THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 239, No. 7, July 1964

Printed in U.S.A.

The Carbon Monoxide-binding Pigment of Liver Microsomes

I. EVIDENCE FOR ITS HEMOPROTEIN NATURE

TSUNEO OMURA AND RYO SATO

From the Institute for Protein Research, Osaka University, Osaka, Japan

(Received for publication, January 3, 1964)

The presence in mammalian liver microsomes of a carbon monoxide-binding pigment has been reported by Klingenberg (1) and by Garfinkel (2). The CO compound of the reduced pigment has an intense absorption band at 450 rnl.c and thus can be readily detected in dithionite-treated microsomes by difference spectrophotometry. The CO difference spectrum of reduced microsomes is, however, unusual in that it shows no peaks other than that at 450 rnp and, therefore, provides no clue to the nature of the pigment. The elucidation of its nature has further been hampered by the reported lability of the microsomal pigment to detergents, low pH, and enzymatic digestion (1,2). In addition, the CO compound has been reported as not photodissociable (1).

In preliminary communications (3, 4), we have reported evidence for the hemoprotein nature of the microsomal CO- binding pigment, provisionally called “P-450,” and shown that it can be converted into a solubilised form, which we term “P- 420,” by treatment of microsomes anaerobically with snake venom or deosycholate. The solubiliaation is accompanied by an unusual change in the spectral properties of the pigment. Further, the solubilized pigment has been partly purified, free from cytochrome 55, and shown to possess absorption spectra characteristic of hemoproteins (5). The present paper gives a detailed account of the investigations on rabbit liver microsomes and crude microsomal digests, which have led us to postulate the hemoprotein nature of the pigment. Purification and properties of the solubilized hemoprotein will be reported in the accom- panying paper (6).

EXPERIMENTAL PROCEDURE

Mtioscnnal Preparation-s-The microsomes used were pre- pared from rabbit liver by a modification of the method of Mitoma et al. (7). Male rabbits (2.5 to 3 kg) were starved for 24 hours and killed by injecting air into the ear vein. The thoracic cavity was opened, and the liver was thoroughly per- fused in situ with more than 200 ml of 0.9% NaCl solution. The liver was excised, finely chopped with a razor, and homog- enized with 4 volumes of isotonic (1.150/0) KC1 solution in a Potter glass homogenizer equipped with a Teflon pestle. The homogenate was centrifuged at 12,000 x g for 25 minutes in a refrigerated centrifuge, and the precipitate was discarded. The microsomes were sedimented by centrifugation at 78,000 x g for 90 minutes in a Hitachi model 40P preparative ultracen- trifuge. The firmly packed pellet of microsomes was resus- pended in isotonic KC1 solution with the Potter homogenizer and again centrifuged as above. The washed microsomes were finally suspended in isotonic KCl, usually at a concentration of 10 mg of protein per ml. The resultant microsomal suspensions

were stored at 4” and used within 2 to 3 days. In these prepara- tions isotonic KC1 (ionic strength, about 0.15) was employed, instead of the more usual 0.25 M sucrose, so as to minimize the adsorption of hemoglobin on microsomes (8). The microsomal preparations thus obtained were found to be practically free of adsorbed hemoglobin, when examined by zone electrophoresis as described by Paigen (8). AS shown below, the absence of hemoglobin could be confirmed further spectrophotometricahy.

Measurements of Di$erence Spectra-Difference spectra of microsomal preparations were measured in a Cary model 14 spectrophotometer with cuvettes of l-cm optical path. JIicro- somal preparations, usually containing 2 mg of protein per ml of 0.1 M phosphate buffer (pH 7.0), were placed in both the sample and reference cells. After recording the base-line, the content of sample cell was treated with various reagents, and the spectral difference thereby induced was measured. When CO was used, it was carefully bubbled through the sample for about 20 seconds; this was sufficient to saturate the sample with the gas. Reduction of samples with dithionite was effected with a few milligrams of solid NazS204. Thunberg-type cuvettes, also of l-cm optical path, were employed when it was necessary to keep the sample anaerobic, or under a gas mixture of definite composition. Anaerobic conditions were established simply by evacuating the cell, or by replacing the gas phase with osygen- free nitrogen. In some experiments the sample and reference cells were treated differently, and the difference spectrum was measured. All spectrophotometric measurements were made at room temperature (20-25”). Further details are given in legends of figures.

Photodissociation of CO Compound of P-QO-Microsomes suspended in 0.1 M phosphate buffer, pH 7.0, were placed in a l-cm square cell and reduced by a few milligrams of solid sodium dithionite. A suitable amount of water saturated with CO was added to the suspension to convert reduced P-450 to the CO compound. The final concentration of CO was calculated from the solubility of CO in water at 20”. The sample cell was then placed in the sample compartment of a Cary model 14 spec- trophotometer and illuminated from the side with a IOO-watt tungsten lamp through a set of a condenser lens and a red glass filter (Toshiba R-52). The filter transmitted only red light of wave lengths longer than 520 rnp. An interference filter with a sharp transmittance maximum at about 450 rnp was inserted between the sample cell and the measuring phototube to absorb the red light and to pass the light from the monochromater (450 mr). The reference cell containing the same suspension was placed in the reference cell compartment without illumina- tion. The photodissociation was measured by following the

2370

by guest on March 8, 2020

http://ww

w.jbc.org/

Dow

nloaded from

July 1964 T. Omura and R. Sate 2371

absorbance change at 450 rnM caused by illumination of the sample with the red light. Control esperiments using reduced microsomes were carried out without CO addition to eliminate the effect of any red light leaking through the blue interference filter or blue light leaking through the red glass filter.

Solutilization Treatments-Solubilisation of the microsomal CO-binding pigment and cytochrome bs was usually effected with venom of the snake, Trimeresurus jZavo&icZis, treated with heat as described previously (9). A mixture containing microsomes (3 to 6 mg of protein per ml), 0.1 M Tris buffer (pH 8.5), and 0.1% (in terms of unheated venom) of heated venom was incubated anaerobically in a Thunberg tube at 4” for 15 to 20 hours. Pro- longed incubation up to 48 hours was also carried out, with results similar to those obtained on shorter incubation. Un- heated snake venom was equally effective in solubilizing the pigments. The hemoproteins could also be effectively solubilieed by addition of 0.5% (final concentration) of sodium deoxycholate to the microsomal suspensions in 0.1 M Tris or phosphate buffer under anaerobic conditions (pH 8.0) ; solubilization was maximal within 10 to 20 minutes at room temperature. Both solubiliza- tion treatments resulted in liberation of 70 to 90% of the micro- somal protein in a form not sedimentable by centrifugation at 105,000 x g for 60 minutes. Both the CO-binding pigment and cytochrome ba appeared almost quantitatively in the super- natant. As described below, the CO-binding pigment thus solubilized (P-420) possessed spectral properties considerably different from those of the microsomal bound form (P-450).

