- Email: [email protected]
Use of high pressure to study elementary steps in P450 and nitric oxide synthase
Journal of Inorganic Biochemistry 87 (2001) 191–195 www.elsevier.com / locate / jinorgbio
Use of high pressure to study elementary steps in P450 and nitric oxide synthase a, a b c d Reinhard Lange *, Nicole Bec , Pavel Anzenbacher , Andrew W. Munro , Antonius C.F. Gorren , Bernd Mayer d a INSERM U128, IFR24, 1919 Route de Mende, 34293 Montpellier, France Institute of Pharmacology, Faculty of Medicine, Palacky University, 775 15 Olomouc, Czech Republic c University of Strathclyde, The Royal College, Glasgow, UK d ¨ Pharmakologie und Toxikologie, Karl-Franzens-Universitat ¨ , Universitatsplatz ¨ 2, 8010 Graz, Austria Institut f ur b
Received 9 January 2001; received in revised form 2 March 2001; accepted 9 April 2001 Dedicated to Professor Alexander I. Archakov on the occasion of his 60th birthday
Abstract Chemical reactions are often highly pressure-dependent. A perturbation of the elementary steps by pressure therefore offers the possibility of a detailed characterization of enzyme mechanisms. We used this method to study distinct steps in the reaction of nitric-oxide synthase (NOS), and compared them to analogous steps in the reaction of cytochrome P450 BM3 (BM3). Our results indicate that, in BM3, electron transfer depends on electrostatic interactions. In NOS, pressure, similarly to chemical denaturants, can mimick the structural effects of Ca / calmodulin. This helps to better understand the structural basis of the regulatory effect of Ca / calmodulin. Furthermore, stopped-flow kinetics under high pressure show that CO binding to the heme iron is hindered by substrate in NOS, but not in BM3. This indicates a relatively large or flexible substrate binding site in BM3, and a more narrow and rigid binding site in NOS. 2001 Elsevier Science B.V. All rights reserved. Keywords: Nitric-oxide synthase; Cytochrome P450; Calmodulin; High pressure; Electron transfer; CO binding
1. Introduction Nitric oxide synthase (NOS) and cytochrome P450 BM3 (BM3) are self-sufficient oxygenases which contain in a single polypeptide a P450-like oxygenase domain, with a thiolate ligated to the haem iron, and a flavin (FAD, FMN) containing reductase domain which has similarities to cytochrome P450 reductase [1–3]. In contrast to the monomeric BM3, NOS contains several additional cofactors: Ca / calmodulin (Cam), tetrahydrobiopterin (BH4), and one zinc ion per homodimer. Both enzymes catalyze the reductive activation of molecular oxygen, leading to substrate oxidation (fatty acid in the case of BM3, Larginine in the case of NOS), and to the formation of water. The reaction mechanism of both enzymes is complex and understood only partially [4,5]. Since both *Corresponding author. Tel.: 133-467-613-365; fax: 133-467-523681. E-mail address: [email protected] (R. Lange).
enzymes are self-sufficient, but do not have the same requirement for cofactors, a closer comparison of their reaction mechanisms is of interest. In this study we focussed on two reactions. Firstly, we investigated the elementary steps of the electron transfer within the reductase domain, with emphasis on the role of Cam as a regulator of the electron flow. Towards that goal, we measured the reduction rates of external electron acceptors cytochrome c and dichlorophenol-indophenol (DCPIP) after mixing the enzyme with NADPH. Secondly, we studied the interaction of the active site with substrate: for that, we measured the CO binding kinetics in the presence and absence of substrate. Our approach to studying these reactions was by high pressure stopped-flow kinetics, with pressures ranging from 0.1 to 200 MPa, i.e. up to 2 kbar. As a matter of fact, high pressure is an elegant way to perturb reversibly chemical equilibria and reactions [6–8]. Another advantage of using the pressure parameter is that reactions are slowed or accelerated depending on the type of chemical
0162-0134 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0162-0134( 01 )00330-0
192
R. Lange et al. / Journal of Inorganic Biochemistry 87 (2001) 191 – 195
interaction involved. For instance, pressure weakens electrostatic interactions, but stimulates some hydrophobic interactions, such as stacking between aromatic residues [9,10]. Similarly to the activation enthalpy, DH ‡ , obtained from kinetics as a function of temperature, experiments under pressure yield an activation volume, DV ‡ . When the activation volume is positive, the reaction is slowed, and when it is negative, the velocity is increased by pressure. This effect was used to obtain information about the structural basis underlying the elementary reaction steps of BM3 and NOS.
2. Experimental
2.1. Materials Recombinant rat brain nNOS was purified from baculovirus-infected insect cells as described previously [11] in its full-length form. It contained tightly bound BH 4 in the ratio of 0.5 per heme. CYP102 (cytochrome P450 BM-3) holoenzyme was prepared by expressing the plasmid construct with an appropriate gene in E. coli. Procedures for gene isolation, plasmid construction and expression have been described previously [12]. BH 4 and NOHLA were from Alexis Biochemicals (Switzerland). Sodium dithionite and glycerol were obtained from Fluka. All other chemicals were of the highest quality available commercially.
