Mechanism of interaction between cytochromes P-450 RLM5 and b5: Evidence for an electrostatic mechanism involving cytochrome b5 heme propionate groups

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 245, No. 2, March, pp. 512-522,1986

Mechanism

of interaction between Cytochromes P-450 RLMS and b5: Evidence for an Electrostatic Mechanism Involving Cytochrome b5 Heme Propionate Groups PAUL

Department

P. TAMBURINI

AND JOHN

B. SCHENKMAN

of Pharmacology, University of Connecticut Health Centi, Farmingtan,

Connecticut 06032

Received August 1,1985, and in revised form November 12,1985

The role of cytochrome bs heme propionate groups in the functional interactions between cytochromes P-450 RLM5 and bs has been investigated by comparing the capacity of RLM5 to interact with both native bs and a b5 derivative in which the native heme was replaced with ferric protoporphyrin IX dimethyl ester (DME-b5). Both forms of bs interacted with RLM5 causing an increase in the RLM5 spin state from 28 to 68% highspin RLM5 at saturation, as judged using uv-visible spectrophotometry. However, DME-br, exhibited a 7-fold weaker affinity for RLM5. The apparent dissociation constant (&) for the interaction between RLM5 and b5 was also shown to be a strong function of ionic strength, in a manner consistent with the involvement of electrostatic attraction in complex formation. Reconstitution of b5 into an RLM&dependent monooxygenase system stimulated the pnitroanisole demethylase rate about 25-fold and 7-ethoxycoumarin deethylase about 6-fold. DME-b,, however, produced only 30% of the stimulation of RLM&dependent turnover of pnitroanisole observed at equivalent concentrations turnover was 50% diminof native bswithout a change in Km. With 7-ethoxycoumarin, ished. The diminished capacity of DME-b5 to stimulate RLMS-dependent substrate turnover was shown not to be due to impairment of electron flow between NADPH-cytochrome P-450 reductase and DME-br,, since the Km of reductase for DME-b5 is 2.5-fold lower, and the V,, is actually increased, but rather to an impairment of some aspect of functional interaction between the DME-b5 and RLM5. The data show that complex formation between cytochrome P-450 and & involves electrostatic attraction mediated in part by cytochrome bs heme propionate groups. Q 19s Academic PMI, IW.

The hepatic microsomes of mammalian liver contain both NADH-driven cytochrome b5-cytochrome bs reductase and NADPH-driven cytochrome P-450-cytochrome P-450 reductase electron transfer pathways. The former pathway functions as an electron donor to a variety of microsomal enzymes including those involved in fatty acid desaturation (1) and elongation (2) while the latter pathway services the cytochrome(s) P-450-dependent oxidation of a variety of compounds including fatty acids (3), steroids (4), prostaglandins (5), and foreign compounds (6). A plethora of studies performed both on 0003-9861/86 $3.00 Copyright All righta

0 1996 by Academic Press., Inc. of reproduction in any form reserved.

intact microsomes and, later, on purified reconstituted cytochrome P-450 systems have shown an interaction between these electron transfer pathways during some substrate oxidations [see (7-10) and references cited therein]. However, the mechanism of this interaction has been elusive. A cytochrome br,-mediated stimulation of substrate oxidation catalyzed by some reconstituted cytochrome P-450 isozymes (11-14) has been attributed both to an effector role of cytochrome bson one or more of the reaction steps in the cytochrome P450 catalytic cycle (11,15-17) and/or to an efficient direct electron transfer between 512

CYTOCHROME

P-45OKYTOCHROME

the cytochromes (10, 11,16,18,19). Indeed, direct electron transfer from br, to oxyferrous LM2 (8) and LM4 (20) has been shown. In either case, complex formation between the two hemoproteins would seem a necessary step, and has been shown using a number of techniques including a study of cytochrome br,-induced spectral changes in cytochrome P-450 (9,13,21). This spectral change is characterized by a shift in P-450 Soret absorbance from 417 to 390 nm upon b5 addition and, as with similar substrateinduced perturbations, has been correlated with a cytochrome P-450 spin state change from S = l/2 to S = 512 (22,23). The nature of such protein-protein interactions is of particular interest since they may provide information as to the mechanism of b5 stimulation of P-450 turnover. We have previously shown (24), using selective chemical modification, that some of the protein side chain carboxyl groups on cytochrome b5 are essential for causing spin state changes in cytochrome P-450 as well as for mediating both electron flow between the cytochromes and a stimulation of cytochrome P-450 metabolic turnover. In the work described here we have carried the study further and examined the role of b5 heme propionate groups in the interactions with RLM5.l The results indicate a participation of these heme carboxy1 groups both in complex formation with RLM5 and in cytochrome b5-mediated turnover of RLM5. MATERIALS

AND

METHODS

Purification of microsomal enzm. RLM5 was purified to electrophoretic homogeneity from the hepatic microsomes of untreated male CD Sprague-Dawley rats (200 g) by the method of Cheng and Schenkman (4). The final preparation exhibited a specific content ’ Abbreviations used: DLPC, L-a-dilauroylphosphatidylcholine; DME-beme, ferric protoporphyrin IX dimethyl ester; DME-4, cytochrome 4 dimethyl ester; native reconstituted 4,4 formed from the reconstitution of apo-4 with native heme; FPt, NADPH-cytochrome P-450 reductase; DMSO, dimethyl sulfoxide; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis, 4, cytochrome 4; RLMS, cytochrome P-450 RLM5, RLM5a, another constitutive form of P-450, found in male and female rats (44).

