Selective upregulation of β1-adrenergic receptors and dephosphorylation of troponin I in end-stage heart failure patients supported by ventricular assist devices

Journal of Molecular and Cellular Cardiology 41 (2006) 441–450www.elsevier.com/locate/yjmcc

Original article

Selective upregulation of β1-adrenergic receptors and dephosphorylation oftroponin I in end-stage heart failure patients supported by

ventricular assist devices

Hendrik Milting a,⁎, Christian Scholz c, Latif Arusoglu a, Meike Freitag b, Ramona Cebulla a,Kornelia Jaquet b, Reiner Körfer a, Dirk v. Lewinski c, Astrid Kassner a, Otto-Erich Brodde d,

Harald Kögler c, Aly El Banayosy a,1, Burkert Pieske c,1

a Ruhr-Universität Bochum, Herz- and Diabeteszentrum NRW, Erich and Hanna Klessmann-Institut für Kardiovaskuläre Forschung and Entwicklung,32545 Bad Oeynhausen, Germany

b Berufsgenossenschaftliche Kliniken Bergmannsheil, Abtlg. Molekulare Kardiologie, 44791 Bochum, Germanyc Georg-August Universität, Abteilung Kardiologie und Pneumologie, Robert Koch Str. 40, 37075 Göttingen, Germany

d Universität Essen-Duisburg, Universitätsklinikum, Klinik für Nieren und Hochdruckkrankheiten und Institut für Pathophysiologie, Hufelandstr. 55,45122 Essen, Germany

Received 31 August 2005; received in revised form 27 March 2006; accepted 11 April 2006Available online 12 June 2006

Abstract

In terminal failing hearts, adrenergic receptors are downregulated and intracellular adrenergic signal transduction is inhibited. Mechanicalcirculatory support by ventricular assist devices (VAD) is used to bridge patients to heart transplantation. Mechanical unloading by VAD mayinduce reverse remodeling in heart transplantation (HTx) candidates. However, little is known on β-adrenergic receptor subtype regulation andadrenergic signal transduction under VAD-support. We investigated paired myocardial samples from 16 VAD-supported patients and 9 non-failingdonor hearts. We analyzed β-adrenergic receptor subtype regulation by real-time PCR and radioligand binding and cardiac troponin Iphosphorylation (by phospho-cTnI-specific antibodies). We found that the β1-adrenergic receptor (β1AR) is downregulated at VAD-implantationon mRNA and protein levels whereas the β2-adrenergic receptor (β2AR) was not. After VAD-support, β1AR protein but not its mRNA wasupregulated, whereas the degree of cTnI-phosphorylation was reduced. Upregulation of β1AR was enhanced by beta blocking medication duringVAD-support. However, in 9 out of 15 patients, β1AR-density remained below the 0.25 percentile of donor hearts. VAD-support is associated withpartial normalization of the βAR-signal transduction pathways. This beneficial effect is related to a posttranscriptional increase in β1AR-density.© 2006 Elsevier Inc. All rights reserved.

Keywords: Heart failure; Reverse remodeling; Ventricular assist device; Troponin I; Beta receptor

1. Introduction

Chronic heart failure (CHF) is associated with increasedstimulation of the β-adrenergic system, which appears to bethe most important modulator of cardiac contractility [1]. InCHF, signal transduction of the β-adrenergic system is

⁎ Corresponding author. Herz- and Diabeteszentrum NRW, Georgstr. 11,32545 Bad Oeynhausen, Germany. Tel.: +49 5731 97 3510; fax: +49 5731 971819.

E-mail address: [email protected] (H. Milting).1 Both authors contributed equally to this work.

0022-2828/$ - see front matter © 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.yjmcc.2006.04.010

desensitized due to downregulation of beta receptors (βAR)[2,3] and blunting of their intracellular signal-transmission[4]. In human myocardium, at least two types of βAR (β1

and β2) are expressed. The β1AR is the dominant βAR-subtype with respect to receptor density and inotropic effectsand is coupled to the Gs-adenylyl-cyclase-protein kinase A(PKA) pathway. In rodent cardiomyocytes, the β2ARactivates in addition the p38 mitogen-activated proteinkinase (p38 MAPK) and Giα2- and Giα3-proteins [5,6].The β1AR and β2AR are coupled to different signaltransduction systems [6,7] and might have different effectson cardiomyocyte survival and apoptosis [5,8]. With β-

442 H. Milting et al. / Journal of Molecular and Cellular Cardiology 41 (2006) 441–450

adrenergic stimulation, functional proteins such as cardiactroponin I (cTnI) are phosphorylated by PKA, whichdecreases myofilament Ca2+ sensitivity [9]. Thus, phosphor-ylation of cTnI represents one endpoint of the β-adrenergicsignal transduction cascade at the level of the sarcomericproteins.

It is generally accepted that β1ARs are downregulated inCHF, but conflicting data exist on β2AR-downregulationpossibly depending on the etiology of heart failure [10–12].Furthermore, as a result of impaired β-adrenergic signalingcTnI-phosphorylation may be reduced [13].

Mechanical unloading of failing hearts using ventricularassist devices (VAD) to bridge patients to transplantation(HTx) can result in reverse remodeling on the organ andcellular level (for review, see [14]). However, little is knownon restoration of βAR-regulation during VAD-support. Ininitial experiments, the functional response to β-adrenoceptoragonists was partially restored in isolated human cardiomyo-cytes [15], and total βAR-density was increased after VAD-support [16,17]. However, βAR-subtype-specific receptorregulation and cTnI-phosphorylation during VAD-supporthave not been assessed. This may also be of relevancesince β2AR-stimulation during mechanical unloading mightbe beneficial [18,19].

