J Biol Chem, Vol. 275, Issue 10, 6741-6748, March 10, 2000

N-t-Butyl Hydroxylamine, a Hydrolysis Product of -Phenyl-N-t-butyl Nitrone, Is More Potent in Delaying Senescence in Human Lung Fibroblasts^*

Hani Atamna, Andrés Paler-Martínez, and Bruce N. Ames

From the Division of Biochemistry and Molecular Biology, Department of Molecular and Cell Biology, University of California, Berkeley, California 94720-3202

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

-Phenyl-N-t-butyl nitrone (PBN), a spin trap, scavenges hydroxyl radicals, protects tissues from oxidative injury, and delays senescence of both normal human lung fibroblasts (IMR90) and senescence-accelerated mice. N-t-butyl hydroxylamine and benzaldehyde are the breakdown products of PBN. N-t-Butyl hydroxylamine delays senescence of IMR90 cells at concentrations as low as 10 µM compared with 200 µM PBN to produce a similar effect, suggesting that N-t-butyl hydroxylamine is the active form of PBN. N-Benzyl hydroxylamine and N-methyl hydroxylamine compounds unrelated to PBN were also effective in delaying senescence, suggesting the active functional group is the N-hydroxylamine. All the N-hydroxylamines tested significantly decreased the endogenous production of oxidants, as measured by the oxidation of 2',7'-dichlorodihydrofluorescin and the increase in the GSH/GSSG ratio. The acceleration of senescence induced by hydrogen peroxide is reversed by the N-hydroxylamines. DNA damage, as determined by the level of apurinic/apyrimidinic sites, also decreased significantly following treatment with N-hydroxylamines. The N-hydroxylamines appear to be effective through mitochondria; they delay age-dependent changes in mitochondria as measured by accumulation of rhodamine-123, they prevent reduction of cytochrome C^FeIII by superoxide radical, and they reverse an age-dependent decay of mitochondrial aconitase, suggesting they react with the superoxide radical.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

-Phenyl-N-t-butyl nitrone (PBN)¹ is one of the most widely used spin-trapping agents for investigating the existence of free radicals in biological systems. PBN reverses the age-related oxidative changes in the brains of old gerbils (1, 2) and delays senescence in senescence-accelerated mice (3) and in normal mice (4). PBN also delays senescence in the normal human lung fibroblast cell line IMR90 (5). In addition, PBN reverses mitochondrial decay in the liver of old rats (6) and exerts a neuroprotective effect in gerbils (1, 7) and rats (8, 9) after oxidative damage from ischemia/reperfusion injury. The mechanism underlying the biological activity of PBN is still controversial. However, PBN is a well known scavenger of radical species, although a variety of other well known spin traps or antioxidants do not mimic its anti-senescence activity in IMR90.² PBN at relatively high concentrations reduces the production of hydrogen peroxide in mitochondrial preparations of cerebral cortex (10) and therefore may exert similar properties in vivo. This suggests that PBN possesses special properties that do not exist in other spin traps or antioxidants.

In the course of our study of the effect of PBN on IMR90 cells we observed that old solutions were more effective than fresh solutions in delaying senescence of IMR90 cells. This raised the question about the interaction of the decomposition products of PBN with IMR90 cells. This encouraged us to test the anti-senescent effect of the PBN decomposition products, N-t-butyl hydroxylamine and benzaldehyde (Scheme 1) on IMR90 cells. PBN (or PBN/^·OH) has been reported to decompose to N-t-butyl hydroxylamine or N-t-butyl hydronitroxide and benzaldehyde (11-13). PBN as purchased often contains N-t-butyl hydroxylamine (14). Benzaldehyde is both mutagenic (15) and carcinogenic (16). N-t-Butyl hydroxylamine is a primary hydroxylamine that can be oxidized under certain conditions (such as with UV or Fe⁺³) to N-(t-butyl)aminoxyl (also referred as N-t-butyl hydronitroxide (10-12). N-(t-Butyl)aminoxyl and the corresponding N-hydroxylamine (Scheme 1) are primary amines and are, thus, different from the well known cyclic nitroxides/cyclic hydroxylamines (see references herein). The antioxidative and protective features of some cyclic nitroxides/cyclic hydroxylamines are known. Probably the most important feature in this regard is their ability to catalyze superoxide radical dismutation to form H₂O₂ (17-21). In vitro cyclic nitroxides can either be oxidized to oxo-ammonium cation or reduced to the corresponding hydroxylamine by superoxide radical, depending on the type of cyclic nitroxide. Thus cyclic hydroxylamine or the corresponding oxo-ammonium cation are intermediates during the dismutation of superoxide radical by nitroxide. Interestingly, the oxo-ammonium cation species is reduced to the corresponding cyclic hydroxylamine by the cellular reductant NADH, which suggests that cyclic hydroxylamine can be the dominant form inside the cells. In addition the cyclic nitroxide species can undergo one electron reduction to the corresponding cyclic hydroxylamine, a reaction proposed to be mediated by mitochondrial coenzyme Q and ascorbic acid (21-23). Mitochondrial cytochrome c oxidase can also oxidize the cyclic hydroxylamine to the corresponding nitroxide (24). Thus, it appears that mitochondria can contribute to the cycling of cyclic nitroxides/cyclic hydroxylamines, which in turn can facilitate dismutation of superoxide radical to H₂O₂. The N-t-butyl hydroxylamine and the other N-hydroxylamines tested in this study are primary N-hydroxylamines that have not been previously examined as antioxidants, although in comparison with the cyclic hydroxylamines described, one might expect similarities in their biochemical properties.

