Substrates modulate the rate-determining step for CO binding in cytochrome P450cam (CYP101) A high-pressure stopped-flow study Christiane Jung 1 , Nicole Bec 2 and Reinhard Lange 2 1 Max-Delbru ¨ ck-Center for Molecular Medicine, Protein Dynamics Laboratory, Berlin, Germany; 2 Institut National de la Sante ´ et de la Recherche Me ´ dicale, Unite ´ 128, IFR24, Montpellier, France The high-pressure stopped-flow technique is applied to study the CO binding in cytochrome P450cam (P450cam) bound with homologous substrates (1R-camphor, camphane, nor- camphor and norbornane) and in the substrate-free protein. The activation volume DV # of the CO on-rate is positive for P450cam bound with substrates that do not contain methyl groups. The k on rate constant for these substrate complexes is in the order of 3 · 10 6 M )1 Æs )1 . In contrast, P450cam complexed with substrates carrying methyl groups show a negative activation volume and a low k on rate constant of % 3 · 10 4 M )1 Æs )1 .Byrelatingk on and DV # with values for the compressibility and the influx rate of water for the heme pocket of the substrate complexes it is concluded that the positive activation volume is indicative for a loosely bound substrate that guarantees a high solvent accessibility for the heme pocket and a very compressible active site. In addition, subconformers have been found for the substrate-free and camphane-bound protein which show different CO binding kinetics. Keywords: high-pressure stopped-flow; cytochrome P450; CO ligand binding; protein dynamics. Cytochromes P450 represent a big superfamily of heme-type monooxygenases that catalyze the conversion of diverse substrates [1]. Besides the main route of the reaction cycle from the substrate to the product there are side reactions which lead to the production of cytotoxic oxygen species such as hydrogen peroxide or of water in the oxidase reaction. These so-called uncoupling processes have been observed in many cytochrome P450 systems [2]. However, the structural parameters of the protein and the substrate which are responsible for the uncoupling process are not well understood. Data are increasingly accumulated indica- ting that the dynamics of the protein structure and in particular the accessibility of the active site for water molecules are very important [3]. In the oxidized form of P450 the high-spin/low-spin state equilibrium reflects a time-averaged population of water molecules at the sixth iron co-ordination site. This equilibrium can be monitored using the heme Soret band [4]. However, for the iron- reduced form there is no spectral signal that could be used directly to monitor the water exchange. An indirect method is a water replacement technique using a probe molecule. In a large number of studies [4–9] using different approaches we found that the CO iron ligand is a good probe for the polarity and therefore for the presence of water molecules in the heme environment of cytochrome P450cam. To get a further insight into the dynamics of the water exchange process in different substrate P450 complexes we used the high-pressure stopped-flow technique [10,11]. The activation volume as well as the rate constant for the CO on-reaction obtained from such studies should allow us to quantitate dynamic properties of the heme pocket when P450 complexed with homologous substrates is studied. High-pressure flash photolysis studies on ferrous heme model complexes and heme proteins with imidazole as proximal ligand show that the sign of the activation volume for the overall on-reaction depends on the nature of the ligand indicating two main steps, the iron-ligand bond formation (negative DV # ) and the entry of the ligand into the protein (positive DV # ), which can be rate-limiting [12]. It was found that the overall activation volume for the CO ligand binding in heme proteins with histidine proximal ligand is always negative indicating that the bond formation is the rate-limiting step. Considering these results it was surprising that cytochromes P450 do not seem to show the same behaviour. Lange et al. [11] have determined the activation volumes for the CO binding in several cyto- chromes P450 in the absence of a substrate using the stopped-flow technique under high pressure. It turned out that all the proteins which have a cysteine as proximal ligand have a small positive activation volume of (+1)– (+6) cm 3 Æmol )1 . It was concluded that the transition state in the sulfur ligand class proteins is structurally very close to the ground state and that the negatively charged sulfur from the cysteine ligand produces specific electronic properties which may be the origin for this behaviour. However, flash photolysis studies under pressure for P450cam in the presence of various substrate analogues [13] indicate that even negative activation volumes are possible. Due to the Correspondence to C. Jung, Max-Delbru ¨ ck-Center for Molecular Medicine, Protein Dynamics Laboratory, Robert-Ro ¨ ssle-Strasse 10, 13125 Berlin, Germany. Fax: + 49 30 94063329, Tel.: + 49 30 94063370, E-mail: cjung@mdc-berlin.de Abbreviations: P450, cytochrome P450; P450cam, 1R-camphor- hydroxylating P450 from Pseudomonas putida (CYP101); P420, denatured and nonactive form of P450; TMCH, 3,3,5,5- tetramethylcyclohexanone; FTIR, Fourier transform infrared (Received 28 January 2002, revised 9 April 2002, accepted 2 May 2002) Eur. J. Biochem. 