Vanadate-stimulated NADH oxidation in plasma membrane

The rate of NADH oxidation with oxygen as the acceptor is very low in mouse liver plasma membrane and erythrocyte membrane. When vanadate is added, this rate is stimulated 10- to 20-fold. The absorption spectrum of vanadate does not change with the disappearance of NADH. The reaction is inhibited by superoxide dismutase, and there is no activity under an argon atmosphere. This indicates that oxygen is the electron acceptor and the reaction is mediated by superoxide. The vanadate stimulation is not limited to plasma membrane. Golgi apparatus and endoplasmic reticulum show similar increase in NADH oxidase activity when vanadate is added. The endomembranes have significant vanadate-stimulated activity with both NADH and NADPH. The vanadate-stimulated NADH oxidase in plasma membrane is inhibited by compounds, which inhibit NADH dehydrogenase activity: catechols, anthracycline drugs and manganese. This activity is stimulated by high phosphate and sulfate anion concentrations.     

Inhibition of transplasma membrane electron transport by transferrin-adriamycin conjugates.

Transplasma membrane electron transport from HeLa cells, measured by reduction of ferricyanide or diferric transferrin in the presence of bathophenanthroline disulfonate, is inhibited by low concentrations of adriamycin and adriamycin conjugated to diferric transferrin. Inhibition with the conjugate is observed at one-tenth the concentration required for adriamycin inhibition. The inhibitory action of the conjugate appears to be at the plasma membrane since (a) the conjugate does not transfer adriamycin to the nucleus, (b) the inhibition is observed within three minutes of addition to cells, and (c) the inhibition is observed with NADH dehydrogenase and oxidase activities of isolated plasma membranes. Cytostatic effects of the compounds on HeLa cells show the same concentration dependence as for enzyme inhibition. The adriamycin-ferric transferrin conjugate provides a more effective tool for inhibition of the plasma membrane electron transport than is given by the free drug.     

Relationship between IAA-induced elongation growth and plasma membrane NADH oxidase in soybean (Glycine max Merr.) hypocotyls

A relationship between the activity of NADH oxidase of the plasma membrane and the IAA-induced elongation growth of hypocotyl segments in etiolated soybean (Glycine max Merr.) seedlings was investigated. The plasma membrane NADH oxidase activity increased in parallel to IAA effect on elongation growth in hypocotyl segments. Actually, NADH oxidase activity was stimulated 3-fold by 1 u,M IAA, and the elongation rate of segments was stimulated 10-fold by 10 iM IAA. The short-term elongation growth kinetics, however, showed that the IAA-induced elongation of hypocotyl segments was completely inhibited by plasma membrane redox inhibitors such as actinomycin D and adriamycin, at 80 μM and 50 μM respectively. In addition, 1 mM actinomycin D inhibited the IAA-stimulated NADH oxidase activity by about 80%. However, adriamycin had no effect on NADH oxidase activity of plasma membrane vesicles. Based on these results, the plasma membrane redox reactions seemed to be involved in IAA-induced elongation growth of hypocotyls, and the redox component responding to IAA was suggested to be NADH oxidase.     

Purification and properties of a peptic heme peptide from cytochrome c1.

Highly purified mouse liver plasma membranes have been used to define the properties of an NADH dehydrogenase activity associated with plasma membrane. The NADH indophenol reductase activity is two-fold stimulated at 5 × 10^?8 M glucagon and the stimulation is inhibited by atebrin. Corresponding activity in endoplasmic reticulum is not stimulated by glucagon. The NADH indophenol reductase is 90% inhibited by insulin at 7 × 10^?11M and shows a return to the original activity at higher insulin concentrations. NADH dehydrogenase activity in endoplasmic reticulum is inhibited up to 50% by insulin at a similar concentration. Triiodothyronine at 10^?7M also inhibits the plasma membrane dehydrogenase whereas thyroxine has little effect. The response of this dehydrogenase to hormones suggests a role in regulation of cellular function.     