Therefore, we analyzed in paired left ventricular myocardialsamples from 16 VAD-supported patients βAR-subtype-dependent regulation on the mRNA and protein level and itseffect on the PKA-dependent cTnI-phosphorylation. Wecompared our findings to non-failing donor hearts (NF),which could not be implanted for technical reasons. We alsostratified our data with respect to β-blocker-medication duringVAD-support.

2. Material and methods

2.1. Patients and cardiac muscle samples

Transmural left ventricular muscle samples were collectedat VAD-implantation (apex) and later from the same patientat heart transplantation. Samples were immediately frozen inliquid nitrogen and stored at −80 °C.

The study was performed in agreement with the declarationof Helsinki and was approved by the local ethics committee. Allpatients gave informed consent and were from the transplant-program at the Heart and Diabetes center, Bad Oeynhausen,Germany. For clinical data of the patients, see Table 1. Non-failing myocardium was available from donor hearts, whichcould not be transplanted for technical reasons and had noapparent signs of cardiac failure.

2.2. Isolation of total RNA

Isolation of total RNA was done with the RNeasy kit(Qiagen, Hilden, Germany). For RNA extraction, 30 mgmyocardial tissue was used. Eluted total RNA was quantifiedat 260 nm and tested for genomic DNA contamination byPCR.

2.3. Real-time PCR of β-adrenergic receptors

Reverse transcription of myocardial RNA was done using250 ng total RNA and 50 units superscript II (Invitrogen,Netherlands) after random priming with hexamers. mRNA ofadrenergic receptors was analyzed on a LightCycler (Roche,Basel, Switzerland) using real-time PCR and gene-specificstandard-RNA. The PCR-primers for real-time experimentswere: β1AR sense CCATCTCGGCCCTGGTGTC; antisenseGAAACGGCGCTCGCAGCTGTCG (PCR conditions: 95 °C,2′; 64 °C, 15″; 72 °C, 15″; 95 °C, 39″; 40 cycles; 3 mMMgCl2;5 pmol/μl primer; fluorescence data acquisition at 90 °C);β2AR sense ATCGTCCTGGCCATCGTGTTTG, antisenseCGGGCCTTATTCTTGGTCAGCA; hybprobes donor GCCA-GCATTGAGACCCTGTGCGTacceptor red640 TCGCAGTG-GAT-CGCTACTTTGCCATTA (PCR conditions were identicalto β1AR; however, annealing was done at 60 °C).

2.4. Preparation of standard RNA for calibration

For βAR-mRNA quantification, PCR-products of β1ARand β2AR were cloned using TOPO-TA cloning (Invitrogen,Netherlands) and in vitro transcribed using 10 U T7-polymerase after linearization of the vector. Sequences ofβAR PCR-fragments were confirmed by sequencing(ABI310; Applied Biosystems, Foster City, USA). StandardRNA was isolated (RNeasy, Qiagen, Hilden, Germany) withon column DNase I digestion. RNA was eluted in 30 μlDEPC-H2O and analyzed for DNA contamination by PCR.The concentration of the standard RNAs was calculated fromtriplicate absorption measurements at 260 nm with respect tonucleotid composition.

2.5. Radioligand binding studies

Radioligand binding studies were done as previouslydescribed [12,20]. The βAR-density was assessed by(−)-[125I]-iodocyanopindolol (ICYP) binding at six concentra-tions ranging from 5 to 150 pM for 90 min at 37 °C. Non-specific binding of ICYP was defined as binding tomembranes that was not displaced by 1 μM of the nonselective β-adrenoceptor antagonist (±)-CGP 12177. Specificbinding of ICYP was total binding minus non-specific bindingand amounted usually to 70–80% at 50 pM of ICYP. Kd

values of all groups of patients were not significantly different(NF, VAD-IP and VAD-HTx, means±SD: 27.3±18.6, 27.5±20.6, 50.4±56.2 pM, respectively). However, the Kd at thetime of VAD-HTx was significantly higher in patients withbeta-blocking medication during mechanical assist comparedto those without (P<0.05).

To determine the relative amount of β1- and β2-adrenocep-tors, membranes were incubated with ICYP (100 pM) and 12concentrations (10−10–10−3 M) of the highly β1-adrenoceptorselective antagonist CGP 20712 A, and specific binding wasassessed as described above. CGP 20712 A competition curveswere analyzed using GraphPad Prism 4.03 (GraphPad SoftwareInc., San Diego, USA).

Table 1Clinical data at the time of VAD-implantation

Patient no. Time of VAD-support [d]

Disease Age[44]

VADtype

LVEDD LVESD EF CI PCW CVP PAP Gender Medicationpre VAD

Medicationpre HTx

#1 551 DCM 21 TCI 80 75 16 2.28 21 16 29 m CL, AI, DG CL, AI, DG#2 440 ICM 60 THO-B 61 57 19 n.a. n.a. n.a. n.a. m CL, AR, DG CL, AI, AR#3 91 DCM 60 NOV 61 55 25 1.4 16 15 20 f CL, AI, DG CL, AI, DG#4 650 ICM 57 TCI 75 66 32 2.89 24 11 39 m CA, AI, PI#5 189 DCM 68 THO-L 77 70 25 2.1 21 8 36 m CA, AR, PI#6 181 ICM 52 NOV 90 79 30 1.6 33 8 40 m CA, DG, PI AI, AR#7 467 DCM 36 NOV 88 80 25 1.7 23 16 30 m CL, CA, AI, AR CL, AI, AR#8 30 ICM 52 NOV 77 30 22 2.2 28 13 33 m CL, AI, DG, PI AI, AR, DG#9 272 DCM 66 NOV 77 66 20 1.5 18 6 28 m CL, CA, AI, AR,