269, 2989–2996 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.02980.x fact that in all these substrate complexes the cysteine ligand is the same, the specific electronic structure of the proximal ligand cannot be the origin for the positive activation volume observed for some substrate complexes. To be sure that this result is not only specific for CO rebinding induced by flash photolysis we extended the high-pressure stopped- flow study on P450cam by a homologous series of camphor analogues (1R-camphor, camphane, norcamphor and nor- bonane) and substrate-free protein. These camphor ana- logues lack characteristic groups which are relevant for a fit of the substrate into the heme pocket (Fig. 1). It will be shown that the activation volume of the CO on-rate is positive for P450cam bound with substrates which lack methyl groups, are loosely bound, have a higher water influx rate [3] and form a more compressible active site [7]. In addition, subconformers have been found for the substrate- free and camphane-bound protein which show different CO binding kinetics. MATERIALS AND METHODS Cytochrome P450cam from Pseudomonas putida expressed in Escherichia coli TB1 was isolated and purified as described [14]. The absorbance ratio e 392nm /e 280nm of the purified protein was 1.3. Substrate removal was performed by dialysis against 50 m M Tris/HCl buffer, pH 7.4 and Sephadex G-25 (medium) gel chromatography and final dialysis against 100 m M potassium phosphate buffer, pH 7. The concentrated substrate-free P450cam stock solution was 1.1 m M . To have comparable conditions to previous other experiments we used 100 m M potassium phosphate buffer, pH 7.3 (20 °C), 10% (w/w) glycerol to which aliquots of the P450cam stock solution were added. 1R-camphor was from Sigma. Camphane, norcamphor and norbornane were from Aldrich. Substrate analogues were added to the substrate-free protein as few microliters aliquot of an ethanolic stock solution. Because the substrates have different dissociation constants [3] the substrate concentration was chosen such that substrate complex was completely formed. The amount of high-spin state content at 20 °Cwasestimatedfromthe Soret band spectrum of the oxidized protein using the fit procedure described earlier [4]. The P450cam concentration before mixing was 5–6 l M in all experiments. We always mixed equal volumes of an enzyme solution with the CO solution. The buffer and substrate composition was the same in both volumes. Both solutions were carefully deoxy- genated by purging with argon before the experiment, and the same amount of sodium dithionite was added to each syringe to have always a constant final dithionite concen- tration of 1.7 m M . This dithionite concentration guaranteed that P450 remained reduced during the stopped-flow experiment. The CO containing solution was prepared by adding an appropriate volume of a CO saturated buffer stock solution to the syringe. The CO stock buffer solution is % 1m M at 20 °C calculated by the Henry’s law [15]. Because the binding kinetics strongly differ for the different substrate complexes the final CO concentration has to be varied to stay in a time window which can be resolved by the stopped-flow-spectrometer. All stopped-flow experiments were carried out between 3.8 °Cand5.6°C. The tempera- ture was stable during the experiment (± 0.2 °). After each stopped-flow experiment the recovered protein solution was checked for possible P420 formation using the CO differ- ence spectrum. There was no spectral difference to the solution at the beginning indicating that P420 was not formed during the high-pressure stopped-flow experiment. The high-pressure stopped-flow apparatus used is inter- faced with the Aminco DW2 spectrometer and is described in [10,11]. All kinetic traces were recorded in the dual- wavelength mode of the Aminco using the wavelength of the maximum at k 1 ¼ 446 nm and the minimum at k 2 ¼ 406 nm. We have previously found for P450s that the observed rate constant for the CO binding is linearly related to the CO concentration indicating bimolecular binding kinetics [11,16]. To get the k on rate constants the time curves for the absorbance difference DA(t) were fitted with bimolecular kinetics as described recently [5] (Eqn 1). [P450] 0 , [P450CO] 1 ,[CO] 0 , e,andl are the initial P450 concentration, the final P450-CO concentration, the initial CO concentration, the extinction coefficient at 446 nm, and the optical pathlength, respectively. eÆl, [P450CO] 1 ,and k on,i were used as fit parameters. The subscript letter i indicates the first or second phase in case of two-phase kinetics (see below). Fig. 1. Structure of the active site of cytochrome P450cam and of camphor analogues. Top, heme and amino acids contacting the sub- strate 1R-camphor, PDB accession no. 3cpp; bottom, substrate ana- logues used in the high-pressure stopped-flow study. DA i ðtÞ¼e Á ‘ Á " P450½ 0 exp k on;i Á t Áð½CO 0 À½P450 0 Þ Âà À 1 exp k on;i Á t Áð½CO 0 À½P450 0 Þ Âà À ½P450 0 ½CO 0 À½P450CO 1 # ð1Þ 2990 C. Jung et al.