Purification and Characterization of a Doxorubicin-inhibited NADH-quinone (NADH-ferricyanide) Reductase from Rat Liver Plasma Membranes*

Plasma membrane-associated redox systems play important roles in regulation of cell growth, internal pH, signal transduction, apoptosis, and defense against pathogens. Stimulation of cell growth and stimulation of the redox system of plasma membranes are correlated. When cell growth is inhibited by antitumor agents such as doxorubicin, capsaicin, and antitumor sulfonylureas, redox activities of the plasma membrane also are inhibited. A doxorubicin-inhibited NADH-quinone reductase was characterized and purified from plasma membranes of rat liver. First, an NADH-cytochromeb 5 reductase, which was doxorubicin-insensitive, was removed from the plasma membranes by the lysosomal protease, cathepsin D. After removal of the NADH-cytochromeb 5 reductase, the plasma membranes retained a doxorubicin-inhibited NADH-quinone reductase activity. The enzyme, with an apparent molecular mass of 57 kDa, was purified 200-fold over the cathepsin D-treated plasma membranes. The purified enzyme had also an NADH-coenzyme Q0 reductase (NADH: external acceptor (quinone) reductase; EC 1.6.5..) activity. Partial amino acid sequence of the enzyme showed that it was unique with no sequence homology to any known protein. Antibody against the enzyme (peptide sequence) was produced and affinity-purified. The purified antibody immunoprecipitated both the NADH-ferricyanide reductase activity and NADH-coenzyme Q0 reductase activity of plasma membranes and cross-reacted with human chronic myelogenous leukemia K562 cells and doxorubicin-resistant human chronic myelogenous leukemia K562R cells. Localization by fluorescence microscopy showed that the reaction was with the external surface of the plasma membranes. The doxorubicin-inhibited NADH-quinone reductase may provide a target for the anthracycline antitumor agents and a candidate ferricyanide reductase for plasma membrane electron transport.

Membrane-bound oxidoreductases and enzymatic electron transfer chains are common to all living organisms (1,2). In eukaryotic cells, chemically and functionally different electron transport systems are localized in different subcellular membranes. Some of the electron carriers may be loosely bound to the membrane or even soluble in the cytoplasm; others are structured as integral proteins of the membrane (1). The mitochondrial inner membrane and outer membranes, endoplasmic reticulum, Golgi apparatus, plasma membrane, and membranes derived from these structures all contain redox systems.
Evidence for a growth-related NADH-ferricyanide (FeCN 1 ; Fe(CN) 6 Ϫ3 ) reductase traces its origin to the work of Ellem and Kay (3). When melanoma cells were cultured at low concentrations of fetal calf serum, cell growth was Ͻ20% of the optimum growth. If low concentrations of impermeable ferricyanide were added (0.01-0.1 mM), the growth was increased to 80% of the maximum rate with optimum fetal calf serum (3). HeLa cells also were stimulated to proliferate by micromolar concentrations of ferricyanide in serum free-media (2). Indeed, there appears to be a direct connection between increases in transplasma membrane electron transport activity and cell growth (1). Stimulation of growth by external oxidants is not limited to iron compounds because HeLa cell growth also is stimulated by hexamine ruthenium III (4) and by indigotetrasulfonate (5).
Inhibitors of the NADH-ferricyanide reductase and NADH oxidase of plasma membranes have been reported. The most interesting inhibitors from the view of both selectivity for tumor cells and significance are certain antitumor drugs such as anthracyclines (6,7), cis-platinum (8), bleomycin (9), capsaicin (10), and antitumor sulfonylureas (11).
The growth response of several cell types in culture to both external impermeable oxidants and antitumor drugs indicates that the electron transport system of the transplasma membrane transcends iron uptake for nutritive purposes. A broader role of plasma membrane electron transport related to mechanisms of cell growth and possibly cell differentiation is indicated. In this study, we report isolation and characterization of NADH-ferricyanide reductase from rat liver plasma membrane, which is inhibited by the antitumor agent doxorubicin.

