Cytochrome P-450 from Bovine Adrenocortical Mitochondria IMMUNOCHEMICAL PROPERTIES AND PURITY*

Antibodies were induced against side chain cleavage P- 450 from bovine adrenocortical mitochondria by injecting the enzyme into rabbits. Double diffusion in agarose gels revealed single bands between rabbit anti-P-450 antiserum and adrenal cytochrome P-450 side chain cleavage prepared by two different methods and also with P-450 of different molecular weights called protein 4, protein 8, and large P- 450. In addition, the same anti-serum showed a line of identity with an ammonium sulfate fraction which is an impure precursor of pure cytochrome P-450 side chain cleavage in two methods used for its preparation. The IgG fraction a single P-450

A cytochrome P-450 responsible for catalyzing the ratedetermining step in steroid synthesis has been prepared from bovine adrenocortical mitochondria (1-5). The reaction catalyzed is called side chain cleavage of cholesterol (cholesterol + pregnenolone); this reaction requires TPNH and two ancillary proteins adrenodoxin and adrenodoxin reductase (6, 7). At least four methods are available for purifying the enzyme (l-5) but only two of these methods appear to produce proteins with identical properties (1, 5). Moreover, it is clear that the cytochrome purified to homogeneity according to rigorous * This work was supported by Grants CA14638-04 and AM15621-03 from the National Institutes of Health (to P.F.H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. physicochemical criteria can exist in a variety of forms of M, = 210,000, 420,000, 850,000 (8), and larger aggregates that have been less well characterized (5); the larger aggregates are collectively referred to as large P-450. All these forms of the enzyme give rise to a single band on electrophoresis on polyacrylamide gels with SDS'; this band corresponds to a subunit of M, = 52,000 to 53,000. The single species of protein subunit is not degraded to smaller forms by reactions which cleave S-S bonds (9), so that the forms referred to above are known as protein 4, protein 8, and protein 16 according to the number of subunits present. However, the best preparations contain variable amounts of phospholipid and the amount of phospholipid influences the behavior of the protein on electrophoresis in SDS gels (9). Moreover, on reconstituting the multienzyme system (adrenodoxin reductase, adrenodoxin, and P-450) in aqueous phases from the pure membrane proteins, reduction of P-450 by TPNH and ancillary proteins or reduction of P-450 alone by dithionite proceeds extremely slowly. This problem makes for difficulty in determining the heme content of the enzyme by CO-difference spectroscopy (1). This difficulty has in turn been misinterpreted as a reflection on the purity of the enzyme (4). It is, therefore, of considerable importance to eliminate the possibility that a heme protein which is not a cytochrome P-450 may be present as a contaminant in sufficient amounts to cause disparity between the heme content determined as the pyridine hemochromogen and the heme content determined from the CO complex. Such a possibility would require that the contaminant and the P-450 were indistinguishable by the physicochemical methods used to establish the purity of the P-450 (1). In order to clarify this point, to determine the immunochemical relationships between different forms of cytochrome P-450 in adrenal mitochondria and as a preliminary step in developing a radioimmunoassay for the side chain cleavage P-450, we have prepared antibodies to this protein. For studies of enzyme activity in the presence of antiserum or IgG, enzyme and antibodies were incubated for 3 h at 4" before the assay was performed.

Preparation of Antibody
Female New Zealand White rabbits were immunized subcutaneously in the nuchal region with 100 n.g of cytochrome P-450 prepared by affinity chromatography or on DEAE-cellulose (protein 16). The protein was emulsified with an equal volume of Freund's complete adjuvant (11). The animals were reimmunized every 3 weeks on opposite sides of the neck with 100 pg of the same protein emulsified in Freund's complete adjuvant for a total of six injections. Serum was collected before immunization and once a week after the 6th week. Animals were bled from the central artery of the ear and sera were stored in aliquots of 2 ml each at -20". In some cases, P-450 was boiled for 5 min in SDS at a final concentration of 0.5% (w/v) to prepare protein subunits which were injected into rabbits (after cooling) according to the above schedule. The preparation of goat anti-rabbit IgG has been described previously (11).
When anti-P-450 see antiserum was purified by chromatography on DEAE-cellulose, antibody activity toward P-450 see was observed only in the IgG fraction. In addition, the purified antibody cross-reacted with goat anti-rabbit IgG on double diffusion but not against other species of antibody and migrated like other rabbit IgG fractions on immunoelectrophoresis (data not shown). The following studies used both the crude antiserum and the IgG fraction prepared in this way.

Studies
of P-450lAnti-P-450 Reaction by Double Diffusion

Double Diffusion and Immunoelectrophoresis
Micro-Ouchterlony double diffusion tests were performed on microscope slides coated with 4.0 ml of 1.25% agarose in NaCl (0.15 M), potassium phosphate (10 mM, pH 7.21, and 0.02% sodium azide (12). All samples were incubated from 1 to 5 days at 25" and at 37". In the accompanying data, the total amount of protein in each well of the gels is given.