(Eur. J. Biochem. 269) Ó FEBS 2002 Three shots at each pressure were taken and fitted. The k on values for the same pressure were averaged. The averaged values were used for further analysis to get the activation volumes DV # according to Eqn. 2. @ ln k on @P         T ¼À DV # RT ; @ ln K @P         T ¼À DV RT ð2Þ For the camphane complex and the substrate-free complex two bimolecular processes were required to get a reasonable fit indicating subconformer equilibrium. We used a linear combination as the simplest approximation (Eqn 3). The fraction w for one phase is used to estimate the equilibrium constant K ¼ w/(1–w) for the subconformer equilibrium. The reaction volume DV between the subcon- former is calculated according to Eqn. 2. DAðtÞ¼w Á DA 1 ðtÞþð1 À wÞÁDA 2 ðtÞð3Þ RESULTS Figure 2 shows the typical time traces obtained from the stopped-flow measurements. There is a delay time of 0.622 s after the trigger signal was initiated and before the two volumes with the enzyme and the CO are mixed. At the lowest absorbance the time is set to zero for fitting. The data for the different substrate complexes show that the binding rate can decrease or increase with increasing pressure. As an example, Fig. 2 demonstrates the results for 1R-camphor, where the rate increases with pressure, and for norbornane, where the rate decreases with pressure. The bimolecular rate constants given in Fig. 2 are obtained by nonlinear least- square fitting the time curves using a single bimolecular process according to Eqn (1). Figure 3 shows the plot of the logarithm of the rate constant vs. the pressure which is linear. The activation volume, obtained from the slope of this linear dependence, is strongly negative by % )19.6 cm 3 Æmol )1 for the CO binding in the 1R-camphor-bound P450cam. In contrast, the activation volumes are positive for the norcam- phor-bound as well as for the norbornane-bound proteins (% +8 cm 3 Æmol )1 for both, Table 1). While the CO binding in 1R-camphor-bound P450cam is very slow (k on % 3 · 10 4 M )1 Æs )1 ) the rate is significantly increased for both of the other substrates (k on % 381 · 10 4 M )1 Æs )1 , norcamphor and k on % 332 · 10 4 M )1 Æs )1 , norbornane). In contrast to P450cam bound with 1R-camphor, nor- camphor and norbornane, the time curves for substrate-free P450cam and P450cam bound with camphane could not be fitted satisfactorily with only one bimolecular process. Figure 4 shows the data for the camphane complex as an example. As the simplest approximation we used a linear combination of two bimolecular processes to fit the curves. At 1 bar the fractions of the slow and the fast phases are approximately equal. For the camphane complex the fraction of the slow phase is almost constant up to 1150 bar (52–55%) but increases to 75% at further pressure elevation up to 1380 bar (Fig. 5, Table 1). The activation volumes DV # for both binding phases are negative. The absolute value of DV # for the fast phase is approxi- mately twice that of the slow phase ()18.2 cm 3 Æmol )1 vs. )10.6 cm 3 Æmol )1 , Table 1). In the pressure range higher than 1150 bar, the activation volumes become even more negative (Table 1). For substrate-free P450cam the fraction w of the fast phase gradually increases from % 54% at 1 bar to %65% at 1000 bar. The plot of ln(w/(1 ) w)), which corresponds to the logarithm of the equilibrium constant between the fast phase conformer to the slow-phase conformer, vs. the pressure, allows the estimation of the reaction volume DV ¼ V fast ) V slow to be approximately +11 cm 3 Æmol )1 (Fig. 6). The activation volumes, DV # are positive in both phases (10.4 cm 3 Æmol )1 for the fast phase and % 4.7 cm 3 Æmol )1 for the slow phase; Table 1, Fig. 6). The binding rate constants for both phases in substrate-free P450cam are significantly higher (k on,slow % 29.5 · 10 4 M )1 Æs )1 and k on,fast % 297 · 10 4 M )1 Æs )1 ) compared to the respective values for camphane (k on,slow % 1.6 · 10 4 M )1 Æs )1 and k on,fast % 7.8 · 10 4 M )1 Æs )1 ) andalsofor1R-camphor (k on $ 3 · 10 4 M )1 Æs )1 ). In com- parison to the other substrate complexes which also have positive activation volumes (norcamphor, norbornane), the rate constant for the fast phase in substrate-free P450cam is similar or slightly lower, while for the slow phase it is approximately 10 times smaller (Table 1). DISCUSSION The high-pressure stopped-flow study on the CO binding in cytochrome P450cam revealed two important results: (a) The substrate complexes studied can be divided into two Fig. 2. Time-dependent absorbance change at 446 nm recorded at low and high pressure in the stopped-flow experiment on cytochrome P450cam bound with two different substrates. Bottom, 1R-camphor; top, norbornane. The curves for norbornane were offset for better view. The rate constants k on are obtained by fitting the curves with a bimolecular kinetics as described in Materials and methods. The k on mean values are given with their ± SD. Experimental conditions are summarized in Table 1. Ó FEBS 2002 High-pressure stopped-flow for P450 CO binding (Eur. J. Biochem. 