Preparation of Rat Liver Plasma Membranes by Aqueous Two-phase Partition
Rat liver homogenates were prepared as described by Bruno et al. (12). Homogenates then were resuspended in 1 mM sodium bicarbonate solution and fractionated by using a 16-g aqueous two-phase system (13,14). The contents of the two-phase system were mixed thoroughly and separated by centrifugation. Plasma membranes partitioned into the upper phase were collected and stored at Ϫ70°C until used.

Enzyme Assays
For assay of NADH-ferricyanide reductase activity, the assay medium contained 50 mM Tris-HCl, pH 7.2, 200 M NADH, 200 M K 3 Fe(CN) 6 , and 0.02 to 0.1 mg of protein and/or inhibitors as indicated in a final volume of 2.5 ml. Ferricyanide reduction was monitored at 420 nm with a reference beam at 600 nm using a SLM Aminco DW 2000 dual wavelength spectrophotometer or by the decrease in the absorbance at 420 nm using Hitachi U3210 spectrophotometers. The assay medium containing all the reagents except potassium ferricyanide was preincubated for 5 min at 37°C, and the assay was started by the addition of potassium ferricyanide and continued for two consecutive 5-min periods. Decrease in absorbance during the second 5-min period was used to calculate the specific activity of ferricyanide reduction. A blank rate was determined in the absence of proteins and subtracted. The extinction coefficient for ferricyanide reduction was 1.0 mM Ϫ1 cm Ϫ1 . For assay of NADH-quinone reductase activity, enzyme activity was measured as described above, except that 200 M coenzyme Q 0 was added in place of potassium ferricyanide, and the decrease in absorbance at 410 nm was measured. The extinction coefficient for coenzyme Q 0 was 0.80 mM Ϫ1 cm Ϫ1 .

Treatment of Plasma Membranes with Cathepsin D
Rat liver plasma membranes were resuspended in 15 ml of 100 mM Tris-HCl, pH 5.0, containing 0.25 unit of cathepsin D and incubated at 37°C for 4 h with occasional agitation. After incubation, the pH was adjusted to 7.0 with 1 N HCl, and the preparations were centrifuged at 100,000 ϫ g for 1 h to collect the membranes. The membrane pellet was resuspended in 20 ml of 100 mM Tris-HCl, pH 7.0.

Solubilization of Proteins from Plasma Membranes
Peripheral proteins and loosely bound cytosolic proteins were first removed by incubating plasma membranes (ϳ76 mg of protein) in a medium containing 50 mM Tris-HCl, pH 7.2, and 1 mM EDTA with gentle stirring at 4°C, followed by centrifugation for 60 min at 100,000 ϫ g. After centrifugation, pellet was collected and treated with cathepsin D. The cathepsin D-treated plasma membranes were resuspended and solubilized in 20 ml of Tris-HCl, pH 7.0, containing 1 mM ␣-phenylmethylsulfonyl fluoride, 1 mM EDTA, 0.5% Triton X-100, and 10% glycerol. The plasma membranes were incubated at 4°C for 4 h with stirring and then centrifuged at 100,000 ϫ g for 1 h. The supernatant was collected.

Purification of NADH-quinone Reductase
NADH-quinone reductase was purified from rat liver plasma membranes by hydroxyapatite, anion-exchange, and gel filtration chromatographies, as follows.
Purification by Hydroxyapatite Chromatography-Triton X-100-solubilized proteins of cathepsin D-treated rat liver plasma membranes were fractionated by hydroxyapatite HPLC. Samples were loaded onto a hydroxyapatite column (TSK Gel HA-1000). All purifications including subsequent steps were performed at room temperature by using a Waters HPLC system. The column was prewashed with elution buffer (0.5 M NaH 2 PO 4 , pH 6.8, 10% glycerol, and 0.3% CHAPS) and preequilibrated with equilibration buffer (10 mM NaH 2 PO 4 , pH 6.8, 10% glycerol, and 0.3% CHAPS). After loading of the sample, proteins were eluted with a linear gradient of equilibration buffer to elution buffer in 60 min at a flow rate of 0.7 ml/min with monitoring absorbance at 280 nm. NADH-ferricyanide reductase activity was assayed in each fraction, and fractions with enzyme activity were pooled and concentrated for further purification.
Purification by Anion-exchange Chromatography-The pool from hydroxyapatite chromatography was further purified by anion-exchange HPLC. The pool was loaded onto an anion-exchange column (LiChrospher 1000 DEAE). The column was prewashed with elution buffer (50 mM Tris-HCl, pH 7.5, 0.3% CHAPS, and 1.0 M NaCl) and pre-equilibrated with equilibration buffer (50 mM Tris-HCl, pH 7.5, and 0.3% CHAPS). Proteins were eluted with a linear gradient of equilibration buffer to elution buffer in 100 min at a flow rate of 0.5 ml/min. Fractions with the highest NADH-ferricyanide reductase activity were concentrated and further purified.
Purification by Gel Filtration Chromatography-The pool from anion-exchange chromatography was purified by gel filtration HPLC by using TSK G3000SWXL column. Proteins were eluted with 20 mM Tris-HCl, pH 7.0, containing 100 mM NaCl and 0.3% CHAPS at a flow rate of 0.5 ml/min. Fractions were collected and assayed for NADHferricyanide reductase activity.

Determination and Analysis of Partial Amino Acid Sequence of NADH-quinone Reductase
Partial amino acid sequence of NADH-quinone reductase was determined from a peptide generated by cyanogen bromide cleavage. About 7 g of the purified NADH-quinone reductase were cleaved by cyanogen bromide, and peptides were separated by SDS-PAGE as described by Shagger and Jagow (15). After SDS-PAGE, peptides were transferred to polyvinylidene difluoride membrane. One of the cleaved peptides was subjected to N-terminal amino acid sequencing. The sequencing was carried out using an automated pulsed-liquid protein sequencer, courtesy of Dr. Gerald W. Becker (Eli Lilly Research Laboratories, Indianapolis, IN). The sequence obtained was compared with known protein sequences using the BLAST program of the NCBI/GenBank TM data base.

Preparation of Anti-NADH-quinone Reductase Antibody
A synthetic peptide (CMVADKANIDK) derived from the amino acid sequence of NADH-quinone reductase was used to generate rabbit anti-NADH-quinone reductase antibodies. Synthesis and analysis of the peptide and production of the peptide antibody were performed by Immunodynamics, Inc. (La Jolla, CA). Briefly, the synthesized peptide was conjugated to carrier proteins, keyhole limpet hemocyanin and bovine serum albumin. Antibodies were made by immunizing two rabbits and maintaining them for a period of 10 weeks. Rabbits were boosted at t ϭ 3 weeks and t ϭ 6 weeks. Bleeds were taken at t ϭ 5 weeks, t ϭ 7 weeks, and t ϭ 9 weeks and monitored for specificity and relative antibody titer by enzyme-linked immunosorbent assay. Bleeds were pooled, and antibody was affinity-purified. For affinity purification, the antibody was precipitated with 33% ammonium sulfate, resuspended, dialyzed against 20 mM Tris-HCl, pH 8.0, containing 30 mM NaCl and 0.02% NaN 3 , and passed through the DEAE-Affi-Gel blue column. The antibody, unbound to the column at 30 mM NaCl, was collected and further purified by the peptide-linked affinity resin, which was prepared by linking peptide to the activated affinity resin (Sulfolink) according to the procedure provided by the manufacturer. Antibody was applied to the affinity column, and anti-NADH-quinone reductase antibody was eluted with 100 mM glycine, pH 2.8, and neutralized by 1 M Tris-HCl, pH 8.0.

Immunoprecipitation of NADH-quinone Reductase
Rat liver plasma membranes (2.5 mg of protein) were solubilized in 400 l of lysis buffer (10 mM Tris, 10 mM sodium phosphate, 140 mM NaCl, 1 mM dithiothreitol, 1 mM ␣-phenylmethylsulfonyl fluoride, and 0.5% Nonidet P-40, pH 8.0) at 4°C for 16 h with shaking and centrifuged at 100,000 ϫ g for 30 min, and supernatant was collected. Supernatant (150 l containing 550 g of protein) was diluted to 1.2 ml with Tris-phosphate-buffered saline (10 mM Tris, 10 mM sodium phosphate, and 140 mM NaCl, pH 8.0) and precleared with 100 l of protein A-Sepharose (70 mg resin/ml) by incubation for 4 h at 4°C. Three aliquots of precleared solution (300 l; 120 g of protein) were prepared. Two of these aliquots were incubated with the affinity-purified anti-NADH-quinone reductase antibody (20 l) for 4 h at 4°C; the other aliquot was incubated with 20 l of 10 mM Tris-HCl buffer, pH 8.0, or preimmune IgG as controls. Protein A-Sepharose (200 l) was added and incubated for an additional 2 h at 4°C. The solution was centri-fuged, and supernatant was collected for the measurement of enzyme activities.

SDS-PAGE and Western Blot Analysis
SDS-PAGE was according to the method of Laemmli (16), and protein bands were stained with silver as described by Oakley et al. (17). For Western blot analysis, proteins were transferred to a nitrocellulose membrane after SDS-PAGE, and the membrane was blocked with 3% bovine serum albumin. The membrane was incubated with anti-NADHquinone reductase peptide antibody for 2 h at 4°C and then incubated with goat alkaline phosphatase-conjugated anti-rabbit IgG antibody. The blot was developed using nitro blue tetrazolium and 5-bromo-4chloro-3-indolyl phosphate.

Cell Culture
Human chronic myelogenous leukemia K562 cells and doxorubicinresistant K562R cells were grown as described by Barabas and Faulk (18). Cell viability was measured by trypan blue exclusion and was Ͼ95%.

Localization of NADH-quinone Reductase by Fluorescence Microscopy
K562R cells (1 ϫ 10 7 ) were washed twice with washing solution and resuspended in 0.5 ml of washing solution. Cells were incubated with affinity-purified anti-NADH-quinone reductase antibody or preimmune IgG for 30 min at 4°C. Cells were washed twice and incubated with rhodamine-labeled goat F(abЈ) 2 antibody to rabbit IgG for 30 min at 4°C. Cells were washed twice and resuspended in washing solution containing 80% glycerol. Cells were transferred to slides, and overlaid with coverslips. Localization was determined with a Leitz microscope fitted with an HBO-100 mercury arc lamp, interference optics, and a Ploem epi-illuminator containing filters appropriate for fluorescein and rhodamine detection (19).

Isolation of Plasma Membranes by Aqueous Two-phase
Partition-Intact cells have been shown to reduce the externally added impermeable electron acceptor ferricyanide (20,21). The observation that whole cells can reduce impermeable ferricyanide implies that there may be a transmembraneous component of the NADH dehydrogenase. The presence of NADHferricyanide reductase activity in plasma membranes has been demonstrated (22), and its activity has been reported to be inhibited by the antitumor agent doxorubicin (7).
The total NADH-ferricyanide reductase activity of rat liver plasma membranes prepared by aqueous two-phase partition was largely not inhibited by doxorubicin and was even stimulated at high concentrations ( Fig. 1, E). Because plasma membranes contain more than one redox enzyme, the doxorubicininhibited activity may have been masked by the NADHcytochrome b 5 reductase that is not inhibited by doxorubicin. To test this possibility, the NADH-cytochrome b 5 reductase was removed from the membranes without loss of enzymatic activity by cathepsin D (23,24). When the plasma membranes were treated with cathepsin D, the enzyme activity was reduced to 7% of the initial rate. The activity that remained, however, was strongly inhibited by doxorubicin, with maximum inhibition at 1 M doxorubicin (Fig. 1, q). A 10-fold higher concentration was less inhibitory (Fig. 1, E). The cathepsin D-treated plasma membranes were used as the starting material for purification and characterization of the doxorubicininhibited NADH-quinone reductase.
Purification of NADH-quinone Reductase-The major activity of NADH-ferricyanide reductase was eluted at concentrations between 0.14 and 0.18 M sodium phosphate upon hydroxyapatite chromatography (fractions 6 -9 in Fig. 2A). Fractions 6 -9 were pooled, concentrated, and purified by anion-exchange chromatography. NADH-ferricyanide reductase was eluted at concentrations between 0.22 and 0.28 M NaCl (Fig. 2B). Fractions 3-5 had enzyme activity, and fraction 4 had the highest enzyme activity. Fraction 4 was concentrated, applied to a gel filtration column, and eluted. Enzymatic activity was concentrated in fractions 2-4 (Fig. 2C).
The protein composition and purity of fractions 2-4 of gel filtration chromatography were analyzed by SDS-PAGE. Equal amounts were subjected to SDS-PAGE, and band intensity was compared with enzyme activity. A 57-kDa protein band (arrows in Fig. 3A) correlated with NADH-ferricyanide reductase activity (Fig. 3B). Although minor protein bands were visible, no correlation between band intensity of the minor bands and enzyme activity was noted.
The purified enzyme also reduced CoQ 0 (NADH-CoQ 0 reductase) (NADH: external acceptor (quinone) reductase; EC 1.6.5..), indicating that the natural substrate for the enzyme may be coenzyme Q (ubiquinone). NADH-ferricyanide reductase was purified 198-fold with a yield of 5% over that of cathepsin D-treated plasma membranes (Table I). Purification of NADH-CoQ 0 reductase was 186-fold, and the yield was 5% (Table II).
Fractions 3 and 4 from gel filtration chromatography were pooled, concentrated, cleaved by cyanogen bromide, and electrotransferred to polyvinylidene difluoride membrane. They yielded a peptide band with the N-terminal sequence: Val-X-Asp-Lys-Ala-Asn-Ile-Asp/Ser/Glu-Lys-Gln/Glu-Asp/Thr/Glu-X-Asp/Gln-X-Val. Peptide sequence homology search (December 2001) showed no sequence homology with known proteins, but a segment of the peptide sequence had homology with aldehyde dehydrogenase, pyruvate:ferredoxin oxidoreductase, and trans-2-enoyl-acyl carrier protein reductase II (Fig. 4A). Western Blot Analysis and Immunoprecipitation of NADHquinone Reductase-Antibody was generated by using peptide (CMVADKANIDK) deduced from the partial amino acid sequence, and the antibody was affinity-purified. Cysteine was added to the N terminus for the purpose of conjugating the peptide to activated affinity resin. Because cyanogen bromide cleaves after methionine at the N terminus, methionine was added to the deduced amino acid sequence. Alanine replaced the unknown amino acids because alanine has low antigenicity. For Western blot analysis, rat liver plasma membranes were electrophoresed, and proteins were transferred to a nitrocellulose membrane. Proteins were reacted with affinity-purified antibody and detected with alkaline phosphatase-conjugated anti-rabbit antibody using nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate as substrates. Affinity-purified antibody reacted with a 57-kDa protein (arrow in Fig. 4B).
After immunoprecipitation of NADH-quinone reductase from cathepsin D-treated and solubilized plasma membranes, sample was centrifuged, and the supernatant was collected for the measurement of reductase activity. Both the NADH-ferricyanide reductase and NADH-CoQ 0 reductase activities were depleted from the supernatant, although the depletion was incomplete (Table III). Both the lack of complete depletion by immunoprecipitation and the lack of complete inhibition by doxorubicin may be due to residual NADH-cytochrome b 5 reductase remaining in the cathepsin D-treated membranes.
Localization of NADH-quinone Reductase-Localization of NADH-quinone reductase was studied by fluorescence microscopy. K562R cells were incubated with affinity-purified antibody, followed by incubation with rhodamine-labeled anti-rabbit IgG. Detection was by fluorescence microscopy. The antibody reacted with plasma membranes of doxorubicin-resistant human chronic myelogenous leukemia (K562R) cells, and the reaction was localized to the external surface of the plasma membrane as patches (Fig. 5A), which indicates that the purified NADH-quinone reductase was from plasma membranes.
Quantitation of NADH-quinone Reductase by Flow Cytometry-K562 and K562R cells were reacted with affinity-purified antibody, followed by reaction with fluorescein isothiocyanatelabeled anti-rabbit IgG. Quantification was measured by flow cytometry. No fluorescence channel shift occurred with cells incubated with preimmune IgG (3.44 Ϯ 0.61 in K562 cells and 3.89 Ϯ in K562R cells) (Fig. 5B, a and c), but fluorescence channel shift occurred with cells incubated with affinity-purified antibody (7.09 Ϯ 1.36 in K562 cells and 24.24 Ϯ 1.16 in K562R cells) (Fig. 5B, b and d). K562 cells expressed NADHquinone reductase on the plasma membrane, and K562R cells expressed NADH-quinone reductase at a level 1 order of magnitude greater than that of K562 cells. DISCUSSION Although physiological roles of the redox system are not yet completely understood, functions in which plasma membrane redox systems have been implicated are proton extrusion and control of internal pH (25), generation of superoxide (26), reduction of ferric irons and iron uptake (27,28), control of cell growth and proliferation (5,29,30), and biological timekeeping (31,32). Comprehensive reviews of the plasma membrane redox system are found elsewhere (33)(34)(35).
In this study, we have described purification and characterization of doxorubicin-inhibited NADH-quinone (CoQ 0 ) reductase from rat liver plasma membranes. Although redox enzymes have been purified previously from plasma membranes of rat liver, pig liver, Ehrlich tumor, HeLa cells, or human erythrocytes (22, 36 -39), this is the first report of purification and characterization of a doxorubicin-inhibited NADH-quinone (CoQ 0 ) reductase from rat liver plasma membranes. The findings demonstrate it to be a unique reductase that is distinct from NADH-cytochrome b 5 reductase. Preparation of plasma membranes by two-phase partition yielded highly pure plasma membranes, which is essential to prevent contamination of endogenous membranes for characterization and purification of plasma proteins (40).
NADH-cytochrome b 5 reductase (34 kDa) has been known to be present on the cytoplasmic surface of the plasma membrane (23), and it has an NADH-ferricyanide reductase activity, but it was not inhibited doxorubicin (22). Cathepsin D has been used to remove NADH-cytochrome b 5 reductase from plasma mem-

FIG. 4. Comparison of the amino acid sequence (A) and Western blot analysis of NADH-quinone reductase (B).
A, a segment of the peptide sequence had sequence homology with aldehyde dehydrogenase, pyruvate:ferredoxin oxidoreductase, and trans-2-enoyl-ACP reductase II. B, plasma membranes were subjected to SDS-PAGE and transferred to a nitrocellulose membrane. Proteins were reacted with the affinity-purified antibody and detected with the alkaline phosphatase-conjugated anti-rabbit IgG antibody in the presence of 5-bromo-4-chloro-3-indolyl phosphate and nitro blue tetrazolium. Lane std, protein molecular mass standards; lane 1, plasma membranes of rat liver. A 57-kDa protein band reacted with the affinity-purified antibody (arrow). branes of erythrocytes as a soluble form without losing its activity (24). Selective digestion of NADH-cytochrome b 5 reductase by cathepsin D and resistance of doxorubicin-inhibited NADH-quinone reductase to cathepsin D digestion were critical to unmask and identify the doxorubicin-inhibited NADHquinone reductase of rat liver plasma membranes. After cathepsin D treatment, the remaining NADH-ferricyanide reductase activity was about 7% of the initial activity of the untreated plasma membranes, and the activity was inhibited by doxorubicin.
Anthracycline antitumor drugs inhibit cell growth by interaction with DNA or RNA. However, there are analogs of doxorubicin that do not react well with DNA, yet they inhibit cell growth. Israel et al. (41) have shown that N-trifluoroacetyl-14-valerate doxorubicin (AD32) does not intercalate with DNA but does in-hibit cell growth (41), NADH-ferricyanide reductase (42) and of plasma membranes. It has been shown that doxorubicin is cytotoxic even without entering cells and targets the plasma membrane (43). Transferrin conjugates of doxorubicin are cytotoxic without intercalating nuclear DNA (19), target primarily plasma membranes (44), and inhibit plasma membrane oxidoreductase of K562 cells (45). Sun et al. (6,7,42) reported that ferricyanide stimulated HeLa cell growth and that doxorubicin and its anthracycline derivatives inhibited cell growth as well as NADHferricyanide reductase activity.
Three chromatographic steps enabled us to purify cathepsin D-resistant and doxorubicin-inhibited NADH-quinone reductase from rat liver plasma membranes with an apparent molecular mass of 57 kDa, which differs from the 32-kDa NADH dehydrogenase purified from Ehrlich tumor cells (36), the 34-kDa coen- 5. Localization (A) and quantification of NADH-quinone reductase (B). A, K562R cells were incubated with affinity-purified antibody, followed by incubation with rhodamine-labeled goat anti-rabbit antibody. Localization was analyzed by fluorescence microscopy. The antibody reacted with the external surface of the plasma membrane as patches. B, K562 and K562R cells were incubated with affinity-purified anti-NADH-FeCN quinone reductase antibody, followed by incubation with fluorescein isothiocyanate-labeled goat anti-rabbit antibody. Quantitation was by flow cytometry. Preimmune IgG did not react with NADH-FeCN quinone reductase (a and c), but affinity-purified anti-NADH-FeCN/quinone reductase antibody reacted with NADH-FeCN/quinone reductase (b and d). Reaction with K562R cells was even greater. zyme Q reductase from pig liver plasma membranes (39), the 40-kDa oxidoreductase purified from human erythrocyte plasma membrane (37), the 33.5-kDa protein with capsaicin-inhibited NADH oxidase activity from total defined culture media conditioned by growth of HeLa cells (38), the 45-kDa protein from plasma membranes of spinach leaves (46) ,and the auxinstimulated NADH oxidase purified from soybean plasma membranes (36-, 52-, and 72-kDa complex) (47). The partial amino acid sequence of the NADH-quinone reductase revealed no sequence homology to any known proteins and appears to be unique.
The purified NADH-ferricyanide reductase also had an NADH-quinone reductase activity, which suggests that the natural substrate of the enzyme is coenzyme Q. Coenzyme Q has been found in the plasma membranes of HeLa cells (48 -50), and the redox system utilizing coenzyme Q as a substrate has been reported to be present in the plasma membrane (29,39,51,52). NADH-diferric transferrin reductase also is present in the plasma membranes (28,53), and its activity is inhibited by doxorubicin (6). Affinity-purified antibody showed that the NADH-quinone reductase was localized to the plasma membrane of both K562 and K562R cells.
When these studies are combined with the fact that four of the most commonly used anticancer drugs (doxorubicin, bleomycin, actinomycin D, and cis-platinum) can inhibit the plasma membrane redox system, it is clear that the redox system can be an important element in control of cell growth. The growth response of cells to a variety of impermeable external oxidants and anticancer drugs indicates that electron transport systems of the plasma membrane are not simply involved in iron uptake for nutritive purposes but have a large role in stimulating cell growth and possibly cell differentiation. As such, this protein may represent the long-sought ferricyanide reductase of the plasma membrane electron transport system.