The effect of diabetes, nutritional factors, and sex on rat liver and kidney mevalonate kinase, mevalonate-5-phosphate kinase, and mevalonate-5-pyrophosphate decarboxylase

Isopycnic sucrose gradient separation of rat liver organelles revealed the presence of two distinct branched-chain α-keto acid decarboxylase activities; a mitochondrial activity, which decarboxylates the three branched-chain α-keto acids and requires CoA and NAD⁺ and a cytosolic activity, which decarboxylates α-ketoisocaproate, but not α-ketoisovalerate, or α-keto-β-methylvalerate. The latter enzyme does not require added CoA or NAD⁺. Assay conditions for the cytosolic α-ketoisocaproate decarboxylase activity were optimized and this activity was partially characterized. In rat liver cytosol preparations this activity has a pH optimum of 6.5 and is activated by 1.5 m ammonium sulfate. The decarboxylase activity has an apparent K_m of 0.03 mm for α-ketoisocaproate when optimized assay conditions are employed. Phenylpyruvate is a very potent inhibitor. α-Ketoisovalerate, α-keto-β-methylvalerate, α-ketobutyrate, and α-ketononanoate also inhibit the α-ketoisocaproate decarboxylase activity. The data indicate that the soluble α-ketoisocaproate decarboxylase is an oxidase. Rat liver cytosol preparations consumed oxygen when either α-ketoisocaproate or α-keto-γ-methiolbutyrate were added. None of the other α-keto acids tested stimulated oxygen consumption. 1-¹⁴C-Labeled α-keto-γ-methiolbutyrate is also decarboxylated by cytosol preparations. The α-ketoisocaproate oxidase was purified 20-fold from a 70,000g supernatant fraction of a rat liver homogenate. In these preparations the activity was increased 4-fold by the addition of dithiothreitol, ferrous iron, and ascorbate. The major product of this enzyme activity is β-hydroxyisovalerate. Isovalerate is not a free intermediate in the reaction. The data indicate an alternative pathway for metabolism of α-ketoisocaproate which produces β-hydroxyisovalerate.     

THE BIOSYNTHESIS OF 3-METHOXY-4-HYDROXYPHENYLGLYCOL SULPHATE BY LIVER AND BRAIN

—3-Methoxy-4-hydroxyphenylglycol (MHPG) formed a sulphate conjugate when incubated with ATP, Mg²⁺ ions, Na₂³⁵SO₄ and the high-speed supernatant preparations of rabbit or rat brain. The same reactions could be catalysed by similar enzyme preparations from liver. The sulphated product was separated and identified by paper chromatography. On acid hydrolysis, it released both Na₂³⁵SO₄ and the free glycol. The measurement of this labelled sulphate was used as a specific assay procedure for determining the overall sulphoconjugatory process. The pH optimum of the reaction is 7.8. For rabbit brain, the K_m for Na₂SO₄ determined for the activating system is 3.6 × 10⁻⁴m , and that for MHPG for the sulphotransferase reaction is 1.05 × 10⁻⁴m . The specific enzyme activity, expressed as nmol ³⁵SO₄ incorporated/h/mg protein for a 30-min assay is as follows: rat brain, 2.8; rabbit brain, 1.6; rat liver, 33.4and rabbit liver, 15.0. Dithiothreitol at 3 mm concentration had no significant effect on the sulphation of MHPG in all these preparations.     

Mechanism and stereospecificity of alpha-hydroxylation of lignoceric acid in rat brain.

Cerebronic acid (2-hydroxytetracosanoic acid) is the major fatty acid component of cerebrosides and sulfatides in mammalian brain. Our previous communication demonstrated the synthesis of cerebronic acid from lignoceric acid (tetracosanoic acid) by a rat brain preparation in the presence of molecular oxygen and a reduced pyridine nucleotide (Hoshi, M., and Kishimoto, Y. (1973) J. Biol. Chem., 248, 4123–4130). The present'studies on the conversion of (RS)-[2-³H]-, (RS)-[3-³H]-, (R)-[2-³H]-, and (S)-[2-³H]lignoceric acids to cerebronic acid by rat brain preparations establish that the pro-R hydrogen at the α-carbon of lignoceric acid is replaced by a hydroxyl group with overall retention of configuration.     