View larger version (16K): [in this window] [in a new window]   Scheme 1.   R1, N- or O- methoxyamine; R2, N- or O-t-butylhydroxylamine; R3, N- or O-benzylhydroxylamine.

We demonstrate in the present study the ability of N-t-butyl hydroxylamine to delay the senescence of IMR90 cells. The ability of N-t-butyl hydroxylamine to exert this effect at concentrations much lower than that used for PBN together with increased potency of PBN preparations with longer storage time suggests that this decomposition product mediates the action of PBN on IMR90 cells. Benzaldehyde was without effect, and in high concentrations, was toxic to the cells. Related N-hydroxylamines, N-benzyl hydroxylamine, and N-methyl hydroxylamine have also been found to be active.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- N-t-Butyl hydroxylamine, N-benzyl hydroxylamine, N-methyl hydroxylamine (and the corresponding O-hydroxylamines), nitroso-tert-butane, 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO), 4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl (4-OH-TEMPO), 3-carba-moyl2,2,5,5-tetramethylpyrrolidin-1-yloxy (3-CP), octanesulfonic acid, and methanesulfonic acid were purchased from Aldrich. Cytochrome c, Rho123 and xanthine, NADP, fluorocitrate, isocitrate dehydrogenase, and PBN were from Sigma. Xanthine oxidase was from Roche Molecular Biochemicals. Aldehyde-reactive probe was from Dojindo (Kumamoto, Japan). DCFH was from Molecular Probes (Eugene, OR). N,N'-bis(3,3'-(dimethylamino) propylamine)-3,4,9,10-perylene3-carbamoyl-tetracar3-carbamoyl-boxylic diimide (DAPER) was from Pierce. The ABC kit was from Victor labs (Burlington, CA). The DNA isolation QIAamp kit was from QIAgen (Valencia, CA).

Cultivating IMR90 Cells in Culture-- Normal human epithelial fibroblasts (IMR90) cells were obtained from the Coriell Institute for Medical Research at a population doubling level (PDL) of 10.85. The PDLs were calculated as log₂(D/D_o), where D and D_o are defined as the density of cells at the time of harvesting and seeding, respectively. Stock cultures were grown in 100-mm Corning tissue culture dishes containing 10 ml of Dulbecco's modified Eagle's medium supplemented with 10% (V/V) fetal bovine serum (Hyclone). Stock cultures were split once a week when near confluence. Cells were harvested by trypsinization for 5 min at 37 °C, immediately collected in 5 ml of complete Dulbecco's modified Eagle's medium, washed once with 5 ml of complete Dulbecco's modified Eagle's medium, and incubated for 10-15 min at 37 °C to allow the cells to recover.

To test the effect of hydroxylamines on replicative life span, IMR90 cells were seeded at 0.5 × 10⁶/100-mm dish. N-Hydroxylamines (N-t-butyl hydroxylamine, N-benzyl hydroxylamine, N-methyl hydroxylamine) were added either individually (final concentration 10 or 100 µM) or in a combination of all three N-hydroxylamines (30 µM each). The cultures were split after 7 days and seeded with fresh medium supplemented with the hydroxylamines described above. In other experiments the medium of the cultures was replaced after 3 days of seeding with fresh medium, and with fresh N-hydroxylamines, the split was done as usual after 7 days from seeding. The effect of other chemicals used in this study on the life span of the cells was tested as described above, and the effect of PBN on life span was tested as in (5).

To determine the effect of H₂O₂ on the replicative life span, cells were first seeded with fresh medium with or without N-hydroxylamines (see above) for a week. Next, cells grown with or without N-hydroxylamines were each split into two additional groups and then either 1) treated with 20 or 30 µM H₂O₂ or 2) left untreated, with H₂O₂.

Analysis of Aconitase Activity in Tissue Culture Treated with N-Hydroxylamine-- Aconitase was measured as described by Gardner et al. (25). Briefly, IMR90 cells were grown with or without N-hydroxylamine as described above. After 12 weeks of treatment, the cells were washed twice by cold PBS and scraped from the dishes with a cell scraper. The cells (3-4 × 10⁶) were collected by centrifugation and resuspended into 200 µl of ice-cold 50 mM Tris, pH 7.4, 0.6 mM MnCl₂/20 µM fluorocitrate supplemented with antiprotease mixture (leupiptin, pepstatine, and phenylmethylsulfonyl fluoride, 1 µg each). The cells were disrupted by three cycles of sonication for 3-5 s at low output separated by 1 min of incubation in ice. Then the lysate was spun at 12,000 × g for 5 min in 4 °C, and the supernatant was used to measure total soluble protein and aconitase activity. In general, 60-100 µg of protein are adequate to readily detect aconitase activity as described (25).