Putting together a plasma membrane NADH oxidase: A tale of three laboratories

The observation that high cellular concentrations of NADH were associated with low adenylate cyclase activity led to a search for the mechanism of the effect. Since cyclase is in the plasma membrane, we considered the membrane might have a site for NADH action, and that NADH might be oxidized at that site. A test for NADH oxidase showed very low activity, which could be increased by adding growth factors. The plasma membrane oxidase was not inhibited by inhibitors of mitochondrial NADH oxidase such as cyanide, rotenone or antimycin. Stimulation of the plasma membrane oxidase by iso-proterenol or triiodothyronine was different from lack of stimulation in endoplasmic reticulum. After 25 years of research, three components of a trans membrane NADH oxidase have been discovered. Flavoprotein NADH coenzyme Q reductases (NADH cytochrome b reductase) on the inside, coenzyme Q in the middle, and a coenzyme Q oxidase on the outside as a terminal oxidase. The external oxidase segment is a copper protein with unique properties in timekeeping, protein disulfide isomerase and endogenous NADH oxidase activity, which affords a mechanism for control of cell growth by the overall NADH oxidase and the remarkable inhibition of oxidase activity and growth of cancer cells by a wide range of anti-tumor drugs. A second trans plasma membrane electron transport system has been found in voltage dependent anion channel (VDAC), which has NADH ferricyanide reductase activity. This activity must be considered in relation to ferricyanide stimulation of growth and increased VDAC antibodies in patients with autism.     

Is the Drug-Responsive NADH Oxidase of the Cancer Cell Plasma Membrane a Molecular Target for Adriamycin?

Enhanced growth inhibition and antitumor responses to adriamycin have been observed repeatedly from several laboratories using impermeant forms of adriamycin where entry into the cell was greatly reduced or prevented. Our laboratory has described an NADH oxidase activity at the external surface of plasma membrane vesicles from tumor cells where inhibition by an antitumor sulfonylurea, N-(4-methylphenylsulfonyl)-N-(4-chlorophenyl)urea (LY181984), and by the vanilloid, capsaicin (8-methyl-N-vanillyl-6-noneamide) correlated with inhibition of growth. Here we report that the oxidation of NADH by isolated plasma membrane vesicles was inhibited, as well, by adriamycin. An external site of inhibition was indicated from studies where impermeant adriamycin conjugates were used. The EC₅₀ for inhibition of the oxidase of rat hepatoma plasma membranes by adriamycin was several orders of magnitude less than that for rat liver. Adriamycin cross-linked to diferric transferrin and other impermeant supports also was effective in inhibition of NADH oxidation by isolated plasma membrane vesicles and in inhibition of growth of cultured cells. The findings suggest the NADH oxidase of the plasma membrane as a growth-related adriamycin target at the surface of cancer cells responsive to adriamycin. Whereas DNA intercalation remains clearly one of the principal bases for the cytotoxic action of free adriamycin, this second site, possibly related to a more specific antitumor action, may be helpful in understanding the enhanced efficacy reported previously for immobilized adriamycin forms compared to free adriamycin.     

Adriamycin tolerance in human mesothelioma lines and cell surface NADH oxidase

Adriamycin tolerant human mesothelioma cell lines derived from a single tumor prior to either chemotherapy or radiation therapy and a susceptible cell line were investigated. Not only was growth resistant to low doses of adriamycin but an unusual pattern of resistance was encountered in which cells seemed to better tolerate high adriamycin doses than intermediate doses. The differential growth susceptibility of the tolerant lines compared to A549 lung carcinoma and the bimodal dose response correlated with differences in the specific activity of a plasma membrane-associated NADH oxidase (NOX). Plasma membrane fractions of high purity were isolated by aqueous two-phase partition and assayed directly. The NADH oxidase activity of the plasma membranes for the susceptible cell line was maximally inhibited by 1 microM adriamycin whereas the NADH oxidase activity of the tolerant lines was less and was maximally inhibited by 0.1 microM adriamycin with 1 and 10 microM adriamycin being less inhibitory than 0.1 microM adriamycin. The findings suggest a relationship between the growth response to adriamycin of the adriamycin tolerant mesothelioma lines and the activity of the plasma membrane-associated NADH oxidase activity of the cell surface in these cell lines.     

Diferric transferrin reductase in the plasma membrane is inhibited by adriamycin

Diferric transferrin which is often necessary for growth of cells is reduced by the transplasma membrane electron transport system of HeLa cells with release of ferrous iron outside the cell. Reduction of external diferric transferrin is reflected in oxidation of internal NADH. Adriamycin, an antitumor drug, inhibits diferric transferrin reduction by the HeLa cells and inhibits concomittant oxidation of cytosolic NADH at concentrations, 10(-8)-10(-6)M, which inhibit cell growth. Isolated liver plasma membranes have an NADH diferric transferrin reductase activity which is inhibited by similar adriamycin concentrations. We propose that inhibition of cell growth by adriamycin can be based on inhibition of transplasmalemma diferric transferrin reductase.     