2.2. Procedures
atmospheric pressure. CO solutions were then prepared by diluting this stock solution with oxygen-free buffer. Maximum / minimum wavelengths of absorption of the CO complex were 444 / 406 nm and 449 / 412 nm for nNOS and BM-3, respectively. Final concentrations of enzyme and CO after mixing in the stopped-flow apparatus were 1 and 100 mM for nNOS, and 0.3 and 5 mM for BM-3. The temperature was maintained at 88C.
2.2.2. Cytochrome c or DCPIP reduction under hydrostatic pressure The reduction of cytochrome c and DCPIP was monitored at 550 nm minus 565 nm (D´5502565 5 21 500 M 21 cm 21 ) and 600 nm minus 750 nm (D´6002750 5 12 610 M 21 cm 21 ), respectively at 258C. The nNOS reductase assays were performed in a buffer solution (see above) containing 1 mM CHAPS. We used 7 nM enzyme, 25 mM cytochrome c or 40 mM DCPIP, 20 mg / ml calmodulin, 0.5 mM CaCl 2 , and 200 mM NADPH. In the absence of calmodulin and CaCl 2 , the enzyme concentration was 70 nM. In experiments performed at atmospheric pressure in the presence of urea, the nNOS concentration was 1.75 nM. The reduction of cytochrome c and DCPIP was 21 cm 21 for cytoquantified using D´550 5 17 520 M 21 21 chrome c and D´600 5 13 000 M cm for DCPIP. The BM3 reductase assays were performed using 7 nM enzyme, 25 mM cytochrome c or 40 mM DCPIP, and 200 mM NADPH. After mixing the components, the kinetic parameters of reductase activity for both enzymes were determined by measuring the initial velocities, vi , expressed in mmol substrate / min / nmol enzyme.
Kinetic measurements at atmospheric pressure were carried out using a thermostated Uvikon 941 spec¨ trophotometer (Kontron, Zurich, Switzerland). Experiments under high pressure were performed in a highpressure stopped-flow apparatus built in our laboratory [13]. Control spectroscopic (UV and visible absorbance) measurements showed that pressure (within the range used) did not have any detectable structural effect on either BM3 or NOS. Kinetics were started by mixing equal volumes (60 ml) of the enzyme and substrate solutions in a thermostated high-pressure cell placed in an Aminco DW2 spectrophotometer operating in dual-wavelength mode. The buffers were: 50 mM TRIS pH 7.5 (NOS), and 20 mM MOPS pH 7.4, 100 mM KCl, 20% glycerol (v / v) (BM3). All experiments were performed in triplicate.
2.2.3. Analysis of high-pressure kinetic data From the pressure-dependent kinetic constants (k obs or vi ) the activation volumes (DV ‡ ) were determined according to
2.2.1. CO binding under hydrostatic pressure The enzyme solutions were deoxygenated by blowing argon over their surface and then reduced by sodium dithionite (1.2 mM final concentration) under argon for 30 min. A CO stock solution was prepared by bubbling oxygen-free argon through the buffer solution for 30 min and then flushing with CO for 45 min. The concentration of this stock solution was taken to be 1 mM CO at 208C at
3.1. Electron transfer within the reductase domain
‡
(d ln k obs /d P) T 5 2 DV /RT
(1)
(d ln vi /d P) T 5 2 DV ‡ /RT
(2)
where P is the hydrostatic pressure, T the absolute temperature and R the gas constant (8.2 ml MPa mol 21 K 21 ).
3. Results
NOS and BM3, in the absence of substrate, were reduced by NADPH, and then mixed in the stopped-flow apparatus with cytochrome c or DCPIP, the reduction of which was followed as described in the Experimental section. Linear kinetics were obtained, allowing the determination of vi . Very similar results were obtained
R. Lange et al. / Journal of Inorganic Biochemistry 87 (2001) 191 – 195
193
Table 1 Activation volumes of BM3 and NOS catalyzed cytochrome c and DCPIP reduction DV ‡ (ml / mol) (pressure range)
Cytochrome c 1Ca / calmodulin 2Ca / calmodulin DCPIP 1Ca / calmodulin 2Ca / calmodulin
Fig. 1. Effect of pressure on the rate of cytochrome c reduction by BM3. The rates vi are expressed in mmol cytochrome c / min / nmol BM3.
