Chem. Res. Toxicol. 1992,5, 425-431 (14)Crowell, T., and Hammett, L. P. (1948)Kinetics of the reactions of thiosulphate ions with ethyl, propyl and isopropyl bromides. J. Am. Chem. SOC. 70,3444-3450. (15) Chamulitrat, W., and Mason, R. P. (1989)Lipid peroxyl radical intermediates in the peroxidation of polyunsaturated fatty acids by lipoxygenase. Direct electron spin resonance investigation. J. Biol. Chem. 264,968-973. (16) Neidlemann, S.L., and Geigert, J. (1986)Compendium of in vitro haloperoxidase reactions. in Biohalogenation, Principles, basic roles and applications. Ellis Horwood Series in Organic Chemistry, pp 61-69, Ellis Horwood Publishers, Halstead Press, Chichester, England, and John Wiley and Sons, New York.
425
(17) Struku, G.,Sinigalia, R., Zanardo, A., Pinna, F., and Michelin, R. (1989)Selective oxidation of olefins catalyzed by Pt(I1) complexes. Inorg. Chem. 28,554-559. (18) Burkus, J., and Eckert, C. F. (1958)The kinetics of the triethylamine-catalyzed reaction of diisocyanates with l-butanol in 80,5948-5950. toluene. J. Am. Chem. SOC. (19) McMillan, W.G. (1957)Determination of rate constant ratio in competitive consecutive second-order reactions. J.Am. Chem. SOC. 79,4838-4839. (20) Masako, T., Toshiyuki, M., Takeshi, M., and Komei, M. (1973) Oxidation products derived from methyl linoleate and their toxicities in mice and in chick embryos. Yukagaku 22, 259-264.
Inactivation of a1-Proteinase Inhibitor by Peroxynitrite Juan J. Moreno and William A. Pryor* Department of Chemistry and Biodynamics Institute, Louisiana State university, Baton Rouge, Louisiana 70803 Received December 23, 1991
We here report the reactions of peroxynitrite (O=NOO-) with al-proteinase inhibitor (alPI) and with a synthetic decapeptide (MER10) containing the sequence of amino acids found in Peroxynitrite inactivates the active site of alp1 (Pro-Met-Ser-Ile-Pro-Pro-Glu-Val-Lys-Phe). alp1 a t both pH 7.4 and 12.0. Thiourea and methionine protect alp1 against inactivation by peroxynitrite, while mannitol and benzoate fail to afford effective protection. The major product isolated from the reaction between peroxynitrite and MER10 was analyzed by NMR, mass spectrometry, and amino acid analysis. These analyses indicate that peroxynitrite primarily oxidizes the methionine residue in the peptide. We detect neither smaller molecular weight peptides, which would indicate cleavage of MER10, nor hydroxylation or nitration of the phenylalanine residue. Our results suggest that peroxynitrite is capable of oxidizing methionine residues in proteins without the involvement of the hydroxyl radical or nitrogen dioxide. The implications of these observations on lung diseases attributed to cigarette smoke are discussed.
I ntroductlon The suppression of antiprotease activity in the lung by cigarette smoke is one of the most widely accepted explanations for the pathogenesis of pulmonary emphysema in smokers (I), but the mechanism(s) by which cigarette smoke causes (cause) this suppression is (are) still not well understood. It has been shown that pulmonary alveolar macrophages (PAM)' collect in the bronchiole of smokers before and during tissue destruction and fibrosis (2,3)and that superoxide (Of)production by PAM is significantly increased in the lungs of chronic smokers over nonsmokers (4,5).The presence of cigarette smoke in the lung also causes increased concentrations of neutrophils, suggesting that emphysematous lesions may result from the destruction of lung tissue by these cells (6). Both PAM and neutrophils are capable of producing and releasing nitric oxide (NO) and 0,-(7-9),which could react to form peroxynitrite2 (O=NOO-), a powerful oxidant (10,11). The reaction of NO and 0; is diffusion-controlled in the gas phase (12),and a rate constant of about 3.7 X lo7 M-l s-l has been measured in water under physiological conditions (13). Peroxynitrite also could be formed in the lungs of smokers from the reaction of nitric oxide present in cigarette smoke and endogenous superoxide, which is produced in increased amounts in the lungs of smokers (4,5).Nitric
* Addreas correspondence to this author a t Biodynamics Institute, 711 Choppin, Louisiana State University, Baton Rouge, LA 708031800.
oxide is present in the gas phase of cigarette smoke at concentrations of up to 1000 parts per million (ppm) depending on the type of tobacco from which cigarettes are made (14). Nitric oxide is slowly oxidized to nitrogen dioxide (NO,) in cigarette smoke (15);fresh smoke does not contain NOz (16).The half-life of 1000 ppm NO in air is about 3 min (17), and it has been suggested that the rate of oxidation of NO in the respiratory airways is similar (18). Since smoke is held in the mouth and lungs of smokers for short periods of time, it is likely that little NO2 is formed. To investigate if peroxynitrite could play a role in the pathogenesis of emphysema, we studied the reaction of peroxynitrite with al-proteinase inhibitor (alPI). alp1 is the most abundant extracellular antiprotease in the lung and provides most of the protection against neutrophil elastase in the lower respiratory tract (19,20).We also report on the reaction of a synthetic decapeptide (MERlO), Abbreviations: PAM, pulmonary alveolar macrophages; alPI, human al-proteinase inhibitor; EIC,elastase inhibitory capacity; MER10, a synthetic decapeptide, H-Pro-Met-Ser-Ile-Pro-Pro-Glu-Val-Lys-PheOH; BSA, bovine serum albumin; SANA, N-succinyl(L-alanyl)s-p-nitroanilide; PDMS, plasma desorption mass spectrometry; NMR,nuclear magnetic resonance; TSP, sodium 3-(trimethylsilyl)propionate-2,2,3,3-d4; MetSO, methionine sulfoxide. The recommended name for the ion O=NOO- is oxoperoxonitrate (48). For simplicity, in this paper we have chosen to use the common name of peroxynitrite. Since our data do not distinguish between the anion and the protonated species as the oxidant in our system, for convenience we have used the term peroxynitrite or the symbol O=NOOto refer to either the anion or ita conjugate acid, hydrogen oxoperoxonitrate (HOON=O), commonly named pernitrous acid.