269) 2991 groups. The one group is characterized by a positive activation volume and a fast CO binding (substrate-free, norcamphor and norbornane). The other group shows a negative activation volume and slow CO binding kinetics (1R-camphor and camphane). (b) There are two complexes which show two-phase CO binding kinetics (substrate-free, camphane). In the following both these findings will be discussed. The presence of methyl groups in the substrate changes the rate-determining step for CO binding Unno et al. [13] reported CO flash photolysis experiments under high pressure on cytochrome P450cam bound with various camphor analogues and on the substrate-free protein. They found that 1R-camphor, fenchone, 3-endo- bromocamphor and 3,3,5,5-tetramethylcyclohexanone show negative activation volumes and slow rebinding kinetics while the substrates norcamphor and adamantane and the substrate-free protein have positive activation volumes and fast rebinding kinetics. Stopped-flow and flash photolysis studies should give comparable results at normal temperatures (> 5 °C). Indeed, our data confirm qualita- tively the finding by Unno et al. although other camphor analogues except norcamphor have been used. Combining the data from the flash photolysis and the stopped-flow studies, we sort the substrate analogues into two classes: classI(negativeDV # ,smallk on :1R-camphor, camphane, fenchone, 3-endo-bromocamphor and 3,3,5,5-tetramethyl- cyclohexanone) and class II (positive DV # , large k on : norcamphor, norbornane and adamantane and the sub- strate-free protein). All class I substrates possess methyl groups while class II substrates do not. We conclude that the methyl groups present in the substrate are the relevant structural entities which modulate significantly the CO binding properties of P450cam. The crystal structure for 1R-camphor-bound protein [17] shows that 1R-camphor is held in an optimal orientation by (a) the hydrogen bond Table 1. Activation volume DV # and binding rate constant k on for the CO binding in cytochrome P450cam bound with different substrates obtained from stopped-flow measurements as function of the hydrostatic pressure monitored at the Soret band. Experimental conditions a k on b DV #b Substrate High-spin (%at20°C) P450cam (l M ) Substrate (l M ) CO (l M ) T °C Pressure(bar) Slow phase (10 4 M )1 Æs )1 ) Fast phase (10 4 M )1 Æs )1 ) Pressure range (bar) Slow phase (cm 3 Æmol )1 ) Fast phase (cm 3 Æmol )1 ) Substrate-free 5.1 3.05 – 20 5.0 1 29.50 ± 0.70 (46%) 297 ± 7 (54%) 35–1140 4.6 ± 2.0 10.2 ± 2.1 Norcamphor 41.1 2.94 4000 10 4.5 1 – 381 ± 43 (100%) 206–1214 – 7.6 ± 2.0 Norbornane 69.9 3.18 2000 5 3.8 1 – 332 ± 14 (100%) 208–1418 – 8.4 ± 0.8 Camphane 91.8 2.68 400 50 4.8 1 1150 1380 1.60 ± 0.02 (52%) 2.50 ± 0.04 (55%) 6.00 ± 0.10 (75%) 7.8 ± 0.1 (48%) 24.2 ± 0.4 (45%) 117.9 ± 1.8 (25%) 11–1150 1150–1490 )10.6 ± 1.1 )53.1 ± 21.9 )18.2 ± 3.8 )138.8 ± 27.8 1R-Camphor 96.5 2.80 400 50 5.6 20.0 1 1 3.00 ± 0.03 (100%) 2.95 ± 0.04 (100%) – – 14–1515 4–1311 )19.6 ± 0.9 )13.2 ± 0.8 – – a 100m M potassium phosphate buffer, pH 7.3, 10% (w/w) glycerol, substrate dissociation constants [3]: norcamphor (345 l M ), norbornane (47 l M ), camphane (1.1 l M ), 1R-camphor (0.8 l M ), 1.7 m M sodium dithionite, values for concentrations correspond to the mixture. b The mean values for k on and DV # are given with their ± SD. Fig. 3. Plot of lnk on against the pressure for cytochrome P450cam bound with different substrates. The experimental conditions are given in Table 1. r 2 is the regression coefficient for the linear regression ana- lysis: 1R-camphor (0.97); norcamphor (0.71) and norbornane (0.93). 2992 C. Jung et al.(Eur. J. Biochem. 269) Ó FEBS 2002 between its keto group and the hydroxyl group of the amino-acid residue Tyr-96, and (b) by hydrophobic contacts of its methyl groups C-8, C-9 to Val295 and Asp297 in the b 3 sheet, and of the methyl group C-10 to Val247 in the I helix and Thr185 in the F helix (Fig. 1). Disturbing these interactions leads to a higher substrate mobility and acces- sibility of the heme pocket for water molecules [5–9,18,19 1 ]. We have recently found for the same homologous series of camphor analogues used for this stopped-flow study that the amount of high-spin state content which can be trapped by a negative temperature jump (fast freezing) from 297 K to 77 K depends strongly on the presence of substrate methyl groups and correlates with the initial high-spin state content at 297 K [3]. The slope of the loss of the high-spin- state content DHS with the temperature change DT (from 297 K to 77 K within 10 min) represents a water influx rate for the heme pocket. The inverse value of the water influx rate has been defined in [3] as rigidity factor. As seen in Table 2 the water influx rate is clearly smaller for