whether the electron acceptor was cytochrome c or DCPIP. For NOS, as well as for BM3, vi was proportional to the enzyme concentration, but did not depend on the concentration of NADPH or of the electron acceptor. The measured rate thus reflected the steady state electron flow within the reductase domain. The effect of pressure is evident from Figs. 1 and 2. In the case of BM3 (Fig. 1), vi decreased strongly as a function of pressure. The activation volume (Table 1) was rather high and positive: DV ‡ 5 1 2162 ml / mol, suggesting the involvement of electrostatic interactions for this electron transfer reaction. The case of NOS is more complicated. In the presence of Ca / calmodulin, vi increased up to 50 MPa, but decreased upon a further rise in pressure. In the absence of Ca / calmodulin, the electron transfer rate was about 20 times smaller than in its presence, as expected from the regulatory role of cal-
Fig. 2. Effect of pressure on the rate of cytochrome c reduction by NOS, in the presence (d) and absence (s) of Ca / calmodulin. The rates vi are expressed in mmol cytochrome c / min / nmol NOS.
nNOS
BM3
21867 (0 to 50 MPa) 14862 (60 to 160 MPa) 27065 (0 to 80 MPa) 21165 (100 to 160 MPa)
– – 12162 (0 to 160 MPa)
21364 (0 to 50 MPa) 13161 (60 to 160 MPa) 25462 (0 to 80 MPa) 21464 (100 to 160 MPa)
– – 12462 (0 to 160 MPa)
modulin. However, under high pressure, the rate increased strongly (DV ‡ 5 2 7065 ml / mol), and at 150 MPa, a similar value of vi was obtained as that measured in the presence of Ca / calmodulin. The 20-fold increase in the reduction rate was fully reversible after pressure release. These results demonstrate that, for this elementary step, high pressure can substitute for Ca / calmodulin. Interestingly, a similar accelerating effect has been reported for urea [14]. We could confirm this effect by an experiment carried out under our conditions (Fig. 3).
3.2. The interaction of substrate with the substrate binding site of the oxygenase domains The CO binding kinetics of BM3, as well as of NOS, were monoexponential, thus allowing the determination of the binding rate constant k obs . As shown in Fig. 4, pressure did not affect the rate constant of BM3, indicating an activation volume DV ‡ close to zero. In that respect, BM3 behaved like a ‘classical’ P450: similar results have been
Fig. 3. Effect of urea on the rate of cytochrome c reduction by NOS in the absence of Ca / calmodulin. The initial velocity vi is expressed in mmol cytochrome c / min / nmol NOS.
194
R. Lange et al. / Journal of Inorganic Biochemistry 87 (2001) 191 – 195
high pressure can be explained by the substrate being expelled from the binding site.
4. Discussion
Fig. 4. Effect of pressure on the rate of CO binding by BM3, in the presence (d) and absence (s) of 100 mM palmitate. The rate constant k obs is expressed in s 21 .
reported for other P450s [15]. Furthermore, the kinetics were not substrate-dependent. This indicates that the active site of BM3 is relatively large (or flexible), allowing the binding of CO to the heme iron in the presence of substrate. This confirms our earlier results on the compressibility of the BM3 active site [16]. The situation with NOS is different. As shown in Fig. 5, pressure did not affect the CO binding kinetics in the absence of substrate. In the presence of substrate, the rate was much smaller (eight-fold) than in its absence, in accordance with previous reports [17,18]. However, the application of high pressure induced a strong increase of the rate constant. At 200 MPa (the limit of our instrument), the rate was nearly that in the absence of substrate. This suggests that CO binding is strongly hindered by the presence of substrate. The removal of this hindrance at
Fig. 5. Effect of pressure on the rate of CO binding by NOS, in the presence of 200 mM L-arginine (d), 200 mM L-arginine and an excess of 25 mM BH4 (j), 500 mM NOHLA (m) and in the absence of any substrates (s). The rate constant k obs is expressed in s 21 .
Our results show that pressure is a very useful parameter to study the elementary steps of complex enzyme reactions. The application of this approach to BM3 and NOS revealed quite different mechanisms of two elementary reactions, shedding some light on structure–function relationships. Let us first consider the electron flow within the reductase domain. In the case of BM3, our results are consistent with electrostatic interactions being important for electron transfer within the reductase domain. Indeed, the strongly negative activation volume suggests the involvement of several charged groups, perhaps also the participation of water molecules. For NOS, on the contrary, the pressure effect is more complicated and, in the presence of Ca / calmodulin, suggests two competing mechanisms: one prevailing at low pressure, and another above 50 MPa. The fact that pressure can substitute for Ca / calmodulin is also intriguing. Generally, pressure acts on proteins as a denaturant, i.e. it favors the unfolding of proteins. In our case, a strong and global protein unfolding effect can be excluded, as this would have resulted in a loss of flavins [14]. The effect of pressure on NOS structure can therefore be expected to be small, and restricted locally. As pressure and urea have a similar effect in substituting for Ca / calmodulin, it therefore appears that one effect of Ca / calmodulin can be understood as a weakening or partial unfolding of the reductase domain in the region of the ‘autoinhibitory loop’. Deletion of this peptide sequence has been shown to induce a similar effect [14]. Let us now turn to the CO binding experiment. Again, BM3 and NOS behave very differently, suggesting different structural features of their substrate binding sites. BM3 behaves like a classical P450 [15], and bimolecular CO binding is not influenced by the substrate. The pressure effect on NOS, however, reveals that, if substrate is bound, CO can bind only with difficulty, as suggested also by the experiments of Abu-Soud et al. [17]. This indicates a very tight substrate binding, leaving no place for the relatively small CO molecule. CO binding is often taken as a model for O 2 binding. Interestingly, for NOS the mechanism of CO and oxygen binding appears to be different: the presence of substrate affects very much the rate of CO binding, but only little O 2 binding [19]. These results show the potential interest in high pressure to study the elementary reaction steps of BM3 and NOS. This approach will now be applied to the study of further steps of their reaction cycles. Furthermore, studies of the effects of pressure on the monomer / dimer ratio and NO synthesis of NOS and the unfolding ratio of BM3 will be the subject of future efforts.
