- Email: [email protected]
Right ventricular upregulation of the Ca2+ binding protein S100A1 in chronic pulmonary hypertension
Biochimica et Biophysica Acta 1500 (2000) 249^255 www.elsevier.com/locate/bba
Right ventricular upregulation of the Ca2 binding protein S100A1 in chronic pulmonary hypertension Philipp Ehlermann a , Andrew Remppis a; *, Oliver Guddat b , Jo«rg Weimann c , Philipp A. Schnabel b , Johann Motsch c , Claus W. Heizmann d , Hugo A. Katus
a
a
Medizinische Klinik II, Medizinische Universita«t zu Lu«beck, Ratzeburger Allee 160, D-23538 Lu«beck, Germany Pathologisches Institut, Universita«t Heidelberg, Im Neuenheimer Feld 220/221, D-69120 Heidelberg, Germany c Klinik fu«r Ana«sthesiologie, Universita«t Heidelberg, Im Neuenheimer Feld 110, D-69120 Heidelberg, Germany Abteilung fu«r Klinische Chemie und Biochemie, Kinderspital Zu«rich, Steinwiesstr. 75, CH-8032 Zu«rich, Switzerland b
d
Received 27 July 1999; received in revised form 22 October 1999; accepted 10 November 1999
Abstract The Ca2 binding protein S100A1 increases the Ca2 release from the sarcoplasmatic reticulum by interacting with the ryanodine receptor. In order to understand whether this effect might be operative in the early course of hypertrophy, when myocardium is able to meet increased workload, we investigated the expression of S100A1 in a model of moderate right ventricular hypertrophy. The pulmonary arteries of nine pigs were embolised three times with Sephadex G-50. After 70 days, all pigs showed a moderate pulmonary hypertension. Right ventricular tissue of embolised animals showed a significant increase of connective tissue and enlargement of myocyte diameters. In controls, we found a differential expression of S100A1 with significantly lower S100A1 protein levels in right ventricular compared to left ventricular tissue. In pulmonary hypertension, S100A1 expression increased significantly in hypertrophied right ventricles while it was unchanged in left ventricular tissue. No change was observed in the expression of SERCA2a and phospholamban. Our data show, for the first time, that moderate pressure overload results in an upregulation of S100A1. This may reflect an adaptive response of myocardial Ca2 homeostasis to a higher workload. ß 2000 Elsevier Science B.V. All rights reserved. Keywords: Calcium-binding protein; S100A1; Myocardial hypertrophy; Pulmonary hypertension; SERCA; Phospholamban
1. Introduction The intracellular Ca2 cycling is largely determined by the activity of the sarcoplasmatic reticulum (SR) [1]. Since Ca2 cycling is compromised in end stage heart failure, studies have focussed on the functional state of the SR in patients that underwent
* Corresponding author. Fax: +49-451-500-6437; E-mail: [email protected]
cardiac transplantation [2]. Mulieri et al. [3] and Pieske et al. [4] demonstrated a reversed force-frequency relationship and an altered post-rest potentiation in isolated muscle ¢bres, indicating a dysfunction of the SR in end stage heart failure. However, the underlying molecular mechanisms are not fully understood. Diminished systolic Ca2 peaks and increased diastolic Ca2 levels [5,6] indicate an altered activity of the sarcoplasmatic Ca2 -release channel (ryanodine receptor, RyR) and Ca2 ATPase (SERCA), respectively. An altered expression of SR proteins has therefore been discussed as one of the pos-
0925-4439 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 5 - 4 4 3 9 ( 9 9 ) 0 0 1 0 6 - 4
BBADIS 61906 20-1-00
250
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255
sible molecular causes of SR dysfunction. Results, however, are still contradictory, possibly due to different methodological approaches [7,8]. Since the activity of RyR and SERCA are tightly regulated [9], an altered regulation of Ca2 transients might be due to SR dysfunction in end-stage heart failure. It has been shown that Ca2 binding proteins of the EF-hand type di¡erentially regulate SR Ca2 transients. Calmodulin is known to activate SERCA via calmodulin kinase-dependent phosphorylation of phospholamban [10] and it may also regulate the activity of the ryanodine receptor (RyR) [11]. However, it is unknown whether calmodulin protein concentrations are altered in the human cardiomyopathic heart since only mRNA data are available [12]. Like calmodulin, S100 proteins also belong to the large family of Ca2 binding proteins of the EF-hand type. So far 19 S100 proteins have been discovered which are involved in the regulation of cell di¡erentiation, cell cycling, signal transduction and cytoskeletal architecture [13,14]. Due to their tissue speci¢c expression pattern, S100 proteins transduce the ubiquitous Ca2 signal into cell-speci¢c signals by hydrophobic interaction with speci¢c intracellular target proteins. Among these proteins, the S100A1 protein displays a speci¢c expression in slow skeletal and cardiac muscle and is involved in the activation of the Ca2 -induced Ca2 release [15]. The EF-hand type Ca2 binding protein S100A1 is strongly associated with the SR [16] revealing a direct protein interaction with the ryanodine receptor [17] and possibly with the SERCA/phospholamban complex, as has been proposed by Western blot overlay studies [18]. Due to this subcellular compartmentation of S100A1 and its high expression level in cardiac myocytes, we hypothesised that S100A1 might play a pivotal role in the cardiospeci¢c regulation of SR function. While a regulation of the SERCA/phospholamban complex by S100A1 still appears hypothetical, S100A1 activates the Ca2 activated Ca2 release in SR vesicle preparations [19] and in permeabilised ¢bres [17]. Our ¢nding of a signi¢cant downregulation of this protein in endstage heart failure [20] therefore may causally be related to a defective excitation^contraction coupling as shown by Gomez et al. [21], who demonstrated reduced Ca2 releases evoked by plasma membrane
