Differential effects of organic hydroperoxides and hydrogen peroxide

Differential effects of organic hydroperoxides and hydrogen peroxide on proteolysis in human erythrocytes. Melissa Runge-Morris, Patricia Frank, and R...
0 downloads 0 Views 1MB Size
Chem. Res. Toxicol. 1989,2, 76-83

76

Art ic 1es Differential Effects of Organic Hydroperoxides and Hydrogen Peroxide on Proteolysis in Human Erythrocytes Melissa Runge-Morris, Patricia Frank, and Raymond

F. Novak*

Departments of Pharmacology and Molecular Biology, Northwestern University Medical and Dental Schools, Chicago, Illinois 60611, and The Institute of Chemical Toxicology, Wayne State University, Detroit, Michigan 48201 Received November 9,1988

The effects of tert-butyl hydroperoxide, cumene hydroperoxide, and hydrogen peroxide on proteolysis in human red blood cells have been examined. The organic hydroperoxides effectively stimulated the rate of protein degradation in red cells and in hemolysate; in contrast, H 2 0 2 addition was without significant effect in either system. tert-Butyl hydroperoxide or cumene hydroperoxide (8 mM) increased the rate of protein degradation in red cells 2.3- and 4-fold, respectively, relative to control as monitored by tyrosine release. In hemolysate, tert-butyl hydroperoxide and cumene hydroperoxide, present a t 8 mM, produced a 2- and 3-fold increase in the rate of protein degradation, respectively, as compared to controls. Hydroperoxide-stimulated proteolysis in red cells or in hemolysate was concentration-dependent and reached saturation at 8 mM hydroperoxide. The reaction was linear for 2 h after which a plateau was reached. In contrast to the results observed for the organic hydroperoxides, H 2 0 2(100 or 200 mM) addition either alone or in the presence of the catalase inhibitor 3-amino-1,2,4-triazole (50-200 mM), failed to stimulate proteolysis. N-Acetylcysteine (20 mM) and dimethylthiourea (50 mM) inhibited the rate of hydroperoxide-stimulated proteolysis in red cells by -50 and -35%, respectively, and in hemolysate by 25 and 4070, respectively. The hydroxyl radical scavengers methyl sulfoxide (50 mM) or dimethylfuran (50 mM), metal ion chelators, or spin traps failed to decrease significantly the rate of organic hydroperoxide stimulated proteolysis. In addition, inhibitors of the calpain/procalpain system in red cell or hemolysate incubations challenged by organic hydroperoxide were without significant effect on the rate of proteolysis. HPLC analysis of red cell incubations was employed to quantify the release of histidine and tyrosine. Whereas cumene hydroperoxide produced a concentration- and time-dependent release of these amino acids, H202failed to stimulate amino acid release. These data c o n f i i the findings of the fluorometric tyrosine assay. The results of this study reveal that organic hydroperoxides produce protein damage in red cells and hemolysate as a result of alkyl or alkoxy1 free radicals generated during hydroperoxide decomposition and that HPLC provides a quantitative approach for monitoring proteolysis in these systems. Proteolysis, which has been suggested as being a defense mechanism for the removal of protein damaged by xenobiotic insult, may also be a rapid and novel index of xenobiotic-mediated cellular injury.

Introduction It is well recognized that a number of drugs such as nitrofurantoin and certain foods (e.g., fava beans) are capable of producing oxidant stress in red blood cells from individuals with deficiencies in key metabolic enzymes which are required for maintenance of red cell homeostasis, such as glucose-6-phosphate dehydrogenase, glutathione peroxidase, glutathione reductase, phosphogluconic dehydrogenase, and the enzymes active in glutathione synthesis (I). The ultimate response to drug- or food-induced *Address correspondence to this author at The Institute of Chemical Toxicology, Wayne State University, 2727 Second Ave., Detroit, MI 48201.

oxidant streas is the development of nonimmune hemolytic anemia (I). In the erythrocytes, oxidant stress is thought to result in the oxidation of Hb to metHb with the subsequent denaturation of metHb to hemichrome and nucleation of protein aggregates at the membrane (2). The ultimate recognition of such damage by the spleen results in cell lysis ( I ) . Red cells contain a complex system of proteolytic enzymes capable of degrading intracellular protein which is damaged as a consequence of either biosynthetic error, senescent degeneration, or oxidant stress. The ATP-dependent proteolytic pathway initially described for rabbit red blood cells (3) has been shown to play a relatively minor role in the degradation of abnormal protein (4). By use of inhibitors that block the ATP-dependent pathway 0 1989 American Chemical Society

