THEJOURNAL OF BroLoC1c.a CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Ine
Vol. 269,No. 46,Issue of November 18,pp. 28564-28590, 1994 Printed in U.S.A.
Binding of a New Ca2+Sensitizer, Levosimendan, to Recombinant Human Cardiac Troponin C A MOLECULAR MODELLING, FLUORESCENCE PROBE, AND PROTON NUCLEAR MAGNETIC RESONANCE STUDY* (Received for publication, June 21, 1994, and in revised form, August 24, 1994)
Piero PolleselloSO, Martti OvaskaS,Juha Kaivolai, Carola TilgmannS, Kenneth LundstromS, Nisse Kalkkinenn, Ism0 UlmanenS, Erkki NissinenS,and Jyrki TaskinenS From $.Orion Corp., Orion-Farmos, Orion Research, P 0. Box 65, FIN-02101 Espoo, Finland and the IZnstitute of Biotechnology, University of Helsinki, r! 0. Box 17, FIN-00014 Helsinki, Finland
The binding of a new calcium sensitizer, levosimenof both skeletal and cardiac TnC’ has been elucidated (11-13). dan, to human cardiac troponin C (cTnC) is described. It has been well established that a Ca2+-induced conformational Fluorescence studies done on dansylated recombinant change in the NH,-terminal domain of TnC regulates muscle humancTnCand a site-directed mutant showed that contraction (14-17). When Ca2+is bound, a hydrophobic patch levosimendan modulated the calcium-induced conforis formed in the N-domain. This newly exposed surface has mational change in cTnC, and revealed the role ofAsp-88 in the binding of hydrophobic been proposed to participate of the drug to the NHz-terminal domain of drugs. Additionally, i t has been suggested that Aspaa plays an in the binding cTnC. Furthermore,NlMR studies performed on the NH,- important role in the bindingof drugs that modulate troponin terminal fragment of cTnC showed a spatial proximity C function (18). Models for the binding of calcium sensitizers to between levosimendan and Mets1, MetBS, and Phe” in the cardiac troponin C have been described (18, 19). drug-protein complex. These data were used to build an optimized model of the drug-protein complex, in which This paper is aimed to assess whether levosimendan binds to levosimendan bindscTnC at the hydrophobic pocket of the NH,-terminal domain of cTnC. For this purpose, the interof the bindingof le- action between levosimendan and recombinant human cTnC the NHz-terminal domain. The role vosimendan to cTnC in the pharmacological action of was studied. Fluorescence spectroscopy was used to study the effect of levosimendan on the calcium-induced conformational this drugin vivo is discussed. change of the protein. In addition, fluorescence studies were also done ona site-directed mutantof cTnC to better delineate the binding siteof levosimendan in the NH,-terminal domain. Congestive heart failure occurs in a majority of patients with Moreover, to better characterizethe binding, a n NH,-terminal acute myocardial infarction. It is the most common cause of cTnC fragment was produced, and its interaction with levosideath in the acute phase of the disease, due to an impairment mendan studied by ‘H NMR spectroscopy. Finally, a binding of cardiac function(1).It has been proposed that diseased myo- site for levosimendan at the hydrophobic pocket of the regulacardiumisdesensitizedto Ca”, i.e. thecardiaccontractile tory N-domainof cTnC is proposed by applying molecular modstructures respondin an impaired manner to activator Ca2+ (2). elling techniques. This may be a consequence of either a reduction in intracellular MATERIALSANDMETHODS pH during cardiac hypoxia or ischemia, or an increase in the Cloning and Expression ofRecombinant Human Cardiac Doponin C CAMP-dependent phosphorylation of troponin I: both of which of human lead to a reduction in the Ca2+ affinity of troponin C (3, 4). and Its Amino-terminal Fragment-Thecodingsequence Therefore, increasing the Ca2+ sensitivity of the myocardial cTnC