Ca2 currents in failing myocardium. Data from D'Agnolo et al. [22] who demonstrated a markedly increased ca¡eine threshold in idiopathic dilated cardiomyopathy would be in favour of this hypothesis. If S100A1 protein levels indeed correlate with the Ca2 release, an increased expression of this protein should therefore be found in the setting of compensated myocardial hypertrophy. 2. Materials and methods The pulmonary arteries of nine male pigs were embolised in three sessions each with 15 mg/kg b.wt. Sephadex G-50. Haemodynamic measurements were performed with a pulmonary artery catheter after percutaneous catheterisation of the femoral vein. The hearts of nine untreated animals served as controls. Myocardial specimens were taken from the free left-ventricular wall and the right-ventricular outlet tract of embolised animals and untreated controls after coronary perfusion of the hearts with 60 ml/kg b.wt. Eurocollins cardioplegic solution. The present investigation represents a substudy of Weimann et al. [23] and conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institute of Health (NIH Publication No. 85-23, revised 1996). Five semithin sections of subendocardial tissue blocks of each animal were evaluated according to a systematic random embedding, random sectioning and random sampling method [24]. Morphometric measurements according to a point-counting method were performed by light-microscopy as described by Oberholzer et al. [25], using an ocular integrated test system (Integrationsplatte, Zeiss, Oberkochen, Germany). The volume densities of the endomysial connective tissue indicating the state of ¢brosis were evaluated at 100-fold magni¢cation on Ladewigstained sections [26]. The minimal myocyte diameters at the cell nucleus level were evaluated at 1000-fold magni¢cation on HE-stained sections and calculated as mean value of 100 cross-sectioned myocytes using a semi-automatic analysis system (Kontron, Eching, Germany) and served as histological indicator for myocyte hypertrophy [27]. Myocardial samples (100 mg) were homogenised in 5 vols. (v/w) of extraction bu¡er (25 mM Tris-HCl,
BBADIS 61906 20-1-00
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255
50 mM KCl, 5 mM EDTA, 0.1 mM PMSF, pH 7.5) and centrifuged (50 000Ug, 15 min). The protein content of resultant supernatants were measured in order to compare the total extractable protein amounts from control and hypertrophied myocardium using the DC Protein Assay (Bio-Rad, Hercules, Ca, USA) as outlined by the manufacturer. The supernatants were then adjusted to 1 mg/ml to optimise ammonium sulfate (AS) precipitation and equilibrated with AS (50% saturation) for 15 min at room temperature. After centrifugation (50 000Ug, 15 min) the AS supernatants were injected onto an Octyl-Sepharose column (Pharmacia, 1 ml) equilibrated with loading bu¡er (25 mM Tris-HCl, 2 mM CaCl2 , pH 7.5). A step gradient was run with elution bu¡er (25 mM Tris-HCl, pH 9.5, 5 mM EGTA, pH 9.5). Eluting proteins were collected in aliquots of 5 ml and then loaded onto an UnoQ-1 anion-exchanger (Bio-Rad) equilibrated with a low-salt bu¡er (25mM Tris-HCl, pH 7.5). Proteins were eluted by a linear gradient from 0 to 40% high-salt bu¡er (25 mM Tris-HCl, 1 M NaCl, pH 7.5) within 40 min (£ow rate 1 ml/min). UV monitoring was performed at 220 nm. Relative protein amounts were calculated by peak integration. The S100A1 peak was identi¢ed by Western blotting and electronspray-ionisationmass-spectrometry (ESI-MS) as described by Pedrocchi et al. [28]. For SDS-PAGE, the samples were homogenised in urea bu¡er (50 mM Tris-HCl, 8 M urea, 5 mM EGTA, pH 8.6) and centrifuged (50 000Ug, 15 min). The resulting supernatants were diluted 1:10 (v/v) in sample bu¡er (0.5 M Tris-HCl, 5% SDS, 10% glycerol, 0.05% Bromphenol blue, 5% L-mercaptoethanol, pH 6.8) and boiled for 3 min. Ten-microlitre aliquots were separated on SDS-polyacrylamide gels (15% T, 3% C) using the Tris-tricine bu¡er-system [29]. The proteins were transferred to PVDF sheets with a semi-dry transfer system [30]. The immunodetection of SERCA2a and phospholamban was performed using the Western Light Plus-kit (Tropix) employing the CSPD chemiluminescence according to the manufacturer's guidelines. Primary antibodies (Biomol) were diluted 1:10 000 in blocking bu¡er, while the working dilution of the biotinlabelled secondary antibody and the streptavidin^alkaline phosphatase conjugate was 1:20 000. Proteins were quanti¢ed by densitometric peak integration.
251
Actin served as an internal standard for equivalent protein load. The mean values of integrated peaks of S100A1, SERCA2a and phospholamban of left ventricular controls was set 100%. Protein amounts are therefore given as percent of control left ventricles þ standard deviation (% þ S.D.). P-values were calculated by the unpaired Student's t-test. 3. Results All pigs showed a moderate pulmonary hypertension after 70 days with three sessions of embolisations. Systolic pulmonary artery pressure increased to 30.7 þ 4.3 mmHg (PAPmean 21 þ 5.5 mmHg) as compared to controls 19 þ 0.6 mmHg (PAPmean 12 þ 0.4 mmHg) (Fig. 1). Right ventricular tissue of embolised animals showed a signi¢cant increase of myocyte diameters at cell nucleus level from 10.41 þ 1.40 to 12.89 þ 2.05 Wm (P 6 0.01) (Fig. 2A,B). Although perimysial connective tissue increased from 1.12 þ 0.44% to 2.39 þ 0.94% (P 6 0.001) the total protein recovery from control and hypertrophied myocardial samples did not di¡er signi¢cantly (2.86 þ 0.17 vs. 2.80 þ 0.23 g/l). Furthermore, no difference was seen in perimysial connective tissue content and myocyte diameters of left ventricular samples. Since commercially available S100A1 antibodies show a considerable crossreactivity with other S100 proteins, such as S100A4 and S100A6 (data not
Fig. 1. Haemodynamic measurements : systolic (PA sys) and mean pulmonary pressures (PA mean) in embolised animals as compared to controls. Bars are mean values with standard deviation.
BBADIS 61906 20-1-00
252
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255
Fig. 2. Histological analyses. (A) Connective tissue content in PAH animals and controls in percent of volume density. (B) Myocyte diameters at cell nucleus level in right ventricles of embolised animals and controls in Wm.
shown), the quantitation of S100A1 in total tissue extracts by Western blot analyses did not seem applicable. We therefore employed a combined HPLC and mass spectrometry methodology to purify and quantitate S100A1. After EDTA extraction of tissue samples, the hydrophobic interaction chromatography (HIC) resulted in an elution of mainly three proteins that interacted in a calcium-dependent manner with the octyl-Sepharose matrix (Fig. 4, lane 2). These proteins were further fractionated by anion exchanger chromatography UnoQ-1 (Fig. 3), revealing molecular masses of 66 734 Da (annexin VI), 10427 Da (S100A1), and 16 789 Da (calmodulin) as detected by ESI-MS. The porcine S100A1 protein
Fig. 3. Elution pro¢le. Tissue extracts of 100 mg from UnoQ-1 anion exchanger (1.3 ml). Annexin VI (4 and 5), S100A1 (6), and calmodulin (7) peaks as con¢rmed by Western blotting and ESI-MS.