Analysis of the Age-dependent Changes in the Steady State Level of Oxidants and Mitochondrial Membrane Potential in IMR90 Cells by FACS-- IMR90 cells were trypsinized and resuspended into complete Dulbecco's modified Eagle's medium. For each condition, two tubes were prepared with 1 × 10⁶ cells each. Tubes were then spun at 250 × g for 10 min at room temperature, and supernatant was replaced with 1 ml of Hanks' balanced salt solution without Ca²⁺ or Mg²⁺. Rho123 (20 µl of 525 µM stock; 10.5 µM final concentration) was added to one tube, and DCFH (20 µl of 1.25 mM stock; 25 µM final concentration) was added to the other tube. The cells were then incubated in the dark in a water bath at 37 °C for 30 min followed by cell resuspension and centrifugation at 250 × g for 10 min at room temperature. The supernatant (500 µl) was removed from each tube, and the cells were resuspended in the remaining 500 µl before FACS analysis on a FACSort analyzer (Becton Dickinson, San Jose, CA). Cell Quest was used for data acquisition and analysis. The data are reported as the mean of the channel of the fluorescence histogram obtained. Fluorescence output was calibrated with LinearFlow Green Flow cytometry intensity calibration particles (Molecular Probes, Eugene, OR).

Measurement of AP Sites in IMR90 Cells-- AP sites were measured according to H. Atamna and B. N. Ames.³ Briefly, IMR90 cells (1-2 × 10⁶) in 0.5 ml PBS, 5 mM glucose were incubated with 3 mM aldehyde-reactive probe (ARP (26)) for 60 min at 37 °C. The cells were then collected by centrifugation at room temperature and washed twice with 1 ml of PBS. DNA was isolated by the QIAamp blood kit, as suggested by the manufacturer. DNA was quantified by Picogreen, and 1 µg was transferred into 200 µl of elution buffer (10 mM Tris, pH 8.9), mixed with 14 µl of 5 M NaCl (the mole ratio of NaCl/dNTP should be 25,000-30,000), and incubated for 60 min at room temperature with 30 µl of freshly prepared avidin-HRP (ABC kit), prepared as described by the manufacturer but with avidin-HRP concentrations diluted 1:3 and the incubation volumes scaled down to 1 ml. The DNA-avidin-HRP complex (DNA-HRP) was separated from unbound avidin-HRP by gently mixing 65 µl of 1 mM DAPER with the DNA and incubated at room temperature for 5 min (the mole ratio of DAPER/dNTP should be 66). The DNA-DAPER precipitate was then collected by centrifugation for 5 min at 12,500 × g and washed twice with 1.5 ml of 0.17 M NaCl, 20 mM Tris, 0.25% Tween 20, 1% bovine serum albumin, pH 8. The precipitate of DNA-HRP was suspended in 500 µl of ice-cold 50 mM sodium citrate, pH 5.3 and sonicated at output 1-2 watts for 5 s (Sonifier Cell Disruptor, model w185D, Branson) and cooled immediately. HRP activity was measured as an indicator of AP sites in DNA-HRP by using the chromogenic ImmunoPure TMB or the fluorogenic QuantaBlu substrate kits. The background control was established by performing a parallel analysis on calf thymus DNA. The standard curve for AP sites was constructed with 100 ng of DNA standard containing a known amount of uracil suspended in 50 µl of 10 mM Na₂HPO₄, pH 7.5. The standard DNA was incubated with 25 µM spermine for 3 min and then with 3 units of uracil-DNA N-glycosylase for 20 min at 37 °C to catalyze the removal of uracil residues and generate AP sites. The resulting "AP-enriched" DNA was incubated with 3 mM ARP for 45 min at 37 °C. The standard DNA-ARP adducts were isolated from unbound ARP by QIAamp columns (without the protease step) and quantified. The number of AP sites was corrected for the loss of DNA during isolation (10-20% loss). The biotinylated DNA was incubated with avidin-HRP and processed as above.

Reduction of cyt C^FeIII by Superoxide Radical-- Superoxide radical was generated by the reaction of xanthine (120 µM) with xanthine oxidase (0.06 units). The reaction was performed at 25 °C in a final volume of 1 ml of PBS containing 40 µM cyt C^FeIII. The reaction was started by the addition of the substrate xanthine. N-Hydroxylamines were added just before the addition of xanthine. The initial rate of reduction of cyt C^FeIII was determined based on the linear change in absorbance at 550 nm.

To test the effect of N-hydroxylamines on the spontaneous oxidation of cytochrome c, a complete reduction of cyt C^FeIII was achieved by incubating the xanthine/xanthine oxidase system for 4-5 min at 25 °C. Auto-oxidation of cyt C^FeII is associated with a decrease in absorbance at 550 nm. Reduced cytochrome c was incubated at 25 °C with or without 2 or 3 mM N-hydroxylamines, and the auto-oxidation was followed by spectrophotometer. The rate of reduction of cytochrome c by different concentrations of each N-hydroxylamine was measured by the increase in absorbance at 550 nm.