Q893-228~/92 f 27Q5-Q425$Q3.OQ/Q0 1992 American Chemical Society
426 Chem. Res. Toxicol., Vol. 5, No.3, 1992
a model for the active site of alPI, with peroxynitrite. The use of this smaller molecule in place of alp1 in the reaction with peroxynitrite enabled us to conduct product analysis by NMR, mass spectrometry, and amino acid analysis.
Experimental Section Chemicals. Human alp1 and porcine pancreatic elastase (EC 3.4.21.36) were purchased from Calbiochem Co. (San Diego, CA). Bovine liver catalase (EC 1.11.1.6), bovine serum albumin, mannitol, thiourea, methionine, sodium benzoate, amino acid standard solution (2.5 mM; 0.1 N HCl), cyanogen bromide, and 1,4-dithiothreitol were all purchased from Sigma Chemical Co. (St. Louis, MO) and used without further purification. Cyanogen bromide is highly toxic and volatile and should be handled in a fume hood with hand protection. Hydrochloric acid (6 N) was purchased from Pierce (Rockford, IL). Chelex-100 resin was purchased from Bio-Rad Laboratories (Richmond, CA). MERlO (Pro-Met-Ser-Ile-Pro-Pro-Glu-Val-Lys-Phe) was synthesized in a Milligen/Biosearch SAM 2 automatic protein synthesizer apparatus (Milligen/Biosearch, San Rafael, CA). This synthetic peptide contains the amino acid sequence of the inhibitory active site of alp1 from Pro357to Phe366(21). The elastase inhibitory capacity (EIC) of the MERlO peptide was tested as described below, and the peptide was found not to be an inhibitor of porcine pancreatic elastase. All other chemicals were of reagent grade and were used without further purification. Unless otherwise indicated sodium phosphate pretreated with Chelex-100 was used as buffer. Synthesis of Peroxynitrite. Peroxynitrite was synthesized using a quenched-flow reactor as described by Reed et al. (22). The purpose of the flow reactor is to quench the acid-catalyzed reaction between NaNOz and H 2 0 2with an excess of NaOH in order to obtain the stabilized peroxynitrite anion. Solutions consisting of 0.6 M NaNOZ,0.7 M HzOz in 0.6 M HC1, and 1.5 M NaOH were pressurized with nitrogen and passed through the reactor at equal flow rates (25 i 5 mL/min). T h e product was collected in a flask immersed in ice water. Excess H202was removed by passing the reaction mixture through a column (10 X 1.5 cm) packed with manganese oxide (MnO,); the resulting solution containing a large excess of alkali was stored a t -20 "C for periods no longer than 2 weeks. T h e use of MnO, in the removal of excess H 2 0 2 resulted in loss of peroxynitrite; this decomposition of peroxynitrite has previously been attributed to MnO, catalyzing a reaction with H 2 0 2 to give nitrite as a product (23). Peroxynitrite slowly isomerizes t o nitrate a t p H 12-13 (23). Therefore, our peroxynitrite solutions were contaminated with nitrite and nitrate ions. Peroxynitrite solutions form a deep yellow top layer by freeze fractionation; aliquots obtained from this top layer were used for the experiments. T h e concentration of peroxynitrite, determined spectrophotometrically a t 302 nm using an extinction coefficient of 1670 M-' cm-l (24), was typically of about 0.10 & 0.02 M. A cell containing previously-decomposed peroxynitrite in phosphate buffer was used to correct for interference by other compounds (e.g., nitrite) absorbing a t 302 nm. Peroxynitrite Reaction with alPI. Solutions of alp1 (125 pg/mL, 2.4 pM using a MW of 52K) were incubated with peroxynitrite (1-2 mM) in 0.1 or 0.5 M sodium phosphate buffer under constant stirring for 12 h a t 25 "C. Other additives were added as indicated in table footnotes. T h e p H of solutions was measured a t the end of each incubation period and found t o be 10.4-11.5 f 0.5 using 0.1 M buffer (pH 7.4) and 7.4 f 0.1 using 0.5 M buffer (pH 7.4). Although 0.1 M sodium phosphate (initial p H 7.4) is not sufficiently concentrated to completely buffer the reaction with peroxynitrite, these conditions were used to reproduce those previously utilized to study the reaction of alp1 with cigarette smoke (25). Reactions with the synthetic peptide, MER10, were carried out in a 2:1 molar ratio of peptide:peroxynitrite in 0.1 M buffer with incubation periods of 24 h a t 37 "C. Protein or peptide samples used as controls were incubated and treated under the same conditions as peroxynitrite-treated samples but in the absence of peroxynitrite. Elastase Inhibition Assay. Assays for inhibitory activity were performed by measuring the decrease in elastase enzymatic activity resulting from preincubation with native or peroxynitrite-treated