R. Lange et al. / Journal of Inorganic Biochemistry 87 (2001) 191 – 195
5. Abbreviations nNOS BH 4 NOHLA DCPIP NADPH CHAPS TRIS MOPS
neuronal nitric-oxide synthase (6R)-5,6,7,8-tetrahydro-L-biopterin N v -hydroxy-L-arginine 2,6-dichlorophenolindophenol reduced b-nicotinamide adenine dinucleotide phosphate 3-[(3-cholamidopropyl)dimethylammonio]-1propanesulfonic acid tris(hydroxymethyl)aminomethane 4-morpholinepropanesulfonic acid
Acknowledgements This work was accomplished in the frame of the French–Austrian AMADEUS program. Part of the work was supported by grants 13013-MED and 13586-MED of ¨ the Fonds zur Forderung der Wissenschaftlichen Forschung ¨ in Osterreich. Financial support from the Czech Grant Agency (203 / 99 / 0277) and the Ministry of Education (MSM 151100003) is gratefully acknowledged.
References [1] S. Pfeiffer, B. Mayer, B. Hemmens, Angew. Chem. Int. Ed. 38 (1999) 1714–1731.
195
[2] C.S. Raman, H. Li, P. Martasek, V. Kral, B.S.S. Masters, T.L. Poulos, Cell 95 (1998) 939–950. [3] K.G. Ravichandran, S.S. Boddupalli, C.A. Hasemann, J.A. Peterson, J. Deisenhofer, Science 261 (1993) 731–736. [4] B.S.S. Masters, Annu. Rev. Nutr. 14 (1994) 131–145. [5] M.A. Noble, L. Quaroni, G.D. Chumanov, K.L. Turner, S.K. Chapman, R.P. Hanzlik, A.W. Munro, Biochemistry 37 (1998) 15799–15807. [6] J.L. Silva, G. Weber, Annu. Rev. Phys. Chem. 44 (1993) 89–113. [7] V.V. Mozhaev, K. Heremans, J. Frank, P. Masson, C. Balny, Proteins Struct. Funct. Genet. 24 (1996) 81–91. [8] M. Gross, R. Jaenicke, Eur. J. Biochem. 221 (1994) 617–630. [9] E. Mombelli, M. Afshar, P. Fusi, M. Mariani, P. Tortora, J.P. Connelly, R. Lange, Biochemistry 36 (1997) 8733–8742. [10] J. Torrent, J.P. Connelly, M.G. Coll, M. Ribo, R. Lange, M. Vilanova, Biochemistry 38 (1999) 15952–15961. [11] C. Harteneck, P. Klatt, K. Schmidt, B. Mayer, Biochem. J. 304 (1994) 683–686. [12] J.S. Miles, A.W. Munro, B.N. Rospendowski, W.E. Smith, J. McKnight, A.J. Thomson, Biochem. J. 288 (1992) 503–508. [13] C. Balny, J.L. Saldana, N. Dahan, Anal. Biochem. 139 (1984) 178–189. [14] R. Narayanasami, J.S. Nishimura, K. McMillan, L.J. Roman, T.M. Shea, A.M. Robida, P.M. Horowitz, B.S. Masters, Nitric Oxide 1 (1997) 39–49. [15] R. Lange, I. Heiber-Langer, C. Bonfils, I. Fabre, M. Negishi, C. Balny, Biophys. J. 66 (1994) 89–98. ´ N. Bec, P. Anzenbacher, J. Hudecek, P. Soucek, [16] E. Anzenbacherova, C. Jung, A.W. Munro, R. Lange, Eur. J. Biochem. 267 (2000) 2916–2920. [17] H.M. Abu-Soud, C. Wu, D.K. Ghosh, D.J. Stuehr, Biochemistry 37 (1998) 3777–3786. [18] C. Tetreau, M. Tourbez, A. Gorren, B. Mayer, D. Lavalette, Biochemistry 38 (1999) 7210–7218. [19] H.M. Abu-Soud, R. Gachhui, F.M. Raushel, D.J. Stuehr, J. Biol. Chem. 272 (1997) 17349–17353.
