Sequencing, expression, and characterization of cDNA expressed flavin-containing monooxygenase 2 from mouse

The cDNA clone of mouse flavin-containing monooxygenase 2 (FMO2) was obtained as an expressed sequence tag (EST) isolated from a female mouse kidney cDNA library from the I.M.A.G.E. consortium (I.M.A.G.E. CloneID 1432164). Complete sequencing of the EST derived a nucleotide sequence for mouse FMO2, which contains 112 bases of 5' flanking region, 1607 bases of coding region, and 309 bases of 3' flanking region. This FMO2 sequence encodes a protein of 535 amino acids including two putative pyrophosphate binding sequences (GxGxxG/A) beginning at positions 9 and 191. Additionally, this mouse FMO protein sequence shows 87 and 86% homology to rabbit and human FMO2 respectively. The mouse FMO2 sequence was subcloned into the expression vector pJL-2, a derivative of pKK233-2 and used to transform XL1-Blue Escherichia coli. FMO activity in particulate fractions isolated from isopropyl-beta-D-thiogalactopyanoside (IPTG) induced cells was heat stable (45 degrees C for 5 min) and demonstrated optimal activity at a relatively high pH of 10.5. The expressed FMO2 enzyme showed catalytic activity towards the FMO substrate methimazole and further analysis of E. coli fractions utilizing NADPH oxidation demonstrated that the mouse FMO2 enzyme also exhibits catalytic activity towards thiourea, trimethylamine, and the insecticide phorate.     

Functional activity of the mouse flavin-containing monooxygenase forms 1, 3, and 5

Three functional mouse flavin-containing monooxygenases (mFMOs) (i.e., mFMO1, mFMO3, and mFMO5) have been reported to be the major FMOs present in mouse liver. To examine the biochemical features of these enzymes, recombinant enzymes were expressed as maltose-binding protein fusion proteins (i.e., MBP-mFMO1, MBP-mFMO3, and MBP-mFMO5) in Escherichia coli and isolated and purified with affinity chromatography. The substrate specificity of these three mouse hepatic FMO enzymes were examined using a variety of substrates, including mercaptoimidazole, trimethylamine, S-methyl esonarimod, and an analog thereof, and a series of 10-(N,N-dimethylaminoalkyl)-2-(trifluoromethyl)phenothiazine analogs. The kinetic parameters of the three mouse FMOs for these substrates were compared in an attempt to explore substrate structure--function relationships specific for each mFMO. Utilizing a common phenothiazine substrate for all three enzymes, we compared the pH dependence for the recombinant enzymes under similar conditions. In addition, thermal stability for mFMO1, mFMO3, and mFMO5 enzymes was examined in the presence and absence of NADPH. The results revealed unique features for mFMO5, suggesting possible impact on the functional significance of this abundantly expressed FMO5 isoform in both human and mouse liver.     

Potential roles of flavin-containing monooxygenases in sulfoxidation reactions of l-methionine, N-acetyl-l-methionine and peptides containing l-methionine

Flavin-containing monooxygenases (FMOs) are microsomal enzymes that catalyze the NADPH-dependent oxidation of a large number of sulfur-, selenium-, and nitrogen-containing compounds. Five active isoforms (FMO1-5) have been identified and shown to be differently expressed in various mammalian tissues. Previous work from this laboratory has shown l-methionine to be S-oxidized by rat, rabbit and human FMO1-4, with FMO3 exhibiting the highest stereoselectivity for the formation of the d-diastereomer of methionine sulfoxide. In this report, we describe new studies aimed at determining if N-acetyl-l-methionine and peptides containing l-methionine can be substrates for FMOs. Experiments were carried out using either rabbit liver microsomes or human cDNA-expressed FMOs. The results show that while N-acetyl-l-methionine and peptides with a modified methionine amino group may not function as substrates for FMOs, peptides containing a free N-terminal methionine may act as FMO substrates. With human cDNA-expressed FMO1, FMO3, and FMO5, both FMO1 and FMO3 exhibited activity with the active peptides whereas FMO5 was inactive. With FMO3, the activity measured with methionine was similar (1 mM) or higher (5 mM) than the activity measured with H-Met-Val-OH and H-Met-Phe-OH. With FMO1, H-Met-Phe-OH and methionine exhibited similar activities whereas activity with H-Met-Val-OH was much lower. Collectively, the results show that FMOs can oxidize peptides containing a free N-terminal methionine. Thus, the role of FMOs in the oxidation of methionine in larger peptides or proteins warrants further investigation.     