Hydroperoxide-Stimulated Proteolysis in Red Cells in reticulocytes and reticulocyte lysate, an ATP-independent proteolytic pathway was shown to be involved primarily in the degradation of protein damaged by phenylhydrazine ( 4 ) . It has been reported recently (5,6)that rabbit red blood cells exposed to various oxygen freeradical-generating systems exhibited a rapid increase in proteolysis, as monitored by alanine release, which preceded stimulation of lipid peroxidation. Evidence has been provided for the role of a calcium-dependent neutral protease on the degradation of oxidant-damaged proteins, and it has been suggested that this system may play a role in the prevention of Heinz body formation and consequent cell lysis (7, 8). The treatment of red cells and red cell ghosts with organic hydroperoxides has been studied as a potential model for oxidative stress. tert-Butyl hydroperoxide (BHP)l addition to red cells has been shown previously to result in Hb oxidation, lipid peroxidation (9-12), and echinocyte formation (12). It has been suggested that, in the red cell, Hb decomposes the hydroperoxide by homolytic scission to tert-butoxyl radicals which presumably bind macromolecules such as Hb (11). Both BHP and cumene hydroperoxide (CHP) have been shown to function as tumor promoters in mouse skin (13,14), and more recently the formation of alkoxyl and alkyl (ethyl, methyl) free radicals has been detected in murine basal keratinocytes following exposure to tert-butyl or cumene hydroperoxide (15). While the organic hydroperoxides are active as tumor promoters, H202exhibits only marginal activity (14). Since BHP and CHP may serve as models for lipid hydroperoxides generated in cell membranes and yield alkoxyl and alkyl radicals, the ability of these agents to stimulate protein degradation in the red cell was characterized. In the present paper, the relative effects of BHP, CHP, and H202on proteolysis in the human red cell and red cell hemolysates are examined. Our results show that whereas these organic hydroperoxides produced a pronounced concentration-dependent stimulation in the rate of proteolysis which occurred rapidly and proceeded linearly with time to 2 h, H202addition alone, or in the presence of the catalase inhibitor 3-AT, failed to stimulate proteolysis. In addition, data are presented that demonstrate the feasibility of examining proteolysis in cellular preparations via a highly sensitive HPLC analysis of amino acids which are released in red cell or hemolysate incubations at the picomolar level.

Experimental Procedures Chemicals. tertButy1 hydroperoxide, cumene hydroperoxide, 3-amino-1,2,4-triazole,the chelator diethylenetriaminepentaacetic acid (DETAPAC), cycloheximide, and the inhibitors leupeptin, antipain, chymostatin, and hemin as well as amino acid standards N-oxide (DMPO) were and the spin trap 5,5-dimethyl-l-pyrroline purchased from Sigma Chemical Co., St. Louis, MO. Blood Sample Preparation. Units of freshly isolated packed human red blood cells were obtained from the blood bank or individual donors and prepared as described previously (16,17). Aliquots of packed red cells were washed three times in normal saline and then diluted to yield a final hematocrit (HCT) of 35% with a buffer consisting of 115 mM NaC1, 4 mM KCl, 1 mM MgCl,, 0.3 mM Na@04, and 20 mM monobasic sodium phosphate. The buffer was then adjusted to pH 7.45 with NaOH and made 10 mM in glucose. Hemolysates were prepared from red cells by using deionized water and were diluted with buffer to a final concentration of Hb consistent with that present in a 35% HCT. Abbreviations: BHP, tert-butyl hydroperoxide;CHP, cumene hydroperoxide; HCT, hematocrit; metHb, methemoglobin:3-AT, 3-aminol,2,4-triazole;DETAPAC, diethylenetriaminepentaaceticacid; DMPO, 5,5-dimethyl-l-pyrrolineN-oxide.