Analytical Procedures-The protoheme content of microsomes was measured spectrophotometrically after converting the heme into pyridine-hemochromogen in the presence of 0.1 N NaOH and 20% pyridine. In computing the concentration, a value of 32.4 cm-1 mr0 was used for the difference in molar extinction between 557 and 575 rnp in the dithionite-reduced minus oxidized difference spectrum of the hemochromogen. This value was obtained from the molar extinction coefficient of reduced pyridine protohemochromogen (Esb7 m,, = 34.7 cm-1 rnK1) reported by Paul, Theorell, and Akesson (10) and the dithionite-reduced minus oxidized difference spectrum of pyridine-hemochromogen prepared from crystalline hemin chloride. Cytochrome bs was determined from the difference spectrum between NADH-re- duced and air-saturated microsomes (Fig. 7, Curve A). As will be shown, in NADH-reduced microsomes P-450 was mostly in the oxidized form whereas cytochrome bs was in the reduced form. The increment of molar extinction between 424 and 409 rnp in this difference spectrum was assumed to be 185 cm-1 rnM-1. This value was obtained with a purified preparation of cyto- chrome b6 isolated in this laboratory’ and is different from those reported previously (1). Total iron was analyzed colorimetri- tally with o-phenanthroline as the reagent after digesting the samples with boiling sulfuric acid (11). Protein was determined by the method of Lowry et aZ. (12) with bovine serum albumin as the standard.

ChemicaZsNAD and NADP were purchased from the Sigma Chemical Company. NADH was prepared from NAD by the procedure described by Bonnichsen (13), but yeast alcohol dehydrogenase, purified according to Racker (14), was used instead of the liver enzyme. NADPH was prepared from NADP by the isocitrate dehydrogenase method (15). CO was pre- pared from formic acid and concentrated sulfuric acid by the con-

1 K. Hirota and T. Omura, unpublished observations.

ventional method, and purified by bubbling through a KOH solution. Ethyl isocyanide was synthesized from silver cyanide and ethyl iodide and redistilled twice (16). Lyophilized venom of Trimeresurus jZavoviridis collected in 1960 on Amami Oshima Island, Kyushu, Japan was used.

RESULTS

Carbon Monoxide Diference Spectra of Alicrosomes-The addition of dithionite to liver microsomes produced, as previously observed (1, 2), a spectral change that could be mostly, if not entirely, accounted for by the reduction of cytochrome b6 and possibly also of flavins. There was scarcely any sign of the presence of any other reducible pigment. In confirmation of previous work (1,2), however, a difference spectrum with a peak at 450 rnp appeared when CO was bubbled into the dithionite- treated microsomes (Fig. 1, Curve A) indicating the presence of P-450, the microsomal CO-binding pigment. Unlike usual hemoprotein compounds, this pigment showed no peaks in the CO difference spectrum between 500 and 650 rnp; the only feature in this region was a broad and shallow trough around 500 rnp. The addition of CO to air-saturated microsomes, on the other hand, induced no significant change in spectrum (Fig. 1, Curve B) indicating that P-450 was a reducible pigment and only the reduced form could bind CO (cf. (2)).

The spectra shown in Fig. 1 indicated further that the micro- somes used were virtually free from hemoglobin and methemo- globin, because neither peak nor shoulder could be seen at 420 rnp where a sharp absorption maximum of carboxyhemoglobin was to be expected. Because cytochrome bs does not combine with CO (17), it was evident that P-450 was not this cytochrome.

The peak at 450 rnp was lowered gradually when the dithionite- treated sample in a CO-filled cuvette was evacuated, but resumed the original height on readmission of CO, which indicated revers- ibility of CO binding. The reaction of CO with reduced micro- somes was slow (1); a preliminary value for the first order velocity constant was determined to be 5 mm-‘. This may be compared with a t+ value of 10 seconds (corresponding to a velocity constant of 4.2 min-l) reported for the CO-binding reaction of rat liver microsomes (1).

As shown in Fig. 2, the CO 450-rnp band intensity diminished

+ 0.3

0.2 >

3 s 0.1

&z 2 0.0

2 0.1

- 0.2

I 400 450’ 500 550 600 650

WAVE LENGTH my)

FIG. 1. Carbon monoxide difference spectra of liver micro- somes. Curve A, dithionite-reduced microsomes. Both the sample and reference cells contained a microsomal suspension (2 mg of protein per ml, 0.1 M phosphate, pH 7.0) that had been treated with dithionite. The content of the sample cell was then saturated with CO, and the difference spectrum was measured after several minutes. Curve B, aerobic microsomes. The differ- ence spectrum was measured in the absence of dithionite.

by guest on March 8, 2020

http://ww

w.jbc.org/

Dow

nloaded from

2372 Microsomal CO-binding Pigment. I

4 0.2 s > Y s 0: 0.1 5: z

Q 0.0 I I - I

107 IO- IO-'

CO CONCENTRATION (MI

FIG. 2. Effect of carbon monoxide concentration on formation of carbon monoxide compound of P-450. The CO difference spec- trum of dithionite-reduced microsomes was observed under the same conditions as specified for Curve A of Fig. 1, except that the final CO concentration was varied as indicated. The CO concen- tration was varied by the addition of a varying amount of water saturated with CO. Dilution effect caused by the addition of aqueous CO solution was corrected.

on decreasing the concentration of CO in the medium. It was determined from these data that the concentration of unbound CO in the presence of half-saturated pigment was 3 NM. Assum-

ing a simple dissociation equilibrium, this value could be regarded as the dissociation constant of the CO compound.

As reported by Garlinkel (2), addition of cyanide (1 mM) to dithionite-reduced microsomes gave a small absorption peak around 440 mp. However, in contrast to Garfinkel (2), cyanide showed no observable effect on the CO difference spectrum.