substrate complexes with negative activation volume for CO binding (camphor and camphane) compared to those substrate complexes with positive activation volumes (norcamphor, norbornane). In addition, the resulting CO complex has a smaller compressibility for substrates causing a negative activation volume compared to those with positive activa- tion volume (Table 2). It has been discussed in various papers [12,13,23] that a positive activation volume indicates that the entry of CO into the protein is the rate-limiting step of CO binding. In contrast, a negative activation volume points to the Fe-CO bond formation as the rate-limiting step. However, the Fe-CO bond formation step itself (geminate binding) is very fast and independent of CO concentration [24] if the CO molecule has found the optimal place close to the iron. It is Fig. 6. Plot of lnk on against the pressure for substrate-free cytochrome P450cam. Inset: logarithm of the equilibrium constant K ¼ w/(1 ) w) with w being the fraction of the fast phase. The activation volume DV # (10.9 ± 0.8 cm 3 Æmol )1 ) is obtained from the slope of the linear fit. r 2 is the regression coefficient for the linear regression analysis: slow phase (0.37, this slow regression coefficient is caused essentially by the extreme points around 35 bar and 300 bar); fast phase (0.71). Fig. 5. Plot of lnk on against the pressure for cytochrome P450cam bound with camphane. Inset: fraction of the slow phase. r 2 is the regression coefficient for the linear regression analysis: slow phase (0.92 for P < 1200 bar, 0.75 for P > 1200 bar); fast phase (0.72 for P < 1000 bar, 0.93 for P > 1000 bar). Fig. 4. Time-dependent absorbance change at 446 nm recorded in the stopped-flow experi- ment on cytochrome P450cam bound with camphane. The experimental curve is fitted with a single and with two bimolecular bind- ing processes according to Eqns (1) and (2). Only two processes fit the experimental curve well. Experimental conditions are summarized in Table 1. Ó FEBS 2002 High-pressure stopped-flow for P450 CO binding (Eur. J. Biochem. 269) 2993 the probability of finding this optimal place which causes the rate limitation. This suggestion can be explained for P450cam using the values for the activation enthalpy DH on # and activation entropy DS on # of CO binding in substrate-free and 1R-camphor-bound P450cam determined from flash photolysis studies by Kato et al.[25].DH on # is 31.8 kJÆmol )1 for camphor-bound P450cam. This value is increased to 61.9 kJÆmol )1 in substrate-free protein. The activation enthalpy may be written as DH on # ¼ DE on # + PÆDV # [15] where DE on # is the internal energy of activation which may be assigned to the energy needed to break bonds or other contacts (e.g. hydrogen bonds) or to induce a conformational change accompanied with forming the transition state for CO binding. PÆDV # is the volume work which has to be applied to the system for CO binding. The energetic contribution of this volume work to DH on # is, however, negligibly small ()0.0019 kJÆmol )1 for camphor-bound; +0.00046 kJÆmol )1 and +0.000102 kJÆmol )1 for substrate- free P450cam at 1 bar using the activation volumes from Table 1). The larger value for DH on # in substrate-free P450cam indicates therefore that stronger bonds or more bonds have to be broken during CO binding which let one expect a slower binding rate compared to camphor- bound P450cam. However, DS on # is )43.6 JÆK )1 Æmol )1 for camphor-bound and +98.7 JÆK )1 Æmol )1 for substrate-free P450cam. The energetic contribution of the entropic term (–TÆDS)at5°C to the free enthalpy of activation DG on # is 12.13 kJÆmol )1 and )27.45 kJÆmol )1 for camphor-bound and substrate-free P450cam, respectively. Therefore, DG on # for substrate-free P450cam is lower (34.44 kJÆmol )1 )than DG on # for camphor-bound P450cam (43.93 kJÆmol )1 )mean- ing that k on (free) > k on (bound). Thus, the entropic part is the major contribution which makes CO binding in substrate- free P450cam faster than in the presence of camphor [23]. The large positive activation entropy in substrate-free P450cam may indicate that the CO molecule travels along many pathways to the heme iron. Along each pathway, however, many contacts (e.g. contacts to many water molecules) have to be broken, reflected in the large positive activation enthalpy. The higher flexibility, respective stronger compres- sibility (Table 2) of the structure in substrate-free P450cam is in agreement with this view. In contrast, in the presence of camphor the activation state is highly ordered as seen by the negative activation entropy. Camphor makes the protein and the heme pocket more rigid (smaller compressibility, Table 2) and the CO molecule has few or even only one pathway to approach the heme iron where it immediately sticks in the right position for bond formation leading to volume contraction (negative activa- tion volume). Along each of these few pathways obviously only a small number of contacts are necessary to cleave (low positive value for DH on # , e.g. because less water molecules are present). In