4

INTERACTIONS

518

of 12 nmol P-450 mg-’ protein and an A&Am ratio of 1.2, and was free of Emulgen 911 and P-420. The hemoprotein was catalytically active in the reconstituted demethylation of benzphetamine, displaying rates of 32 nmol formaldehyde/nmol P-4501min. Hepatic microsomal NADPH-cytochrome P-450 reductase (EC 1.6.2.4) was purified to homogeneity by the method of Yasukochi and Masters (25). The rats were pretreated with phenobarbital (0.1% w/v, in their drinking water for 6 days prior to killing). The final preparation exhibited a specific content of 12.5 nmol reductase/mg protein. Cytochrome 4 was purified to homogeneity from the hepatic microsomes of phenobarbital pretreated rats using a modification of the procedure of Spatz and Strittmatter (26) essentially as described previously (9).

Preparation of native reconstituted and fewic protoporphyrin IX dimethyl ester reconstituted c&+ chromes b5. Apo-cytochrome bswas prepared by acid/ acetone extraction of native 4 as described by Cinti and 0~01s (27), except that a second extraction of resolubilized apo-b, was performed to ensure complete native heme removal. Native reconstituted cytocbrome 4 was prepared by the addition of repurified native heme (hemin, Eastman, 3.1 mM in DMSO) to a solution of apo-4 [176 pM in 0.7 ml of 100 mM Tris-HCl (pH 8.1), containing 0.4% sodium deoxycholate] to attain a molar ratio (heme/apo-4) of 2.4. After 2 h, excess heme was removed from the holoby passage of the mixture through a column of Sephadex G-25 (1.3 X 15 cm) previously equilibrated with 100 mM TrisHCI (pH 8.3) containing 0.5% sodium deoxycholate; this was followed by passage through a second, similar column of G-25 equilibrated with 50 mM sodium phosphate (pH 7.25), which removes all of the deoxycholate. The final eluate was dialyzed versus two successive 300-ml volumes of the last buffer and then concentrated to 110 pM stock concentration on an Amicon PM-10 membrane. Ferric protoporphyrin IX dimethyl ester reconstituted 4 (DME-4) was prepared by the dropwise addition of ferric protoporphyrin IX dimethyl ester (3.1 mM in DMSO, Porphyrin Products, Logan, Utah) to a gently stirring solution of apo-4 [approximately 307 pM in 0.69 ml of 40 mM Tris-HCl (pH 8.3), containing 0.2% sodium deoxycholate], at 4”C, to attain a molar ratio (heme/apo-protein) of 2.1. At various times following mixing, spectrophotometric assays were performed upon the mixture to determine the extent of reconstitution. Preliminary studies showed that maximal reconstitution occurred within 3 days. No further increases occurred when the preparations were incubated for 11 days. Excess heme was separated from the reconstituted 4 using a modification of the procedure of Reid et al (28); the mixture was chromatographed on a column of Sephadex G-100 (2.6 X 70 cm) previously equilibrated with 10 mM Tris-

514

TAMBURINI

AND

acetate (pH 8.1), containing 0.4% deoxycholate and 0.1 mrd EDTA. Fractions containing DME-4, and exhibiting an All./Azao ratio > 2.1 were pooled and passed through a column of Sephadex G-25 (21 X 1 cm) previously equilibrated with 50 mM sodium phosphate buffer (pH 7.25). The eluted protein was then dialyzed versus 2 X 200 ml of the same buffer for consecutive 8-h periods, and then concentrated 11-fold to 114 pM on an Amicon PM-10 membrane. The total final yield of DME-4 from the starting cytochrome was 23%. Analytical procedures. Spectral titrations of RLM5 (0.56 PM) with native or DME-4 were performed at 25°C in the presence of DLPC vesicles (30 pg ml-‘), in sodium phosphate buffers, pH 7.25, containing 25% glycerol, as described previously (24). RLM5-dependent pnitroanisole o-demethylation was studied at 25“C in a reconstituted system containing RLM5 (0.19 pM), NADPH-cytochrome P-450 reductase (0.19 pM), DLPC vesicles (10 pg ml-‘), and pnitroanisole (0.92 mM, Eastman), in 50 mm sodium phosphate buffer (pH 7.25), containing 25% glycerol. The incubation mixture (0.45 ml) was equilibrated at 25°C for 5 min prior to the addition of NADPH (1.1 mM final), and the rate of pnitrophenol production from p-nitroanisole was followed spectropnotometrically according to Netter and Seidel(29). In some studies the o-deethylation of 7-ethoxycoumarin (1 mM, Aldrich) was examined, using the fluorometric assay of Ullrich and Weber (30). Electron transfer from NADPH-cytochrome P-450 reductase to cytochrome(s) 4 was performed as described in the figure legends. NADPH-cytochrome P450 reductase was quantified using a value of EM nm