Purification of human liver cytochrome P-450 and comparison to the enzyme isolated from rat liver

Human liver cytochrome P-450 was isolated from autopsy samples using cholate extraction and chromatography on n-octylamino-Sepharose 4B, hydroxylapatite, and DEAE-cellulose gels. Purified preparations contained as much as 14 nmol cytochrome P-450 mg^?1 protein, were free of other hemoproteins, and were active in the mixed-function oxidation of d-benzphetamine and 7-ethoxycoumarin when coupled with either rat or human liver NADPH-cytochrome P-450 reductase. Some of the preparations were apparently homogeneous as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis; apparent subunit M_rs estimated for several preparations were 53,000 or 55,500. The amino acid composition of one preparation was determined and found to resemble those of rat liver cytochromes P-450, although some variations were noted. Rabbit antibodies raised to phenobarbital-treated rat liver cytochrome P-450 were more effective in inhibiting d-benzphetamine N-demethylase activity in human liver microsomes than were antibodies raised to 3-methylcholanthrene-treated rat liver cytochrome P-450. These antibodies also inhibited benzo(a)pyrene hydroxylation in human liver microsomes, although the inhibition patterns did not follow a general pattern as in the case of benzphetamine demethylase activity. Microsomes prepared from three different human liver samples were more effective in eliciting complement fixation with antibodies raised to phenobarbitalthan to 3-methylcholanthrene-treated rat liver cytochrome P-450. Complement fixation in such systems appears to result from similarity of certain rat and human liver cytochrome P-450 antigenic determinants, as fixation could be inhibited by removal of cytochrome P-450-directed antibodies from the total immunoglobulin population and purified human cytochrome P-450 was more effective (on a protein basis) than liver microsomes in producing fixation. Human liver microsomes prepared from five different individuals all produced ≥ 90% complement fixation, but variations were observed in the fixation curves plotted either versus microsomal protein or versus spectrally detectable microsomal cytochrome P-450.These results indicate that human liver microsomal cytochromes P-450 can be isolated using modifications of techniques developed for laboratory animals and that human and rat liver cytochromes P-450 share certain features of structural, functional, and immunological similarity. The available data suggest the existence of multiple forms of human liver microsomal cytochrome P-450, but possible artifacts associated with the use of autopsy samples suggest caution in advancing such a conclusion.     

GABA and glutamate fluxes in developing nerve endings

Radiolabeled GABA and glutamate transport into 7 day, 14 day and adult cortical nerve ending preparations was examined. Transport was measured at several Na⁺ concentrations, 19, 27, 43 and 121 mM, and at two temperatures, 15 and 30°C. K_m and V_max values were calculated for all experimental conditions by means of Wilkinson (1961) analysis. A comparison of the day 14 and adult data shows higher K_m values at all Na⁺ concentrations on day 14 for both GABA and glutamate transport. In addition, the temperature dependence of transport was attenuated in the day 14 preparation. Finally, the specificity of GABA transport, as measured by the use of the transport inhibitors β-alanine and 2,4-diaminobutyric acid, was not different between the day 14 and adult preparations. Overall, it is concluded that both GABA and glutamate transport into day 14 nerve endings behave as if “adult” transporter molecules were existing in a more fluid lipid environment, which is the situation found in synaptic membranes prepared from day 14 nerve endings (Hitzemann and Johnson, 1983).Glutamate and GABA transport into 7 day nerve endings is complex and shows marked differences from the day 14 and adult data. Day 7 GABA transport was significantly more sensitive to β-alanine inhibition. Day 7 transport was more sensitive to Na⁺ manipulation and the temperature dependent kinetics show unique Na⁺ effects not seen in the day 14 or adult preparations. For example, at 19 mM Na⁺, 7 day glutamate transport was more temperature dependent than adult transport but as the Na⁺ concentration was increased the reverse was true. The opposite situation for temperature-Na⁺ effects was seen for GABA transport. Finally, no Ca⁺²-dependent component of GABA release could be found in 7 day nerve endings while a significant component was found at day 14. Overall, it is concluded that both glutamate and GABA fluxes in 7 day nerve endings differ both qualitatively from that seen in both day 14 and adult nerve endings.     

