Biochemical Distinctions between the Nuclear and Microsomal Membranes from Rat Hepatocytes

SUMMARY The nuclear membrane from rat hepatocytes has been examined for components of the DPNH and TPNH electron transport chains as well as two drug-metabolizing enzymes, hydroxylase and an aminoazo dye N-de-methylase, in control and phenobarbital-treated animals. The enzyme induction phenomenon characteristic of the microsomes was not observed with the nuclear membrane for any of the enzymes or pigments studied. The nuclear electron transport enzyme DPNH-cytochrome c reductase and cytochrome bs were found to follow the same pattern as their microsomal counterparts. Thus, in each case, the reductase level was depressed in the treated animals while the cytochrome bg level was not significantly altered. In control animals, the specific activity of TPNH-cyto-chrome c reductase in the nuclear membrane was approximately one-third that found for the same enzyme associated with the microsomes. The specific activity of this enzyme remained unchanged

portance to define in a precise manner the chemical and biochemical properties of the nuclear membrane. Questions regarding nuclear function frequently require a detailed knowledge of membrane structure and function. Studies from our laboratory (1,2) and elsewhere (3)(4)(5)(6)(7)(8)(9)(10) have dealt with the isolation, composition, morphology, and biochemistry of this unique bileaflet structure. Morphologically, the outer leaflet of the nuclear membrane forms a continuum with the endoplasmic membranes (ll), and it is of extreme interest to study the biochemical response of the nuclear membrane under various physiological conditions in an attempt better to define its functional role in the cell.
The main objective of this study was to examine further the biochemical interrelationships between the nuclear and microsomal membranes under conditions of enzyme induction. The response of various membrane-associated enzymes to phenobarbital administration is rather well characterized in the case of the microsomal system (12). For example, phenobarbital induces a spectrum of drug-metabolizing enzymes which require TPNH and molecular oxygen. Concomitant with this increase in drug-metabolizing capacity is an increase in the level of the components of,: the TPNH electron transport chain which includes TPNH-cytechrome c reductase (thought to be equivalent to cytochrome P-450 reductase) and cytochrome P-450 (13)(14)(15). This behavior is in contrast to the microsomal DPNH electron transport chain which is not induced by phenobarbital (14).
This communication describes the systematic examination of the nuclear membrane for the DPNH-and TPNH-dependent electron transport systems as well as for two representatives of the drug-metabolizing enzymes, N-demethylase and benzo[a]pyrene hydroxylase. METHODS

AND MATERLALS
Animals-Male Holtzman rats weighing 50 to 60 g were used in all studies. Induction of drug-metabolizing enzymes was accomplished by one daily intraperitoneal injection of 100 mg of phenobarbital per kg of body weight (14). Control animals received 0.9% NaCl by the same route. All animals were injected for 5 successive days in the mornings so as not to disturb feeding and sleeping cycles (16).
Preparation of Nuclear and Microsomal Membranes-Animals were fasted for a period of 20 hours and killed by decapitation. All subsequent operations were carried out at either ice bath temperature or 3", and maximum values are given for the relative centrifugal force. The livers were homogenized in 2 vol- Nuclei were sedimented through the high density sucrose by centrifuging at 106,000 X g for 65 min.
The recovery of DNA in the washed nuclear pellet was in the range of 50 to 70% of the DNA present in the homogenate.
Nuclear membrane comprising both the inner and outer leaflets was prepared as previously described (2). The distribution of nuclear membrane between the density (d) 1.16 to 1.18 and d 1.18 to 1.20 g per cc interfaces was approximately 10 : 1 on a protein basis.
Both fractions were combined for the purpose of this study.
Routinely, 100 g of liver were processed per preparation and the average yield of nuclear membrane from control and phenobarbital-treated rats was 10 mg and 6 mg of membrane protein per 100 g of liver.
The supernatant recovered from the 3,000 x g run was recentrifuged at 14,000 x g for 10 min in order to remove nonmicrosomal contaminants.
Crude microsomal membrane was obtained by centrifugation of the postmitochondrial supernatant at 106,000 X g for 60 min.