Use of high pressure to study elementary steps in P450 and nitric oxide synthase a, a b c d Reinhard Lange *, Nicole Bec , Pavel Anzenbacher , Andrew W. Munro , Antonius C.F. Gorren , Bernd Mayer d a INSERM U128, IFR24, 1919 Route de Mende, 34293 Montpellier, France Institute of Pharmacology, Faculty of Medicine, Palacky University, 775 15 Olomouc, Czech Republic c University of Strathclyde, The Royal College, Glasgow, UK d ¨ Pharmakologie und Toxikologie, Karl-Franzens-Universitat ¨ , Universitatsplatz ¨ 2, 8010 Graz, Austria Institut f ur b
Received 9 January 2001; received in revised form 2 March 2001; accepted 9 April 2001 Dedicated to Professor Alexander I. Archakov on the occasion of his 60th birthday
Abstract Chemical reactions are often highly pressure-dependent. A perturbation of the elementary steps by pressure therefore offers the possibility of a detailed characterization of enzyme mechanisms. We used this method to study distinct steps in the reaction of nitric-oxide synthase (NOS), and compared them to analogous steps in the reaction of cytochrome P450 BM3 (BM3). Our results indicate that, in BM3, electron transfer depends on electrostatic interactions. In NOS, pressure, similarly to chemical denaturants, can mimick the structural effects of Ca / calmodulin. This helps to better understand the structural basis of the regulatory effect of Ca / calmodulin. Furthermore, stopped-flow kinetics under high pressure show that CO binding to the heme iron is hindered by substrate in NOS, but not in BM3. This indicates a relatively large or flexible substrate binding site in BM3, and a more narrow and rigid binding site in NOS. 2001 Elsevier Science B.V. All rights reserved. Keywords: Nitric-oxide synthase; Cytochrome P450; Calmodulin; High pressure; Electron transfer; CO binding
1. Introduction Nitric oxide synthase (NOS) and cytochrome P450 BM3 (BM3) are self-sufficient oxygenases which contain in a single polypeptide a P450-like oxygenase domain, with a thiolate ligated to the haem iron, and a flavin (FAD, FMN) containing reductase domain which has similarities to cytochrome P450 reductase [1–3]. In contrast to the monomeric BM3, NOS contains several additional cofactors: Ca / calmodulin (Cam), tetrahydrobiopterin (BH4), and one zinc ion per homodimer. Both enzymes catalyze the reductive activation of molecular oxygen, leading to substrate oxidation (fatty acid in the case of BM3, Larginine in the case of NOS), and to the formation of water. The reaction mechanism of both enzymes is complex and understood only partially [4,5]. Since both *Corresponding author. Tel.: 133-467-613-365; fax: 133-467-523681. E-mail address: [email protected] (R. Lange).
enzymes are self-sufficient, but do not have the same requirement for cofactors, a closer comparison of their reaction mechanisms is of interest. In this study we focussed on two reactions. Firstly, we investigated the elementary steps of the electron transfer within the reductase domain, with emphasis on the role of Cam as a regulator of the electron flow. Towards that goal, we measured the reduction rates of external electron acceptors cytochrome c and dichlorophenol-indophenol (DCPIP) after mixing the enzyme with NADPH. Secondly, we studied the interaction of the active site with substrate: for that, we measured the CO binding kinetics in the presence and absence of substrate. Our approach to studying these reactions was by high pressure stopped-flow kinetics, with pressures ranging from 0.1 to 200 MPa, i.e. up to 2 kbar. As a matter of fact, high pressure is an elegant way to perturb reversibly chemical equilibria and reactions [6–8]. Another advantage of using the pressure parameter is that reactions are slowed or accelerated depending on the type of chemical
0162-0134 / 01 / $ – see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S0162-0134( 01 )00330-0
192
R. Lange et al. / Journal of Inorganic Biochemistry 87 (2001) 191 – 195
interaction involved. For instance, pressure weakens electrostatic interactions, but stimulates some hydrophobic interactions, such as stacking between aromatic residues [9,10]. Similarly to the activation enthalpy, DH ‡ , obtained from kinetics as a function of temperature, experiments under pressure yield an activation volume, DV ‡ . When the activation volume is positive, the reaction is slowed, and when it is negative, the velocity is increased by pressure. This effect was used to obtain information about the structural basis underlying the elementary reaction steps of BM3 and NOS.
2. Experimental
2.1. Materials Recombinant rat brain nNOS was purified from baculovirus-infected insect cells as described previously [11] in its full-length form. It contained tightly bound BH 4 in the ratio of 0.5 per heme. CYP102 (cytochrome P450 BM-3) holoenzyme was prepared by expressing the plasmid construct with an appropriate gene in E. coli. Procedures for gene isolation, plasmid construction and expression have been described previously [12]. BH 4 and NOHLA were from Alexis Biochemicals (Switzerland). Sodium dithionite and glycerol were obtained from Fluka. All other chemicals were of the highest quality available commercially.