Efect of Light on CO Compound of P-&O-In agreement with Klingenberg’s observation (l), it was not possible in the present investigation to observe any photodissociation of the CO com- pound of microsomal P-450. The final concentration of CO was varied from 0.1 mM to as low as 5 PM, a concentration in which P-450 was only about 50% saturated with CO (Fig. 2).

Under all these conditions, however, no detectable decrease was observed in absorbance at 450 rnp on illumination of samples with the red light of wave lengths longer than 520 rnp. On the other hand, the same illumination device, with an interference filter having a transmittance maximum at 420 rnM, was sufficient to demonstrate the reversible photodissociation of CO-hemo- globin in hemolyzed rabbit erythrocytes. In this case, the extent of dissociation reached more than 30% when the concentration of CO was lowered to a level where 70% of hemoglobin was in combination with CO in the dark. The estent of photodissocia- tion became less as the CO concentration was increased.

Ethyl Isocyanide Deference Spectra of Microsomes-Fthyl iso- cyanide combines reversibly with hemoglobin and myoglobin to give characteristic spectra (18). When this reagent was added to dithionite-reduced microsomes, a spectral change as shown in Fig. 3 occurred, indicating the presence of an ethyl isocyanide- biding pigment. In contrast to the corresponding CO spec-

trum, the isocyanide spectrum exhibited features characteristic of hemochromogens and suggested the hemoprotein nature of the pigment involved. This pigment was again different from cytochrome bg; purified cytochrome bs showed no affinity for ethyl is0cyanide.l

The isocyanide spectrum, although characteristic of hemochro-

mogens, was unusual in that it showed a small peak at 455 rnF in addition to typical LY, ,8, and Soret peaks at 560,530, and 430 rnp, respectively. As seen in Fig. 4, however, the bands at 430 and 455 mp diminished in the same proportion when the con- centration of ethyl isocyanide was decreased. It was therefore concluded that one hemoprotein was responsible for the two absorption peaks,

Fig. 4 also permitted an estimation of the affinity of ethyl

+ 0.4

> 0 0.3

f m 0.2

i 0.1

4:

f

t

0.0 _ I

J -0.q J

I 8 400 450 500 550 600 650

WAVE LENGTH (mpl FIG. 3. Ethyl isocyanide difference spectrum of liver micro-

somes. The conditions were the same as specified for Curve A of Fig. 1, except that ethyl isocyanide (final concentration, 1 mM), instead of CO, was added to the reduced microsomes in the sample cell.

0.4

+ 0:3

> 0.2

z a z

0.1

i 0.0

a

0.1

- 0.2

A- 6 _-----

0 -.- .- .- .

fJ - . .._._

400 450 500 WAVELENGTH (mp)

0.4

+ 0.3

‘0 0.2 z

: LT 0.1 0

3 0.0

- 0.1

0.2

I --

400 450 500

WAVELENGTH (mp)

FIG. 4 (left). Effect of ethyl isocyanide concentration on difference spectrum. The conditions were the same as specified in the legend of Fig. 3, except that the final concentration of ethyl isocyanide was varied from 100 PM to 1 PM. Curve A, 100 PM; Curve B, 10 FM; Curve C, 3 PM; Curve D, 1 PM.

FIG. 5 (right). Competition between carbon monoxide and ethyl isocyanide for P-450. Both the sample and reference cells con- tained a microsomal suspension (2 mg of protein per ml, 0.1 M phosphate, pH 7.0) that had been reduced by dithionite. The content of the sample cell was saturated with CO by bubbling the gas for 2 minutes, and then treated with ethyl isocyanide. The difference spectrum was measured after about 10 minutes. During this equilibration time, CO was occasionally bubbled into the sample cell to secure the saturation of the gas. The final concentrations of ethyl isocyanide used were as follows: Curve A, 0 (CO difference spectrum); Curve B, 0.3 mm; Curve C, 3 mM; Curve D, 1 mM (ethyl isocyanide difference spectrum). Curve D was measured in the absence of CO. The concentration of CO in the saturated solution was assumed to be 1 m&r.

by guest on March 8, 2020

http://ww

w.jbc.org/

Dow

nloaded from

July 1964 T. Omura and R. Sato 2373

isocyanide for the microsomal pigment. Thus, the concentra- tion of unbound isocyanide in the presence of half-saturated pigment was determined to be 1 PM, a value smaller than that reported for hemoglobin (19).

Competition between Carbon Monoxide and Ethyl Isocyanide- Fig. 5 and Table I indicate that there was competition between CO and ethyl isocyanide for dithionite-reduced microsomes. Hence, it was concluded that the microsomal CO- and ethyl- isocyanide-binding pigments were identical with each other. This, in turn, suggested that P-450 was a hemoprotein, although the spectrum of its CO compound was unusual for a hemoprotein.

The competition data, coupled with the dissociation constant of the isocyanide compound, permitted further an estimation of t,he affinity of CO for reduced P-450. The dissociation constant of the CO compound of P-450 was thus determined to be 2 PM.

This value was in agreement with that obtained above from the effect of CO concentration on the development of the 450-rnp band (Fig. 2).

Reduction and Oxidution of P-QO-Microsomal P-450 was reported to be reduced incompletely by NADH, based on partial appearance of the 450-rnp band of CO (1,2). In a CO-O2 atmos- phere (80:20, by volume), the 450-rnp band of NADH-treated microsomes was in fact only 10 to 30% as high as that attainable by dithionite treatment. As seen in Fig. 6, however, the band could be intensified to the dithionite level if the gas phase was replaced by pure CO (or a CO-N2 mixture). When the gas phase was again replaced by the CO-O2 mixture, the intensified band returned quickly to the low level. Similar changes were also observed in the presence of NADPH. These findings suggested that P-450 was in fact fully reducible by both NADH and NADPH under anaerobic conditions, but was rapidly re- oxidizable on exposure to oxygen.