The microsomal membrane was further purified by density gradient centrifugation on a discontinuous sucrosecitrate gradient by the method of Kashnig and Kasper (2). The above procedure permitted the isolation of nuclear and microsomal membranes from the same liver homogenate.
Enzyme Assays-DPNH cytochrome c reductase was assayed by the method of Mackler and Green (17) with 0.50 ml of cocktail.
The final volume after the addition of enzyme and cofactor was 0.58 ml.
For the determination of TPNH-cytochrome c reductase the same assay was used but 100 1.18 of TPNH were substituted for DPNH. All incubations were carried out at 30" in a thermostated cuvette chamber of a Beckman DB-G spectrophotometer and the amount of cytochrome c reduced was calculated with the extinction coefficient at 550 mp of 18.5 cm+ m~-l (18). TPNH and DPNH were obtained as the tetrasodium salts from Sigma. N-Demethylase activity was determined by measuring the oxidative demethylation of 3-methyl&monomethylaminoazobenzene by a modification of the procedure of Mueller and Miller (19). The assay was conducted in 0.05 M Tris-HCl buffer, pH 7.5, containing 25 pg of 3-methyl-4-monomethylaminoazobenzene, 522 pg of MgCL, 2 mg of KCl, 500 pg of TPNH, and 0.5 mg to 2 mg of microsomal or nuclear membrane protein in a final volume of 1.1 ml.
The reaction was initiated by the addition of TPNH and the incubation was carried out aerobically with shaking at 37" for 30 min.
A TPNH-regenerating system was not required.
The reaction was stopped by the addition of 1.5 volumes of acetone and 1.5 volumes of benzene.
After vigorous mixing, 1.0 ml of the organic phase was removed and the solvent was evaporated in a stream of nitrogen. The residue was redissolved in 25 ~1 of methanol and a 5+1 aliquot was applied to an activated previously coated silica gel plate (Eastman, type 6061). The chromatogram was developed at room temperature for 4 hours with a methanol-hexane solvent (4:96). After removal of the solvent, the visualization of the dyes was aided by brief exposure of the thin layer chromatographic plate to HCl vapor.
The area of the chromatogram containing each of these compounds was cut out and transferred to individual conical centrifuge tubes.
Methanol (0.6 ml) was added to each tube and the silica gel was dislodged from the Mylar backing by vortex mixing.
This step also extracted the dyes from the silica gel. An equal volume of 7 N HCl was added to the methanol extract with mixing, the Mylar backing was removed, and the contents of the tube were centrifuged to obtain an optically clear solution.
Concentrations were determined by comparison to a standard curve after the optical density was measured at 505 rnp with a Beckman DU-2 spectrophotometer. Benzo[a]pyrene hydroxylase was assayed by the procedure of Nebert and Gelboin (20 in the reference and sample cuvettes with sodium dithionite, carbon monoxide was gently passed through the contents of the sample cuvette for 20 to 30 set prior to recording the difference spectrum. An extinction difference coefficient of 91 cm-1 rnM+ between 450 rnp and 490 rnp was used to calculate the amount of CO-heme complex (22).
Protein was determined by the Folin procedure (23) with four times crystallized ovalbumin as the standard.

AND DISCUSSION
Nuclear and microsomal membranes were prepared from the same animals so that precise correlations of enzymic activities could be made.
As isolated, the nuclear membrane is a complex comprising both the inner and outer leaflets. In control animals, DPNH-cytochrome c reductase was found in the nuclear membrane at a level approximately 56% that found for the microsomal membrane (Table I) ; this finding is consistent with earlier results from our laboratory which also showed the rotenone insensitivity of nuclear-DPNH-cytochrome c reductase (2). Berezney, Funk, and Crane (10)   The pigment, cytochrome bs, a component of the DPNH electron transport chain was found in nuclear membrane from control animals at 37% of the level occurring in the microsomes ( Table  I).
The cytochrome b6 content of 0.183 mpmole per mg of nuclear membrane protein is greater than 5 times the content reported by Franke et al. (7) for the same membrane.
It is well known that the nucleus contains proteolytic enzymes (25), and it is possible that the prolonged incubation of a sonicated nuclear suspension (7) could have resulted in the release of some membrane proteins through the action of proteases.
