with atrial fibrillation (AF)and sinus rhythm (SR); n=10 per group; *pb0.05.
680 O. Adam et al. / Journal of Molecular and Cellular Cardiology 50 (2011) 678–685
2.10. Immunofluorescence analysis
Indirect immunofluorescencewas performed onparaffin-embeddedsections of the human left atria applying monoclonal antibodiesagainst α-sarcomeric actin (cardiomyocytes, Sigma-Aldrich, Germany),LOX (Abcam,) and vimentin (fibroblasts, Dako Cytomation, CA, USA).FITC- or TRITC-conjugated (Dianova, Germany) anti-mouse IgM, anti-mouse IgG, anti-goat IgG were used as secondary antibodies. Sectionswere counterstained with DAPI (Calbiochem, Germany). Immunostain-ingwith FITC- or TRITC-conjugated anti-mouse IgM, anti-mouse IgG andanti-goat IgG as a negative control for non-specific staining wasperformed in parallel sections.
2.11. Immunohistochemical studies
To distinguish between cross-linked (insoluble) and noncross-linked (soluble) collagen, a colorimetric procedure was used [16,17].Total collagen content per mg of protein was quantified in tissuesections prepared from formalin-fixed paraffin-embedded samplesstained with sirius red F3BA (collagen components) and Fast greenFCF (noncollagenous components), respectively. Both dyes wereeluted readily and simultaneously with NaOH-methanol and theabsorbances obtained at 540 and 605 nm were used to determine theamount of collagen and protein [17]. The myocardial pepsin-solublecollagens were extracted overnight with 0.1 mg/mL pepsin in0.5 mol/L acetic acid. The soluble collagen was quantified with aSircol collagen assay kit with soluble type I collagen as a standard(Tebu-Bio, USA). The amount of insoluble collagen was calculated bysubtraction of soluble collagen from total collagen. The degree ofcross-linking was calculated as the ratio between the insoluble andthe soluble forms of collagen.
2.12. Statistical analysis
Band intensities were analyzed by densitometry. Mean, standarderror of the mean (SEM) and statistical analyses were calculated bySigma Stat software, version 2.0. Unpaired Student's t tests and in caseof failing normality test, Mann–Whitney Rank Sum test, for singlecomparison, was performed. For multiple comparisons, ANOVAfollowed by Newman–Keuls post-hoc analysis for multiple compar-isons was applied. Differences were considered significant at pb0.05.
3. Results
3.1. Increased content and cross-linking of collagen in left atria ofpatients with AF
Left atria of patients with AF are characterized by increasedinterstitial fibrosis [11,12]. To further characterize these previousfindings, the collagen composition and collagen crosslinking in thehuman samples were quantitated. LA of patients with AF showedincreased content of total collagen (n=10 per group, 73±5 vs. 38±2 μg/mg protein, pb0.05) characterized by an increased content ofinsoluble collagen (70±6 vs. 33±3 μg/mg protein, pb0.05) com-pared to SR whereas the amount of soluble collagen content wassimilar (Fig. 1A), demonstrating increased collagen crosslink in AFpatients (26±5 vs. 9±2 insoluble/soluble collagen, pb0.05, Fig. 1B).
3.2. Increased LOX expression in left atria of patients with AF
Lysyl oxidase is a key enzyme for collagen cross-linking [10], butthe role of LOX in atrial fibrillation has not been investigated.Screening of ~25000 genes by Affymetrix arrays [12] showed thatLOX is one out of 122 genes displaying a greater than 2-fold changes inLA of patients with AF compared to SR (pb0.01). Upregulation of LOXmRNA expression was confirmed by RT-PCR (Fig. 2A, 218±42%,
pb0.05 compared to SR) andwestern blot analysis (Fig. 2B, 253±11%,pb0.001). Immunofluorescence staining of LA from patients with AFshowed a higher expression of LOX compared to SR primarily due toincreased expression in cardiac fibroblasts, where as cardiac myocytesshowed only weak LOX expression (Fig. 2C–E).