Rat brain aspartate β-decarboxylase. A comparative study with the liver enzyme

Aspartate β-decarboxylase (AspD), which catalyses the β-decarboxylation of aspartate (Asp) to alanine (Ala), was found in significant quantities only in the brain, kidney and liver. This enzyme has an optimum pH at 7.4. Addition of exogenous pyridoxal 5′-phosphate did not increase enzyme activity presumably because of firmly bound cofactor. However, aminooxyacetic acid is a potent inhibitor.There is an apparent 8-fold variation in AspD in the seven brain regions studied, with the highest activities in the cortex and the lowest in the striatum and hippocampus. In the presence of α-ketoglutarate, the production of ¹⁴CO₂ from [¹⁴C]Asp may no longer represent AspD activity due to active transamination of Asp, presumably by aspartate aminotransferase, to oxaloacetate. Under such conditions, comparable AspD activities were observed in all seven brain regions.Kinetic analysis showed that the liver and kidney enzymes have identical affinity for Asp (K_m = 3.5 mM) while the brain enzyme has a higher affinit (K_m = 1.3 mM). The V_max values obtained indicated that the enzyme populations in liver, kidney and brain are in the ratio 18:4:1. Various amino acids were found to inhibit both brain and liver AspD. Serine, however, activated the liver enzyme but inhibited competitively the kidney and brain enzymes. These results indicate that AspD may exist as two or more isozymes.     

Galactosyltransferase activities in pancreas, liver and gut of the developing rat

Two galactosyltransferase activities (1 and 2) were measured in the pancreas, liver and gut of the developing rat embryo. 1. N-Acetylglucosamine:Galactosyltransferase. UDP [¹⁴C]galactose + N-acetylglucosamine → [¹⁴C]galactosyl-β-(1 → 4)-N-acetylglucosamine + UDP. 2. N-Acetylgalactosamine-protein:Galactosyltransferase. UDP [¹⁴C]galactose + N-acetylgalactosamine-protein → [¹⁴C]galactosyl-β-(1 → 3)-N-acetylgalactosamine-protein + UDP. Galactosyltransferases 1 and 2 increased in the pancreas, about 10- and 40-fold in specific activity, respectively, from 11 to 12 days in utero to birth. During this period the activities of both transferases in the liver were somewhat variable, but showed no definite trend. A drop in the level of galactosyltransferase 1 in the pancreas occurred at birth or shortly thereafter. The “Golgimarker” enzyme for liver, galactosyltransferase 1, may be absent or present at low levels in adult rat pancreas.Zymogen granule membrane preparations apparently are devoid of these galactosyltransferase activities. Bromodeoxyuridine, which inhibits the development of the synthetic capability of the specific exocrine proteins, had essentially no effect on the normal accretion of the galactosyltransferase activities in organ cultures of pancreatic rudiments from 13-day rat embryos.     

The effect of 3-mercaptopicolinic acid on phosphoenolpyruvate carboxykinase (GTP) in the rat and guinea pig.

In vitro, 3-mercaptopicolinic acid inhibited phosphoenolpyruvate carboxykinase activity in supernatant fractions of liver, kidney cortex, and adipose tissue obtained from fasted rats. 3-Mercaptopicolinic acid also inhibited enzymatic activity in the mitochondrial and supernatant fractions of liver obtained from fasted guinea pigs. In the fasted rat, the oral administration of 3-mercaptopicolinic acid increased liver carboxykinase activity even though the blood glucose concentrations decreased. Kidney cortex carboxykinase decreased while adipose tissue enzyme was unchanged. In the fasted guinea pig, the oral administration of 3-mercaptopicolinic acid lowered blood glucose concentrations but had no effect on liver mitochondrial or supernatant carboxykinase activity. The elevation in rat liver enzymatic activity appears to be due to protein synthesis, since the concurrent administration of cycloheximide prevents the increase in enzyme activity. 3-Mercaptopicolinic acid appears to be noncompetitive with respect to Mn²⁺.     

Incorporation of chorismic acid and 4-aminobenzoic acid into the 4-hydroxyaniline moiety of N-(γ-L-glutamyl)-4-hydroxyaniline in Agaricusbisporus

Agaricus bisporus contains the unique aniline derivative, $\text{N-(γ-}\text{L}\text{-}\text{glutamyl}\text{)-4-}\text{hydroxyaniline}$. ¹⁴C-labelled chorismic acid was quantitatively incorporated into the 4-hydroxyaniline moiety of this aniline derivative, whereas ¹⁴C-labelled prephenic acid and anthranilic acid were not incorporated into 4-hydroxyaniline. These observations indicate the branch point of the biosynthetic route of 4-hydroxyaniline in the shikimic acid pathway to be chorismic acid. Moreover, 4-aminobenzoic acid proved to be an effective precursor of 4-hydroxyaniline.     