Characterization of the purified microsomal FAD-containing monooxygenase from mouse and pig liver

The FAD-containing monooxygenase (FMO) has been purified from both mouse and pig liver microsomes by similar purification procedures. Characterization of the enzyme from these two sources has revealed significant differences in catalytic and immunological properties. The pH optimum of mouse FMO is slightly higher than that of pig FMO (9.2 vs. 8.7) and, while pig FMO is activated 2-fold by n-octylamine, mouse FMO is activated less than 20%. Compounds, including primary, secondary and tertiary amines, sulfides, sulfoxides, thiols, thioureas and mercaptoimidazoles were tested as substrates for both the mouse and pig liver FMO. Km- and Vmax-values were determined for substrates representative of each of these groups. In general, the mouse FMO had higher Km-values for all of the amines and disulfides tested. Mouse FMO had Km-values similar to those of pig FMO for sulfides, mercaptoimidazoles, thioureas, thiobenzamide and cysteamine. Vmax-values for mouse FMO with most substrates was approximately equal, indicating that as with pig FMO, breakdown of the hydroxyflavin is the rate limiting step in the reaction mechanism. Either NADPH or NADH will serve as an electron donor for FMO, however, NADPH is the preferred donor. Pig and mouse FMOs have similar affinity for NADPH (Km = 0.97 and 1.1 microM, respectively) and for NADH (Km = 48 and 73 microM, respectively). An antibody, prepared by immunizing rabbits with purified pig liver FMO, reacts with purified pig liver FMO but not with mouse liver FMO, indicating structural differences between these two enzymes. This antibody inhibited pig FMO activity up to 60%.     

Molecular Cloning,Sequence, and Expression of Mouse Flavin-Containing Monooxygenases 1 and 5 (FMO1 and FMO5)

Full-length cDNA clones encoding FMO1 and FMO5 have been isolated from a library constructed with mRNA from the liver of a female CD-1 mouse. The derived sequence of FMO1 contains 2310 bases: 1596 in the coding region, 301 in the 5′-flanking region, and 413 in the 3′-flanking region. The sequence for FMO5 consists of 3168 bases; 1599 in the coding region, 812 in the 5′-flanking region, and 757 in the 3′-flanking region. The sequence of FMO1 encodes a protein of 532 amino acids with a predicted molecular weight of 59.9 kDa and shows 83.3% identity to human FMO1 and 83–94% identity to other FMO1 homologs. FMO5 encodes a protein of 533 amino acids with a predicted molecular weight of 60.0 kDa and 84.1% identity to human FMO5 and 83–84% identity to other FMO5 orthologs. Two GxGxxG putative pyrophosphate binding domains exist beginning at positions 9 and 191 for FMO1, and 10 and 192 for FMO5. Mouse FMO1 and FMO5 were expressed in E. coli and show similar mobility to the native proteins as determined by SDS-PAGE. The expressed FMO1 protein showed activity toward methimazole, and FMO5 was active toward n -octylamine. In addition, FMO1 was shown to metabolize radiolabeled phorate, whereas FMO5 showed no activity toward phorate. © 1998 John Wiley & Sons, Inc. J Biochem Toxicol 12: 205–212, 1998     

Glutathione reductases from a variety of sources are inhibited by physiological levels of glutathione.

1. Glutathione reductase from human platelets, bovine intestinal mucosa, yeast and E. coli were inhibited in vitro by physiological levels of reduced glutathione with IC50s of 6.61 mM, 2.92 mM, 2.40 mM and 12.11 mM, respectively. 2. A steady-state kinetic examination revealed that glutathione inhibited the NADPH oxidation (at constant [glutathione-disulphide]) catalysed by the eucaryotic enzymes uncompetitively, whereas the E. coli enzyme appeared unaffected by glutathione concentrations of up to 10 mM. 3. With respect to glutathione inhibition of glutathione-disulphide reduction (at constant [NADPH]), the human enzyme was inhibited uncompetitively; the bovine and yeast enzymes displayed apparent mixed hyperbolic inhibition; the E. coli enzyme was inhibited competitively.     