I

100 - A 9 s

d

1

g

SCHENKMAN = 21.4 rnrd-’ cm-’ for the oxidized flavoprotein (31). and cytochrome P-450 was quantified using a difference extinction coefficient (450-490 nm) of 91 mM-’ cm-‘, for the carbonmonoxy ferrous versus ferrous spectrum (32), or using a value of Emnm = 90 mrd cm-’ for the oxidized protein. Cytochrome 4 was quantified using a difference extinction coefficient (424-490 nm) of 112 mM cm-’ (33). SDS-polyacrylamide gel electrophoresis was performed by the method of Laemmli (34), and protein was determined as described by Lowry et al (35). Conductivity was measured with a YSI Model 32 conductance meter (Yellow Springs Instrument Co.).

RESULTS

The time course of reconstitution of apocytochrome b5 and heme dimethyl ester is shown in Fig. 1A. Unlike the reconstitution of amphipathic b6with native heme, which occurs within the mixing time, the incorporation of DME-heme was slow, proceeding to completion in 2.5 days. The addition of further amounts of DME-heme after this time produced no further increase in spectrophotometrically detectable holocytochrome (Fig. 1A) showing that effective heme concentration was not limiting. The slow reaction time was suspected to be due to a high state of aggregation of the

I

.--

i A -E

I

05

I.0

c r8

50

2 g4 ae 0

I 5 DAYS

I ID

3

8 e!!

0

IO

20 FRACTION

30

40

50

60

NUMBER

FIG. 1. Time course of reconstitution of ferric protoporphyrin IX dimethyl ester separation of the holoenzyme. (A) Time course of reconstitution of apo-cytochrome with a 2.1-fold molar excess of heme dimethyl ester. The percentages reconstitution theoretical apo-bs concentrations calculated assuming a complete recovery of apo-4 acetone extraction of native b6. Addition of more DME-heme at the arrow was (B) Elution profile of the above incubation mixture following chromatography on the separation of unincorporated (excess) heme (peak I) from holocytochrome 4 (peak in A and B were as described under Materials and Methods.

with

apo-& and (212 nmol), are based upon from the acid/ without effect. G-100, showing II). Conditions

4

CYTOCHROME

P-450KYTOCHROME

DME-heme resulting in an extremely low concentration of DME-heme monomer under the experimental conditions. A high state of aggregation of DME heme was confirmed by data showing that the DMEbs eluted behind the unincorporated DMEheme during chromatography on Sephadex G-100 (Fig. 1B); the elution position suggested aggregates of DME heme with an apparent molecular weight greater than 18,000. In contrast, excess native heme eluted much later than native reconstituted bs on Sephadex G-25, indicating an apparent molecular weight much smaller than 18,000. Figure 2 shows the uv-visible spectra of oxidized and reduced DME-b,. The spectra were highly similar to those of native bs. However, minor differences were observed, including a shift in the ferric/ferrous isosbestic point from 438.5 to 440 nm, and the resolution of the (Y band of ferrous DME-b5 into two peaks at 554 and 559 nm. Similar splitting of the (Yband of ferrous cytochrome bs heme peptide reconstituted with DME-heme (28) and of microsomal cytochrome bs at liquid nitrogen temperatures (36) has been reported. Native bs and native reconstituted bs were spectrally in-

FIG. 2. Ultraviolet-visible spectra of DME-4. Spectra of the oxidized (solid line) or sodium dithionitereduced (broken line) DME-b, (0.82 PM) are shown. Spectra were obtained in 50 mM sodium phosphate buffer, pH 7.25, containing 25% glycerol, at 5°C.

4 INTERACTIONS

513

distinguishable. From a comparison of either the ASoret/AZw in the ferric state or the specific content of native b6 (A~AzxI = 2.3, specific content = 40 nmol mg-‘) and DME-br, (A&Am = 2.1, specific content = 37 nmol mg-‘), it was calculated that less than 10% of apo-b5 was present in the DME-b5 preparation. Qualitatively, the shape of the ABO peak in DME-b5 was indistinguishable from that of the native protein, indicating that no appreciable destruction of tryptophan and tyrosine residues occurred during acid/acetone preparation of the apo-b5. The DME-b5 co-migrated as a single band with native bb in SDS-PAGE showing that DME-b5 had not undergone proteolysis during its preparation. To evaluate the role of cytochrome b6 heme propionate groups in complex formation with cytochrome P-450, native b5, native reconstituted bs and DME-heme reconstituted bs were compared in their capacity to interact with ferric RLM5 and to cause a spectrally detectable conformational change. The dependence of such spectral changes upon either native bs or DME-b5 concentration is shown in Fig. 3A. The addition of native bs induced substantial spectral changes in RLM5, the magnitudes of which were substantially greater than the corresponding spectral changes elicited by equal amounts of DME-bS. Spectral titration of RLM5 with native reconstituted & produced spectral changes of the same magnitude as observed with native bs. Representative Hanes-Woolf plots for the analysis of the spectral titration data are shown in Fig. 3B, and the resultant spectral binding parameters (Kd and AA,,,) for the titration with native b5, native reconstituted b5, and DME-& are listed in Table I. The results clearly showed that heme propionate esterification resulted in a six- to sevenfold weakening of the interaction between cytochrome bs and RLM5 as judged from the apparent Kd values. Replacement of DME-heme with native heme yielded results identical to those obtained with native b5, showing that the altered interaction observed with DME-b, was not an artifact produced as a result of the reconstitution methodology employed, nor

