Epidemiological study of urinary 6beta-hydroxycortisol to cortisol ratios and breast cancer risk.

The ratio of urinary 6beta-hydroxycortisol:cortisol is a measure of the activity of cytochrome p450 3A4 (CYP3A4). CYP3A4 catalyzes the formation of the genotoxic estrogen, 16alpha-hydroxyestrone. It is also involved in the activation of many other mammary carcinogens, such as the polycyclic aromatic hydrocarbons and heterocyclic amines. We evaluated the association between urinary cortisol ratios and breast cancer risk in a subgroup of women who participated in a population-based case-control study in Shanghai. Overnight urine samples from 246 case-control pairs were assayed for 6beta-hydroxycortisol (6beta-OHC) to cortisol. The urine samples from all of the breast cancer patients were collected before any chemotherapy or radiotherapy. In-person interviews were conducted to obtain comprehensive information on dietary habits, reproductive history, and other lifestyle factors. The median levels of 6beta-OHC:cortisol ratios were 2.61 in cases and 2.16 in controls, a 20.8% difference (P < 0.001). The case-control difference was larger in women over 45 years of age (31.3% difference; P < 0.001) than younger women (6.0%; P = 0.45). After adjusting for confounding variables, the risks of breast cancer were increased from 1.0 (reference) to 1.6 [95% confidence interval (CI), 0.9-3.1], 2.2 (95% CI, 1.1-4.2), and 3.7 (95% CI, 1.9-7.4; P for trend, <0.001) with increasing levels of 6beta-OHC:cortisol ratios. The positive association was more pronounced among older women (>45 years) than among younger women (< or = 45 years). The adjusted odds ratios associated with the highest cortisol ratio were 6.0 (95%CI, 2.2-16.1) among older women and 2.2 (95%CI, 0.8-6.1) among younger women. The association of the 6beta-OHC:cortisol ratio was stronger among older women who had a high body mass index, late age at menopause, and early age at menarche (factors related to high endogenous estrogen exposure) than those who did not have these factors. These findings are consistent with the role of CYP3A4 in estrogen and carcinogen metabolism and suggest that high CYP3A4 activity may be a risk factor for breast cancer risk.

Physiologic concentrations of homocysteine inhibit the human plasma GSH peroxidase that reduces organic hydroperoxides.

The plasma reduced glutathione (GSH) selenoperoxidase is a highly conserved enzyme. Furthermore, a small clinical study reported that patients with severe atherosclerosis had low peroxidase activities. Together these observations suggest that the peroxidase is important in preventing atherosclerosis. Yet others have reported that when the assay was run in Tris buffer, it was inactive with the concentrations of GSH found in the plasma. Second, it is known that hyperhomocysteinemia increases the rate of atherogenesis. Because there is some homology between homocysteine and the cysteine in GSH, the question is whether the hyperhomocysteinemia effect may be due to inhibition of the peroxidase. We purified the peroxidase from human plasma and determined its activity by a coupled spectrophotometric assay and a substrate disappearance chemiluminescence assay. When the peroxidase activity was determined in phosphate-buffered saline solution (PBS), there was significant activity with the reported plasma GSH concentrations (5 to 20 micromol/L). The peroxidase was exclusively in the HDL fraction. There was no correlation between the peroxidase activity and the HDL or LDL cholesterol concentrations. Finally, at physiologic concentrations of GSH (9 micromol/L), the peroxidase was inhibited by physiologic, free homocysteine concentrations (1 to 5 micromol/L). These data suggest that the peroxidase is active in vivo and may be important in protecting the endothelium from atherosclerosis by preventing oxidant injury. The homocysteine inhibition of the peroxidase suggests a possible biochemical basis for the observed association between hyperhomocysteinemia and cardiovascular disease. Our studies imply that low concentrations of this peroxidase may be an independent risk factor for atherosclerosis.

Effect of methoxychlor administration to male rats on hepatic, microsomal iodothyronine 5'-deiodinase, form I.