Moreno and Pryor a l P I . Elastase activity was measured according t o the method described by Bieth et al. (26). Briefly, in 952 WLof 0.2 M Tris-HC1 buffer (pH 8.0), 10 pL of elastase (0.14 unit/mL) is incubated with 30 p L of alp1 (3.75 Kg/mL) for 5 min at 25 "C. After the incubation period, 8 KL of a 125 m M solution of N-succinyl-(L-alanyl),-p-nitroanilide (SANA) in 1-methyl-2-pyrrolidone is added and the increase in absorbance at 410 nm is continuously measured for the first 2 min against a reference consisting of buffer and SANA only. T h e initial linear rate of absorbance change due to uninhibited elastase was taken as 100% elastase activity; the decrease in this rate due to native alp1 was taken as 100% EIC. HPLC. T h e products of the reaction between peroxynitrite and MER10 were isolated by semipreparative reverse-phase HPLC in an Econosil C18 1OU Alltech column (10 pm, 250 X 10 mm) (Alltech Associates, Inc., Deerfield, IL). The mobile phase consisted of a ternary system of water, acetonitrile, and 2-propanol; all solvents contained 0.1% trifluoroacetic acid. The componenta of the reaction mixture were separated using a flow rate of 3 mL/min and an initial eluent composition of 80% water, 15% acetonitrile, and 5 % 2-propanol. The mobile-phase composition was linearly increased to 95% acetonitrile over a period of 45 min, keeping 2-propanol constant a t 5%. Two major fractions were collected from the reaction of MERlO and peroxynitrite comprising 65% (unreacted peptide) and 35% of the starting material using 210 nm for detection. Isolated fractions were evaporated under vacuum and lyophilized prior t o molecular weight determination, NMR, and amino acid analysis. Amino Acid Analysis. Peptide samples were hydrolyzed with 6 N HC1 at 110 "C for 24 h in vacuo. Amino acids were separated and quantitated by HPLC in an AminoQuant column (Hewlett-Packard) using precolumn derivatization according to Schuster and Apfel(27). Oxidation of methionine to methionine sulfoxide was determined by the method described by Shechter et al. (28). Briefly, lyophilized peptide samples were dissolved in 80% formic acid and allowed t o react with cyanogen bromide (0.1 M) for 24 h a t room temperature. The reaction was stopped by addition of an equal volume of water, and the samples were then frozen and lyophilized. The cyanogen bromide peptides were hydrolyzed as described above but in the presence of lP-dithiothreitol. NMR Analysis of MERIO. All NMR spectra were recorded at 25 "C in D,O on a Bruker AM-400 spectrometer operating at 400.13 MHz equipped with an Aspect 3000 computer. The chemical shifts were relative to internal sodium 3-(trimethylsilyl)propionate-2,2,3,3-d4 (TSP)in DzO.Spectra were collected with 8K complex data points over a 3600-5000-Hz sweep width with a minimum of 64 and a maximum of 640 scans. The residual HOD signal from the solvent (D20)was suppressed by low-power presaturation prior to data acquisition. Time-of-FlightMS Analysis. All mass spectra were obtained on a plasma desorption time-of-flight Bio-Ion 20 mass spectrometer (Bio-Ion, Uppsala, Sweden). Plasma desorption mass spectrometry (PDMS) utilizes a 262Cfsource to ionize molecules, which subsequently are analyzed using the time-of-flight technique, and provides optimal soft ionization preventing fragmentation of biological molecules.
Results Peroxynitrite inactivates alp1 in a dose-dependent manner at pH 10-11.5 (Figure 1). Solutions of peroxynitrite are contaminated with nitrite and nitrate ions, and possibly HzOz. Although most of the HzOz is probably removed by treating the crude peroxynitrite solution with MnO, (see Experimental Section), traces of hydrogen peroxide left after this treatment could oxidize the methionine residue in the active site of alp1 (Metw), causing its inactivation (21). In order to assure ourselves that alp1 is inactivated by peroxynitrite and not by either H202 or a mixture of nitrite and nitrate ions, alp1 was incubated with peroxynitrite, an equivalent solution of peroxynitrite thermally decomposed at 37 "C prior to addition, and a mixture of nitrite and nitrate ions. Figure 2 shows that addition of 2 mM peroxynitrite causes complete inactivation, while an equivalent concentration of previouslydecomposed peroxynitrite does not inactivate alPI. Sim-
Chem. Res. Toxicol., Vol. 5 , No. 3, 1992 427
Methionine Oxidation by Peroxynitrite
Table I. Protection by Hydroxyl Radical Scavengers and Catalase against Inactivation of alp1 by Peroxynitritea
I\
90.0 '05.0
ts I
75.0
concn in (O=NOO-) comDound final svstem (mM) catalasec 100 unita/mL 2 300 units/mL 1 BSA 5.33 pg/mL 1 methionine 20 mM 2 25 mM 1 thiourea 20 mM 2 20 mM 1 50 mM 1 benzoate 20 mM 1 50 mM 1 mannitol 50 mM 1
L
30.0 15.0
0.0 0.0
0.4 0.7
1.4
1.1
[ON=OO-]
1.8
2.5
2.1
(mM)
Figure
1. Inactivation of alp1 as a function of peroxynitrite concentration. al-Proteinase inhibitor (125 p g / m L ) was exposed to different concentrations of peroxynitrite for 12 h at 37 "C in 0.1 M phosphate buffer (initial p H 7.4, final p H 10-11.5 f 0.5) and assayed for antielastase activity according to the method of Bieth e t al. (26). Each data point is the average of duplicate measurements of two independent determinations.
040
1
T
0.35 030
A
G
025 0.20
0
.