Functional characterization of genetic variants of human FMO3 associated with trimethylaminuria

Impaired conversion of trimethylamine to trimethylamine N-oxide by human flavin containing monooxygenase 3 (FMO3) is strongly associated with primary trimethylaminuria, also known as 'fish-odor' syndrome. Numerous non-synonymous mutations in FMO3 have been identified in patients suffering from this metabolic disorder (e.g., N61S, M66I, P153L, and R492W), but the molecular mechanism(s) underlying the functional deficit attributed to these alleles has not been elucidated. The purpose of the present study was to determine the impact of these disease-associated genetic variants on FMO3 holoenzyme formation and on steady-state kinetic parameters for metabolism of several substrates, including trimethylamine. For comparative purposes, several common allelic variants not associated with primary trimethylaminuria (i.e., E158K, V257M, E308G, and the E158K/E308G haplotype) were also analyzed. When recombinantly expressed in insect cells, only the M66I and R492W mutants failed to incorporate/retain the FAD cofactor. Of the remaining mutant proteins P153L and N61S displayed substantially reduced (<10%) catalytic efficiencies for trimethylamine N-oxygenation relative to the wild-type enzyme. For N61S, reduced catalytic efficiency was solely a consequence of an increased K(m), whereas for P153L, both K(m) and k(cat) were altered. Similar results were obtained when benzydamine N-oxygenation was monitored. A homology model for FMO3 was constructed based on the crystal structure for yeast FMO which places the N61 residue alone, of the mutants analyzed here, in close proximity to the FAD catalytic center. These data demonstrate that primary trimethylaminuria is multifactorial in origin in that enzyme dysfunction can result from kinetic incompetencies as well as impaired assembly of holoprotein.     

Characterization and modulation by drugs of sheep liver microsomal flavin monooxygenase activity

The flavin monooxygenases (FMO) catalyse the NADPH and oxygen-dependent oxidation of a wide range of nucleophilic nitrogen-, sulfur-, phosphorus-, and selenium heteroatom-containing chemicals, drugs, and agricultural agents. In the present study, sheep liver microsomal FMO activity was determined by measuring the S-oxidation rate of methimazole and the average specific activity obtained from different microsomal preparations was found to be 3.8 +/- 1.5 nmol methimazole oxidized min(-1) mg(-1) microsomal protein (mean +/- SE, n = 7). The presence of 0.1% Triton X-100 in the reaction mixture caused an increase of specific sheep liver microsomal FMO activity towards methimazole to 6.1 +/- 1.4 nmol methimazole oxidized min(-1) mg(-1) microsomal protein (mean +/- SE, n = 6). Metabolism of imipramine and chlorpromazine was measured by following the oxidation of cofactor NADPH spectrophotometrically at 340 nm. Sheep liver microsomal FMO activity towards imipramine and chlorpromazine was found to be 10.7 and 12.3 nmol NADPH oxidized min(-1) mg(-1) microsomal protein, respectively. Characterization of sheep liver enzyme was carried out using methimazole as substrate and the maximum FMO enzyme activity was detected at 37 degrees C and at pH 8.0. The apparent K(m) value of sheep liver microsomal FMO for methimazole was 0.118 mM. Effects of the detergents Triton X-100, Cholate, and Emulgen 913, on FMO activity were determined and FMO activity was found to increase with the addition of detergents to the reaction medium. Sheep liver microsomal FMO-catalysed methimazole oxidation was inhibited by imipramine and chlorpromazine when these drugs were used at high concentrations. Western blot-immunochemical analysis revealed the presence of FMO3 in sheep liver microsomes.     

A cell-free preparation of human neutrophils catalyzing NADPH-dependent conversion of leukotriene B4

The sonicate of human neutrophils converted leukotriene B4 to a polar product in aerobic condition in the presence of NADPH at a rate comparable to that of the intact cells. NADH could scarcely replace NADPH. The conversion was not observed in anaerobic conditions and was inhibited by carbon monoxide (CO/O2 = 4/1) or by 1 mM p-chlormercuribenzoate, while it was not affected by 1 mM KCN, 5 mM NaN3, 200 micrograms/ml catalase, 100 mM mannitol, and 10 micrograms/ml superoxide dismutase. These observations suggest that the myeloperoxidase-H2O2-halide system and active oxygen species are not involved in the reaction. The activity was observed in the 100,000xg supernatant from the homogenate, in which cytochrome P-450 was not detected.     

Phospholipid hydroperoxides are substrates for non-selenium glutathione peroxidase.