Experiments with CO-02 mixtures of varying composition further indicated that at a given partial pressure of oxygen a steady state was established with regard to the reduction and reoxidation of P-450. However, the extent of reduction in a steady state could not be determined accurately, because the

reduction of P-450 was detectable only indirectly by the absorp- tion of its CO compound which was in equilibrium with the reduced pigment. In view of the very high affinity of CO for

TABLE I

Competition between carbon monoxide and ethyl isocyanide for microsomal P-&O

The experimental conditions were the same as specified in the

legend of Fig. 5. The four experiments recorded correspond to Curves A, B, C, and D in Fig. 5. The concentration of CO was

calculated from its solubility in water. The percentages of P-450 in the forms of CO and ethyl isocyanide compounds were calculated from the absorbance increments at 450 and 429 m, respectively.

Reagent concentration Absorbance increment state of P-450 as

I I

7nH %

1 0 0.372 -0.048 100 0

1 0.3 0.250 0.149 56 41

1 3 0.146 0.355 18 83 0 1 0.096 0.439 0 100

z I I 0 2 4 6 d IO 12

TIME (min)

FIQ. 6. Effect of oxygen on carbon monoxide difference spec- trum of NADH-reduced microsomes. Each 3 ml of a microsomal suspension (2 mg of protein per ml, 0.1 M phosphate buffer, pH 7.5) was placed in a Thunberg-type sample cell and in an open reference cell. The suspension contained 1 mM KCN to prevent NADH oxidation by contaminating mitochondria, if any. The side arm of the sample cell received 0.05 ml of 10 mM NADH. The sample cell was first filled with a CO-02 mixture (80:20, by volume). At Arrow 1, 0.05 ml of 10 mM NADH was added to the reference cell, and, at the same time, the NADH was tipped from the side arm into the microsomal suspension in the sample cell. At Arrow 2, the gas phase of the sample cell was replaced by pure CO, and then at Arrow 9, the gas phase was again replaced by the CO-O2 mixture (80:20). The absorbance change during these treatments was measured.

+ 0.3- A-

> 0 0.2 - 6 -----

5 g 0.1 -

i2 * 0.0 - a

- 0.1

400 450 500 550 600 650

WAVELENGTH (my I

Fm. 7. NADH-reduced minus oxidized and dithionite-reduced minus NADH-reduced difference spectra of liver microsomes. The microsomal suspension contained 2 mg of protein per ml in 0.1 M phosphate, pH 7.0, and was saturated with air before the addition of NADH or dithionite. The final concentration of NADH was 0.2 mM. Curve A, NADH-reduced minus oxidized difference spectrum; Curve B, dithionite-reduced minus NADH- reduced difference spectrum.

reduced P-450, the true extent of reduction may not be very large as compared with the value estimated from the CO band, e.g. 10 to 30% in the presence of 20 volume y. of oxygen.

In contrast to P-450, cytochrome ba in microsomes could be reduced almost fully by both NADH and NADPH and main- tained in the reduced state even in the presence of dissolved

oxygen (20). Cytochrome bb was also reduced partly by ascor- bate and cysteine, whereas these reagents failed to reduce P-450 even in the absence of oxygen. This was in contrast to Gar- finkel’s finding (2) and suggested an oxidation potential for P-450 lower than that of cytochrome bs.

Spectral Difwence between Dithionite- and NADH-reduced Microscmes-As stated above, the addition of NADH to micro- somes under aerobic conditions resulted in almost full reduction of cytochrome b5, but left P-450 mostly in the oxidized state. Accordingly, the spectral change thereby induced (Fig. 7, Curve

by guest on March 8, 2020

http://ww

w.jbc.org/

Dow

nloaded from

2374 Microsomal CO-binding Pigment. I Vol. 239, No. 7

A) was nearly identical with the reduced minus oxidized differ- ence spectrum of purified cytochrome bs. On addition of dithi- onite, on the other hand, both cytochrome bs and P-450 were fully reduced. The spectral difference between dithionite- and NADH-reduced microsomes, therefore, should have pro- duced a spectrum attributable to the difference between the reduced and oxidized forms of P-450. Curve B in Fig. 7 shows the difference spectrum thus obtained. This spectrum closely resembled that reported for rat liver microsomes (1) and showed two broad peaks around 550 and 440 rnp. Although Klingenberg (1) attributed these peaks to the presence of methemoglobin, the microsomes used here were free from both hemoglobin and methemoglobin as already mentioned. It was more likely that this spectrum was equivalent, at least partly, to the spectral difference between the reduced and oxidized forms of P-450. It was, however, rather ambiguous in shape and therefore provided little information concerning the nature of the pigment.

Spectral Changes Induced by Solubilization Treatments--It was reported that treatment of microsomes with bile salts, digitonin, or pancreatin either destroyed or greatly decreased the CO- binding capacity (1, 2). Klingenberg (1) further observed a very slow formation of a “CO-protohemochromogen” on addition of CO to cholate-treated microsomes. Such lability of the CO- binding capacity to solubilization treatments was confirmed in the present study. Thus, no peak appeared at 450 rnp in the CO-difference spectrum of microsomes after they had been treated with 0.5% sodium deoxycholate (pH 8.0,lO minutes, 20”) or with 0.1% heated venom of T. $uvoviridis (pH 8.5, 15 hours, 4”). However, it was noticed that the treated microsomes showed a feeble CO difference spectrum with a small peak at 420 rnp and a trough at 435 rnp. This CO spectrum might have been the same one which Klingenberg had attributed to a “CO- protohemochromogen,” although its formation was not very slow.

It was found that this feeble CO spectrum was greatly in- tensified by carrying out the solubilization treatments under

0.4

+ 0.3 I

g 0.2

5 m 0.1 a

g 0.0 m I

400 450 500 550 600 650

WAVE LENGTH (mp)

FIG. 8. Carbon monoxide difference spectra of venom-digested and untreated microsomes. Liver microsomes were treated anaerobically with heated snake venom as described under “Ex- perimental Procedure,” and the digest was diluted 3-fold with 0.15 M phosphate, pH 7.0, to give a protein concentration of 2 mg per ml. The diluted digest was placed in two Thunbern-tvne v “I cuvettes, one serving as the sample and the other as the reference. The sample cell was filled with pure CO, and the reference cell with pure Nz. Then the digest in both cells was reduced with dithi- onite anaerobically, and the difference spectrum was measured. The spectrum of untreated microsomes is the same with Curve A of Fig. 1. Curve A, untreated microsomes; Curve B, venom- digested microsomes.