3.3. Angiotensin II regulates lysyl oxidase expression in neonatal cardiacfibroblasts via activation of Rac1
ELISA showed increased tissue concentrations of Angiotensin II(Ang II) in LA of AF patients (Fig. 3A, 700±77% vs. SR, pb0.001). TheRac1-GTPase is a mediator of Ang II in the cardiovascular system[13,18]. In order to test whether the Rac1-GTPase plays a causal rolefor the observed effects, neonatal rat cardiac fibroblasts werestimulated with Ang II 1 μM for 3 h after preincubation with orwithout the Rac1 specific small molecule inhibitor NSC23766(100 nM, 21 h). NSC prevents Rac1 activation by inhibition of theRac-specific guanine nucleotide exchange factors (GEFs) TrioN andTiam1 without affecting Cdc42 or RhoA activation. Western blotanalysis showed upregulation of LOX expression after stimulationwith Ang II to 214±16% which was completely prevented bypreincubation with the small molecule inhibitor of Rac1 (84±16%,Fig. 3B, pb0.05). These data show that Rac1-GTPase is important forthe angiotensin II induced regulation of lysyl oxidase.
3.4. Lysyl oxidase expression is regulated via activation of CTGF by Rac1
Because Rac1 regulates CTGF expression in cardiac fibroblasts [12],we tested whether LOX is regulated by CTGF. Cultured neonatal ratcardiac fibroblasts were treated with recombinant CTGF (1 ng/ml; 1 h)
Fig. 2. Increased LOX expression in left atria of patients with AF. (A) Quantification and representative RT-PCR of lysyl oxidase (LOX) mRNA related to GAPDH expression in leftatrium of patients with SR or AF; *pb0.05. (B) Quantification and representative western blot of LOX protein related to tubulin expression in the left atrium of patients with SR or AF;n=10, **pb0.001. (C) Representative immunofluorescent imaging of LOX protein (red), the myocyte marker α-sarcomeric actin (green) and cell nuclei (DAPI blue) in the leftatrium of patients with SR or AF at 10-fold magnification. (D) Representative immunofluorescent imaging of LOX protein (red), the myocyte marker α-sarcomeric actin (green) andcell nuclei (DAPI blue) in the left atrium of patients with SR or AF at 100-fold magnification. (E) representative immunofluorescent imaging of LOX protein (red), the fibroblastmarker vimentin (green) and cell nuclei (DAPI blue) in the left atrium of patients with SR or AF at 100-fold magnification.
681O. Adam et al. / Journal of Molecular and Cellular Cardiology 50 (2011) 678–685
which resulted in an upregulation of LOX protein expression (Fig. 4A,218±34% of vehicle treatment, pb0.05). In contrast, expression ofRac1-GTPase was not affected by CTGF treatment (Fig. 4B). To clarifythe role of CTGF in this signal transduction, neonatal cardiac fibroblastswere transfected with siRNA for CTGF. SiRNA-mediated knockdown ofCTGF prevented angiotensin-mediated upregulation of LOX (84±10%,pb0.05 vs Ang+sc-CTGF; Fig. 4C).
3.5. Collagen content, collagen cross-linking and LOX expression isdecreased via reduction of Rac1 activity by statins in vivo
To further characterize the interaction of Rac1, LOX and collagencrosslink during AF in vivo, transgenic mice with cardiac over-expression of constitutively active (V12) Rac1 under the control of theαMHC promoter (RacET) were studied. These mice spontaneouslydevelop AF at high age [11]. RacET are characterized by markedupregulation of Rac1 activity compared to WT [11]. Treatment withHMG-CoA reductase inhibitors (statins) is able to inhibit Rac1activation by inhibiting its geranylgeranylation and membranetranslocation [11,18,19]. Therefore we tested the effect of treatment
with rosuvastatin 0.4 mg/d po for 10 months on LOX expression,collagen composition and collagen crosslink. Interestingly, RacETexhibited a marked increase of LOX expression (Fig. 5A; 187±7% vs.WT; pb0.001), which was observed in RacET in AF as well as in RacETthat were still in SR. Statin treatment reduced LOX expression to97±17% compared to WT controls. Furthermore, statin treatmentresulted in reduced deposition of unsoluable collagen in 10 month oldRacET mice (Fig. 5B, RacET 28±2 vs. RacET+Statin 22±0.8 μg/mgprotein, pb0.05) and in reduced collagen cross-linking (Fig. 5C,RacET 10±0.7 vs. RacET+Statin 6±0.3 insoluble/soluble collagen,pb0.05).