This study investigated phospholipid hydroperoxides as substrates for non-selenium GSH peroxidase (NSGPx), an enzyme also called 1-Cys peroxiredoxin. Recombinant human NSGPx expressed in Escherichia coli from a human cDNA clone (HA0683) showed GSH peroxidase activity with sn-2-linolenoyl- or sn-2-arachidonoyl-phosphatidylcholine hydroperoxides as substrate; NADPH or thioredoxin could not substitute for GSH. Activity did not saturate with GSH, and kinetics were compatible with a ping-pong mechanism; kinetic constants (mM(-1) min(-1)) were k(1) = 1-3 x 10(5) and k(2) = 4-11 x 10(4). In the presence of 0.36 mM GSH, apparent K(m) was 120-130 microM and apparent V(max) was 1.5-1.6 micromol/min/mg of protein. Assays with H(2)O(2) and organic hydroperoxides as substrate indicated activity similar to that with phospholipid hydroperoxides. Maximal enzymatic activity was at pH 7-8. Activity with phospholipid hydroperoxide substrate was inhibited noncompetitively by mercaptosuccinate with K(i) 4 miroM. The enzyme had no GSH S-transferase activity. Bovine cDNA encoding NSGPx, isolated from a lung expression library using a polymerase chain reaction probe, showed >95% similarity to previously published human, rat, and mouse sequences and does not contain the TGA stop codon, which is translated as selenocysteine in selenium-containing peroxidases. The molecular mass of bovine NSGPx deduced from the cDNA is 25,047 Da. These results identify a new GSH peroxidase that is not a selenoenzyme and can reduce phospholipid hydroperoxides. Thus, this enzyme may be an important component of cellular antioxidant defense systems.     

A novel flavin-containing monooxygenase from Methylophaga sp strain SK1 and its indigo synthesis in Escherichia coli

We cloned a gene from Methylophaga sp. strain SK1. This gene was responsible for producing, the blue pigment, indigo. The complete open reading frame was 1371 bp long, which encodes a protein of 456 amino acids. The molecular mass of the encoded protein was 105 kDa, consisting of homodimer of 54 kDa with an isoelectric point of 5.14. The deduced amino acid sequence from the gene showed approximately 30% identities with flavin-containing monooxygenases (FMOs) of human (FMO1-FMO5), Arabidopsis, and yeast. It contained three characteristic sequence motifs of FMOs: FAD binding domain, FMO-identifying sequence motif, and NADPH binding domain. In addition, the biochemical properties such as substrate specificities and absorption spectra were similar to the eukaryotic FMO families. Thus, we assigned the enzyme to be a bacterial FMO. The recombinant Escherichia coli expressing the bacterial FMO produced up to 160 mg of indigo per liter in the tryptophan medium after 12h cultivation. This suggests that the recombinant strain has a potential to be applied in microbial indigo production.     

Noxa1 is a central component of the smooth muscle NADPH oxidase in mice

NADPH oxidase is the most important source of oxygen-derived radicals (ROS) in the vascular wall. In vascular smooth muscle cells (VSMC), NADPH oxidase is characterized by the expression of the membrane subunit Nox1, which is activated by cytoplasmic proteins binding to its activation domain. We set out to identify the cytoplasmic protein involved in NADPH oxidase activation in mouse VSMC. Western blot analysis revealed that human endothelial cells and leukocytes but not VSMC from the aorta of the rat and the mouse express the classic NADPH oxidase activator p67phox. In mouse VSMC, however, the p67phox homologue Noxa1 was detected. Using antibodies generated against mouse Noxa1, the protein was observed in the cytosolic fraction of mouse VSMC with a molecular weight of about 51 kDa. Immunohistochemistry revealed that Noxa1 is expressed in the smooth muscle layer but not in endothelium or the adventitia of the mouse carotid artery. Fluorescent fusion proteins of Noxa1 were observed to be expressed in the cytoplasm of VSMC and coexpression of the NADPH oxidase organizer Noxo1 targeted the complex to membrane. An antisense plasmid of Noxa1 attenuated the endogenous Noxa1 protein expression in VSMC. This plasmid attenuated the ROS formation in mouse VSMC as detected using L012 chemiluminescence and prevented the agonist-induced ROS production in response to basic fibroblast growth factor and epidermal growth factor. In conclusion, these data indicate that Noxa1 replaces p67phox in VSMC and plays a central role in the activation of the NADPH oxidase in the vascular wall.     