We previously reported that methoxychlor administration inhibits the activity of the hepatic, microsomal iodothyronine 5'-deiodinase, form I (ID-I; ). Our data further suggested that the inhibition was due to the covalent binding of a methoxychlor metabolite to a 56-kDa protein identified as ID-I (; ). This protein is 98% homologous to the thiol:protein disulfide oxidoreductase, form Q5 (ERp55;; ). Although at the time there was some controversy, most studies now suggest that ID-I is actually catalyzed by a 27-kDa selenoprotein that does not form adducts with methoxychlor (;; ). Because the 27-kDa protein is considered to be ID-I instead of ERp55, we have further examined the basis for the decreased ID-I activity observed after methoxychlor administration. Male, 150- to 200-g Sprague-Dawley rats were given methoxychlor (0-100 mg/kg/day) in corn oil by gavage for 14 days. ID-I was determined by a thyronine-specific immunoassay. Treated rats showed a significant 15% decline in total hepatic, microsomal protein at all doses. The ID-I-specific activity showed a linear decrease with increasing log doses of methoxychlor. The maximum decrease was 42% at 100 mg/kg/day. The 27-kDa protein specific content declined 37%. In rats given methoxychlor the ratios of the 27-kDa protein mRNA to the 18S ribosomal RNA declined from 2.2 +/- 0.27 x 10(-3) (controls) to 0.99 +/- 0.09 x 10(-3) (100 mg/kg/day). These data suggest that the decreased ID-I observed with chronic methoxychlor administration was due to decreased transcription or stability of the mRNA encoding the 27-kDa protein.

Within-person variability of the ratios of urinary 2-hydroxyestrone to 16alpha-hydroxyestrone in Caucasian women.

The ratio of urinary 2-hydroxyestrone (2-OHE1) to 16alpha-hydroxyestrone (16alpha-OHE1) has been suggested as a potential biomarker for breast cancer risk. We evaluated within-person variability of this biomarker in ten healthy Caucasian women aged 23-58 years. Each study participant was asked to provide an overnight fasting morning urine sample once a week for an average of 8 weeks. These urine samples were assayed for 2-OHE1 and 16alpha-OHE1 by using competitive enzyme immunoassay kits purchased from the ImmunaCare Corporation. The coefficients of variation for urinary 2-OHE1/16alpha-OHE1 over the study period ranged from 13.7 to 59.6% (mean, 33.3%) in our study participants. There was a good correlation between the level of the urinary 2-OHE1/16alpha-OHE1 ratio in any single urine sample and the average ratio over the 8-week study period from the same woman, with the mean correlation coefficient of 0.85. These results indicated that the within-person variation of the 2-OHE1 to 16alpha-OHE1 ratio for most women was moderate and the level of this ratio in a single urine sample, in general, reflects reasonably well the level of this biomarker over a 2-month period.

Catalysis of the cysteine conjugation and protein binding of acetaminophen by microsomes from a human lymphoblast line transfected with the cDNAs of various forms of human cytochrome P450.

We have previously found that for acetaminophen kinetic differences exist between the hepatic microsomal catalyzed protein binding and cysteine conjugation. We have also observed that the protein binding of acetaminophen is only to intralumenal proteins. Together these data suggested that two pools of the reactive metabolite, N-acetyl-p-benzoquinone imine (NABQI), are formed during the oxidative metabolism of acetaminophen: one on the cytosolic surface and the other within the lumen of the microsomes. This would indicate that some of forms of cytochrome P450 (CYP) catalyzing NABQI formation have their active site on the cytosolic surface and others on the lumenal surface. We have examined this question by comparing the rates of cysteine conjugation and protein binding of acetaminophen by microsomes from lymphoblasts transfected with the cDNAs for human CYPs. We found that CYP2D6 catalyzed only cysteine conjugation; CYP1A2 and 3A4 catalyzed only protein binding; CYP2E1 catalyzed both; and CYP1A1, CYP2A6 and CYP2B6 catalyzed neither. These data suggest that CYP2D6 has its active site only on the cytosolic surface; CYP1A2 and CYP3A4 only on the lumenal surface; and CYP2E1 has catalytic sites on both the lumenal and cytosolic surfaces of the membrane. In mouse studies we have found that ethanol administration increased acetaminophen protein binding by 265% but cysteine conjugation by only 61%. CYP2E1 and CYP2B increased, whereas CYP3A decreased and the others did not change. These data suggest that in control mice CYP2E1 catalyzes the bulk of protein binding, whereas CYP2D catalyzes slightly more cysteine conjugation than does CYP2E1.

Studies comparing the kinetics of cysteine conjugation and protein binding of acetaminophen by hepatic microsomes from male mice.

