Effect of Fenbufen on the Entry of New Quinolones, Norfloxacin and Ofloxacin, into the Central Nervous System in Rats

The entry of two new quinolone antibacterial agents, norfloxacin and ofloxacin, into the central nervous system (CNS) of rats, and the effect of fenbufen on this was investigated. At various times after the administration of a bolus intravenous dose of norfloxacin or ofloxacin (10 mg kg−1) with or without fenbufen (20 mg kg−1), serum and cerebrospinal fluid (CSF) samples and whole brain were collected from the rats and the concentration of norfloxacin or ofloxacin in each sample was determined. Serum concentrations of both quinolones declined biexponentially with time and were significantly elevated by coadministration with fenbufen at the terminal phase. The fractions of these quinolones bound to serum protein were not altered by coadministration with fenbufen. Coadministered fenbufen raised the brain concentrations of both quinolones but did not affect their brain to serum unbound concentration ratios. In contrast, CSF to serum unbound concentration ratios as well as CSF concentrations of norfloxacin and ofloxacin were elevated by coadministration with fenbufen. Apparent diffusional clearances between blood and CSF of norfloxacin and ofloxacin estimated by the physiological model analysis increased by 1·9 and 2·6 times, respectively, after coadministration with fenbufen. These findings suggest that coadministered fenbufen may facilitate the entry of norfloxacin and ofloxacin into the CNS.

Salivary excretion of mexiletine after bolus intravenous administration in rats

Salivary excretion of mexiletine was investigated following bolus intravenous administration (10 mg kg−1) in rats. Parotid and mandibular saliva was collected separately by stimulating salivation with constant rate infusion of pilocarpine (3 mg kg−1 h−1). The mexiletine levels in blood plasma and parotid and mandibular saliva declined biexponentially with time in almost parallel fashion. Although the mexiletine levels in both types of saliva were lower than that in plasma, the drug level in parotid saliva was always higher than that in mandibular saliva. Significant correlations were observed when all data relating mexiletine concentration in plasma and saliva were included (P < 0·001). The saliva/plasma drug concentration ratios (S/P ratios) did not vary to a large extent (0·56 ± 0·10 for parotid saliva, 0·21 ± 0·06 for mandibular saliva), but there was a consistent tendency for the higher plasma drug levels in the distribution phase to produce relatively high S/P ratios for both parotid and mandibular saliva. Moreover, the plasma mexiletine levels calculated by the equation of Matin et al (1974) employing the observed values for the saliva drug level, saliva pH and free fraction of mexiletine in plasma were significantly higher than the observed drug levels. Therefore, it is suggested that the salivary excretion of mexiletine could not be explained quantitatively by simple, passive secretion based on pH-partition theory.

Incidence and risk factors of falling in ambulatory patients with rheumatoid arthritis: a prospective 1-year study

A prospective 1-year study showed that fall incidence was 50% in women with rheumatoid arthritis. Multivariate analysis identified swollen joint count, use of antihypertensives or diuretics, one-leg standing time, and sway area measured by stabilometer as significant parameters associated with falls.

INTRODUCTION

Patients with rheumatoid arthritis (RA) may be at increased risk of falling because they frequently experience muscle weakness and stiff or painful joints. The aim of this study was to use a prospective design to determine the incidence of falls and their risk factors in women with RA.

METHODS

Eighty-four women aged 50 and over who had RA were enrolled. The mean age was 64.1 years. We evaluated postural stability, physical performance related to falls, disease activity, muscle volume, and bone density. The occurrence of falls was assessed every month for 1 year. Among 84 patients, 80 completed a 1-year observation.

RESULTS

Forty patients (50.0%) reported one or more falls, and two of them (5.0%) had fractures during the follow-up period. The fall group had more swollen joints and took more antihypertensives and/or diuretics. The fall group also had lower postural stability and tended to have reduced physical performance. The one-leg standing time was shorter, and the step-up-and-down test score was lower in the fall group. The sway area was larger in the fall group.

DISCUSSION

Multiple logistic regression analysis identified that number of swollen joints, use of antihypertensives or diuretics, shorter time standing on one foot, and the sway area were the most significant parameters associated with falls.

CONCLUSION

We concluded that fall rates in RA patients were higher than in the general population and that balance impairment or side effects of drugs may play a role in increasing the risk of falls.

