Free Radical Biology & Medicine, Vol. 10, pp. 305-313, 1991 Printed in the USA. All rights reserved.
0891-5849/91 $3.00 + .00 Copyright © 1991 PergamonPress plc
Original Contribution ON RADICAL PRODUCTION BY PMA-STIMULATED NEUTROPHILS MONITORED BY LUMINOL-AMPLIFIED CHEMILUMINESCENCE
AS
AMRAM SAMUNI,* C. MURALI KRISHNA, JOHN COOK, CHRISTOPHER D . V . BLACK, and ANGELO RUSSO Radiation Oncology Branch, Clinical Oncology Program, Division of Cancer Treatment, National Cancer Institute, National Institute of Health, Bethesda, MD 20892; *Molecular Biology, Hebrew University Medical School, Jerusalem, Israel (Received 31 August 1990; Revised and Accepted 28 November 1990)
A b s t r a c t - - T h e means by which neutrophils within the body ward off infectious and neoplastic processes by the activation of molecular oxygen, as well as how such mechanisms dysfunction, is the subject of extensive ongoing research. Most previous studies of neutrophil activation indicate that there is a transient production of reactive oxygen species. Lurninol-amplified chemiluminescence surveillance of 02 and H202 supported these general findings. Yet, recent studies showed that production of reactive oxygen species by PMA-stimulated neutrophils is not transient but persistent; however, luminol-dependent methods do not corroborate such findings. The kinetics of O 2 production by human neutrophils were studied using luminol-amplified chemiluminescence (CL), spin trapping combined with electron spin resonance detection, and ferricytochrome c reduction. The effects of pH and 02 level on luminol-amplified CL were determined using hypoxanthine/xanthine oxidase to produce 02 and H202 in cell-free systems. As we have found by electron spin resonance and ferricytochrome c reduction, stimulated neutrophils continued to generate O2 for several hours, yet when luminol-amplified CL was used to continuously follow radical production, CL was shortly lost. Similar loss of CL was observed with continuous enzymatic formation of 02 and H202. The failure of the CL assay to report 02 and H202 formation results from some luminol reaction product which interferes with the light reaction. Our results show that the cells are operative for long periods indicating that cell exposure to prolonged 02 fluxes does not terminate radical production, and even when pH, [02], and reagents are optimized, the use of luminol-amplified CL is not a valid assay for continuous monitoring of 02 and H202generated by either stimulated neutrophils or in cell-free systems. Keywords--Spin trapping, Superoxide radical, Cytochrome c, Respiratory burst, Polymorphonuclear leukocytes, Electron spin resonance, Hydrogen peroxide, Free radicals
INTRODUCTION
kinetics of the cell-derived O 2 a n d H202 .2-3 The reaction mechanisms leading to luminol-amplified CL have been investigated extensively,4--6 but to date, the exact nature of the light-producing reactions is still not fully understood. Nevertheless, luminol is widely used to probe cellular events. 7-9 Indeed, luminol has been proposed as a tool for routine clinical practice, t° and is considered superior to other luminogenic agents because of its ability to enter the cell and report on intracellular radical production. 8 Potential difficulties intrinsic in the use of luminolamplified CL for studying cell activation have been identified previously:9 i) CL can result from reactions of luminol with H202 catalyzed by peroxidase, tt-t4 and with 02; 15 therefore, discrimination between the various ROS responsible for CL may be o b s c u r e d . 9'16 ii) The time courses of H 2 0 2 production and the release of peroxidative enzymes such as MPO are not necessarily concerted; therefore, CL does not validly reflect a 2 0 2 formation when MPO availability is delayed, or vice versa. ~7 iii) Luminol can enter the neutrophils; conse-
An important mechanism of higher organisms' protection against bacterial and parasitic infections involves the activation of PMN to produce oxygen-derived cytotoxic species, t Cell stimulation triggers degranulation and release of MPO into the extracellular space. This is accompanied by activation of a membrane-bound NADPH oxidase that catalyzes reduction of oxygen to 02, that rapidly dismutates to yield H202, which is utilized by MPO together with halides such as CI- to produce active C10- species. Stimulants such as PMA, N-formylmethionyMeucyl-phenylalanine, or opsonized zymosan have been used to trigger these events in vitro to study the biochemistry of neutrophil-derived ROS, particularly 02 and H202. Because of convenience and exquisite sensitivity, CL enhanced by luminogenic agents such as luminol is commonly used to detect and study formation
Address correspondence to Angelo Russo, Building #10, Room B3-B69, National Institutes of Health, 9000 Rockville, MD 20892. 305
306
A. SAMUNIet
quently, the exact environmental site from which ROS derive may be obfuscated. 16,18 iv) H20 2 may decay both via a light emitting reaction with peroxidase and through a nonradiative process such as with catalase; ~3't4 hence, neither absolute quantitation nor the kinetics of its production can be validly studied. Neutrophil stimuli fall into two broad categories -those which cause a rapid but brief burst of superoxide (e.g., N-formyl-methionyl-leucyl-phenylalanine or arachidonate) and those which cause more prolonged flux of radicals (PMA, zymosan). Nevertheless, in most studies of PMA-stimulated neutrophils, a transient CL emission lasting 5-40 min has been observed. 2'12'13'18 This apparent short time course has been construed to mean that there is only a brief production of RO5.2'8' 11,12,16,18,19 The perception of cellular production of ROS as a transient event was further corroborated by similar observations using other methodologies, 2°-22 and the apparent termination of ROS production ascribed to various physiological reasons, 23 e.g., i) inactivation of the oxidase system by intracellular inhibitor of G proteins; ii) depletion of cellular NADPH; iii) production and accumulation of NADPH oxidase inhibitors; and iv) impairment of neutrophil function caused by products of the respiratory burst. We recently studied stimulated PMN using ESR and found them capable of producing O~ for several hours. 24 On the other hand, cellular production of ROS studied using the luminol-amplified CL assay appears to be transitory. 8"~6'~9 It is possible that the apparent ephemeral time course of ROS formation observed when assaying with CL might result from methodological artifacts. Therefore, to better understand the CL reaction, its use in biochemical studies, and neutrophil research, and the apparent contradiction between results obtained using ESR and CL we reinvestigated the luminol-amplified luminescent reaction and compared it with two other assays of ROS in both cellular and cell-free systems. The results of the present study indicate that a reaction product of luminol accumulates and interferes with the light reaction and ultimately inhibits the CL without stopping actual radical production; hence, luminol-amplified CL does not truly reflect cellular or enzymatically generated oxygen-derived species and cannot be used to study accurately the long-term kinetics of such processes. MATERIALS AND METHODS
Superoxide dismutase, diethylenetriamine pentaacetic acid (DTPA), horseradish peroxidase, xanthine oxidase EC 1.2.3.2 xanthine: oxygen oxidoreductase, grade III from buttermilk, bis-N-methylacridinium nitrate, ferricytochrome c, and 5-amino-2,3-dihydro-l,4-phthala-
al.
zinedione were obtained from Sigma Chemical Co. (St. Louis, MO), 3-amino phthalate hydrochloride was purchased from Eastman, catalase and hypoxanthine were obtained from Calbiochem-Boehringer Co., phorbol-12myristate- 13-acetate from Chemical Dynamics, Inc.; 5,5 dimethyl-l-pyrroline-N-oxide was purchased from Aldrich Chemical Co. (Milwaukee, WI) The DMPO was purified by distillation in the presence of charcoal. Xanthine oxidase was further purified on a G25 sephadex column. Luminol was recrystallized from 0.1M KOH. All other chemicals were prepared and used without further purification. Distilled-deionized water and controlled temperatures were used throughout all experiments. The reaction systems contained 50 IzM DTPA to minimize unwanted transition metal effects.
Polymorphonuclear leukocytes PMN were prepared from the blood of normal volunteers by sedimentation on a ficol/sodium diatriazoate gradient at a density of 1.083, and neutrophils were separated from contaminating erythrocytes by further sedimentation on 3% dextran followed by hypotonic lysis to remove residual red cells. 25 The neutrophils were resuspended in HBSS without Ca 2+ and Mg 2+, counted, and then kept a room or ice temperature until use. PMA was dissolved in ethanol (1 mg/mL) and before use diluted in phosphate-buffered saline.
Myeloperoxidase activity To assess MPO activity of donors, blood smears were stained using the benzidine dihydrochloride method 26 and examined under light microscopy. Blood donated by leukapheresis was routinely stained prior to donation with an automated staining procedure designed not only to reject abnormal leukocytes but also to denote MPOdeficient donors. None of the leukocytes used in these experiments were deficient in MPO activity.
Electron Spin Resonance 100 p~L samples were drawn by a syringe into a gaspermeable, 0.8 mm inner diameter teflon capillary tube, thus enabling the desired oxygen level to be maintained throughout the experiments. Each capillary was inserted into a 2.5 mm ID quartz ESR tube (open at both ends) and then placed horizontally (to minimize settling down of cells out of the sensitive volume of the ESR cavity) with a modified microwave guide for measurement of the ESR spectra while air was flowed around the sample within the spectrometer cavity. ESR spectra were recorded on a Varian E4 (or E9) X-band spectrometer, with field set at 3357 G, modulation frequency of 100 KHz, modulation amplitude of 1 G, and nonsaturating
Superoxide production by neutrophils
307
100% oxygen
10 6 •
20% oxygen 10 5 •
20/. oxygen
LL]
10 4.