A large body of evidence indicates that acetaminophen toxicity is mediated through the formation of the reactive metabolite, N-acetyl-p-benzoquinone imine (NABQI). Two assays have been employed to monitor NABQI formation by hepatic microsomes: the conjugation with a thiol trap, such as cysteine or glutathione, and the binding of NABQI to microsomal proteins. Studies from our laboratory with rat hepatic microsomes have suggested that the two assays may not be equivalent. We now find with mouse hepatic microsomes that there are also marked differences between these two assays. Among these the rate of cysteine conjugation was almost three orders of greater than that of protein binding. Furthermore, ethanol feeding increased protein binding by 97%, but cysteine conjugation by only 33%. Protein binding was linear for 20 min while cysteine conjugation was linear for only 5 min. CO, imidazole and metyrapone inhibited cysteine conjugation much more than protein binding while SKF-525A and KCN had similar effects on both reactions. Both reactions increased linearly with increasing [NADPH] up to 0.32 mM. At higher concentrations, the rate of cysteine conjugation markedly decreased while the rate of protein binding plateaued. The addition of equimolar concentrations of NADH decreased protein binding, but had no effect on cysteine conjugation. NADPH reduced the protein binding of added NABQI while NADH had little effect. The reduction of NABQI back to acetaminophen was equal for NADPH and NADH. These data indicate that the formation of the cysteine conjugate of NABQI has markedly different kinetic characteristics than the microsomal protein binding. These data might suggest that the two reactions are catalyzed in part by different isoforms of cytochrome P-450.

The covalent binding of [14C]acetaminophen to mouse hepatic microsomal proteins: the specific binding to calreticulin and the two forms of the thiol:protein disulfide oxidoreductases.

Numerous in vitro studies have indicated that acetaminophen is activated by mouse hepatic microsomal cytochrome P450 to form N-acetylbenzoquinone imine. This in turn covalently binds through a Michael addition to protein sulfhydryl and amino groups. Although acetaminophen adducts of several cytosolic proteins have been purified after its administration in vivo, no adducts of specific microsomal proteins have been reported. We find that, after the in vitro incubation of mouse hepatic microsomes with [ring-14C] acetaminophen in the presence of an NADPH generating system, 95% of the bound radioactivity was associated with adducts to three intraluminal microsomal proteins: calreticulin and the two forms of thiol:protein disulfide oxidoreductase, Q2 and Q5. The acetaminophen bound to 0.35, 1.32, and 0.25 mol/mol of the three proteins, respectively. Sequencing of the 14C-labeled tryptic peptides indicated that the acetaminophen bound to lysine 103 of Q2, lysines 202, 209 or 210 and 354 of Q5 and lysines 233 or 239 of calreticulin. No adducts of cysteine residues were observed. Our data might suggest that acetaminophen hepatotoxicity results from the formation of the reactive metabolite within the endoplasmic reticulum. This then binds to these essential proteins and blocks the posttranslational modification of secretory and membrane proteins. This inhibition could then lead to cellular injury and death.

Metabolism of phenytoin by the gingiva of normal humans: the possible role of reactive metabolites of phenytoin in the initiation of gingival hyperplasia.

Gingival hyperplasia is a well-known complication of therapy with cyclosporine, calcium channel blockers, and phenytoin. It is characterized by the presence of inflammation and a marked fibrotic response. The mechanism of this adverse reaction is unknown. We propose that it may be initiated by the metabolic activation of these drugs to form reactive metabolites. These then cause cellular injury and lead to the gingival hyperplasia. To evaluate this hypothesis we examined phenytoin metabolism and the cytochrome P450 contents of gingival tissues from 10 patients undergoing surgery for various periodontal conditions. We found that microsomes obtained from the gingiva show significant phenytoin hydroxylase activity as determined by the production of 5-(4'-hydroxyphenyl)-5-phenylhydantoin (HPPH) (range, 12.8 pmol HPPH/min.mg microsomal protein to 276.9 pmol HPPH/min.mg microsomal protein; rat control, 133.7 +/- 11.5 pmol HPPH/min.mg microsomal protein). We also found that CYP1A1, CYP1A2, CYP2C9, CYP2E1, and CYP3A4 were present in these microsomes. We detected no CYP2B6 or CYP2D6. We believe that these data support our hypothesis that the proliferative inflammation observed with drugs such as phenytoin, nifedipine, and cyclosporine may be initiated by the formation of reactive metabolites and that the formation of these metabolites may be catalyzed by one or more CYPs found in the gingiva. These metabolites may then cause cellular injury and induce a reactive inflammatory response, followed by fibroblastic proliferation. This proliferation leads to the excess collagen deposition observed with gingival hyperplasia.