Catalytic activity of three variants (Ile, Leu, and Thr) at amino acid residue 359 in human CYP2C9 gene and simultaneous detection using single-strand conformation polymorphism analysis

This study evaluated the catalytic activity of three variants (Ile, Leu, and Thr) at codon 359 of CYP2C9 enzymes expressed in a yeast cDNA expression system, and then established single-strand conformation polymorphism (PCR-SSCP) analysis for simultaneous detection as a screening method. Diclofenac was used for the in vitro experiment, and its hydroxy metabolite (4'-hydroxydiclofenac) was measured by HPLC. To discuss the in vivo effect of the Thr359 variant on the pharmacokinetics of phenytoin, a case report is presented. The efficiency of the SSCP method was evaluated by analyzing DNA samples from a homozygote for Ile359 and a heterozygote for Leu359 or Thr359. To evaluate the interaction between the P450 level and reductase activity, two batches of the Thr359 variant with a different P450:reductase activity ratio (1:4.0 and 1:1.4) were used. The in vitro study revealed that recombinant Ile359, Leu359, and Thr359 (2 batches) possessed a mean Km of 2.0, 16.5 and (3.8 and 2.9) micromol and Vmax of 12.4, 17.9 and (4.4 and 5.1) nmol/min/nmol P450, respectively. Although the magnitude of the change in catalytic efficiency for the Thr359 variant was close to that of the Leu359 variant, the effect of the two variants on diclofenac 4'-hydroxylation appears to be different because Leu359 variant was associated with a high Km, and Thr359 with a low Vmax. No significant differences in the kinetic data were observed between the two Thr359 enzymes, suggesting that low reductase activity in the Thr359 enzyme was not a major determinant in the present in vitro experiment. Estimated pharmacokinetic parameters of phenytoin obtained by the Bayesian method in an epileptic patient who was a heterozygote carrier for Thr359 variant were: Km = 6.45 microg/mL, Vmax = 5.77 mg/kg/d, and Vmax/Km = 0.89 L/kg/day. The Vmax/Km value in this patient was similar to the population mean value (0.90 L/kg/day) in Japanese heterozygotes for the Leu359 variant. Results for PCR-SSCP were in complete agreement with those obtained using established methods. Thus, the PCR-SSCP approach is useful for identifying these three variants of the CYP2C9 gene.

High-performance liquid chromatographic determination of nalidixic acid in rat serum, brain and cerebrospinal fluid

A high-performance liquid chromatographic method for the determination of nalidixic acid (NA) in rat serum, brain and cerebrospinal fluid (CSF) was developed. NA in rat serum and brain homogenate was extracted and injected onto a reversed-phase column. CSF was directly analysed without extraction procedure. The limits of detection were 0.05 microgram ml-1 for serum, 0.07 microgram g-1 for brain and 0.02 microgram ml-1 for CSF, respectively. Calibration curves were linear over the concentration ranges 0.1-50 micrograms ml-1 for serum, 0.12-9 micrograms g-1 for brain and 0.05-10 micrograms ml-1 for CSF, respectively. The reproducibility of NA assay in rat biological media ranged from 1 to 4% relative standard deviations (RSD). The recoveries of NA added to serum and brain were higher than 96% with an RSD of less than 4%. The present method was found to be applicable to pharmacokinetic study of NA in rat serum, brain and CSF.

Dependency of salivary excretion of mexiletine on the plasma concentration in rats

The effect of steady-state plasma concentrations on the salivary excretion of mexiletine was investigated following simultaneous bolus intravenous injection of the loading dose (2.7 or 16.1 mg kg-1) and constant-rate intravenous infusion of the maintenance dose (15 or 102 micrograms min-1 kg-1) in male Wistar rats. Parotid and mandibular saliva was collected separately by stimulating salivation with a constant-rate infusion of pilocarpine (50 micrograms kg-1 min-1) in each rat. The low and high steady-state levels of mexiletine in blood plasma were attained at 0.259 +/- 0.123 and 1.616 +/- 0.475 micrograms mL-1, respectively, within the first 1-2 h after drug administration. Similarly, the two different steady-states in both parotid and mandibular saliva were attained. Although the mexiletine levels in both types of saliva were lower than that in plasma, the drug level in parotid saliva was always higher than that in mandibular saliva at any steady-state (P < 0.001 or 0.01). In parotid saliva, the high steady-state produced greater saliva to plasma drug concentration ratios (S/P ratio, 0.475 +/- 0.160) than that (0.386 +/- 0.131) at the low steady-state (P < 0.05). The S/P ratio for mandibular saliva at the high (0.204 +/- 0.060) steady-state was also greater than that at the low (0.158 +/- 0.050) steady-state (P < 0.01). These changes in the S/P ratio could not be explained by the pH for either parotid or mandibular saliva, but partially by the change in the unbound fraction of the drug which tended to be consistent with that in the ratio for both salivary glands. These findings suggest that the salivary excretion of mexiletine may be dependent on the plasma unbound concentration in rats.