AND

TAMBURINI

CYTOCWRCM

b3

SCHENKMAN

(,M

FIG. 3. The effect of cytochrome bh heme propionate esterification on the spectral interactions between cytochrome 4 and cytochrome P-450 RLMS. (A) Spectral binding curves for the titration with either native 4 (closed circles) or DME-4 (open circles). Data points in each titration represent the averages of three determinations, and the lines drawn were calculated from the best-fit spectral binding parameters obtained using an iterative Hanes-Woolf analysis of the mean data points. Experiments were performed in 50 mrd phosphate buffer as described under Materials and Methods. (B) Hanes-Woolf plots for the spectral titration of RLM5 with either native 4 (0), native reconstituted 4 (A), or DME-4 (0).

the result of an increase in conformational heterogeneity in the relative orientation of heme methyl and vinyl groups in the bs heme binding domain (37). The &I,, value for the spectral titration of RLM5 with bb (Table I) was unaffected by heme propionate esterification; the capacity of cytochrome b5, once bound, to elicit spectrally detectable spin state changes in RLM5 does not involve heme propionate groups. Since methyl esterification of the &, heme propionate groups causes anionic charge neutralization, presumably with a minimum of steric alteration, the increased Kd

value obtained (Table I) is most likely the result of diminished electrostatic interactions between cytochrome bs and cytochrome P-450. The ionic strength dependence of protein-protein interactions has been used extensively to evaluate the role of electrostatic attraction. Generally, a strong inhibition of complex formation on increasing ionic strength has been cited as evidence for the involvement of electrostatic interaction (38, 39). Accordingly, we investigated the ionic strength dependence of cytochrome b-RLM5 interactions. Spectral titrations of RLM5 with native bs

TABLE THE EFFECT

OF CYTOCHROME SPECTRAL

Ligand Native Native DME-4

reconstituted

4

MODIFICATION ON THE BINDING PARAMETEIU COMPLEX FORMATION BETWEEN CYTOCHROME

(2)

4 4

a The correlation coefficients data. *The number of determinations

I

0.028 + 0.002 (3) 0.027 + 0.002 (3) 0.027 * 0.002 (3)

are the averages are given

of those

in parentheses.

obtained

DESCRIBING AND RLM5

THE CAPACITY

Spin state in the complex (5% Fe!+)

A&ax

0.42 + 0.02 (3)* 0.33 + 0.03 (3) 2.8 + 0.67 (3)

4

from

63+3 67 f 3 67 + 3 the best-fit

Hanes-Woolf

FOR

Correlation coefficient” (r) 0.999 0.998 0.967 plots

to the

CYTOCHROME

4

P-450KYTOCHROME TABLE

THE

EFFECT

OF IONIC

STRENGTH

UPON

CYTOCHROME Sodium phosphate Cm@

“The

correlation

AND

are those

INTERACTION

BETWEEN

NATIVE

at pH 7.25

(2)

0.36 2.53 4.80 8.92 coefficients

RLM5

Correlation coefficient”

Conductivity (ma, 25°C)

5.6 50 500 189

II

THE SPECTRAL

4

517

INTERACTIONS

(r)

63 358 517 2322 obtained

for the best-fit

were performed over a range of sodium phosphate concentrations, at constant pH. The resultant spectral binding parameters are summarized in Table II. Increasing the ionic strength from 5.6 to 189 mM sodium phosphate caused a 37-fold weakening of the interaction between the proteins, without significantly altering AA,,,. The effects are qualitatively similar to those observed due to heme propionate esterification, and support the involvement of electrostatic attraction in complex formation between these two proteins. Studies on the ionic strength dependence of the interaction between the tryptic catalytic fragment of cytochrome bs and either cytochrome c (38) or methemoglobin (39) have revealed a strong inverse relationship between the association constant and ionic strength, a result expected for interactions between oppositely charged proteins. In the present studies, a linear relationship was observed between In Kd and the square root of conductivity for cytochromes bs and RLM5 (data not shown), a relationship predicted by the Debye-Huckel equation for electrostatic interactions between macromolecules. Protein side chain carboxyl groups were shown previously to be important in the cytochrome &-mediated stimulation of pnitroanisole demethylation catalyzed by cytochrome(s) P-450 (24). Consequently, the importance of b5 heme propionate groups in RLM&dependent substrate oxidation was examined. DME-b, and native bs were compared in their capacity to stimulate p-nitroanisole demethylation over a range of concentrations. The results are