3.6. Fibronectin expression is regulated via LOX by Rac 1 and CTGF
Fibronectin has been characterized as an extracellular matrixglycoprotein that regulates several cellular functions such asproliferation, differentiation, migration, adhesion and apoptosis[20,21]. The formation of the fibronectin matrix glycoprotein hasbeen shown to be critical for the subsequent assembly of types I and IIIcollagen fibrils [22,23]. Previous evidence suggests that there is an
Fig. 3. Angiotensin II regulates lysyl oxidase expression in neonatal cardiac fibroblastsvia activation of Rac1. (A) Quantification of angiotensin II (Ang II) concentration inatrial myocardium detected by ELISA in left atrium of patients with sinus rhythm (SR)or atrial fibrillation (AF) (n=5, **pb0.001). (B) Effects of treatment with angiotensin II(1 μM, 3 h) and the specific Rac1 small molecule inhibitor NSC23766 (NSC, 100 nM;21 h) on LOX expression in relation to tubulin in neonatal rat cardiac fibroblasts, n=5,*pb0.05 vs. control (C), #pb0.05 vs. Ang II.
Fig. 4. Lysyl oxidase expression is regulated via activation of CTGF by Rac1.Quantification and representative western blots showing the effects of treatmentwith recombinant connective tissue growth factor (CTGF, 1 ng/ml, 1 h) on (A) lysyloxidase and (B) Rac1 protein expression in neonatal rat cardiac fibroblasts related totubulin; n=5, *pb0.05 vs control (C). (C) Quantification and representative westernblots showing the effect of CTGF knockdown by siRNA-transfection (si-CTGF;sc-CTGF=scrambled RNA) on LOX expression in neonatal rat cardiac fibroblastsrelated to tubulin, n=4, *pb0.05.
682 O. Adam et al. / Journal of Molecular and Cellular Cardiology 50 (2011) 678–685
interaction between LOX and fibronectin in human neonatal skinfibroblasts [24]. Western analysis revealed that LA of AF patients arecharacterized by increased fibronectin expression (Fig. 6A; 236±38%vs. SR, pb0.05). Interestingly, RacET mice exhibited increasedfibronectin protein expression as well (Fig. 6B, 627±146% vs. WT,pb0.05) which was reduced by statin treatment (206±54% vs. WT,pb0.05).
To clarify the underlying signaling, neonatal rat cardiac fibroblastswere treated with recombinant CTGF (1 ng/ml; 1 h) or Ang II (1 μMfor 3 h) after pre-incubation with or without the LOX specific smallmolecule inhibitor b-aminopropionitrile (BAPN, 100 nM, 21 h). BothAng II and CTGF increased fibronectin protein expression (Fig. 6C).Both effects were completely prevented by preincubation with BAPN.These data indicate that fibronectin expression is regulated via Rac1,CTGF and LOX.
4. Discussion
The data demonstrate for the first time that lysyl oxidase contri-butes to structural remodeling in atrial fibrillation. Specifically,angiotensin II activates LOX via activation of the small G proteinRac1-GTPase and the pro-fibrotic connective tissue growth factor. Thisnovel mechanism contributes to the signal transduction of structuralremodeling in left atrial myocardium leading to collagen cross-linkingand interstitial fibrosis.
Atrial fibrosis plays a fundamental role in the pathology of atrialfibrillation [2,3]. Extracellular matrix (ECM) volume and compositioncorrelate with AF persistence [25]. ECM proteins are produced pre-dominantly by cardiac fibroblasts that surround the cardiac myocytes,however, the detailed mechanism(s) leading to atrial fibrosis areincompletely understood. Several growth factors regulate fibroblastproliferation and ECM synthesis. The Cu-dependent lysyl oxidase(LOX) is essential for the formation of ECM proteins including type I,II, and III procollagens [26,27]. LOX is responsible for cross-linking ofcollagens and mediates soluble collagen molecules into insoluble
fibrous organization [26,28,29]. An increased expression and activityof LOX has been demonstrated in left ventricular myocardial fibrosis[27] and in left ventricles of patients with dilated cardiomyopathy[30]. However, the role of LOX in atrial myocardium has not beenstudied. LOX is synthesized as a 50 kDa pro-enzyme that undergoespost-translational modification in the endoplasmic reticulum andGolgi apparatus and is subsequently secreted into the extracellularspace, where it is processed to form the 30 kDa mature active enzyme[31,32]. Here, expression analysis revealed that the atria of patientswith AF are characterized by up-regulation of LOX expression. This isassociated with increased collagen content and collagen crosslinking.