Measurement of Cellular Levels of GSH and GSSG in IMR90 Cells-- Cultivated IMR90 cells (3 × 10⁶) were washed once in cold PBS and resuspended in 200 µl of ice-cold methanesulfonic acid (0.2 M methanesulfonic acid, 0.5 mM DTPA) and allowed to stand for 10 min at room temperature. Denatured proteins were removed by centrifugation, and the supernatant was filtered with 30,000 M_r cut-off Ultrafree filters (Millipore) before injection. Fifty microliters were injected and separated on an HPLC column (3-µm 0.46 × 15-cm Suplecosil LC18-DB, Supelco, Bellefonte, PA) with a flow rate of 1 ml/min using a mobile phase consisting of 25 mM NaH₂PO₄, 5 mM octane sulfonic acid, and 2% acetonitrile adjusted to pH 2.7 with phosphoric acid (27). An ESA model 5100A Coulochem detector, 5020 guard cell, and model 5010 analytical cell combination was used for analysis. Oxidation potentials of 900, 400, and 880 mV were used for guard cell and electrodes 1 and 2, respectively. Full-scale output was 10 µA, and peak areas were compared using commercial GSH and GSSG as standards.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

N-t-Butyl hydroxylamines and Other N-Hydroxylamines Delay Senescence of IMR90 Cells-- N-t-Butyl hydroxylamine, N-benzyl hydroxylamine, and N-methyl hydroxylamine (N-hydroxylamines, Scheme 1) at 100 µM (added once per 7 days) delay senescence of IMR90 cells by at least 17-20 PDLs (Fig. 1). The concentration of PBN required to achieve a similar gain in PDLs is eight times higher than N-hydroxylamines (Table I and Ref. 5). The minimal concentration of N-hydroxylamines required to achieve a gain of 5-7 PDLs above the untreated control was 20 times lower than that for PBN (200 µM) to achieve 2-3 PDLs (Table I and Ref. 5). For each of the three N-hydroxylamines, when they were added to IMR90 cells every 3 days at 25 µM, they were twice as efficient as 100 µM every 7 days (data not shown). None of the N-hydroxylamines tested were toxic at the concentration applied to the cells, as measured by PDL, whereas benzaldehyde, the co-product of PBN hydrolysis, was without effect or toxic at high concentrations (data not shown). All the N-hydroxylamines at the concentrations tested were much more effective than PBN in delaying senescence. The N-hydroxylamines studied appear to be equally efficient in delaying senescence, with a variation only when cells were close to senescence (late PDLs, Fig. 1A). We have developed a new HPLC-electrochemical detection method for N-hydroxylamines. We find, using this method, that N-t-butyl hydroxylamine rapidly enters cells and reaches saturating concentrations after 1-2 min (55-65 nmol/mg of protein (12 × 10⁶ cells)), which is approximately 5-fold the extracellular concentration, 1 mM.³

View larger version (13K): [in this window] [in a new window]   Fig. 1.   A, N-hydroxylamines delay senescence of human lung fibroblasts cells (IMR90) in tissue culture. Cells were cultivated continuously with N-hydroxylamines. IMR90 cells were seeded in 100-mm dishes at an initial density of 0.5 × 106 and treated once per week with 100 µM each of N-hydroxylamine, except for the control, as described under "Experimental Procedures." The cells were harvested every week and counted, and PDLs were calculated. A portion of the cells were used to seed new dishes in fresh medium or fresh medium with N-hydroxylamines. Control (), N-t-butyl hydroxylamine (), N-benzyl hydroxylamine (), N-methyl hydroxylamine (). B, the effect of O-hydroxylamines on the life span of human lung fibroblasts cells (IMR90) in tissue culture. The cells were treated as described in A. Control (), O-t-butyl hydroxylamine (), O-benzyl hydroxylamine (), O-methyl hydroxylamine ().

A simultaneous treatment of the cells with all three of the N-hydroxylamines (30 µM each) yielded results similar to single treatments, delaying senescence by 14-17 PDLs. In contrast, at concentrations equivalent to the N-hydroxylamines, the isomeric O-hydroxylamines were either without effect (O-t-butyl hydroxylamine) or induced a small decline in the final PDL (O-benzyl hydroxylamine and O-methyl hydroxylamine) (Fig. 1B). We tested the ability of the cyclic nitroxides TEMPO, 4-OH-TEMPO, and 3-CP to delay senescence of IMR90 cells. None of these nitroxides delayed senescence at 25 µM; moreover, a decline of about 8-9 PDLs was observed at 100 µM. Also, we tested nitroso-tert-butane (tNB), a potential generator of nitric oxide (NO), for its effect on the life span of IMR90 cells. We found that tNB is toxic at 50 µM and has no effect on the cells at concentration of 10 µM.

N-t-Butyl Hydroxylamines and Other N-Hydroxylamines Delay Senescence-dependent Change in Mitochondria-- Senescencedependent change in mitochondria of IMR90 cells was estimated by Rho123. The Rho123 fluorescence that accumulated in the cells was measured weekly by FACS for a total of at least 8 weeks and plotted against the current PDL (i.e. age of the cells). The age-dependent incorporation of Rho123 in IMR90 is biphasic (Fig. 2). This is characterized by a slow and linear increase at early PDLs, followed by a shorter and steeper phase at late PDLs. Linear regression analysis was used to calculate the rate of Rho123 accumulation as a function of PDL (Table II). The regression analysis was based on early PDLs only; late PDLs were not included in the analysis (Fig. 2). The increment in accumulation of Rho123 indicates a senescence-dependent change in the mitochondria of IMR90 cells as they became senescent. N-Hydroxylamines delay these changes in mitochondria, and the rate of Rho123 accumulation as a function of senescence decreased by 70, 69, and 52% for N-t-butyl hydroxylamine, N-benzyl hydroxylamine, and N-methyl hydroxylamine, respectively (Table II).