A growth factor- and hormone-stimulated NADH oxidase from rat liver plasma membrane.

NADH oxidase activity (electron transfer from NADH to molecular oxygen) of plasma membranes purified from rat liver was characterized by a cyanide-insensitive rate of 1 to 5 nmol/min per mg protein. The activity was stimulated by growth factors (diferric transferrin and epidermal growth factor) and hormones (insulin and pituitary extract) 2- to 3-fold. In contrast, NADH oxidase was inhibited up to 80% by several agents known to inhibit growth or induce differentiation (retinoic acid, calcitriol, and the monosialoganglioside, GM3). The growth factor-responsive NADH oxidase of isolated plasma membranes was not inhibited by common inhibitors of oxidoreductases of endoplasmic reticulum or mitochondria. As well, NADH oxidase of the plasma membrane was stimulated by concentrations of detergents which strongly inhibited mitochondrial NADH oxidases and by lysolipids or fatty acids. Growth factor-responsive NADH oxidase, however, was inhibited greater than 90% by chloroquine and quinone analogues. Addition of coenzyme Q10 stimulated the activity and partially reversed the analogue inhibition. The pH optimum for NADH oxidase was 7.0 both in the absence and presence of growth factors. The Km for NADH was 5 microM and was increased in the presence of growth factors. The stoichiometry of the electron transfer reaction from NADH to oxygen was 2 to 1, indicating a 2 electron transfer. NADH oxidase was separated from NADH-ferricyanide reductase, also present at the plasma membrane, by ion exchange chromatography. Taken together, the evidence suggests that NADH oxidase of the plasma membrane is a unique oxidoreductase and may be important to the regulation of cell growth.     

Ceruloplasmin stimulates NADH oxidation of pig liver plasma membrane.

NADH oxidation by pig liver plasma membranes is stimulated by ceruloplasmin (CUP) reaching a maximal value at 50 U/ml of CUP. NADH oxidation activated by CUP is proportional to the amount of protein. Concanavalin A (Con A) which recognizes the glucidic residues of the CUP required for binding to the receptor inhibits the NADH oxidation in a dose-responsive manner. Both adriamycin and bathophenantroline disulfonate (BPS), previously reported as transplasma membrane electron transport inhibitors, also inhibit the CUP-stimulated NADH oxidation of pig liver plasma membranes. Our results show a clear interaction between CUP and the NADH oxidase of plasma membrane, which supports an oxidative role for CUP in its growth effect.     

NADH-ascorbate free radical and -ferricyanide reductase activities represent different levels of plasma membrane electron transport

Plasma membranes isolated from rat liver by two-phase partition exhibited dehydrogenase activities for ascorbate free radical (AFR) and ferricyanide reduction in a ratio of specific activities of 1 : 40. NADH-AFR reductase could not be solubilized by detergents from plasma membrane fractions. NADH-AFR reductase was inhibited in both clathrin-depleted membrane and membranes incubated with anti-clathrin antiserum. This activity was reconstituted in plasma membranes in proportion to the amount of clathrin-enriched supernatant added. NADH ferricyanide reductase was unaffected by both clathrin-depletion and antibody incubation and was fully solubilized by detergents. Also, wheat germ agglutinin only inhibited NADH-AFR reductase. The findings suggest that NADH-AFR reductase and NADH-ferricyanide reductase activities of plasma membrane represent different levels of the electron transport chain. The inability of the NADH-AFR reductase to survive detergent solubilization might indicate the involvement of more than one protein in the electron transport from NADH to the AFR but not to ferricyanide.     

A transmembranous NADH-dehydrogenase in human erythrocyte membranes

Evidence is presented for a transmembranous NADH-dehydrogenase in human erythrocyte plasma membrane. We suggest that this enzyme is responsible for the ferricyanide reduction by intact cells. This NADH-dehydrogenase is distinctly different from the NADH-cytochromeb ₅ reductase on the cytoplasmic side of the membrane. Pretreatment of erythrocytes with the nonpenetrating inhibitor diazobenzene sulfonate (DABS) results in a 35% loss of NADH-ferricyanide reductase activity in the isolated plasma membrane. Since NADH and ferricyanide are both impermeable, the transmembrane enzyme can only be assayed in open membrane sheets with both surfaces exposed, and not in closed vesicles. The transmembrane dehydrogenase has affinity constants of 90 µM for NADH and 125 µM for ferricyanide. It is inhibited byp-chloromercuribenzoate, bathophenanthroline sulfonate, and chlorpromazine.     