Fig. 4. Electrophoresis. (A) Silver stained SDS-PAGE and (B) Western blot with monoclonal anti-S100A1 antibody (Sigma, Munich) of protein peaks eluted from UnoQ-1 anion exchanger. Lane 1, EGTA tissue extract; lane 2, protein sample injected on UnoQ-1; lanes 3^7, protein peaks eluted from UnoQ-1 as numbered in Fig. 3. Protein standards: (A) Bio-Rad prestained low (19^106 kDa) and (B) Tropix biotinylated MW markers (6.5^200 kDa).
peak revealed only one detectable mass (10 427 kDa) that correlated with the calculated molecular mass of rat S100A1 (10 428 kDa). Western blotting con¢rmed the presence of annexin VI (not shown), S100A1 (Fig. 4B) and calmodulin (not shown), respectively. Quantitation of S100A1 by peak integration in control animals revealed a di¡erential expression in left and right ventricular tissues. Right ventricular tissue of control animals showed signi¢cantly lower
Fig. 5. S100A1 protein levels. Left (LV) and right ventricular (RV) tissues of embolised animals and controls. The di¡erences are highly signi¢cant between right ventricular tissues of embolised and control animals (**P 6 0.001). The di¡erential expression of S100A1 is signi¢cant between left and right ventricular tissues of controls (*P 6 0.01) while insigni¢cant in embolised animals.
BBADIS 61906 20-1-00
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255
253
S100A1 protein content for a given myocyte diameter. In our experimental setting, we could not observe any signi¢cant change in the expression of the SR proteins SERCA2a and phospholamban (Fig. 7). 4. Discussion
Fig. 6. S100A1 protein-expression and myocyte diameters: correlation of S100A1 protein-expression with myocyte diameters in control and embolised animals.
relative protein levels of S100A1 amounting to 58 þ 21% compared to LV with 100 þ 31% (P 6 0.01). Embolised animals, however, showed a signi¢cant increase of S100A1 levels in right ventricular tissues with a relative amount of 122 þ 33% compared to control right ventricular tissue (58 þ 21%; P 6 0.001). The signi¢cant di¡erence in S100A1 concentration between right and left ventricular tissue was lost in the embolised animals (Fig. 5). Fig. 6 depicts the correlation of S100A1 protein expression in right ventricular myocardium with the myocyte diameter as found for control (open circles) and embolised (¢lled squares) animals. While in both groups S100A1 expression appears to be negatively correlated with myocyte diameter, pressure overload of the right ventricle leads to a signi¢cant increase of
Fig. 7. SR proteins. Protein levels of phospholamban and SERCA2a in right and left ventricular tissues from embolised animals and controls as determined by Western blotting.
Gwathmey et al. [2] showed, for the ¢rst time, that intracellular Ca2 handling is abnormal in end-stage heart failure re£ecting a defective SR function. Several authors reported a decreased activity of SERCA in end-stage heart failure using cell homogenates [31]. Highly puri¢ed membrane fractions of the sarcoplasmatic reticulum, however, revealed an unchanged stimulated SERCA activity in failing hearts as compared to controls [32]. These con£icting results not only re£ect di¡erent methodological approaches, but also might indicate that factors regulating SERCA activity, such as Ca2 binding proteins of the EFhand type, might be extracted during microsomal preparations. If this hypothesis holds true, a di¡erential expression of Ca2 binding proteins, that are involved in the regulation of SR function, might explain di¡erential Ca2 handling by the SR in crude homogenates and microsomal preparations of hypertrophied and failing hearts. We chose a new model of repetitive pulmonary embolism in order to mimic a gradual development of right ventricular hypertrophy in pigs with a preserved systolic contractility [23]. All experimental animals included showed criteria of compensated hypertrophy with moderately increased systolic pressures, signi¢cantly increased myocyte diameters and concomitant ¢brosis while protein expression levels of SERCA and phospholamban remained una¡ected. As recently reported [20], we found a di¡erential expression of S100A1 in control animals with signi¢cantly higher protein levels in left ventricular tissue as compared to the right ventricular and atrial tissue. Interestingly, Minajeva et al. [33] found that in ca¡eine-induced contracture, atrial ¢bres produce a lower peak tension with a substantial tonic component re£ecting a lower and slower Ca2 release from SR than left ventricular ¢bres. Since S100A1 increases the Ca2 release from the sarcoplasmatic reticulum, a di¡erential expression of this Ca2 binding protein might correlate with a di¡erential regulation
BBADIS 61906 20-1-00
254
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255
of the ryanodine receptor in di¡erent areas of the heart. In the present study, pulmonary hypertension induced a selective and signi¢cant increase of S100A1 protein expression in right ventricular tissues. Left ventricular expression levels were una¡ected, insinuating a locally controlled expression of S100A1. To date, there are no data available as to the regulation of S100A1 gene expression in mammals. Promoter studies of the rat S100A1 gene [34], however, suggest a possible mechanism for an upregulation of S100A1 in hypertrophy. Since the rat S100A1 promoter contains several putative cAMP regulatory sequences on a positive regulatory segment, a stimulation of the Ladrenergic pathway [35] as found in early and compensated myocardial hypertrophy, might induce overexpression of the Ca2 binding protein S100A1. The transition from compensated hypertrophy to heart failure in turn was reported to be correlated with decreased cAMP levels [36] and thus would explain the observed downregulation of S100A1 in explanted myocardium from patients with end-stage heart failure. Interestingly, S100A1 expression was negatively correlated with the myocyte diameter both in the control and embolised animal group, while the latter group showed higher S100A1 levels for a given myocyte diameter. The molecular basis of this ¢nding is still unclear. Data from Baudier et al. [37], however, show that S100A1 inhibits the phosphorylation of protein kinase C target proteins. Since the PKC pathway is involved in the formation of cardiac hypertrophy [38], S100A1 might therefore also exert an antihypertrophic e¡ect. Recently Tsutsui et al. [39] showed that an inhibition of microtubule polymerisation by chronic colchicine administration attenuates the progression of myocardial hypertrophy in spontaneously hypertensive rats independently of blood pressure. Since Donato [40] documented an inhibition of microtuble polymerisation by S100A1, an increased expression of S100A1 might show an antihypertrophic e¡ect by modulating the cytoskeleton. In conclusion, our study shows for the ¢rst time that the Ca2 binding protein S100A1 is di¡erentially expressed in right and left ventricular myocardium, revealing a signi¢cant upregulation upon pressure overload after pulmonary artery embolisation. Since S100A1 is involved in the regulation of the Ca2
release channel of the SR, a di¡erential expression of S100A1 in di¡erent states of myocardial hypertrophy would be an attractive explanation for changes in Ca2 -handling found in the transition from compensated hypertrophy to heart failure. Acknowledgements Muscle specimens derived from an experimental heart^lung transplantation program and were kindly provided by Prof. H. Jakob, Department of Heart Surgery, University of Heidelberg, Germany.