0.025 0.027 0.017 0.023

0.999 0.999 0.989 0.937

Hanes-Woolf

plots

to the biological

data.

shown in Fig. 4. In the absence of b5, pnitroanisole demethylation rates were extremely low. Increasing concentrations of native cytochrome bs caused a progressive increase in demethylase activity reaching an optimal 26-fold stimulation at a 2.‘7-fold molar excess of bs. The same concentrations of DME-b5 produced a much smaller stimulation of demethylase activity, although optimal stimulation was observed at the same relative bs concentration. Native reconstituted cytochrome bs was much more effective than DME-b, in stimulating p-nitroanisole demethylation, exhibiting

0

I

2 W-AR

3

4

I 5

I 6‘

RATIO ( b5/P450)

FIG. 4. The effect of cytochrome 4 heme propionate esterification on the cytochrome &-mediated stimulation of RLM5-dependent pnitroanisole demethylation. pNitroanisole demethylation was assayed at 25°C as described under Materials and Methods and is shown as a function of either native 4 (closed circles), DME-g (open circles), or native reconstituted 4 (A) concentration.

518

TAMBURINI

AND

about 80% of the stimulation afforded by native cytochrome & (Fig. 4). The impaired capacity of DME-bs to stimulate reconstituted RLM5-dependent substrate turnover is not limited to p-nitroanisole demethylation. The use of native or native reconstituted br, in levels equimolar to RLM5 stimulated 7-ethoxycoumarin turnover 4to 6-fold, from 0.08 min-’ in the absence of bs to 0.35 and 0.48 mine1 respectively. In contrast, an equivalent concentration of DME-bS produced only a 2-fold stimulation, to 0.22 min-‘. Cytochrome & and NADPH-cytochrome P-450 reductase interact with one another to form an electron transfer complex (40). Therefore, since electron transfer from the reductase to cytochrome P-450 mediated by cytochrome bs may be involved in the bsmediated stimulation of RLM5 turnover, the data in Fig. 4 could be attributed to altered interactions between the reductase and cytochrome bs rather than between cytochrome bs and cytochrome P-450. To test this possibility, we compared electron transfer from the reductase to native b5, native reconstituted b5, and DME-b5, at a ratio of reductase to bs of 1:l at the same lipid-to-reductase ratio as used in substrate turnover studies. The results are shown in Fig. 5. Native and native reconstituted bs were both reduced at similar rates, with complete reduction occurring in 50 s. In contrast, DME-b, was rapidly reduced to completion within 5 s after mixing. The results observed were not expected in view of the earlier report (40) that methylamidation of cytochrome bs heme and protein carboxyl residues resulted in elevated Km but unaltered V,,, values for electron transfer from the reductase to bs. Since those studies were performed at higher phosphate buffer concentration, we examined the kinetics of bs reduction by NADPH-cytochrome P-450 reductase at phosphate buffer concentration used both in the current study and in the earlier report. The results indicate, using conditions of the current study (Fig. 6A), that the affinity of the reductase for cytochrome bs is enhanced by a factor of about 2 following

SCHENKMAN

I

0

ID

I

zcl SECONDS

I

30

I

40

FIG. 5. The effect of cytochrome 4 heme propionate esterification upon cytochrome 4 reduction by NADPH-cytochrome P-450 reductase. NADPH-cytochrome P-450 reductase, cytochrome 4, and DLPC vesicles were preequilibrated for 30 min, diluted with 50 mM sodium phosphate, pH 7.25, 25% glycerol and transferred to a plunger cuvette. The final concentrations of each component were NADPH-cytochrome P-450 reductase (0.56 PM), cytochrome 4 (0.56 PM), and DPLC (30 pg ml-‘). After 5 min preequilibration at 25°C the reactions were initiated by the rapid addition of NADPH (1.0 mM final), and the reduction was followed in the dual-wavelength mode as an increase in AAwith time. At this wavelength couple the contribution of reductase to absorbance is less than 5% and bleaching occurs before a significant reduction of 4 occurs. AC for 4 is 110 rn& cm-‘. The reductions of native 4 (solid line), DME-4 (broken line), and native reconstituted 4 (dotted line) are shown.

heme propionate esterification. The V,,, was also increased (from 59 to 230 min-‘) by almost a factor of 5. Increasing the buffer concentration to 0.3 M, keeping all other conditions as in Fig. 6A, markedly enhanced turnover (Fig. 6B). The V,,, values observed, however, were essentially the same for reduction of native reconstituted bs and DME-b5, 6400 and 5600 min-I, respectively. Once again, the apparent Km for the DME-& was lower than the K, for native bs by about a factor of 2. Interestingly, ionic strength did not appreciably affect either Km value. DISCUSSION