10 3.
10 2.
101,
10 o
5
i
I
i
i
|
6
7
8
9
10
pH Fig. 1. Effects of pH and 02 level on luminol-amplified chemiluminescence. The intensity of CL produced at 25°C by 50 p,M luminol + 6.7 U/ rnL HRP in 50 mM aerated phosphate buffer containing 50 p,M DTPA at various pH values upon injection of 0.1 nmole of H202 into 1 mL. The total CL was determined by integrating the areas under the CL peaks.
microwave power. Radical concentrations were standardized against known concentrations of stable nitroxide spin labels.
Ferricytochrome c assay P M N (107 ceUs/mL) were stimulated and maintained at 25°C in either HBSS or RPMI 1640 free of phenol red. At timed intervals, an aliquot of the stimulated cells was added to each of the cuvettes in a temperaturecontrolled SLM-DW-C2 dual beam spectrophotometer, where the solution in the reference cuvette contained 92 units/mL SOD. Both sample and reference cuvettes contained 65 U/mL catalase to minimize possible effects of HzO2. The assay was started by the addition of the stimulated cells to both cuvettes containing 100 p~M Cyt-cm. 27 Rates of 02 production were determined from the initial slopes of the traces using esso = 21 r a M - l c m - 1 and calculated per 106 PMN/min.
H20 e assay YSI model 27 Industrial Analyzer equipped with H 2 0 2 selective electrode from Yellow Spring Instrument
was used for the assay. Aliquots of 25 p,L were sampled from the reaction system at varying time points and injected into the analyzer. I-I202 was determined after the instrument was calibrated with known concentrations of g e n u i n e H 2 0 2.
Chemiluminescence ( CL) assay PMN were stimulated at 25°C in HBSS or phenol red-free RPMI 1640 medium supplemented with 50 p,M DTPA, 50-250 p,M luminol, and 30 mM HEPES buffer,
under continuous supply of the saturating gas, with or without gentle stirring. Following temperature equilibration, the experiments were started by adding PMA. CL emission was measured by photon counting using an SLM 8000 spectrofluorimeter. Gases of desired composition were flushed above or bubbled into the reaction cuvette. For most experiments, stirring of a micro magnetic bar within the reaction mixture was used to ensure equilibration with defined oxygen concentrations. RESULTS
The pH-dependence of luminol-amplified CL PMN stimulation is accompanied by both intra- and extracellular transient pH changes. 28 The buffering capacity, however, of HBSS and many cell media commonly used for in vitro experiments is not always sufficient to maintain the extracellular pH at its initial value. To study the pH dependence of the CL, solutions of 50 p~M luminol and 6.7U/mL HRP in 50 mM phosphate buffer at different pH values were placed in the spectrofluorimeter, known amounts of H20 z (0.1lnmole) were injected, and the CL emitted was recorded and integrated. Under these experimental conditions and in the absence of 02, only H2Oz-derived CL was emitted. The pH-dependence of CL intensity is illustrated in Fig. I showing a 10a-fold increase in CL intensity over pH range of 6-9.
The dependence of luminol-amplified CL on oxygen level In many experimental procedures, particularly when CL is measured by means of scintillation counters, nei-
308
A. SAMtrN[et al.
// _ 4o104 e-
2.10 4 -J
L)
0
0
5
10
15
i
i
25
30
Time (rain)
Fig. 2. Time-dependenceof luminol-amplifiedchemiluminescenceof PMA-stimulated human neutrophils. CL generated by 106 neutrophils/mL stimulated by 1 p.g/mLPMA at 25°C in phenol red-free RPMI 1640 mediumcontaining 50 ixM luminol, 50 p.M DTPA, and 6.7U/mL HRP, and measured using the photon counting mode of an SLM 8000 spectrofluorimeter.
ther desired temperature nor constant [02] can be readily achieved and maintained. Since cells, particularly stimulated PMN, rapidly consume available oxygen, the CL will vary if the quantum yield of the luminol light reaction is sensitive to [02]. To check that possibility, the dependence of H202-derived CL on [02] was studied. The experiments described above were repeated where buffered luminol solutions containing HRP were saturated with gas mixtures of various 02 concentrations. Aliquots of H202 (0.1-1nmole) were injected, and the recorded CL emitted was integrated. The results (Fig. 1) indicate that luminol-amplified CL increases almost 50 fold when going from 2% to 21% oxygen, whereas further increase beyond 21% was practically without effect.
The transient nature of luminol-amplified CL Since both pH and [02] affect the CL intensity, all subsequent experiments were carried out under controlled 0 2 levels, and the media used were supplemented with 30 mM HEPES buffer pH 7.4 to increase the extracellular buffering capacity. Under such conditions the extracellular pH did not change during the experiment. In addition, the temperature was carefully controlled. The time course of luminol-amplified CL emitted by PMA-stimulated neutrophils (106 cells/mL) at 25°C in phenol red-free RPMI 1640 (or HBSS) medium was studied. As reported by others, 8,16'm the CL intensity decayed shortly after the initial burst, despite careful regulation of temperature, [O2], and pH. A similar progressive CL loss was observed in the presence of 6.7 U/mL HRP (Fig. 2) with or without 0.5 mM azide, and also when the cells were stimulated in HBSS (data not shown). The CL loss may reflect cessation of ROS pro-
duction; but it might also stem from depletion of any component essential for the light reaction. Therefore, fresh PMA, HRP, and luminol were separately added following CL decay. Likewise, air was flushed above (or bubbled inside) the stirred reaction mixture. No CL restoration could be achieved in any of these cases.
Chemiluminescence loss in cell-free system To determine whether a similar time-dependent CLloss occurs in a cell-free system, the experiment was repeated using, in place of neutrophils, the HX/XO superoxide-generating reaction. In the absence of peroxidase, with or without catalase, the luminol-amplified CL can only be due to 0 2. With 4 mM HX and 0.03 U/mL XO, the production of O 2 persists for several hours as has been previously shown by spin-trapping 29 and Cytc m reduction (data not shown). Therefore, a steady CL intensity would have been anticipated. However, as seen in Fig. 3A, the luminol-amplified O2-induced CL rapidly decayed, and no CL restoration was achieved upon adding fresh luminol or XO as indicated by arrows in Fig. 3A. Similar CL-loss was observed when catalase (65-1300 U/mL) was included in the reaction mixture to eliminate any possible effect of accumulating H20 2. For comparison, the experiment was repeated with 250 IxM lucigenin instead of luminol. 3° The resulting CL persisted for over 1.5 h and a typical time course is displayed also in Fig. 3A. The luminol-amplified CL induced by H20 2 is far more intense than that originating from 0 2 . 16 To test if luminol-amplified CL validly assays H2Oz production, the HX/XO reaction was carried out in the presence of both 6.7 U/mL HRP and 50 U/mL SOD. Under such experimental conditions, the CL should only reflect
Superoxide productionby neutrophils
309
A r~'~', '" ....
3.10 3
/'~
//
"-'~MI~I~II4j~L~..~. d J._ _ , = u . . ~ / ~ ==, . . . . l . . , . . ~ , u . L...~, _=, • . . . . . . ~. . . . -- -'~" + ' " ' " ' " " m " r~r I " r ' W ~ R ~ ' " ' W " a ' " ~1'"' I r I W P " " n ~ / ~ 4 ~ A I I ~ I ~
t'>.
2.10 3
LUMINOL 50 #M
.IQ
~j
1-103
l 0
0
_.~
, //~,
, 5
10
15
XO SOD 0.0i Utml 10 i/ml ,--,---~,..//
35
90
4O
Time (min)
B 6- 10 4
A ,'- 4 . 1 0 4
.j
2-10
4
+XO
(J
l
f 0 0
5
i 15
10
• i 20
T i m e (min) Fig. 3. Time-dependenceof CL inducedby 02 and H202 formed by hypoxanthine/ xanthineoxidase. CL emitted upon XO-catalyzedoxidationof HX under air, at 25°C, in pH 7.8, 50 mM phosphate containing4 mM hypoxanthine,50 I~M DTPA, and 50 ~M luminol. Reactionswere started by adding 0.03U/mL xanthineoxidase. The reactionsystems contained:a) 50 U/mL catalase with either 250 ixM lucigeninor 50 ~M luminol;additional aliquotes of luminol, XO, SOD were injected at indicatedtime points (arrows). b) 6.7 U/mL HRP + 50 U/mL SOD + 50 p.M luminol. Additional XO was injectedat 16 min. H20 2 flux. As seen in Fig. 3B the CL did not persist. Similar CL-loss was observed in the absence of SOD where both H2Oe and 0 2 contribute to luminol-amplifled CL, as previously observed. 15 Again, additional fresh aliquots of luminol, HRP, or XO did not restore the CL. The arrow drawn in Fig. 3B designates the time point when an additional aliquot of fresh XO was introduced to the reaction mixture. CL-loss may result from a progressive inactivation of XO and cessation of 0 2 and H20 2 formation. To examine this possibility the accumulation of H202 produced by HX/XO in the presence of 50-250 IJ,M luminol in a
constantly aerated solution was assayed. Aliquots (25 IxL) of HX (4 mM) / XO (0.03 U/mL) reaction mixture were sampled at different times and injected into a YSI
NH2
O
NH2
O
~ 0 Lurninol
o" 0 3-amino phthalate
A. SAMtrNret al.
310
AO A ° Ao
:E A0
O nO
•
o • [2.