Evidence that casein kinase 2 phosphorylates hepatic microsomal calcium-binding proteins 1 and 2 but not 3.

We have extensively purified three of the hepatic microsomal intralumenal Ca2+-binding proteins, CBP1, CBP2, and CBP3, which were originally described by Van et al. [(1989) J. Biol. Chem. 264, 17494-17501]. These apparently homogeneous preparations showed only single 45Ca2+ binding bands. On the basis of the peptide sequence, CBP2 was found to be highly homologous with the previously described protein ERp72. Similarly, CBP3 was identical to calreticulin and CBP1 had some homology to calmodulin. Contrary to the report of Van et al. (1989), we found that CBP2 had little thiol:protein disulfide oxidoreductase activity. Of the three purified preparations, only CBP2 exhibited apparent intrinsic protein kinase activity. This activity was found to be due to contamination of the CBP2 preparation by an extremely low concentration of tightly bound casein kinase 2 (CK2). In line with this observation, the phosphorylation was inhibited by heparin, removed by antibody to CK2, and stimulated by spermine. Furthermore, CBP2 was readily phosphorylated in vitro by added CK2 but only slowly phosphorylated by several other protein kinases. Thus, the persistence of CK2 in a highly purified preparation of CBP2 along with several other lines of evidence presented in this study might suggest that the protein CBP2 is a physiologically relevant substrate for CK2. Furthermore, these data suggest that CK2 might be localized in the lumen of the endoplasmic reticulum and that the phosphorylation of CBP2 in the lumen may play a role in the chaperone activity attributed to this protein.

Cytochrome P450 catalyzed covalent binding of methoxychlor to rat hepatic, microsomal iodothyronine 5'-monodeiodinase, type I: does exposure to methoxychlor disrupt thyroid hormone metabolism?

The insecticide methoxychlor is estrogenic in birds and mammals and interferes with sexual development and reproduction, but it is not known whether this toxicity is due solely to its estrogenicity. We now have found that during hepatic, microsomal metabolism of [ring-14C]- or [3H-OCH3]methoxychlor, their metabolite primarily binds to iodothyronine 5'-monodeiodinase, type I (5'-ID1). The purified, radiolabeled protein reacted with antibodies against protein disulfide isomerase, isoform Q5, which is highly homologous to 5'-ID1. Sequencing of the radiolabeled tryptic peptide indicated that methoxychlor bound to cysteine 372 or 375 or to lysine 376 of 5'-ID1. Treatment of rats with methoxychlor for 4 days decreased hepatic, microsomal 5'-ID1 activity from 2.94 to 2.20 nmol/min-mg prot (P < 0.02). Since 5'-ID1 catalyzes thyroxine conversion to the biologically active triiodothyronine, these data suggest that methoxychlor may interfere with thyroid hormone metabolism. This may be an additional factor in its environmental toxicity.

Pharmacokinetic interaction between isradipine and lovastatin in normal, female and male volunteers.

We have examined the pharmacokinetic interaction between isradipine and lovastatin in six male and six female, healthy, normotensive, human subjects after a single dose and after treatment for 5 days. The isradipine plasma concentrations were determined by a radioimmunoassay and the lovastatin serum concentrations by gas chromatography-mass spectrometry (GC/MS) and by the inhibition of the 3-hydroxy-3-methylglutaric-coenzyme reductase activity. We found that the apparent serum concentrations of lovastatin were 4- to 6-fold higher in the reductase-inhibition assay than the GC/MS assay, suggesting that the bulk of the reductase inhibition is due to active metabolites. The peak and the time-to-peak concentrations were unaffected by the treatments, either after the first dose or after continued administration. In male subjects, after repeated doses of isradipine, the lovastatin area under the time-concentration curves (AUCs) decreased by 40% as determined by the GC/MS assay (P < .001) and 20% as determined by the reductase-inhibition assay (P < .0022). In the female subjects, isradipine treatment decreased the lovastatin AUCs as determined by the GC/MS assay, but this was not statistically significant due to a high variance. Furthermore, in the female subjects, isradipine had no effect on the lovastatin AUCs as determined by the reductase-inhibition assay. Because the lovastatin peak and the time-to-peak concentrations were unaffected by isradipine treatment, the decreased lovastatin AUCs were probably not due to altered intestinal absorption. More likely, because lovastatin has a high hepatic clearance, the decreased AUCs seen after isradipine treatment could be due to increases in the clearance of lovastatin secondary to increased hepatic blood flow.