References [1] D.M. Bers, J.W.M. Bassani, R.A. Bassani, Cardiovasc. Res. 27 (1993) 1772^1777. [2] J.K. Gwathmey, L. Copelas, R. MacKinnon, F.J. Schoen, M.D. Feldman, W. Grossman, J.P. Morgan, Circ. Res. 61 (1987) 70^76. [3] L.A. Mulieri, G. Hasenfuss, B. Leavitt, P.D. Allen, N.R. Alpert, Circulation 85 (1992) 1743^1750. [4] B. Pieske, M. Su«tterlin, M. Schmidt-Schweda, K. Minami, M. Meyer, M. Olschewski, C. Holubarsch, H. Just, G. Hasenfuss, J. Clin. Invest. 98 (1996) 764^776. [5] D.J. Beuckelmann, M. Nabauer, E. Erdmann, Circulation 85 (1992) 1046^1055. [6] J.K. Gwathmey, L.A. Bentivegna, B.J. Ransil, W. Grossman, J.P. Morgan, Cardiovasc. Res. 27 (1993) 199^203. [7] M. Arai, N.R. Alpert, D.H. MacLennan, P. Barton, M. Periasamy, Circ. Res. 72 (1993) 463^469. [8] M.A. Movsesian, M. Karimi, K. Green, L.R. Jones, Circulation 90 (1994) 653^657. [9] J.J. Mackrill, R.A.J. Challiss, D.A. O'Connell, F.A. Lai, S.R. Nahorski, Biochem. J. 327 (1997) 251^258. [10] W.A. Jackson, J. Colyer, Biochem. J. 316 (1996) 201^207. [11] A.J. Lokuta, T.B. Rogers, W.J. Lederer, H.H. Valdivia, J. Physiol. 487 (1995) 609^622. [12] C.D. Jeck, R. Zimmermann, J. Schaper, W. Schaper, J. Mol. Cell. Cardiol. 26 (1994) 99^107. [13] B.W. Scha«fer, C.W. Heizmann, Trends Biochem. Sci. 21 (1996) 134^140. [14] R. Donato, Biochim. Biophys. Acta 1450 (1999) 191^231. [15] S. Treves, E. Scutari, M. Robert, S. Groh, M. Ottolia, G. Prestipino, M. Ronjat, F. Zorzato, Biochemistry 36 (1997) 11496^11503. [16] A. Mandinova, D. Atar, B.W. Scha«fer, M. Spiess, U. Aebi, C.W. Heizmann, J. Cell Sci. 111 (1998) 2043^2054. [17] A. Remppis, P. Ehlermann, P. Most, C. Weber, C.W. Heizmann, R.H.A. Fink, H.A. Katus, Eur. Heart J. 18 (Suppl.) (1997) 643.
BBADIS 61906 20-1-00
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255 [18] D.B. Zimmer, Cell Motil. Cytoskeleton 20 (1991) 325^337. [19] G. Fano, V. Marsili, P. Angelella, M.C. Aisa, I. Giambanco, R. Donato, FEBS Lett. 255 (1989) 381^384. [20] A. Remppis, T. Greten, B.W. Scha«fer, P. Hunziker, P. Erne, H.A. Katus, C.W. Heizmann, Biochim. Biophys. Acta 1313 (1996) 253^257. [21] A.M. Gomez, H.H. Valdivia, H. Cheng, M.R. Lederer, L.F. Santana, M.B. Cannell, S.A. McCune, R.A. Altschuld, W.J. Lederer, Science 276 (1997) 800^806. [22] A. D'Agnolo, G.B. Luciani, A. Mazzucco, V. Gallucci, G. Salviati, Circulation 85 (1992) 518^525. [23] J. Weimann, W. Zink, P. Schnabel, H. Jakob, M.M. Gebhard, E. Martin, J. Motsch, Crit. Care, in press. [24] E.R. Weibel, Stereological Methods for Biological Morphometry, Academic Press, New York, 1980. [25] M. Oberholzer, R.A. Ettlin, G. Kloppel, M. Lohr, M.W. Buser, M.J. Mihatsch, H. Christen, U. Marti, P.U. Heitz, Anal. Quant. Cytol. Histol. 9 (1987) 123^132. [26] P.A. Schnabel, R. Lange, K. Amann, T.M. Bingold, H. Jakob, M. Kampmann, A. Koch, B. Koch, A. Magener, H. Mehmanesch, Z. Niemir, F.U. Sack, M. Schmid, A. Schmiedl, R. Waldherr, R. Zimmermann, S. Hagl, H.F. Otto, Z. Herz-, Thor-, Gefa«Mchir. 11 (Suppl.) (1997) 7^22. [27] P.A. Schnabel, K. Amann, H. Ghaderipour, R. Lange, F.U. Sack, S. Knapp, A. Magener, R. Zimmermann, S. Hagl, H.F. Otto, J. Heart Lung Transplant. S13 (1994) 81. [28] M. Pedrocchi, C.R. Hauer, B.W. Scha«fer, P. Erne, C.W. Heizmann, Biochem. Biophys. Res. Commun. 197 (1993) 529^535.
255
[29] H. Scha«gger, G. von Jagow, Anal. Biochem. 166 (1987) 368^ 379. [30] H. Towbin, T. Staehlin, J. Gordon, Proc. Natl. Acad. Sci. USA 76 (1979) 4350^4354. [31] M.A. Movsesian, R.H. Schwinger, Cardiovasc. Res. 37 (1998) 352^359. [32] M.A. Movsesian, M.R. Bristow, J. Krall, Circ. Res. 65 (1989) 1141^1144. [33] A. Minajeva, A. Kaasik, K. Paju, E. Seppet, A.M. Lompre, V. Veksler, R. Ventura-Clapier, Am. J. Physiol. 273 (1997) H2498^H2507. [34] W. Song, D.B. Zimmer, Brain Res. 721 (1996) 204^216. [35] T. Yamazaki, I. Komuro, Y. Zou, S. Kudoh, T. Mizuno, Y. Hiroi, I. Shiojima, H. Takano, K.I. Kinugawa, O. Kohmoto, T. Takahashi, Y. Yazaki, J. Mol. Cell. Cardiol. 29 (1997) 2491^2501. [36] M. Bo«hm, B. Reiger, R.H. Schwinger, E. Erdmann, Cardiovasc. Res. 28 (1994) 1713^1719. [37] J. Baudier, E. Bergeret, N. Bertacchi, H. Weintraub, J. Gagnon, J. Garin, Biochemistry 34 (1995) 7834^7846. [38] N.S. Dhalla, Y.J. Xu, S.S. Sheu, P.S. Tappia, V. Panagia, J. Mol. Cell. Cardiol. 29 (1997) 2865^2871. [39] H. Tsutsui, Y. Ishibashi, M. Takahashi, T. Namba, H. Tagawa, K. Imanaka-Yoshida, A. Takeshita, J. Mol. Cell. Cardiol. 31 (1999) 1203^1213. [40] R. Donato, FEBS Lett. 162 (1983) 310^313.