Pharmacokinetics of [6]-gingerol after intravenous administration in rats with acute renal or hepatic failure

The pharmacokinetics of [6]-gingerol were investigated in rats with acute renal failure induced by bilateral nephrectomy, or those with acute hepatic failure induced by a single oral administration of carbon tetrachloride (CCl4), to clarify the contribution of the kidney and liver to the elimination process of [6]-gingerol. After bolus intravenous administration, a plasma concentration-time curve of [6]-gingerol was illustrated by a two-compartment open model. There was no significant difference in either the plasma concentration-time curve or any pharmacokinetic parameters between the control and nephrectomized rats. It is suggested, therefore, that renal excretion does not contribute at all to the disappearance of [6]-gingerol from plasma in rats. In contrast, hepatic intoxication with CCl4 elevated the plasma concentration of [6]-gingerol at the terminal phase. Its elimination half-life increased significantly, from 8.5 to 11.0 min, in CCl4-intoxicated rats. The extent of [6]-gingerol bound to serum protein was more than 90% and was affected very slightly by the CCl4-intoxication. These aspects indicate that [6]-gingerol is eliminated partly by the liver.

Enhanced entry of ciprofloxacin into the rat central nervous system induced by fenbufen

The effects of fenbufen on the transport of ciprofloxacin (CPFX) into the brain and cerebrospinal fluid (CSF) were investigated in rats. Periodically after a bolus i.v. dose of CPFX (10 mg/kg), alone or with fenbufen (20 mg/kg), to rats, aliquots of CSF and blood were collected and then the whole brain was readily excised from the animal after sacrifice by microwave irradiation. Serum levels of CPFX in the terminal phase were significantly elevated by the coadministration with fenbufen. However, the extent of CPFX binding to serum protein was not affected by fenbufen. Immediately after the coadministration with fenbufen, brain and CSF levels of CPFX were raised by about 15 to 70% and 70 to 100%, respectively. Both brain/serum unbound and CSF/serum unbound level ratios were increased by fenbufen at relatively early periods after drug injection. Analysis based on physiological models indicated that fenbufen significantly increased the apparent diffusion clearances of CPFX across blood-brain and blood-CSF barriers. These findings suggest that coadministered fenbufen may facilitate the entry of CPFX into the central nervous system not only by elevation of serum level but also by enhancement of permeability across the blood-brain or blood-CSF barrier.

Salivary excretion of mexiletine in normal healthy volunteers

To investigate the kinetics and correlation between serum and saliva levels of mexiletine, serum (total and unbound) and saliva drug concentration-time courses have been analysed in five normal healthy volunteers after administration of a single oral dose (200 mg) of the drug. Mexiletine levels in saliva were always higher than those in serum. The drug concentration-time curve in each sample was analysed according to the non-linear least squares regression program MULTI, for a two-compartment model with first-order absorption. The saliva drug concentration in the post-absorption phase was found to be well correlated with either corresponding serum total or serum unbound drug level in four of the subjects. Although there was a large inter-individual variation in the ratio of saliva to serum drug concentrations as well as in the pharmacokinetic parameters, an almost consistent ratio was obtained in each individual.

Pharmacokinetics of [6]-gingerol after intravenous administration in rats

A high-performance liquid chromatographic method to determine [6]-gingerol, a pungent constituent of ginger, in rat plasma was developed and a pharmacokinetic study was performed in rats. Quantitative analysis with high reproducibility was achieved for [6]-gingerol over the concentration range of 0.2-40 micrograms/ml. After bolus intravenous administration at a dose of 3 mg/kg, the plasma concentration-time curve was described by a two-compartment open model. [6]-Gingerol was rapidly cleared from plasma with a terminal half-life of 7.23 min and a total body clearance of 16.8 ml/min/kg. Serum protein binding of [6]-gingerol was 92.4%.