was clonedby using 2.5 pg of the human heart poly(A)tRNA (purchased from Clontech, Palo Alto,CA) as template for the cDNA contractile structures has been proposed as an ideal treatment synthesis with the synthetic oligonucleotide 5”ACGTGAATTCfor heart failure (2, 5, 6 ) . Drugs that are able to elicit suchan CTACTCCACACCCTTCATGA-3’acting as the 3‘ primer. In the followeffect have been designated “calcium sensitizers.” ing PCR reaction (20) the 5’ PCR primer was oligonucleotide5”ACGTLevosimendan (Scheme I) is a new drug that augments the GGATCCATGGATGACATCTACAAGGC-3‘and the 3’ PCR primer was calcium-induced tension in skinned fibers,and produces posi- the same oligonucleotide used for the cDNA synthesis. The amplified tive inotropic action in the paced papillary muscle at micromo- DNA fragment was digested with BamHI and EcoRI, purified, and subcloned to the pGEM3-vector (Promega) (20). The cloned cTnC DNA lar concentrations(7). Preliminary experimentsusing troponin fragment was sequenced as described previously (21-23). For protein affinity chromatography indicated a possible direct calciumexpression, the cTnC insert from pGEM3 was isolated and ligated to dependent binding of levosimendan to this protein(8). BamHI-EcoRI-digestedglutathione S-transferase fusion protein vector pGEX-ZT (Pharmacia Biotech Inc.) to yield the clone pOGL501. Troponin C is the calcium-binding protein in the thin filaIn Vitro Mutagenesis and Expression of Doponin C-The 5”primer ment of skeletal and cardiac muscle. The x-ray structure of skeletal troponinC is known(9, lo), andthe molecular function and subsequent cloning procedures werethe same as described above. Amino acid substitution Aspa8+ Alaa8 (pOGL511) wascarried out by applying the “megaprimer”technique as described (24,25).The mutant oligonucleotide was 5’-TGCATGAAGGACGCCAGCAGGGAAATCT* The costs of publication of this article were defrayedin part by the payment of page charges. This article must therefore be hereby marked The abbreviations used are: TnC, troponin C; cTnC,cardiac troponin “advertisement”in accordance with 18 U.S.C. Section 1734 solely to C; sTnC, skeletal troponin C; TnI, troponin I; PCR, polymerase chain indicate this fact. 5 To whom correspondence should be addressed: Orion Corp., Orion- reaction; PAGE, polyacrylamidegel electrophoresis; NOE, nuclear Farmos, Orion Research, Chemical Research Dept., P. 0. Box 65, FIN- Overhauser effect;NOESY, two-dimensional nuclear Overhauser en02101 Espoo, Finland. Tel.: 358-0-429-4191; Fax: 358-0-429-2924; hancement and exchange spectroscopy;COSY, correlated spectroscopy; dansyl, 5-dimethylaminonaphthalene-1-sulfonyl;RP, reversed phase. e-mail:
[email protected].
28584
Binding of Levoaimendan to Recombinant Human cTnC
SCIIEME
c
j A
H
a
1.5:
a b
;
;
I
3'. The cloning of the PCR fragment to yield the glutathione S-transferase fusion expressinn construct was performed a s described above. The accuracyof the mutantclones was verified hy DNA sequencing. The expression construct was introduced into Eschrrirhin coli BL21 tDE3) or D H d cells (Promega) which were gmwn and induced with isopropylP-wgalactopyranoside (0.5 mlc final concentration) a s described by j! B Lundstmm rt al. (26,. The expression construct coding for the aminoterminal fragment of cTnC (amino acids 1-91; pOGL522) was made by 0.2; PCR using as the 3' primer an oligonucleotide I:','-AGCTGAAT"CTCACCCTITGCTGTCGTCCT-3') that incorporates the translation stopcodon alter the sequence coding for the amino acid 91. 