2.2. Procedures
atmospheric pressure. CO solutions were then prepared by diluting this stock solution with oxygen-free buffer. Maximum / minimum wavelengths of absorption of the CO complex were 444 / 406 nm and 449 / 412 nm for nNOS and BM-3, respectively. Final concentrations of enzyme and CO after mixing in the stopped-flow apparatus were 1 and 100 mM for nNOS, and 0.3 and 5 mM for BM-3. The temperature was maintained at 88C.
2.2.2. Cytochrome c or DCPIP reduction under hydrostatic pressure The reduction of cytochrome c and DCPIP was monitored at 550 nm minus 565 nm (D´5502565 5 21 500 M 21 cm 21 ) and 600 nm minus 750 nm (D´6002750 5 12 610 M 21 cm 21 ), respectively at 258C. The nNOS reductase assays were performed in a buffer solution (see above) containing 1 mM CHAPS. We used 7 nM enzyme, 25 mM cytochrome c or 40 mM DCPIP, 20 mg / ml calmodulin, 0.5 mM CaCl 2 , and 200 mM NADPH. In the absence of calmodulin and CaCl 2 , the enzyme concentration was 70 nM. In experiments performed at atmospheric pressure in the presence of urea, the nNOS concentration was 1.75 nM. The reduction of cytochrome c and DCPIP was 21 cm 21 for cytoquantified using D´550 5 17 520 M 21 21 chrome c and D´600 5 13 000 M cm for DCPIP. The BM3 reductase assays were performed using 7 nM enzyme, 25 mM cytochrome c or 40 mM DCPIP, and 200 mM NADPH. After mixing the components, the kinetic parameters of reductase activity for both enzymes were determined by measuring the initial velocities, vi , expressed in mmol substrate / min / nmol enzyme.
Kinetic measurements at atmospheric pressure were carried out using a thermostated Uvikon 941 spec¨ trophotometer (Kontron, Zurich, Switzerland). Experiments under high pressure were performed in a highpressure stopped-flow apparatus built in our laboratory [13]. Control spectroscopic (UV and visible absorbance) measurements showed that pressure (within the range used) did not have any detectable structural effect on either BM3 or NOS. Kinetics were started by mixing equal volumes (60 ml) of the enzyme and substrate solutions in a thermostated high-pressure cell placed in an Aminco DW2 spectrophotometer operating in dual-wavelength mode. The buffers were: 50 mM TRIS pH 7.5 (NOS), and 20 mM MOPS pH 7.4, 100 mM KCl, 20% glycerol (v / v) (BM3). All experiments were performed in triplicate.
2.2.3. Analysis of high-pressure kinetic data From the pressure-dependent kinetic constants (k obs or vi ) the activation volumes (DV ‡ ) were determined according to
2.2.1. CO binding under hydrostatic pressure The enzyme solutions were deoxygenated by blowing argon over their surface and then reduced by sodium dithionite (1.2 mM final concentration) under argon for 30 min. A CO stock solution was prepared by bubbling oxygen-free argon through the buffer solution for 30 min and then flushing with CO for 45 min. The concentration of this stock solution was taken to be 1 mM CO at 208C at
3.1. Electron transfer within the reductase domain
‡
(d ln k obs /d P) T 5 2 DV /RT
(1)
(d ln vi /d P) T 5 2 DV ‡ /RT
(2)
where P is the hydrostatic pressure, T the absolute temperature and R the gas constant (8.2 ml MPa mol 21 K 21 ).
3. Results
NOS and BM3, in the absence of substrate, were reduced by NADPH, and then mixed in the stopped-flow apparatus with cytochrome c or DCPIP, the reduction of which was followed as described in the Experimental section. Linear kinetics were obtained, allowing the determination of vi . Very similar results were obtained
R. Lange et al. / Journal of Inorganic Biochemistry 87 (2001) 191 – 195
193
Table 1 Activation volumes of BM3 and NOS catalyzed cytochrome c and DCPIP reduction DV ‡ (ml / mol) (pressure range)
Cytochrome c 1Ca / calmodulin 2Ca / calmodulin DCPIP 1Ca / calmodulin 2Ca / calmodulin
Fig. 1. Effect of pressure on the rate of cytochrome c reduction by BM3. The rates vi are expressed in mmol cytochrome c / min / nmol BM3.