DG, PICL, AI

#10 224 DCM 54 THO-L 69 61 31 2.4 21 7 29 m CA, AI, AR, DG, PI AR#11 380 DCM 61 NOV 82 77 20 1.36 24 11 42 f CA, AI, PI CL, AI, AR, DG#12 359 DCM 56 TCI 85 77 20 3.4§ 23 9 36 m CL, CA, AI, DG, PI CL, AI#13 66 DCM 12 THO-B 74 67 18 2.2 24 10 33 m CA, PI ML#14 94 DCM 46 THO-B n.a. n.a. n.a. n.a. n.a. n.a. n.a. m CA#15 77 ICM 68 TCI 2 83 79 14 1.5 28 6 36 m CL, CA, AI, AR,

DG, PICL, CA, AI,AR, DG, PI

#16 176 DCM 63 THO-L 78 71 25 2.6 12 12 30 m CL, AI, DG, PI CL#17 NF 18 n.a. n.a. n.a. n.a. n.a. n.a. n.a. f#18 NF 61 n.a. n.a. n.a. n.a. n.a. n.a. n.a. m#19 NF 40 n.a. n.a. n.a. n.a. n.a. n.a. n.a. f#20 NF 43 n.a. n.a. n.a. n.a. n.a. n.a. n.a. m#21 NF 53 n.a. n.a. n.a. n.a. n.a. n.a. n.a. f#22 NF 51 n.a. n.a. n.a. n.a. n.a. n.a. n.a. m#23 NF 55 n.a. n.a. n.a. n.a. n.a. n.a. n.a. f#24 NF 52 n.a. n.a. n.a. n.a. n.a. n.a. n.a. m

Abbreviations: ICM=ischemic cardiomyopathy; DCM=idiopathic dilative cardiomyopathy; THO-L, THO-B=Thoratec left (L) or bi(B)-VAD; NOV-L=Novacor left VAD; LVEDD=left ventricular end-diastolic diameter; LVESD=left ventricular end-systolic diameter; EF=ejection fraction; CI=cardiacindex; PCW=pulmonary capillary wedge pressure; PAP=pulmonary artery pressure; IABP=intraaortic ballon pump. n.a.=not available. NF=non-failingdonor heart. §at syst. vasc. resist. of 542 dyn. Medication: AI=angiotensin converting enzyme inhibitor; AR=antiarrhythmias; CA=catecholamines;CL=carvedilol; DG=digitalis; ML=metoprolol; PI=phosphodiesterase inhibitor. Drugs interfering with βAR signal transduction printed in bold; m=male,f= female.

443H. Milting et al. / Journal of Molecular and Cellular Cardiology 41 (2006) 441–450

2.6. Monoclonal antibodies against phospho- anddephospho-cTnI

The phospho-cTnI antibodies were kindly provided by Dr.Filatov (Lomonosov Moscow State University), the dephosphoantibodies [21,22] were from HyTEST (Turku, Finland).Monoclonal antibodies 22B11, 7E10 and 1G11 were obtainedusing standard protocols [21]. The specificity of the antibodieswas published recently [23]. Total cTnI was analyzed byimmunoblotting using a commercially available antibody(Hytest, Turku Finland, #4T21).

2.7. Western blotting

The antibody against the PKA phosphorylation sites ofcTnI was used for quantification of cTnI phosphorylation.Tissue samples (about 100 mg wet weight) were crushed inliquid nitrogen and transferred to 1000 μl extraction buffercontaining [mM]: 10 imidazol, 300 sucrose, 10 NaF, 1EDTA, 0.5 DTT, proteinase inhibitors aprotinin, pepstatin A,leupeptin and PMSF, pH 7. Samples were homogenized withan Ultra Turrax (Ika, Germany) and centrifuged at 12,000×g.Protein (7.5 μg) of the supernatant was separated in 10%SDS-polyacrylamid electrophoresis and electrotransferred

(Trans-blot SD, Biorad, USA) to nitrocellulose sheets(Biorad, USA, #162-0145). For immunodetection, theantibody against phospho-cTnI and cTnI was diluted1:1000 and 1:500 in TBS-Tween, respectively. The secondHRP-coupled antibody was anti-mouse (1:1000, Pharmingen#554002). As an HRP-substrate, Chemiglow (Alpha Innotec,USA) was used and recorded by a CCD-camera system(Alpha Innotec, USA). Paired samples pre- and post-VAD-support from the same patient were analyzed in triplicate forphospho-cTnI on the same gel under standardized conditionsafter determination of the linear range with the Western blot.The total cTnI was determined under the same conditions induplicate.

2.8. Laser scanning microscopy

The distribution of phosphorylated and dephosphorylatedcTnI was examined by laser scanning microscopy of cryosectedmyocardial samples of patients #3 and #15 with a Zeiss LSM510 (Zeiss Jena, Germany). Tissue slides (15 μm) were dried at40 °C for 60 min and fixed in ice-cold 100% ethanol (−20 °C).After 3 washing steps with PBS, and incubation for 30 min witha saturation solution, the phospho-cTnI antibody (1:40 in PBS;labeled with Alexa fluor 488) was incubated at room

444 H. Milting et al. / Journal of Molecular and Cellular Cardiology 41 (2006) 441–450

temperature (4 h). After washing off excess antibody with PBS,the dephospho-cTnI antibody (Alexa fluor 594, dilution 1:150)was incubated overnight at room temperature. After washingwith PBS, the slides were coated with 25 μl ProLong Antifade(Molecular Probes Inc., Leiden, Netherlands) for bleachingprotection.