conclusion, CO binding in camphor-bound P450cam is statistically disfavoured and therefore slow. Table 2. Comparison of k on and DV # for the CO binding in cytochrome P450cam bound with class I and II substrates obtained from stopped-flow (SF, Table 1), flash photolysis (F [13]), and FTIR-flash photolysis (F-FTIR [5]), studies. Substrate Method T (°C) k on (10 4 M )1 Æs )1 ) DV # (cm 3 Æmol )1 ) Water influx rate DHS%/ (KÆ10 min) [3] b a (GPa )1 ) [7] Class I 1R-Camphor SF 5.6 3.0 )19.6 0.147 0.00713 SF 20.0 2.95 )13.2 F 20.0 10.0 )31.0 F-FTIR b 26.8 9.8 (1939.1 cm )1 )– Camphane SF 4.8 1.6 (52%) & 7.8 (48%) )10.6 & )18.2 0.302 0.00638 F-FTIR b 26.8 10.4 (61%; 1939.4 cm )1 ) & 49.7 (25%; 1949.2 cm )1 ) – Bromocamphor F 20.0 55 )32.0 – 0.00981 TMCH F 20.0 75 )14.0 – – Fenchone F 20.0 150 )20.0 – 0.01271 Class II Substrate-free SF 5.0 29.5 (46%) & 297 (54%) 4.6 & 10.2 – 0.01228 F 20.0 850 4 F-FTIR b 26.8 158.1 (54%; 1941.1 cm )1 ) & 132.5 (8%; 1951.9 cm )1 ) & 381.3 (31%; 1960.1 cm )1 ) – Norbornane SF 3.8 332 8.4 0.538 – F-FTIR b 26.8 343.8 (1953.3 cm )1 )– Norcamphor SF 4.5 381 7.6 0.571 0.01445 F 20.0 1000 3.0 F-FTIR b 26.8 340.8 (1946.1 cm )1 )– Adamantane F 20.0 1300 7.0 – 0.0113 a b is the isothermal compressibility determined from the following equation using the absolute value for the slope of the linear pressure- induced red-shift of the Soret band maximum m in P450cam-CO. m 0 is the Soret band maximum extrapolated to 1 bar using the regression parameters for the particular substrate complex given in [7]. const has been assumed to be equal to 1. b ¼À 1 V ½ @V @P  T  const: Á 1 m o ½ @m @P  T ; b The values in parantheses give the percentage population and the CO stretch mode frequency of the substate. The FTIR data are obtained for a D 2 O buffer solution [5]. 2994 C. Jung et al.(Eur. J. Biochem. 269) Ó FEBS 2002 Extending this conclusion to all the other substrate complexes of P450cam studied, we note that for the class I substrates CO binding is disfavoured because of a rigid heme pocket and the search for the optimal place near the heme for CO-iron bond formation appears to be rate- limiting. In contrast, the CO entry into the protein and the CO migration through the protein to the heme iron favours statistically CO binding for class II substrates. The lack of methyl groups in the substrate and the higher substrate mobility and water accessibility are the relevant structural parameters which allow that another step besides diffusion becomes rate-determining when going from the Ôhypothet- icalÕ protein-free heme to the protein. This step is purely entropically driven. The positive activation volume in P450 is therefore indicative rather for a high solvent accessibility of the heme pocket than for a diffusion limited process. Subconformers of P450cam have different k on and D V # for CO binding The CO binding time traces for substrate-free and cam- phane-bound P450cam had to be fitted with two processes. Biphasic kinetics were also observed for substrate-free P450cam in the flash photolysis study under pressure by Unno et al. [13]. At a first glance one could suppose that cytochrome P420 was formed during the experiment as discussed by Unno et al. However, in our studies the spectral analysis before and after the stopped-flow experi- ments as well as a spectral comparison with the substrate complexes with mono-phase behaviour clearly excludes this possibility (data not shown). Because biphasic kinetics are observed already at ambient pressure we conclude that rather an equilibrium of subconformers with different CO binding behaviour exists than a pressure dependence of the activation volume for the pressure range lower than % 1100 bar. Indeed, conformational substates in P450cam have been observed and extensively studied by FTIR using the CO stretch vibration mode as spectroscopic probe [6,8,9, 14,26]. Many of the substrate complexes of P450cam-CO studied reveal conformational substates at low temperatures (< 160 K). At room temperature however, the transitions between substates become rather fast resulting in an averaged CO stretch infrared band or in shift of the equilibrium to only one substate. Many of the substrate complexes appear therefore as a single substate at room temperature [9] (e.g. 1R-camphor, norcamphor, norborn- ane). In contrast, the infrared spectra of substrate-free and camphane-bound P450cam-CO are an overlap of several subconformer bands even at room temperature which can be merged into two main subconformer ensembles [sub- strate-free: at % 1940 cm )1 (% 60%) and % 1952–1963 cm )1 (% 40%) and camphane: at % 1941 cm )1 (% 60%) and 1955–1962 cm )1 (% 40%)] [6,9]. This subconformer beha- viour could explain the biphasic CO binding kinetics observed in our stopped-flow studies. The fractions of slow and fast phases match approximately the population of the main subconformer ensembles in both P450 complexes. Because the activation volumes for both phases in substrate- free, respective camphane-bound, P450cam are qualitatively similar (positive for substrate-free and negative for cam- phane) we exclude