whether the electron acceptor was cytochrome c or DCPIP. For NOS, as well as for BM3, vi was proportional to the enzyme concentration, but did not depend on the concentration of NADPH or of the electron acceptor. The measured rate thus reflected the steady state electron flow within the reductase domain. The effect of pressure is evident from Figs. 1 and 2. In the case of BM3 (Fig. 1), vi decreased strongly as a function of pressure. The activation volume (Table 1) was rather high and positive: DV ‡ 5 1 2162 ml / mol, suggesting the involvement of electrostatic interactions for this electron transfer reaction. The case of NOS is more complicated. In the presence of Ca / calmodulin, vi increased up to 50 MPa, but decreased upon a further rise in pressure. In the absence of Ca / calmodulin, the electron transfer rate was about 20 times smaller than in its presence, as expected from the regulatory role of cal-
Fig. 2. Effect of pressure on the rate of cytochrome c reduction by NOS, in the presence (d) and absence (s) of Ca / calmodulin. The rates vi are expressed in mmol cytochrome c / min / nmol NOS.
nNOS
BM3
21867 (0 to 50 MPa) 14862 (60 to 160 MPa) 27065 (0 to 80 MPa) 21165 (100 to 160 MPa)
– – 12162 (0 to 160 MPa)
21364 (0 to 50 MPa) 13161 (60 to 160 MPa) 25462 (0 to 80 MPa) 21464 (100 to 160 MPa)
– – 12462 (0 to 160 MPa)
modulin. However, under high pressure, the rate increased strongly (DV ‡ 5 2 7065 ml / mol), and at 150 MPa, a similar value of vi was obtained as that measured in the presence of Ca / calmodulin. The 20-fold increase in the reduction rate was fully reversible after pressure release. These results demonstrate that, for this elementary step, high pressure can substitute for Ca / calmodulin. Interestingly, a similar accelerating effect has been reported for urea [14]. We could confirm this effect by an experiment carried out under our conditions (Fig. 3).
3.2. The interaction of substrate with the substrate binding site of the oxygenase domains The CO binding kinetics of BM3, as well as of NOS, were monoexponential, thus allowing the determination of the binding rate constant k obs . As shown in Fig. 4, pressure did not affect the rate constant of BM3, indicating an activation volume DV ‡ close to zero. In that respect, BM3 behaved like a ‘classical’ P450: similar results have been
Fig. 3. Effect of urea on the rate of cytochrome c reduction by NOS in the absence of Ca / calmodulin. The initial velocity vi is expressed in mmol cytochrome c / min / nmol NOS.
194
R. Lange et al. / Journal of Inorganic Biochemistry 87 (2001) 191 – 195
high pressure can be explained by the substrate being expelled from the binding site.
4. Discussion
Fig. 4. Effect of pressure on the rate of CO binding by BM3, in the presence (d) and absence (s) of 100 mM palmitate. The rate constant k obs is expressed in s 21 .
reported for other P450s [15]. Furthermore, the kinetics were not substrate-dependent. This indicates that the active site of BM3 is relatively large (or flexible), allowing the binding of CO to the heme iron in the presence of substrate. This confirms our earlier results on the compressibility of the BM3 active site [16]. The situation with NOS is different. As shown in Fig. 5, pressure did not affect the CO binding kinetics in the absence of substrate. In the presence of substrate, the rate was much smaller (eight-fold) than in its absence, in accordance with previous reports [17,18]. However, the application of high pressure induced a strong increase of the rate constant. At 200 MPa (the limit of our instrument), the rate was nearly that in the absence of substrate. This suggests that CO binding is strongly hindered by the presence of substrate. The removal of this hindrance at
Fig. 5. Effect of pressure on the rate of CO binding by NOS, in the presence of 200 mM L-arginine (d), 200 mM L-arginine and an excess of 25 mM BH4 (j), 500 mM NOHLA (m) and in the absence of any substrates (s). The rate constant k obs is expressed in s 21 .
Our results show that pressure is a very useful parameter to study the elementary steps of complex enzyme reactions. The application of this approach to BM3 and NOS revealed quite different mechanisms of two elementary reactions, shedding some light on structure–function relationships. Let us first consider the electron flow within the reductase domain. In the case of BM3, our results are consistent with electrostatic interactions being important for electron transfer within the reductase domain. Indeed, the strongly negative activation volume suggests the involvement of several charged groups, perhaps also the participation of water molecules. For NOS, on the contrary, the pressure effect is more complicated and, in the presence of Ca / calmodulin, suggests two competing mechanisms: one prevailing at low pressure, and another above 50 MPa. The fact that pressure can substitute for Ca / calmodulin is also intriguing. Generally, pressure acts on proteins as a denaturant, i.e. it favors the unfolding of proteins. In our case, a strong and global protein unfolding effect can be excluded, as this would have resulted in a loss of flavins [14]. The effect of pressure on NOS structure can therefore be expected to be small, and restricted locally. As pressure and urea have a similar effect in substituting for Ca / calmodulin, it therefore appears that one effect of Ca / calmodulin can be understood as a weakening or partial unfolding of the reductase domain in the region of the ‘autoinhibitory loop’. Deletion of this peptide sequence has been shown to induce a similar effect [14]. Let us now turn to the CO binding experiment. Again, BM3 and NOS behave very differently, suggesting different structural features of their substrate binding sites. BM3 behaves like a classical P450 [15], and bimolecular CO binding is not influenced by the substrate. The pressure effect on NOS, however, reveals that, if substrate is bound, CO can bind only with difficulty, as suggested also by the experiments of Abu-Soud et al. [17]. This indicates a very tight substrate binding, leaving no place for the relatively small CO molecule. CO binding is often taken as a model for O 2 binding. Interestingly, for NOS the mechanism of CO and oxygen binding appears to be different: the presence of substrate affects very much the rate of CO binding, but only little O 2 binding [19]. These results show the potential interest in high pressure to study the elementary reaction steps of BM3 and NOS. This approach will now be applied to the study of further steps of their reaction cycles. Furthermore, studies of the effects of pressure on the monomer / dimer ratio and NO synthesis of NOS and the unfolding ratio of BM3 will be the subject of future efforts.