BBADIS 61906 20-1-00
Right ventricular upregulation of the Ca2 binding protein S100A1 in chronic pulmonary hypertension Philipp Ehlermann a , Andrew Remppis a; *, Oliver Guddat b , Jo«rg Weimann c , Philipp A. Schnabel b , Johann Motsch c , Claus W. Heizmann d , Hugo A. Katus
a
a
Medizinische Klinik II, Medizinische Universita«t zu Lu«beck, Ratzeburger Allee 160, D-23538 Lu«beck, Germany Pathologisches Institut, Universita«t Heidelberg, Im Neuenheimer Feld 220/221, D-69120 Heidelberg, Germany c Klinik fu«r Ana«sthesiologie, Universita«t Heidelberg, Im Neuenheimer Feld 110, D-69120 Heidelberg, Germany Abteilung fu«r Klinische Chemie und Biochemie, Kinderspital Zu«rich, Steinwiesstr. 75, CH-8032 Zu«rich, Switzerland b
d
Received 27 July 1999; received in revised form 22 October 1999; accepted 10 November 1999
Abstract The Ca2 binding protein S100A1 increases the Ca2 release from the sarcoplasmatic reticulum by interacting with the ryanodine receptor. In order to understand whether this effect might be operative in the early course of hypertrophy, when myocardium is able to meet increased workload, we investigated the expression of S100A1 in a model of moderate right ventricular hypertrophy. The pulmonary arteries of nine pigs were embolised three times with Sephadex G-50. After 70 days, all pigs showed a moderate pulmonary hypertension. Right ventricular tissue of embolised animals showed a significant increase of connective tissue and enlargement of myocyte diameters. In controls, we found a differential expression of S100A1 with significantly lower S100A1 protein levels in right ventricular compared to left ventricular tissue. In pulmonary hypertension, S100A1 expression increased significantly in hypertrophied right ventricles while it was unchanged in left ventricular tissue. No change was observed in the expression of SERCA2a and phospholamban. Our data show, for the first time, that moderate pressure overload results in an upregulation of S100A1. This may reflect an adaptive response of myocardial Ca2 homeostasis to a higher workload. ß 2000 Elsevier Science B.V. All rights reserved. Keywords: Calcium-binding protein; S100A1; Myocardial hypertrophy; Pulmonary hypertension; SERCA; Phospholamban
1. Introduction The intracellular Ca2 cycling is largely determined by the activity of the sarcoplasmatic reticulum (SR) [1]. Since Ca2 cycling is compromised in end stage heart failure, studies have focussed on the functional state of the SR in patients that underwent
* Corresponding author. Fax: +49-451-500-6437; E-mail: [email protected]
cardiac transplantation [2]. Mulieri et al. [3] and Pieske et al. [4] demonstrated a reversed force-frequency relationship and an altered post-rest potentiation in isolated muscle ¢bres, indicating a dysfunction of the SR in end stage heart failure. However, the underlying molecular mechanisms are not fully understood. Diminished systolic Ca2 peaks and increased diastolic Ca2 levels [5,6] indicate an altered activity of the sarcoplasmatic Ca2 -release channel (ryanodine receptor, RyR) and Ca2 ATPase (SERCA), respectively. An altered expression of SR proteins has therefore been discussed as one of the pos-
0925-4439 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 5 - 4 4 3 9 ( 9 9 ) 0 0 1 0 6 - 4
BBADIS 61906 20-1-00
250
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255
sible molecular causes of SR dysfunction. Results, however, are still contradictory, possibly due to different methodological approaches [7,8]. Since the activity of RyR and SERCA are tightly regulated [9], an altered regulation of Ca2 transients might be due to SR dysfunction in end-stage heart failure. It has been shown that Ca2 binding proteins of the EF-hand type di¡erentially regulate SR Ca2 transients. Calmodulin is known to activate SERCA via calmodulin kinase-dependent phosphorylation of phospholamban [10] and it may also regulate the activity of the ryanodine receptor (RyR) [11]. However, it is unknown whether calmodulin protein concentrations are altered in the human cardiomyopathic heart since only mRNA data are available [12]. Like calmodulin, S100 proteins also belong to the large family of Ca2 binding proteins of the EF-hand type. So far 19 S100 proteins have been discovered which are involved in the regulation of cell di¡erentiation, cell cycling, signal transduction and cytoskeletal architecture [13,14]. Due to their tissue speci¢c expression pattern, S100 proteins transduce the ubiquitous Ca2 signal into cell-speci¢c signals by hydrophobic interaction with speci¢c intracellular target proteins. Among these proteins, the S100A1 protein displays a speci¢c expression in slow skeletal and cardiac muscle and is involved in the activation of the Ca2 -induced Ca2 release [15]. The EF-hand type Ca2 binding protein S100A1 is strongly associated with the SR [16] revealing a direct protein interaction with the ryanodine receptor [17] and possibly with the SERCA/phospholamban complex, as has been proposed by Western blot overlay studies [18]. Due to this subcellular compartmentation of S100A1 and its high expression level in cardiac myocytes, we hypothesised that S100A1 might play a pivotal role in the cardiospeci¢c regulation of SR function. While a regulation of the SERCA/phospholamban complex by S100A1 still appears hypothetical, S100A1 activates the Ca2 activated Ca2 release in SR vesicle preparations [19] and in permeabilised ¢bres [17]. Our ¢nding of a signi¢cant downregulation of this protein in endstage heart failure [20] therefore may causally be related to a defective excitation^contraction coupling as shown by Gomez et al. [21], who demonstrated reduced Ca2 releases evoked by plasma membrane