Acetyl transport mechanisms in the nervous system. The oxoglutarate shunt and fatty acid synthesis in the developing rat brain

Abstract Radioactive acetyl groups and lipids are produced from dl -[5-¹⁴C]glutamate. Degradation studies indicate that approximately 90 per cent of the radioactivity is localized in the original carboxyl groups of the two carbon unit. Since these results are shown not to be due to a ¹⁴CO₂ fixation, it is concluded that the oxoglutarate shunt as an acetyl group transport system is functional in brain. The highest ratio of fatty’acid activity/CO₂ activity in this pathway is found in the newborn rat brain and steadily decreases with development. This pattern is observed with incubations of brain slices with labelled glutamate or citrate and is similar to the changes observed in the activity of the citrate cleavage enzyme with brain maturation. In contrast to the previous studies with liver preparations, the conversion of [2-¹⁴C]- and [5-¹⁴C]glutamate to fatty acids is relatively small. This is particularly true during the period of maximal lipid synthesis.     

Isolation and characterization of PZ-peptidase in developing rat brain

Changes in the activity of a collagen peptidase, PZ-peptidase, acting on a synthetic substrate [4-Phenylazobenzyloxycarbonyl(PZ)-l-Pro-l-Leu-Gly-l-Pro-l-Arg] for bacterial collagenase were examined in developing rat brain regions. The hypothalamus, pons-medulla, colliculi, cerebellum, ceerbrum, midbrain and pituitary gland were studied in rats ranging in age from 1 week to adult; PZ-peptidase activity continuously decreased with maturation in all of the brain regions examined except the hypothalamus. The pituitary gland showed the highest activity in all of the brain regions. PZ-peptidase activity in crude mitochondrial and supernatant fractions from rat whole brain had an optimum pH between 7.5–8.0. It was strongly inhibited by p-chloromercuribenzoic acid, N-ethylmaleimide or EDTA. whereas iodoacetic acid did not affect the enzyme activity. Among various metal ions, the enzyme activity was inhibited by Zn⁺² or Cu⁺² but not by Mn⁺², Ca⁺², Mg⁺² or Na⁺. There is no inhibition of the activity by serine protease inhibitors, including diisopropylfluorophosphate and phenylmethylsulphonyl fluoride. An approximate molecular weight of this enzyme was estimated to be 68,000 by gel filtration. Since these properties of rat brain PZ-peptidase were similar to those of other peripheral PZ-peptidases, we suppose that PZ-peptidase in the brain may be the same molecule as the enzyme which hydrolyses collagen peptides in peripheral tissues, but it may have some different physiological roles.     

Control of glutamate oxidation in brain and liver mitochondrial systems

1. Glutamate oxidation in brain and liver mitochondrial systems proceeds mainly through transamination with oxaloacetate followed by oxidation of the α-oxoglutarate formed. Both in the presence and absence of dinitrophenol in liver mitochondria this pathway accounted for almost 80% of the uptake of glutamate. In brain preparations the transamination pathway accounted for about 90% of the glutamate uptake. 2. The oxidation of [1-¹⁴C]- and [5-¹⁴C]-glutamate in brain preparations is compatible with utilization through the tricarboxylic acid cycle, either after the formation of α-oxoglutarate or after decarboxylation to form γ-aminobutyrate. There is no indication of γ-decarboxylation of glutamate. 3. The high respiratory control ratio obtained with glutamate as substrate in brain mitochondrial preparations is due to the low respiration rate in the absence of ADP: this results from the low rate of formation of oxaloacetate under these conditions. When oxaloacetate is made available by the addition of malate or of NAD⁺, the respiration rate is increased to the level obtained with other substrates. 4. When the transamination pathway of glutamate oxidation was blocked with malonate, the uptake of glutamate was inhibited in the presence of ADP or ADP plus dinitrophenol by about 70 and 80% respectively in brain mitochondrial systems, whereas the inhibition was only about 50% in dinitrophenol-stimulated liver preparations. In unstimulated liver mitochondria in the presence of malonate there was a sixfold increase in the oxidation of glutamate by the glutamate-dehydrogenase pathway. Thus the operating activity of glutamate dehydrogenase is much less than the `free' (non-latent) activity. 5. The following explanation is put forward for the control of glutamate metabolism in liver and brain mitochondrial preparations. The oxidation of glutamate by either pathway yields α-oxoglutarate, which is further metabolized. Since aspartate aminotransferase is present in great excess compared with the respiration rate, the oxaloacetate formed is continuously removed by the transamination reaction. Thus α-oxoglutarate is formed independently of glutamate dehydrogenation, and the question is how the dehydrogenation of glutamate is influenced by the continuous formation of α-oxoglutarate. The results indicate that a competition takes place between the α-oxoglutarate-dehydrogenase complex and glutamate dehydrogenase, probably for NAD⁺, resulting in preferential oxidation of α-oxoglutarate.     