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Accumulation of Rho123 over the life span of IMR90 cells. IMR90 cells (1 × 106 cells) were sampled for Rho123 accumulation once per week from PDL  25 until reaching senescence at PDL  55. Cells were incubated for 30 min in the dark with 10 µM Rho123, and fluorescence was measured using a FACSort. Data are expressed as the mean channel of the fluorescence histogram obtained. Data are from a representative experiment.

N-t-Butyl Hydroxylamines and Other N-Hydroxylamines Decrease Formation of Oxidants and Oxidative DNA Damage in IMR90 Cells-- The level of oxidants was measured each week by estimating the oxidation of DCFH (28) in the living cells. Measurements of fluorescence of oxidized DCFH were made weekly by FACS for total of at least 8 weeks and plotted against the current PDL (i.e. age of the cells), and a biphasic curve similar to that seen with Rho123 fluorescence was observed with DCFH fluorescence (data not shown). A linear regression analysis was used to calculate the initial linear rate of DCFH oxidation as a function of PDL (Table II). The regression analysis was based on early PDLs only. Late PDLs were not included in the analysis. IMR90 cells treated continuously with N-hydroxylamines exhibit a slower rate of formation of oxidants compared with control cells. The percent of decrease in the rate of oxidant formation from control are 88, 95, and 79% for N-t-butyl hydroxylamine, N-benzyl hydroxylamine, and N-methyl hydroxylamine, respectively (Table II). The level of AP sites in DNA can be used as a measure of the level of oxidative damage. IMR90 cells treated simultaneously with the three N-hydroxylamines (30 µM each) showed a 52% reduction in AP sites compared with PDL matched control cells (Fig. 3).

View larger version (10K): [in this window] [in a new window]   Fig. 3.   The effect of N-hydroxylamines on the level of AP sites in IMR90 cells. IMR90 cells were collected by trypsinization and counted. Cells (2 × 106) were washed in PBS and incubated in 0.5 ml of PBS, 5 mM glucose with 3 mM ARP at 37 °C for 1 h. Free ARP was separated from the cells by centrifugation, DNA was isolated, and AP sites were determined as described under "Experimental Procedures."

N-t-Butyl Hydroxylamines and the Other N-Hydroxylamines Increase the Activity of Aconitase in IMR90 Cells-- A 2-3-fold age-dependent decline in the activity of aconitase is seen in old (high PDL) compared with young (low PDL) IMR90 cells (Fig. 4). The age-dependent decline in the activity of aconitase was largely prevented when the cells were grown with N-hydroxylamines. The efficiency of protecting aconitase from inhibition was as follows; N-t-butyl hydroxylamine (90%)  N-benzyl hydroxylamines (75%) N-methyl hydroxylamine (17%, Fig. 4). This order is comparable with the relative efficiencies for the ability to delay senescence (Fig. 1A).

View larger version (19K): [in this window] [in a new window]   Fig. 4.   N-Hydroxylamines prevent the age-dependent decline in the activity of aconitase. IMR90 cells (3-4 × 106) were harvested and washed with ice-cold PBS. The cells were suspended in ice-cold 50 mM Tris, pH 7.4, 0.6 mM MnCl2, 20 µM fluorocitrate. Cells were disrupted by three short (2-3 s) cycles of sonication separated by a 1-min incubation in ice. The cellular lysate was spun at 12000 × g for 5 min, and the supernatant was used for protein quantification and the aconitase assay as described under "Experimental Procedures." NMHA, N-methyl hydroxylamine; NBHA, N-benzyl hydroxylamine; NtBHA, N-t-butyl hydroxylamine. Data are the mean ± S.D. of one representative experiment.

N-Hydroxylamines Increase the GSH/GSSG Ratio in IMR90 Cells-- The three N-hydroxylamines tested improved the glutathione status in IMR90 cells. The GSH/GSSG ratio increased by 75, 90.4, and 94% for N-t-butyl hydroxylamine, N-benzyl hydroxylamine, and N-methyl hydroxylamine, respectively (Table III). The GSH/GSSG ratio increased because of a decrease in the level of GSSG in treated cells compared with untreated cells. No change in the level of GSH was observed between the treated and control groups (Table III). When the cells were treated simultaneously with the three N-hydroxylamines a similar effect on GSH metabolism was observed (data not shown).

IMR90 Cells Treated with N-t-Butyl Hydroxylamine and N-Benzyl Hydroxylamine Are Resistant to Hydrogen Peroxide-- Hydrogen peroxide, applied at low concentrations (20 or 30 µM in fresh medium) once a week to control IMR90 cells, accelerated senescence (Fig. 5). The H₂O₂-induced senescence was attenuated when these cells were continuously treated with N-t-butyl hydroxylamine, N-benzyl hydroxylamine, or both compounds + N-methyl hydroxylamine (Fig. 5, inset).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   N-Hydroxylamines protect IMR90 cells from the toxicity of hydrogen peroxide. IMR90 cells were seeded at an initial density of 0.5 × 106/dish and grown for a week in the presence of N-hydroxylamines before the start of hydrogen peroxide treatment as described under "Experimental Procedures." Every week the cells were harvested, counted, and seeded with or without N-hydroxylamines and/or hydrogen peroxide. , control; , 20 µM H2O2; , H2O2 + N-t-butyl hydroxylamine; , H2O2 + N-benzyl hydroxylamine. Inset: , control; , 30 µM H2O2; , H2O2 + three N-hydroxylamines.