DME-bs has been prepared by the reconstitution of apo-b5 with ferric protopor-

CYTOCHROME

A

Km (/IN)

‘h hid)

III

P-450XYTOCHROME

B

FIG. 6. The effect of cytochrome 4 heme propionate esterification on NADPH-cytochrome P-450 reductase-mediated 4 reduction at 50 and 300 mM phosphate. Native reconstituted 4 or DME-b, was preequilibrated with NADPH-cytochrome P-450 reductase and DLPC vesicles (5 mg ml-’ stock) for 2 h at 25’C, then diluted with sodium phosphate buffer, pH 7.3, to attain the following concentrations: 4 (0.22 to 7.12 PM), reductase (0.08 PM), DLPC (58 pg ml-‘). The solution was transferred to a drive syringe of a Durrum stopped-flow spectrophotometer, Model D-110. Equal volumes of the reconstituted system (0.25 ml) and a solution of NADPH (1 mM in the same sodium phosphate buffer) contained in the second drive syringe were rapidly mixed (dead time, ca. 2 ms) and the time course of reduction was followed as an increase in Am nm at 25°C through a 2-cm light path. Turnover rates of the reductase were obtained from first-order plots of the data. (A) Double-reciprocal plot of reductase turnover in 50 mM sodium phosphate buffer (determined in the first 2 s of the reaction). (B) Corresponding double-reciprocal plot of turnover at 300 mM sodium phosphate buffer (determined in the first 0.2 s of the reaction). Data for native reconstituted be (0) and DME-b, (0) are shown.

phyrin IX dimethyl ester, and used to study the role of cytochrome bs heme propionate groups in the interactions between b6 and RLM5, and b5 and NADPH-cytochrome P450 reductase. The similarity in spectral properties, specific content, and electrophoretic mobility between DME-b, and native bs show that the DME-b5 is an authentic cytochrome bs model with which to perform these studies. In particular, the esterification of heme propionate groups causes anionic charge neutralization. DME-b5 exhibited a much weaker affinity for RLM5 than did native bs thereby showing an involvement of the br,-propio-

br, INTERACTIONS

519

nate groups in complex formation with RLM5. The strong dependence of Kd for complex formation on ionic strength demonstrates the electrostatic nature of this interaction. Further, the data may provide an explanation for the results of Noshiro et al. (19) that increased ionic strength elevates the steady-state level of oxy-P-450 in microsomes; this effect might occur by an impairment of functional interaction between b5 and oxy-P-450 culminating in an inhibition of electron transfer between the two proteins. Interestingly, neither heme propionate esterification nor ionic strength alteration led to a change in AA,,, for the interaction between the cytochromes. In a preceding report, the capacity of bs to induce spectrally-detectable conformational changes in cytochrome P-450 was rationalized in terms of a difference in the binding affinity of bs for low- and high-spin cytochrome P-450 (9). Using similar reasoning, it is clear that since AA,,, is unaffected by heme propionate modification, the heme propionate groups on bs are involved in binding to both low-spin and high-spin cytochrome P-450 proportionally to the same degree. The heme propionate groups of cytochrome 4 are also important in stimulation of both RLM5-dependent p-nitroanisole demethylation and 7-ethoxycoumarin dealkylation. Substitution of DME-b5 for native bs caused a marked depression of these activities. In preliminary studies we have observed a similar requirement for intact bs heme propionate groups in the bs stimulation of rabbit liver LM2 and rat liver RLM5a-dependent p-nitroanisole and 7ethoxycoumarin dealkylation (data not shown). The interpretation of the dependence of RLM5-catalyzed p-nitroanisole metabolism upon cytochrome bs concentration is complicated by the fact that bs will form an electron transfer complex with both NADPH-cytochrome P-450 reductase and P-450 therefore obscuring the molecular significance of the apparent Km for bs in metabolism. Indeed, the apparent decline in the demethylase activity at above a molar ratio (b,/P-450) of 3 may reflect