Electrochemical sensing of trimethylamine based on polypyrrole-flavin-containing monooxygenase (FMO3) and ferrocene as redox probe for evaluation of fish freshness

Amperometric and impedimetric biosensor for detecting trimethylamine (TMA) which represents good parameters for estimating fish freshness has been developed. The biosensor is based on a conducting polypyrrole substituted with ferrocenyl, where flavin-containing monooxygenase 3 (FMO3) enzyme was immobilised by covalent bonding. FMO3 catalyzes the monooxygenation TMA to trimethylamine N-oxide (TMO). For catalysis FMO require flavin adenine (FAD) as a prosthetic group, NADPH as a cofactor and molecular oxygen as cosubstrate. Ferrocenyl group substituted on the polypyrrole matrix will serve as redox probe for monitoring the response of the biosensor to TMA. The construction of the biosensor was characterized by FT-IR, cyclic voltammetry and impedance measurements. Detection is done through the analysis of the current of oxidation signal of the ferrocenyl groups and compared to the measurement of impedance related to the electrical properties of the layers. Amperometric and impedimetric response were measured as a function of TMA concentration in range of 0.4 μgm L(-1)-80 μgm L(-1) (6.5 μmol L(-1)-1.5 mmol L(-1)). Amperometric measurements show a decrease in current response which is in correlation with the increase of the charge transfer resistance demonstrated by impedance. Calibration curve obtained by impedance spectroscopy shows a high sensitivity with a dynamic range from (0.4 μgm L(-1) to 80 μgm L(-1)). We demonstrated, using ferrocene as redox probe for catalytic reaction of FMO3, that high sensitivity and dynamic range was obtained. The biosensor was stable during 16 days. The biosensor shows high selectivity and its sensitivity to TMA in real samples was evaluated using fish extract after deterioration during storage.     

The house musk shrew (Suncus murinus): a unique animal with extremely low level of expression of mRNAs for CYP3A and flavin-containing monooxygenase

Expression of drug-metabolizing enzymes including cytochrome P450 (CYP) and flavin-containing monooxygenase (FMO) in various tissues of Suncus murinus (Suncus) were examined. Northern blot analysis showed that mRNAs hybridizable with cDNAs for rat CYP1A2, human CYP2A6, rat CYP2B1, human CYP2C8, human CYP2D6, rat CYP2E1, human CYP3A4 and rat CYP4A1 were expressed in various tissues from Suncus. The mRNA level of CYP2A in the Suncus lung was very high. Furthermore, it was found that the level of CYP2A mRNA in the Suncus lung was higher compared to the Suncus liver. The expression level of mRNA hybridizable with cDNA for human CYP3A4 was very low. The presence of CYP3A gene in Suncus was proven by the induction of the CYP with dexamethasone. Very low expression levels of mRNAs hybridizable with cDNAs for rat FMO1, rat FMO2, rat FMO3 and rat FMO5 were also seen in Suncus liver. No apparent hybridization band appeared when human FMO4 cDNA was used as a probe. The hepatic expression of mRNAs hybridizable with cDNAs for UDP-glucuronosyltransferase 1*6, aryl sulfotransferase, glutathione S-transferase 1, carboxyesterase and microsomal epoxide hydrolase in the Suncus were observed. These results indicate that the Suncus is a unique animal species in that mRNAs for CYP3A and FMO are expressed at very low levels.     

Induction of hepatic drug-metabolising enzymes and tamoxifen metabolite profile in relation to administration route during low-dose treatment in nude rats

Tamoxifen is the most used anticancer drug and is approved for chemoprevention. Little is known about the enzyme inducing properties of low-dose regimens and the influence of route of administration. In this study, nude rats received 5 mg/kg/day of tamoxifen orally or a 50 mg continuous-release pellet subcutaneously. The mRNAs for cytochrome P450-enzymes (CYPs), flavin-containing monooxygenase 1 (FMO1) and phase II drug-metabolising enzymes were quantified by real-time RT-PCR. Tamoxifen and metabolite concentrations were measured using HPLC. We observed a significant increase in CYP3A18 and FMO1 mRNA expression levels in the orally treated animals, whereas the increase in CYP3A2 expression did not reach statistical significance (p = 0.057). No significant induction of enzyme expression was observed in rats that received subcutaneous (S.c.) treatment. After 33 days the serum levels of 4-hydroxytamoxifen (4OHtam), tamoxifen and N-desmethyltamoxifen (NDtam) in orally treated animals were 1.8 ± 0.7, 11.1 ± 3.2 and 11.4 ± 3.8 ng/ml, respectively. In subcutaneously treated animals, tamoxifen and N-desmethyltamoxifen were detected in tissues, but not in serum. These data demonstrate that in contrast to the subcutaneous administration, low-dose oral tamoxifen induced tamoxifen-metabolising enzymes. Furthermore, the different routes of administration resulted in different serum and tissue levels of tamoxifen and metabolites.     