One year experience of elderly hypertensive patients with isradipine therapy.

This is the first report of long-term use (one year) of isradipine, a new dihydropyridine calcium channel blocker, in the treatment of elderly patients with essential hypertension. Patients completing a three month, double-blind, multicentre study comparing isradipine to hydrochlorothiazide (HCTZ) were eligible to enroll in this open-label, continuation study. At initial baseline, patients were at least 60 years of age and had DBP from 95 mmHg to 120 mmHg. Patients were titrated when necessary every two weeks with isradipine, 5 mg to 15 mg once daily or 2.5 mg to 10 mg twice daily, to maintain sitting DBP < or = 90 mmHg. HCTZ, 12.5 mg to 50 mg once daily, could be added for better BP control. A total of 136 patients completed the one year, open-label phase. One hundred and fourteen patients (84%) received isradipine as monotherapy (mean dose, 9.7 mg/day); 22 received concomitant HCTZ therapy at one year. Reduction in DBP was significant and similar among all age groups and races (mean change of -19 mmHg). Reduction in SBP was similar among all age groups. Ninety-four per cent of those receiving isradipine monotherapy achieved BP control during the last four months of treatment. Twenty-six patients (16%) withdrew from the study: 11 (7%) had adverse reactions (one with headache, two with pedal oedema, eight with other problems); 11 (7%) had nondrug-related problems; and in four (2%), the drugs were ineffective. Based on these observations, isradipine is a well-tolerated, safe and effective agent for long-term BP control in elderly patients with essential hypertension.

Induction of cytochrome CYPIA1 and formation of toxic metabolites of benzo[a]pyrene by rat aorta: a possible role in atherogenesis.

Cigarette smoking is a leading risk factor for atherosclerosis. Endothelial injury may be the initial event in this process. The carcinogenic metabolites of the polycyclic aromatic hydrocarbons found in cigarette smoke tars could cause this injury. We tested this model by examining the effect of 3-methylcholanthrene administration on aortic polycyclic aromatic hydrocarbon metabolism. Immunoblotting with a monoclonal antibody (mAb 1-7-1) specific for cytochromes CYPIA1 and CYPIA2 showed that aortic microsomes from treated, but not from control, animals contained CYPIA1; the CYPIA1 was primarily in the endothelium. Aortic microsomes from induced animals metabolized benzo[a]pyrene (BaP) to the 7R,8S,9,10-tetrahydrotetrol-, 7,8-dihydrodiol-, 1,6 quinone-, 3,6 quinone-, 6,12 quinone-, 3-hydroxy-, and 9-hydroxy-BaP. mAb 1-7-1 inhibited the formation of the tetrahydrotetrol, the dihydrodiol-BaP, and the 3-hydroxy-BaP but did not inhibit the quinones or the 9-hydroxy-BaP. Arachidonic acid did not affect metabolism. These data suggest that the aortas of induced animals metabolize the BaP in cigarette smoke to carcinogenic and toxic products and that this metabolism may initiate vessel injury and lead to the accelerated atherosclerosis seen in cigarette smokers.

The reported cDNA sequence for phospholipase C alpha encodes protein disulfide isomerase, isozyme Q-2 and not phospholipase-C.

We have previously shown that the cDNA sequence published for the rat gene encoding phosphatidylinositol specific phospholipase C alpha (1) may be identical to the sequence of a new isoform of protein disulfide isomerase, Isozyme Q2 (2). To determine whether the presumed phospholipase C alpha gene encodes phospholipase C or protein disulfide isomerase, we have now expressed the cDNA in Escherichia coli and find that the 55.5 kDa protein that it synthesized contains no phospholipase C activity. Moreover, an extract of Escherichia coli containing the expressed protein shows a concentration dependent increase in protein disulfide isomerase activity. The expressed protein is identical to protein disulfide isomerase, Isozyme Q2 on SDS-PAGE and by Western blot analysis. These current studies confirm that the published sequence of phospholipase C alpha is, in fact, the sequence of protein disulfide isomerase, Isozyme Q2.