Chem. Res. Toxicol., Vol. 2, No.2, 1989 77 Incubations. Incubations were performed in 50-cm3glass vials in a 37 "C shaking water bath for various time intervals which ranged from 30 min to 21 h. The vials were sealed under a positive oxygen pressure. Red blood cell suspensions or hemolysates were incubated in the absence or presence of organic hydroperoxide (1.0-8.0 mM) and contained cycloheximide (50 mM) as described previously (3). Studies using the catalase inhibitor 3-AT (50,100, or 200 mM), the antioxidants, DETAPAC, DMPO, or protease inhibitors involved the preincubation of red cell suspensions or lysates with the respective agent for 2 h at 37 "C prior to HzOz or organic hydroperoxide addition. Studies using 3-AT employed a 5-8% HCT, and inhibition of catalase activity after HzOzaddition was confirmed as described previously (17). After the incubation,aliquots were taken for tyrosine assay or HPLC amino acid analysis. Tyrosine Assay. A fluorescence assay for tyrosine content was used as an indicator of proteolysis (18,19). A 1-mL aliquot of the red cell suspension was precipitated with an equal volume of 20% trichloroacetic acid and left to stand a t 4 "C overnight. Samples were then vortexed and centrifuged at high speed in a clinical centrifuge for 15 min. One milliliter of supernatant from each sample was treated with 1.0 mL of 0.2% 1-nitroso-Znaphthol and 1.0 mL of acid nitrite reagent (10 mg/mL NaNOz in 20% HNOJ, respectively, and then incubated for 30 min at 37 "C. After incubation, samples were extracted with ethylene dichloride and centrifuged for 10 min a t high speed. The fluorescence of the aqueous phase was determined with emission at 570 nm and excitation at 460 nm. Tyrosine content of the sample was determined from a standard curve. A standard curve corrected for reagent or inhibitor (H202, hydroperoxide, etc.) effects on fluorescence was employed for treated samples. Hz02 addition was without significant effect (510%) on tyrosine standards. Since 3-AT quenches the tyrosine-naphthol chromophore with an extremely steep concentration dependence, separate parallel controls for untreated, 3-AT-treated, and 3-AT-peroxide-treated samples had to be used concurrently to evaluate 3-AT-Hz02 effects on tyrosine production. Unless otherwise stated, experiments were performed in triplicate on blood from two or three different donors or units obtained from commercial sources and results presented as the mean f standard deviation. HPLC Amino Acid Analysis. A 0.5-1.0-mL volume of red cells was added dropwise to an equal volume of acetonitrile and vortexed vigorously. Following centrifugation (2000g for 10 min) in a clinical centrifuge, the supernatant was removed and derivatized directly by using a modification of a method described previously (20). A 10-pL volume of the supernant was placed in a Reacti-Vial (Pierce Chemical), and to this was added 20 pL of 50 mM NaHC03 followed by 40 pL of 4 mM 4-(dimethylamino)azobenzene-4'sulfonyl chloride (dabsyl chloride) in acetonitrile. The sample vials were sealed and heated at 70 "C in a water bath for 10 min after which the samples were cooled to room temperature and analyzed by HPLC. HPLC analysis was performed on an LDC/Milton Roy system with Gradient Master and Spectromonitor I11 variable-wavelength detector or on a Bio-Rad Model 400 HPLC system. A 5-pm CIBcolumn (Spheri-5, Brownlee Labs) fitted with a guard column was used. A flow rate of 1.9 mL/min was employed and sample detection accomplished at 436 nm using 0.02 AUFS. The solvent gradient consisted of solvent A (26 mM sodium acetate in 4% dimethylformamide) and solvent B (100% acetonitrile) with an increase from 10 to 35% solvent B attained in 20 min followed by an increase from 35 to 55% solvent B in 12 min. The appearance of tyrosine and histidine in the HPLC chromatograms was confirmed by using derivatized standards, and the linearity of the peak amplitude with concentrations from 20 to 75 pmol was confirmed by using a standard curve.

Results When BHP and CHP were incubated in the presence of red blood cells at a 35% hematocrit for 2 h, a concentration-dependent increase in proteolysis was observed. Of the two hydroperoxides examined, CHP was clearly the most efficacious in producing protein damage. CHP at 3 and 8 mM produced a 2.6- and 4.1-fold increase in the rate of proteolysis in red cells, respectively, relative to control,

Runge-Morris et al.

78 Chem. Res. Toxicol., Vol. 2, No. 2, 1989 Table I. Effects of Organic Hydroperoxides on Proteolysis in Human Red Cells and Red Cell Hemolysate addition tyrosine release: nmol/(mL of RBCsSh) % control Red Cells none 11.4 f 0.4 100 +CHP 1 mM 16.0 f 0.6 140 3 mM 29.7 f 6.2 261 46.6 f 1.8 409 8 mM

+BHP 1 mM 3 mM 8 mM

none +CHP 1mM 3 mM 8 mM

E

0

E

C

*

/

20

L

14.4 f 0.6 26.8 f 2.4 33.4 f 2.2

100 186 231

Hemolysate 12.2 f 2.8

100

16.4 f 4.6 30.6 f 8.4 35.2 f 10.2

143 251 288

13.8 f 3.4 23.8 f 5.4 26.4 f 5.4

113 195 216

+BHP

1 mM 3 mM 8 mM

-

/$

I-

$\:/f-f-

10

t 0

1

2

3

Time (hr)

b

I

Data represent the mean ISD of three determinations on red cells from two or three different donors.