High-performance liquid chromatographic determination of ciprofloxacin in rat brain and cerebrospinal fluid

A novel high-performance liquid chromatographic method for the fluorometric determination of a newer quinolone, ciprofloxacin (CPFX), in rat brain and cerebrospinal fluid (CSF) was developed. CPFX in brain homogenate was extracted and injected onto a reversed-phase column without fluorescence derivatization. CSF was directly analyzed without the extraction procedure. Calibration curves were linear over the concentration ranges of 10 to 500 ng/g for brain and 5 to 500 ng/ml for CSF. The recoveries of CPFX added to brain were more than 97% with a coefficient of variation of less than 4%. The present method was sensitive and reliable enough to be utilized for detailed pharmacokinetic studies of CPFX in rat brain and CSF.

A possible reduction in the renal clearance of ciprofloxacin by fenbufen in rats

The change in plasma concentration-time profile, serum protein binding and renal and biliary clearances of ciprofloxacin caused by coadministration of fenbufen has been studied in rats administered an intravenous dose of ciprofloxacin (5 mg kg-1) alone or with fenbufen (10 mg kg-1). Coadministered fenbufen significantly prolonged the plasma elimination half-life of ciprofloxacin from 40.5 to 57.6 min and tended to reduce the total body clearance of this quinolone by about 20%. The extent of ciprofloxacin binding to rat serum protein was not affected by fenbufen, nor did it affect the biliary clearance of the quinolone. However, fenbufen tended to reduce renal clearance and significantly decreased the cumulative renal excretion of the quinolone during at least the first 3 h after drug administration. These results suggest a possible reduction of ciprofloxacin clearance owing to inhibition of renal excretion by fenbufen.

A minor possibility of pharmacokinetic interaction between enoxacin and fenbufen in rats

In order to clarify the possibility of pharmacokinetic interaction between quinolone and fenbufen, the plasma concentration-time profiles and serum protein binding of enoxacin, fenbufen and its active metabolite, felbinac, were investigated in rats. The rats were administered an intravenous dose of enoxacin (5 mg/kg) and fenbufen (10 mg/kg) alone or concomitantly. Coadministration with fenbufen tended to prolong the plasma elimination half-life of enoxacin by about 20%, whereas it showed no effect on the area under plasma concentration-time curve, total body clearance or distribution volume of enoxacin. The extent of enoxacin binding to rat serum tended to be slightly reduced by fenbufen in vivo and in vitro. Plasma concentration-time curves, pharmacokinetic parameters and serum protein binding of fenbufen and felbinac were not affected at all by the coadministration with enoxacin. These aspects suggest that there may be only a minor possibility of the pharmacokinetic interaction between enoxacin and fenbufen.

High-performance liquid chromatographic procedure for the simultaneous determination of norfloxacin, fenbufen and felbinac in rat plasma

A high-performance liquid chromatographic method for the simultaneous determination of norfloxacin, fenbufen and felbinac extracted from 50 microliters of rat plasma is described. Chromatography was performed on a reversed-phase column with ultraviolet detection. By the present method, quantitative and reproducible determinations were possible for norfloxacin, fenbufen and felbinac over the concentration ranges of 0.2-20, 0.2-120 and 0.4-40 micrograms/ml, respectively. The recoveries of norfloxacin, fenbufen and felbinac added to plasma were nearly 100% with a coefficient of variation of less than 8.0%. This method was found to be applicable to pharmacokinetic studies of each drug after the concomitant administration of norfloxacin and fenbufen.

Absence of pharmacokinetic interaction between ofloxacin and fenbufen in rats

The possible pharmacokinetic interaction between a new quinolone and fenbufen was investigated by comparing the plasma concentration-time profiles and serum protein binding of ofloxacin, fenbufen and its active metabolite, felbinac, in rats. The rats were administered intravenous doses of ofloxacin (5 mg kg-1), fenbufen (10 mg kg-1) alone or concomitantly. The plasma elimination half-lives were about 55 min in both groups. A slight elevation of plasma concentration of ofloxacin and a small decrease of its total body clearance were observed after its coadministration with fenbufen. The extent of ofloxacin binding to rat serum tended to be slightly reduced by fenbufen which coexisted at relatively high concentrations. Plasma concentration-time curves, pharmacokinetic parameters and serum protein binding of fenbufen and felbinac were not affected by the coadministration with ofloxacin. These results suggest that any substantive pharmacokinetic interaction may be unlikely after the concomitant administration of ofloxacin and fenbufen.