Potentiation of acute acetaminophen lethality by selenium and vitamin E deficiency in mice.

The effects of selenium, vitamin E, and DL-methionine deficiency on the acute lethality and hepatotoxicity of acetaminophen in male CD-1 mice were studied. Vitamin E and selenium deficiencies led to an increase in the acute lethality of acetaminophen, with a decrease in the LD50 from 376 to 84 mg/kg. These dietary deficiencies impaired the inducibility of the hepatic microsomal mixed function oxidase system by phenobarbital, but on the basis of the covalent binding of acetaminophen to microsomes, these treatments did not alter the activation of acetaminophen to a reactive intermediate by this system. Addition of methionine to the deficient diet restored hepatic glutathione content to control levels but did little to protect against the acute lethality of acetaminophen. In methionine-supplemented animals, the addition of either selenium or vitamin E increased the LD50 of acetaminophen to 167 and 200 mg/kg, respectively. Administration of a sublethal, toxic dose of acetaminophen (LD30) to the methionine-supplemented and selenium- and vitamin E-deficient mice did not produce any hepatic damage as evidenced by a lack of plasma aminotransferase elevation. In view of the known antioxidant effects of vitamin E and selenium, these data suggest the involvement of a reactive radical in the acute lethality of acetaminophen and further suggest that death from acute acetaminophen overdose in chronic selenium- and vitamin E-deficient mice may be unrelated to liver necrosis.

Purification and characterization of a new isozyme of thiol:protein-disulfide oxidoreductase from rat hepatic microsomes. Relationship of this isozyme to cytosolic phosphatidylinositol-specific phospholipase C form 1A.

Thiol:protein-disulfide oxidoreductase catalyzes the GSH reduction of protein disulfides to sulfhydryls. Chromatography of solubilized hepatic microsomes on Mono Q yielded two peaks, Q-2 and Q-5, which contained all the thiol:protein-disulfide oxidoreductase activity. These were further purified by chromatofocusing giving specific activities of 14.4 and 45.9 nmol/mg of protein/min, respectively with purifications of 45.0- and 143.6-fold. Amino acids 1-18 of Q-5 were the same as previously reported for Thiol:protein-disulfide oxidoreductase (Edman, J. C., Ellis, L., Blacher, R. W., Roth, R. A., and Rutter, W. J. (1985) Nature 317, 267-270), except amino acid 1 was leucine instead of aspartate and amino acid 6 was asparagine instead of glutamate. The N-terminal amino acid sequence of Q-2 differed markedly from Q-5 but Q-2 showed 100% identity at amino acids 25-54, 258-269, 285-310, 347-350, 412-419, and 434-463 for the reported sequence of rat, hepatic, cytosolic phosphatidylinositol-specific phospholipase C form 1a (PLC) (Bennett, C. F., Balcarek, J. M., Varrichio, A., and Crooke, S. T. (1988) Nature 334, 268-270). PLC activity was found in the elution from the Mono Q column, but none was found in purified Q-2 or Q-5. Antibodies to Q-5 reacted with Q-2, but anti-Q-2 did not react with Q-5. Anti-Q-2 antibody showed immunoreactivity with 55- and 60-kDa microsomal proteins, whereas Q-5 antibody reacted with a number of microsomal proteins. Although Q-2 was immunoreactive with a polyclonal antibody to guinea pig, uterine cytosolic PLC, partially purified PLCs from rat liver cytosol did not react to this antibody. Our data would suggest that the published sequence for PLC form 1a may actually be the sequence for Q-2.

The effect of substrates and inhibitors of cytochrome P-450 on the NADPH inhibition of the ATP-dependent, hepatic, microsomal calcium pump.

In hepatic microsomes one or more isozymes of cytochrome P-450 inhibits the ATP-dependent Ca2+ pump. This inhibition is reversible by GSH and appears to be due to a direct oxidation of the pump proteins by the oxygenated cytochrome. To determine which isozyme mediates this inhibition, we have examined the effect of various substrates and inhibitors on the NADPH inhibition of Ca2+ uptake. We find that aminopyrine, benzphetamine and SKF-525A reverse this inhibition while a number of other substrates do not. This pattern suggests that a previously unreported isozyme of cytochrome P-450 mediates the Ca2+ pump inhibition.