Catalytic Activity of Partially Purified Renal 25-Hydroxyvitamin D Hydroxylases from Vitamin D-deficient and Vitamin D-replete Rats*

A method based on affinity and hydrophobic chromatography has been developed for the partial purifi- cation of renal mitochondrial cytochrome P-450 (P- 450). 2,4-Dichloro-6-phenylphenoxyethylamine coupled to Sepharose 4B provides the chromatographic medium which in the presence of two detergents, emulgen 911 (0.06%) and cholate (0.16%). retained P-450 but not most other mitochondrial proteins. P-450 was eluted from the column by increasing the detergent concentrations to 0.2% emulgen 911 and 0.5% cholate. Because of removal of 80% of the mitochondrial proteins and most of the chromophores, the method allows for the spectral quantitation of the mitochondrial P-450 and provides starting material for further purification. The specific content of the P-450 eluted from the 2,4-dichloro-6- phenylphenoxyethylamine columns varied between 0.5 and 2 nmol/mg of protein. The concentration of P-450 in the renal mitochondria of control rats was 0.26 f 0.2 nmol/g of kidney or 0.016 nmol/mg of mitochondrial protein; renal mitochondria from vitamin D-deficient rats contained similar amounts of P-450. The solubilized partially purified P-450 fraction catalyzed both the la- and 24-hydroxylations of 25-hy- droxyvitamin DS. In vitamin D-replete animals, the turnover numbers of the la- and 24-hydroxylation reactions were 0.38 and 0.18 pmol/pmol of

A method based on affinity and hydrophobic chromatography has been developed for the partial purification of renal mitochondrial cytochrome P-450 (P-450). 2,4-Dichloro-6-phenylphenoxyethylamine coupled to Sepharose 4B provides the chromatographic medium which in the presence of two detergents, emulgen 911 (0.06%) and cholate (0.16%). retained P-450 but not most other mitochondrial proteins. P-450 was eluted from the column by increasing the detergent concentrations to 0.2% emulgen 911 and 0.5% cholate. Because of removal of 80% of the mitochondrial proteins and most of the chromophores, the method allows for the spectral quantitation of the mitochondrial P-450 and provides starting material for further purification. The specific content of the P-450 eluted from the 2,4-dichloro-6phenylphenoxyethylamine columns varied between 0.5 and 2 nmol/mg of protein. The concentration of P-450 in the renal mitochondria of control rats was 0.26 f 0.2 nmol/g of kidney or 0.016 nmol/mg of mitochondrial protein; renal mitochondria from vitamin D-deficient rats contained similar amounts of P-450.
The solubilized partially purified P-450 fraction catalyzed both the la-and 24-hydroxylations of 25-hydroxyvitamin DS. In vitamin D-replete animals, the turnover numbers of the la-and 24-hydroxylation reactions were 0.38 and 0.18 pmol/pmol of P-450/30 min. In vitamin D deficiency, there was an increase in the turnover number of both the la-and 24-hydroxylations Renal mitochondria and placental homogenates catalyze the conversion of 25(OH)D3' to two metabolites, la,25(OH)zD3 and 24,25(OH)zD3 (1)(2)(3). la,25(OH)~D;$ is involved in calcium absorption from the gut (4,5) and mobilization of calcium from bone (6). The role of 24,25(OH)zD3 is less clear, but there is evidence that it is involved in bone ossification (7).
The two hydroxylation reactions are closely regulated by serum levels of calcium, phosphate, 25-hydroxyvitamin D, and parathyroid hormone. In vitamin D deficiency, in hypocalcemia, and in hypophosphatemia, there is an increase in the rate of 1a,25(OH)2D3 formation by the kidney, while in vitamin D-replete animals with a normal phosphate and calcium dietary intake, 24,25(OH)zD3 is the predominant metabolite (8)(9)(10). Both the la-and 24-hydroxylases are cytochrome P-450 mixed function oxidases (11,12) but the biochemical mechanisms involved in the regulation of these enzymes are unknown. The quantitative and qualitative changes which occur in renal mitochondrial P-450s under conditions known to affect the la-and 24-hydroxylases have been difficult to assess for several reasons. First, the concentration of the enzyme in the membrane is low (estimated to be between 0.01-0.04 nmol/mg of mitochondrial protein (13,14); second, the presence of other chromophores prevents the spectral quantitation of P-450; third, the enzyme recovery is low and the catalytic rates of the enzyme are slow after solubilization from the membrane (15,16).