Ca2 currents in failing myocardium. Data from D'Agnolo et al. [22] who demonstrated a markedly increased ca¡eine threshold in idiopathic dilated cardiomyopathy would be in favour of this hypothesis. If S100A1 protein levels indeed correlate with the Ca2 release, an increased expression of this protein should therefore be found in the setting of compensated myocardial hypertrophy. 2. Materials and methods The pulmonary arteries of nine male pigs were embolised in three sessions each with 15 mg/kg b.wt. Sephadex G-50. Haemodynamic measurements were performed with a pulmonary artery catheter after percutaneous catheterisation of the femoral vein. The hearts of nine untreated animals served as controls. Myocardial specimens were taken from the free left-ventricular wall and the right-ventricular outlet tract of embolised animals and untreated controls after coronary perfusion of the hearts with 60 ml/kg b.wt. Eurocollins cardioplegic solution. The present investigation represents a substudy of Weimann et al. [23] and conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institute of Health (NIH Publication No. 85-23, revised 1996). Five semithin sections of subendocardial tissue blocks of each animal were evaluated according to a systematic random embedding, random sectioning and random sampling method [24]. Morphometric measurements according to a point-counting method were performed by light-microscopy as described by Oberholzer et al. [25], using an ocular integrated test system (Integrationsplatte, Zeiss, Oberkochen, Germany). The volume densities of the endomysial connective tissue indicating the state of ¢brosis were evaluated at 100-fold magni¢cation on Ladewigstained sections [26]. The minimal myocyte diameters at the cell nucleus level were evaluated at 1000-fold magni¢cation on HE-stained sections and calculated as mean value of 100 cross-sectioned myocytes using a semi-automatic analysis system (Kontron, Eching, Germany) and served as histological indicator for myocyte hypertrophy [27]. Myocardial samples (100 mg) were homogenised in 5 vols. (v/w) of extraction bu¡er (25 mM Tris-HCl,
BBADIS 61906 20-1-00
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255
50 mM KCl, 5 mM EDTA, 0.1 mM PMSF, pH 7.5) and centrifuged (50 000Ug, 15 min). The protein content of resultant supernatants were measured in order to compare the total extractable protein amounts from control and hypertrophied myocardium using the DC Protein Assay (Bio-Rad, Hercules, Ca, USA) as outlined by the manufacturer. The supernatants were then adjusted to 1 mg/ml to optimise ammonium sulfate (AS) precipitation and equilibrated with AS (50% saturation) for 15 min at room temperature. After centrifugation (50 000Ug, 15 min) the AS supernatants were injected onto an Octyl-Sepharose column (Pharmacia, 1 ml) equilibrated with loading bu¡er (25 mM Tris-HCl, 2 mM CaCl2 , pH 7.5). A step gradient was run with elution bu¡er (25 mM Tris-HCl, pH 9.5, 5 mM EGTA, pH 9.5). Eluting proteins were collected in aliquots of 5 ml and then loaded onto an UnoQ-1 anion-exchanger (Bio-Rad) equilibrated with a low-salt bu¡er (25mM Tris-HCl, pH 7.5). Proteins were eluted by a linear gradient from 0 to 40% high-salt bu¡er (25 mM Tris-HCl, 1 M NaCl, pH 7.5) within 40 min (£ow rate 1 ml/min). UV monitoring was performed at 220 nm. Relative protein amounts were calculated by peak integration. The S100A1 peak was identi¢ed by Western blotting and electronspray-ionisationmass-spectrometry (ESI-MS) as described by Pedrocchi et al. [28]. For SDS-PAGE, the samples were homogenised in urea bu¡er (50 mM Tris-HCl, 8 M urea, 5 mM EGTA, pH 8.6) and centrifuged (50 000Ug, 15 min). The resulting supernatants were diluted 1:10 (v/v) in sample bu¡er (0.5 M Tris-HCl, 5% SDS, 10% glycerol, 0.05% Bromphenol blue, 5% L-mercaptoethanol, pH 6.8) and boiled for 3 min. Ten-microlitre aliquots were separated on SDS-polyacrylamide gels (15% T, 3% C) using the Tris-tricine bu¡er-system [29]. The proteins were transferred to PVDF sheets with a semi-dry transfer system [30]. The immunodetection of SERCA2a and phospholamban was performed using the Western Light Plus-kit (Tropix) employing the CSPD chemiluminescence according to the manufacturer's guidelines. Primary antibodies (Biomol) were diluted 1:10 000 in blocking bu¡er, while the working dilution of the biotinlabelled secondary antibody and the streptavidin^alkaline phosphatase conjugate was 1:20 000. Proteins were quanti¢ed by densitometric peak integration.
251
Actin served as an internal standard for equivalent protein load. The mean values of integrated peaks of S100A1, SERCA2a and phospholamban of left ventricular controls was set 100%. Protein amounts are therefore given as percent of control left ventricles þ standard deviation (% þ S.D.). P-values were calculated by the unpaired Student's t-test. 3. Results All pigs showed a moderate pulmonary hypertension after 70 days with three sessions of embolisations. Systolic pulmonary artery pressure increased to 30.7 þ 4.3 mmHg (PAPmean 21 þ 5.5 mmHg) as compared to controls 19 þ 0.6 mmHg (PAPmean 12 þ 0.4 mmHg) (Fig. 1). Right ventricular tissue of embolised animals showed a signi¢cant increase of myocyte diameters at cell nucleus level from 10.41 þ 1.40 to 12.89 þ 2.05 Wm (P 6 0.01) (Fig. 2A,B). Although perimysial connective tissue increased from 1.12 þ 0.44% to 2.39 þ 0.94% (P 6 0.001) the total protein recovery from control and hypertrophied myocardial samples did not di¡er signi¢cantly (2.86 þ 0.17 vs. 2.80 þ 0.23 g/l). Furthermore, no difference was seen in perimysial connective tissue content and myocyte diameters of left ventricular samples. Since commercially available S100A1 antibodies show a considerable crossreactivity with other S100 proteins, such as S100A4 and S100A6 (data not
Fig. 1. Haemodynamic measurements : systolic (PA sys) and mean pulmonary pressures (PA mean) in embolised animals as compared to controls. Bars are mean values with standard deviation.
BBADIS 61906 20-1-00
252
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255
Fig. 2. Histological analyses. (A) Connective tissue content in PAH animals and controls in percent of volume density. (B) Myocyte diameters at cell nucleus level in right ventricles of embolised animals and controls in Wm.
shown), the quantitation of S100A1 in total tissue extracts by Western blot analyses did not seem applicable. We therefore employed a combined HPLC and mass spectrometry methodology to purify and quantitate S100A1. After EDTA extraction of tissue samples, the hydrophobic interaction chromatography (HIC) resulted in an elution of mainly three proteins that interacted in a calcium-dependent manner with the octyl-Sepharose matrix (Fig. 4, lane 2). These proteins were further fractionated by anion exchanger chromatography UnoQ-1 (Fig. 3), revealing molecular masses of 66 734 Da (annexin VI), 10427 Da (S100A1), and 16 789 Da (calmodulin) as detected by ESI-MS. The porcine S100A1 protein
Fig. 3. Elution pro¢le. Tissue extracts of 100 mg from UnoQ-1 anion exchanger (1.3 ml). Annexin VI (4 and 5), S100A1 (6), and calmodulin (7) peaks as con¢rmed by Western blotting and ESI-MS.