•
500 AO
0
O
0
O
&O
0
0
0 "O
A0
"I" 0" 0
5'0
1 '0 0
Time (min) Fig. 4. H202 formationby HX/XO in the presence of luminol. Hydrogenperoxide generated upon XO-catalyzedoxidation of HX under air, at 25°C, in pH 7.8 50 mM phosphate buffer containing 5 mM hypoxanthine, 50 ~M DTPA, and 250 IxM luminol, was determinedusing YSI model 27 industrial analyzer equipped with a selective electrode for H20z. Reactions were started by adding 0.03 U/mL xanthine oxidase. Control (o); with luminol (A).
model 27 industrial analyzer equipped with a selective electrode for H20 2. The H20 2 concentrations were determined and found (Fig. 4) to increase linearly throughout the experiment. The rate of H20 2 formation was unaffected by luminol both at the start and after 1.5 h of the reaction, thus excluding the possibility of progressive XO inactivation. This conclusion was further corroborated by spectrophotometrically monitoring the accumulation of uric acid (data not shown). The failure to restore the CL suggests an accumulation of some reaction product that irreversibly terminates the light reaction. It is noteworthy that 3-amino phthalate, the major oxidation product of luminol, 5 had no effect on the CL intensity and, therefore, is not the inhibitory product.
Superoxide production demonstrated by spin-trapping and ESR detection Alternatively, the presumed accumulating reactionproduct might compete for 0 2, thus inhibiting CL. To test the possibility that O~ is removed from the system, spin trapping coupled with ESR detection was used. The reaction mixture (containing 4 mM HX, 130 U/mL catalase, and 250 ~M luminol) was sampled 5 min after the addition of 0.03 U/mL XO, at the CL peak, and again 45 min later when CL had fully decayed. An aliquot of DMPO spin trap was added to each sample (0.1 M final concentration) then was placed within an ESR cavity under a constant air supply and scanned for DMPO spin adducts. Both samples gave rise to the same ESR signals of similar intensities composed of DMPO/OOH (the spin adduct of O~ ) spectrum, with its characteristic hyperfine coupling constants (a N = 14.2 G, a n = 11.3 G, aH~/ = 1.3G) accompanied by a small signal of its decomposition product DMPO/OH with hyperfine coupling constants a ~ = 14.9G; aH = 14.9G) (Fig. 5).
A small amount of the methyl adduct of DMPO, that is, DMPO/CH 3 is also observed with ar~= 16.0 G and all= 23.0G. The methyl adduct arises from DMSO in which luminol had been dissolved and stored. Evidently, the solution which is incapable of producing luminolamplified O2-induced CL, gave rise to a constant O~ flux, as detected by spin trapping, similar to that generated during light emission.
Luminol effect on 02 formation by PMA-stimulated PMN To further ascertain that neither luminol nor its reaction products affect 0 2 production by the cells, the rate of radical formation was monitored using the SOD-inhibitable Cyt-c m reduction assay. To that effect, 5 × 10 6 PMN/mL were incubated in RPMI at 25°C and stimulated by 0.5 p~g/mL PMA. At various time points 100 ~L aliquots from the stimulated cell suspension were placed into 2 mL of RPMI containing 100 ~M Cyt-c m and assayed for O 2 formation rate as detailed in Methods. It is also noteworthy that with this method of sampiing, there is no chance of exhausting the Cyt-c m reporter chromophore within the reaction mixture. Following a few minutes lag, the formation rate of 0 2 increased, peaked, declined to about 25-40% of its maximal value, and persisted for several hours. No essential difference was detected when the cells were stimulated and incubated in the presence of 250 p~M luminol (Fig. 6). This result rules out the possibility that luminol or its reaction product interferes with the enzymatic source of cell-derived 0 2 production. DISCUSSION Although better physiological stimuli are commonly used, PMA is widely employed for studying neutrophil
Superoxide production by neutrophils
311
, lOG
Fig. 5. Spin-trapping and ESR detection of 02 formed by HX/XO in the presence of luminol and its reaction product. ESR spectra obtained upon XO (0.03 U/mL) catalyzed oxidation of 4 mM HX in the presence of 250 p,M luminol in 50 mM phosphate buffer pH 7.8 and 50 I~M DTPA at 25°C, under a constant supply of air. DMPO (0.1M) was added 45 min after the start of the reaction after the luminol-amplified O2-induced chemiluminescence had fully decayed. The ESR signal is composed of the spectra of DMPO/OOH (the spin-adduct of 02) accompanied by a small signal of its decomposition product, DMPO/OH. The carbon centered radical (arrows) results from methyl radical from DMSO adding to DMPO to form DMPO/CH3.
action and, despite known experimental limitations, the luminol-amplified CL assay is a particularly useful method because of its simplicity and sensitivity. This assay has been previously used to study the HX/XO reaction, in which 02 is continuously formed and H 2 0 2 accumulates. 15'31'32 Those studies generally only covered the first minutes of enzyme activity, before there is any significant accumulation of reaction products.31 Yet, it has been previously observed that luminol-amplified CL from HX/XO stops after a few minutes although formation of H202 and uric acid persists. 32 In the presence of SOD and HRP, H202 but not O 2 was responsible for luminol-amplified CL (Fig. 3). Conversely, in
(=0
*
1.0
o~
o
a. •1 o
oo
c
"~ ._o
~E m o
0 0
0.5
v
E 0.0 0
100
200
Time (rain) Fig. 6. The effect of luminol reaction-product on 02 formation by PMA-stimulated neutrophils. The production rate of 02 was determined using the Cyt-cm reduction assay. PMNs were incubated in RPMI at 25°(2 in the presence of 250 ixM luminol and stimulated by 0.5 ~g/mL PMA. At various time points aliquots of 100 IxL were sampled into 2 mL RPMI containing 50 p,M DTPA, 100 tzM Cyt-cm, and 65 U/mL catalase. The reference cuvette contained 90 U/mL SOD.
the presence of catalase only 02 gives rise to the CL. Luminol-amplified CL monitors both H202 (in the presence of peroxidase) and superoxide, the formation rates of H202 and 02 appear to be distinguishable by virtue of the simple addition of either catalase or SOD. Despite reports that question luminol use for biological studies, 7'9'32 luminol-amplified CL has been assumed to reflect the formation rate of ROS such as O 2 and H 2 0 2 and the time course of CL has been widely used to investigate subtle cellular events following neutrophil stimulation. Such an experimental approach is valid provided the effects of pH, [02], temperature, and media composition are carefully controlled and only if the assay does not perturb the system it purports to measure. As seen in Fig. l, luminol-amplified CL is extremely pH dependent, hence, slight pH changes can greatly modify CL and erroneously imply changes in ROS flux. One might predict such a situation during cell stimulation. 2s In the present study, however, care was taken to minimize the above-mentioned potential artifacts. Although the pH of the extracellular milieu can be controlled by the use of buffers, the intracellular environment which is not accessible to the buffer would not be subject to defined control. Such pH independence of the intracellular space has the potential to impact on the amount of CL resulting from the PMN stimulation. However, it would be predicted that since the superoxide anion radical would be produced on the extrafacial surface of the membrane and that the anion radical does not readily cross into the intracellular space, there would not be a large intracellular contribution to CL by superoxide anion radical. This may not be the case for H 2 0 2 , since it can readily partition between the intracellular and extracellular space. If after stimulation of the neutrophil the pH of the intracellular environment were to