Opiate target site N-demethylase enzymes: differences from the liver N-demethylase.

Mixtures of morphine-6³H and morphine-N-¹⁴CH₃ were incubated with rat brain subcellular fractions. Isotope ratio measurements served as the marker for identification, purification and quantitation of N-nor products which were shown to consist almost solely of N-normorphine. The microsomal, synaptosomal and mitochondrial, but not the supernatant brain preparations yielded N-normorphine. The microsomal incubations were then repeated in the presence of cytochrome P-450 inhibitors which suppressed the liver reaction but did not affect the brain biotransformation. The brain N-dealkylase is therefore different from the one in the liver and is not a cytochrome P-450 linked enzyme.     

THE DISTRIBUTION OF GLUTAMATE DECARBOXYLASE IN RAT TISSUES; ISOTOPIC VS FLUORIMETRIC ASSAYS

—The activity of glutamate decarboxylase (GAD, EC 4.1.1.15) in normal and neoplastic rat tissues was determined by two assay methods, one based on the production of ¹⁴CO₂ from [¹⁴C]glutamic acid and the other on the fluorimetric measurement of γ-aminobutyric acid (GABA) formation. Activities obtained with the isotopic assay were high in every tissue (ranging from over 800 in liver and brain to 107nmol CO₂/min/g in lung). They were drastically diminished by Triton X-100, by an oxygen-free atmosphere or by the mitochondrial electron transport inhibitors, rotenone and antimycin A. Activities measured fluorimetrically were significant in only a few tissues and were stimulated by Triton (e.g. from 299 to 569 nmol GABA/min/g brain) but were unaffected by rotenone. For several tissues after Triton treatment the fluorimetric and isotopic assays (in air) gave the same results (i.e. the two end products, CO₂ and GABA were in stoichiometric agreement); however, the fluorimetric assay remains the more reliable measure of GAD activity since Triton may not inhibit completely the non-GAD dependent decarboxylation of glutamate in all types of tissue preparations. The hepatic, renal and mammary tumours tested were devoid of GAD; among non-neural normal tissues, kidney, liver and, possibly, adrenal gland contained significant GAD activity. In kidney and liver the activity was 15 and 10 per cent of that in brain.     

Aldrin epoxidation by plant root extracts

Aldrin epoxidase activity of the cell-free pea and bean root preparations was located in the particulate fraction. High speed centrifugation at 250 000 g for 2 hr resulted in a pellet with almost all the activity of the crude cell-free preparations. While the epoxidase was stimulated by NADPH generating system in these cell-free root preparations, that in high speed centrifugation pellets was not. Aldrin epoxidase activity of the dwarf bean root homogenates was increased by addition of p-aminobenzoic acid (10^?4 M). This increase in activity is above that already manifested by Polyclar AT. No activity was detected in the dormant or germinating Alaska peas or dwarf beans until the 5th and 6th day, respectively.     

Biosynthesis of naphthoqinones and anthraquinones in streptocarpus dunnii cell cultures

Administration of ^p13C- and ^p2H-labelled precursors to Streptocarpus dunnii cell cultures demonstrated that the naphthoquinones formed through aunique prenylation mode are biosynthesized via 4-(2'-carboxyphenyl)-4-oxobutanoic acid, 1,4-dihydroxy-2-naphthoic acid, lawsone and lawsone 2-prenyl ether, and that the anthraquinones are biosynthesized through prenylation of 2-carboxy-4-oxo-1-tetralone at the carboxy-bearing carbon atom to form 2-carboxy-2-prenyl-4-oxo-1-tetralone,or through ipso attack of the prenyl group on the corresponding carbon atom of 1,4-dihydroxy-2-naphthoic acid.     

Calcium-activated proteolytic activity in rat liver mitochondria

Soluble extracts from sonicated rat liver mitochondria and rat liver cytosol were each chromatographed on DEAE-cellulose columns, and the fractions assayed for Ca²⁺-activated proteolytic activity using ¹⁴C-casein as a substrate. The mitochondrial preparations were shown to be free of cytosolic and microsomal contamination by the lack of alcohol dehydrogenase activity, a cytosolic marker enzyme, and by a lack of cytochrome P-450 activity, a microsomal marker enzyme. Two peaks of Ca²⁺-activated neutral endoprotease activity were resolved from the mitochondrial fractions. One protease was half-maximally activated with 25 μM Ca²⁺, and the other by 750 μM Ca²⁺. Rat liver cytosol contained only a high Ca²⁺-requiring protease peak. This is the first demonstration of Ca²⁺-activated proteases in mitochondria.     