that one of the two phases in the camphane complex is caused by a fraction of P450 that has not bound camphane. Recently, we have found by CO flash photolysis time-resolved FTIR studies [5] that the subcon- formers have different CO rebinding rate constants. This finding agrees with the observation in the present stopped- flow study. Within the same P450 complex the subcon- formers with the higher CO stretching mode frequency generally rebind faster (Table 2). In addition, in substrate-free P450cam-CO the popula- tion and the CO stretching mode frequency shift of the subconformers with higher CO stretch frequencies show an inverse behaviour on changes of hydrostatic and osmotic pressure [6]. This indicates that the CO ligand in these subconformers is more influenced by the solvent, which is in line with the higher positive activation volume for the fast phase compared to the slow phase of the CO binding curves obtained in the stopped-flow experiments (Table 1). In the static pressure dependence study [6] the population of the subconformer with the higher CO stretching mode fre- quency increases by % 11% with increasing pressure (from % 62% at 1 bar to % 73% at 1600 bar) and the reaction volume is in the order of 9 cm 3 Æmol )1 . In the present stopped-flow experiment we found that the fraction w of the fast phase increases by % 11% (from % 54% at 1 bar to % 65% at 1000 bar) which may reflect a pressure-induced shift of the subconformer equilibrium to a higher-frequency (faster CO binding) subconformer. The reaction volume (DV ¼ V fast ) V slow ) obtained from the plot of ln(w/(1 ) w)) vs. pressure is approximately +11 cm 3 Æmol )1 and seems to be in reasonable agreement with the value of the static high- pressure study. In contrast to substrate-free protein, the fast phase in stopped-flow CO binding kinetics of the camphane com- plex, which we assign to the fast rebinding in the FTIR flash photolysis experiment and to the higher-frequency CO stretching mode, shows a more negative activation volume ()18.2 cm 3 Æmol )1 ) than the slow phase ()10.6 cm 3 Æmol )1 ). This behaviour is different to substrate-free P450cam. This might indicate that the subconformers in the camphane complex do not originate from different solvent accessibility but for example from different orientations of the substrate itself within the heme pocket. The strong increase of the negative value of the activation volume at pressures higher than % 1100 bar (Fig. 5) might indicate that the volume is actually pressure dependent or the compressibility is changed, for example, due to substrate rearrangement in the heme pocket. Summarizing the outcome of the present high-pressure stopped-flow study under consideration of the different flash photolysis studies and diverse other studies on P450cam we suggest that the accessibility of the protein for water molecules is a relevant property which is modulated by substrate binding. The positive sign of the activation volume for CO binding is rather indicative for solvent accessibility and flexibility of the protein than for diffusion-controlled CO binding or for a specific electronic structure of the thiolate proximal ligand compared to the imidazole proximal ligand as earlier assumed [11]. Con- cerning the functional significance one may conclude at least for the camphor-hydroxylating cytochrome P450cam sys- tem that a suboptimal fit of the substrate in the heme pocket increases the mobility of the substrate, facilitates the access for water molecules and makes the heme pocket more compressible. Under these conditions the tight structural coupling for a specific proton transfer is disturbed which Ó FEBS 2002 High-pressure stopped-flow for P450 CO binding (Eur. J. Biochem. 269) 2995 may favour the formation of hydrogen peroxide or of water in the oxidase reaction over the substrate hydroxylation [26,27]. For example, with 1R-camphor only % 3–7% of the consumed dioxygen is released as hydrogen peroxide while with norcamphor 20–40% of H 2 O 2 is formed [26,28]. Both substrate P450cam complexes show monophasic CO bind- ing but with different sign of the activation volume (negative for camphor and positive for norcamphor). ACKNOWLEDGEMENTS We thank Dieter Schwarz for critical reading of the manuscript. Financial support from the Deutsche Forschungsgemeinschaft (Sk35/ 3–1,2,4), the Institut National de la Sante ´ et de la Recherche Me ´ dicale and the Deutscher Akademischer Austauschdienst in the frame of the PROCOPE programme (312/pro-ms) is acknowledged. REFERENCES 1. Lewis, D.F.V. (1996) Cytochromes P450 ) Structure, Function and Mechanism. Taylor & Francis Ltd, London. 2. Bast, A. (1986) Is formation of reactive oxygen by cytochrome P-450 perilous and predictable? Trends Pharm. Sci. 7, 266–270. 3.Schulze,H.,HuiBonHoa,G.&Jung,C.(1997)Mobilityof norbornane-type substrates and water accessibility in cytochrome P-450cam. Biochim. Biophys. Acta 1338, 77–92. 4. Jung, C., Ristau, O. & Rein, H. (1991) The high-spin/low-spin equilibrium in cytochrome P-450 – a new method for determin- ation of the high-spin content. Biochim. Biophys. Acta 1076, 130–136. 