0.3 -

>

z 0.2 -

m

f5 0.1 - fn

: 0.0 --

- 0.1

t

A- 6 -----_

p> 0 -.-.-._

I I

400 450 500

WAVELENGTH (my)

FIG. 9. Changes in carbon monoxide difference spectrum of microsomes during digestion with heated snake venom. Micro- somes (2 mg of protein per ml) in 0.1 M Tris-acetate buffer, pH 8.5, were digested with 0.05% (final concentration) of heat-treated snake venom at 0” under pure N2. At desired time, a portion of the digest was withdrawn, and the CO difference spectrum was measured with the procedures described in the legend of Fig. 8. Curve A, 1 minute after addition of heated venom; Curve B, 5 minutes; Curve C, 20 minutes; Curve D, 69 minutes.

anaerobic conditions. A still further intensification of the spec- trum occurred when anaerobically treated microsomes were first saturated with CO and then reduced by dithionite, as com- pared with the sample which had been first reduced by dithionite and then subjected to CO bubbling. Curve B in Fig. 8 shows the CO difference spectrum, measured with the precautions cited, of the microsomes digested anaerobically with heated snake venom. The corresponding spectrum of untreated microsomes (Curve A) is included in the figure to emphasize the profound spectral change induced by the solubilization treatment. As can be seen, the CO spectrum of digested microsomes showed, in addition to the intense peak at 420 rnp and the deep trough at 435 rnp, two small peaks at 540 and 575 rnp and a trough at 560 rnl.r. It was clear from the shape of spectrum that a hemo- protein was responsible for the CO binding in digested micro- somes. This hemoprotein was called “P-420.”

P-490 as a Solub-ilized Form of P-@O-Although the CO spec- trum of digested microsomes resembled closely that of hemo- globin, it was unlikely that the microsomes employed were con- taminated by detectable amounts of the blood pigment, as repeatedly mentioned. The possibility was also excluded that P-420 was a denatured product of cytochrome b6, because a purified preparation of the cytochrome acquired no CO-binding capacity even after prolonged treatment with heated venom.1 Instead, it seemed very likely that P-420 had been formed at the expense of P-450, because the emergence of the 420-ml peak in the CO spectrum was accompanied by the disappearance of the 450-rnp peak. This possibility was in fact confirmed by a quantitative experiment shown in Fig. 9 and Table II. It may be seen that the amount of P-450 which disappeared in the course of the venom treatment was always proportional to that of P-420 which appeared. This could be explained by assuming that P-420 was a product of P-450, although their spectral proper- ties differed considerably from each other. Similar parallelism was also observed with deoxycholate as the solubilizing agent.

by guest on March 8, 2020

http://ww

w.jbc.org/

Dow

nloaded from

July 1964 T. Omura and R. Sato 2375

When the microsomal digests were centrifuged at 105,000 x 9 for 60 minutes, both P-420 and cytochrome bg were recovered in the supernatant almost quantitatively. This indicated further that P-420 was in fact a solubilized form of P-450.

Properties of P-@O-As suggested by the intensification of the 420-rnp band of CO by solubilization under anaerobic condi- tions, the microsomal CO-binding pigment was labile when solubilized in the presence of oxygen. But once solubilized anaerobically and converted to P-420, it became more stable to aeration. P-420 was, however, extremely labile to oxygen in the presence of dithionite. Anaerobic conditions were, therefore, required for the reduction of P-420. For this reason, the CO spectrum of anaerobically digested microsomes had been in- tensified by first saturating the sample with CO and then reducing by dithionite.

P-420, when anaerobically reduced by dithionite, combined also with ethyl isocyanide to give a difference spectrum illustrated in Fig. 10. This spectrum with an intense Soret peak at 433 rnp again differed substantially from the corresponding spectrum of P-450 (Fig. 3). The solubilized pigment in microsomal digests could be reduced slowly by both NADH and NADPH under anaerobic conditions, as evidenced by the appearance of the 420-rnp band in the presence of CO. Like P-450, however, the reduced pigment could be very rapidly reoxidized by molecular oxygen. Only a small portion of P-420 seemed to be in the reduced form in the presence of air. Cytochrome ba in the digests, on the other hand, could be rapidly and fully reduced by reduced pyridine nucleotides and maintained in the reduced state even in the presence of oxygen.

Cytochrome Nature of P-&O-As in the case of P-450, the difference in the oxidation-reduction behaviors of P-420 and cytochrome ba in microsomal digests permitted the measurement of a reduced minus oxidized difference spectrum of P-420 itself. When digested microsomes were kept anaerobic in the presence of dithionite, both P-420 and cytochrome bs were fully reduced. When NADH was added to the preparation under aerobic con- ditions, only cytochrome ba was fully reduced and P-426 re- mained mostly oxidized. The spectral difference between these two samples should, therefore, have been equivalent to a reduced minus oxidized difference spectrum of P-420, ignoring the con-

TABLE II

Conversion of P-&O into P-.@?O by treatment of microsomes with heated snake venom

The experimental conditions were the same as specified in the legend of Fig. 9. Four out of the six experiments recorded corre- spond to Curves A, B, C, and D in Fig. 9. The percentages of CO-binding pigment in the forms of P-456 and P-420 were calcu- lated from the absorbance increments at 459 and 420 q, respec- tively.

Incubation time

-

I

min

0 1

5 10 20 60

Absorbance increment in CO difference spectrum

450 mp

0.290 0.256

0.151 0.096

0.640 -0.006

-

--

-

420 mp

-0.110 -0.045

0.133 0.177 0.248

0 341

- I State of CO-binding

pigment as

P-450 P-420

%

199 0 88 14 52 53

33 64

14 80

0 100

400 450 500 550 600 650 WAVE LENGTH (mp)

1

- 0.1

400 450 500 550 600 650 WAVELENGTH ( my)

FIG. 10 (upper). Ethyl isocyanide difference spectrum of venom- digested microsomes. The microsomal digest described in the legend of Fig. 8 was diluted a-fold with 0.2 M phosphate, pH 7.0, to give a protein concentration of 3 mg per ml. The diluted digest was placed in two Thunberg-type cells, and, after replacing the gas phase by pure Nz, the contents of both cells were reduced anaerobically by dithionite. Ethyl isocyanide (final concentra- tion, 1 mM) was then added from the side arm into the digest in the sample cell, and the difference spectrum was measured after a few minutes.