312
A. SAMtmlet al.
change dramatically toward a more acidic state, chemiluminescence would be expected to decrease (Fig. 1) even though the flux of H202 would remain the same. However, the intracellular pH upon stimulation of the neutrophil has been reported to increase. 28 As in previous studies, 2'12'13'18'33 the time-course of the CL produced by the cells appeared as a peak (Fig. 2) despite the fact that ROS production continued as demonstrated in Fig. 6. The present results do not agree with reports that 0 2 production by PMN of normal donors, unlike MPO-deficient ones, is undetectable beyond 6-7 min. 34 Findings obtained by ESR support the conclusion that 0 2 generation by stimulated neutrophils of normal donors is not terminated shortly, 23 but rather persists for long durations. 24 The failure of luminol-amplified CL to report on O 2 production is demonstrable in cell-free system. The HX/ XO reaction, which is free from many complications associated with cell preparation, has been selected for that purpose. As seen in Fig. 3, the CL emitted by enzymatically generated ROS shortly stopped, whereas the production of 0 2 (Fig. 5) and of H202 (Fig. 4) continued. Consequently, luminol cannot therefore serve for monitoring ROS produced in the absence of cells. Figure 6 shows that neither luminol nor its reactionproducts affect ROS formation by the cells. Likewise, no effect is evident on 0 2 or H20 2 formation by HX/ XO (Figs. 4 and 5). Instead, the progressive CL decay reflects an artifact resulting from some luminol reaction-product that interferes with the light reaction. The major oxidation product of luminol is the 3-amino phthalate, yet it did not appear to affect CL, indicating that some other product is responsible for the loss of CL. Under our experimental conditions, only a small fraction of the luminol undergoes chemical change, 6 so the presumed reaction product might be very effective in quenching the light. Yet, its chemical structure is not known and is currently under study. In conclusion, the present results prove that i) contrary to previous conclusions, PMA-stimulated neutrophils from normal donors remain active for long periods of time and that continuous high flux of 0 2 does not terminate their capability of producing 02; ii) luminolamplified CL can neither serve to study in a continuous manner the formation kinetics nor be used as a valid continuous assay of ROS such as 02 and H20 2 generated by stimulated neutrophils; iii) luminol speciously report on H20 2 or 0 2 production even in a cell-free system such as HX/XO. REFERENCES 1. Babior B. M. The respiratory burst of phagocytes. J. Clin. Invest. 73"599-601; 1984. 2. Alien, R. C.; Mead, M. E.; Kelly, J. L. Phagocyte oxygenation activity measured by chemiluminescence and chemiluminegenic
3.
4.
5. 6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17. 18. 19.
20. 21. 22. 23.
probing. In Greenwald, R.A., ed. CRC handbook of methods for oxygen radical research. Boca Raton, FL: CRC Press, Inc. 1985: 343-351. Mills, E. L.; Rholl, K. S.; Quie, P. G. Luminol-amplified chemiluminescence: a sensitive method for detecting the carder state in chronic granulomatous disease. J. Clin. Microbiol. 12: 52-56; 1980. Cormier, M. J.; Prichard, P. M. An investigation of the mechanism of the luminescent peroxidation of luminol by stopped-flow technique. J. Biol. Chem. 243:4706--4714; 1968. Lind, J.; Merenyi, G.; Eriksen, T. E. Chemiluminescence mechanism of cyclic hydrazides such as luminol in aqueous solutions. J. Am. Chem. Soc. 105:7655-7661; 1983. Merenyi, C.; Lind, J.; Eriksen, T. E. The reactivity of superoxide (02) and its ability to induce chemiluminescence with luminol. Photochem. Photobiol. 41:203-208; 1985. Falck, P. Investigation of the chemiluminescence response of human neutrophils and mononuclear cells. Folia biologica (Praha)32:103-115; 1986. Dahlgren, C. Polymorphonuclear leukocytes chemiluminescence induced by formylmethionyl-leucyl-phenylalanine and phorbol myristate acetate: effects of catalase and superoxide dismutase. Agents and Actions 21:104-112; 1987. Vilim, V.; Wilhelm, J. What do we measure by a luminol-dependent chemiluminescence of phagocytes? Free Radic. Biol. Med. 6:623-629; 1989. Bruchelt, G.; Schmidt, K. H. Comparative studies on the oxidative processes during phagocytosis measured by luminol-dependent chemiluminescence. J. Clin. Chem. Clin. Biochem. 22: 1-13; 1984. Aniansson, H.; Stendahl, O.; Dahlgren, C. Comparison between luminol- and lucigenin-dependent chemiluminescence of polymorphonuclear leukocytes. Acta Path. Microbiol. Immunol. Scand. Sect. C 92:357-361; 1984. Dahlgren, C.; Stendahl, O. Role of myeloperoxidase in luminoldependent chemiluminescence of polymorphonuclear leukocytes. Infection and Immunity 39:736-741; 1983. Wymann, M. P.; von Tscharner, V.; Deranleau, D. A.; Baggiolini, M. Chemiluminescence detection of H202 produced by human neutrophils during the respiratory burst. Anal. Biochem. 65: 371-378; 1987. Lock, R.; Johansson, A.; Orselius, K.; Dahlgren, C. Analysis of horseradish peroxidase-amplified chemiluminescence produced by human neutrophils reveals a role for the superoxide anion in the light emitting reaction. Anal. Biochem. 173:450--455; 1988. Misra, H. P.; Squatrito, P. M. The role of superoxide anion in peroxidase-catalyzed chemiluminescence of luminol. Arch. Biochem. Biophys. 215:59-65; 1982. Dahlgren, C, Aniansson, H.; Magnusson, K.-E. Pattern of formylmethionyl-leucyl-phenylalanine-induced luminol- and lucigenin-dependent chemiluminescence in human neutrophils. Infection and Immunity 47:326-328; 1985. Dahlgren, C. Is lysosomal fusion required for the granulocyte chemiluminescence reaction? Free Radic. Biol. Med. 6:399--403; 1989. Bender, J. G. ; Van Epps, D. E. Analysis of the bimodal chemiluminescence pattern stimulated in human neutrophils by chemotactic factors. Infection and Immunity 41:1062-1070; 1983. Gyllenhammar, H. Mechanisms for luminol-augmented chemiluminescence from neutrophils induced by leukotriene B,, and Nformyl-methionyl-leucyl-phenylalanine. Photochem. Photobiol. 49:217-223; 1989. Test, S. T.; Weiss, S. J. Quantitative and temporal characterization of the extracellular H202 pool generated by human neutrophils. J. Biol. Chem. 259:399-405; 1984. Kleinhans, F. W.; Barefoot, S. T. Spin trap determination of free radical burst kinetics in stimulated neutrophils. J. Biol. Chem. 262"12452-12457; 1987. Holman, R. G.; Maier, R. V. Superoxide production by neutrophils in a model of adult respiratory distress syndrome. Arch. Surg. 123"1491-1495; 1988. Whitin, J. C.; Cohen, H. J. Disorders of respiratory burst termination. Hematol. Oncol. Clin. North Am. 2:289-299; 1988.
Superoxide production by neutrophils 24. Samuni, A.; Black, C. D. V.; Krishna, C. M.; Malech, H, L.; Bernstein, E. F.; Russo, A. Hydroxyl radical production by stimulated neutrophils reappraised. J. Biol. Chem. 263:13797-13801; 1988. 25. Boyum, A. Isolation of mononuclear cells and granulocytes from human blood. Scan. J. Clin. Lab. lnvest. 21 (Suppl. 97):77-89; 1968. 26. Kaplow, L. S. Simplified myeloperoxide stain using benzidine dihydrochloride. Blood 26:215-219; 1965. 27. Fridovich, I. Quantitative aspects of the production of superoxide anion radical by milk xanthine oxidase. J. Biol. Chem. 245: 4053-4057; 1970. 28. Simchowitz, L.; Cragoe, E. J., Jr. Regulation of human neutrophil chemotaxis by intracellular pH. J. Biol. Chem. 261: 6492--6500; 1986. 29. Samuni, A.; ICrishna, C. M.; Riesz, P.; Finkelstein, E.; Russo, A. Superoxide reaction with nitroxide spin-adducts. Free Radic. Biol. Med. 6:141-148; 1989. 30. Muller-Peddinghaus, R. Comparison of luminol- and lucigeninamplified chemiluminescence (CL) of phagocytes. In: Bors, W.; Saran, M.; Tait, D., eds. Oxygen radicals in chemistry and biology. Berlin: Walter de Gruyter & Co.; 1984: 881-882. 31. Hodgson, E. K.; Fridovich, I. The role of 02 in the chemiluminescence of luminol. Photochem. Photobiol. 18:451-455; 1973. 32. Wilhelm, J.; Vilim, V. Variables in xanthine oxidase-initiated luminol chemiluminescence implications for chemiluminescence measurements in biological systems. Anal. Biochem. 158: 201-210; 1986. 33. Dahlgren, C.; Lock, R. The limitation of the human neutrophil chemiluminescence response by extracellular peroxidase is stimulus dependent: effect of added horseradish peroxidase on the re-
313
sponse induced by both soluble and particulate stimuli. J. Clin. Lab. Immunol. 26:49-53; 1988.
34. Edwards, S. W.; Swan, T. F. Regulation of superoxide generation by myeloperoxidase during the respiratory burst of human neutrophils. Biochem. J. 237:601-604; 1986.
ABBREVIATIONS
CL-- chemiluminescence Cyt-cm-- ferricytochrome c DMPO--5,5-dimethyl-l-pyrroline-N-oxide DTPA--diethylenetriamine pentaacetic acid ESR--electron spin resonance HBSS--Hank's balanced salt solution HRP--horseradish peroxidase HX--hypoxanthine lucigenin--bis-N-methylacridinium nitrate luminol--5-amino-2,3-dihydro-l,4-phthalazinedione MPO-- myeloperoxidase PMA-- phorbol- 12-myristate- 13-acetate PMN-- polymorphonuclear leukocytes ROS--reactive oxygen species SOD-- superoxide dismutase XO--xanthine oxidase
0891-5849/91 $3.00 + .00 Copyright © 1991 PergamonPress plc
Original Contribution ON RADICAL PRODUCTION BY PMA-STIMULATED NEUTROPHILS MONITORED BY LUMINOL-AMPLIFIED CHEMILUMINESCENCE
AS
AMRAM SAMUNI,* C. MURALI KRISHNA, JOHN COOK, CHRISTOPHER D . V . BLACK, and ANGELO RUSSO Radiation Oncology Branch, Clinical Oncology Program, Division of Cancer Treatment, National Cancer Institute, National Institute of Health, Bethesda, MD 20892; *Molecular Biology, Hebrew University Medical School, Jerusalem, Israel (Received 31 August 1990; Revised and Accepted 28 November 1990)
A b s t r a c t - - T h e means by which neutrophils within the body ward off infectious and neoplastic processes by the activation of molecular oxygen, as well as how such mechanisms dysfunction, is the subject of extensive ongoing research. Most previous studies of neutrophil activation indicate that there is a transient production of reactive oxygen species. Lurninol-amplified chemiluminescence surveillance of 02 and H202 supported these general findings. Yet, recent studies showed that production of reactive oxygen species by PMA-stimulated neutrophils is not transient but persistent; however, luminol-dependent methods do not corroborate such findings. The kinetics of O 2 production by human neutrophils were studied using luminol-amplified chemiluminescence (CL), spin trapping combined with electron spin resonance detection, and ferricytochrome c reduction. The effects of pH and 02 level on luminol-amplified CL were determined using hypoxanthine/xanthine oxidase to produce 02 and H202 in cell-free systems. As we have found by electron spin resonance and ferricytochrome c reduction, stimulated neutrophils continued to generate O2 for several hours, yet when luminol-amplified CL was used to continuously follow radical production, CL was shortly lost. Similar loss of CL was observed with continuous enzymatic formation of 02 and H202. The failure of the CL assay to report 02 and H202 formation results from some luminol reaction product which interferes with the light reaction. Our results show that the cells are operative for long periods indicating that cell exposure to prolonged 02 fluxes does not terminate radical production, and even when pH, [02], and reagents are optimized, the use of luminol-amplified CL is not a valid assay for continuous monitoring of 02 and H202generated by either stimulated neutrophils or in cell-free systems. Keywords--Spin trapping, Superoxide radical, Cytochrome c, Respiratory burst, Polymorphonuclear leukocytes, Electron spin resonance, Hydrogen peroxide, Free radicals
INTRODUCTION
kinetics of the cell-derived O 2 a n d H202 .2-3 The reaction mechanisms leading to luminol-amplified CL have been investigated extensively,4--6 but to date, the exact nature of the light-producing reactions is still not fully understood. Nevertheless, luminol is widely used to probe cellular events. 7-9 Indeed, luminol has been proposed as a tool for routine clinical practice, t° and is considered superior to other luminogenic agents because of its ability to enter the cell and report on intracellular radical production. 8 Potential difficulties intrinsic in the use of luminolamplified CL for studying cell activation have been identified previously:9 i) CL can result from reactions of luminol with H202 catalyzed by peroxidase, tt-t4 and with 02; 15 therefore, discrimination between the various ROS responsible for CL may be o b s c u r e d . 9'16 ii) The time courses of H 2 0 2 production and the release of peroxidative enzymes such as MPO are not necessarily concerted; therefore, CL does not validly reflect a 2 0 2 formation when MPO availability is delayed, or vice versa. ~7 iii) Luminol can enter the neutrophils; conse-
An important mechanism of higher organisms' protection against bacterial and parasitic infections involves the activation of PMN to produce oxygen-derived cytotoxic species, t Cell stimulation triggers degranulation and release of MPO into the extracellular space. This is accompanied by activation of a membrane-bound NADPH oxidase that catalyzes reduction of oxygen to 02, that rapidly dismutates to yield H202, which is utilized by MPO together with halides such as CI- to produce active C10- species. Stimulants such as PMA, N-formylmethionyMeucyl-phenylalanine, or opsonized zymosan have been used to trigger these events in vitro to study the biochemistry of neutrophil-derived ROS, particularly 02 and H202. Because of convenience and exquisite sensitivity, CL enhanced by luminogenic agents such as luminol is commonly used to detect and study formation
Address correspondence to Angelo Russo, Building #10, Room B3-B69, National Institutes of Health, 9000 Rockville, MD 20892. 305
306
A. SAMUNIet
quently, the exact environmental site from which ROS derive may be obfuscated. 16,18 iv) H20 2 may decay both via a light emitting reaction with peroxidase and through a nonradiative process such as with catalase; ~3't4 hence, neither absolute quantitation nor the kinetics of its production can be validly studied. Neutrophil stimuli fall into two broad categories -those which cause a rapid but brief burst of superoxide (e.g., N-formyl-methionyl-leucyl-phenylalanine or arachidonate) and those which cause more prolonged flux of radicals (PMA, zymosan). Nevertheless, in most studies of PMA-stimulated neutrophils, a transient CL emission lasting 5-40 min has been observed. 2'12'13'18 This apparent short time course has been construed to mean that there is only a brief production of RO5.2'8' 11,12,16,18,19 The perception of cellular production of ROS as a transient event was further corroborated by similar observations using other methodologies, 2°-22 and the apparent termination of ROS production ascribed to various physiological reasons, 23 e.g., i) inactivation of the oxidase system by intracellular inhibitor of G proteins; ii) depletion of cellular NADPH; iii) production and accumulation of NADPH oxidase inhibitors; and iv) impairment of neutrophil function caused by products of the respiratory burst. We recently studied stimulated PMN using ESR and found them capable of producing O~ for several hours. 24 On the other hand, cellular production of ROS studied using the luminol-amplified CL assay appears to be transitory. 8"~6'~9 It is possible that the apparent ephemeral time course of ROS formation observed when assaying with CL might result from methodological artifacts. Therefore, to better understand the CL reaction, its use in biochemical studies, and neutrophil research, and the apparent contradiction between results obtained using ESR and CL we reinvestigated the luminol-amplified luminescent reaction and compared it with two other assays of ROS in both cellular and cell-free systems. The results of the present study indicate that a reaction product of luminol accumulates and interferes with the light reaction and ultimately inhibits the CL without stopping actual radical production; hence, luminol-amplified CL does not truly reflect cellular or enzymatically generated oxygen-derived species and cannot be used to study accurately the long-term kinetics of such processes. MATERIALS AND METHODS
Superoxide dismutase, diethylenetriamine pentaacetic acid (DTPA), horseradish peroxidase, xanthine oxidase EC 1.2.3.2 xanthine: oxygen oxidoreductase, grade III from buttermilk, bis-N-methylacridinium nitrate, ferricytochrome c, and 5-amino-2,3-dihydro-l,4-phthala-
al.
zinedione were obtained from Sigma Chemical Co. (St. Louis, MO), 3-amino phthalate hydrochloride was purchased from Eastman, catalase and hypoxanthine were obtained from Calbiochem-Boehringer Co., phorbol-12myristate- 13-acetate from Chemical Dynamics, Inc.; 5,5 dimethyl-l-pyrroline-N-oxide was purchased from Aldrich Chemical Co. (Milwaukee, WI) The DMPO was purified by distillation in the presence of charcoal. Xanthine oxidase was further purified on a G25 sephadex column. Luminol was recrystallized from 0.1M KOH. All other chemicals were prepared and used without further purification. Distilled-deionized water and controlled temperatures were used throughout all experiments. The reaction systems contained 50 IzM DTPA to minimize unwanted transition metal effects.
Polymorphonuclear leukocytes PMN were prepared from the blood of normal volunteers by sedimentation on a ficol/sodium diatriazoate gradient at a density of 1.083, and neutrophils were separated from contaminating erythrocytes by further sedimentation on 3% dextran followed by hypotonic lysis to remove residual red cells. 25 The neutrophils were resuspended in HBSS without Ca 2+ and Mg 2+, counted, and then kept a room or ice temperature until use. PMA was dissolved in ethanol (1 mg/mL) and before use diluted in phosphate-buffered saline.
Myeloperoxidase activity To assess MPO activity of donors, blood smears were stained using the benzidine dihydrochloride method 26 and examined under light microscopy. Blood donated by leukapheresis was routinely stained prior to donation with an automated staining procedure designed not only to reject abnormal leukocytes but also to denote MPOdeficient donors. None of the leukocytes used in these experiments were deficient in MPO activity.
Electron Spin Resonance 100 p~L samples were drawn by a syringe into a gaspermeable, 0.8 mm inner diameter teflon capillary tube, thus enabling the desired oxygen level to be maintained throughout the experiments. Each capillary was inserted into a 2.5 mm ID quartz ESR tube (open at both ends) and then placed horizontally (to minimize settling down of cells out of the sensitive volume of the ESR cavity) with a modified microwave guide for measurement of the ESR spectra while air was flowed around the sample within the spectrometer cavity. ESR spectra were recorded on a Varian E4 (or E9) X-band spectrometer, with field set at 3357 G, modulation frequency of 100 KHz, modulation amplitude of 1 G, and nonsaturating
Superoxide production by neutrophils
307
100% oxygen
10 6 •
20% oxygen 10 5 •
20/. oxygen
LL]
10 4.
10 3.
10 2.
101,
10 o
5
i
I
i
i
|
6
7
8
9
10
pH Fig. 1. Effects of pH and 02 level on luminol-amplified chemiluminescence. The intensity of CL produced at 25°C by 50 p,M luminol + 6.7 U/ rnL HRP in 50 mM aerated phosphate buffer containing 50 p,M DTPA at various pH values upon injection of 0.1 nmole of H202 into 1 mL. The total CL was determined by integrating the areas under the CL peaks.
microwave power. Radical concentrations were standardized against known concentrations of stable nitroxide spin labels.
Ferricytochrome c assay P M N (107 ceUs/mL) were stimulated and maintained at 25°C in either HBSS or RPMI 1640 free of phenol red. At timed intervals, an aliquot of the stimulated cells was added to each of the cuvettes in a temperaturecontrolled SLM-DW-C2 dual beam spectrophotometer, where the solution in the reference cuvette contained 92 units/mL SOD. Both sample and reference cuvettes contained 65 U/mL catalase to minimize possible effects of HzO2. The assay was started by the addition of the stimulated cells to both cuvettes containing 100 p~M Cyt-cm. 27 Rates of 02 production were determined from the initial slopes of the traces using esso = 21 r a M - l c m - 1 and calculated per 106 PMN/min.
H20 e assay YSI model 27 Industrial Analyzer equipped with H 2 0 2 selective electrode from Yellow Spring Instrument
was used for the assay. Aliquots of 25 p,L were sampled from the reaction system at varying time points and injected into the analyzer. I-I202 was determined after the instrument was calibrated with known concentrations of g e n u i n e H 2 0 2.
Chemiluminescence ( CL) assay PMN were stimulated at 25°C in HBSS or phenol red-free RPMI 1640 medium supplemented with 50 p,M DTPA, 50-250 p,M luminol, and 30 mM HEPES buffer,
under continuous supply of the saturating gas, with or without gentle stirring. Following temperature equilibration, the experiments were started by adding PMA. CL emission was measured by photon counting using an SLM 8000 spectrofluorimeter. Gases of desired composition were flushed above or bubbled into the reaction cuvette. For most experiments, stirring of a micro magnetic bar within the reaction mixture was used to ensure equilibration with defined oxygen concentrations. RESULTS
The pH-dependence of luminol-amplified CL PMN stimulation is accompanied by both intra- and extracellular transient pH changes. 28 The buffering capacity, however, of HBSS and many cell media commonly used for in vitro experiments is not always sufficient to maintain the extracellular pH at its initial value. To study the pH dependence of the CL, solutions of 50 p~M luminol and 6.7U/mL HRP in 50 mM phosphate buffer at different pH values were placed in the spectrofluorimeter, known amounts of H20 z (0.1lnmole) were injected, and the CL emitted was recorded and integrated. Under these experimental conditions and in the absence of 02, only H2Oz-derived CL was emitted. The pH-dependence of CL intensity is illustrated in Fig. I showing a 10a-fold increase in CL intensity over pH range of 6-9.
The dependence of luminol-amplified CL on oxygen level In many experimental procedures, particularly when CL is measured by means of scintillation counters, nei-
308
A. SAMtrN[et al.
// _ 4o104 e-
2.10 4 -J
L)
0
0
5
10
15
i
i
25
30
Time (rain)
Fig. 2. Time-dependenceof luminol-amplifiedchemiluminescenceof PMA-stimulated human neutrophils. CL generated by 106 neutrophils/mL stimulated by 1 p.g/mLPMA at 25°C in phenol red-free RPMI 1640 mediumcontaining 50 ixM luminol, 50 p.M DTPA, and 6.7U/mL HRP, and measured using the photon counting mode of an SLM 8000 spectrofluorimeter.
ther desired temperature nor constant [02] can be readily achieved and maintained. Since cells, particularly stimulated PMN, rapidly consume available oxygen, the CL will vary if the quantum yield of the luminol light reaction is sensitive to [02]. To check that possibility, the dependence of H202-derived CL on [02] was studied. The experiments described above were repeated where buffered luminol solutions containing HRP were saturated with gas mixtures of various 02 concentrations. Aliquots of H202 (0.1-1nmole) were injected, and the recorded CL emitted was integrated. The results (Fig. 1) indicate that luminol-amplified CL increases almost 50 fold when going from 2% to 21% oxygen, whereas further increase beyond 21% was practically without effect.
The transient nature of luminol-amplified CL Since both pH and [02] affect the CL intensity, all subsequent experiments were carried out under controlled 0 2 levels, and the media used were supplemented with 30 mM HEPES buffer pH 7.4 to increase the extracellular buffering capacity. Under such conditions the extracellular pH did not change during the experiment. In addition, the temperature was carefully controlled. The time course of luminol-amplified CL emitted by PMA-stimulated neutrophils (106 cells/mL) at 25°C in phenol red-free RPMI 1640 (or HBSS) medium was studied. As reported by others, 8,16'm the CL intensity decayed shortly after the initial burst, despite careful regulation of temperature, [O2], and pH. A similar progressive CL loss was observed in the presence of 6.7 U/mL HRP (Fig. 2) with or without 0.5 mM azide, and also when the cells were stimulated in HBSS (data not shown). The CL loss may reflect cessation of ROS pro-
duction; but it might also stem from depletion of any component essential for the light reaction. Therefore, fresh PMA, HRP, and luminol were separately added following CL decay. Likewise, air was flushed above (or bubbled inside) the stirred reaction mixture. No CL restoration could be achieved in any of these cases.
Chemiluminescence loss in cell-free system To determine whether a similar time-dependent CLloss occurs in a cell-free system, the experiment was repeated using, in place of neutrophils, the HX/XO superoxide-generating reaction. In the absence of peroxidase, with or without catalase, the luminol-amplified CL can only be due to 0 2. With 4 mM HX and 0.03 U/mL XO, the production of O 2 persists for several hours as has been previously shown by spin-trapping 29 and Cytc m reduction (data not shown). Therefore, a steady CL intensity would have been anticipated. However, as seen in Fig. 3A, the luminol-amplified O2-induced CL rapidly decayed, and no CL restoration was achieved upon adding fresh luminol or XO as indicated by arrows in Fig. 3A. Similar CL-loss was observed when catalase (65-1300 U/mL) was included in the reaction mixture to eliminate any possible effect of accumulating H20 2. For comparison, the experiment was repeated with 250 IxM lucigenin instead of luminol. 3° The resulting CL persisted for over 1.5 h and a typical time course is displayed also in Fig. 3A. The luminol-amplified CL induced by H20 2 is far more intense than that originating from 0 2 . 16 To test if luminol-amplified CL validly assays H2Oz production, the HX/XO reaction was carried out in the presence of both 6.7 U/mL HRP and 50 U/mL SOD. Under such experimental conditions, the CL should only reflect
Superoxide productionby neutrophils
309
A r~'~', '" ....
3.10 3
/'~
//
"-'~MI~I~II4j~L~..~. d J._ _ , = u . . ~ / ~ ==, . . . . l . . , . . ~ , u . L...~, _=, • . . . . . . ~. . . . -- -'~" + ' " ' " ' " " m " r~r I " r ' W ~ R ~ ' " ' W " a ' " ~1'"' I r I W P " " n ~ / ~ 4 ~ A I I ~ I ~
t'>.
2.10 3
LUMINOL 50 #M
.IQ
~j
1-103
l 0
0
_.~
, //~,
, 5
10
15
XO SOD 0.0i Utml 10 i/ml ,--,---~,..//
35
90
4O
Time (min)
B 6- 10 4
A ,'- 4 . 1 0 4
.j
2-10
4
+XO
(J
l
f 0 0
5
i 15
10
• i 20
T i m e (min) Fig. 3. Time-dependenceof CL inducedby 02 and H202 formed by hypoxanthine/ xanthineoxidase. CL emitted upon XO-catalyzedoxidationof HX under air, at 25°C, in pH 7.8, 50 mM phosphate containing4 mM hypoxanthine,50 I~M DTPA, and 50 ~M luminol. Reactionswere started by adding 0.03U/mL xanthineoxidase. The reactionsystems contained:a) 50 U/mL catalase with either 250 ixM lucigeninor 50 ~M luminol;additional aliquotes of luminol, XO, SOD were injected at indicatedtime points (arrows). b) 6.7 U/mL HRP + 50 U/mL SOD + 50 p.M luminol. Additional XO was injectedat 16 min. H20 2 flux. As seen in Fig. 3B the CL did not persist. Similar CL-loss was observed in the absence of SOD where both H2Oe and 0 2 contribute to luminol-amplifled CL, as previously observed. 15 Again, additional fresh aliquots of luminol, HRP, or XO did not restore the CL. The arrow drawn in Fig. 3B designates the time point when an additional aliquot of fresh XO was introduced to the reaction mixture. CL-loss may result from a progressive inactivation of XO and cessation of 0 2 and H20 2 formation. To examine this possibility the accumulation of H202 produced by HX/XO in the presence of 50-250 IJ,M luminol in a
constantly aerated solution was assayed. Aliquots (25 IxL) of HX (4 mM) / XO (0.03 U/mL) reaction mixture were sampled at different times and injected into a YSI
NH2
O
NH2
O
~ 0 Lurninol
o" 0 3-amino phthalate
A. SAMtrNret al.
310
AO A ° Ao
:E A0
O nO
•
o • [2.
•
500 AO
0
O
0
O
&O
0
0
0 "O
A0
"I" 0" 0
5'0
1 '0 0
Time (min) Fig. 4. H202 formationby HX/XO in the presence of luminol. Hydrogenperoxide generated upon XO-catalyzedoxidation of HX under air, at 25°C, in pH 7.8 50 mM phosphate buffer containing 5 mM hypoxanthine, 50 ~M DTPA, and 250 IxM luminol, was determinedusing YSI model 27 industrial analyzer equipped with a selective electrode for H20z. Reactions were started by adding 0.03 U/mL xanthine oxidase. Control (o); with luminol (A).
model 27 industrial analyzer equipped with a selective electrode for H20 2. The H20 2 concentrations were determined and found (Fig. 4) to increase linearly throughout the experiment. The rate of H20 2 formation was unaffected by luminol both at the start and after 1.5 h of the reaction, thus excluding the possibility of progressive XO inactivation. This conclusion was further corroborated by spectrophotometrically monitoring the accumulation of uric acid (data not shown). The failure to restore the CL suggests an accumulation of some reaction product that irreversibly terminates the light reaction. It is noteworthy that 3-amino phthalate, the major oxidation product of luminol, 5 had no effect on the CL intensity and, therefore, is not the inhibitory product.
Superoxide production demonstrated by spin-trapping and ESR detection Alternatively, the presumed accumulating reactionproduct might compete for 0 2, thus inhibiting CL. To test the possibility that O~ is removed from the system, spin trapping coupled with ESR detection was used. The reaction mixture (containing 4 mM HX, 130 U/mL catalase, and 250 ~M luminol) was sampled 5 min after the addition of 0.03 U/mL XO, at the CL peak, and again 45 min later when CL had fully decayed. An aliquot of DMPO spin trap was added to each sample (0.1 M final concentration) then was placed within an ESR cavity under a constant air supply and scanned for DMPO spin adducts. Both samples gave rise to the same ESR signals of similar intensities composed of DMPO/OOH (the spin adduct of O~ ) spectrum, with its characteristic hyperfine coupling constants (a N = 14.2 G, a n = 11.3 G, aH~/ = 1.3G) accompanied by a small signal of its decomposition product DMPO/OH with hyperfine coupling constants a ~ = 14.9G; aH = 14.9G) (Fig. 5).
A small amount of the methyl adduct of DMPO, that is, DMPO/CH 3 is also observed with ar~= 16.0 G and all= 23.0G. The methyl adduct arises from DMSO in which luminol had been dissolved and stored. Evidently, the solution which is incapable of producing luminolamplified O2-induced CL, gave rise to a constant O~ flux, as detected by spin trapping, similar to that generated during light emission.
Luminol effect on 02 formation by PMA-stimulated PMN To further ascertain that neither luminol nor its reaction products affect 0 2 production by the cells, the rate of radical formation was monitored using the SOD-inhibitable Cyt-c m reduction assay. To that effect, 5 × 10 6 PMN/mL were incubated in RPMI at 25°C and stimulated by 0.5 p~g/mL PMA. At various time points 100 ~L aliquots from the stimulated cell suspension were placed into 2 mL of RPMI containing 100 ~M Cyt-c m and assayed for O 2 formation rate as detailed in Methods. It is also noteworthy that with this method of sampiing, there is no chance of exhausting the Cyt-c m reporter chromophore within the reaction mixture. Following a few minutes lag, the formation rate of 0 2 increased, peaked, declined to about 25-40% of its maximal value, and persisted for several hours. No essential difference was detected when the cells were stimulated and incubated in the presence of 250 p~M luminol (Fig. 6). This result rules out the possibility that luminol or its reaction product interferes with the enzymatic source of cell-derived 0 2 production. DISCUSSION Although better physiological stimuli are commonly used, PMA is widely employed for studying neutrophil
Superoxide production by neutrophils
311
, lOG
Fig. 5. Spin-trapping and ESR detection of 02 formed by HX/XO in the presence of luminol and its reaction product. ESR spectra obtained upon XO (0.03 U/mL) catalyzed oxidation of 4 mM HX in the presence of 250 p,M luminol in 50 mM phosphate buffer pH 7.8 and 50 I~M DTPA at 25°C, under a constant supply of air. DMPO (0.1M) was added 45 min after the start of the reaction after the luminol-amplified O2-induced chemiluminescence had fully decayed. The ESR signal is composed of the spectra of DMPO/OOH (the spin-adduct of 02) accompanied by a small signal of its decomposition product, DMPO/OH. The carbon centered radical (arrows) results from methyl radical from DMSO adding to DMPO to form DMPO/CH3.
action and, despite known experimental limitations, the luminol-amplified CL assay is a particularly useful method because of its simplicity and sensitivity. This assay has been previously used to study the HX/XO reaction, in which 02 is continuously formed and H 2 0 2 accumulates. 15'31'32 Those studies generally only covered the first minutes of enzyme activity, before there is any significant accumulation of reaction products.31 Yet, it has been previously observed that luminol-amplified CL from HX/XO stops after a few minutes although formation of H202 and uric acid persists. 32 In the presence of SOD and HRP, H202 but not O 2 was responsible for luminol-amplified CL (Fig. 3). Conversely, in
(=0
*
1.0
o~
o
a. •1 o
oo
c
"~ ._o
~E m o
0 0
0.5
v
E 0.0 0
100
200
Time (rain) Fig. 6. The effect of luminol reaction-product on 02 formation by PMA-stimulated neutrophils. The production rate of 02 was determined using the Cyt-cm reduction assay. PMNs were incubated in RPMI at 25°(2 in the presence of 250 ixM luminol and stimulated by 0.5 ~g/mL PMA. At various time points aliquots of 100 IxL were sampled into 2 mL RPMI containing 50 p,M DTPA, 100 tzM Cyt-cm, and 65 U/mL catalase. The reference cuvette contained 90 U/mL SOD.
the presence of catalase only 02 gives rise to the CL. Luminol-amplified CL monitors both H202 (in the presence of peroxidase) and superoxide, the formation rates of H202 and 02 appear to be distinguishable by virtue of the simple addition of either catalase or SOD. Despite reports that question luminol use for biological studies, 7'9'32 luminol-amplified CL has been assumed to reflect the formation rate of ROS such as O 2 and H 2 0 2 and the time course of CL has been widely used to investigate subtle cellular events following neutrophil stimulation. Such an experimental approach is valid provided the effects of pH, [02], temperature, and media composition are carefully controlled and only if the assay does not perturb the system it purports to measure. As seen in Fig. l, luminol-amplified CL is extremely pH dependent, hence, slight pH changes can greatly modify CL and erroneously imply changes in ROS flux. One might predict such a situation during cell stimulation. 2s In the present study, however, care was taken to minimize the above-mentioned potential artifacts. Although the pH of the extracellular milieu can be controlled by the use of buffers, the intracellular environment which is not accessible to the buffer would not be subject to defined control. Such pH independence of the intracellular space has the potential to impact on the amount of CL resulting from the PMN stimulation. However, it would be predicted that since the superoxide anion radical would be produced on the extrafacial surface of the membrane and that the anion radical does not readily cross into the intracellular space, there would not be a large intracellular contribution to CL by superoxide anion radical. This may not be the case for H 2 0 2 , since it can readily partition between the intracellular and extracellular space. If after stimulation of the neutrophil the pH of the intracellular environment were to
312
A. SAMtmlet al.
change dramatically toward a more acidic state, chemiluminescence would be expected to decrease (Fig. 1) even though the flux of H202 would remain the same. However, the intracellular pH upon stimulation of the neutrophil has been reported to increase. 28 As in previous studies, 2'12'13'18'33 the time-course of the CL produced by the cells appeared as a peak (Fig. 2) despite the fact that ROS production continued as demonstrated in Fig. 6. The present results do not agree with reports that 0 2 production by PMN of normal donors, unlike MPO-deficient ones, is undetectable beyond 6-7 min. 34 Findings obtained by ESR support the conclusion that 0 2 generation by stimulated neutrophils of normal donors is not terminated shortly, 23 but rather persists for long durations. 24 The failure of luminol-amplified CL to report on O 2 production is demonstrable in cell-free system. The HX/ XO reaction, which is free from many complications associated with cell preparation, has been selected for that purpose. As seen in Fig. 3, the CL emitted by enzymatically generated ROS shortly stopped, whereas the production of 0 2 (Fig. 5) and of H202 (Fig. 4) continued. Consequently, luminol cannot therefore serve for monitoring ROS produced in the absence of cells. Figure 6 shows that neither luminol nor its reactionproducts affect ROS formation by the cells. Likewise, no effect is evident on 0 2 or H20 2 formation by HX/ XO (Figs. 4 and 5). Instead, the progressive CL decay reflects an artifact resulting from some luminol reaction-product that interferes with the light reaction. The major oxidation product of luminol is the 3-amino phthalate, yet it did not appear to affect CL, indicating that some other product is responsible for the loss of CL. Under our experimental conditions, only a small fraction of the luminol undergoes chemical change, 6 so the presumed reaction product might be very effective in quenching the light. Yet, its chemical structure is not known and is currently under study. In conclusion, the present results prove that i) contrary to previous conclusions, PMA-stimulated neutrophils from normal donors remain active for long periods of time and that continuous high flux of 0 2 does not terminate their capability of producing 02; ii) luminolamplified CL can neither serve to study in a continuous manner the formation kinetics nor be used as a valid continuous assay of ROS such as 02 and H20 2 generated by stimulated neutrophils; iii) luminol speciously report on H20 2 or 0 2 production even in a cell-free system such as HX/XO. REFERENCES 1. Babior B. M. The respiratory burst of phagocytes. J. Clin. Invest. 73"599-601; 1984. 2. Alien, R. C.; Mead, M. E.; Kelly, J. L. Phagocyte oxygenation activity measured by chemiluminescence and chemiluminegenic
3.
4.
5. 6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17. 18. 19.
20. 21. 22. 23.
probing. In Greenwald, R.A., ed. CRC handbook of methods for oxygen radical research. Boca Raton, FL: CRC Press, Inc. 1985: 343-351. Mills, E. L.; Rholl, K. S.; Quie, P. G. Luminol-amplified chemiluminescence: a sensitive method for detecting the carder state in chronic granulomatous disease. J. Clin. Microbiol. 12: 52-56; 1980. Cormier, M. J.; Prichard, P. M. An investigation of the mechanism of the luminescent peroxidation of luminol by stopped-flow technique. J. Biol. Chem. 243:4706--4714; 1968. Lind, J.; Merenyi, G.; Eriksen, T. E. Chemiluminescence mechanism of cyclic hydrazides such as luminol in aqueous solutions. J. Am. Chem. Soc. 105:7655-7661; 1983. Merenyi, C.; Lind, J.; Eriksen, T. E. The reactivity of superoxide (02) and its ability to induce chemiluminescence with luminol. Photochem. Photobiol. 41:203-208; 1985. Falck, P. Investigation of the chemiluminescence response of human neutrophils and mononuclear cells. Folia biologica (Praha)32:103-115; 1986. Dahlgren, C. Polymorphonuclear leukocytes chemiluminescence induced by formylmethionyl-leucyl-phenylalanine and phorbol myristate acetate: effects of catalase and superoxide dismutase. Agents and Actions 21:104-112; 1987. Vilim, V.; Wilhelm, J. What do we measure by a luminol-dependent chemiluminescence of phagocytes? Free Radic. Biol. Med. 6:623-629; 1989. Bruchelt, G.; Schmidt, K. H. Comparative studies on the oxidative processes during phagocytosis measured by luminol-dependent chemiluminescence. J. Clin. Chem. Clin. Biochem. 22: 1-13; 1984. Aniansson, H.; Stendahl, O.; Dahlgren, C. Comparison between luminol- and lucigenin-dependent chemiluminescence of polymorphonuclear leukocytes. Acta Path. Microbiol. Immunol. Scand. Sect. C 92:357-361; 1984. Dahlgren, C.; Stendahl, O. Role of myeloperoxidase in luminoldependent chemiluminescence of polymorphonuclear leukocytes. Infection and Immunity 39:736-741; 1983. Wymann, M. P.; von Tscharner, V.; Deranleau, D. A.; Baggiolini, M. Chemiluminescence detection of H202 produced by human neutrophils during the respiratory burst. Anal. Biochem. 65: 371-378; 1987. Lock, R.; Johansson, A.; Orselius, K.; Dahlgren, C. Analysis of horseradish peroxidase-amplified chemiluminescence produced by human neutrophils reveals a role for the superoxide anion in the light emitting reaction. Anal. Biochem. 173:450--455; 1988. Misra, H. P.; Squatrito, P. M. The role of superoxide anion in peroxidase-catalyzed chemiluminescence of luminol. Arch. Biochem. Biophys. 215:59-65; 1982. Dahlgren, C, Aniansson, H.; Magnusson, K.-E. Pattern of formylmethionyl-leucyl-phenylalanine-induced luminol- and lucigenin-dependent chemiluminescence in human neutrophils. Infection and Immunity 47:326-328; 1985. Dahlgren, C. Is lysosomal fusion required for the granulocyte chemiluminescence reaction? Free Radic. Biol. Med. 6:399--403; 1989. Bender, J. G. ; Van Epps, D. E. Analysis of the bimodal chemiluminescence pattern stimulated in human neutrophils by chemotactic factors. Infection and Immunity 41:1062-1070; 1983. Gyllenhammar, H. Mechanisms for luminol-augmented chemiluminescence from neutrophils induced by leukotriene B,, and Nformyl-methionyl-leucyl-phenylalanine. Photochem. Photobiol. 49:217-223; 1989. Test, S. T.; Weiss, S. J. Quantitative and temporal characterization of the extracellular H202 pool generated by human neutrophils. J. Biol. Chem. 259:399-405; 1984. Kleinhans, F. W.; Barefoot, S. T. Spin trap determination of free radical burst kinetics in stimulated neutrophils. J. Biol. Chem. 262"12452-12457; 1987. Holman, R. G.; Maier, R. V. Superoxide production by neutrophils in a model of adult respiratory distress syndrome. Arch. Surg. 123"1491-1495; 1988. Whitin, J. C.; Cohen, H. J. Disorders of respiratory burst termination. Hematol. Oncol. Clin. North Am. 2:289-299; 1988.
Superoxide production by neutrophils 24. Samuni, A.; Black, C. D. V.; Krishna, C. M.; Malech, H, L.; Bernstein, E. F.; Russo, A. Hydroxyl radical production by stimulated neutrophils reappraised. J. Biol. Chem. 263:13797-13801; 1988. 25. Boyum, A. Isolation of mononuclear cells and granulocytes from human blood. Scan. J. Clin. Lab. lnvest. 21 (Suppl. 97):77-89; 1968. 26. Kaplow, L. S. Simplified myeloperoxide stain using benzidine dihydrochloride. Blood 26:215-219; 1965. 27. Fridovich, I. Quantitative aspects of the production of superoxide anion radical by milk xanthine oxidase. J. Biol. Chem. 245: 4053-4057; 1970. 28. Simchowitz, L.; Cragoe, E. J., Jr. Regulation of human neutrophil chemotaxis by intracellular pH. J. Biol. Chem. 261: 6492--6500; 1986. 29. Samuni, A.; ICrishna, C. M.; Riesz, P.; Finkelstein, E.; Russo, A. Superoxide reaction with nitroxide spin-adducts. Free Radic. Biol. Med. 6:141-148; 1989. 30. Muller-Peddinghaus, R. Comparison of luminol- and lucigeninamplified chemiluminescence (CL) of phagocytes. In: Bors, W.; Saran, M.; Tait, D., eds. Oxygen radicals in chemistry and biology. Berlin: Walter de Gruyter & Co.; 1984: 881-882. 31. Hodgson, E. K.; Fridovich, I. The role of 02 in the chemiluminescence of luminol. Photochem. Photobiol. 18:451-455; 1973. 32. Wilhelm, J.; Vilim, V. Variables in xanthine oxidase-initiated luminol chemiluminescence implications for chemiluminescence measurements in biological systems. Anal. Biochem. 158: 201-210; 1986. 33. Dahlgren, C.; Lock, R. The limitation of the human neutrophil chemiluminescence response by extracellular peroxidase is stimulus dependent: effect of added horseradish peroxidase on the re-
313
sponse induced by both soluble and particulate stimuli. J. Clin. Lab. Immunol. 26:49-53; 1988.
34. Edwards, S. W.; Swan, T. F. Regulation of superoxide generation by myeloperoxidase during the respiratory burst of human neutrophils. Biochem. J. 237:601-604; 1986.
ABBREVIATIONS
CL-- chemiluminescence Cyt-cm-- ferricytochrome c DMPO--5,5-dimethyl-l-pyrroline-N-oxide DTPA--diethylenetriamine pentaacetic acid ESR--electron spin resonance HBSS--Hank's balanced salt solution HRP--horseradish peroxidase HX--hypoxanthine lucigenin--bis-N-methylacridinium nitrate luminol--5-amino-2,3-dihydro-l,4-phthalazinedione MPO-- myeloperoxidase PMA-- phorbol- 12-myristate- 13-acetate PMN-- polymorphonuclear leukocytes ROS--reactive oxygen species SOD-- superoxide dismutase XO--xanthine oxidase