Carvedilol inhibits the exogenous NADH dehydrogenase in rat heart mitochondria

There are several reports on the oxidation of external NADH by an exogenous NADH dehydrogenase in the outer leaflet of the inner membrane of rat heart mitochondria. Until now, however, little was known about its physiological role in cellular metabolism. The present work shows that carvedilol (?1-[carbazolyl-(4)-oxy]-3-[2-methoxyphenoxyethyl)amino]-pro - panol-(2)?) is a specific inhibitor of an exogenous NADH dehydrogenase in rat heart mitochondria. Carvedilol does not affect oxygen consumption linked to the oxidation of succinate and internal NADH. It is also demonstrated that the inhibition of exogenous NADH dehydrogenase by carvedilol is accompanied by the inhibition of alkalinization of the external medium. In contrast to the addition of glutamate/malate or succinate, exogenous NADH does not generate a membrane potential in rat heart mitochondria, as observed with a TPP(+) electrode. It is also demonstrated that the oxygen consumption linked to NADH oxidation is not due to permeabilized mitochondria, but to actual oxidase activity in the inner membrane. The enzyme has a K(m) for NADH of 13 microM. Carvedilol is a noncompetitive inhibitor of this external NADH dehydrogenase with a K(i) of 15 microM. Carvedilol is the first inhibitor described to this organospecific enzyme. Since this enzyme was demonstrated to play a key role in the cardiotoxicity of anticancer drugs of the anthracycline family (e.g., adriamycin), we may suggest that the administration of carvedilol to tumor patients treated with adriamycin might be of great help in the prevention of the cardioselective toxicity of this antibiotic.     

Studies of pyridine nucleotide oxidizing enzymes from human neutrophils

An NADH cytochrome c reductase has been identified in plasma membrane fractions from neutrophils in addition to the superoxide producing NADPH oxidase which has been extensively studied by other investigators. Activation of neutrophils resulted in increased enzyme activities but to different degrees; the NADH cytochrome c reductase increased 2 fold in specific activity and the NADPH oxidase 30 fold. Treatment of the plasma membrane fraction with sonication and differential centrifugation yielded a particulate fraction (R2) with a 2 fold increase in specific activities of both enzymes and concentrations of cytochrome b and FAD. The cytochrome b in the preparation was not reduced under anaerobic conditions by either NADH or NADPH. Treatment of preparations of R2 with deoxycholate or potassium thiocyanate separated the two enzymes yielding particulate preparations with only NADPH oxidase or NADH cytochrome c reductase activity, respectively.     

Intraparticulate localization and some properties of a clofibrate-induced peroxisomal aldehyde dehydrogenase from rat liver

A study was made of the effect of chronic administration of the hypolipidemic drug clofibrate on the activity and intracellular localization of rat liver aldehyde dehydrogenase. The enzyme was assayed using several aliphatic and aromatic aldehydes. Clofibrate treatment caused a 1.5 to 2.3-fold increase in the liver specific aldehyde dehydrogenase activity. The induced enzyme has a high Km for acetaldehyde and was found to be located in peroxisomes and microsomes. Clofibrate did not alter the enzyme activity in the cytoplasmic fraction. The total peroxisomal aldehyde dehydrogenase activity increased 3 to 4-fold under the action of clofibrate. Disruption of the purified peroxisomes by the hypotonic treatment or in the alkaline conditions resulted in the release of catalase from the broken organelles, while aldehyde dehydrogenase as well as nucleoid-bound urate oxidase and the peroxisomal membrane marker NADH:cytochrome c reductase remained in the peroxisomal 'ghosts'. At the same time, treatment by Triton X-100 led to solubilization of the membrane-bound NADH:cytochrome c reductase and aldehyde dehydrogenase from intact peroxisomes and their 'ghosts'. These results indicate that aldehyde dehydrogenase is located in the peroxisomal membrane. The peroxisomal aldehyde dehydrogenase is active with different aliphatic and aromatic aldehydes, except for formaldehyde and glyceraldehyde. The enzyme Km values lie in the millimolar range for acetaldehyde, propionaldehyde, benzaldehyde and phenylacetaldehyde and in the micromolar range for nonanal. Both NAD and NADP serve as coenzymes for the enzyme. Aldehyde dehydrogenase was inhibited by disulfiram, N-ethylmaleimide and 5,5'-dithiobis(2-nitrobenzoic)acid. According to its basic kinetic properties peroxisomal aldehyde dehydrogenase seems to be similar to a clofibrate-induced microsomal enzyme. The functional role of both enzymes in the liver cells is discussed.     