The fist objective of this present work was to develop a method for solubilization and purification of renal mitochondrial P-450. In this report, such a method is described; it permits estimation of the P-450 content of individual rat kidneys and characterization of the enzymes by their catalytic activities.
The second objective was to compare renal mitochondrial P-450s from vitamin D-deficient and vitamin D-replete rats in order to determine whether the increase in the la-hydroxylase which occurs in vitamin D deficiency results from an increased enzyme concentration, an activation of existing enzymes, or the synthesis of a new form of P-450. MATERIALS  with a PT20 probe, for 5 s at speed setting No. 5. Mitochondrial fractions were prepared by differential centrifugation. The nuclear fraction was removed by centrifugation at 1000 X g for 10 min. The mitochondrial fraction was obtained by centrifuging the postnuclear fraction a t 9OOO X g for 10 min. Preparation of DPEA Columns-Aliquots (5 g) of CH-Sepharose were swollen, washed, and coupled to DPEA as directed by the manufacturers. This involved incubating a 3-fold molar excess of DPEA with the swollen gel a t room temperature for 2 h in 100 mM sodium bicarbonate buffer (pH 8.0) containing 1 M NaCI. After coupling, the gel was washed with three cycles of 100 mM Tris buffer, pH 8.0, 100 mM sodium acetate buffer, pH 4.5, both containing 1 M NaCI. The gel was then washed with 1 liter of deionized water and equilibrated in 30 mM sodium phosphate buffer, pH 7.4, containing 20% glycerol, O.M% emulgen 911, 0.16% sodium cholate (starting buffer). Equilibrated gel was packed in columns (0.5 X 10 cm) in starting buffer.
DPEA Chromatography-Mitochondrial fractions were resuspended with the use of a glass hand homogenizer in 30 mM sodium phosphate buffer, pH 7.4, containing 20% glycerol, 0.2% emulgen 911, 0.5% sodium cholate, 0.1 mM EDTA (solubilization buffer). The final suspension contained the equivalent of 1 g of tissue, wet weight/ml, and was allowed to stand at room temperature for a t least 1 h prior to chromatography. Such suspensions can be left at room temperature for 4 h or a t 4 for 24 h with no loss of P-450. Before application to DPEA columns, the suspensions were diluted 3-fold with 30 mM sodium phosphate buffer containing 20% glycerol. This step brings the concentrations of the detergents to those in the starting buffer. These concentrations were found to be optimal for binding of P-450 to the columns. Samples were applied to DPEA columns at a flow rate of 0.5 ml/min. Following application of the sample, columns were washed with starting buffer containing 100 mM NaCI. The P-450 was eluted with solubilization buffer.
Spectral Measurements-The concentration of P-450 was determined from the CO-reduced difference spectrum after reduction with dithionite according to the method of Omura and Sat0 (17). An extinction coefficient of 91 mM " cm I was used. Protein was measured by the method of Lowry et al. (18) with bovine serum albumin as standard. Cytochrome c reductase was measured according to Phillips and Langdon (19) with 1 unit of activity defined as the reduction of 1 nmol of cytochrome c h i n a t 27 "C.