FIG. 11 (lower). Reduced minus oxidized difference spectrum of P-420. Venom-digested microsomes as used in the experiment described in Fig. 10 (final protein concentration, 3 mg per ml) was placed in a Thunberg-type cell and reduced by dithionite under pure Nz. On the other hand, the digest in an open cuvette was treated with NADH (final concentration, 0.1 mM) in the presence of air. The difference spectrum between the former and the latter samples was measured after a few minutes.

tribution of flavins to the spectrum. The resultant difference spectrum, shown in Fig. 11, was that of a typical hemochromogen with (Y, 8, and Soret peaks at 560,530, and 427 rnp, respectively, and closely similar to that of bacterial cytochrome bl (21, 22). This finding, together with the fact that all the heme in micro- somes could be extracted with acid-acetone; strongly suggested that this pigment was a cytochrome of the b type.

Composition of Microsomes-If the CO-binding pigment is a hemoprotein other than cytochrome bg as suggested above, liver microsomes should contain more heme than can be explained by cytochrome bs. Actually, previous workers (1, 2, 23) reported the presence of such extra heme in liver microsomes. To con- firm this, the microsomes employed in the present study were analyzed for several components and the results are recorded in Table III. As will be seen, the total heme content of the micro- somes was in fact more than twice as high as that of cytochrome

2 T. Omura, unpublished observations.

by guest on March 8, 2020

http://ww

w.jbc.org/

Dow

nloaded from

2376 Microscrmal CO-binding Pigment. I Vol. 239, No. 7

TABLE III

Composition of rabbit liver microsomes

Average of 40 determinations on microsomes obtained from different rabbits. Figures indicate the contents in millimicro- moles per mg of microsomal protein.

Component content

Protoheme....................... Cytochrome bg................... P-450............................ Total iron.

* Millimicroatoms per mg of protein.

2.55 1.12

1.55 140*

bs on molar basis. From the alkaline pyridine-hemochromogen spectrum, it was further concluded that protoheme was the only heme present in the microsomes. The content of P-450 was determined from the CO difference spectrum of dithionite- reduced microsomes, with a molar extinction difference of 91 cm-1 m~d-1 between 450 and 490 rnp reported for the pigment (5, 6). It is of interest to note that the content of P-450 was roughly equal to that of protoheme which was present in excess of the amount associated with cytochrome bs. It was also noticed that the contents of total heme, cytochrome bg, and P-450 in different preparations were fairly constant. Finally, the rnicro- somes contained large amounts of nonheme iron, the content of which varied considerably from one preparation to the other. It is probable that most of the nonheme iron is due to the pres- ence of ferritin in the microsomal preparations (24).

DISCUSSION

The results reported in this paper suggest that the CO-binding pigment (P-450) in liver microsomes is a hemoprotein with unusual properties. In the reduced form it combines not only with CO but also with ethyl isocyanide, although the absorption spectra of these addition compounds are somewhat atypical for hemoprotein derivatives. P-450 can be converted into a sol- ubilized form (P-420) by treatment of microsomes anaerobically with heated snake venom or deoxycholate. Unlike P-450, the solubilized form behaves spectrophotometrically as a typical hemoprotein.

As the CO difference spectrum of P-420 (Fig. 8, Curve B) is very similar to that of hemoglobin, it might be attributed to contamination by the blood pigment. However, the absence of hemoglobin and methemoglobin in the microsomes used can be concluded from zone electrophoresis and difference spectra (Fig. 1) of the preparations. Further, the difference spectrum that can be ascribed to the difference between the reduced and oxi- dized forms of P-420 (Fig. 11) is not like those of hemoglobin (hemoglobin minus oxyhemoglobin and hemoglobin minus methemoglobin). The possibility that P-420 is a denaturation product of cytochrome & can also be ruled out, because purified cytochrome bs acquired no CO-binding capacity even after prolonged treatment with heated venom,’ and the microsomal content of NADH-reducible cytochrome b6 did not change detectably during the solubilization treatment.* Therefore, it may be concluded that P-420 (hence also P-450) is a new hemo- protein differing from any of the known hemoproteins and their derivatives. This was further confirmed by purification of P- 420 as described in the accompanying paper (6).

The absorption spectra of P-450 are uncommon for a hemo-

protein and differ considerably from those of P-420. The fact that the conversion of P-450 into P-420 is effected by solubilizing agents capable of attacking phospholipids suggests that micro- somal phospholipids may be responsible for the peculiar spectral properties of P-450. It seems that P-450 is embedded in micro- somal lipids. However, further studies are needed before one can be sure of the state of P-450 in microsomes. Conversely, such unusual spectra of P-450 seem to provide an interesting problem in the physical chemistry of hemoproteins. No hemo- proteins have so far been shown to change their spectra consid- erably on liberation from the structures to which they are bound in viva.

Other variant hemoproteins are known. An example is the Rhodospirillum rubrum hemoprotein (25)) which contains 2 moles of heme per mole and is related to a cytochrome of the c type (26,27). Unlike usual c-type cytochromes, reduced RHPa shows a spectrum with only a broad band in the cr- and P-band region, and the Soret band of the oxidized form is not markedly in- tensified on reduction. In the reduced form it binds CO and this binding results in a marked intensification of the Soret peak. Under alkaline conditions, this protein shows profound changes in its spectra and the reduced form shows a spectrum character- istic of c-type cytochromes. Cryptocytochrome c, recently purified and crystallized from Pseudomonas &nitriIcans, also has properties similar to those of RHP (28).

Klingenberg (1) reported the failure of his attempts to induce photodissociation of the microsomal CO addition compound. In the present study it was also unsuccessful to observe the photo- dissociation of the CO compound of P-450 under conditions in which light showed a detectable effect on CO-hemoglobin. Since all the CO compounds of hemoproteins so far known, including that of RHP, are dissociable in the light, these findings might be regarded as evidence against the hemoprotein nature of P-450. However, all the other evidence reported in this paper seems to indicate rat,her decisively that the pigment is in fact a hemopro- tein and, as will be reported later (6), a highly purified prepara- tion of its solubilized form (P-420) did yield protoheme on treat- ment with acid-acetone. It seems, therefore, possible that the photodissociation can be detected by improving the experi- mental conditions such as light intensity. Recently, Estabrook, Cooper, and Rosenthal (29) have reported that CO inhibits the steroid C-21 hydroxylation reaction in adrenal cortex microsomes and that this inhibition is reversed by light. They have further found that the photochemical action spectrum obtained by examining the effect of light of various wave lengths in reversing the CO inhibition shows a maximum at 450 ml.c. Since the presence of P-450 in adrenal cortex microsomes has been con- firmed (29), these findings indicate the photodissociability of the CO compound of adrenal P-450, as well as the involvement of the pigment in the steroid C-21 hydroxylase system.