520

TAMBURINI

AND

the emergence of a secondary effect involving a competition between the two hemoproteins for NADPH-cytochrome P-450 reductase. Since b5 binds tightly to both proteins (24) (also Table I and Fig. 6), and an appreciable portion of the bs is complexed, it is not possible to obtain a precise value for the Km of bs during metabolism. However, simple inspection of the data (Fig. 4) shows that optimal stimulation of pnitroanisole demethylation occurred at the same b5/P-450 ratio for both native and DME-br,. The calculated Km values for cytochrome bs and DME-br, in p-nitroanisole demethylation, using the Hanes-Woolf analysis, up to a b5/P-450 ratio of 3, were 45 and 55 nM, respectively, indicating little, if any, effect of the heme propionate carboxy1 groups on Km. According to the Briggs-Haldane steadystate equation, the Kd and Km for b,/P-450 interaction should be equal only if the rate of enzyme-ligand dissociation is much faster than the rate of substrate metabolism. The simplest interpretation of the different effects of heme propionate esterification on the Kd and Km for bs interactions, therefore, is that the rate of dissociation of the functional bJP-450 complex is slow relative to a subsequent rate-limiting step in product formation, and that during turnover, bs remains largely bound to cytochrome P-450 as a stable catalytic complex. Accordingly, the decreased rate of pnitroanisole demethylation or ‘I-ethoxycoumarin deethylation in the presence of DME-bs compared with native b5 is consistent with a decreased functional competence of the DME-bJP-450 complex. In view of the reported elevation of the redox potential of the b5-heme peptide following heme propionate esterification (28) it is tempting to attribute the decreased turnover to the increased redox potential of the b5, resulting in a decreased electromotive force for electron transfer from bs to P450. This change in bs redox potential could also explain the enhanced rate of electron transfer from NADPH-cytochrome P-450 reductase to cytochrome bs (Fig. 5). However, the similarity of V,, values for the reductase with bs and DME-b5 at high ionic

SCHENKMAN

strength (Fig. 6B) makes this suggestion less tenable. The decreased turnover of P-450 in the presence of DME-b5 as compared to the turnover in the presence of native bs could not be attributed to diminished electron flow between bs and the reductase since at low buffer strengths DME-b5 exhibited a lower Km for the reductase and a much higher V,,, for reduction (Fig. 6). This result, and the observed stimulation of b5 reduction on methylamidation of bj protein carboxyls (24), is at variance with the results of Dailey and Strittmatter (40) who reported an increased K, for bs in reductase assays following heme and protein carboxyl group methylamidation. That study was performed in 0.3 M phosphate buffer. Since the reductase turnover of bs is known to be stimulated by high ionic strength (41), a phosphate buffer concentration similar to that of Dailey and Strittmatter (40) was also used. The results indicated that DME-b5 is reduced more efficiently than native reconstituted bs regardless of buffer concentration; however, the V,,, values are essentially the same at the high buffer concentration. Of interest, although the Km for DME-b, and native reconstituted b6 differ, high ionic strength did not influence the values. Based upon our observations the interaction between bs and NADPH-cytochrome P-450 reductase must be reconsidered. The completely contrasting effects of cytochrome bs heme propionate esterification or protein carboxyl methylamidation (24) and ionic strength on & interactions with either NADPH-cytochrome P-450 reductase or RLM5 show that the mechanisms of complex formation are completely different. Based upon our results, heme propionate charge destabilizes functional interaction with the reductase possibly due to like charge repulsion between heme propionate carboxyl residue and an anionic charge on the reductase. In support of this is the decrease in Km for DME-b6 (Fig. 6). Further support includes the observed stimulation of turnover of the reductase with bs in high ionic strength; increases in ionic strength usually impairs electrostatic interactions.

CYTOCHROME

P-450/CYTOCHROME

In all of the studies reported above with DME-b,, native heme insertion into apo-b6 served as a control for insertion of DMEheme insertion. Thus, the observed effects are due to the heme ester and are not due either to a modification of polypeptide structure during DME-bS formation or to an increased conformational heterogeneity in orientation of heme methyl and vinyl groups relative to the protein (37) following heme reconstitution. In brief, the data indicate that in complex formation with cytochrome P-450, the heme of cytochrome bs is in close contact with cytochrome P450. Since both propionate groups were esterified in DME-b5, it is not clear whether one or both carboxyl groups is involved in bs binding to cytochrome P-450. According to an existing model (42), one of the heme propionate groups in ferric cytochrome bs is involved in control of redox potential through charge stabilization of ferric iron leaving the other available for interaction with redox partners. From the difference in Kd of binding of native bs and DME-b5 to RLM5 (Table I), we calculate a Gibbs free energy contribution from the bs heme propionate groups in binding to RLM5 of about 1 kcal mall’, which is of the order expected for a single electrostatic pairing (43). ACKNOWLEDGMENTS This study was supported in part by U. S. Public Health Service Grants GM-26548 and GM-26114 from the National Institutes of Health. REFERENCES 1. OSHINO, N., IMAI, Y., AND SATO, R. (1971) J. Biochem. (Tokyo) 69,155-167. 2. NAGI, M., COOK, L., PRASAD, M. R., AND CINTI, D. L. (1983) J. Biol. Chem. 258,14823-14828. 3. LU, A. Y. H., AND COON, M. J. (1968) J. Biol Chem 243,1331-1332. 4. CHENG, K.-C., AND SCHENKMAN, J. B. (1982) J. Bid Chem 257,2378-2385. 5. VATSIS, K. P., THEOHARIDES, A. D., KUPFER, D., AND COON, M. J. (1982) J. Biol Chem 257,1122111229. 6. Lu, A. Y. H., LEVIN, W., WEST, S. B., JACOBSON, M., RYAN, D., KUNTZMAN, R., AND CONNEY, A. H. (1973) J. Biol. Chem 248,456-460.