14
T O
005
oo
1
4 0
1
€c
io @O
0
0
25
0 *,@ 0
0
0 0
03
0
0
0
'
I '' 5c 75 1 oc
TIME
(DH10.4-11.5)
(DH7.4) NDd ~
0.0
ND ND
9.2 f 1.5 4.8 f 0.8
99.6 f 1.4 100.0 f 2.8 99.0 0.7
ND
*
97.7 f 0.6
ND 10.0 f 5.2 11.8 f 0.5
ND ND ND 99.0 f 0.1 104.0 f 0.0 19.3 f 5.7 15.6 f 2.4 10.9 f 4.8
"Reaction mixtures containing alp1 (125 pg/mL), added compounds in the indicated concentrations, and peroxynitrite (1-2 mM) were incubated for periods of 3 h (25 "C) in 0.5 M phosphate buffer (final pH 7.40 f 0.05) or 12 h (25 "C) in 0.1 M phosphate buffer (final pH 10.4-11.5). bPercent protection was calculated as 1 W % damage, and 90 damage = 100 X [(lo0 - %EICA)/(lOO- %EICB)],where '%EICA" is the EIC measured for alp1 plus peroxynitrite plus additive and "%EICB"is the EIC measured for alp1 plus peroxynitrite only. Values are presented as average % protection f standard deviation calculated from duplicate runs of at least two independent experimenta. elOO unita/mL catalase corresponds to 7.1 pg/mL, and 300 unita/mL, to 21.3 pg/mL. d'ND" indicates value not determined. Table 11. Rate Constants and Calculated Reaction Rates between the Hydroxyl Radical and Various Components Present in Incubation Mixtures of alPI and Peroxynitrite reaction of the rate constanta concnb reaction hydroxyl radical with (L mol-' s-l) (mol L-l) ratee (s-l) 10" 2.3 X 10" 2 X lo6 protein hydrogen phosphate 1.5 X lo6 0.4 6 X lo4 dihydrogen phosphate 2 x lo4 0.1 2 x 103 thiourea 3.9 X lo9 2.0 X lo-* 8 X 10' benzoate 5.9 x 109 5.0 x io-* 3 x 10s 1.7 x 109 mannitol 5.0 x 10-2 Q x 107 8.3 x 109 2.0 x io-* 2 x 108 methionine "The rate constants (pH 7,25 "C)were obtained from Buxton et al. (47) and used to calculate the predicted rate of reaction between the hydroxyl radical and the component listed in column l present in the alp1 incubation mixtures. These rates are used to show that certain scavengers (e.g., benzoate and mannitol) would have scavenged HO' had it been present. *Maximum concentrations used for the various components present in the incubation mixture. 'Estimated rates for the reactions of HO' with the species in column 1.
I-
0 10 j
% % protectionb protection
(5)
Figure 2. Effect of previously-decompod peroxynitrite, nitrate, and nitrite ions on the antielastase activity of alPI. Elastase activity (increase in absorbance a t 410 nm, see Experimental Section) in the presence of alp1 (125 rg/mL) solutions (final volume 250 pL) incubated at 25 "C for 3 h with 2 m M peroxynitrite (m), 0.3 M nitrate 0.3 M nitrite (o),or 2 mM decomposed peroxynitrite (0)in 0.5 M phosphate buffer (pH 7.4). Decomposed peroxynitrite was prepared by incubating it in buffer at 37 "C until the absorbance band a t 302 nm had disappeared. Uninhibited elastase ( 0 )and untreated alp1 ( 0 )are shown for comparison purpoeea. Error bars appear only where they are larger than symbols,and they represent the SD of duplicate runs of two determinations.
+
ilarly, the antielastase activity of alp1 is not affected upon incubation with nitrite and nitrate ions. Moreover, Table I shows that addition of catalase provides minimum protection (ca. 9%) against inactivation of alPI. About the
same degree of protection observed with catalase is obtained with BSA (ca. 5%) and with heat-denatured catalase (data not shown), suggesting that catalase protecta by a sacrificial role rather than by ita catalytic activity. These results illustrate that inactivation of the protein is caused by peroxynitrite and not by a mixture of nitrite and nitrate ions or HzOz. Pernitrous acid has a pK, of 6.8 (IO); peroxynitrite (O=NOO-) is stable in alkaline solutions (24), but ita conjugate acid, pernitrous acid, rapidly decompcses. It has been proposed (11, 29) that the decomposition of peroxynitrite generates a strong oxidant with reactivity similar to that of the hydroxyl radical. A number of hydroxyl radical scavengers were studied to determine if they inhibit the inactivation of alp1 by peroxynitrite. Table I shows that, of the scavengers tested, thiourea affords complete protection, while benzoate and mannitol do not protect alp1 significantly even at a 50-fold excess with respect to peroxynitrite. On the other hand, methionine, which is not considered a specific scavenger of the hydroxyl radical, completely prevents alp1 inactivation. Similar results are obtained at basic and physiological pH (Table I).
Moreno and Pryor
428 Chem. Res. Toxicol., Vol. 5, No. 3, 1992
Table 111. Amino Acid Composition of MERlO Exposed to Peroxynitrite MERlO + MERlO + residue" MERlO O=NOO- * residue" MERlO O=NOO-b Glu 10.5 f 0.4 10.4 f 0.7 Ile 10.0 f 0.1 10.3 f 0.2 Ser 9.9 f 0.6 8.6 f 0.3 Phe 10.0 f 0.1 9.8 f 0.6 Met 10.4 f 0.7 0.4 f 0.3 Lys 10.5 f 0.5 9.2 f 0.9 Val 9.8 f 0.4 10.3 f 0.2 Pro 29.3 f 3.0 32.5 f 2.1 MetSO' 0.1 f 0.1 8.5 f 0.6
lA
I om0
eooo; n
Residue amounts expressed as mol % f sd of duplicate runs of two independent experiments. Major product isolated from the reaction between MERlO and peroxynitrite (see HPLC section in Experimental Section). Methionine sulfoxide was determined by the cyanogen bromide method (see Experimental Section).