Effect of Antigen Concentration
-When a certain volume of antiserum to P-450 see was placed in the center well of an agarose slide and increasing concentrations of P-450 were placed in the peripheral wells, a single band was seen in the gels ( Fig. 1). As the concentration of P-450 was increased, the single narrow band became broader as expected, since at higher concentrations side chain cleavage P-450 becomes heterogeneous with respect to molecular weight in buffered solutions or during various purification procedures (1, 5, 8). However, no additional bands were seen at higher concentration.

Radioimmunoassay
of Cytochrome P-450 Cytochrome P-450 (100 pg) was iodinated by means of horseradish peroxidase attached to PA beads (Bio-Gel P-200) as described by Thorell and Larsson (13). Following iodination, the preparation was centrifuged to remove Bio-Gel beads and aqueous potassium iodide (10 ~1, 0.1 M) was added. The mixture was dialyzed against potassium phosphate (100 mM, pH 7.0) containing 0.1 mM EDTA and ion exchange resin AG l-X2 (200 to 400 mesh; Bio-Rad; 5 g/liter) overnight with five changes of buffer. The protein was then applied to a column of Sephadex G-100 (15 x 0.4 cm) and eluted with potassium phosphate buffer (100 mM; pH 7.0) with EDTA (0.1 mM). A single peak of l""I-bound protein was observed; the peak was eluted in 1.5 ml of eluate. This fraction contained the ""I-bound P-450 used for radioimmunoassay (7 x 10' cpm/ml). Electrophoresis on SDS gels revealed a single peak of ly;'I which was associated with a single protein band (P-450 see subunit).
Second antibody (sheep anti-rabbit serum) (150 ~1; 1:lO dilution) was then added. After standing at 4" for 8 h, the tubes were centrifuged at 2,500 x g for 20 min. The supernatant layer was removed and the pellet was examined for determination of '""I in a Nuclear Chicago y scintillation counter (model 1185).
Specificity of Anti-P-450 xc Antiserum - Fig. 2 shows that antiserum to P-450 see reacts with side chain cleavage P-450 producing a single band, but not with highly purified 11/3hydroxylase P-450 (10). Moreover, demonstrable antibodies do not appear within 1 week (Fig. 2, right gel, Well B) and the antibodies to P-450 see do not cross-react with antiserum prepared by injecting into another rabbit preparations of subunits of P-450 see prepared in SDS (Fig. 2, right gel, Well A). Normal rabbit serum and bovine serum albumin do not cross-react with anti-P-450 (data not shown).

Interaction
of Antisera Prepared against Protein 16 with Various Forms of P-450-Rabbit antiserum to protein 16 prepared by pregnenolone-Sepharose affinity chromatography (5) reacts with P-450 from DEAE-cellulose (l), with P-450 see prepared as a by-product during preparation of H/3-hydroxylase (lo), with large P-450 from affinity chromatography and with proteins 4, 8, and 16 from DEAE-cellulose (1, 8) (Fig. 3). It is important to notice the halo effect around wells containing large amounts of P-450. This effect is due to aggregation of P-450 in the gel and is not seen at low concentrations of this protein but is seen when P-450 is studied at similar concentrations without antiserum in the center well.

Purification of Antiserum
Serum from rabbits immunized against P-450 was purified by chromatography on DEAE-cellulose using a standard procedure (15).
Reaction of Anti-P-450 Serum with Purified and Crude Preparations of P-450 -The crude ammonium sulfate fraction (30 to 60% saturation) prepared from adrenal mitochondria contains many impurities (5); at least 15 bands are seen on SDS gels (data not shown). This fraction serves as a precursor of P-450 see prepared by various methods (1, 5, 10). As can be seen in Fig. 4, a single line of precipitation is observed with both the ammonium sulfate fraction and with highly purified P-450. Moreover, the junction of these two lines shows a classical line of identity, indicating that only a single species of antigen can be detected by this method. ies were raised against P-450 see prepared by chromatography on DEAE-cellulose.
In double diffusion this anti-P-450 antiserum gave single bands with P-450 prepared on DEAE-cellulose and that prepared by affinity chromatography (data not shown).

P-450
When anti-P-450 IgG was subjected to immunoelectropho-resis with P-450 see, a single band of precipitation was observed (Fig. 5).