Simultaneous determination of ofloxacin, fenbufen and felbinac in rat plasma by high-performance liquid chromatography

Ofloxacin, fenbufen and its active metabolite, felbinac, were simultaneously extracted from 50 microliters of rat plasma and analysed by high-performance liquid chromatography on a reversed-phase column. Quantitative and reproducible determinations were possible for ofloxacin, fenbufen and felbinac over the concentration ranges 0.15-40, 0.3-80 and 0.45-45 micrograms/ml, respectively. The detection limits for all the drugs were lower than those reported previously. The recovery of ofloxacin, fenbufen and felbinac added to plasma was nearly 100% with a coefficient of variation of less than 3.0%. This method was found to be applicable to pharmacokinetic studies of each drug after the concomitant administration of ofloxacin and fenbufen.

Inhibition of hepatic 17 beta- and 3 alpha-hydroxysteroid dehydrogenases by antiinflammatory drugs and nonsteroidal estrogens

Antiinflammatory agents and estrogens have been tested as inhibitors of two isozymes of guinea pig liver testosterone 17 beta-dehydrogenase (NADP) 1.1.1.64) and rat liver 3 alpha-hydroxysteroid dehydrogenase (EC 1.1.1.50). Antiinflammatory steroids and estradiols were highly inhibitory to 3 alpha-hydroxysteroid dehydrogenase and one isozyme of testosterone 17 beta-dehydrogenase, respectively, but nonsteroidal antiinflammatory agents and nonsteroidal estrogens such as hexestrol, dienstrol, diethylstilbestrol and zearalenone showed potent inhibitions on all the enzymes. Although the inhibitory potency of indomethacin for one isozymes of testosterone 17 beta-dehydrogenase and 3 alpha-hydroxysteroid dehydrogenase decreased with changing pH from 9.7 to 7.0, that of the nonsteroidal estrogens for all the enzymes was little affected by pH. No additive effect in double inhibitor experiments with indomethacin and the nonsteroidal estrogens was observed, and the compounds were all competitive inhibitors with respect to steroidal substrate. The results suggest that there is a very similar region in substrate binding sites of the enzymes.

Dihydrodiol dehydrogenases in guinea pig liver

Four major and four minor dihydrodiol dehydrogenases, with similar apparent molecular weights of 28,000 to 34,000 but with different charges, were purified from male guinea pig liver cytosol. One of the minor enzymes catalyzed only the oxidation of benzene dihydrodiol with a high Km value of 5.0 mM and was identified immunologically with aldehyde reductase. The other enzymes oxidized xenobiotic alicyclic alcohols and 17 beta-hydroxysteroids as well as benzene dihydrodiol. These enzymes exhibited higher affinity for 17 beta-hydroxysteroids than for alicyclic alcohols and benzene dihydrodiol, and immunologically cross-reacted with testosterone 17 beta-dehydrogenase purified from the same source. Four major enzymes and one minor with Km values for benzene dihydrodiol of about 0.2 mM, possessed specificity for 5 beta-androstane--17 beta-hydroxysteroids and dual cofactor requirement, whereas the other two minor enzymes with high Km values of over 5 mM showed apparent NADP and 5 alpha-androstane specificity. The dihydrodiol dehydrogenase activity was localized in the cytosol of liver. The results indicate that the hepatic oxidation of dihydrodiols in the guinea pig is mediated by cytosolic testosterone 17 beta-dehydrogenase isozymes and aldehyde reductase. Testosterone 17 beta-dehydrogenase immunologically identical to the liver enzymes was detected only in kidney, whereas aldehyde reductase was detected in all tissues of the guinea pig.

Immunological identification of soluble carbonyl reductase with testosterone 17 beta-dehydrogenase (NADP+) in guinea pig liver and kidney

Antiserum was produced against one of two carbonyl reductases purified from guinea pig liver cytosol to identify the enzymes as testosterone 17 beta-dehydrogenase isozymes. Immunoelectrophoresis and immunoprecipitation with the antiserum indicated that the two reductases had common antigenic sites. The antiserum inhibited most of both carbonyl reductase and testosterone 17 beta-dehydrogenase activities in the purified reductases and in cytosols of liver and kidney.