Excitation of the fluorescence labels was done at 488 nm forthe phospho-cTnI antibody and at 543 nm for the dephospho-cTnI antibody. Detection of the fluorescence was done withfilters passing 505–530 nm in case of the phospho-cTnI(‘green’) and 560–615 nm for the dephospho cTnI antibody(‘red’).

2.9. Statistics

Statistical evaluation was done with ANOVA and Bonferro-ni's multiple comparison posttest, regarding P<0.05 assignificant. Box and whiskers plots: the boxes extend fromthe 25th to the 75th percentile, with a line at the median. Thewhiskers extend above and below the box to show the highestand lowest values.

Fig. 1. Expression of β1AR during VAD-support and in non-failing donor hearts. (a) Eand in paired samples at VAD-implantation (VAD-IP) and at VAD-transplantation (circles) in non-failing hearts (left) and during VAD-support (right). Lines connectmyocardium at the time of VAD-implantation, in some hearts despite clearly reduce

3. Results

3.1. Regulation of β-adrenergic receptors

We have analyzed the mRNA of the β1- and β2AR in pairedmyocardial samples of patients at the time of VAD-implantation(VAD-IP), -transplantation (VAD-HTx) and in eight non-failing(NF) hearts. The mRNA content of the β1AR in myocardialsamples of NF hearts, at VAD-IP and VAD-HTx, was (amol/μgtotal RNA, mean±SD) 18.6±7.76, 9.44±6.28 and 7.07±3.55,respectively. The mean mRNA content between VAD-IP andVAD-HTx was not significantly different, but samples fromVAD-patients revealed a significantly lower β1AR-mRNAcontent compared to NF hearts, indicating downregulation ofthe β1AR-mRNA (Fig. 1a). The mRNA of β2AR was (amol/μgtotal RNA, mean±SD) 7.1±3.47 in NF, 7.65±3.2 at VAD-IPand 6.94±2.81 at the time of VAD-HTx. The β2AR-mRNAcontent of NF was not different from VAD-samples (Fig. 2a).Thus, the β1AR-mRNA was downregulated at the time ofVAD-IP and VAD-HTx, but the β2AR-mRNA was not alteredwith heart failure (Figs. 1 and 2). The ratio of the means

xpression of β1AR-mRNA (left) and -density (right) in non-failing hearts (NF),VAD-HTx). (b) Expression of β1AR-mRNA (closed circles) and density (opendata derived from the same hearts. β1AR-mRNA was in the range of controld β1AR-density.

445H. Milting et al. / Journal of Molecular and Cellular Cardiology 41 (2006) 441–450

between β2- and β1AR-mRNA in NF, VAD-IP and VAD-HTxsamples was 0.38, 0.81 and 0.98, respectively. Individual βARratios between VAD-IP and VAD-HTx were not significantlydifferent. The βAR-mRNA concentrations were similar topublished data [24].

However, total βAR-density was significantly affected byVAD.Maximum binding of [125I]-iodocyanopindolol at the timeof VAD-IP was significantly lower (14.4±5 fmol/mg protein;means±SD) compared to the time of VAD-HTx (32.23±14.72;P<0.01) and to samples from NF hearts (34.58±11.2;P<0.0001). The means of the Kd values in samples from NFand VAD-IP and VAD-HTx without beta-blocker medicationwere [pM]: 27.3, 27.5 and 20.8. Kd values in samples of beta-blocker treated patients were significantly higher at VAD-HTx(48.7; P<0.05) confirming retrospectively collected data on themedication during VAD-support. Maximum binding betweenVAD-HTx and NF-samples was not significantly different,indicating upregulation of β-adrenergic binding sites aftermechanical unloading of the failing heart.

We further analyzed the portion of β1- and β2AR by CGP20712A displacement experiments. β1AR-density was (fmol/mg±SD) 27.63±12.34 in NF, 9.79±3.95 at VAD-IP and 18.33±

Fig. 2. Expression of β2AR during VAD-support and in non-failing donor hearts. (a) Eand in paired samples at VAD-implantation (VAD-IP) and at VAD-transplantation (Vcircles) in non-failing hearts (left) and during VAD-support (right). Lines connect d

12.54 in VAD-HTx (Fig. 1a), and the β2AR-density was 6.95±4.93, 3.84±3.22 and 6.83±5.97 in NF, VAD-IP and VAD-HTx-samples, respectively (Fig. 2a). The β1AR-density wassignificantly lower at VAD-IP compared to VAD-HTx(P<0.05) and to NF-samples (P<0.01; Fig. 1a). However, theβ2AR-density did not differ between sample groups (Fig. 2a).The ratio of β2AR/β1AR protein in NF, VAD-IP and VAD-HTxsamples was 0.25, 0.39 and 0.37, respectively. Thus, down-regulation of βAR at the time of VAD-IP and upregulation afterVAD-support were restricted to the β1AR-protein. We alsocompared the relative expression ofβAR-mRNA and -protein inindividual patients. Interestingly, as compared to the NF-myocardium in some patients, downregulation of β1AR at thetime of VAD-implantation was not associated with reducedcontent of the corresponding mRNA (Figs. 1b and 2b).

We observed a trend towards more pronounced β1ARupregulation in VAD-patients receiving beta-blocking medica-tion during mechanical support (data not shown). We did notfind any correlation of βAR-regulation with the time of VAD-support or the VAD-type. However, the only patient (#15)supported by an axial flow pump with device failure andcombined catecholamine and phosphodiesterase inhibitor

xpression of β2AR-mRNA (left) and -density (right) in non-failing hearts (NF),AD-HTx). (b) Expression of β2AR-mRNA (closed circles) and -density (openata derived from the same hearts.

446 H. Milting et al. / Journal of Molecular and Cellular Cardiology 41 (2006) 441–450

medication before VAD-HTx did not reveal β1AR upregulation(see also Figs. 3b and 4c and d).

We estimated the number of β1- and β2AR translated fromthe corresponding mRNA using the data from samplepreparation, radioligand-binding and mRNA-analysis. Wecalculated for the β1AR in NF-samples and at the time ofVAD-IP and VAD-HTx [mol receptor/mol mRNA; mean±SEM] 49±14, 184±85 and 333±135, respectively. Therelationship of the receptor protein per mRNA for the β2ARtended to be lower: 26±7, 57±23 and 82±29. However,differences in the ratios for both βAR between VAD-IP andVAD-HTx were not significantly different.

3.2. Phosphorylation of cardiac troponin I

We further analyzed the PKA-dependent phosphorylationsites in cTnI at serine-22 (S22) and -23 (S23) by Westernblotting (Fig. 3a, b and c) of paired samples. The PKA-dependent phosphorylation of cTnI was comparable betweennon-failing donor hearts and failing samples (Fig. 3c). Samplesfrom patients at VAD-transplantation were less phosphorylatedat S22/S23 of cTnI compared to NF. This was confirmed byisoelectric focusing of isolated cTnI [25] in a different set ofsamples (data not shown) and by laser scanning microscopy in asubset of patients (Fig. 4). The cTnI phosphorylation of VAD-supported hearts was lower compared to VAD-IP (P<0.05) orNF-samples (P<0.01), respectively (Fig. 3a, b and c). TotalcTnI was analyzed in parallel and not regulated between non-failing donor hearts and during VAD-support, respectively (Fig.3a and c).

3.3. Relative regulation of β-adrenergic receptors and troponinI phosphorylation

In 8 patients, paired samples could be analyzed for therelative regulation of βAR, cTnI-phosphorylation and β1AR-mRNA during VAD-support. Interestingly, 6 of these patientsshowed upregulation of the β1AR-protein, but dephosphoryla-tion of cTnI and no upregulation in β1AR-mRNA.

4. Discussion

Patients suffering from end-stage heart failure may bebridged to transplantation by ventricular assist devices [26].Mechanical unloading of the failing ventricles leads toreduction of ventricular wall tension and partial normalization

Fig. 3. (a) Western blot analysis of total (left) and phosphorylated (right) cardiac tropaired samples pre (3 lanes left) and post (3 lanes right) VAD-support and duplicate anin the same gel-run and immunoblotted in parallel under standardized conditions. Cseparated in the same gel and immunoblotted for phosphorylated cTnI. (a) electiveldonor hearts without (b) and with (c) catecholamine medication, from paired samplpatient #15 (f) shows intensive phosphorylation of cTnI due to device failure andinverted for presentation. (c) Densitometric evaluation of immunoblots for the quantcardiac troponin I during ventricular assist device support. Phosphorylation of cTnIconnect paired samples pre and post VAD-support. At VAD-IP, cTnI-phosphorylatiphosphorylation was reduced as compared to VAD-IP. Paired data reveal dephosphoregulated.

of neurohumoral overstimulation [27]. It was speculated thattransient unloading of terminal failing hearts by ventricularassist devices may even induce reverse remodeling andmyocardial recovery [28] (for review, see [14]). However, theconcept of bridging to recovery in end-stage heart failure is stilla matter of debate [29,30].

Related to sustained catecholamine stimulation, signaltransduction of the β-adrenergic system is blunted [12] bydownregulation and desensitization of the βAR in CHF. TotalβAR-downregulation was primarily due to reduced expressionof β1ARs with no or minor changes in β2AR-mRNA [31] or-density [12]. Nevertheless, β2ARs may be regulated in thehuman heart under specific clinical [10] or therapeutic [32]conditions. However, little is known on subtype-specific βAR-regulation during VAD-support in patients with end-stage heartfailure. This might be of specific clinical relevance since β2AR-stimulation had beneficial effects on cardiac regeneration[18,19].

In line with previous studies [11,12,31,33], we founddownregulation of β1AR-mRNA and -density in end-stageCHF before mechanical support as compared to NF humanmyocardium. Downregulation was exclusively related toreduced expression of β1AR-mRNA and reduced β1AR-protein with no changes in the β2AR. As previouslyreported, relative abundancy of the β2AR was significantlylower than β1AR in NF myocardium (ratio β2AR/β1ARwas 0.25 for binding and 0.38 for mRNA experiments,respectively). However, this ratio increased in failingmyocardium (to 0.39 and 0.81, respectively) due to theselective downregulation of β1ARs (P<0.05 vs. failing).These data are consistent with previously published reports[24,31] and may reflect the response to chronic sympatheticoverstimulation.

We then assessed the effects of VAD-support on βAR-expression on the mRNA and protein level. We observed a clearupregulation of total βAR-density during VAD-support, whichwas exclusively related to an increase in β1AR-binding with nochange in β2AR-binding. However, interestingly, the increasein β1AR-density was not related to an increase in β1AR-mRNA. These data point to a posttranscriptional regulation ofthe βAR-mRNA or -protein turnover during mechanicalunloading. We estimated the βAR-receptors translated percorresponding mRNA molecule from data derived from RNAand membrane preparations. The ratio of β1AR-proteintranslated per β1AR-mRNA molecule increased 1.8-times.We also found that the number of receptor molecules translated

ponin I. Triplicate determinations of phosphorylated cTnI (23.5 kDa shown) inalyses for total cTnI were shown. Samples from the same patient were separatedhemiluminescence data were inverted for presentation. (b) Myocardial samplesy transplanted patient (clinical data not shown), (b and c) from two non-failinges (patient #11) at VAD-IP (d) and VAD-HTx (e). The myocardial sample fromcatecholamine medication during VAD-support. Chemiluminescence data wereification of total cTnI (left) and phosphorylated cTnI (right). Phosphorylation ofin NF-hearts (hatched bars) and VAD-supported myocardium (open bars). Lineson was not significantly lower compared to NF-samples. At VAD-HTx, cTnI-rylation of cTnI in all but one patient during VAD-support. Total cTnI was not

447H. Milting et al. / Journal of Molecular and Cellular Cardiology 41 (2006) 441–450

from the corresponding mRNA was higher for the β1ARcompared to the β2AR. Thus, the efficiency of β1AR-mRNAtranslation was higher compared to the β2AR-mRNA, and theβ1AR-mRNA and the β1AR-protein appeared to be indepen-

dently regulated during mechanical unloading of terminalfailing hearts.

The stability of β1AR- and β2AR-mRNA [34,35] mayaffect agonist-dependent βAR-regulation [36]. It was reported

Fig. 4. Immunohistology of cardiac phospho-troponin I. Laser scanning microscopy of paired myocardial cryosections of VAD-patients labeled with monoclonalantibodies against phosphorylated (‘green’) and unphosphorylated (‘red’) cTnI. Upper panels: Patient #3 at VAD-implantation (a) and -transplantation (b) revealsdephosphorylation of cTnI during VAD-support typical for the cohort as shown in Fig. 3a–c. Lower panels: Patient #15 suffering from VAD failure and need forpharmacological support with catecholamines and phosphodiesterase inhibitors until transplantation. Phosphorylated cTnI dominates at VAD-implantation (c) andeven appears to increase at transplantation (d).

448 H. Milting et al. / Journal of Molecular and Cellular Cardiology 41 (2006) 441–450

that the stability of the β1AR-mRNA in a recombinant systemis dependent on its 3′ untranslated region [37]. Agonist-dependent degradation of the mRNA is also critical for theβ2AR due to an AU-rich element located at the 3′ end of themRNA [38,39]. Thus, we suppose that, in the unloaded terminalfailing heart, expression of the β1AR is regulated on thetranscriptional and/or posttranscriptional level by a not yetidentified mechanism.

Selective upregulation of β1AR after mechanical unloadingwas not observed in all patients. Especially patient #15, whoexperienced device failure during VAD-support and need forinotropic stimulation did not show upregulation of the β1AR.Of note, 9 out of 15 patients showed β1AR-densities at the timeof transplantation which were below the 0.25 percentile of non-failing donor hearts (<17.3 fmol/mg). The regulation of theβ1AR was not dependent on the duration of VAD-support, theVAD-type or clinical variables except medication (see below).Lack of correlation to the time of VAD-support was alsoreported by Ogletree-Hughes et al. [16]. However, in that study,only patients with a comparable short VAD-support time wereincluded.

We further analyzed the effects of mechanical unloading onthe phosphorylation of cTnI. At the time of VAD-IP, cTnI-

phosphorylation was not significantly lower compared to NF-samples, which is in agreement with some reports [40,41].However, most of the VAD-IP samples revealed lower levelsof cTnI-phosphorylation compared to the means of NF-samples. In contrast, we found in samples from electivelytransplanted patients (n=5) a significantly lower degree ofcTnI-phosphorylation compared to samples from NF andpatients with need for mechanical circulatory support (data notshown). These findings may be related to blunting of the β-adrenergic signaling pathways in failing myocardium despiteelevated catecholamine levels. However, some patientsrevealed a high degree of cTnI-phosphorylation, whichmight be due to the medication of these patients withinotropes before VAD-IP (see Table 1). The high cTnI-phosphorylation in some non-failing hearts may contribute todiscrepant differences in the literature [13], since these brain-dead organ-donors usually have excessive endogenouscatecholamine concentrations. In this study, all but one ofthe donors analyzed for cTnI-phosphorylation receivedcatecholamines due to a low cardiac output before removalof the donor heart.

Interestingly, despite an increase in βAR-density, cTnI-phosphorylation decreased during VAD-therapy in the present

449H. Milting et al. / Journal of Molecular and Cellular Cardiology 41 (2006) 441–450

study, which might be explained by the reduced catecholaminestress during VAD-support. Dephosphorylation of cTnI duringVAD-support may contribute to the reported normalization offunctional responsiveness to catecholamines [15]. Dephosphor-ylation of cTnI in VAD-supported myocardium was alsoconfirmed by isoelectric focusing of isolated cTnI [25] (datanot shown) and could be related to the reported normalization ofplasma catecholamine levels during VAD therapy [42]. Beta-blocker therapy was associated with significantly higher β1AR-density after VAD, if tested against the implant samples.Therefore, under the condition of mechanical unloading,upregulation of β1AR on the protein level may be supportedby beta-blocker therapy.

However, no conclusive data on substantial clinical normal-ization of the failing heart have been obtained in this study.Moreover, the regulation of the βAR appeared to beindependent of the total collagen regulation in the patientsinvestigated [43]. Therefore, the challenging concept of reverseremodeling of the failing human heart and especially thepharmacological management under mechanical support clearlydeserves further study.

Acknowledgments

This work was supported by Grants of the Erich and HannaKlessmann-Stiftung, Halle/Westf., Förderverein Herzchirurgie,Bad Oeynhausen, Germany and by the Deutsche Forschungs-gemeinschaft (DFG PI 414/1–4).

References

[1] Zerkowski HR, Broede A, Kunde K, Hillemann S, Schafer E, VogelsangM, et al. Comparison of the positive inotropic effects of serotonin,histamine, angiotensin II, endothelin and isoprenaline in the isolatedhuman right atrium. Naunyn-Schmiedeberg's Arch Pharmacol 1993;347:347–52.

[2] Bristow MR, Ginsburg R, Minobe W, Cubicciotti RS, Sageman WS, LurieK, et al. Decreased catecholamine sensitivity and beta-adrenergic-receptordensity in failing human hearts. N Engl J Med 1982;307:205–11.

[3] Brodde O-E. Beta 1- and beta 2-adrenoceptors in the human heart:properties, function, and alterations in chronic heart failure. PharmacolRev 1991;43:203–42.

[4] Ungerer M, Böhm M, Elce JS, Erdmann E, Lohse MJ. Altered expressionof beta-adrenergic receptor kinase and beta 1-adrenergic receptors in thefailing human heart. Circulation 1993;87:454–63.

[5] Zhu WZ, Zheng M, Koch WJ, Lefkowitz RJ, Kobilka BK, Xiao RP. Dualmodulation of cell survival and cell death by beta(2)-adrenergic signalingin adult mouse cardiac myocytes. Proc Natl Acad Sci U S A 2001;98:1607–12.

[6] Daaka Y, Luttrell LM, Lefkowitz RJ. Switching of the coupling of thebeta2-adrenergic receptor to different G proteins by protein kinase A.Nature 1997;390:88–91.

[7] Steinberg SF. The molecular basis for distinct beta-adrenergic receptorsubtype actions in cardiomyocytes. Circ Res 1999;85:1101–11.

[8] ZauggM, XuW, Lucchinetti E, Shafiq SA, Jamali NZ, Siddiqui MA. Beta-adrenergic receptor subtypes differentially affect apoptosis in adult ratventricular myocytes. Circulation 2000;102:344–50.

[9] Wattanapermpool J, Guo X, Solaro RJ. The unique amino-terminal peptideof cardiac troponin I regulates myofibrillar activity only when it isphosphorylated. J Mol Cell Cardiol 1995;27:1383–91.

[10] Brodde O-E, Zerkowski HR, Doetsch N, Motomura S, Khamssi M, MichelMC. Myocardial beta-adrenoceptor changes in heart failure: concomitant

reduction in beta 1- and beta 2-adrenoceptor function related to the degreeof heart failure in patients with mitral valve disease. J Am Coll Cardiol1989;14:323–31.

[11] Steinfath M, Geertz B, Schmitz W, Scholz H, Haverich A, Breil I, et al.Distinct down-regulation of cardiac beta 1- and beta 2-adrenoceptors indifferent human heart diseases. Naunyn-Schmiedeberg's Arch Pharmacol1991;343:217–20.

[12] Brodde O-E, Vogelsang M, Broede A, Michel-Reher M, Beisenbusch-Schafer E, Hakim K, et al. Diminished responsiveness of Gs-coupledreceptors in severely failing human hearts: no difference in dilatedversus ischemic cardiomyopathy. J Cardiovasc Pharmacol 1998;31:585–94.

[13] van der Velden J, Papp Z, Zaremba R, Boontje NM, de Jong JW, Owen VJ,et al. Increased Ca2+-sensitivity of the contractile apparatus in end-stagehuman heart failure results from altered phosphorylation of contractileproteins. Cardiovasc Res 2003;57:37–47.

[14] Pieske B. Reverse remodeling in heart failure—Fact or fiction? Eur Heart J2004;6:D66–78.

[15] Dipla K, Mattiello JA, Jeevanandam V, Houser SR, Margulies KB.Myocyte recovery after mechanical circulatory support in humans withend-stage heart failure. Circulation 1998;97:2316–22.

[16] Ogletree-Hughes ML, Stull LB, Sweet WE, Smedira NG, McCarthy PM,Moravec CS. Mechanical unloading restores beta-adrenergic responsive-ness and reverses receptor downregulation in the failing human heart.Circulation 2001;104:881–6.

[17] Klotz S, Barbone A, Reiken S, Holmes JW, Naka Y, Oz MC, Marks AR,Burkhoff D. Left ventricular assist device support normalizes left and rightventricular beta-adrenergic pathway properties. J Am Coll Cardiol 2005;45:668–76.

[18] Tevaearai HT, Eckhart AD, Walton GB, Keys JR, Wilson K, Koch WJ.Myocardial gene transfer and overexpression of beta2-adrenergic receptorspotentiates the functional recovery of unloaded failing hearts. Circulation2002;106:124–9.

[19] Hon JK, Yacoub MH. Bridge to recovery with the use of leftventricular assist device and clenbuterol. Ann Thorac Surg 2003;75:S36–41.

[20] Brodde O-E, Zerkowski HR, Schranz D, Sitz Broede A, Reher Michel M,Beisenbusch Schafer E, et al. Age-dependent changes in the beta-adrenoceptor-G-protein(s)-adenylyl cyclase system in human right atrium.J Cardiovasc Pharmacol 1995;26:20–6.

[21] Katrukha AG, Bereznikova AV, Esakova TV, Pettersson K, Lovgren T,Severina ME, et al. Troponin I is released in bloodstream of patients withacute myocardial infarction not in free form but as complex. Clin Chem1997;43:1379–85.

[22] Filatov VL, Katrukha AG, Bereznikova AV, Esakova TV, Bulargina TV,Kolosova OV, et al. Epitope mapping of anti-troponin I monoclonalantibodies. Biochem Mol Biol Int 1998;45:1179–87.

[23] Deng Y, Schmidtmann A, Kruse S, Filatov V, Heilmeyer Jr LM, JaquetK, et al. Phosphorylation of human cardiac troponin I G203S andK206Q linked to familial hypertrophic cardiomyopathy affectsactomyosin interaction in different ways. J Mol Cell Cardiol 2003;35:1365–74.

[24] Lowes BD, Gilbert EM, Abraham WT, Minobe WA, Larrabee P,Ferguson D, et al. Myocardial gene expression in dilated cardiomyop-athy treated with beta-blocking agents. N Engl J Med 2002;346:1357–65.

[25] Ardelt P, Dorka P, Jaquet K, HeilmeyerJr LM, Körtke H, Körfer R, et al.Microanalysis and distribution of cardiac troponin I phospho species inheart areas. Biol Chem 1998;379:341–7.

[26] El-Banayosy A, Deng M, Loisance DY, Vetter H, Gronda E, Loebe M,et al. The European experience of Novacor left ventricular assist (LVAS)therapy as a bridge to transplant: a retrospective multi-centre study. EurJ Cardio-thorac Surg 1999;15:835–41.

[27] Milting H, El-Banayosy A, Kassner A, Fey O, Sarnowski P, Arusoglu L,et al. The time course of natriuretic hormones as plasma markers ofmyocardial recovery in heart transplant candidates during ventricularassist device support reveals differences among device types. J HeartLung Transplant 2001;20:949–55.

450 H. Milting et al. / Journal of Molecular and Cellular Cardiology 41 (2006) 441–450

[28] Müller J, Wallukat G, Weng Y-G, Dandel M, Spiegelsberger S, Semrau S,et al. Weaning from mechanical cardiac support in patients with idiopathicdilated cardiomyopathy. Circulation 1997;96:542–9.

[29] El-Banayosy A, Arusoglu L, Kizner L, Fey O, Sarnowski P, MorshuisM, et al. Hemodynamic exercise testing reveals a low incidence ofmyocardial recovery in LVAD patients. J Heart Lung Transplant 2001;20:209–10.

[30] Mancini DM, Beniaminovitz A, Levin H, Catanese K, Flannery M,DiTullio M. Low incidence of myocardial recovery after left ventricularassist device implantation in patients with chronic heart failure. Circulation1998;98:2383–9.

[31] Ihl-Vahl R, Eschenhagen T, Kubler W, Marquetant R, Nose M, Schmitz W,et al. Differential regulation of mRNA specific for beta 1- and beta 2-adrenergic receptors in human failing hearts. Evaluation of the absolutecardiac mRNA levels by two independent methods. J Mol Cell Cardiol1996;28:1–10.

[32] Michel MC, Pingsmann A, Beckeringh JJ, Zerkowski HR, Doetsch N,Brodde O-E. Selective regulation of beta 1- and beta 2-adrenoceptors in thehuman heart by chronic beta-adrenoceptor antagonist treatment. Br JPharmacol 1988;94:685–92.

[33] BohmM, Beuckelmann D, Brown L, Feiler G, Lorenz B, Nabauer M, et al.Reduction of beta-adrenoceptor density and evaluation of positiveinotropic responses in isolated, diseased human myocardium. Eur HeartJ 1988;9:844–52.

[34] Conne B, Stutz A, Vassalli JD. The 3′ untranslated region of messengerRNA: a molecular ‘hotspot’ for pathology? Nat Med 2000;6:637–41.

[35] Bevilacqua A, Ceriani MC, Capaccioli S, Nicolin A. Post-transcriptionalregulation of gene expression by degradation of messenger RNAs. J CellPhysiol 2003;195:356–72.

[36] Blaxall BC, Port JD. Adrenergic receptor protocols. In: Machida C, editor.Adrenergic receptor protocols. Totowa, New Jersey: Humana Press Inc.;2000. p. 453–65.

[37] Mitchusson KD, Blaxall BC, Pende A, Port JD. Agonist-mediateddestabilization of human beta1-adrenergic receptor mRNA: role of the 3′untranslated translated region. Biochem Biophys Res Commun 1998;252:357–62.

[38] Hadcock JR, Wang HY, Malbon CC. Agonist-induced destabilization ofbeta-adrenergic receptor mRNA. Attenuation of glucocorticoid-inducedup-regulation of beta-adrenergic receptors. J Biol Chem 1989;264:19928–33.

[39] Danner S, Frank M, Lohse MJ. Agonist regulation of human beta2-adrenergic receptor mRNA stability occurs via a specific AU-rich element.J Biol Chem 1998;273:3223–9.

[40] Bartel S, Stein B, Eschenhagen T, Mende U, Neumann J, Schmitz W, et al.Protein phosphorylation in isolated trabeculae from nonfailing and failinghuman hearts. Mol Cell Biochem 1996;157:171–9.

[41] El-Armouche A, Pamminger T, Ditz D, Zolk O, Eschenhagen T. Decreasedprotein and phosphorylation level of the protein phosphatase inhibitor-1 infailing human hearts. Cardiovasc Res 2004;61:87–93.

[42] James KB, McCarthy PM, Thomas JD, Vargo R, Hobbs RE, Sapp S, et al.Effect of the implantable left ventricular assist device on neuroendocrineactivation in heart failure. Circulation 1995;92:II191–5.

[43] Milting H, Jacob M, Kassner A, Heimann P, Mannherz HG, BeckerG, et al. The structural examination of myocardial samples frompatients with end-stage heart failure supported by ventricular assistdevices using electron microscopy and amino acid analysis revealslow degree of reverse remodeling. J Heart Lung Transplant 2004;23:396–404.