whereas BHP at 3 and 8 mM produced a 1.9- and 2.3-fold increase in the rate of tyrosine release, respectively, as compared to untreated red cells (Table I). The maximal rate of hydroperoxide-stimulated tyrosine release was monitored at 8 mM CHP with a slight decrease in rate monitored at higher concentrations while BHP produced a maximal increase in the rate of proteolysis at 8 mM. The concentration dependence for stimulation of proteolysis by the hydroperoxides in red cell hemolysate was less striking than that monitored for red cells with a maximal rate of tyrosine release monitored at -3 mM CHP or BHP. Although CHP was also more efficacious in stimulating protein degradation in hemolysate, the difference between CHP and BHP was not as dramatic as that observed in red cells. Concentrations of 3 and 8 mM CHP stimulated the rate of proteolysis 2.5- and 2.9-fold, respectively, while BHP at identical concentrations increased proteolysis by 2.0- and 2.2-fold, respectively, as compared to control rates (Table I). Interestingly, the organic hydroperoxides remained effective in stimulating the rate of proteolysis even in the absence of cellular integrity and a lipid environment. Hydroperoxide-stimulated proteolysis in red cell suspensions or hemolysate proceeded linearly with time and reached a plateau 3 h posttreatment (Figure 1) while control red cells or hemolysate failed to show a significant change in the rate of protein degradation over the 3-hour interval. The increased rate of proteolysis associated with hydroperoxide-treated red cells or hemolysate was found to continue for over 6 h, demonstrating that although the maximal rate (and protein damage) was achieved rapidly and reached a plateau in -3 h, proteolysis continued, suggesting that this system continues to degrade damaged protein. Because of the significant potential for interference of inhibitors in the fluorometric assay as well as limitations with respect to sample size, preparation time, specificity, and sensitivity, analysis of the rate of amino acid release in treated and untreated samples was accomplished by using an HPLC technique essentially according to previously published methods (20). Although some 16 individual amino acids were observed in the samples from control and treated cells (Runge-Morris and Novak, unpublished results), the data presented will focus on

0

1

2

3

T i m e (hr)

Figure 1. Time-dependent stimulation of proteolysis by hydroperoxide in red cells or hemolysate. (a) Time-dependent increase in proteolysis (tyrosine release) monitored for control (v)and 4 mM cumene hydroperoxide treated ( 0 )red cells. (b) Time-dependent increase in proteolysis monitored for control (v) and 4 mh4 cumene hydroperoxide treated (0) hemolysate. Results are presented as the mean hSD of duplicate determinations. Incubations were performed as described under Experimental Procedures. quantifying changes in histidine and tyrosine levels since these peaks have been unambiguously assigned in the observed HPLC chromatograms, were readily resolved and standard curves of peak amplitude versus concentration were linear over the concentration range of interest. Both histidine and tyrosine levels increased in the samples treated with 5 mM CHP as compared to controls. Panel A in Figure 2 contains the HPLC chromatogram of histidine and tyrosine standards, each present at 71 pmol, with retention times of 30 and -32.5 min, respectively. Panel B of Figure 2 represents an amino acid profile of the same chromatographic region obtained from red cells incubated for 2 h at 37 "C under an atmosphere of 100% oxygen in the absence of hydroperoxide. The levels of histidine and tyrosine present in the sample correspond to rates of 11 and 8 nmol/(mL of RBCseh), respectively, in agreement with values of 15 to 7 nmol tyrosine/(mL of RBCs-h) obtained from preparations of red cells from different donors using the fluorometric assay. The effects of 3 and 5 mM CHP on amino acid release are shown in Figure 2, panels C and D, respectively. As can be observed readily, there is a substantial increase (-3-fold) in the histidine and tyrosine peaks relative to control for red cells treated with 3 mM CHP for 2 h (Figure 2, panel C). Moreover, there is an even greater increase in histidine and tyrosine peak amplitudes between control and red cells

-

Chem. Res. Toxicol., Vol. 2, No. 2, 1989 79

Hydroperoxide-Stimulated Proteolysis in Red Cells 01

B

2s

ili

1

L L r L a's

30

MINUTES

Ib

M I NUT€ S

MINUTES

MINUTES

01

Y

U

Y

Y

U

U

z

/I1

z