Inhibition of Side Chain Cleavage
Enzyme Activity by Anti-P-450 Serum and by Anti-P-450 IgG Side chain cleavage of cholesterol is inhibited by antiserum to P-450 see (Fig. 6). Inhibition shows a curvilinear relationship to amount of added antiserum whether the side chain cleavage is measured by radioimmunoassay of pregnenolone (14) or by conversion of 17a-3Hlcholesterol to 13Hlpregnenolone (1). Normal rabbit serum (preimmunization) (Fig.  6A) and bovine serum albumin (Fig.  6B) stimulated the side chain cleavage reaction. This effect of albumin has been noticed previously (7), although the mechanism concerned is uncertain.
Presumably normal rabbit serum acts to increase enzyme activity as the result of a similar effect by rabbit serum protein.
When the serum used was taken from rabbits after the second injection of P-450, inhibition was less pronounced (Fig. 6A)  Immunochemistry of Cytochrome P-450 no visible bands with P-450 in double diffusion (data not shown); clearly the inhibition of enzyme activity is a much more sensitive indication of antigen-antibody reaction than double diffusion under the conditions used. Antiserum to P-450 XC was without obvious effect on lip-or 1%hydroxylase enzyme activities (data not shown).
When purified IgG prepared from antiserum to P-450 see was incubated with the enzyme, inhibition of the enzyme reaction was observed as expected (Fig. 6C). In this case, inhibition by IgG occurs without the complicating effect of those serum proteins which stimulate the enzyme activity (Fig. 6, A and B). It can also be seen that IgG from normal rabbit serum (no immunization) was without effect on the enzyme activity (Fig. 6C).

Reaction
of '2~T-labeled P-450 with Anti-P-450 Fig. 7 shows the result of studies in which iodinated P-450 was added to antiserum and unlabeled P-450. It can be seen that P-450 competes with '*"I-labeled P-450 in combining with anti-P-450 antibodies in a concentration-dependent manner. The assay procedure is capable of detecting 0.01 pg of cytochrome P-450 see (p < 0.1) and binding is linear in the range of 0.025 to 0.25 pg using a log scale for amount of P-450 added. Addition of lip-hydroxylase P-450 causes some decrease in bound IzaI, but within the limits tested, evidence of specific binding was not observed since a parallel displaced curve was not seen. The lip-hydroxylase contains bound Tween 20 and, although this substance does not affect the binding of P-450 with antibody when added without H/3-hydroxylase in amounts equivalent to those accompanying the enzyme additions shown in Fig. 7, higher concentrations of Tween do modify the combination of P-450 with antibody. For these reasons, addition of very large amounts of the enzyme cannot be made without nonspecific effects.
The present studies lend strong support to the conclusion that the side chain cleavage P-450 prepared by chromatography on DEAE-cellulose (1) and that prepared by affinity chromatography on pregnenolone-Sepharose (4) are highly purified. Each of these preparations at various concentrations shows a single band with rabbit anti-P-450 antiserum in double diffusion. Moreover, only a single band of precipitate (line of identity) was seen in double diffusion with anti-P-450 serum and a crude ammonium sulfate precipitate, which represents a precursor of pure P-450 in both methods of purification.
The enzyme mixed with purified antibody (IgG fraction) also gives a single band on immunoelectrophoresis. In addition, the various forms of the cytochrome P-450 which differ in molecular weight also show a single band with the anti-P-450 antiserum. This statement applies to proteins 4, 8, and 16 prepared by chromatography on DEAE-cellulose and to protein 16 and large P-450 prepared by affinity chromatography. Evidently, these various forms of the enzyme do not result from binding of smaller forms by a contaminating protein. These findings indicate that the various forms of the enzyme reveal the same antigenic determinants.
Since several available methods of preparing P-450 see result in enzymes with significant differences in physicochemical properties (l-51, it was of interest to demonstrate that the two preparations developed in this laboratory (based upon DEAE-cellulose and affinity chromatography, respectively) are immunochemically identical since both preparations have been used extensively in published studies.
Purification of anti-P-450 antiserum reveals the presence of antibody activity only in the IgG fraction.
Evidently, the physicochemical evidence (SDS gels, analytical ultracentrifugation of the enzyme as subunit or as native enzyme, various forms of chromatography, and isoelectric focusing) (1, 8) are accurate indications of homogeneity of this cytochrome P-450 and the possibility (4) of significant contamination by other heme proteins can be excluded. Moreover, the various forms of side chain cleavage P-450, which show the same physicochemical properties except for molecular weight, are apparently immunochemically indistinguishable and represent various states of aggregation of the pure protein. Evidently, the cause of difficulty encountered in determining the heme content of the enzyme does not result from the presence of impurities containing heme and must be sought elsewhere.
Although the anti-P-450 see reacts with all these various forms of the side chain cleavage enzyme, it does not react with 11/3-, 18-hydroxylase P-450 in double diffusion; this enzyme is prepared from the same mitochondria.
Again, the anti-P-450 see inhibits side chain cleavage enzyme activity but not lipand 18-hydroxylase activities. The 11/3-, 18-hydroxylase shows limited competition with '""I-labeled P-450 see for the anti-P-450 antiserum, but the possibility that this competition is nonspecific cannot be eliminated at this time (Fig. 7). Evidently the antigenic determinants of native P-450 are not revealed by the subunits of the enzyme, since the serum of animals treated with the subunits does not cross-react with P-450 see in double diffusion (Fig. 2).
Finally, the cytochrome P-450 has been successfully iodinated and unlabeled P-450 competes with the iodinated enzyme for binding to the antibody. This reaction has provided the basis for a sensitive radioimmunoassay for P-450 see which can detect 0.01 pg of the enzyme. It is proposed to increase the sensitivity of the assay by suitable modification of the iodination procedure.