Intracellular localization and properties of 3 beta-hydroxysteroid dehydrogenase/isomerase in the adrenal cortex.

The intracellular locations and properties of the 3/3hydroxysteroid dehydrogenase/isomerase were studied in subcellular fractions isolated from homogenates of rat adrenal and calf adrenal cortex. The goal was to determine whether the enzyme has a dual localization in the microsomes and mitochondria in the adrenal cortex. Mitochondrial fractions isolated from these homogenates were found to contain 26% of the total 3phydroxysteroid dehydrogenase/isomerase activity of the homogenate. Sucrose density gradient centrifugation demonstrated a co-migration of SD-hydroxysteroid dehydrogenase/isomerase activity with an adrenal mitochondrial marker enzyme. No a&factual attachment or transfer of 3P-hydroxysteroid dehydrogenase/isomerase activity to liver mitochondria could be demonstrated when liver and adrenal tissues were jointly homogenized. A portion of the 3P-hydroxysteroid dehydrogenase/isomerase activity associated with the mitochondrial fraction was active with matrix space NAD’. The microsomal enzyme was inactive in the absence of exogenous NAD’. When levels of intramitochondrial NAD+ were decreased either through reduction to NADH or by elution from calcium-swollen (10 mM CaCh) mitochondria, mitochondrial3/&hydroxysteroid dehydrogenase/isomerase activity was inhibited. 3fl-Hydroxysteroid dehydrogenase/isomerase activity in adrenal cortex microsomes was abolished by mersalyl (40 PM). The mitochondrial enzyme was unaffected by mersalyl except when the mitochondria were in a calcium-swollen condition. It is concluded that 3P-hydroxysteroid dehydrogenase/isomerase has a dual intracellular location in the adrenal cortex.

The intracellular locations and properties of the 3/3hydroxysteroid dehydrogenase/isomerase were studied in subcellular fractions isolated from homogenates of rat adrenal and calf adrenal cortex. The goal was to determine whether the enzyme has a dual localization in the microsomes and mitochondria in the adrenal cortex. Mitochondrial fractions isolated from these homogenates were found to contain 26% of the total 3phydroxysteroid dehydrogenase/isomerase activity of the homogenate.
Sucrose density gradient centrifugation demonstrated a co-migration of SD-hydroxysteroid dehydrogenase/isomerase activity with an adrenal mitochondrial marker enzyme. No a&factual attachment or transfer of 3P-hydroxysteroid dehydrogenase/isomerase activity to liver mitochondria could be demonstrated when liver and adrenal tissues were jointly homogenized.
A portion of the 3P-hydroxysteroid dehydrogenase/isomerase activity associated with the mitochondrial fraction was active with matrix space NAD'.
The microsomal enzyme was inactive in the absence of exogenous NAD'. When levels of intramitochondrial NAD+ were decreased either through reduction to NADH or by elution from calcium-swollen (10 mM CaCh) mitochondria, mitochondrial3/&hydroxysteroid dehydrogenase/isomerase activity was inhibited.
3fl-Hydroxysteroid dehydrogenase/isomerase activity in adrenal cortex microsomes was abolished by mersalyl (40 PM). The mitochondrial enzyme was unaffected by mersalyl except when the mitochondria were in a calcium-swollen condition.
(2), testes (3), and placenta (4). The enzyme is involved in steroidogenesis and catalyzes the oxidative conversion of steroids with the A5,3/3-hydroxy structure to the A4,3-keto configuration; e.g. pregnenolone + NAD' -+ progesterone + NADH + H+. The intracellular location of HSD in bovine adrenal cortex tissue was examined by Beyer and Samuels in 1956 (1). Their initial enzyme distribution study revealed HSD activity in the microsomes (39%) as well as the nuclear (26%) and mitochondrial (14%) fractions. The HSD activity found in the nuclear fraction could be reduced to zero by repeated washing of the pellet through resuspension and recentrifugation.
Applying the same technique to the microsomes and mitochondria was without effect and in these organelles the enzyme was removable only through detergent solubilization.
On the basis of phase-contrast microscopic examination of the mitochondria Beyer and Samuels (1) concluded that HSD activity in this fraction could be accounted for by microsomal contamination and that in intact adrenal cortex tissue the HSD was a constituent only of the endoplasmic reticulum.
These reports included both enzyme distribution studies and the indirect evidence that mitochondria from steroid-producing tissues incubated with cholesterol produce progesterone as well as pregnenolone.
Additionally, when rat adrenal cortex tissue slices were stained for HSD activity, the product of the reaction was found in the mitochondria as well as the smooth endoplasmic reticulum (15). Nevertheless, other investigators disputed the presence of HSD in the mitochondria either as microsomal contamination (16), or as an enzyme redistribution artifact that occurred during homogenization (17). Consequently, the exact intracellular localization of HSD has remained unsettled.
In this report we show that adrenal cortex mitochondria contain a substantial and consistent amount of the total homogenate HSD activity, and that this activity cannot be removed from the mitochondria either by repeated washings, or by sucrose density gradient centrifugation.
Moreover, we show that the HSD activity associated with the mitochondria has characteristics which clearly differentiate it from the microsomal enzyme. After 15 min incubation the reaction was stopped by the addition of 2 ml of heptane followed by a 10-s blending on a Vortex mixer. The aqueous and organic layers were separated by centrifugation, and the absorbance due to the A4,3-keto configuration of progesterone was measured at 233 nm. The product of the incubation extracted into the organic solvent was identified as progesterone by thin layer chromatography. The extraction efficiency of progesterone standards from Medium B was 88% -+ 2 (X + S.E.), n = 26. Progesterone concentration was estimated using a molar absorptivity of 17,000 at 240 nm with the steroid dissolved in methanol (21). In heptane the absorbance maxima of progesterone is shifted to 233 with no change in the extinction coefficient.
Production of progesterone was linearly related to the enzyme concentration.

Measurement of HSD Activity by Pregnenolone
Disappearance-The procedure of Philpott and Peron (22)  Assays-The following marker enzyme assays were used in the characterization of cell fractions obtained by differential centrifugation: HSD and the steroid 21-hydroxylase' as markers for ' The assay medium for steroid 21-hydroxylase activity included 1 mM KCN, and a NADPH-regenerating system that consisted of 0.5 mM NADPH, 3.3 mM glucose-6-P, and 1 unit of glucose-6-P dehydrogenase.

microsomes
(23); cytochrome c oxidase as a marker for mitochondria (24); and glucose-6-PO4 dehydrogenase as a marker for cytosol (25 to 37°C in a shaking water bath. Five milligrams of mitochondrial protein were added and allowed to equilibrate for 2 min before the reaction was initiated by the addition of 2 mM CaC12. After various times, Ca*+ uptake was stopped by the addition of 20 mM EGTA (32), followed by chilling in ice. The complete reaction mixture, usually about 2.5 ml, was then added to the top of a 15 to 55% w/w continuous sucrose gradient and centrifuged for 3 h at 25,000 rpm in a SW 25.1 swinging bucket rotor in a Spinco model L centrifuge. Fractions were collected from the bottom of the tube and were either frozen or assayed immediately.
Sucrose concentrations were determined using a Goldberg T/C* refractometer. Ca" uptake associated with O2 consumption was measured using a Clark oxygen electrode.
Calf adrenal cortex mitochondria (4 mg of protein) were added to an oxygen-saturated medium contained in a l-ml oxygen electrode chamber.
The incubation medium used was that developed by Carafoli and Lehninger (33) and contained 225 m&r mannitol, 55 mM sucrose, 5 mu succinate, 5 mM KHzPO,, 0.1% w/v bovine serum albumin, and 10 mM Tris/HCl, pH 7.4. Oxygen consumption was recorded and the effect of Ca*' (700 pM) on mitochondrial respiration was observed. After incubation at 30°C the complete mixture was layered atop a 15 to 55% w/w continuous sucrose density gradient.
Centrifugation, fraction collection, and assay were as described for the massive Ca'+-loading experiment. Sucrose Density Gradient Centrifugation for the Separation of Liver and Adrenal Mitochondria-Liver and adrenals were removed from rats killed by decapitation and placed in ice-cold 0.25 M sucrose. The adrenal glands were trimmed of adherent fat, blotted on tissue paper, and weighed.
The liver was rinsed of blood prior to blotting and weighing.
The cleaned adrenals (150 mg) were combined with an equal amount of liver tissue, minced together with scissors, and homogenized in Medium A. The homogenization procedure and isolation of the mitochondria were as described for the adrenal alone. The resulting mitochondrial pellet, which contained both liver and adrenal mitochondria, was washed twice by resuspension and recentrifugation (10 min at 5200 x g). The final washed mitochondria were resuspended in 2 ml of Medium A and added to the top of a 15 to 55% w/w continuous sucrose gradient. Centrifugation, fraction collection, and assay were as described for Ca"'-loading experiments.

Enzyme
Distribution- Table  I lists the results of three experiments on the subcellular distribution of HSD in rat adrenal cortex tissue. As shown, adrenal cortex mitochondria washed four times contain 26% of the total homogenate HSD activity. Contribution to the mitochondrial HSD activity by microsomes was judged not to be a significant factor. When the mitochondria were assayed for microsomal contamination only 6% of the microsomal marker enzyme, 21-hydroxylase, was found. The ratio of HSD activity to 21-hydroxylase activity was calculated in both subcellular fractions. This is an important comparison, for if mitochondrial HSD activity resulted simply from microsomal contamination one should expect the ratio of HSD to 21-hydroxylase in the mitochondria to parallel its presumed microsomal source closely. In the microsomes the ratio of the two enzymes is 2.&l whereas in the mitochondria it is 19.6:1. Moustafa and Koritz (16) suggested that this disparity is the result of microsomal heterogeneity, with microsomes richer in HSD activity selectively contaminating the mitochondrial fraction. To our knowledge differences in enzyme distributions have not been demon-Adrenal Cortex 3/3-Hydroxysteroid Dehydrogenase/Isomerase &rated in adrenal cortex microsomes. In a preliminary report Kream and Sauer (14) suggested that rat adrenal cortex mitochondria could utilize the NAD' in the matrix space as a cofactor for HSD activity. Using this information we reexamined the intracellular distribution of HSD, this time in the absence of added NAD' (Table II). HSD activity, active with endogenous NAD', is clearly associated with the mitochondria.
Even HSD activity found in the crude nuclear and microsomal fractions appears to be of mitochondrial origin. In the nuclear and microsomal fractions the percentage of mitochondrial contamination, as judged by cytochrome c oxidase activity, is almost identical with the percentage of HSD activity (with endogenous NAD+). Two separate controls were employed to ensure that HSD activity in the absence of added NAD+ resulted solely from the utilization of matrix space NAD+. In one, glucose-6-PO4 and NAD'-linked glucose-6-PO4 dehydrogenase were added to reduce available NAD+ to NADH. In the other experiment NADase (N. crassa) was added. Each one of these enzyme systems abolished or severely reduced HSD activity in microsomes supplemented with NAD+ (not shown) and therefore should have inhibited HSD activity in the mitochondrial fraction if the activity were due to microsomes and exogenous NAD'. The results of these experiments are also shown in Table II. Neither system influenced the mitochondrial HSD distribution, suggesting that the endogenous NAD+ used by the mitochondria remained enclosed within and protected by the inner membrane for the duration of the incubation and did not exit and become available to a microsomal contaminant.
(31) have shown that intact mitochondria are able to accumulate large amounts of Ca2+ in an energy-dependent process. Mitochondria sequestering Ca2' in this manner become more dense and can be clearly differentiated from control mitochondria when both are subjected to sucrose density gradient centrifugation. This aspect of mitochondrial function was chosen to further define the nature of the mitochondrial HSD. Sucrose density gradient centrifugation is a well recognized technique for increasing the homogeneity of subfractionated organelles. Therefore, combining this technique with the known mitochondrial capacity of Ca2+ uptake could further differentiate between mitochondrial HSD activity and HSD activity thought to originate through microsomal contamination.
In these experiments calf adrenal cortex mitochondria were first incubated under conditions for massive Ca'+-loading (see "Methods") and then centrifuged to equilibrium in sucrose density gradients.    1. a, photograph of the isopycnic centrifugation of Cal+loaded calf adrenal cortex mitochondria in a 15 to 55% w/w continuous sucrose density gradient. Calf adrenal cortex mitochonclria (5 mg of protein) were incubated under conditions described under "Methods" for massive Ca*'-loading and were allowed to accumulate calcium for 0 (A), 45 (B), and 90 (C) s. Centrifugation was for 3 h at 25,090 rpm in a SW 25.1 swinging bucket rotor. An inhibitor of oxidative phosphorylation (oligomycin, 1 pg/ml) and of electron transport (antimytin A, 1 pg/ml) added to the mitochondria during preincubation and prior to the addition of Ca2+ inhibited movement down the gradient shows the distribution of HSD activity and succinate-cytochrome c reductase among fractions collected from the tubes containing mitochondria Ca"-loaded for 0 and 90 s. The mitochondria incubated with Ca*+, as measured by the succinate-cytochrome c reductase marker, have shifted to a new isopycnic density position. More importantly, HSD activity also moved to the same new position.
The capability of mitochondria to take up smaller amounts of Ca2+ using energy derived from respiration has also been documented (33). Calcium accumulation by this pathway also increased mitochondrial density. In these experiments calf adrenal cortex mitochondria, incubated in an oxygen electrode chamber, were treated with two successive additions of CaC12. Each addition resulted in a stimulation of respiration.
In the control experiment an identical amount of mitochondria were treated with a single addition of ADP. At the conclusion of the incubation each reaction mixture was removed, placed on a sucrose density gradient and centrifuged as described under "Methods." Fig. 2 shows the distribution of HSD activity and succinate-cytochrome c reductase among fractions collected from the sucrose density gradients. Mitochondria to which Ca*+ has been added had a slightly increased isopycnic density position over that of the controls, and there was an attendant shift in HSD activity to a greater density. The results of these experiments demonstrate that mitochondrial HSD activity remains attached to the mitochondria during sucrose density gradient centrifugation; when adrenal cortex mitochondria were induced to a new equilibrium density position as a result of Ca*' accumulation, the HSD activity remained with the mitochondria.
Test for Microsomal HSD Adsorption to  suggested that the HSD associated with the mitochondria of bovine adrenal cortex tissue was located in the outer mitochondrial membrane. They speculated that this was probably not the normal location for the enzyme in uiuo but that microsomal HSD, solubilized during homogenization, had become attached to .the outer mitochondrial membrane. To test if HSD were transferred among organelles during homogenization, or fractionation, or both, we made homogenates of rat adrenal and rat liver tissues combined. This was performed to ascertain whether liver mitochondria could acquire HSD activity from microsomes of adrenal cortex tissue. A mitochondrial fraction containing both liver and adrenal mitochondria was isolated from the combined tissue homog-(not shown). b, the distribution of HSD and succinate-cytochrome c reductase activities versus fraction density following isopycnic sucrose density gradient centrifugation of the 0 (M)-and 90 (M)s Ca'+-loaded calf adrenal cortex mitochondria shown in a. HSD activity was measured as described under "Methods," under "Measurement of HSD Activity by Progesterone Formation" except that the incubation medium consisted of 50 mM Tris/HCl pH 7.4, 5 mM EGTA, 1 mM KCN, 0.012% sodium deoxycholate, 0.05 mM NAD', 1 mM p.yruvate, 1 unit of lactic dehydrogenase, and the incubation was and Ca'+-loaded (M) calf adrenal cortex mitochondria. Mitochondria were loaded with calcium by the procedure described under "Methods," under "Ca*+ Loading and Sucrose Density Gradient Centrifugation of Adrenal Cortex Mitochondria." Calf adrenal cortex mitochondria (4 mg of protein) were incubated in a l-ml oxygen electrode chamber and the rate of mitochondrial respiration was stimulated by two additions of 350 nmol of CaC12. These Ca'+-treated mitochondria and mitochondria from a control incubation which received an ADP addition (300 nmol, no CaCb) were centrifuged to equilibrium in a 15 to 55% w/w continuous sucrose density gradient for 3 h at 25,000 rpm in a SW 25.1 swinging bucket rotor. HSD activity was measured under the conditions described in the legend for Fig. lb. enate by differential centrifugation.
Resolution of this fraction into liver and adrenal mitochondrial bands was accomplished by sucrose density gradient centrifugation.
This was possible because of the different equilibrium densities of liver (p P 1.20) and adrenal cortex mitochondria (p = 1.15). The results of this procedure as well as the enzymatic profiles of individually isolated rat adrenal and rat liver mitochondria are shown in Fig. 3. As indicated by the bottom profile, no detectable HSD activity was transferred to the heavier liver mitochondria. While the results of this experiment are not absolute proof against HSD transfer between microsomes and mitochondria in adrenal cortex tissue, the complete absence of enzyme transfer to liver mitochondria makes it difficult to see how over one-fourth of the total homogenate HSD activity (see Table I) found in adrenal cortex mitochondria could arise by the same process.

Adrenal Cortex 3jSHydroxysteroid
Dehydrogenase/Isomerase Effect of Reduction of Intramitochondrial NAD(P)+ on Mitochondrial HSD Activity-Mitochondrial HSD activity was found to be decreased by over 75% by the addition of citric acid cycle Substrates (14) suggesting that reduction of the pyridine nucleotides in the matrix space inhibited mitochondrial HSD. To test this mechanism in more detail we measured the mitochondrial pyridine nucleotide content in rat adrenal cortex mitochondria in the presence of an uncoupler of oxidative phosphorylation ("1799"), and in the presence FIG. 3. The distribution of HSD and succinate-cytochrome c reductase activities uer.sus fraction density following isopycnic sucrose density gradient centrifugation of rat adrenal mitochondria (A), rat liver mitochondria (B), and combined rat adrenal and liver mitochondria (C). The enzymatic profiles shown in C resulted from mitochondrial fractionation of a homogenate of rat adrenal and liver. Mitochondrial fractions isolated from the three different homogenates were layered on separate 15 to 55% w/w continuous sucrose density gradients and centrifuged for 3 h at 25,COO rpm in a SW 25.1 swinging bucket rotor. Fractions were collected and assayed for HSD and succinate-cytochrome c reductase activity as described in the legend for Fig. lb. of pregnenolone as well as pregnenolone + succinate and aketoglutarate.
The measured NADP+ + NADPH and NAD' + NADH values agreed well with the values reported by Purvis et al. (34). Mitochondria (1.1 mg of protein/ml) incubated with 4 pM "1799" (pregnenolone was not added) contained 2.2, 1.1, 2.9, and 0 nmol/mg of protein of NADP', NADPH, NAD', and NADH, respectively. Mitochondria incubated with 0.1 XIIM pregnenolone contained 2.5, 1.1, 2.9, and 0 nmol/mg of protein of NADP+, NADPH, NAD+, and NADH, respectively. Mitochondria incubated with 0.1 mM pregnenolone plus 10 mu a-ketoglutarate and 10 mu succinate contained 0, 2.5, 0.4, and 2.7 nmol/mg of protein of NADP', NADPH, NAD+, and NADH, respectively. Pregnenolone did not alter the oxidation-reduction state of the pyridine nucleotides probably because the rate of HSD activity (3.0 nmol of progesterone. min-' . mg-') is only about one-seventeenth that of the rate of respiration linked to NADH oxidation measured under these conditions (18). Fig. 4 shows that a-ketoglutarate and succinate inhibited progesterone formation by over 80%. Under these conditions only about 0.4 nmol of NAD+/mg of protein remained available to the HSD. Fig. 4 also shows that malonate, a competitive inhibitor of succinic dehydrogenase, reverses the inhibition of mitochondrial HSD activity due to succinate. This result is further evidence for the mitochondrial location of the HSD activity. These results suggest that the rate of mitochondrial HSD activity is dependent on the oxidation-reduction state of intramitochondrial NAD+, the dynamics of which are regulated through substrate selection and availability as well as energy requirements of the cell. Moustafa and Koritz (16) examined the intracellular distribution of HSD in the rat adrenal and reported that mitochondria contain 13% of the total homogenate HSD activity, a value somewhat smaller than our result (Table I). The assay procedure used by Moustafa and Koritz (16) employed an NAD+-regenerating system consisting of lactic dehydrogenase and pyruvate. These agents removed the inhibitor NADH and sustained V,,,,, HSD activity in microsomes (20). However, pyruvate is capable of reducing intramitochondrial NAD' which in turn could inhibit mitochondrial HSD activity. As Fig. 4 shows, the addition of pyruvate to a preparation of intact adrenal cortex mitochondria did inhibit mitochondrial HSD activity by about 50%. Thus, reduction of intramitochondrial NAD+ following addition of pyruvate may explain  5 (right). The effect of Ca" and NAD' on mitochondrial were incubated for 0, 5, 10, and 15 min in Medium B containing 2 mM HSD activity. Duplicate samples containing rat adrenal mitochondria potassium phosphate and 0.1 mu pregnenolone. Some samples also (119 pg of protein) were incubated in Medium B containing 0.1 mM contained the substrates as indicated. Assays were performed in pregnenolone for 0, 7L/2, and 15 min. Samples contained either no duplicate and the rate of progesterone formation was calculated by a additions or the substances indicated. The rates of progesterone least squares method (correlation coefficients were between 0.934 to formation were calculated by a least squares method (correlation 0.998). The values shown represent means + S.E. of from three to coefficients were 0.929 to 0.999).
Adrenal Cortex 3/3-Hydroxysteroid Dehydrogenase/Isomerase 6629 the lower percentage of HSD activity found in the mitochondrial fraction by these investigators. Rat adrenal cortex mitochondria can form ll/?-hydroxyprogesterone from progesterone in the presence of citric acid cycle substrates (35). This steroid, which is more polar and therefore is not extracted by heptane as well as is progesterone, could lead to errors in measurement of HSD activity based on extractable progesterone. Therefore, we also assessed the effect of citric acid cycle substrates on mitochondrial HSD activity as measured by the rate of pregnenolone disappearance (22). No differences were noted in the results obtained by either assay method (not shown). Citric acid cycle substrates inhibited mitochondrial HSD activity measured by pregnenolone disappearance as well as by progesterone formation.
Effect of Mitochondrial Swelling on HSD Actiuity-Calcium is known to promote swelling of the mitochondrial inner membrane (36). Through this action otherwise impenetrable metabolites have free access to, and exit from, the mitochondrial matrix area. For example, adrenal cortex mitochondria in the presence of 10 mM CaClz can utilize exogenous NADPH to support high rates of steroid hydroxylation (37). Pyridine nucleotides do not normally penetrate intact mitochondria. As Fig. 5 shows, rat adrenal cortex mitochondria swollen by 10 IIIM CaClz lose most of their HSD activity (active with endogenous NAD'). We interpret the loss in activity to be the result of exit and subsequent dilution of intramitochondrial NAD+, a conclusion reinforced by the observation that HSD activity can be restored to the swollen mitochondria by the addition of exogenous NAD+. When NAD' was added to "intact" adrenal cortex mitochondria HSD activity was enhanced 2-to 3-fold. The reason for this increase in HSD activity is not completely understood. Certainly, a part of the HSD activity (23% of the increment as calculated from Tables  I and II) dependent on added NAD' is due to microsomal contamination.
Microsomes plus a contribution from damaged mitochondria could be the reason for this added HSD activity. Further research on mitochondrial HSD activity in the presence of added NAD+ will be presented in a subsequent paper.
Mersalyl is an organic mercurial that has been used as a tool to demonstrate the presence and location of mitochondrial inner membrane carrier systems (38). Because of its charge and size mersalyl does not rapidly penetrate an intact mitochondrial inner membrane. Therefore, loss in enzymatic or transport activity through the use of mersalyl suggests that the active site is on the outside of the inner membrane.  6. A, effect of mersalyl on rat adrenal microsomal HSD activity in the presence or absence of Ca*+. Rat adrenal microsomes (43 pg of protein) were incubated for 15 min in Medium B containing 0.5 mM NAD' and 0.1 mM pregnenolone. Mersalyl was present at the concentration indicated on the abscissa. The points connected by the dashed line represent incubations that also contained 10 IIIM CaC12. B, effect of mersalyl on rat adrenal mitochondrial HSD activity in the presence or absence of Ca'+. Rat adrenal mitochondria (119 pg of may be interpreted to mean that succinate, by reducing intramitochondrial NAD', inhibited the mitochondrial HSD and decreased conversion of pregnenolone to progesterone. The importance of mitochondrial HSD in controlling the product of cholesterol side chain cleavage that leaves the mitochondrion cannot be determined at this time. However, HSD is located at a branch point in steroidogenesis through which steroid precursors for gluco-and mineralocorticoid synthesis must pass. Therefore, a mitochondrial location for HSD in close proximity to the initial rate limiting, cholesterol side chain cleavage step, could be a deciding factor in determining the ultimate steroid products of the adrenal cortex.
Finally, because of their different intracellular locations and properties, mitochondrial inner membrane HSD and microsomal HSD may be considered as isoenzymes. The proteins may be coded for by different genes and be inserted into their respective membrane sites by different mechanisms. Consequently, it will be of great interest to determine whether the two HSD apoenzymes are identical or different.