Measurement of Catalytic Actiuity-Because of the increased instability of P-450 after removal of emulgen 91 1, catalytic activity was assessed within 4 h of removal of the detergent. The assay mixture consisted of Tris-acetate, 30 mM, pH 7.4, glucose &phosphate, 1 mM, glucose-6-phosphate dehydrogenase, 1 unit, NADPH, 100 p~, P-450 fraction, 0.1-1 ml, 30p1 of ["H]25(OH)D:, (248,000 cpm), and unlabeled 25(OH)Da between 10 and 500 nM as indicated in individual experiments. The incubation volume was 5 ml. The substrate dissolved in ethanol was added to a 20-ml Erlenmeyer flask. The ethanol was evaporated under a stream of N?. All other components of the system except NADPH were then added to the flask. The reaction was started with addition of NADPH. Incubations were done at 37 "C. The reactions were stopped by decanting the contents of the Erlenmeyer flasks into 50-ml tubes containing 10 ml of chloroform/methanol (2:l). The tubes were vortexed and centrifuged, and the lower phase was removed. The extraction procedure was repeated. Both organic phases were pooled and evaporated under a stream of N,. This procedure resulted in extraction of 95' % of the radioactivity from the incubation medium. The residue was resuspended in 20 pl of isopropyl alcohol/hexane (1:9) and an aliquot of this was assayed for vitamin D metabolites on a Beckman Model 110 A liquid chromatograph.
Assay of Metabolites-Vitamin D metabolites were separated on a Zorbax-Si1 column (4.6 mm X 25 cm) (DuPont) equilibrated with isopropyl alcohol/hexane (1:9) according to Jones and DeLuca (23). The flow rate was maintained a t 1 ml/min and 1.5-ml fractions were collected. Radioactivity in each fraction was measured in 10 ml of Aquasol by liquid scintillation counting. Fig. 1 illustrates a typical elution profile from a PDEA column in which a bed volume of 10 ml was used to isolate mitochondrial P-450 from the kidneys of five control rats. Smaller columns of 1-ml bed volume have been used to isolate P-450 from individual rat kidneys. The yield of mitochondrial protein was 16.4 f 1.2 mg/g of kidney. Upon DPEA chromatography of solubilized renal mitochondria, 85% of the protein was removed by washing the column with starting buffer containing 100 mM NaCI. Elution of P-450 was achieved with solubilization buffer. The specific content of P-450 eluted from DPEA columns varied between 0.5 and 2.0 nmol/mg of protein. The calculated mitochondrial P-450 content was 0.26 * 0.02 nmol/g of kidney or 0.016 nmol/mg of mitochondrial protein. Rechromatography of the unbound protein fraction from the DPEA column wash did not result in further binding of P-450 to the column. A typical CO-reduced difference spectrum of the P-450 eluted from DPEA columns is illustrated in Fig. 2. The absorption maximum is at 451 nm and there was very little conversion to P-420. In vitamin D deficiency, there was no change in the renal mitochondrial content of P-450 and no difference in the CO-reduced difference spectral maximum ( Table I). The P-450 was assessed in three separate groups of animals which were maintained on vitamin D-deficient diets on three separate occasions. The P-450 content was found to be the same as in control animals whether  individual kidneys were examined or kidneys from groups of up to 20 animals were pooled and chromatographed together.

Comparison of spectral and kinetic properties of solubilized P-450 from vitamin D-deficient and vitamin D-replete rats
SDS Electrophoresis-A comparison of the SDS electrophoretic profiles of P-450 fractions from vitamin D-replete and D-deficient animals provided no evidence that the forms of P-450 present in these two groups of animals are different (Fig. 3). All known P-450s have M , = 60,OOO-45,OOO (24). In this molecular weight range of the gel, bands of 60,000 and 55,000 are present. Although none of these bands has been identified as P-450, the intensity of staining of the proteins reflects the concentration of P-450 protein applied to the gels.
Catalytic Actiuity-Upon removal of the detergents, the P-450 fraction catalyzed the hydroxylation of 25(OH)DZI at both the la-and 24-positions. Fig. 4 is an example of a high pressure liquid chromatography elution profile of the chloroform/ methanol extract after incubation. The system exhibited no requirement for added adrenodoxin or adrenodoxin reductase. Since the cytochrome c reductase activity of the P-450 fraction was 0.1 unit of cytochrome c reduced/min/pmol of P-450, it can be concluded that renal ferredoxin and renal ferredoxin reductase (14) are eluted from the DPEA column along with P-450. Both hydroxylations were inhibited by emulgen 911 concentrations above 0.002%. The hydroxylation rates were linear with time (Fig. 5) and P-450 concentration (Fig. 6) for 30 min. At a substrate concentration of 500 nM, P-450 fractions from vitamin D-replete animals catalyzed the la-and 24hydroxylations of 25(OH)D:, at rates of 0.38 and 0.18 pmol/ pmol of P-450/30 min, respectively. With P-450 from vitamin D-deficient rats, the turnover numbers increased to 4.00 for the la-hydroxylase and 1.75 for the 24-hydroxylase ( Table I).