The microsomal CO-binding pigment is labile when solubilized under aerobic conditions, and the solubilized pigment (P-420) is extremely unstable to oxygen in the presence of dithionite. Such instability of the pigment appears to be related to aerobic peroxidation of microsomal unsaturated lipids (30). It is likely that the peroxidation is accompanied by oxidative decomposition of the heme moiety of P-420. It was in fact found that P-420, when further purified, becomes more stable to oxygen even in the presence of dithionite, probably owing to removal of as-

s The abbreviation used is: RHP, Rhodospirillum rubrum hemoprotein.

by guest on March 8, 2020

http://ww

w.jbc.org/

Dow

nloaded from

July 1964 T. Omura and R. Sato 2377

sociated lipid constituents (5). In contrast to P-420, cytochrome ba is not affected by oxygen under all the conditions examined. This appears to reflect the difference in the state of protoheme in the two microsomal hemoproteins.

It confirmation of previous work (1, 2, 23), it was found that liver microsomes contained protoheme in excess of the amount associated with cytochrome b5. As is evident from Table III, it seems that most of the extra heme is present in the form of the CO-binding pigment. There is no spectral evidence sug- gesting the occurrence of any hemoproteins other than cyto- chrome bS and P-450. The microsomal preparations used were, however, found to possess a weak activity of catalase, but no peroxidatic activity could be detected.*

Microsomal bound P-450 is reducible by both NADH and NADPH. As will be reported later (6), however, purified P-420 is not reducible by both NADH and NADPH. The addition of purified NADH-cytochrome bS reductase (31) in the presence of NADH also fails to reduce the purified pigment. It is likely that the reduction of P-450 by NADH and NADPH is catalyzed by specific reductases. Reduced P-450, on the other hand, is rapidly reoxidizable by oxygen, and neither the reduction nor reoxidation are influenced by cyanide.2 These findings indicate that P-450 can mediate aerobic oxidation of both NADH and NADPH by liver microsomes. It has in fact been reported that liver microsomes catalyze slow cyanide- insensitive oxidation of reduced pyridine nucleotides by oxygen (32, 33). In rabbit liver microsomes, the rate of NADH oxida- tion has been reported to be 1 to 2 mpmoles per minute per mg of protein and that of NADPH oxidation to be 3 to 5 mpmoles per minute per mg of protein at 20” and pH 6.5 (34,35). It has further been shown that the slow autoxidation of microsomal bound cytochrome bs can account for only 30% or less of the microsomal oxidation of reduced pyridine nucleotides (36). The rest of oxidation, therefore, seems to be catalyzed by a mechanism involving P-450. It is, however, dubious to assign an important role to this slow oxidative activity of P-450. As mentioned above, Estabrook et al. (29) have reported evidence suggesting that P-450 in adrenal cortex microsomes functions in oxygen activation for the steroid C-21 hydroxylase reaction. It appears therefore likely that P-450 in liver microsomes is also functioning in similar mixed function oxidase reactions such as hydroxylation and N-demethylation of foreign aromatic compounds (7, 9, 37). Hashimoto, Yamano, and Mason (38) have further found by electron spin resonance spectroscopy that liver microsomes con- tain an entity called “microsomal Fe,” which is capable of alternate reduction and oxidation. The properties reported for this entity, which has not yet been solubilized, seem to be similar to those of P-450.

The spectrum representing the difference between the reduced and oxidized forms of P-420 (Fig. 11) is similar to that of bacterial cytochrome bl (21, 22) and suggests that this pigment is a cyto- chrome of the b type. In agreement with this, protoheme is the only heme present in microsomes and all the microsomal heme can be extracted with acid-acetone. Moreover, the fact that P-420 is capable of undergoing reversible reduction and oxidation qualifies the pigment as a cytochrome. However, no b-type cytochromes so far known have been reported to combine with CO.

The occurrence in liver microsomes of two hemoproteins, P- 450 and cytochrome b5, may suggest that there is an interaction between them in the microsomal electron transfer mechanism.

Two observations, however, appear to provide evidence against this possibility. First, microsomal bound cytochrome b6 is partly reducible by ascorbate, whereas this reagent is completely incapable of reducing P-450 even in the absence of oxygen. The flow of electrons from cytochrome b5 to P-450 seems to be im- probable from these findings. Secondly, the oxidation of reduced P-450 by molecular oxygen proceeds much faster than that of cytochrome bg. This latter fact seems to exclude the possibility that oxidation of reduced P-450 takes place by way of cytochrome bg. Thus the two hemoproteins appear to behave independently in oxidative functions, though they are present in the same sub- cellular fraction of liver.

SUMMARY

Spectrophotometric evidence is presented that “P-450,” the CO-binding pigment, in rabbit liver microsomes is a new hemo- protein with unusual properties. Microsomal bound P-450, when reduced, binds CO and ethyl isocyanide. Although the CO difference spectrum of the reduced pigment is unusual for a hemoprotein, the isocyanide difference spectrum is characteristic of a hemoprotein compound. In microsomes P-450 is reducible by both NADH and NADPH under anaerobic conditions, but the reduced form is very rapidly reoxidizable in the presence of molecular oxygen.

When microsomes are incubated anaerobically with heated snake venom or with deoxycholate, the pigment is converted to a soluhilized form called “P-420.” The solubilization is, how- ever, accompanied by profound changes in the spectral properties of the pigment. The CO compound of reduced P-420 shows a difference spectrum characterized by an intense Soret peak at 420 rnp. P-420 is labile to aeration, especially in the presence of dithionite. In microsomal digests it can be reduced by NADH and NADPH in the absence of oxygen, but most of the pigment is maintained in the oxidized form under aerobic conditions even in the presence of NADH or NADPH. Cytochrome bs in the digests, on the other hand, is fully reduced under comparable conditions. Taking advantage of such difference in oxidation- reduction properties of P-420 and cytochrome bg, it is possible to measure a reduced minus oxidized difference spectrum of P-420 alone. The spectrum thus obtained suggests that P-420 is a cytochrome of the b type.