Cytochrome bg is released in good yield from the microsomal membrane by mild proteolytic digestion (2528). Phenobarbital induction did not significantly alter the cytochrome b6 content of the nuclear membrane and in this respect is quite similar to the pattern obtained for the microsomal membrane (Table I; Reference 14). Presumably, the induction process has little effect on the rate of cytochrome bg synthesis in microsomes (24). Furthermore, cytochrome 66 degradation is completely prevented during phenobarbital administration (29). Our results would suggest that a similar situation may exist for the nuclear membrane. TPNH-cytochrome c reductase was found to be an integral part of the nuclear membrane with a specific activity close to one-third of the microsomal membrane. This reductase is generally assumed to be identical with the enzyme that transfers reducing equivalents from TPNH to cytochrome P-450 (30). Other investigators have also reported the occurrence of this enzyme in rat liver nuclear membrane (6,7). However, in the case of bovine liver nuclear membrane, TPNH-cytochrome c reductase appears to be absent (10). A most important distinction exists between the. nuclear and microsomal membranes. The TPNH-cytochrome c reductase associated with the nuclear membrane is not inducible by phenobarbital while the level of the microsomal enzyme is increased 2-to 4-fold over control values (Table I; Reference 14). This finding clearly illustrates a marked difference in the effect of phenobarbital on two intracellular membranes.
Analysis of the nuclear membrane for cytochrome P-450 yielded variable results. Ten individual preparations from control animals were examined and negative results were obtained with seven of the samples.
In three preparations, cytochrome P-450 was present at the level of 0.08, 0.16, and 0.22 mpmole per mg of protein.
These values correspond to 13, 26, and 35% of the value obtained for the microsomal membrane.
All of the analyses were performed at a protein concentration in the range of 2 to 4.5 mg per ml.
Thus, the level at which the analyses were carried out would have permitted the detection of cytochrome P-450 down to the 10% level in most cases. The reason for not obtaining uniform results in the determination of this pigment is unclear.
No cytochrome P-420 was observed in any of the nuclear membrane spectra. Consequently, the conversion of P-450 to P-420 did not account for the negative results.
Assays performed on four different preparations of nuclear membrane from phenobarbital-treated animals indicated the absence of cytochrome P-450. Franke et al. (7) have reported that rat liver nuclear membrane from untreated animals had a cytochrome P-450 content of 0.025 mpmole per mg which was one-seventh the level present in the microsomes.
In the current study, the microsomal membrane yielded values of 0.62 and 1.57 mpmoles of cytochrome P-450 for control and treated animals, respectively (Table I).
These values are in agreement with those reported by Estabrook and Cohen (30) but are several orders of magnitude larger than values reported by other investigators (7,14,31). It is difficult to make accurate comparisons of specific activity measurements on microsomal constituents conducted in different laboratories since the protein content of the microsomal vesicle can widely vary depending on the method of preparation. It should be emphasized that the microsomes used in this study are disrupted by sonic oscillation and purified by discontinuous gradient centrifugation. This procedure gives predominantly the insoluble membrane matrix free of soluble intravesicular proteins.
Low levels of N-demethylase and aryl hydroxylase activity were found in the nuclear membrane from control animals which corresponded to 11 and 7% of the specific activity obtained for the microsomal membrane, respectively (Table  I).
A most striking feature was the lack of inducibility of these two enzymes in the nuclear membrane by phenobarbital. This is in marked contrast to the microsomal membrane which exhibited a 2. aryl hydroxylase activity. The spectrophotofluorometric determination of hydroxylated benzo[a]pyrene derivatives is extremely sensitive and is capable of detecting as little as lo-l2 mole per ml (20). Consequently, small fluctuations in activity should be easily detectable. For the purposes of discussion, if we assume that the nuclear membrane aryl hydroxylase activity in the control animals was due to a 7% microsomal contamination, the value for the phenobarbital-treated animals should have increased from 259 in the controls to 820 mpmoles per mg per 30 min. The fact that no increase was noted strongly indicates that the aryl hydroxylase and the N-demethylase are actual nuclear membrane constituents. The nuclear membrane also contained an enzyme capable of the reductive cleavage of the azo linkage in 3-methyl-4-aminoazobenzene.