5. Contzen, J. & Jung, C. (1998) Step-scan time-resolved FTIR spectroscopy of cytochrome P-450cam carbon monoxide complex: asaltlinkinvolvedintheligandrebindingprocess.Biochemistry 37, 4317–4324. 6. Jung, C., Ristau, O., Schulze, H. & Sligar, S.G. (1996) The CO stretching mode infrared spectrum of substrate-free cytochrome P-450cam-CO: effect of solvent conditions, temperature, and pressure. Eur. J. Biochem. 235, 660–669. 7. Jung, C., Hui Bon Hoa, G., Davydov, D., Gill, E. & Heremans, K. (1995) Compressibility of the heme pocket of substrate analogue complexes of cytochrome P-450cam-CO: The effect of hydrostatic pressure on the Soret band. Eur. J. Biochem. 233, 600–606. 8. Schulze, H., Ristau, O. & Jung, C. (1994) The carbon monoxide stretching modes in camphor-bound cytochrome P450cam – the effect of solvent conditions, temperature, and pressure. Eur. J. Biochem. 224, 1047–1055. 9. Jung, C., Schulze, H. & Deprez, E. (1996) The role of the polarity of the heme environment for the CO stretch modes in cytochrome P-450cam-CO. Biochemistry 35, 15088–15094. 10. Balny, C., Saldana, J.L. & Dahan, N. (1984) High pressure stopped flow spectrometry at low temperature. Anal. Biochem. 139, 178–189. 11. Lange, R., Heiber-Langer, I., Bonfils, C., Fabre, I., Negishi, M. & Balny, C. (1994) Activation volume and energetic properties of the binding of CO to hemeproteins. Biophys. J. 66, 89–98. 12. Taube, D.J., Projahn, H D., van Eldik, R., Magde, D. & Traylor, T.G. (1990) Mechansim of ligand binding to heme and hemeproteins. High-pressure study. J. Am. Chem. Soc. 112, 6880– 6886. 13. Unno, M., Ishimori, K., Ishimura, Y. & Morishima, I. (1994) High-pressure flash photolysis study of heme proteins: effect of substrate analogues on the recombination of carbon monoxide to cytochrome P450 CAM . Biochemistry 33, 9762–9768. 14.Jung,C.,HuiBonHoa,G.,Schro ¨ der, K L., Simon, M. & Doucet, J.P. (1992) Substrate analogue induced changes of the CO-stretching mode in cytochrome P-450cam carbon monoxide complex. Biochemistry 31, 12855–12862. 15. Atkins, P.W. (1986) Physical Chemistry, 3nd edn. Oxford Uni- versity Press, Oxford, UK. 16. Contzen, J., Ristau, O. & Jung, C. (1996) Time-resolved Fourier- transform infrared studies of the cytochrome P-450cam carbon- monoxide complex bound with (1R)-camphor and (1S)-camphor. FEBS Lett. 383, 13–17. 17. Poulos, T.L., Finzel, B.C. & Howard, A.J. (1987) High-resolution crystal structure of cytochrome P450 cam . J. Mol. Biol. 195, 687–700. 18. Atkins, W.M. & Sligar, S.G. (1989) Molecular Recognition in cytochrome P-450: alteration of regioselective alkane hydroxyla- tion via protein engineering. J. Am. Chem. Soc. 111, 2715–2717. 19. Das, B., Helms, V., Lounnas, V. & Wade, R.C. (2000) Multicopy molecular dynamics simulations suggest how to reconcile crystallographic and product formation data for camphor enan- tiomers bound to cytochrome P450cam. J. Inorg. Biochem. 81, 121–131. 20. Hasinoff, B.B. (1974) Kinetic activation volumes of the binding of oxygen and carbon monoxide to hemoglobin and myoglobin studied on a high-pressure flash photolysis apparatus. Biochemis- try 13, 3111–3117. 21. Caldin, E.F. & Hasinoff, B.B. (1975) Diffusion-controlled kinetics in the reaction of ferroprotoporphyrin IX with carbon monoxide. J. Chem. Soc. Faraday Trans. I 71, 515–527. 22. Traylor, T.G., Magde, D., Taube, D.J., Jongeward, K.A., Bandyopadhyay, D., Luo, J. & Walda, K.N. (1992) Geminate recombination of carbon monoxide complexes of heme and heme proteins. J. Am. Chem. Soc. 114, 417–429. 23. Stochel, G. & van Eldik, R. (1999) Elucidation of inorganic reaction mechanisms through volume profile analysis. Coord. Chem. Rev. 187, 329–374. 24. Te ´ treau,C.,DiPrimo,C.,Lange,R.,Tourbez,H.&Lavalette,D. (1997) Dynamics of carbon monoxide binding with cytochromes P450. Biochemistry 36, 10262–10275. 25. Kato, M., Makino, R. & Iizuka, T. (1995) Thermodynamic aspects of the CO-binding reaction to cytochrome P-450cam. Relevance with their biological significance and structure. Bio- chim. Biophys. Acta 1246, 179–184. 26. Jung, C. (2000) Insight into protein structure and protein-ligand recognition by Fourier transform infrared spectroscopy. J. Mol. Recognition 13, 325–351. 27. Vidakovic, M., Sligar, S.G., Li, H. & Poulos, T.L. (1998) Understanding the role of the essential Asp251 in cytochrome P450cam using site-directed mutagenesis, crystallography, and kinetic solvent isotope effect. Biochemistry 37, 9211–9219. 28. Atkins, W.A. & Sligar, S.G. (1988) Deuterium isotope effects in norcamphor metabolism by cytochrome P-450cam: Kinetic evi- dence for the two-electron reduction of a high-valent iron-oxo intermediate. Biochemistry 27, 1610–1616. 2996 C. Jung et al.(Eur. J. Biochem. 269) Ó FEBS 2002 . Substrates modulate the rate-determining step for CO binding in cytochrome P450cam (CYP101) A high-pressure stopped-flow study Christiane Jung 1 , Nicole. observed rate constant for the CO binding is linearly related to the CO concentration indicating bimolecular binding kinetics [11,16]. To get the k on rate constants