1 Purifirntion of Rrcomhinnnt Poponin C Pmtrins-After the induc1 I tion, the E. coli cells from a 5-liter flask culture medium were harvested " a s described hy Guan andDixon (27). Thecells were lysed by sonication 0 ! . ~ . . . . . . . by centrifugation a t 100 watts (at 4 ^C) and the lysate was cleared 0 l o ' ' " " " ' 2 o " ' " " " ' " " 30 "" '40 (10,000 x g. 20 min a t 4 "C). The recombinant protein was purifiedby lime(min) glutathione-Sepharose 4B-aflinity chromatography (Pharmacia)as described previously (27). Recombinant cTnC protein was cleaved from FIG.1. Characterization of the purified recombinant human the fusion protein with thrombin, and eluted from the column with the cTnC proteins by RP chromatography and SDS-PAGE. RP chromatography of thc anion cxchangc chromatography fraction containing cleavage huffer. Aliquots from the eluted fractions were analyzed by SDS-PAGE (15% acrylamide) (28) and Coomassic BrilliantBlue stain- ( A ) the full-length cTnC proteinor ( I ] ) the h'I-l,-terminal fragment ing. The fractions containing recornbinant cTnC protein were pooled, (amino acids 1-91) of the cTnC protein. The elution conditionswere as described under "Materials and Methods." The SDS-PAGE (A, 12.53; concentrated, and the buffer changed hy ultrafiltration (Omega Filter NMWL 3K. Filtmn). Further purification was performed on an anion and B, 1 5 3 acrylamide) of the respective RP peaks are shown in the insrrt. The gels were stained with Coomassic Brilliant Blue. The first exchange Mono-Q Chromatography column (HR 5 6 , Pharmacia). The top in A, 43.0. to hottom column was equilibrated with 20mh! Tris-HCI. pH 7.0, and the protein lanes show the molecular mass standards (from 30.0, 20.1, and 14.4 kDa; and in R, 30.0, 20.1. and 14.4 kDa). eluted with a linear gradientof sodium chloride (0-0.7 in 40 min. The peak containing recomhinant cTnC was concentrated and the buffcr measuring timeof 15 s and a hand passof 5 nm. Calcium chloride was exchanged to water by ultrafiltration. of added in incrementsof 1pmol (total calcium), and the concentration Chnrncterizntion of thr Purifird Pmtrins-The recombinant human free Ca2*was calculated according to Fahiato and Fabiato (33). When cTnC was analyzed by reversed phase chromatography (Spherisorh, indicated, either10 ps! trifluoperazine or3 p~ levosimendan was added 300A C,,, 0.3 x 15-cm column). The column was cquilihrated and the protein eluted a s described by Tilgmann and Kalkkinen (29). The ami- (from stock solution in Me,SO) to the protein before calcium titration. no-terminal fragment and the suhstitution mutantof the recombinant Control experiments were performed in the presence of 0.7% Me,SO, cTnC were analyzed on a RP column tTSK TMS 250. C,, 0.21 x 3 cm). and pH was readjusted to 7.00 0.02 with KOH after each Ca" addito Bradford The proteins wrre eluted witha gradient of 3-10Or+ acetonitrile in 60 tion. Protein concentrations were determined according two-tailed min. The collected peaks were further analyzedby SDS-PAGE and the (34). Statistical significances were calculated with Dunnet's gels were stained by Coomassie Brilliant Blue. For analysis of tryptic I test, by using the program ANOVA (repeated measurements). 'HN M R Experirnmts-Samples of the recornbinant amino-terminal peptides the proteins were alkvlated (30) and desalted by RP chromatography a s described (29). The collected proteins were dried by vacuum fragment of cTnC (typically 20 mg) were run on a Chelex column and lyophilized. The freeze-dried protein was reconstituted with 0.4 ml of a centrifugation, dissolved in 0.6 11 urea, 50 my Tris-HCI, pH 8.0, and digcstrd with trypsin( 3 9 , w/w) for 4 h a t 37 "C. The resulting peptides solution containing 10ms! Hepes, 100mM KCI, and 5 msc dithiothreitol, wrre analyzed by capillaryelectrophoresis (Applied Biosystems 270 a t pH 7.4, in 99.V; 'H20 (Aldrich), and againfreeze-dried. Finally, the HT, on a 50-pm x 500-mm (to detector) fused silica capillary column samples were redissolved in 0.4 mlof 99.9953 'H,O (Aldrich). The calcium content of the 'H,O batch (