Quantitative measurements were not made, but the products were detected after thin layer chromatography of the reaction mixture.
The current study clearly illustrates several important points regarding the biochemical interrelationships of the nuclear and microsomal membranes.
1. Rotenone-insensitive DPNH-cytochrome c reductase and cytochrome bg are firmly associated with the nuclear membrane. Under conditions of enzyme induction with phenobarbital, their levels followed the same pattern as the levels of their counterparts in the microsomal membrane.
2. TPNH-cytochrome c reductase is an integral part of the nuclear membrane but differs markedly from the microsomal enzyme in not being induced by phenobarbital.
3. Cytochrome P-450 was not detectable in seven out of 10 nuclear membrane preparations from control animals. In three preparations, the level ranged from 0.08 to 0.22 mpmole per mg. No cytochrome P450 was found in nuclear membrane from phenobarbital-treated animals. 4. N-Demethylase and benzo[a]pyrene hydroxylase were present in the nuclear membrane at low levels and differed from their microsomal counterparts in not being inducible.
Although the mechanism by which phenobarbital functions in enzyme induction is not completely understood, our results suggest that the regulation of enzyme levels in the nuclear membrane with respect to TPNH-cytochrome c reductase and the drug-metabolizing enzymes is under a different control than that which operates in the microsomal membrane. The comparatively high level of TPNH-cytochrome c reductase in relation to the other members of the TPNH electron transport chain raises the question as to its metabolic role in the nuclear membrane. Since it is known that phenobarbital and 3-methylcholanthrene each stimulate the synthesis of a specific type of P-450 (32), the effect of other inducing agents on the level of this pigment in the nuclear membrane will be of interest. It is conceivable that an entirely different electron acceptor may function in the nuclear membrane which is yet unidentified.
The differential response of the nuclear and microsomal membranes to phenobarbital has provided a sensitive means for establishing clear-cut differences between these two closely related intracellular structures.
It has been suggested that the endoplasmic reticulum is derived from the outer leaflet of the nuclear membrane by a delaminating process (33,34). Since phenobarbital administration results in the elevation of the total amount of smooth surfaced membranes in the hepatocyte (12), an increased turnover of the outer leaflet of the nuclear membrane would be expected. Studies examining the incorporation of '"C-choline into nuclear and microsomal phospholipid did not support a precursor-product relationship (35) ; however, the presence of significant amounts of nonmembrane phospholipid in the nucleus (36) renders this interpretation equivocal. Considering the morphological and biochemical heterogeneity of the microsomal vesicles (37,38), the outer nuclear leaflet could represent a segment of endoplasmic reticulum that is undifferentiated with respect to certain physiological and biochemical functions normally associated with the microsomal fraction. Alternatively, the biogenesis of the outer nuclear leaflet and the endoplasmic reticulum may be quite distinct and responsive to different regulatory mechanisms.
In addition to the differences and similarities already discussed, earlier biochemical studies established the presence of glucose 6-phosphatase (1,2) and Mg*-adenosine triphosphatase (2,6,7) in the nuclear membrane of rat hepatocytes. A Na+-K+-adenosine triphosphatase was absent (2,6,7). Claims (6, 7) that glucose 6-phosphatase is not associated with the nuclear membrane should be critically examined. In one instance (7), the rat liver nuclei used to prepare nuclear membrane contained glucose 6-phosphatase at a level which was 26% the level found in the microsomes. Although the isolated nuclear membrane contained a negligible amount of the enzyme, an attempt to account for the fate of the initial enzymic activity present in the nuclei was not made. This is particularly important since attention has already been called to the fact that the glucose 6-phosphatase of the nuclear membrane is considerably less stable than the corresponding microsomal enzyme (2). Berezney et at. (lo), working with bovine liver, have reported that glucose 6-phosphatase was present in nuclear membrane at a specific activity one-third that of the microsomal enzyme. Recent histochemical studies (39-41) support the biochemical observations by clearly showing that glucose 6-phosphatase was localized on the nuclear membrane and the endoplasmic reticulum.