The left atria of patients in AF exhibited increased tissueconcentrations of angiotensin II. Ang II is a known activator of fibrosisand regulator of structural remodeling [33–36] and high-densityoligonucleotide arrays showed that angiotensin I-receptor expressionis higher in human atrial myocardium compared to ventricular [37].This could be one reason, why atria are more susceptible to ANG-II-induced fibrosis. The Rac1-GTPase is a mediator of Ang II in thecardiovascular system [13,18]. Previous data from our lab have shownthat the small GTPase Rac1 and the connective tissue growth factor(CTGF) contribute to the signal transduction during the pathogenesisof atrial fibrillation [11,12]. Based on these data, further experimentswere performed to characterize the newly identified role of lysyloxidase in the context of atrial fibrosis. The experiments in neonatalcardiac fibroblasts showed upregulation of LOX by angiotensin II, aknown upstream mediator of CTGF [38]. This effect was completelyinhibited by a Rac1 specific small molecule inhibitor. Importantly,stimulation with recombinant CTGF lead to increased expression of
Fig. 5. Collagen content, collagen cross-linking and LOX expression is decreased via reduction of Rac1 activity by statins in vivo. Transgenic mice with cardiac overexpression ofconstitutively active (V12) Rac1 under the control of the α-myosin heavy chain (MHC) promoter (RacET) were treated with rosuvastatin 0.4 mg/d p.o. (RacET+Statin) or regular chowfor 10 months and compared to 10 months oldwild type controls (WT). (A) Quantification of LOX protein expression related to tubulin; (B) representativewestern blots; n=5 per group,*pb0.05 vsWT, #pb0.05 vs RacET. (C) Quantification of soluble, insoluble and total collagen as well as (D) the degree of collagen cross-linking in RacET and RacET+Statin compared toWT; n=5 per group, *pb0.05 vs. WT, #pb0.05 vs RacET.
683O. Adam et al. / Journal of Molecular and Cellular Cardiology 50 (2011) 678–685
LOX. In addition, transfection of cardiac fibroblasts with smallinterfering RNA for CTGF completely prevented Ang II inducedupregulation of LOX expression. Taken together the data identifyRac1 and CTGF as mediators of angiotensin II induced regulation ofLOX in cardiac fibroblasts.
Our cell culture model of ventricular rat neonatal cells hassignificant limitations regarding the study of the signaling duringatrial fibrillation, which in vivo is regulated e.g. by the interaction ofseveral cell types, mechanical forces and electrical stimulation. Pilotexperiments show a similar upregulation of LOX by CTGF andangiotensin in neonatal atrial fibroblasts compared to the ventricularfibroblasts. In our study, cultured cells are applied to focus on specificsteps in the signal cascade, and we are confident that the neonatalventricular cell can be used for this purpose. However, in order to gaininformation about the in vivo situation in atrial myocardium, we usedthe human samples as well as transgenic mice over-expressing Rac1under the control of the MHC promoter. The RacET mice arecharacterized by marked elevation of CTGF [12]. Western analysisshowed increased LOX expression in LA of RacETmice associated withincreased total collagen content and collagen cross-linking. Impor-tantly, these animals develop atrial fibrillation at older age [11,12].
Rac1 GTPase is inhibited by HMG-CoA reductase inhibitors (statins)which reduce isoprenylation-dependent small G-protein function[13,19,39]. Indeed, statin treatment reduced the expression of LOX
and decreased collagen cross-linking in the RacET mice. This effect wasassociated with the reduction of atrial fibrillation in this animal model.
Recent evidence suggests an interaction between LOX andfibronectin, an extracellular matrix glycoprotein and known regulatorof ECM [20–24]. Fibronectin interacts with several proteins includingcollagens, fibrin, thrombospondin, cell surface integrins, heparinsulfate proteoglycans, and tenascin-C. In tissues, activated fibronectinis deposited and organized into a polymeric matrix, which has beenobserved to be associatedwith collagen fibers. Recently, the formationof the fibronectin matrix has been shown to be critical for thesubsequent assembly of types I and III collagen fibrils in FN knock-outmouse fibroblasts [22,23]. Here, the data show that both LA of AFpatients and RacET mice are characterized by increased fibronectinexpression. These data are in agreement with former studies showingincreased fibronectin expression in rapidly-paced cardiomyocytemedium and in atrial tissue from atrial-tachypaced dogs withventricular rate control [40] as well as in LA of pigs after 3–4 weeksof continuous rapid atrial pacing [41]. Importantly, statin treatmentreduced fibronectin protein expression in our 10 months old RacETmice in vivo. Preincubation with the LOX specific small moleculeinhibitor b-aminopropionitrile completely prevented Ang II and CTGFinduced fibronectin expression in neonatal cardiac fibroblasts. Thesedata indicate that fibronectin expression is regulated via Rac1, CTGFand LOX.
Fig. 6. Fibronectin expression is regulated via LOX by Rac 1 and CTGF. (A) Quantification and representative western blot of fibronectin (FN) protein related to tubulin expression inleft atrium of patients with SR or AF; n=5,*pb0.05, and (B) fibronectin expression in transgenic mice with cardiac overexpression of constitutively active (V12) Rac1 under thecontrol of the α-myosin heavy chain (MHC) promoter (RacET) and RacET treated with rosuvastatin 0.4 mg/d p.o. for 10 months (RacET+Statin) compared to their wild typecontrols (WT), n=5 per group, *pb0.05 vs WT, #pb0.05 vs RacET. (C) Effects of treatment with angiotensin II (Ang II, 1 μM, 3 h), recombinant CTGF and the specific LOX smallmolecule inhibitor b-aminopropionitrile (BAPN, 100 nM; 21 h) on the expression of fibronectin in neonatal rat cardiac fibroblasts, n=5, *pb0.05 vs. control (C), #pb0.05 vs. Ang II ,**pb0.05 vs. CTGF.
684 O. Adam et al. / Journal of Molecular and Cellular Cardiology 50 (2011) 678–685
In summary, left atria of patients with atrial fibrillation arecharacterized by increased expression of lysyl oxidase and collagencross-linking. The studies of cultured cardiac fibroblasts and RacETmice show that angiotensin II increases LOX and fibronectinexpression via activation of Rac1 GTPase and CTGF. Inhibition ofRac1 GTPase by statin treatment reduced LOX, fibronectin andcollagen cross-linking in vivo resulting in a reduced prevalence of AFin RacET mice. We therefore believe that further characterization ofthe signal transduction of atrial structural remodeling in thepathogenesis of AF has the potential to identify targets for theupstream therapy to prevent AF [42,43].
Sources of funding
This study was supported by the Deutsche Forschungsgemeinschaft(KFO 196), the Universität des Saarlandes (HOMFOR), the EuropeanStroke Network (ESN) and the Ministerium für Wirtschaft undWissenschaft des Saarlandes.
Disclosures
All authors have no disclosures or conflict of interest.
Acknowledgments
We thank Simone Jäger and Ellen Becker for their excellenttechnical assistance.
References
[1] Nattel S, Opie LH. Controversies in atrial fibrillation. Lancet 2006;367(9506):262–72.
[2] Allessie M, Ausma J, Schotten U. Electrical, contractile and structural remodelingduring atrial fibrillation. Cardiovasc Res 2002;54(2):230–46.
[3] Li D, Fareh S, Leung TK, Nattel S. Promotion of atrial fibrillation by heart failure indogs: atrial remodeling of a different sort 1. Circulation 1999;100(1):87–95.
[4] Nattel S, Burstein B, Dobrev D. Atrial remodeling and atrial fibrillation:mechanisms and implications. Circ Arrhythm Electrophysiol 2008;1(1):62–73.
[5] Silver MA, Pick R, Brilla CG, Jalil JE, Janicki JS, Weber KT. Reactive and reparativefibrillar collagen remodelling in the hypertrophied rat left ventricle: twoexperimental models of myocardial fibrosis. Cardiovasc Res 1990;24(9):741–7.
[6] Frustaci A, Chimenti C, Bellocci F, Morgante E, Russo MA, Maseri A. Histologicalsubstrate of atrial biopsies in patients with lone atrial fibrillation. Circulation1997;96(4):1180–4.
[7] Boldt A, Wetzel U, Lauschke J, Weigl J, Gummert J, Hindricks G, et al. Fibrosis in leftatrial tissue of patients with atrial fibrillation with and without underlying mitralvalve disease. Heart 2004;90(4):400–5.
[8] Prockop DJ, Kivirikko KI. Collagens: molecular biology, diseases, and potentials fortherapy. Annu Rev Biochem 1995;64:403–34.
685O. Adam et al. / Journal of Molecular and Cellular Cardiology 50 (2011) 678–685
[9] Reiser K, McCormick RJ, Rucker RB. Enzymatic and nonenzymatic cross-linking ofcollagen and elastin. FASEB J 1992;6(7):2439–49.
[10] Kagan HM, Trackman PC. Properties and function of lysyl oxidase. Am J Respir CellMol Biol 1991;5(3):206–10.
[11] Adam O, Frost G, Custodis F, Sussman MA, Schäfers HJ, Böhm M, et al. Role of Rac1GTPase activation in atrial fibrillation. J Am Coll Cardiol 2007;50(4):359–67.
[12] Adam O, Lavall D, Theobald K, Hohl M, Grube M, Ameling S, et al. Rac1-inducedconnective tissue growth factor regulates connexin 43 and N-cadherin expressionin atrial fibrillation. J Am Coll Cardiol 2010;55(5):469–80.
[13] Custodis F, Eberl M, Kilter H, Böhm M, Laufs U. Association of RhoGDI alpha withRac1 GTPase mediates free radical production during myocardial hypertrophy.Cardiovasc Res 2006;71(2):342–51.
[14] Sussman MA, Welch S, Walker A, Klevitsky R, Hewett TE, Price RL, et al. Alteredfocal adhesion regulation correlates with cardiomyopathy in mice expressingconstitutively active rac1. J Clin Invest 2000;105(7):875–86.
[15] Reil JC, Hohl M, Oberhofer M, Kazakov A, Kaestner L, Mueller P, et al. Cardiac Rac1overexpression in mice creates a substrate for atrial arrhythmias characterized bystructural remodelling. Cardiovasc Res 2010;87(3):485–93.
[16] Li YY, Feng Y, McTiernan CF, Pei W, Moravec CS, Wang P, et al. Downregulation ofmatrix metalloproteinases and reduction in collagen damage in the failing humanheart after support with left ventricular assist devices. Circulation 2001;4; 104(10):1147–52.
[17] Lopez-De Leon A, Rojkind M. A simple micromethod for collagen and total proteindetermination in formalin-fixed paraffin-embedded sections. J HistochemCytochem 1985;33(8):737–43.
[18] Liao JK, Laufs U. Pleiotropic effects of statins. Annu Rev Pharmacol Toxicol2005;45:89–118.
[19] Laufs U, Kilter H, Konkol C, Wassmann S, BöhmM, Nickenig G. Impact of HMG CoAreductase inhibition on small GTPases in the heart. Cardiovasc Res 2002;53(4):911–20.
[20] Magnusson MK, Mosher DF. Fibronectin: structure, assembly, and cardiovascularimplications. Arterioscler Thromb Vasc Biol 1998;18(9):1363–70.
[21] Pankov R, Yamada KM. Fibronectin at a glance. J Cell Sci 2002;115(Pt 20):3861–3.[22] Sottile J, Hocking DC. Fibronectin polymerization regulates the composition and
stability of extracellular matrix fibrils and cell–matrix adhesions. Mol Biol Cell2002;13(10):3546–59.
[23] Velling T, Risteli J, Wennerberg K, Mosher DF, Johansson S. Polymerization of type Iand III collagens is dependent on fibronectin and enhanced by integrins alpha11beta 1 and alpha 2beta 1. J Biol Chem 2002;277(40):37377–81.
[24] Fogelgren B, Polgar N, Szauter KM, Ujfaludi Z, Laczko R, Fong KS, et al. Cellularfibronectin binds to lysyl oxidase with high affinity and is critical for its proteolyticactivation. J Biol Chem 2005;280(26):24690–7.
[25] Xu J, Cui G, Esmailian F, Plunkett M, Marelli D, Ardehali A, et al. Atrial extracellularmatrix remodeling and the maintenance of atrial fibrillation. Circulation 2004;109(3):363–8.
[26] Hornstra IK, Birge S, Starcher B, Bailey AJ, Mecham RP, Shapiro SD. Lysyl oxidase isrequired for vascular and diaphragmatic development in mice. J Biol Chem2003;278(16):14387–93.
[27] Maki JM, Rasanen J, Tikkanen H, Sormunen R, Makikallio K, Kivirikko KI, et al.Inactivation of the lysyl oxidase gene Lox leads to aortic aneurysms, cardio-
vascular dysfunction, and perinatal death in mice. Circulation 2002;106(19):2503–9.
[28] Lopez B, Querejeta R, Gonzalez A, Beaumont J, Larman M, Diez J. Impact oftreatment on myocardial lysyl oxidase expression and collagen cross-linking inpatients with heart failure. Hypertension 2009;53(2):236–42.
[29] Smith-Mungo LI, Kagan HM. Lysyl oxidase: properties, regulation and multiplefunctions in biology. Matrix Biol 1998;16(7):387–98.
[30] Sivakumar P, Gupta S, Sarkar S, Sen S. Upregulation of lysyl oxidase and MMPsduring cardiac remodeling in human dilated cardiomyopathy. Mol Cell Biochem2008;307(1–2):159–67.
[31] Lopez B, Gonzalez A, Hermida N, Valencia F, de Teresa E, Diez J. Role of lysyloxidase in myocardial fibrosis. From basic science to clinical aspects. Am J PhysiolHeart Circ Physiol 2010;299(1):H1–9.
[32] Molnar J, Fong KS, He QP, Hayashi K, Kim Y, Fong SF, et al. Structural and functionaldiversity of lysyl oxidase and the LOX-like proteins. Biochim Biophys Acta2003;1647(1–2):220–4.
[33] Li D, Shinagawa K, Pang L, Leung TK, Cardin S, Wang Z, et al. Effects of angiotensin-converting enzyme inhibition on the development of the atrial fibrillationsubstrate in dogs with ventricular tachypacing-induced congestive heart failure.Circulation 2001;104(21):2608–14.
[34] Kawano H, Do YS, Kawano Y, Starnes V, Barr M, Law RE, et al. Angiotensin II hasmultiple profibrotic effects in human cardiac fibroblasts 39. Circulation 2000;101(10):1130–7.
[35] Goette A, Arndt M, Rocken C, Spiess A, Staack T, Geller JC, et al. Regulation ofangiotensin II receptor subtypes during atrial fibrillation in humans 96. Circulation2000;101(23):2678–81.
[36] Saygili E, Rana OR, Meyer C, Gemein C, Andrzejewski MG, Ludwig A, et al. Theangiotensin-calcineurin-NFAT pathway mediates stretch-induced up-regulationof matrix metalloproteinases-2/-9 in atrial myocytes. Basic Res Cardiol 2009;104(4):435–48.
[37] Ellinghaus P, Scheubel RJ, Dobrev D, Ravens U, Holtz J, Huetter J, et al. Comparingthe global mRNA expression profile of human atrial and ventricular myocardiumwith high-density oligonucleotide arrays. J Thorac Cardiovasc Surg 2005;129(6):1383–90.
[38] Ruperez M, Lorenzo O, Blanco-Colio LM, Esteban V, Egido J, Ruiz-Ortega M.Connective tissue growth factor is a mediator of angiotensin II-induced fibrosis.Circulation 2003;108(12):1499–505.
[39] Takemoto M, Node K, Nakagami H, Liao Y, Grimm M, Takemoto Y, et al. Statins asantioxidant therapy for preventing cardiac myocyte hypertrophy. J Clin Invest2001;108(10):1429–37.
[40] Burstein B, Qi XY, Yeh YH, Calderone A, Nattel S. Atrial cardiomyocyte tachycardiaalters cardiac fibroblast function: a novel consideration in atrial remodeling.Cardiovasc Res 2007;1; 76(3):442–52.
[41] Lin CS, Lai LP, Lin JL, Sun YL, Hsu CW, Chen CL, et al. Increased expression ofextracellular matrix proteins in rapid atrial pacing-induced atrial fibrillation.Heart Rhythm 2007;4(7):938–49.
[42] Lendeckel U, Dobrev D, Goette A. Aldosterone-receptor antagonism as a potentialtherapeutic option for atrial fibrillation. Br J Pharmacol 2010;159(8):1581–3.
[43] Dobrev D, Nattel S. New antiarrhythmic drugs for treatment of atrial fibrillation.Lancet 2010;375(9721):1212–23.