R. Lange et al. / Journal of Inorganic Biochemistry 87 (2001) 191 – 195
5. Abbreviations nNOS BH 4 NOHLA DCPIP NADPH CHAPS TRIS MOPS
neuronal nitric-oxide synthase (6R)-5,6,7,8-tetrahydro-L-biopterin N v -hydroxy-L-arginine 2,6-dichlorophenolindophenol reduced b-nicotinamide adenine dinucleotide phosphate 3-[(3-cholamidopropyl)dimethylammonio]-1propanesulfonic acid tris(hydroxymethyl)aminomethane 4-morpholinepropanesulfonic acid
Acknowledgements This work was accomplished in the frame of the French–Austrian AMADEUS program. Part of the work was supported by grants 13013-MED and 13586-MED of ¨ the Fonds zur Forderung der Wissenschaftlichen Forschung ¨ in Osterreich. Financial support from the Czech Grant Agency (203 / 99 / 0277) and the Ministry of Education (MSM 151100003) is gratefully acknowledged.
References [1] S. Pfeiffer, B. Mayer, B. Hemmens, Angew. Chem. Int. Ed. 38 (1999) 1714–1731.
195
[2] C.S. Raman, H. Li, P. Martasek, V. Kral, B.S.S. Masters, T.L. Poulos, Cell 95 (1998) 939–950. [3] K.G. Ravichandran, S.S. Boddupalli, C.A. Hasemann, J.A. Peterson, J. Deisenhofer, Science 261 (1993) 731–736. [4] B.S.S. Masters, Annu. Rev. Nutr. 14 (1994) 131–145. [5] M.A. Noble, L. Quaroni, G.D. Chumanov, K.L. Turner, S.K. Chapman, R.P. Hanzlik, A.W. Munro, Biochemistry 37 (1998) 15799–15807. [6] J.L. Silva, G. Weber, Annu. Rev. Phys. Chem. 44 (1993) 89–113. [7] V.V. Mozhaev, K. Heremans, J. Frank, P. Masson, C. Balny, Proteins Struct. Funct. Genet. 24 (1996) 81–91. [8] M. Gross, R. Jaenicke, Eur. J. Biochem. 221 (1994) 617–630. [9] E. Mombelli, M. Afshar, P. Fusi, M. Mariani, P. Tortora, J.P. Connelly, R. Lange, Biochemistry 36 (1997) 8733–8742. [10] J. Torrent, J.P. Connelly, M.G. Coll, M. Ribo, R. Lange, M. Vilanova, Biochemistry 38 (1999) 15952–15961. [11] C. Harteneck, P. Klatt, K. Schmidt, B. Mayer, Biochem. J. 304 (1994) 683–686. [12] J.S. Miles, A.W. Munro, B.N. Rospendowski, W.E. Smith, J. McKnight, A.J. Thomson, Biochem. J. 288 (1992) 503–508. [13] C. Balny, J.L. Saldana, N. Dahan, Anal. Biochem. 139 (1984) 178–189. [14] R. Narayanasami, J.S. Nishimura, K. McMillan, L.J. Roman, T.M. Shea, A.M. Robida, P.M. Horowitz, B.S. Masters, Nitric Oxide 1 (1997) 39–49. [15] R. Lange, I. Heiber-Langer, C. Bonfils, I. Fabre, M. Negishi, C. Balny, Biophys. J. 66 (1994) 89–98. ´ N. Bec, P. Anzenbacher, J. Hudecek, P. Soucek, [16] E. Anzenbacherova, C. Jung, A.W. Munro, R. Lange, Eur. J. Biochem. 267 (2000) 2916–2920. [17] H.M. Abu-Soud, C. Wu, D.K. Ghosh, D.J. Stuehr, Biochemistry 37 (1998) 3777–3786. [18] C. Tetreau, M. Tourbez, A. Gorren, B. Mayer, D. Lavalette, Biochemistry 38 (1999) 7210–7218. [19] H.M. Abu-Soud, R. Gachhui, F.M. Raushel, D.J. Stuehr, J. Biol. Chem. 272 (1997) 17349–17353.