Fig. 4. Electrophoresis. (A) Silver stained SDS-PAGE and (B) Western blot with monoclonal anti-S100A1 antibody (Sigma, Munich) of protein peaks eluted from UnoQ-1 anion exchanger. Lane 1, EGTA tissue extract; lane 2, protein sample injected on UnoQ-1; lanes 3^7, protein peaks eluted from UnoQ-1 as numbered in Fig. 3. Protein standards: (A) Bio-Rad prestained low (19^106 kDa) and (B) Tropix biotinylated MW markers (6.5^200 kDa).
peak revealed only one detectable mass (10 427 kDa) that correlated with the calculated molecular mass of rat S100A1 (10 428 kDa). Western blotting con¢rmed the presence of annexin VI (not shown), S100A1 (Fig. 4B) and calmodulin (not shown), respectively. Quantitation of S100A1 by peak integration in control animals revealed a di¡erential expression in left and right ventricular tissues. Right ventricular tissue of control animals showed signi¢cantly lower
Fig. 5. S100A1 protein levels. Left (LV) and right ventricular (RV) tissues of embolised animals and controls. The di¡erences are highly signi¢cant between right ventricular tissues of embolised and control animals (**P 6 0.001). The di¡erential expression of S100A1 is signi¢cant between left and right ventricular tissues of controls (*P 6 0.01) while insigni¢cant in embolised animals.
BBADIS 61906 20-1-00
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255
253
S100A1 protein content for a given myocyte diameter. In our experimental setting, we could not observe any signi¢cant change in the expression of the SR proteins SERCA2a and phospholamban (Fig. 7). 4. Discussion
Fig. 6. S100A1 protein-expression and myocyte diameters: correlation of S100A1 protein-expression with myocyte diameters in control and embolised animals.
relative protein levels of S100A1 amounting to 58 þ 21% compared to LV with 100 þ 31% (P 6 0.01). Embolised animals, however, showed a signi¢cant increase of S100A1 levels in right ventricular tissues with a relative amount of 122 þ 33% compared to control right ventricular tissue (58 þ 21%; P 6 0.001). The signi¢cant di¡erence in S100A1 concentration between right and left ventricular tissue was lost in the embolised animals (Fig. 5). Fig. 6 depicts the correlation of S100A1 protein expression in right ventricular myocardium with the myocyte diameter as found for control (open circles) and embolised (¢lled squares) animals. While in both groups S100A1 expression appears to be negatively correlated with myocyte diameter, pressure overload of the right ventricle leads to a signi¢cant increase of
Fig. 7. SR proteins. Protein levels of phospholamban and SERCA2a in right and left ventricular tissues from embolised animals and controls as determined by Western blotting.
Gwathmey et al. [2] showed, for the ¢rst time, that intracellular Ca2 handling is abnormal in end-stage heart failure re£ecting a defective SR function. Several authors reported a decreased activity of SERCA in end-stage heart failure using cell homogenates [31]. Highly puri¢ed membrane fractions of the sarcoplasmatic reticulum, however, revealed an unchanged stimulated SERCA activity in failing hearts as compared to controls [32]. These con£icting results not only re£ect di¡erent methodological approaches, but also might indicate that factors regulating SERCA activity, such as Ca2 binding proteins of the EFhand type, might be extracted during microsomal preparations. If this hypothesis holds true, a di¡erential expression of Ca2 binding proteins, that are involved in the regulation of SR function, might explain di¡erential Ca2 handling by the SR in crude homogenates and microsomal preparations of hypertrophied and failing hearts. We chose a new model of repetitive pulmonary embolism in order to mimic a gradual development of right ventricular hypertrophy in pigs with a preserved systolic contractility [23]. All experimental animals included showed criteria of compensated hypertrophy with moderately increased systolic pressures, signi¢cantly increased myocyte diameters and concomitant ¢brosis while protein expression levels of SERCA and phospholamban remained una¡ected. As recently reported [20], we found a di¡erential expression of S100A1 in control animals with signi¢cantly higher protein levels in left ventricular tissue as compared to the right ventricular and atrial tissue. Interestingly, Minajeva et al. [33] found that in ca¡eine-induced contracture, atrial ¢bres produce a lower peak tension with a substantial tonic component re£ecting a lower and slower Ca2 release from SR than left ventricular ¢bres. Since S100A1 increases the Ca2 release from the sarcoplasmatic reticulum, a di¡erential expression of this Ca2 binding protein might correlate with a di¡erential regulation
BBADIS 61906 20-1-00
254
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255
of the ryanodine receptor in di¡erent areas of the heart. In the present study, pulmonary hypertension induced a selective and signi¢cant increase of S100A1 protein expression in right ventricular tissues. Left ventricular expression levels were una¡ected, insinuating a locally controlled expression of S100A1. To date, there are no data available as to the regulation of S100A1 gene expression in mammals. Promoter studies of the rat S100A1 gene [34], however, suggest a possible mechanism for an upregulation of S100A1 in hypertrophy. Since the rat S100A1 promoter contains several putative cAMP regulatory sequences on a positive regulatory segment, a stimulation of the Ladrenergic pathway [35] as found in early and compensated myocardial hypertrophy, might induce overexpression of the Ca2 binding protein S100A1. The transition from compensated hypertrophy to heart failure in turn was reported to be correlated with decreased cAMP levels [36] and thus would explain the observed downregulation of S100A1 in explanted myocardium from patients with end-stage heart failure. Interestingly, S100A1 expression was negatively correlated with the myocyte diameter both in the control and embolised animal group, while the latter group showed higher S100A1 levels for a given myocyte diameter. The molecular basis of this ¢nding is still unclear. Data from Baudier et al. [37], however, show that S100A1 inhibits the phosphorylation of protein kinase C target proteins. Since the PKC pathway is involved in the formation of cardiac hypertrophy [38], S100A1 might therefore also exert an antihypertrophic e¡ect. Recently Tsutsui et al. [39] showed that an inhibition of microtubule polymerisation by chronic colchicine administration attenuates the progression of myocardial hypertrophy in spontaneously hypertensive rats independently of blood pressure. Since Donato [40] documented an inhibition of microtuble polymerisation by S100A1, an increased expression of S100A1 might show an antihypertrophic e¡ect by modulating the cytoskeleton. In conclusion, our study shows for the ¢rst time that the Ca2 binding protein S100A1 is di¡erentially expressed in right and left ventricular myocardium, revealing a signi¢cant upregulation upon pressure overload after pulmonary artery embolisation. Since S100A1 is involved in the regulation of the Ca2
release channel of the SR, a di¡erential expression of S100A1 in di¡erent states of myocardial hypertrophy would be an attractive explanation for changes in Ca2 -handling found in the transition from compensated hypertrophy to heart failure. Acknowledgements Muscle specimens derived from an experimental heart^lung transplantation program and were kindly provided by Prof. H. Jakob, Department of Heart Surgery, University of Heidelberg, Germany.