The NADH dehydrogenase of the respiratory chain of Escherichia coli. II. Kinetics of the purified enzyme and the effects of antibodies elicited against it on membrane-bound and free enzyme.

The purified respiratory chain NADH dehydrogenase of Escherichia coli oxidizes NADH with either dichlorophenolindophenol (DCIP). ferricyanide, or menadione as electron acceptors, with values for NADH are similar with the three electron acceptors (approximately 50 muM). The purified enzyme contains no flavin and has an absolute requirement for FAD, with Km values around 4 muM. The pH optimum of the enzyme appears to be between 6.5 and 7; the optimum is difficult to establish because of nonenzymatic reduction of DCIP at the lower pH values. Potassium cyanide stimulates the DCIP reductase activity about 2-fold, but has no effect on ferricyanide reductase. The enzyme exhibits hyperbolic kinetics with respect to NADH concentration in both the ferricyanide and DCIP reductase assays, but cooperatively is seen in the menadione reductase reaction. NAD+ is an effective competitive inhibitor of the reaction (Ki congruent to 20 muM); in the presence of NAD+, the NADH saturation curve becomes cooperative, even in the DCIP reductase assay. Many adenine containing nucleotides are competitive inhibitors of the enzyme. The apparent Ki values for these nucleotides as inhibitors of the purified enzyme, the membrane-bound NADH dehydrogenase, and the NADH oxidase are equivalent. An examination of inhibitory effects of a series of adenine nucleotides suggests that the inhibitors act as analogues of NAD+, which is the true physiological inhibitor. The results suggest that the enzyme in situ is always partially inhibited by the levels of NAD- in the E coli cell, and thus behaves in a cooperative fashion to changes in the NAD+/NADH ratio. An antibody has been elicited against the purified NADH dehydrogenase. Immunodiffusion and crossed immunoelectrophoresis show that the antibody is directed principally against the NADH dehydrogenase, with some activity against minor contaminants in the purified preparation. The antibody inhibits NADH dehydrogenase activity 50% at saturating levels. When this antibody preparation is used to examine solubilized membrane preparations, two major immunoprecipitates are found. A parallel inhibition of the membrane-bound NADH dehydrogenase and NADH oxidase activities is seen, supporting the hypothesis that the purified enzyme is indeed a component of the respiratory chain-dependent NADH oxidase pathway.     

NADH oxidase of liver plasma membrane stimulated by diferric transferrin and neoplastic transformation induced by the carcinogen 2-acetylaminofluorene

NADH oxidase of purified plasma membranes (electron transfer from NADH to oxygen) was stimulated by the growth factor diferric transferrin. This stimulation was of an activity not inhibited by cyanide and was not seen in plasma membranes prepared from hyperplastic nodules from liver of animals fed the hepatocarcinogen, 2-acetylaminofluorene, nor was it due to reduction of iron associated with diferric transferrin. With plasma membranes from nodules, the activity was already elevated and the added transferrin was without effect. The stimulation by diferric transferrin did not correlate with the absence of transferrin receptors which were increased at the nodule plasma membranes. With liver plasma membranes, the stimulation by diferric transferrin raised the plasma membrane NADH oxidase specific activity to approximately that of the nodule plasma membranes. In contrast to NADH oxidase, which was markedly stimulated by the diferric transferrin, NADH ferricyanide oxidoreductase or reduction of ferric ammonium citrate by liver plasma membranes was approximately equal to or slightly greater than that of the nodule plasma membrane and unaffected by diferric transferrin. The results suggest the possibility of coupling of NADH oxidase activity to a growth factor response in mammalian cells as observed previously for this enzyme in another system.     

Enzymologic study of the structural organization of the matrix or rat liver peroxisomes]

The effect of ionic strength and pH on the release of some enzymes of the matrix of peroxisomes in rat's liver was studied. Catalase, L ALpha-hydroxy acid oxidase, isocitrate dehydrogenase, glycerophosphate dehydrogenase and lactate dehydrogenase were easily released from the particles during their lysis and treatment with 0.16 M KCl, whereas urate oxidase, NADH cytochrome c reductase and D-amino acid oxidase were not solubilized. After the solubilization of peroxisomal membrane by 0.2% Triton X-100, the remaining core contained about 50% amino acid oxidase activity, and had 1.28--1.30 g/cm3 density. These results suggest that D-amino acid oxidase associates with urate oxidase in the peroxisomal core.     

The NADH oxidase activity of the plasma membrane of synaptosomes is a major source of superoxide anion and is inhibited by peroxynitrite

Plasma membrane vesicles from adult rat brain synaptosomes (PMV) have an ascorbate-dependent NADH oxidase activity of 35-40 nmol/min/(mg protein) at saturation by NADH. NADPH is a much less efficient substrate of this oxidase activity, with a Vmax 10-fold lower than that measured for NADH. Ascorbate-dependent NADH oxidase activity accounts for more than 90% of the total NADH oxidase activity of PMV and, in the absence of NADH and in the presence of 1 mm ascorbate, PMV produce ascorbate free radical (AFR) at a rate of 4.0 +/- 0.5 nmol AFR/min/(mg protein). NADH-dependent *O2- production by PMV occurs with a rate of 35 +/- 3 nmol/min/(mg protein), and is a coreaction product of the NADH oxidase activity, because: (i) it is inhibited by more than 90% by addition of ascorbate oxidase, (ii) it is inhibited by 1 micro g/mL wheat germ agglutinin (a potent inhibitor of the plasma membrane AFR reductase activity), and (iii) the KM(NADH) of the plasma membrane NADH oxidase activity and of NADH-dependent *O2- production are identical. Treatment of PMV with repetitive micromolar ONOO- pulses produced almost complete inhibition of the ascorbate-dependent NADH oxidase and *O2- production, and at 50% inhibition addition of coenzyme Q10 almost completely reverts this inhibition. Cytochrome c stimulated 2.5-fold the plasma membrane NADH oxidase, and pretreatment of PMV with repetitive 10 microm ONOO- pulses lowers the K0.5 for cytochrome c stimulation from 6 +/- 1 (control) to 1.5 +/- 0.5 microm. Thus, the ascorbate-dependent plasma membrane NADH oxidase activity can act as a source of neuronal.O2-, which is up-regulated by cytosolic cytochrome c and down-regulated under chronic oxidative stress conditions producing ONOO-.     

Molecular biology of plasma membrane redox enzymes: A survey of current knowledge

Summary Despite a large body of evidence for enzymatic activities and physiological functions of plasma membrane redox function, few of these enzymes have been characterized in terms of molecular biology. Examples for these with at least some molecular data up to complete sequences, membrane topology and binding sites for substrates and coenzymes or prosthetic groups are NADH-ferricyanide reductase of Ehrlich ascites membranes, NADH-coenzyme Q reductase of liver, NADH oxidase ectoenzyme of liver and HeLa (and possibly other) cells, protein disulfide isomerase which is widespread, and relatives thereof, as well as cytochromes P-450 andb ₅₅₈, NADPH oxidase of fat and thyroid cells and fat cell amine oxidase. Ferricyanide reductase and coenzyme O reductase may be identical, but NADH oxidase ectoenzyme is distinct and possibly functions also as a disulfide and a copper reductase. On the other hand, the plasma-membrane-located protein disulfide isomerase (PDI), despite its similar enzymatic activity, is completely different from the ectooxidase. The latter is shed from the membrane into the surrounding medium by proteolysis, whereas PDI is not an integral membrane protein and is secreted intact. Another disulfide reductase has been demonstrated in THP-1 cells, which again is totally different from the former two. It turns out that enzymatic activities are insufficient to describe redox enzymes. Special forms of cytochrome P-450 can be induced to expression at the cell membrane of liver, where they are transported by the cytoskeleton-associated secretory pathway. Why some isoforms are expressed at the surface and some are not is not yet clear. Cytochromeb ₅₅₈, the flavocytochrome of neutrophils, is described in other cells too, but there are different isoforms, which are genetically distinct. A relative has also been identified in duodenal cells, where it functions as a ferric reductase involved in iron absorption. NADPH oxidase of fat cells has very similar properties, but the identity is unproved, whereas thyroid oxidase is a non-heme protein which is calcium-sensitive and does not need assembly of subunits for activation. Finally, fat cell membranes also possess a quinone-containing amine-oxidase which may be involved in signaling of glucose-transport regulation, as it is also found in GLUT4-containing vesicles. However, the physiological connection has yet to be demonstrated.