b

INTERACTIONS

521

7. JANSSON, I., AND SCHENKMAN, J. B. (1978) Arch. Biochem. Biophys. 185,251-261. 8. BONFILS, C., BALNY, C., AND MAUREL, P. (1981) J. Biol Chem. 256,9457-9466. 9. TAMBURINI, P. P., AND GIBSON, G. G. (1983) J. Biol Chem 258,13444-13452. 10. SUGIYAMA, T., MIKI, N., MIYAKE, Y., AND YAMANO, T. (1982) J. Biochem (Tokyo) 92,1793-1803. 11. IMAI, Y., AND SATO, R. (1977) B&hem. Biophys. Res. Commun. 75,420-426. 12. IMAI, Y. (1981) J. Biochem (Tokyo) 89,351-362. 13. CHIANG, J. Y. L. (1981) Arch. Biochem Biqphys. 211.662-673. 14. WAXMAN, D. J., AND WALSH, C. (1983) Biochemistry 22,4846-4855. 15. BOSTJZRLING, B., TRUDELL, J. R., TREVOR, A. J., AND BENDIX, M. (1982) J. Biol Chem 257,4375-4380. 16. INGELMAN-SUNDBERG, M., AND JOHANSSON, I. (1989) Bicdem. Biophys. Res Cemmun. 97,582589. 17. HLAVICA, P. (1984) Arch Biockem Biophys. 228, 600-608. 18. MORGAN, E. T., AND COON, M. J. (1984) DrzqMefub. D&xx. 12,358-363. 19. NOSHIRO, M., ULLRICH, V., AND OMURA, T. (1981) Eur. J. B&hem. 116.521-526. 20. POMPON, D., AND COON, M. J. (1984) J. Biol Chem. 259,15377-15385. 21. BOSTERLING, B., AND TRUDELL, J. R. (1982) J. Biol. Chem 257,4783-4787. 22. CINTI, D. L., SLIGAR, S. G., GIBSON, G. G., AND SCHENKMAN, J. B. (1979) Biochemistry 18,3642. 23. KUMAKI, K., SATO, M., KON, H., AND NEBERT, D. W. (1978) J. Biol Chem 253,1048-1058. 24. TAMBURINI, P. P., WHITE, R. E., AND SCHENKMAN, J. B. (1985) J. Biol. Chem. 260,4007-4015. 25. YASUKOCHI, Y., AND MASTERS, B. S. S. (1976) J. BioL Chem. 251,5337-5344. 26. SPATZ, L., AND STRITTMA~R, P. (1971) Proc Nati Accd Sci USA 68,1042-1046. 27. CINTI, D. L., AND OZOLS, J. (1975) Biochim Biophgs. Acta 410,32-44. 28. REID, L. S., MAUK, M. R., AND MAUK, A. G. (1984) J. Amer. Chem Sot. 106,2182-2185. 29. NEITER, K. J., AND SEIDEL, G. (1964) J. Pharmacol Exp. T&r. 146.61-65. 30. ULLRICH, V., AND WEBER, P. (1972) Hoppe-Se&r’s J. Physiol Chem. 353.1171-1177. 31. VERMILION, J. L., AND COON, M. J. (1978) J. Biol. Chem 253,2694-2704. 32. OMURA, T., AND SATO, R. (1964) J. Biol. Chem 239, 2379-2385. 33. RAW, I., AND MAHLER, H. R. (1959) J. Biol. Chem. 234,1867-1873. 34. LAEMMLI, U. K. (1970) Nature &mdon) 227,680685.

522

TAMBURINI

AND

35. LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., AND RANDALL, R. J. (1951) J. Bid. Chem 193, 265-275. 36. ESTABROOK, R. E. (1961) in Haematin Enzymes (Falk, J. E., Lemberg, R., and Morton, R. K., eds.), pp. 436-457, Pergamon, Oxford. 37. LAMAR, G. N., BURNS, P. D., JACKSON, T. J., SMITH, K. M., LANGRY, K. C., AND STRITTYAWR, P. (1981) J. BioL Chem 256,6075-6079. 38. MAUK, M. R., REID, L. S., AND MAUK, A. G. (1982) Biochemistr/ 21,1843-1846. 39. MAIJK, M. R., AND MAUK, A. G. (1982) Biochemists 21,4730-4734.

SCHENKMAN 40. DAILEY, H. A., AND STRIWMATTER, Chem 255,5184-5189.

P. (1980) J. Bid

41. BILIMORIA, M. H., AND KAMIN, N. Y Acad Sk 212.428-446.

H. (1973)

42. ARGOS, P., AND MATHEWS, them 250,747-751. 43. CANTO, C. B., AND SCHIMMEL, physical Chemistry, Vol. Francisco.

F. S. (1975)

Ann J. Bi01

P. R. (1980) 1, Freeman,

BioSan

44. JANSSON, I., TAMBURINI, P. P., FAVREAU, L. V., AND SCHENKMAN, J. B. (1985) Drug Metab. Dispos. 13.453-458.