0
2500
1
N N W
0 '
-v
-IN N
, -
, I
h//
lO0Oi
PIjj
-t
5OOj
,
I
-&\ sbo
iH /I
0
RbO
1000
1200
1400 1600 1800
zoo0
425
(17) Struku, G.,Sinigalia, R., Zanardo, A., Pinna, F., and Michelin, R. (1989)Selective oxidation of olefins catalyzed by Pt(I1) complexes. Inorg. Chem. 28,554-559. (18) Burkus, J., and Eckert, C. F. (1958)The kinetics of the triethylamine-catalyzed reaction of diisocyanates with l-butanol in 80,5948-5950. toluene. J. Am. Chem. SOC. (19) McMillan, W.G. (1957)Determination of rate constant ratio in competitive consecutive second-order reactions. J.Am. Chem. SOC. 79,4838-4839. (20) Masako, T., Toshiyuki, M., Takeshi, M., and Komei, M. (1973) Oxidation products derived from methyl linoleate and their toxicities in mice and in chick embryos. Yukagaku 22, 259-264.
Inactivation of a1-Proteinase Inhibitor by Peroxynitrite Juan J. Moreno and William A. Pryor* Department of Chemistry and Biodynamics Institute, Louisiana State university, Baton Rouge, Louisiana 70803 Received December 23, 1991
We here report the reactions of peroxynitrite (O=NOO-) with al-proteinase inhibitor (alPI) and with a synthetic decapeptide (MER10) containing the sequence of amino acids found in Peroxynitrite inactivates the active site of alp1 (Pro-Met-Ser-Ile-Pro-Pro-Glu-Val-Lys-Phe). alp1 a t both pH 7.4 and 12.0. Thiourea and methionine protect alp1 against inactivation by peroxynitrite, while mannitol and benzoate fail to afford effective protection. The major product isolated from the reaction between peroxynitrite and MER10 was analyzed by NMR, mass spectrometry, and amino acid analysis. These analyses indicate that peroxynitrite primarily oxidizes the methionine residue in the peptide. We detect neither smaller molecular weight peptides, which would indicate cleavage of MER10, nor hydroxylation or nitration of the phenylalanine residue. Our results suggest that peroxynitrite is capable of oxidizing methionine residues in proteins without the involvement of the hydroxyl radical or nitrogen dioxide. The implications of these observations on lung diseases attributed to cigarette smoke are discussed.
I ntroductlon The suppression of antiprotease activity in the lung by cigarette smoke is one of the most widely accepted explanations for the pathogenesis of pulmonary emphysema in smokers (I), but the mechanism(s) by which cigarette smoke causes (cause) this suppression is (are) still not well understood. It has been shown that pulmonary alveolar macrophages (PAM)' collect in the bronchiole of smokers before and during tissue destruction and fibrosis (2,3)and that superoxide (Of)production by PAM is significantly increased in the lungs of chronic smokers over nonsmokers (4,5).The presence of cigarette smoke in the lung also causes increased concentrations of neutrophils, suggesting that emphysematous lesions may result from the destruction of lung tissue by these cells (6). Both PAM and neutrophils are capable of producing and releasing nitric oxide (NO) and 0,-(7-9),which could react to form peroxynitrite2 (O=NOO-), a powerful oxidant (10,11). The reaction of NO and 0; is diffusion-controlled in the gas phase (12),and a rate constant of about 3.7 X lo7 M-l s-l has been measured in water under physiological conditions (13). Peroxynitrite also could be formed in the lungs of smokers from the reaction of nitric oxide present in cigarette smoke and endogenous superoxide, which is produced in increased amounts in the lungs of smokers (4,5).Nitric
* Addreas correspondence to this author a t Biodynamics Institute, 711 Choppin, Louisiana State University, Baton Rouge, LA 708031800.
oxide is present in the gas phase of cigarette smoke at concentrations of up to 1000 parts per million (ppm) depending on the type of tobacco from which cigarettes are made (14). Nitric oxide is slowly oxidized to nitrogen dioxide (NO,) in cigarette smoke (15);fresh smoke does not contain NOz (16).The half-life of 1000 ppm NO in air is about 3 min (17), and it has been suggested that the rate of oxidation of NO in the respiratory airways is similar (18). Since smoke is held in the mouth and lungs of smokers for short periods of time, it is likely that little NO2 is formed. To investigate if peroxynitrite could play a role in the pathogenesis of emphysema, we studied the reaction of peroxynitrite with al-proteinase inhibitor (alPI). alp1 is the most abundant extracellular antiprotease in the lung and provides most of the protection against neutrophil elastase in the lower respiratory tract (19,20).We also report on the reaction of a synthetic decapeptide (MERlO), Abbreviations: PAM, pulmonary alveolar macrophages; alPI, human al-proteinase inhibitor; EIC,elastase inhibitory capacity; MER10, a synthetic decapeptide, H-Pro-Met-Ser-Ile-Pro-Pro-Glu-Val-Lys-PheOH; BSA, bovine serum albumin; SANA, N-succinyl(L-alanyl)s-p-nitroanilide; PDMS, plasma desorption mass spectrometry; NMR,nuclear magnetic resonance; TSP, sodium 3-(trimethylsilyl)propionate-2,2,3,3-d4; MetSO, methionine sulfoxide. The recommended name for the ion O=NOO- is oxoperoxonitrate (48). For simplicity, in this paper we have chosen to use the common name of peroxynitrite. Since our data do not distinguish between the anion and the protonated species as the oxidant in our system, for convenience we have used the term peroxynitrite or the symbol O=NOOto refer to either the anion or ita conjugate acid, hydrogen oxoperoxonitrate (HOON=O), commonly named pernitrous acid.