Comparison of P-450s from Vitamin D-deficient a n d Vitamin D-replete Animals by K,,, a n d Sensitivity to Znhibitors-The increased catalytic rate in vitamin D deficiency was not due to an increased affinity for the substrate since the K,,, values for the 24-hydroxylase (25 nM) and the la-hydroxylase (50 nM) were unaffected by the vitamin D status of animals (Fig. 7).

Renal Mitochondrial 25(0H)D3 Hydroxylase
The standard modulators of P-450, SKF 525A and metyrapone, were examined for their effects on solubilized renal mitochondrial P-450 (Fig. 8). No differences were observed between P-450 fractions from vitamin D-deficient or vitamin D-replete animals in their response to SKF 525A M), which inhibited both the la-and 24-hydroxylase by 70%. Metyrapone M) caused a 2-fold stimulation of the 24hydroxylase in the P-450 fraction from vitamin D-replete animals but had no effect on the 24-hydroxylase from vitamin D-deficient animals or on the la-hydroxylase from either animal group.
Effects of Calcium and Phosphate Ions on Catalytic Actiuity-The effects of calcium and phosphate ions were tested in order to determine whether these ions regulate the solubilized vitamin D hydroxylases in a fashion similar to that observed in renal mitochondria (26,27). The results are shown in Table  11. Both the la-and 24-hydroxylases were stimulated by calcium. There was a maximal 2-fold stimulation at a calcium concentration of M. A t concentrations above lo-" M, calcium had no effect. On the other hand, phosphate at concentrations above 2.5 X lo-' M inhibited both hydroxylations. The 24-hydroxylase was completely inhibited and the la-hydroxylase inhibited by 86% at a phosphate concentration of 10" M. Similar results were obtained whether P-450 was derived from renal mitochondria of vitamin D-deficient or vitamin D-replete rats. .

I1
Effects of calcium andphosphate ions on catalytic activity Calcium chloride and sodium phosphate were added to the incubation mixtures at the concentrations indicated. Each incubation contained 5 umol of P-450 from vitamin D-deficient rats.

DISCUSSION
Both the la-and 24-hydroxylations of 25(OH)D3 are catalyzed by mitochondrial P-450s found in the kidney and placenta (1-3, 11, 12). Although it is well established that the lahydroxylation predominates in vitamin D deficiency and the 24-predominates in control animals, very little is known about the biochemical mechanism of this regulation. Nor is it known whether two distinct forms of P-450 catalyze the two hydroxylations. As a first step toward a study of the regulation of these hydroxylases, a method has been developed for the partial purification and characterization of renal mitochondrial P-450s.
Cytochrome P-450 has been solubilized from kidney mitochondria and partially purified to a specific content of between 0.5 and 2.0 nmol/mg of protein. The mitochondrial content of P-450 from control rats is 0.016 nmol/mg of protein. Since the presence of cytochrome oxidase and other chromophores prevented the accurate measurement of P-450 in the intact mitochondria, the percentage of recovery has not been estimated. However, the value reported here of 0.016 nmol/mg of mitochondrial protein is close to the 10 nmol/g of mitochondrial protein reported by Pedersen et al. (14) and the 0.04 nmol/ mg of mitochondrial protein reported by Ghazarian et al. (13). A comparison of the P-450 fractions from control and vitamin D-deficient animals revealed no differences in the P-450 content of mitochondria, the CO-reduced difference spectral maxima, or protein profile on SDS electrophoresis. The solubilized P-450 catalyzed both the la-and 24-hydroxylations of

5(OH)D3
Hydroxylase 12999 25(OH)D3 at rates of 0.38 and 0.18 pmol/pmol of P-450/30 min with the enzyme from control animals, and 4.00 and 1.75 with enzyme from vitamin D-deficient animals. On the basis of the P-450 content of renal mitochondria reported herein, these values can be expressed as picomoles of product/mg of mitochondrial protein/min for comparison with values reported when rat kidney mitochondrial fractions were used (25). These calculated values are 0.18 and 0.09 for the la-and 24-hydroxylases in control animals and 2.1 and 0.93 in vitamin D-deficient animals. The corresponding values reported (25) with rat mitochondrial fractions are 2.9 for the la-hydroxylase in vitamin D-deficient rats and 0.29 for the 24-hydroxylase in control rats. Thus, the catalytic rates of the solubilized P-450 are close to those in intact mitochondria for these two reactions.