References [1] D.M. Bers, J.W.M. Bassani, R.A. Bassani, Cardiovasc. Res. 27 (1993) 1772^1777. [2] J.K. Gwathmey, L. Copelas, R. MacKinnon, F.J. Schoen, M.D. Feldman, W. Grossman, J.P. Morgan, Circ. Res. 61 (1987) 70^76. [3] L.A. Mulieri, G. Hasenfuss, B. Leavitt, P.D. Allen, N.R. Alpert, Circulation 85 (1992) 1743^1750. [4] B. Pieske, M. Su«tterlin, M. Schmidt-Schweda, K. Minami, M. Meyer, M. Olschewski, C. Holubarsch, H. Just, G. Hasenfuss, J. Clin. Invest. 98 (1996) 764^776. [5] D.J. Beuckelmann, M. Nabauer, E. Erdmann, Circulation 85 (1992) 1046^1055. [6] J.K. Gwathmey, L.A. Bentivegna, B.J. Ransil, W. Grossman, J.P. Morgan, Cardiovasc. Res. 27 (1993) 199^203. [7] M. Arai, N.R. Alpert, D.H. MacLennan, P. Barton, M. Periasamy, Circ. Res. 72 (1993) 463^469. [8] M.A. Movsesian, M. Karimi, K. Green, L.R. Jones, Circulation 90 (1994) 653^657. [9] J.J. Mackrill, R.A.J. Challiss, D.A. O'Connell, F.A. Lai, S.R. Nahorski, Biochem. J. 327 (1997) 251^258. [10] W.A. Jackson, J. Colyer, Biochem. J. 316 (1996) 201^207. [11] A.J. Lokuta, T.B. Rogers, W.J. Lederer, H.H. Valdivia, J. Physiol. 487 (1995) 609^622. [12] C.D. Jeck, R. Zimmermann, J. Schaper, W. Schaper, J. Mol. Cell. Cardiol. 26 (1994) 99^107. [13] B.W. Scha«fer, C.W. Heizmann, Trends Biochem. Sci. 21 (1996) 134^140. [14] R. Donato, Biochim. Biophys. Acta 1450 (1999) 191^231. [15] S. Treves, E. Scutari, M. Robert, S. Groh, M. Ottolia, G. Prestipino, M. Ronjat, F. Zorzato, Biochemistry 36 (1997) 11496^11503. [16] A. Mandinova, D. Atar, B.W. Scha«fer, M. Spiess, U. Aebi, C.W. Heizmann, J. Cell Sci. 111 (1998) 2043^2054. [17] A. Remppis, P. Ehlermann, P. Most, C. Weber, C.W. Heizmann, R.H.A. Fink, H.A. Katus, Eur. Heart J. 18 (Suppl.) (1997) 643.
BBADIS 61906 20-1-00
P. Ehlermann et al. / Biochimica et Biophysica Acta 1500 (2000) 249^255 [18] D.B. Zimmer, Cell Motil. Cytoskeleton 20 (1991) 325^337. [19] G. Fano, V. Marsili, P. Angelella, M.C. Aisa, I. Giambanco, R. Donato, FEBS Lett. 255 (1989) 381^384. [20] A. Remppis, T. Greten, B.W. Scha«fer, P. Hunziker, P. Erne, H.A. Katus, C.W. Heizmann, Biochim. Biophys. Acta 1313 (1996) 253^257. [21] A.M. Gomez, H.H. Valdivia, H. Cheng, M.R. Lederer, L.F. Santana, M.B. Cannell, S.A. McCune, R.A. Altschuld, W.J. Lederer, Science 276 (1997) 800^806. [22] A. D'Agnolo, G.B. Luciani, A. Mazzucco, V. Gallucci, G. Salviati, Circulation 85 (1992) 518^525. [23] J. Weimann, W. Zink, P. Schnabel, H. Jakob, M.M. Gebhard, E. Martin, J. Motsch, Crit. Care, in press. [24] E.R. Weibel, Stereological Methods for Biological Morphometry, Academic Press, New York, 1980. [25] M. Oberholzer, R.A. Ettlin, G. Kloppel, M. Lohr, M.W. Buser, M.J. Mihatsch, H. Christen, U. Marti, P.U. Heitz, Anal. Quant. Cytol. Histol. 9 (1987) 123^132. [26] P.A. Schnabel, R. Lange, K. Amann, T.M. Bingold, H. Jakob, M. Kampmann, A. Koch, B. Koch, A. Magener, H. Mehmanesch, Z. Niemir, F.U. Sack, M. Schmid, A. Schmiedl, R. Waldherr, R. Zimmermann, S. Hagl, H.F. Otto, Z. Herz-, Thor-, Gefa«Mchir. 11 (Suppl.) (1997) 7^22. [27] P.A. Schnabel, K. Amann, H. Ghaderipour, R. Lange, F.U. Sack, S. Knapp, A. Magener, R. Zimmermann, S. Hagl, H.F. Otto, J. Heart Lung Transplant. S13 (1994) 81. [28] M. Pedrocchi, C.R. Hauer, B.W. Scha«fer, P. Erne, C.W. Heizmann, Biochem. Biophys. Res. Commun. 197 (1993) 529^535.
255
[29] H. Scha«gger, G. von Jagow, Anal. Biochem. 166 (1987) 368^ 379. [30] H. Towbin, T. Staehlin, J. Gordon, Proc. Natl. Acad. Sci. USA 76 (1979) 4350^4354. [31] M.A. Movsesian, R.H. Schwinger, Cardiovasc. Res. 37 (1998) 352^359. [32] M.A. Movsesian, M.R. Bristow, J. Krall, Circ. Res. 65 (1989) 1141^1144. [33] A. Minajeva, A. Kaasik, K. Paju, E. Seppet, A.M. Lompre, V. Veksler, R. Ventura-Clapier, Am. J. Physiol. 273 (1997) H2498^H2507. [34] W. Song, D.B. Zimmer, Brain Res. 721 (1996) 204^216. [35] T. Yamazaki, I. Komuro, Y. Zou, S. Kudoh, T. Mizuno, Y. Hiroi, I. Shiojima, H. Takano, K.I. Kinugawa, O. Kohmoto, T. Takahashi, Y. Yazaki, J. Mol. Cell. Cardiol. 29 (1997) 2491^2501. [36] M. Bo«hm, B. Reiger, R.H. Schwinger, E. Erdmann, Cardiovasc. Res. 28 (1994) 1713^1719. [37] J. Baudier, E. Bergeret, N. Bertacchi, H. Weintraub, J. Gagnon, J. Garin, Biochemistry 34 (1995) 7834^7846. [38] N.S. Dhalla, Y.J. Xu, S.S. Sheu, P.S. Tappia, V. Panagia, J. Mol. Cell. Cardiol. 29 (1997) 2865^2871. [39] H. Tsutsui, Y. Ishibashi, M. Takahashi, T. Namba, H. Tagawa, K. Imanaka-Yoshida, A. Takeshita, J. Mol. Cell. Cardiol. 31 (1999) 1203^1213. [40] R. Donato, FEBS Lett. 162 (1983) 310^313.
BBADIS 61906 20-1-00