Kinetic mechanisms in the reduction of aldehydes and ketones catalyzed by rabbit liver aldehyde reductases and hydroxysteroid dehydrogenases

The kinetic properties of the NADPH-dependent reduction of aromatic aldehydes and ketones catalyzed by low- and high-molecular-weight aldehyde reductases [alcohol : NADP oxidoreductase, EC 1.1.1.2] and 3 alpha- and 3 beta-hydroxysteroid dehydrogenases [EC 1.1.1.50 and 1.1.1.51] of rabbit liver were compared. Initial velocity measurements with pyridine-4-aldehyde, 4-benzoylpyridine and androstadione as substrates and inhibition studies with their products indicated that all the enzymes followed an ordered Bi Bi reaction mechanism with coenzyme binding first and leaving last. However, phenylpyruvic acid inhibited 3 alpha-hydroxysteroid dehydrogenase and low-molecular-weight aldehyde reductase noncompetitively with respect to either NADPH or substrate, whereas it inhibited 3 beta-hydroxysteroid dehydrogenase and high-molecular-weight aldehyde reductase uncompetitively. Cibacron blue F3GA dye was a dead-end inhibitor of the enzymes, being competitive with respect to NADPH and noncompetitive with respect to the other substrate, but the Ki value of 3 alpha-hydroxysteroid dehydrogenase for this dye was much higher than those of the other enzymes.

Microsomal reductase for aromatic aldehydes and ketones in guinea pig liver. Purification, characterization, and functional relationship to hexose-6-phosphate dehydrogenase

An NADPH-specific aromatic aldehyde-ketone reductase located in guinea pig liver microsomes can be effectively solubilized with nonionic detergents, but not with bile salts and hydrolytic enzymes. Destruction of microsomal membranes by nonionic detergents or acetone treatment leads to significant activation of the reductase, indicating that the enzyme is partly latent in intact microsomes. After solubilization with Triton X-100, the reductase has been highly purified. The purified enzyme catalyzes the NADPH-linked reduction of xenobiotic aromatic aldehydes and ketones as well as 3-ketosteroids, notably 5 alpha- and 5 beta-dihydrotestosterones. The reductase activities for xenobiotic carbonyl compounds and for 3-ketosteroids are each inhibited by addition of the other type of substrate and show the same pH optimum, cofactor requirement, and heat stability, indicating the same enzyme is responsible for the reduction of the two types of substrates. Hexose-6-phosphate dehydrogenase, purified from guinea pig liver microsomes, acts as a more effective NADPH generator for the reductase than yeast and guinea pig liver cytosolic glucose-6-phosphate dehydrogenase. Evidence has been obtained that hexose-6-phosphate dehydrogenase undergoes a functional interaction with the reductase, facilitating the provision of NADPH to the reductase activity both in the reconstituted system and in microsomes.

A possible functional relationship between microsomal aromatic aldehyde-ketone reductase and hexose-6-phosphate dehydrogenase

Aromatic ketone reductase activity of microsomes showed a unique cofactor requirement: Addition of NADP and glucose-6-phosphate was as effective as that of an artificial NADPH generating system, whereas NADPH alone served as a cofactor less efficiently. Microsomal aromatic ketone reductase, purified partially from guinea pig liver microsomes after solubilization with Triton X-100, reduced 5 beta-dihydrotestosterone, aromatic aldehydes, and ketones with NADPH as a cofactor. However, addition of hexose-6-phosphate dehydrogenase, purified from the same source, as an NADPH generator produced about 2 times higher activity than that of yeast glucose-6-phosphate dehydrogenase or NADPH alone.

Guinea pig liver aromatic aldehyde-ketone reductases identical with 17 beta-hydroxysteroid dehydrogenase isozymes

Two NADPH-dependent aromatic aldehyde-ketone reductases purified from guinea pig liver catalyzed oxidoreduction of 17 beta-hydroxysteroids and 17-ketosteroids. One enzyme efficiently oxidized 5 beta-androstanes and reduced 17-ketosteroids of A/B cis configuration, whereas the other enzyme efficiently oxidized 5 alpha-androstanes and equally reduced both 5 alpha-and 5 beta-androstanes of 17-ketosteroids. However, aromatic aldehydes and ketones, and 3-ketosteroids were irreversibly reduced by the two enzymes. The two enzymes utilized NADP+ or NADPH as cofactor, but little activity with NAD+ or NADH was found. Phosphate ions enhanced the NAD+-dependent dehydrogenase activity and NADH-dependent reductase activity of the two enzymes, whereas the activities with NADP+ and NADPH were not affected. The ratios of the two activities of ketone reduction and 17 beta-hydroxysteroid oxidation of the two enzymes were almost constant during the purification steps after the two enzymes had been separated by DEAE-cellulose chromatography. By kinetic studies and electrophoresis and isoelectric focusing experiments it was confirmed that both of the two enzymes were responsile for the reduction aldehydes, ketones, and ketosteroids and for the oxidation of 17 beta-hydroxysteroids. These results indicate that 17 beta-hydroxysteroid dehydrogenases may play important roles in the metabolism of exogeneous aldehydes and ketones as well as steroids.