However, there was a major difference between the catalytic activity in the intact mitochondria and in the solubilized system, namely, the presence of both hydroxylases in approximately the same ratio in the solubilized P-450 from both vitamin D-deficient and vitamin D-replete animals. Previous reports have clearly demonstrated a reciprocal relationship between the two hydroxylases, with the la-predominating in vitamin D deficiency and the 24-in vitamin D-replete chickens. The significance of this difference between the solubilized system and intact mitochondria is not yet clear. Differences may be due to the removal of the enzyme from the mitochondrial membrane with loss of certain regulatory mechanisms, or may reflect a difference between the rat and chicken in the regulation of vitamin D hydroxylases.
Another unexpected but interesting observation of the present experiments was the demonstration that, during vitamin D deficiency, the quantity of P-450 in renal mitochondria is unaltered but the turnover number of the enzyme increases. The increase in the turnover number of P-450 may represent an activation of P-450, or the presence of new forms of the enzyme in vitamin D deficiency. Although in the present study no difference in the SDS electrophoretic profiie of P-450 fractions from control and vitamin D-deficient animals could be detected, the presence of distinct forms of P-450 in vitamin D deficiency cannot yet be ruled out and must await further characterization of the enzymes.
Attempts were made to characterize the solubilized P-450s from vitamin D-deficient and vitamin D-replete animals by a study of their K , values and the effects of the known modulators of P-450: SKF 525A and metyrapone. Vitamin D deficiency did not result in a change in the affinity of the enzymes for the substrate 25(OH)D3; the K , values for the la-and 24hydroxylases were 50 and 25 nM, respectively. These values are an order of magnitude lower than those reported when kidney slices or mitochondrial fractions are used (25). At least two factors may contribute to the lower K,, values reported herein: first, the removal of the mitochondrial membrane in which much of the substrate is sequestered when intact mitochondria are present; and, second, the possible removal of "inhibitory activity" in the form of a substrate binding protein which has been reported to be present in mammalian tissue (25).
No difference was seen between the P-450s from vitamin D-deficient or control animals in their response to SKF 525A which inhibited all the reactions by 70%. Metyrapone, on the other hand, had an effect only on the 24-hydroxylase from vitamin D-replete animals, where it caused a 2-fold stimulation.
Direct effects of calcium and phosphate ions on the lahydroxylase have been postulated on the basis of the effects which these ions elicit in chick kidney mitochondria. Calcium has been reported both to inhibit (27) and stimulate (26) the la-hydroxylase while phosphate was reported to inhibit the enzyme (27). In the present study, addition of phosphate ions to the solubilized P-450 system caused an inhibition of both the la-and 24-hydroxylases while calcium ions stimulated both reactions. Although no physiological significance can yet be attached to these results, it is interesting that the relative proportions of la,25(OH)2D3 and 24,25(OH)2D3 are not influenced by the presence of these ions.
It is not yet known whether the la-and 24-hydroxylations of 25(OH)D3 are catalyzed by two distinct forms of P-450. The difference in the K,,, values and response to metyrapone reported herein may be taken as evidence in support of this view. However, a definitive answer to this question awaits further purification and characterization of the enzymes.