Q893-228~/92 f 27Q5-Q425$Q3.OQ/Q0 1992 American Chemical Society
426 Chem. Res. Toxicol., Vol. 5, No.3, 1992
a model for the active site of alPI, with peroxynitrite. The use of this smaller molecule in place of alp1 in the reaction with peroxynitrite enabled us to conduct product analysis by NMR, mass spectrometry, and amino acid analysis.
Experimental Section Chemicals. Human alp1 and porcine pancreatic elastase (EC 3.4.21.36) were purchased from Calbiochem Co. (San Diego, CA). Bovine liver catalase (EC 1.11.1.6), bovine serum albumin, mannitol, thiourea, methionine, sodium benzoate, amino acid standard solution (2.5 mM; 0.1 N HCl), cyanogen bromide, and 1,4-dithiothreitol were all purchased from Sigma Chemical Co. (St. Louis, MO) and used without further purification. Cyanogen bromide is highly toxic and volatile and should be handled in a fume hood with hand protection. Hydrochloric acid (6 N) was purchased from Pierce (Rockford, IL). Chelex-100 resin was purchased from Bio-Rad Laboratories (Richmond, CA). MERlO (Pro-Met-Ser-Ile-Pro-Pro-Glu-Val-Lys-Phe) was synthesized in a Milligen/Biosearch SAM 2 automatic protein synthesizer apparatus (Milligen/Biosearch, San Rafael, CA). This synthetic peptide contains the amino acid sequence of the inhibitory active site of alp1 from Pro357to Phe366(21). The elastase inhibitory capacity (EIC) of the MERlO peptide was tested as described below, and the peptide was found not to be an inhibitor of porcine pancreatic elastase. All other chemicals were of reagent grade and were used without further purification. Unless otherwise indicated sodium phosphate pretreated with Chelex-100 was used as buffer. Synthesis of Peroxynitrite. Peroxynitrite was synthesized using a quenched-flow reactor as described by Reed et al. (22). The purpose of the flow reactor is to quench the acid-catalyzed reaction between NaNOz and H 2 0 2with an excess of NaOH in order to obtain the stabilized peroxynitrite anion. Solutions consisting of 0.6 M NaNOZ,0.7 M HzOz in 0.6 M HC1, and 1.5 M NaOH were pressurized with nitrogen and passed through the reactor at equal flow rates (25 i 5 mL/min). T h e product was collected in a flask immersed in ice water. Excess H202was removed by passing the reaction mixture through a column (10 X 1.5 cm) packed with manganese oxide (MnO,); the resulting solution containing a large excess of alkali was stored a t -20 "C for periods no longer than 2 weeks. T h e use of MnO, in the removal of excess H 2 0 2 resulted in loss of peroxynitrite; this decomposition of peroxynitrite has previously been attributed to MnO, catalyzing a reaction with H 2 0 2 to give nitrite as a product (23). Peroxynitrite slowly isomerizes t o nitrate a t p H 12-13 (23). Therefore, our peroxynitrite solutions were contaminated with nitrite and nitrate ions. Peroxynitrite solutions form a deep yellow top layer by freeze fractionation; aliquots obtained from this top layer were used for the experiments. T h e concentration of peroxynitrite, determined spectrophotometrically a t 302 nm using an extinction coefficient of 1670 M-' cm-l (24), was typically of about 0.10 & 0.02 M. A cell containing previously-decomposed peroxynitrite in phosphate buffer was used to correct for interference by other compounds (e.g., nitrite) absorbing a t 302 nm. Peroxynitrite Reaction with alPI. Solutions of alp1 (125 pg/mL, 2.4 pM using a MW of 52K) were incubated with peroxynitrite (1-2 mM) in 0.1 or 0.5 M sodium phosphate buffer under constant stirring for 12 h a t 25 "C. Other additives were added as indicated in table footnotes. T h e p H of solutions was measured a t the end of each incubation period and found t o be 10.4-11.5 f 0.5 using 0.1 M buffer (pH 7.4) and 7.4 f 0.1 using 0.5 M buffer (pH 7.4). Although 0.1 M sodium phosphate (initial p H 7.4) is not sufficiently concentrated to completely buffer the reaction with peroxynitrite, these conditions were used to reproduce those previously utilized to study the reaction of alp1 with cigarette smoke (25). Reactions with the synthetic peptide, MER10, were carried out in a 2:1 molar ratio of peptide:peroxynitrite in 0.1 M buffer with incubation periods of 24 h a t 37 "C. Protein or peptide samples used as controls were incubated and treated under the same conditions as peroxynitrite-treated samples but in the absence of peroxynitrite. Elastase Inhibition Assay. Assays for inhibitory activity were performed by measuring the decrease in elastase enzymatic activity resulting from preincubation with native or peroxynitrite-treated