Methionine S-oxidation in human and rabbit liver microsomes: evidence for a high-affinity methionine S-oxidase activity that is distinct from flavin-containing monooxygenase 3.

Methionine has previously been shown to be S-oxidized by flavin-containing monooxygenase (FMO) forms 1, 2, and 3. The most efficient catalyst was FMO3, which has a Km value for methionine S-oxidation of approximately 4 mM, and exhibits high selectivity for formation of the d-diastereoisomer of methionine sulfoxide. The current studies provide evidence for an additional methionine S-oxidase activity in liver microsomes. Human and rabbit liver microsomes exhibited a biphasic response to methionine at concentrations ranging from 0.05 to 10 mM, as indicated by both Eadie-Hofstee plots and nonlinear regression. The low-affinity component of the biphasic response had Km values of approximately 3 and 5 mM for humans and rabbits, respectively, as well as high diastereoselectivity for methionine sulfoxide formation. The low-affinity activity in rabbit liver microsomes was inhibited by methimazole, S-allyl-l-cysteine, and by mild heat treatment, suggesting the activity is FMO3. The high-affinity component of the biphasic response had Km values of approximately 0.07 and 0.04 mM for humans and rabbits, respectively, as well as lower diastereoselectivity for methionine sulfoxide formation. Further characterization of the high-affinity activity in rabbit liver microsomes indicated lack of involvement of cytochrome P450s or reactive oxygen species. The high-affinity activity was inhibited 25% by potassium cyanide and greater than 50% by methimazole and S-allyl-l-cysteine. Mild heat treatment produced 85% inhibition of the low-affinity activity, but only 30% inhibition of the high-affinity activity. Both high- and low-affinity activities were decreased by 85% in flavin-depleted microsomes. Because these results suggested the additional S-oxidase activity has characteristics of an FMO, recombinant human FMO4 was evaluated as a potential catalyst of this activity. Recombinant FMO4 catalyzed S-oxidation of both methionine and S-allyl-l-cysteine, with similar diastereoselectivity to the high-affinity microsomal S-oxidase; however, the Km values for both reactions appeared to be greater than 10 mM. In summary, these studies provide evidence for two microsomal methionine S-oxidase activities. FMO3 is the predominant catalyst at millimolar concentrations of methionine. However, at micromolar methionine concentrations, there is an additional S-oxidase activity that is distinct from FMO3.     

Pyrroline-5-carboxylate synthesis from glutamate by rat intestinal mucosa

The mitochondria of rat intestinal mucosa were found to have an enzymatic activity that converts radioactive glutamate to pyrroline-5-carboxylate (P5C) in the presence of ATP, NADPH, and MgCl2. The product of this enzyme was identified as P5C by the fact that it was converted to proline by chemical reduction with NaBH4 or by enzymatic reduction with NADH in the presence of purified yeast P5C reductase. The product was demonstrated to be P5C rather than pyrroline-2-carboxylate by thin layer chromatography. The presence of the activity in mitochondria prepared from intestinal mucosa of germ-free rats proved that this activity is of mammalian origin. Omission of either ATP, NADPH, or MgCl2 from the reaction mixture resulted in little or no activity. The optimal pH appeared to be about 7.0 under the conditions used. Substrate saturation curves in the presence of an ATP and an NADPH regeneration system gave apparent Km values of 2.5 mM for glutamate, 0.19 mM for ATP, and 6.5 microM for NADPH in the presence of 20 mM MgCl2. The mitochondrial preparation usually produced P5C at a rate of 1.2 to 1.6 nmol/mg/min at 20 degrees C when incubated with 1 mM glutamate, 3 mM ATP, 0.2 mM NADPH, and 20 mM MgCl2.