An investigation into the role of malonyl-coenzyme A in isoprenoid biosynthesis

1. [¹⁴C]Malonyl-CoA was incorporated into isoprenoids by cell-free yeast preparations, by preparations from pigeon and rat liver, and by Hevea brasiliensis latex. 2. In agreement with previous reports the incorporation of acetyl-CoA into isoprenoids was not inhibited by avidin and was not stimulated by HCO₃⁻. In a cell-free yeast preparation addition of HCO₃⁻ stimulated the formation of fatty acids from acetyl-CoA and decreased the incorporation into unsaponifiable lipids. 3. The labelling patterns of β-hydroxy-β-methylglutaryl-CoA formed from [2-¹⁴C]- and [1,3-¹⁴C]-malonyl-CoA in rat and pigeon liver preparations were those that would be expected if malonyl-CoA underwent decarboxylation to acetyl-CoA before incorporation. 4. The labelling pattern of ergosterol formed by cell-free yeast preparations from [2-¹⁴C]malonyl-CoA was also consistent with decarboxylation of malonyl-CoA before incorporation. 5. The incorporation of [2-¹⁴C]malonyl-CoA into mevalonate by rat liver preparations was related to the malonyl-CoA decarboxylase activity present in the preparation.     

DEVELOPMENT OF MITOCHONDRIAL PYRUVATE METABOLISM IN RAT BRAIN

The activities of a number of mitochondrial enzymes involved in the metabolism of pyruvate during development of the rat brain were investigated. The rates of decarboxylation of [1-¹⁴C]pyruvate to ¹⁴CO₂ via pyruvate dehydrogenase and the fixation of H¹⁴CO₃^? in the presence of pyruvate via pyruvate carboxylase by brain homogenates were very low in newborn rats. These rates increased markedly by about four-fold and 15-fold respectively during 10–35 postnatal days. The rates of the fixation of H¹⁴CO₃^? by cerebral homogenates were supported by the development of the activity of pyruvate carboxylase in rat brain. The activities of citrate synthase, aconitase, NAD-malate dehydrogenase, aspartate aminotransferase, alanine aminotransferase and phosphoenol-pyruvate carboxykinase were very low in the particulate fraction of the newborn rat brain. The activities of all these enzymes increased makedly by about three- to 10-fold during 10–35 days after birth. The activity of mitochondrial phosphoenolpyruvate carboxykinase from rat brain was not precipitated by an antibody prepared against rat liver cytosolic phosphoenolpyruvate carboxykinase suggesting that cerebral mitochondrial enzyme is immunologically different from that of the cytosolic form in hepatocytes. The significance of the development of the cerebral mitochondrial metabolism is discussed in relation to biochemical maturation of the brain.     

In vitro transformation of dehydroepiandrosterone or its sulphate into androstenediol or its sulphate by rat brain and blood preparations

Abstract— –Enzymic transformation of [4-¹⁴C]dehydroepiandrosterone or [4-¹⁴C]dehydro-epiandrosterone sulphate to androstenediol or its sulphate occurred when incubated with a microsomal preparation of rat brain or a whole rat blood homogenate. The brain enzyme which appeared to cause this transformation had a pH optimum at 60, was NADPH₂-dependent, and had an apparent K_m of 4·6 × 10^?6m . When the subcellular fractions of rat brain were compared for transformation, microsomes had the highest specific activity, followed by the cytosol. The crude nuclear and mitochondrial fractions had no significant activity. The level of enzymic activity in the brain microsomes increased from that for rats sacrificed at 7 days of postnatal age to a maximum for rats sacrificed at 1 month of age; then the activity appeared to level off in rats older than 1 month. Microsomes obtained from the cerebellum had the highest specific activity in comparison to that obtained from the cerebral cortex, the diencephalon, and the brain stem. The incubated preparations of rat brain also converted dehydroepiandrosterone sulphate to androstenediol sulphate without hydrolysis. The enzyme in rat blood which was similar to that in the brain was also partially characterized. The blood enzyme had a pH optimum at 6–5, was nearly exclusively present in erythrocytes, was also NADPH₂-dependent, and had an apparent K_m of 2·7 × 10^?4m . The developmental pattern of the blood enzyme specific activity was similar to that of the rat brain enzyme. Upon haemolysis, most activity was recovered in the haemolysate.