Moreno and Pryor a l P I . Elastase activity was measured according t o the method described by Bieth et al. (26). Briefly, in 952 WLof 0.2 M Tris-HC1 buffer (pH 8.0), 10 pL of elastase (0.14 unit/mL) is incubated with 30 p L of alp1 (3.75 Kg/mL) for 5 min at 25 "C. After the incubation period, 8 KL of a 125 m M solution of N-succinyl-(L-alanyl),-p-nitroanilide (SANA) in 1-methyl-2-pyrrolidone is added and the increase in absorbance at 410 nm is continuously measured for the first 2 min against a reference consisting of buffer and SANA only. T h e initial linear rate of absorbance change due to uninhibited elastase was taken as 100% elastase activity; the decrease in this rate due to native alp1 was taken as 100% EIC. HPLC. T h e products of the reaction between peroxynitrite and MER10 were isolated by semipreparative reverse-phase HPLC in an Econosil C18 1OU Alltech column (10 pm, 250 X 10 mm) (Alltech Associates, Inc., Deerfield, IL). The mobile phase consisted of a ternary system of water, acetonitrile, and 2-propanol; all solvents contained 0.1% trifluoroacetic acid. The componenta of the reaction mixture were separated using a flow rate of 3 mL/min and an initial eluent composition of 80% water, 15% acetonitrile, and 5 % 2-propanol. The mobile-phase composition was linearly increased to 95% acetonitrile over a period of 45 min, keeping 2-propanol constant a t 5%. Two major fractions were collected from the reaction of MERlO and peroxynitrite comprising 65% (unreacted peptide) and 35% of the starting material using 210 nm for detection. Isolated fractions were evaporated under vacuum and lyophilized prior t o molecular weight determination, NMR, and amino acid analysis. Amino Acid Analysis. Peptide samples were hydrolyzed with 6 N HC1 at 110 "C for 24 h in vacuo. Amino acids were separated and quantitated by HPLC in an AminoQuant column (Hewlett-Packard) using precolumn derivatization according to Schuster and Apfel(27). Oxidation of methionine to methionine sulfoxide was determined by the method described by Shechter et al. (28). Briefly, lyophilized peptide samples were dissolved in 80% formic acid and allowed t o react with cyanogen bromide (0.1 M) for 24 h a t room temperature. The reaction was stopped by addition of an equal volume of water, and the samples were then frozen and lyophilized. The cyanogen bromide peptides were hydrolyzed as described above but in the presence of lP-dithiothreitol. NMR Analysis of MERIO. All NMR spectra were recorded at 25 "C in D,O on a Bruker AM-400 spectrometer operating at 400.13 MHz equipped with an Aspect 3000 computer. The chemical shifts were relative to internal sodium 3-(trimethylsilyl)propionate-2,2,3,3-d4 (TSP)in DzO.Spectra were collected with 8K complex data points over a 3600-5000-Hz sweep width with a minimum of 64 and a maximum of 640 scans. The residual HOD signal from the solvent (D20)was suppressed by low-power presaturation prior to data acquisition. Time-of-FlightMS Analysis. All mass spectra were obtained on a plasma desorption time-of-flight Bio-Ion 20 mass spectrometer (Bio-Ion, Uppsala, Sweden). Plasma desorption mass spectrometry (PDMS) utilizes a 262Cfsource to ionize molecules, which subsequently are analyzed using the time-of-flight technique, and provides optimal soft ionization preventing fragmentation of biological molecules.
Results Peroxynitrite inactivates alp1 in a dose-dependent manner at pH 10-11.5 (Figure 1). Solutions of peroxynitrite are contaminated with nitrite and nitrate ions, and possibly HzOz. Although most of the HzOz is probably removed by treating the crude peroxynitrite solution with MnO, (see Experimental Section), traces of hydrogen peroxide left after this treatment could oxidize the methionine residue in the active site of alp1 (Metw), causing its inactivation (21). In order to assure ourselves that alp1 is inactivated by peroxynitrite and not by either H202 or a mixture of nitrite and nitrate ions, alp1 was incubated with peroxynitrite, an equivalent solution of peroxynitrite thermally decomposed at 37 "C prior to addition, and a mixture of nitrite and nitrate ions. Figure 2 shows that addition of 2 mM peroxynitrite causes complete inactivation, while an equivalent concentration of previouslydecomposed peroxynitrite does not inactivate alPI. Sim-
Chem. Res. Toxicol., Vol. 5 , No. 3, 1992 427
Methionine Oxidation by Peroxynitrite
Table I. Protection by Hydroxyl Radical Scavengers and Catalase against Inactivation of alp1 by Peroxynitritea
I\
90.0 '05.0
ts I
75.0
concn in (O=NOO-) comDound final svstem (mM) catalasec 100 unita/mL 2 300 units/mL 1 BSA 5.33 pg/mL 1 methionine 20 mM 2 25 mM 1 thiourea 20 mM 2 20 mM 1 50 mM 1 benzoate 20 mM 1 50 mM 1 mannitol 50 mM 1
L
30.0 15.0
0.0 0.0
0.4 0.7
1.4
1.1
[ON=OO-]
1.8
2.5
2.1
(mM)
Figure
1. Inactivation of alp1 as a function of peroxynitrite concentration. al-Proteinase inhibitor (125 p g / m L ) was exposed to different concentrations of peroxynitrite for 12 h at 37 "C in 0.1 M phosphate buffer (initial p H 7.4, final p H 10-11.5 f 0.5) and assayed for antielastase activity according to the method of Bieth e t al. (26). Each data point is the average of duplicate measurements of two independent determinations.
040
1
T
0.35 030
A
G
025 0.20
0
.