N-Hydroxylamines Inhibit Reduction of cyt C^FeIII by Superoxide Radical-- N-Hydroxylamines at relatively high concentrations (5-10 mM) were able to inhibit the reduction of cyt C^FeIII by xanthine/xanthine oxidase, a system that generates superoxide radical (Fig. 6). The catalytic activity of the enzyme xanthine oxidase was not inhibited by N-hydroxylamines, as judged from the rate of formation of uric acid (the co-product with O₂) in the presence or absence of N-hydroxylamines (data not shown). Moreover, N-hydroxylamines prevented autooxidation of cyt C^FeII (Fig. 7A). N-Hydroxylamines were able to reduce cyt C^FeIII directly to cyt C^FeII, which explains their ability to delay the oxidation of reduced cyt C^FeII (Fig. 7B). The reduction of cytochrome C by N-hydroxylamines was equally efficient under aerobic and anaerobic condition or in the presence of iron chelator (DTPA). A differential ability to reduce cytochrome c was observed for the three different N-hydroxylamines, N-t-butyl hydroxylamine being somewhat less efficient.

View larger version (12K): [in this window] [in a new window]   Fig. 6.   N-Hydroxylamines inhibit the reduction of cyt CFeIII by superoxide radical. Superoxide radical was generated by a xanthine/xanthine oxidase system, and cytochrome c reduction was followed at 550 nm. N-Hydroxylamines were added immediately before starting of the enzymatic reaction by the addition of the substrate xanthine. Inhibition of the superoxide-dependent reduction of cytochrome c was estimated by the difference in the rates of its reduction in the absence and presence of the N-hydroxylamines (5-10 mM). , N-t-butyl hydroxylamine; , N-benzyl hydroxylamine; , N-methyl hydroxylamine.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Reduction of cytochrome c by N-hydroxylamines. A, N-hydroxylamines inhibit autooxidation of cyt CFeII. Enzymatically reduced cytochrome c (50 µM) by xanthine/xanthine oxidase was incubated with 2 mM N-hydroxylamines in PBS and 0.6 mM DTPA at 25 °C. Autooxidation of cyt CFeII in the presence and absence of the three N-hydroxylamines was followed at 550 nm. , control; , N-hydroxylamines. B, rate of reduction of cytochrome c (50 µM) by different concentrations of N-hydroxylamines in PBS, 0.6 mM DTPA at 25 °C. The amount of cytochrome c reduced was calculated based on the millimolar excitation coefficient of 22.9 at 550 nm. Data from one representative experiment is shown. , N-t-butyl hydroxylamine; , N-benzyl hydroxylamine; , N-methyl hydroxylamine.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

One of the two breakdown products of PBN or PBN/^·OH hydrolysis, N-t-butyl hydroxylamine, but not the other product, benzaldehyde, delays the replicative senescence of human lung fibroblasts at concentrations at least 20 times lower than PBN. Other N-hydroxylamines tested (not related to PBN, i.e. N-benzyl hydroxylamine and N-methyl hydroxylamine, Scheme 1) were also able to delay the senescence of IMR90 cells. Thus, it appears that the N-hydroxylamine functional group is responsible for the biological activity of the three compounds tested. Although PBN is a spin trap and an antioxidant, none of the well known antioxidants studied (ascorbic acid, vitamin E, catalase, 3-CP, 4-OH-TEMPO, and TEMPO) can delay senescence of IMR90 cells as does PBN. These results suggest that the effect of PBN on IMR90 cells is due to the N-t-butyl hydroxylamine breakdown product and not PBN itself.

To gain more insight into the effect of N-hydroxylamines on cells, we assessed the status of different cellular parameters in cells that have been grown continuously with medium supplemented with N-hydroxylamines compared with controls. We show that, concomitant with delayed senescence by N-hydroxylamines, the PDL-dependent formation of oxidants was decreased as estimated by DCFH oxidation (Table II), and there was an increase in the GSH/GSSG ratio (Table III). The age-dependent decay in mitochondria was delayed as estimated by Rho123 accumulation (Table II) and by the inhibition of the age-dependent decline in the activity of aconitase (Fig. 4). The level of AP sites in DNA of cells treated with N-hydroxylamines was also 52% lower than that of the control cells. The increase in the ratio GSH/GSSG by treatment with N-hydroxylamines was due to a decrease in the steady-state level of GSSG without changing the concentration of GSH. In addition N-hydroxylamines prevented the age-dependent decline in aconitase activity in IMR90. Aconitase is an enzyme essential for the Krebs cycle and highly abundant in mitochondria compared with cytosol (29). Its iron-sulfur cluster is known to be damaged by superoxide radical and ONOO (25, 30, 31). The mitochondrial enzyme is more sensitive to inhibition by superoxide radical and oxidative modification compared with the cytosolic enzyme (30, 32). These findings provide evidence that N-hydroxylamines lower the endogenous level of oxidants in mitochondria, thus protecting aconitase and causing less GSH to be oxidized to GSSG. Since aconitase plays an important role in the Krebs cycle, changes in its activity will have a large impact on mitochondrial and cellular metabolic pathways. N-Hydroxylamines also protect IMR90 cells from H₂O₂-induced senescence, probably by acting as mitochondrial antioxidants. This is further supported by the 79-95% decrease in the rate of DCFH oxidation in N-hydroxylamine-treated cells compared with controls. A higher activity of aconitase in conjunction with low level of GSSG in N-hydroxylamine-treated cells suggests that the difference in the rate of DCFH oxidation between control and treated cells stems from differences in oxidant formation by the cells. The complexity of DCFH oxidation has been discussed (33).