Liver microsomes contain protoheme in excess of the amount associated with cytochrome b5. Most of the extra heme seems to be associated with the CO-binding pigment.

Acknowledgments-We are indebted to Professor K. Kaziro of Nippon Medical School, Tokyo, for supplying us with an ethyl isocyanide preparation used in the early phase of the investiga- tion and to Dr. A. Ohsaka of National Institute of Health, Tokyo, for a gift of lyophilized snake venom. Miss Kazu Hirota and Miss Hiroko Nishibayashi have assisted us in some experiments. Their collaboration is also gratefully acknowledged.

REFERENCES

1. KLINQENBERG, M., Arch. Biochem. Biophys., 76, 376 (1968). 2. GARFINKEL, D.,Arch.Biochem. Biophys.,77,493 (1958). 3. SATO, R., AND OMURA, T., Proceedings of the Fifth International

Congress of Biochemistry, Vol. IX, Pergamon Press, Ltd., Oxford, 1961, p. 529.

4. OMURA, T., AND SATO, R., J. Biol. Chem., 237, PC1375 (1962). 5. OMURA, T., AND SATO, R., Biochim. et Biophys. Acta, 71, 224

(1963) . 6. OMMURA, T., AND SATO, R., J. Biol. Chem., 239.2379 (1964).

by guest on March 8, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Microsomal CO-binding Pigmen 2. I Vol. 239, No. 7

7. MITOMA, C., POSNER, H. S.,REITZ, H. C., AND UDENFRIEND, S., 24. KUFF, E. L., AND DALTON, A. J., J. Ultrastruct. Research, 1. Arch. Biochem. Bibphys.; 61, 43i (1956).

8. PAIGEN. K.. Biochim. et Bio~hvs. Acta. 19. 297 (1956). 9. IMAI, k;., AND SATO, R., Biociim. et biophys. Acta; 42, 164

(1960). 10. PAUL, K.-G., THEORELL, H., AND AKESSON, A., Acta Chem.

&and., 7, 1284 (1953). 11. SANDELL, E. B., Calorimetric determination of traces of metals,

Ed. 3, Interscience Publishers, Inc., New York, 1959, p. 541. 12. LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., AND RANDALL,

R. J., J. Biol. Chem., 193, 265 (1951). 13. BONNICHSEN, R. K., Acta Chem. &and., 4, 714 (1950). 14. RACKER, E., J. Biol. Chem., 184, 313 (1950). 15. EVANS, H. J., AND NASON, A., Plant Physiol., 28, 233 (1953). 16. JACKSON, H. L., AND MCKUSICK, B. C., in T. L. CAIRNS (Edi-

tor), Organic syntheses, Vol. 56, John Wiley and Sons, Inc., New York, 1955, p. 62.

17. STRITTMATTER, P., AND VELICK, S. F., J. Biol. Chem., 221,253 (1956).

18. WARBURG, O., NEGELEIN, E., AND CHRISTIAN, W., B&hem. Z., 214, 26 (1929).

19. ST. GEORGE, R. C. C., AND PAULING, L., Science, 114, 629 (1951).

20. CHANCE, B., AND WILLIAMS, G. R., J. Biol. Chem., 208, 945 (1954).

21. TANIGUCHI. S., AND ITAGAKI, E., Biochim. et Biophys. Acta, 44. 263 (196Oj. ’

22. FUJITA, T., ITAGAKI, E., AND SATO, R., J. Biochem. (Tokyo), 63, 282 (1963).

23. STRITTMATTER, C. F., AND BALL, E. G., J. Cellular Comp. Physiol., 43, 57 (1954).

62 (1957). 25. BARTSCH. R. G.. AND KAMEN. M. D.. J. Biol. Chetn.. 230. 41

(1968).’ ’ , ,

26. BARTSCH, R. G., ANDKAJIEN, M. D., J. Biol. Chem., 236, 825 ww.

27. HORIO, T.. AND KAMEN, M. D., Biochim. et Biophvs. Acta, 48. 266 (196i).

_ -

28. SUZUKI, H., AND IWASAKI, H., J. Biochem. (Tokyo), 62, 193 (1962).

29. E&AB$OOH, R. W., COOPER, D. Y., AND ROSENTHAL, O., B&hem. Z., 338, 741 (1963).

30. TAPPEL, A. L., .~ND ZALKIN, H., Nature, 186, 35 (1960). 31. STRITTMATTER, P., AND VELICK, S. F., J. Biol. Chem., 228, 785

(1957). 32. GILLETTE, J. R., BRODIE, B. B., AND LADu, B. N., J. Phar-

macol. Exptl. Therap., 119, 532 (1957). 33. KRISCH, K., AND STAUDINGER, H., Biochem. Z., 334,312 (1961). 34. OMURA, T., NISHIBAYASHI, H., AND S.4~0, R., Koso Kagaku

Shimpoziumu, Vol. 17, Nankodo, Tokyo, 1962, p. 88. 35. SATO, R., NISHIBAYASHI, H., AND OMURA, T., Biochinr. et

Biophys. Acta, 83, 550 (1962). 36. HIROTA, K., OMURA, T., NISHIBAYASHI, H., AND SATO, R.,

Preprints of 16th Koso Kagaku Shimpoziumu, Osaka, 1965, p. 55.

37. BRODIE, B. B., GILLETTE, J. R., AND LADu, B. N., in Ann& review of biochemistry, Vol. lY, Annual Reviews, Inc., Palo Alto, Calif., 1958, p. 427.

38. HA~HIMOTO, Y., YAMANO, T., AND MASON, H. S., J. BioZ. Chem., 237, PC3843 (1962).

by guest on March 8, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Tsuneo Omura and Ryo SatoFOR ITS HEMOPROTEIN NATURE

The Carbon Monoxide-binding Pigment of Liver Microsomes: I. EVIDENCE

1964, 239:2370-2378.J. Biol. Chem. 

  http://www.jbc.org/content/239/7/2370.citation

Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/239/7/2370.citation.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on March 8, 2020

http://ww

w.jbc.org/

Dow

nloaded from