14
T O
005
oo
1
4 0
1
€c
io @O
0
0
25
0 *,@ 0
0
0 0
03
0
0
0
'
I '' 5c 75 1 oc
TIME
(DH10.4-11.5)
(DH7.4) NDd ~
0.0
ND ND
9.2 f 1.5 4.8 f 0.8
99.6 f 1.4 100.0 f 2.8 99.0 0.7
ND
*
97.7 f 0.6
ND 10.0 f 5.2 11.8 f 0.5
ND ND ND 99.0 f 0.1 104.0 f 0.0 19.3 f 5.7 15.6 f 2.4 10.9 f 4.8
"Reaction mixtures containing alp1 (125 pg/mL), added compounds in the indicated concentrations, and peroxynitrite (1-2 mM) were incubated for periods of 3 h (25 "C) in 0.5 M phosphate buffer (final pH 7.40 f 0.05) or 12 h (25 "C) in 0.1 M phosphate buffer (final pH 10.4-11.5). bPercent protection was calculated as 1 W % damage, and 90 damage = 100 X [(lo0 - %EICA)/(lOO- %EICB)],where '%EICA" is the EIC measured for alp1 plus peroxynitrite plus additive and "%EICB"is the EIC measured for alp1 plus peroxynitrite only. Values are presented as average % protection f standard deviation calculated from duplicate runs of at least two independent experimenta. elOO unita/mL catalase corresponds to 7.1 pg/mL, and 300 unita/mL, to 21.3 pg/mL. d'ND" indicates value not determined. Table 11. Rate Constants and Calculated Reaction Rates between the Hydroxyl Radical and Various Components Present in Incubation Mixtures of alPI and Peroxynitrite reaction of the rate constanta concnb reaction hydroxyl radical with (L mol-' s-l) (mol L-l) ratee (s-l) 10" 2.3 X 10" 2 X lo6 protein hydrogen phosphate 1.5 X lo6 0.4 6 X lo4 dihydrogen phosphate 2 x lo4 0.1 2 x 103 thiourea 3.9 X lo9 2.0 X lo-* 8 X 10' benzoate 5.9 x 109 5.0 x io-* 3 x 10s 1.7 x 109 mannitol 5.0 x 10-2 Q x 107 8.3 x 109 2.0 x io-* 2 x 108 methionine "The rate constants (pH 7,25 "C)were obtained from Buxton et al. (47) and used to calculate the predicted rate of reaction between the hydroxyl radical and the component listed in column l present in the alp1 incubation mixtures. These rates are used to show that certain scavengers (e.g., benzoate and mannitol) would have scavenged HO' had it been present. *Maximum concentrations used for the various components present in the incubation mixture. 'Estimated rates for the reactions of HO' with the species in column 1.
I-
0 10 j
% % protectionb protection
(5)
Figure 2. Effect of previously-decompod peroxynitrite, nitrate, and nitrite ions on the antielastase activity of alPI. Elastase activity (increase in absorbance a t 410 nm, see Experimental Section) in the presence of alp1 (125 rg/mL) solutions (final volume 250 pL) incubated at 25 "C for 3 h with 2 m M peroxynitrite (m), 0.3 M nitrate 0.3 M nitrite (o),or 2 mM decomposed peroxynitrite (0)in 0.5 M phosphate buffer (pH 7.4). Decomposed peroxynitrite was prepared by incubating it in buffer at 37 "C until the absorbance band a t 302 nm had disappeared. Uninhibited elastase ( 0 )and untreated alp1 ( 0 )are shown for comparison purpoeea. Error bars appear only where they are larger than symbols,and they represent the SD of duplicate runs of two determinations.
+
ilarly, the antielastase activity of alp1 is not affected upon incubation with nitrite and nitrate ions. Moreover, Table I shows that addition of catalase provides minimum protection (ca. 9%) against inactivation of alPI. About the
same degree of protection observed with catalase is obtained with BSA (ca. 5%) and with heat-denatured catalase (data not shown), suggesting that catalase protecta by a sacrificial role rather than by ita catalytic activity. These results illustrate that inactivation of the protein is caused by peroxynitrite and not by a mixture of nitrite and nitrate ions or HzOz. Pernitrous acid has a pK, of 6.8 (IO); peroxynitrite (O=NOO-) is stable in alkaline solutions (24), but ita conjugate acid, pernitrous acid, rapidly decompcses. It has been proposed (11, 29) that the decomposition of peroxynitrite generates a strong oxidant with reactivity similar to that of the hydroxyl radical. A number of hydroxyl radical scavengers were studied to determine if they inhibit the inactivation of alp1 by peroxynitrite. Table I shows that, of the scavengers tested, thiourea affords complete protection, while benzoate and mannitol do not protect alp1 significantly even at a 50-fold excess with respect to peroxynitrite. On the other hand, methionine, which is not considered a specific scavenger of the hydroxyl radical, completely prevents alp1 inactivation. Similar results are obtained at basic and physiological pH (Table I).
Moreno and Pryor
428 Chem. Res. Toxicol., Vol. 5, No. 3, 1992
Table 111. Amino Acid Composition of MERlO Exposed to Peroxynitrite MERlO + MERlO + residue" MERlO O=NOO- * residue" MERlO O=NOO-b Glu 10.5 f 0.4 10.4 f 0.7 Ile 10.0 f 0.1 10.3 f 0.2 Ser 9.9 f 0.6 8.6 f 0.3 Phe 10.0 f 0.1 9.8 f 0.6 Met 10.4 f 0.7 0.4 f 0.3 Lys 10.5 f 0.5 9.2 f 0.9 Val 9.8 f 0.4 10.3 f 0.2 Pro 29.3 f 3.0 32.5 f 2.1 MetSO' 0.1 f 0.1 8.5 f 0.6
lA
I om0
eooo; n
Residue amounts expressed as mol % f sd of duplicate runs of two independent experiments. Major product isolated from the reaction between MERlO and peroxynitrite (see HPLC section in Experimental Section). Methionine sulfoxide was determined by the cyanogen bromide method (see Experimental Section).
0
2500
1
N N W
0 '
-v
-IN N
, -
, I
h//
lO0Oi
PIjj
-t
5OOj
,
I
-&\ sbo
iH /I
0
RbO
1000
1200
1400 1600 1800
zoo0