PBN has been shown to protect against oxidative damage in different biological models (5, 34-36). Interestingly, PBN inhibits formation of hydrogen peroxide at the level of complex I in mitochondrial preparations, which suggests a direct interaction with mitochondria in vivo (10). The antioxidative effect of N-t-butyl hydroxylamine can be attributed to a similar, although more efficient, inhibition of superoxide formation by mitochondria in vivo, resulting in less hydrogen peroxide being formed. Further studies of the interaction of N-t-butyl hydroxylamine (as representative of the primary N-hydroxylamines used in this study) with mitochondria in IMR90 cells showed that intracellular N-t-butyl hydroxylamine is maintained in the reduced form by mitochondrial NADH and complex I.³ Since N-t-butyl hydroxylamine is stable to autooxidation in a cell-free system, this suggests that N-t-butyl hydroxylamine cycles inside the cells between the oxidized and reduced form. Complex I is a mitochondrial site that is implicated in the formation of superoxide radical. Thus a possible mechanism is the interaction of N-t-butyl hydroxylamine with this site to prevent formation of superoxide radical, as with the interaction of PBN with complex I (10).

The age-related increase in oxidative damage to mitochondrial DNA, proteins, and lipids is thought to be a major factor in organismal aging (6, 37-40). Since mitochondria are assumed to play a major role in the formation of superoxide radicals and suggested to contribute to aging, we compared the senescence-dependent changes in mitochondria in control and N-hydroxylamine-treated cells. A PDL-dependent accumulation of Rho123 is observed in IMR90 cells, which reflects a senescence-dependent change in mitochondria (Table II). Although the reason for this change is not clear, it may be due to age-dependent mitochondrial swelling or changes in the mitochondrial inner membrane that elevate the nonspecific binding of Rho123 to this membrane (37, 41). Accumulation of Rho123 was also observed in one fraction of isolated hepatocytes from livers of old rats over hepatocytes from young rats (37). When IMR90 cells were grown in medium supplemented with N-hydroxylamine, a 52-70% slower rate of the age-dependent accumulation of Rho123 was observed when compared with control cells. This suggests, in conjunction with the protective effect on aconitase, that N-hydroxylamines interact with mitochondria and delay the senescence-dependent changes to mitochondria. Since mitochondria are considered a major source for free radical formation, improving the mitochondrial status could cause a significant decrease in the level of oxidants in the cells (Table II).

We also found that cyt C^FeIII is reduced directly by N-hydroxylamines independently of oxygen or iron, indicating that superoxide radical is not an intermediate in the process. Reduction of cyt C^FeIII by N-hydroxylamines may have physiological significance and suggests that N-hydroxylamines potentially can interact in vivo also with cytochrome c in addition to mitochondrial NADH. Cyclic-N-hydroxylamines/cyclic nitroxides are recycled by mitochondrial ubiquinol and cytochrome oxidase (22, 23), a mechanism of regeneration that may be shared by the primary N-hydroxylamines used in the present study. Our primary data show that mitochondrial NADH is involved in keeping the intracellular N-hydroxylamines in reduced form.³ Although the site of intracellular oxidation of primary N-hydroxylamines is not yet known, it could be either enzyme-mediated, e.g. cytochrome c, or induced by direct interaction with cellular oxidants. Studies are under way to elucidate the exact mechanism of senescence delay by the N-hydroxylamines.

N-Hydroxylamines (5-10 mM) inhibit the reduction of cyt C^FeIII by superoxide radical, which was generated with xanthine/xanthine oxidase. N-Hydroxylamines do not inhibit the catalytic activity of xanthine oxidase since the formation of uric acid (obligatory product with superoxide radical) was not inhibited. This suggests that in vivo, primary N-hydroxylamines (or their corresponding nitroxides), react with superoxide radical, as is known for the cyclic hydroxylamines/cyclic nitroxides. We find that N-t-butyl hydroxylamine rapidly enters cells and is concentrated by approximately 5-fold.³ To test the contribution of superoxide scavenging to the mechanism of senescence delay we tested three cyclic nitroxides as typical nonmetal SOD mimics. The three cyclic nitroxides tested (3-CP, TEMPO, and 4-OH-TEMPO) that form the cyclic hydroxylamines in the cell did not delay the replicative senescence of the cells (at 25 µM) and, at higher concentrations (100 µM), were even toxic. This suggests that there are some differences in the mode of action between cyclic hydroxylamines and the primary N-hydroxylamines. Consequently we suggest that mitochondria are a potential primary target for N-hydroxylamines due to their ability to slow the senescence-dependent changes to mitochondria and lower oxidants and delay senescence of IMR90 cells. Ex-vivo and in vivo studies are currently under way to uncover the molecular details of the interaction of N-hydroxylamines with mitochondrial components. Our initial results provide further evidence that mitochondria are the primary target of N-hydroxylamines.

Nitric oxide was proposed as a product of PBN decomposition and, thus, was suggested to possess a role in the activity of PBN in vivo (11, 13). N-t-Butyl hydroxylamine has also been shown to be oxidized by UV photolysis to produce tNB, which further decomposes to give nitric oxide (11, 13). The in vivo evidence for the formation of N-t-butyl hydroxylamine-dependent (or PBN-dependent) nitric oxide has not been demonstrated, and the evidence is circumstantial or based on in vitro experiments (42, 43). To assess if tNB contributes to the effect of N-t-butyl hydroxylamine on IMR90 cells, the cells were grown in a medium supplemented with tNB. We found that tNB is toxic at 50 µM and has no effect on the cells at much lower concentrations (10 µM). Thus, it seems that tNB plays a negligible role in the mechanism underlying the biological effect of N-t-butyl hydroxylamine, although it is possible that a small fraction of tNB is formed in our system by the effect of other oxidants.

The three N-hydroxylamines used in this study possess different side chains, two alkyl groups and one benzyl group. All the three N-hydroxylamines exhibit the ability to delay senescence of IMR90 cells, and thus, the N-hydroxylamine (R-NHOH) functional group possesses the biological activity. Cyclic N-hydroxylamines (R₂NOH) and their respective nitroxides enhance the clinical recovery of damaged brains in closed-head injury (44) and protect against oxidative damage induced by H₂O₂ (45) but did not delay senescence of IMR90 cells. This emphasizes the remarkable feature of the primary N-hydroxylamines as antioxidants. Harman in 1961 (46) showed that HNHOH (hydroxylamine) possesses anticancer activity and delayed senescence in mice. On the other hand, O-hydroxylamines, which possess a different functional group (R-O-NH₂) but the same alkyl groups (and benzyl group) as N-hydroxylamines, do not affect the rate of senescence, the level of oxidants, or the changes in mitochondria in IMR90 cells. This further indicates that the N-hydroxylamine functional group (R-NHOH) is involved in the effect of delaying senescence in IMR90 cells. It is likely that the alkyl and aromatic groups of the primary N-hydroxylamines could affect their oxidation-reduction potential, as is the case with cyclic nitroxides/cyclic hydroxylamines (22). This ratio is also determined by the oxygen status of the cell (24, 47). In addition, the alkyl groups and their different hydrophobicities may determine the intracellular location of the N-hydroxylamines. We observed differences in the activity of each N-hydroxylamine toward each factor that was measured only at the late PDLs. These topics still need to be studied further.

In summary, the anti-senescence effect of PBN on IMR90 cells can be mimicked efficiently by N-t-butyl hydroxylamine, and other N-hydroxylamines, which suggests that the functional compound in the PBN preparation is the N-hydroxylamine rather than PBN itself. Other N-hydroxylamines were also effective in delaying senescence and protecting IMR90 cells. The results of this study strongly suggest that more studies should be done to assess the relative contribution of PBN and N-hydroxylamines in the protective effect of PBN on different systems (1, 3-5, 7-9). The biological activity of the N-hydroxylamines appears to be due to an antioxidant effect on mitochondria. The use of N-hydroxylamine also avoids the benzaldehyde formed when PBN decomposes (43). The low doses of N-hydroxylamine required make them desirable compounds for delaying aging and protecting from oxidative damage. This is the first time that an anti-aging activity has been attributed to a group of chemicals that share a common functional group.

	ACKNOWLEDGEMENT

We are highly grateful to Irwin Fridovich, Ronald P. Mason, M. Shigenaga, and K. Beckman for criticisms and to Ann Fischer (Tissue Culture Facility, NIEHS Center, University of California, Berkeley) and Ivana Cheung for their assistance.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants AG17140 (NIA), Outstanding Investigator Grant CA39910 (NCI), and Center Grant ES01896 (NIEHS) (to B. N. A.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence and reprint requests should be addressed: Division of Biochemistry and Molecular Biology, CHORI, 5700 Martin Luther King Jr. Way, Oakland, CA 94609. Tel.: 510-450-7625; Fax: 510-597-7128; E-mail: bnames@uclink4.berkeley.edu.

² H. Atamna, A. Paler-Martínez, and B. N. Ames, unpublished observation.

³ H. Atamna and B. N. Ames, manuscript in preparation.

	ABBREVIATIONS

The abbreviations used are: PBN, -phenyl-N-t-butyl nitrone; TEMPO, 2,2,6,6-tetramethylpiperidine-1-oxyl; 4-OH-TEMPO, 4-hydroxy-TEMPO; 3-CP, 3-carbamoyl-2,2,5,5-tetramethylpyrrolidin-1-yloxy; DTPA, diethylenetriamine pentaacetic acid; tNB, nitroso-tert-butane; AP, apurinic/apyrimidinic; IMR90, primary human lung fibroblasts; PDL, population doubling level; cyt C, cytochrome c; FACS, fluorescence-activated cell sorter; DCFH, 2',7'-dichlorofluorescin; DAPER, N,N'-bis(3,3'-(dimethylamino) propylamine)-3,4,9,10-perylene-tetracarboxylic diimide; PBS, phosphate-buffered saline; ARP, aldehyde-reactive probe; HRP, horseradish peroxidase; HPLC, high performance liquid chromatography.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES