From www.bloodjournal.org by guest on January 21, 2015. For personal use only. Insight Into the Nature and Site of Oxygen-Centered Free Radical Generation by Endothelial Cell Monolayers Using a Novel Spin Trapping Technique By Bradley E. Britigan, Tedmund L. Roeder, and D. Michael Shasby Spin trapping, a sensitive and specific means of detecting free radicals, is optimally performed on cell suspensions. This makes it unsuitable for the study of adherent endothelial cell monolayers because disrupting the monolayer t o induce a cell suspension could introduce confounding factors. This problem was eliminated through the use of endothelial cells that were grown t o confluence on microcarrier beads. Using the spin trap 5.5-dimethyl-1-pyrroline-N-oxide (DMPO), the nature of free radical species generated by suspensions of microcarrier bead adherent porcine pulmonary endothelial cells under various forms of oxidant stress was examined. Exposure of these endothelial cells t o paraquat resulted in the spin trapping of superoxide (.O,-). Endothelial cell incubation in the presence of either bolus or continuous fluxes of hydrogen peroxide (H202) yielded spin trap evidence of hydroxyl radical formation, which was preventable by pretreating the cells with deferoxamine. Chromium oxalate which eliminates extracellular electron paramagnetic reso- nance spectrometry (EPR) signals, prevented the detection of DMPO spin adducts generated by paraquat but not H,O,treated endothelial cells. When endothelial cells were coincubated with PMA-stimulated monocytes evidence of both eo2and hydroxyl radical production was detected, whereas with PMA-stimulated neutrophils only eo2-production could be confirmed. Neutrophil elastase, cathepsin G, and the combination of PMA and A23187 have previously been suggested t o induce endothelial cell oxy-radical generation. However, exposure of endothelial cells t o each of these agents did not yield DMPO spin adducts or cyanide-insensitive endothelial cell 0, consumption. These data indicate that endothelial cell exposure: t o paraquat induces extracellular eo2-formation; t o H,O, leads t o intracellular hydroxyl radical production; and t o elastase, cathepsin G, or A23187/PMA does not appear t o cause oxy-radical generation. o 1992by The American Society of Hematology. HE INJURY of pulmonary endothelial cells by both exogenously and endogenously generated oxidant species has been linked to a variety of pathologic lung processes.‘ For example, experimental data has implicated endothelial cell injury by neutrophil-derived superoxide (eo,-)and hydrogen peroxide (H202)in the pathogenesis of increased vascular permeability occurring in the adult respiratory distress ~yndr0me.I.~ Some evidence has been obtained that hydroxyl radical, formed through the iron catalyzed reaction of .02-and H,O, (Haber-Weiss reaction), may be the oxidant species actually responsible for endothelial cell inj~ry.5’~ However, these conclusions are based only on indirect evidence showing the protective effects of “hydroxyl radical scavengers” and iron chelators on neutrophil-mediated endothelial cell injury. One limitation in defining the role of specific oxygencentered free radicals in various forms of endothelial cell injury has been the lack of free radical detection systems that are sensitive and specific for both intracellular and extracellular free radical species.” Both 0 - and hydroxyl radical will react with some nitrone compounds, termed spin traps, with the resultant generation of stable nitroxide free radicals (spin adducts) whose characteristic hyperfine electron paramagnetic resonance spectrometry (EPR) splitting patterns allow identification of the original free radical species present.” Spin trapping has been increasingly applied in recent years to the in vitro and in vivo study of oxygen-centered free radical production by a variety of cellular systems.” One limitation to spin trapping is that, because of the nature of EPR, the technique is most readily applicable to studies of cells in suspension. This poses problems for investigating cells, such as endothelial cells, which under normal physiologic conditions exist as adherent monolayers. Spin trapping has previously been used to study free radical production by endothelial cells, but the cells were studied after they had been released into suspension by Results from such studies may not be readily applicable to adherent cells because the scraping procedure could potentially alter important physical and biochemical cell parameters. Consequently, we developed a new spin trapping system that allows detection of intracellular and extracellular free radical production by adherent endothelial cells. This report describes the use of this technique to investigate the nature and site of free radical production resulting from exposure of pulmonary endothelial cells to pharmacologic agents and human phagocytes. T Blood, Vol79, No 3 (February 1). 1992: pp 699-707 MATERIALS AND METHODS Reagents. 5,s-dimethyl-1-pyrroline-N-oxide (DMPO), 2,2,6,6tetramethylpiperidine-N-oxy1(TEMPO), xanthine, catalase, CuZn superoxide dismutase (SOD), diethylenetriaminepentaacetic acid (DTPA), sodium cyanide (CN-),l,l’-dimethyl-4,4’-bipyridinium From the Research Service and the Department of Intemal Medicine, VA Medical Center; and the Department of Intemal Medicine, University of Iowa College of Medicine, Iowa City. Submitted January 9, 1991; accepted September 27, 1991. Supported in part through the VA Research Service, National Institutes of Health awards HL44275, AI28412, HL33540, and HL42385, The Pfizer Scholars Program, and The Cystic Fibrosis Foundation. It was performed during the tenure of B.E.B. and D.M.S. as a VA Research Associate and VA Clinical Investigator, respectively. Portions of this work have been presented in abstractform at the fifrh biannual meeting of the Intemational Society for Free Radical Research at Pasadena, CA, November 14-20, 1990 (Free Radic Biol Med 9S1:37, 1990) and at the annual meeting of the Midwest section of The American Federation for Clinical Research at Chicago, IL, October 31-November 2, 1990 (Clin Res 38:871A, 1990). Address reprint requests to Bradley E. Britigan, MD, The University of Iowa College of Medicine, Department of Intemal Medicine, Division of Infectious Diseases, SW54, GH, Iowa City, LA 52242. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. section 1734 solely to indicate this fact. 0 1992 by The American Society of Hematology. 0006-4971I921 7903-0033$3.0010 699 From www.bloodjournal.org by guest on January 21, 2015. For personal use only. 700 dichloride (paraquat), hydrogen peroxide (30%, H,O,), glucose oxidase, and deferoxamine mesylate were purchased from Sigma Chemical (St Louis, MO). Additional reagents were obtained as follows: xanthine oxidase from Boehringer-Mannheim (Indianapolis, IN); dimethyl sulfoxide (DMSO) from Fisher Scientific (Fairlawn, NJ); ethanol from Aaper Alcohol and Chemical Co (Shelbeyville, KY); and human neutrophil elastase from Elastin Products (Owensville, MO). Cathepsin G was purified from human neutrophils by previously described method^.'^ Chromium oxalate, synthesized according to the method of Bailar et a],” was kindly provided by Dr Gerald M. Rosen (University of Maryland School of Pharmacy, Baltimore). Endothelial cellculture. Confluent monolayers of porcine pulmonary artery endothelial cells were cultured on microcarrier beads (Qtodex 3; Pharmacia, Piscataway, NJ) as previously described at our institution.16Cells were cultured for 2 to 3 days postconfluence (as determined by experience with this system since its original development). Multiple cell lines were used for these studies, with cells utilized between passage 6 and 9. In each day’s experiment, positive and negative controls were performed with that day’s cell preparation to eliminate any contribution of variation in cell line or passage number to the results. Before use, microcarrier beads were washed free of cell culture media (M199 containing 10% heatinactivated fetal bovine serum; University of Iowa Cancer Center, Iowa City) by transferring the beads to a 15-mL conical tube (Sarstedt, Princeton, NJ) and allowing the beads to precipitate (-5 minutes). Culture media was then removed by aspiration and replaced with 10 mL Hanks’ balanced salt solution without phenol red (HBSS, University of Iowa Cancer Facility). Beads and HBSS were gently mixed and the beads allowed to precipitate again. The HBSS was then replaced with fresh HBSS and the process repeated three times. After the final wash the cells were resuspended in HBSS and all experiments were then performed in this buffer system. Except where indicated, final incubation mixtures contained 25% (vol/vol) microcarrier beads. Based on previous work using the identical system of cell culture,’’ this 25% vol/vol microcarrier beads would yield a final cell concentration of 0.5 to 1.0 x lo7 cells/mL. In some experiments, microcarrier beads lacking cells, but which had been incubated in cell culture media and in all other respects treated identically to cell-containing beads, were used as negative controls to eliminate any potential contribution of the beads and/or culture media to the results. Spin trapping. Desired reaction mixtures (0.5 mL) were assembled in 12 x 75 mm glass tubes and transferred to a quartz EPR flat cell, which was in turn placed into the cavity of the EPR spectrometer (Varian E104A; Varian Instruments, Palo Alto, CA). Resulting EPR spectra were recorded sequentially at 25°C. Unless othenvise noted, spectrometer settings were: microwave power, 20 mW; modulation amplitude, 1.0 G; modulation frequency, 100 kHz; response time, 1 second; and gain 5.0 x lo4. Routinely, reaction mixtures contained 25% (vol/vol) microcarrier beads. The EPR spectrum of the stable nitroxide TEMPO was unchanged with the inclusion of up to 50% (vol/vol) Cytodex 3 microcarrier beads in the solution, with appropriate adjustment of the solution so that the final TEMPO concentration remained lo-’ mol/L. Thus, the presence of the microcarrier beads did not adversely effect the sensitivity of the system. Oxygen consumption. Oxygen consumption was measured as the rate of oxygen depletion from a 3-mL reaction mixture located in the chamber of a Clark oxygen electrode (YSI 53; Yellow Springs Instrument Co, Yellow Springs, OH) and monitored continuously via an attached strip chart recorder. Reaction mixtures routinely contained 1 mmol/L CN- to eliminate cellular 0, consumption resulting from mitochondrial respiration. This cyanideinsensitive respiration has been previously used as a means of BRITIGAN, ROEDER,AND SHASBY measuring univalent reduction to 0, to .O,- induced in cellular systems by redox active pharmacologic compounds such as paraquat.”,’’ RESULTS Free radical generation induced by paraquat. To optimally study pulmonary endothelial cell free radical production in vitro using spin trapping techniques, it was necessary to develop a system in which adherent monolayers could be examined. Accordingly, the possibility that suspensions of microcarrier bead adherent endothelial cells could be used for such a purpose was examined. Exposure to the herbicide paraquat induces an acute lung injury syndrome,” which has been suggested to result from the ability of this compound to undergo cell-mediated redox cycling with resultant generation of .02and H126.96.36.199 Consequently, paraquat was used as an initial means of evaluating the feasibility of spin trapping free radicals generated by microcarrier bead adherent endothelial cells. Porcine pulmonary endothelial cells were grown to confluence on microcarrier beads, after which they were washed free of cell culture media. They were then suspended in HBSS containing 100 mmol/L DMPO and 0.1 mmol/L DTPA, after which paraquat (0.1 to 10 mmol/L) was added. The resulting EPR spectra was that of the hydroxyl radical-derived spin adduct of DMPO, DMPO/ .OH (A, = A, = 15.0 G; Fig 1B). No spectrum above background was observed if either paraquat (Fig 1A) or endothelial cells (not shown) were omitted from the reaction mixture. Zdentifcation of paraquat induced free radical. It has been previously shown by multiple i n v e s t i g a t o r ~ that ~ ’ ~ ~in~ addition to being a product of the reaction of DMPO and hydroxyl radical, DMPO/.OH also can result from the decomposition of the superoxide-derived spin adduct, DMPO/*OOH.This process is more rapid in the presence of some cellular s y ~ t e m s . *Thus, ~ . ~ ~the detection of DMPO/ .OH is not specific for hydroxyl radical, but rather could arise simply from the presence of 0 - alone. To differentiate these two possibilities it is necessary to determine the effect of various oxidant scavengers on the system. Hydroxyl radical reacts rapidly with DMSO or ethanol to generate methyl radical (.CH,) or a-hydroxyethyl radical (CHOHCH,), respectively.2’ In an experimental system in which the concentration of DMSO or ethanol exceed that of DMPO, hydroxyl radical formation results in production of the DMPO spin adducts of C H , or CHOHCH,, respectively, with a concomitant decrease in the expected DMPO/.OH peak amplit~des.~~.’’ Accordingly, microcarrier bead adherent endothelial cells were exposed to paraquat in the presence of DTPA, 100 mmol/L DMPO, and 140 mmol/L DMSO or ethanol. The magnitude of resulting DMPO/.OH peaks was unaffected by the presence of DMSO or ethanol and no new spin adducts were detected (Fig 1E) suggesting that the DMPO/.OH species detected resulted from decomposition of the superoxide derived spin adduct rather than spin trapping of hydroxyl radical. To provide further evidence in support of this interpretation, experiments were repeated in which reac- ’’ From www.bloodjournal.org by guest on January 21, 2015. For personal use only. ENDOTHELIAL CELL OXYGEN RADICAL GENERATION 701 10 0 EC spectrum above background was observed with addition of H,O, to microcarrier beads that lacked endothelial cells but which had been maintained in cell culture media and washed analogous to the adherent endothelial cells or used without washing. Furthermore, even addition of H,O, to similar unwashed “sham” microcarrier bead preparations or a 25%/75% (vol/vol) of the endothelial cell culture media/HBSS solution containing DTPA and DMPO did not yield an EPR signal above background. These data eliminate the possibility that the hydroxyl radical detected with addition of H,O, to the endothelial cells resulted from the presence of iron associated with the microcarrier beads or residual cell culture media rather than the endothelial cells themselves. Because endothelial cells would be unlikely to be exposed in vivo to a large bolus of H,O,, the effect of a continuous flux of H,O, generated by the oxidation of glucose by glucose oxidase was examined. Again, an EPR spectrum comprised of catalase-inhibitable DMPO/-OH was detected (Fig 2, D and E). Previous studies suggested that iron associated with the endothelial cells was responsible for catalyzing the formation of hydroxyl radical resulting from H,O, exposure?23” Consistent with this hypothesis, preincubation of the endothelial cell suspension for 30 minutes with deferoxamine B EC + PQ + CAT EC + PQ + SOD 2 EC + PQ + DMSO n h Fig 1. Free radical species resulting from exposure of endothelial cell monolayers (EC) t o paraquat (Pa). Conditions for each EPR spectrum shown are as follows: (A) 25% (vol/vol) microcarrier bead adherent endothelial cells suspended in the presence of 100 mmol/L DMPO and 0.1 mmol/L DTPA. (E) Same conditions as (A) except that the cells were exposed t o 1 mmol/L paraquat. (C) Paraquat-treated endothelial cells as in (E) except that catalase (CAT, 500 U/mL) was also present. (D) Paraquat-treated endothelial cells as in (E) except that SOD (30 U/mL) was also present. (E) Paraquat-treated endothelial cells as in (E) except that DMSO (140 mmol/L) was also present Identical results were obtained with the use of ethanol instead of DMSO. tion mixtures contained SOD (30 U/mL) or catalase (500 U/mL) during endothelial cell exposure to paraquat. SOD eliminated DMPO/.OH formation, whereas catalase was without effect (Fig 1, C and D), consistent with 0 -rather than hydroxyl radical induced DMPO/.OH formation. Hydrogen peroxide associated free radical formation with endothelial cells. Hydrogen peroxide has been shown to cause endothelial cell i n j ~ r y . ~ , ~The ’ ” ~mechanism of H,O, toxicity has been postulated to involve hydroxyl radical generated by the interaction of H,O, with cellular iron Consequently, we sought evidence for the formation of hydroxyl radical after endothelial cell exposure to H,O,. Microcarrier bead adherent endothelial cells were suspended in HBSS containing DTPA and DMPO, after which H,O, was added. The resulting EPR spectrum was that of DMPO/.OH (Fig 2A) and was detectable at H,O, concentrations 20.1 mmol/L. In contrast to results with paraquat, catalase totally eliminated the spectrum (Fig 2B). Ethanol decreased DMPO/.OH peak amplitudes with a corresponding appearance of the DMPO/CHOHCH, spin adduct (AN= 16.0 G, AH = 23.1 G; Fig 2C). These results are consistent with spin trapping of hydroxyl radical. No I 10Q EC + H202 EC + H202+ CAT B EC + H202+ ETOH C EC + G/GO + CAT E Fig 2. Effect of endothelial cell exposure t o H,O,. Conditions for each EPR spectrum shown are: (A) 25% (vol/vol) microcarrier bead adherent endothelial cells (EC) suspended in 100 mmol/L DMPO and 0.1 mmol/L DTPA after the addition of 1 mmol/L H,O,. (E) H,O,treated endothelial cells as in (A) except that 500 U/mL catalase (CAT) was also present. (C) H,O,-treated endothelial cells as in (A) except that 140 mmol/L ethanol (ETOH) was also present. (D) Endothelial cells exposed t o a continuous flux of H,O, generated by the reaction of glucose and glucose oxidase. (E) Same conditions as in (D) except that catalase was present. Note the location of the 6-line spectrum of DMPO/CHOHCH, in tracing C is designated by the asterisk (*). From www.bloodjournal.org by guest on January 21, 2015. For personal use only. BRITIGAN, ROEDER, AND SHASBY 702 (0.1 mmol/L) before the addition of H 2 0 2 eliminated detectable hydroxyl radical formation (Fig 3B). No inhibition was noted if iron-saturated deferoxamine was used (Fig 3C). It has previously been reported that deferoxamine may have other effects on modulating hydroxyl radical formation beyond simple iron c h e l a t i ~ n ? To ~ - ~help ~ eliminate such an explanation for our results, endothelial cells that had been pretreated with deferoxamine were extensively washed in deferoxamine-free HBSS before the addition of H,02. Once again, no evidence of hydroxyl radical generation was detected (Fig 3D). Induction of hydro& radical by a superoxide generating system. In contrast to the above system in which H20,was the sole oxidant species to which endothelial cells were exposed, the respiratory burst of phagocytes generates a continuous flux of both .O,- and H20,.36The reaction of xanthine oxidase and (hypo) xanthine yields similar prodUCts24.37.38and has been used as a convenient model system for the products of stimulated phago~ytes.~’,~ Therefore, microcarrier bead adherent endothelial cells were suspended in xanthine, DMPO, ethanol, and DTPA. EPR spectra were then obtained after the initiation of continuous 0 - /H,O, production by the addition of xanthine oxidase. The resulting EPR spectra were a composite of three species: the superoxide-derived spin adduct, DMPO/ *OOH (A, = 14.3 G, A, = 11.7 G, A, = 1.3 G); DMPO/ .OH (A, = A, = 15.0 G); and DMPO/CHOHCH, (A, = 16.0 G, A, = 23.1 G, Fig 4A). The detection of DMPO/CHOHCH, as discussed above would be expected I 100 EC + H202+ DF EC + H202+ Washed DF Fig 3. Effect of deferoxamine on hydroxyl radical generated by H,O,-treated endothelial cells. Conditions for each of the EPR spectra shown are: (A) 25% endothelial cells exposed t o 1mmol/L H,O, in the presence of 100 mmol/L DMPO and 0.1 mmol/L DTPA. (B) Same conditions as in (A) except that the endothelial cells were preincubated with 0.1 mmol/L deferoxamine (DF) for 30 minutes before the addition of H,O,. (C) Same conditions as in (B) except that iron-loaded deferoxamine (FeDF) was used. (D) Same conditions as in (B) except that the endothelial cells were extensively washed t o remove the deferoxamine before the addition of HO , ., 10 G EC + WXO EC + XMO + CAT EC + WXO + SOD C I l l OOH OH CH, Fig 4. Detection of free radicals generated after exposure of /HZ02generated by the endothelial cells t o a continuous flux of 0 reaction of xanthine oxidase and xanthine. Conditions for each of the EPR spectra shown are: (A) Endothelial cells suspended in 100 mmol/L DMPO, 0.1 mmol/L DTPA, 140 mmol/L DMSO, and 2 mmol/L xanthine, after which 0.04 U/mL xanthine oxidase was added. (B) Same as (A) except catalase (CAT, 500 U/mL) was present. (C) Same as (A) except that SOD (30 U/mL) was present. High and low field peaks corresponding t o the DMPO spin adducts DMPO/*OOH, DMPO/.OH, and DMPO/.CH, are designated OOH, OH, and CH, respectively. as a consequence of hydroxyl radical spin trapping. Consistent with this, the presence of DMPO/-CHOHCH, was inhibited by both catalase (Fig 4B) and SOD (Fig 4C). In contrast, the other two spin adducts were inhibited by only SOD (Fig 4, B and C ) , confirming that they arose by spin trapping .O,-, the initial product of the xanthine/xanthine oxidase reaction. As we have previously ob~erved,‘~ for reasons that remain unclear, in some cases, the presence of catalase appeared to increase the relative ratio of DMPO/ .OH:DMPO/.OOH resulting from the spin trapping of 0 (Fig 4, A v B). Site of paraquat- and hydrogen peroxide-induced free radical formation. Chromium oxalate induces a concentrationdependent broadening of the EPR spectral lines resulting from solutions of nitroxide free radical^.^' As the concentration of chromium oxalate increases a point is reached in which the peaks are broadened to such an extent that they become indistinguishable from background. For example, 5 mmol/L chromium oxalate essentially eliminated the EPR spectrum resulting from DMPO/.OH (Fig 5, A and B), previously generated via the reaction of H,O, and Fe2+. Because chromium oxalate is able to penetrate cell membranes to only a limited extent, its addition to cellular free From www.bloodjournal.org by guest on January 21, 2015. For personal use only. ENDOTHELIAL CELL OXYGEN RADICAL GENERATION 703 10 0 DMPO I 'OH A DMPO / 'OH + CrOx B m EC + PQ C EC + PQ + CrOX D sure of endothelial cells to activatedphagocytes. Experimental evidence obtained by other investigators has suggested a role for hydroxyl radical in phagocyte-mediated endothelial cell injury.5-'The results described above with exposure to H,O,, glucose/glucose oxidase, and xanthine/xanthine oxidase provide support for these earlier studies. To further examine this hypothesis, additional spin trapping studies were performed in which endothelial cells were incubated in the presence of PMA-stimulated human neutrophils or peripheral blood monocytes, DMPO, DTPA, and ethanol. EPR spectra of the reaction mixture of endothelial cells and monocytes was that of DMPO/.OH and DMPO/ CHOHCH, (Fig 6A). Addition of catalase inhibited predominantly DMPO/CHOHCH, (Fig 6B), whereas SOD eliminated both species (Fig 6C). These results are consistent with previous data4' showing that the spin trapping of .02derived from stimulated mononuclear phagocytes is manifested predominantly as DMPO/.OH rather than DMPO/.OOH due to cell-mediated enhancement of the rate of conversion of DMPO/.OOH to DMPO/.OH. The detection of catalase-inhibitable DMPO/CHOHCH, as observed here is indicative of spin trapping of hydroxyl radical. Because formation of hydroxyl radical after stimulation of human monocytes is not seen in the absence of an exogenous iron catalyst," it is likely that hydroxyl radical production resulted from the presence of catalytic iron provided by the endothelial cells. In contrast to the results with monocytes, the EPR spectra resulting from the exposure of endothelial cells to stimulated neutrophils yielded a predominance of DMPO/.OOH and DMPO/-OH, with only scant detection of DMPO/CHOHCH, (Fig 6D). All species were inhibited by SOD but not catalase (Fig 6, E and F), indicating that the spectrum resulted from the spin trapping of 0 -with subsequent decomposition of the DMPO/.OOH spin adduct to other nitroxides as has been previously noted in neutrophil-containing ~ y s t e m s .Once ~ ~ . ~again, ~ the presence of catalase appeared to increase the peak amplitude ratio of DMPO/.OH:DMPO/.OOH. Inclusion of the myeloperoxidase inhibitor azide in the neutrophil reaction mixture did not alter the EPR spectrum. Experiments with chromium oxalate to localize the site of free radical formation were not attempted because it had previously been shown that the concentration of chromium oxalate necessary to eliminate extracellular EPR signals markedly inhibits phagocyte free radical formation.# Can elastase, cathepsin G, P M , and/or the calcium ionophore A23187 induce endothelial cell free radical formation? The neutrophil-derived proteases elastase and cathepsin G increase permeability of endothelial cell monolayers to a l b ~ m i n ' ~ .by ~ ' ,a~protease-independent ~ mechanism that has been suggested to involve endothelial cell free radical Ph4A and the calcium ionophore A23187 have also been reported to induce endothelial cell .O; f~rmation.~' However, when up to 60 p,g/mL of human elastase or cathepsin G or the combination of A23187 (0.1 mg/L) and PMA (0.2 mmol/L) were added to microcarrier bead adherent endothelial cells in HBSS containing Caz+ (1.67 mmol/L), M$' (4.1 mmol/L), DTPA, and DMPO, no 1 E F Fig 5. Localization of the cellular site of paraquat and H,O,induced free radical production using the EPR line-broadening agent chromium oxalate (CrOx) to eliminate extracellular EPR signals. Conditions for each EPR spectrum shown are: (A) A solution containing the DMPO/-OH spin adduct generated previously by the reaction of H,Oz and Fez+in the presence of DMPO. (6) Same as in (A) except that 5 mmol/L chromium oxalate was added to the solution. (C) Endothelial cells treated with 1 mmol/L paraquat (PO) in the presence of DMPO and DTPA. (D) Same as in (C) except the reaction mixture also contained 5 mmol/L chromium oxalate. (E) Endothelial cells treated with 1 mmol/L HzO, in the presence of DMPO and DTPA. (F) Same as (E) except that 5 mmol/L chromium oxalate was also present. radical generating systems has been used as a means of localizing the site of free radical formation to intracellular or extracellular location^.^^ In the presence of chromium oxalate, extracellularly located spin adducts become undetectable whereas the EPR spectrum of intracellular species remains detectable. Consequently, the effect of chromium oxalate on the EPR spectrum resulting from the addition of paraquat or H,O, to microcarrier bead adherent endothelial cells suspended in the presence of DMPO and DTPA was examined. Chromium oxalate (5 mmol/L) eliminated the DMPO/.OH spectrum resulting from exposure to paraquat (Fig 5, C and D). However, it had only minimal effect on H,O,-induced DMPO/.OH peaks (Fig 5, E and F). These data are consistent with paraquat-mediated induction of extracellular (or in close proximity to the extracellular space) 0 -production. However, hydrogen peroxide exposure appears to lead to intracellular hydroxyl radical formation. Characteristics of free radical formation induced by a p o - From www.bloodjournal.org by guest on January 21, 2015. For personal use only. BRITIGAN, ROEDER, AND SHASBY 704 negative spin trapping data, cyanide-insensitive0,consumption was not increased with exposure of endothelial cells to elastase (Fig 7), cathepsin G (Fig 7), or A23187/PMA (not shown). It is worth noting that in the experiments described earlier, DMPO/.OH detection required endothelial cell exposure to 0.1 mmol/L paraquat, a paraquat concentration that results in consumption of 1 nmol of OJmin by cyanide treated endothelial cells (Fig 7). Thus, the spin trapping system should detect 0 -production in excess of 1 nmol/min. - DISCUSSION Y EC + MC + SOD EC+PMN+SOD , Fig 6. Free radicals generated after endothelial cell exposure to stimulated phagocytes. Conditions for EPR spectra shown are: (A) Endothelial cells incubated in the presence of PMA (100 ng/mL) stimulated peripheral blood monocytes (MC, 5 x 1O6/mL) in the presenceof 100mmol/LDMPO, 140mmol/LethanoI,andO.l mmol/L DTPA. (B) Same as (A) except the reaction mixture contained 500 U/mL catalase (CAT). (C) Same as (A) except 30 U/mL SOD was present. (D) Same as (A) except that PMA-stimulated neutrophils (PMN) were substituted for monocytes. (E) Same as (D) except catalase was present. (F) Same as (D) but with SOD added. Note the receiver gain in scan (F) was twice that of (D) and (E). High and low field peaks corresponding t o DMPO/.OOH, DMPO/.OH, and DMPO/ CHOHCH, are designated OOH, OH, and CH, respectively. EPR spectrum above background was detected over 2 6 0 minutes of observation (data not shown). Cellular .O,- production resulting from exposure to xenobiotics such as paraquat consumes 0, which, in contrast to that resulting from normal cellular respiration, is not inhibited by cyanide (Fig 7).17,18 Consistent with the Spin trapping has become an increasingly popular technique for the investigation of the role of free radical species in human biology." However, a significant limitation to the application of spin trapping to in vitro cellular systems, such as endothelial cell that exist as monolayers, is that spin trapping is most easily performed with cell suspensions. Consequently, we set out to develop a system in which endothelial cell monolayers could be examined intact by conventional spin trapping techniques. Suspensions of microcarrier beads adherent porcine pulmonary endothelial cells did not alter the EPR spectra of the stable nitroxide TEMPO, indicating that such cell preparations are amenable to study by conventional spin trapping techniques. This was confirmed by exposing such endothelial cell monolayers to paraquat. Incubation of microcarrier bead adherent endothelial cells with paraquat in the presence of DMPO yielded the hydroxyl radical spin adduct of DMPO, DMPO/-OH. However, DMPO/.OH formation was inhibited by SOD but not catalase or the hydroxyl radical scavengers DMSO and ethanol. This indicates that the DMPO/.OH spectrum resulted from 0 generation via the well-re~ognized~"~~ decomposition of the superoxide spin adduct DMPO/.OOH to DMPO/.OH rather than spin trapping of hydroxyl radical. Similar conclusions have been reported after paraquat treatment of other cell ~ystems.~~.~' In some4y-51 but not all ~ y s t e m s , ~iron ~ . ' ~ chelation and other means of limiting the potential for hydroxyl radical formation appear to limit paraquat-mediated cytoxicity. We ae 1 Time Fig 7. Induction of endothelial cell cyanide-insensitive 0, consumption. 0, consumption over time by 25% microcarrier bead adherent endothelial cells in HBSS (Control) and in the presence of 1 mmol/L NaCN (+CN). Also shown is the increase in cyanide-insensitive respiration after the addition t o NaCN treated endothelial cells of 1 mmol/L paraquat (+PO, +CN), 60 pg/mL elastase (+EL, +CN), or 60 pg/mL cathepsin G (+CG, +CN). Note paraquat but not elastase increasedthe rate of cyanide-insensitive 0, consumption indicative of '0, - formation. Time elapsed for the peroid shown was 30 minutes. From www.bloodjournal.org by guest on January 21, 2015. For personal use only. ENDOTHELIAL CELL OXYGEN RADICAL GENERATION 705 did not find spin trapping evidence of hydroxyl radical generation during endothelial cell exposure to paraquat. However, our results do not exclude a role for this free radical species at a cellular site where the microenvironment precluded its detection. The cellular site at which oxidants are generated is also likely to be of considerable importance in their relative toxicity. Addition of the EPR line broadening agent chromium oxalate during paraquat treatment of endothelial cells eliminated the expected DMPO/.OH spectrum. Because the concentration of chromium oxalate used did not eliminate the EPR spectra induced by H,O, exposure, these results cannot be attributed to intracellular penetration of chromium oxalate. Rather, they complement earlier work suggesting that .O,- production induced by paraquat occurs predominantly at extracellular locations?o34 Presumably this involves the intracellular reduction of paraquat to the paraquat free radical, which subsequently diffuses outside the cell where it is oxidized by ambient oxygen with regeneration of paraquat and coincident formation of .oz- .20.48 Exposure of endothelial cells monolayers to H,O, results in a decrease in the capacity of the monolayer to function as a diffusion barrier. Incubation of suspensions of adherent endothelial cells to either a bolus or a continuous flux of H,O, +- 0 -resulted in the generation of catalaseinhibitable DMPO/-OH whose magnitude decreased in the presence of DMSO or ethanol with resulting appearance of DMPO/CH, and DMPO/CHOHCH,, respectively. These data confirm other experimental dataG9that exposure of endothelial cells to exogenous H,O, +- -0,-results in the generation of hydroxyl radical. The failure of chromium oxalate to inhibit much of the DMPO/.OH spectrum after endothelial cell incubation with H,O, suggests that a considerable portion of the hydroxyl radical is generated intracellularly. Preincubation of the endothelial cells with deferoxamine prevented DMPO/.OH formation, suggesting that hydroxyl radical is most likely generated through a Fenton reaction catalyzed by an as yet unknown intracellular iron chelate(s). Alternatively, Beckman et a135 have recently reported that the interaction of 0 with endothelial cell-derived peroxynitrite can also generate .OH. Although not applicable to results with exposure to only H,O,, this mechanism could contribute .OH generation observed with endothelial cell exposure to fluxes of .O,- . In vivo phagocytic cells would constitute one of the principal exogenous sources of H,O, and 0 - to which endothelial cells would likely be exposed. When endothelial cells were exposed to PMA-stimulated peripheral blood monocytes spin trap evidence of a small amount of hydroxyl radical formation was detected. We have previously shown that monocytes by themselves do not have the capacity for hydroxyl radical g e n e r a t i ~ n . ~Thus, ~ . ’ ~ it is likely that the source of the iron catalyst in this case was again the endothelial cell. Unfortunately, the location (intracellular v extracellular) of the endothelial cell-derived iron could not be determined. Elimination of monocyte-generated H,O, before its interaction with the endothelial cell monolayer prevented hydroxyl radical production. However, this effect I of catalase would be expected regardless of whether the iron catalyst was located within or on the external surface of the endothelial cell. Experiments with chromium oxalate to eliminate EPR spectra from spin trapping of extracellularly generated hydroxyl radical could not be performed due to the fact that chromium oxalate inhibits the phagocyte respiratory burst. Nevertheless, because monocytes effectively serve as a source of continuous H,O, flux it seems most likely that the source of iron responsible for hydroxyl radical formation is inside the endothelial cell, as is the case with hydroxyl radical formed with endothelial cells treated with H,O, or glucose/glucose oxidase. In contrast to the monocyte results, a reaction mixture composed of endothelial cells and PMA-stimulated neutrophils failed to yield spin trap evidence of hydroxyl radical. These results were somewhat surprising given considerable circumstantial evidence obtained by o t h e d 9 that damage to endothelial cells by neutrophil-derived oxidants is mediated by hydroxyl radical. There appear to be a number of possible explanations for this apparent discrepancy. Release of myeloperoxidase and lactoferrin during the neutrophil respiratory burst have the capacity to inhibit the generation of hydroxyl radical resulting from neutrophil stimulation in the presence of exogenous catalytic iron c h e l a t e ~ . ~In ~ -the ~ ’ present work, it seems possible that the release of these two granule components decreased the magnitude of hydroxyl radical generation to a level below the limits of detection of the spin trapping system used. However, inclusion of the myeloperoxidase inhibitor azide failed to result in spin trap evidence of hydroxyl radical formation. Because neutrophils contain lactoferrin and monocytes do not, lactoferrin release would seem on the surface to be a highly likely explanation. However, as described above, the iron catalyst responsible for hydroxyl radical generation on endothelial cell exposure to sources of H,O, appears to be located intracellularly. Therefore, it seems unlikely that lactoferrin that would remain outside the endothelial cell could influence the result of an intracellular reaction. Alternatively, the stability of hydroxyl radicalderived DMPO spin adducts decreases in the presence of PMA-stimulated n e u t r o p h i l ~ . ~ ’Thus, ~ ~ ~ ~ hydroxyl ’~ radical generation could have occurred during the exposure of endothelial cells to stimulated neutrophils, but the relative instability of the resulting spin adducts precluded their detection. Two new means of spin trapping hydroxyl radical which appear to be relatively immune to this phenomenon have recently been de~cribed’~~’~ and could help clarify the situation. Neutrophils possess a variety of granule proteases suggested to contribute to inflammatory tissue injury. Recently, several l a b o r a t ~ r i e s , ’ ~including , ~ , ~ ’ our 0wn,[email protected] have shown that exposure of endothelial cell monolayers to two different neutrophil granule components, elastase and cathepsin G, leads to cell injury. The ability of elastase and cathepsin G to produce these effects appears to be independent of their enzymatic activity but related to their cationic proper tie^'^.^' and mediated through the release of intracellular calcium.62It has been suggested that the nonenzymatic injury of endothelial cells by elastase and/or cathepsin G From www.bloodjournal.org by guest on January 21, 2015. For personal use only. BRITIGAN, ROEDER, AND SHASBY 706 may result from the induction of cellular free radical formation.46The results of the present study do not support that hypothesis. As measured by both cyanide-insensitive 0, consumption and spin trapping, we detected no evidence of 0 - generation (sensitivity 1 nmol/min) after exposure of microcarrier bead adherent endothelial cells to up to 60 pg/mL elastase or cathepsin G . Similarly, in contrast to an earlier report: no evidence of free radical formation was observed after exposure of a similar concentration of endothelial cells to PMA or A23187. This latter discrepancy could relate to one or more of a variety of differences between the earlier work of work of Matsubara and Ziff4’ and our own including the site of cell harvest and the species used. In addition, the possibility remains that the human umbilical vein preparations in the earlier study could have been contaminated with tissue macrophages. In summary, using a system that allows detection of free radical formation by intact endothelial cell monolayers using conventional spin trapping techniques, we have confirmed previous that exposure of endothelial cells to the redox active compound paraquat induces the generation of .02at, or in close proximity to, the extracellular space. In contrast, exposure of these cells to bolus and - or continuous fluxes of H,O, 2 results in the intracellular generation of hydroxyl radical, catalyzed by cellular iron chelates. These results strongly support the hyp~thesis~~~*”’ that phagocyte-mediated injury to endothelial cells may occur as a consequence of intracellular hydroxyl radical formation. Studies with stimulated monocytes appear to be consistent with this hypothesis, although the exact site of hydroxyl radical generation could not be determined. However, we were surprisingly unable to detect the generation of hydroxyl radical during the exposure of the same endothelial cells to stimulated neutrophils. Further work is required to delineate the reason(s) for this unexpected result. Nevertheless, the ability to quantitate and localize free radical formation in systems containing intact endothelial cell monolayers will likely provided new insights into the mechanism of oxidative injury to these cells. ACKNOWLEDGMENT We acknowledge the technical assistance of Kathy Lindsley in maintaining the endothelial cell cultures, and the help of Kathy Schmuecker and Naomi Erickson with preparation of the manuscript. REFERENCES 1. Heffner JE, Repine JE: Pulmonary strategies of antioxidant defense. Am Rev Respir Dis 140531,1989 2. Till GO, Johnson KJ, Kunkel R, Ward PA: Intravascular activation of complement and acute lung injury: Dependency on neutrophils and toxic oxygen metabolites. J Clin Invest 69:1126, 1982 3. Shasby DM, Vanbenthuysen KM, Tate RM, Shasby SS, McMurthry I, Repine JE: Granulocytes mediate acute edematous lung injury in rabbits and in isolated rabbit lungs perfused with phorbol myristate acetate: role of oxygen radicals. Am Rev Respir Dis 125:443, 1982 4. Johnson KJ, Ward PA: Role of oxygen metabolites in immune complex injury of lung. J Immunol126:2365, 1981 5. Fox RB: Prevention of granulocyte mediated lung injury in rats by a hydroxyl radical scavenger, dimethylthiourea. J Clin Invest 74:1456,1984 6. Ward PA, Till GO, Kunkel R, Beauchamp C Evidence for the role of hydroxyl radical in complement and neutrophildependent tissue injury. J Clin Invest 72:789,1983 7. Till GO, Hatherill JR, Tourtellotte WW, Lutz MJ, Ward P A Lipid peroxidation and acute lung injury after thermal trauma to skin. Am J Pathol 119:376, 1985 8. Kuroda M, Murakami K, lshikawa Y: Role of hydroxyl radicals derived from granulocytes in lung injury induced by phorbol myristate acetate. Am Rev Respir Dis 135:1435,1987 9. Gannon DE, Varani J, Phan SH, Ward JH, Kaplan J, Till GO, Simon RH, Ryan US, Ward P A Source of iron in neutrophilmediated killing of endothelial cells. Lab Invest 57:37,1987 10. Cohen MS, Britigan BE, Hassett DJ, Rosen GM: Do human neutrophils form hydroxyl radical? Evaluation of an unresolved controversy. Free Radic Biol Med 5:81,1988 11. Rosen GM, Finkelstein E: Use of spin traps in biological systems. Adv Free Radic Biol Med 1:345,1985 12. Rosen GM, Freeman BA: Detection of superoxide generated by endothelial cells. Proc Natl Acad Sci USA 81:7269,1984 13. Zweier JL, Kuppusamy P, Lutty GA: Measurement of endothelial cell free radical generation: Evidence for a central mechanism of free radical injury in post-ischemic tissues. Proc Natl Acad Sci USA 854046,1988 14. Peterson MW: Neutrophil cathepsin G increases transendothelia1 albumin flux. J Lab Clin Med 113:297,1989 15. Bailar JC Jr, Jones EM: Trioxalato salts (trioxalatoaluminiate,-ferrate,-chromiate-cobaltiate).Inorganic Synthesis 1:35, 1939 16. Busch C, Cancilla PA, DeBault LE, Goldsmith JC, Owen WG: Use of endothelium cultured on microcarriers as a model for the microcirculation. Lab Invest 47:498,1982 17. Hassan HM, Fridovich I: Intracellular production of superoxide radical and hydrogen peroxide by redox active compounds. Arch Biochem Biophys 196:385,1979 18. Freeman BA, Crapo JD: Hyperoxia increases oxygen radical production in rat lungs and lung mitochondria. J Biol Chem 256:10986,1981 19. Smith L L The response of the lung to foreign compounds that produce free radicals. Annu Rev Physiol48681,1986 20. Hassett DJ, Britigan BE, Svendsen T, Rosen GM, Cohen MS: Bacteria form intracellular free radicals in response to paraquat and streptonigrin: Demonstration of the potency of hydroxyl radical. J Biol Chem 262:13404,1987 21. Finkelstein E, Rosen GM, Rauckman EJ: Spin trapping of superoxide. Mol Pharmacol16:676,1979 22. Yamazaki I, Piette LH, Grover TA Kinetic studies on trapping of superoxide and hydroxyl radicals generated in NADPHcytochrome P-450 reductase-paraquat system: Effect of iron chelates. J Biol Chem 265:652,1990 23. Britigan BE, Rosen GM, Chai Y, Cohen MS: Do human neutrophils make hydroxyl radical? Detection of free radicals generated by human neutrophils activated with a soluble or particulate stimulus using electron paramagnetic resonance spectrometry. J Biol Chem 261:4426,1986 24. Lloyd RV, Mason RP: Evidence against transition metalindependent hydroxyl radical generation by xanthine oxidase. J Biol Chem 265:16733,1990 25. Rosen GM, Britigan BE, Cohen MS, Ellington SP, Barber MJ: Detection of phagocyte-derived free radicals with spin trap- From www.bloodjournal.org by guest on January 21, 2015. For personal use only. ENDOTHELIAL CELL OXYGEN RADICAL GENERATION ping techniques: Effects of temperature and cell metabolism. Biochim Biophys Acta 969:236,1988 26. Samuni A, Black CDV, Krishna CM, Malech HL, Bernstein EF, Russo A: Hydroxyl radical production by stimulated neutrophils reappraised. J Biol Chem 263:13797,1988 27. Shasby DM, Lind SE, Shasby SS, Goldsmith JC, Hunninghake GW: Reversible oxidant-induced increases in albumin transfer across cultured endothelium: Alterations in cell shape and calcium homeostasis. Blood 65:605,1985 28. Weiss SJ, Young J, LoBuglio AF, Slivka A, Nimeh NF: Role of hydrogen peroxide in neutrophil-mediated destruction of cultured endothelial cells. J Clin Invest 68714,1981 29. Ager A, Gordon J L Differential effects of hydrogen peroxide on indices of endothelial cell function. J Exp Med 159:592,1984 30. Kvietys PR, haven W, Bacon BR, Grisham MB: Xanthine oxidase induced injury to endothelium: Role of intracellular iron and hydroxyl radical. Am J Physiol Heart Circ Physiol257H1640, 1989 31. Klebanoff SJ, Waltersdorph AM: Inhibition of peroxidasecatalyzed reactions by deferoxamine. Arch Biochem Biophys 264: 600,1988 32. Vissers MCM, Fantone J: Inhibition of hypochlorous acidmediated reactions by desferrioxamine. Implications for the mechanism of cellular injury by neutrophils. Free Radic Biol Med 8:331, 1990 33. Barankiewicz J, &hen A Impairment of nucleotide metabolism by iron-chelating deferoxamine. Biochem Pharmacol 36: 2343,1987 34. Morehouse KM, Flitter WD, Mason RP: The enzymatic oxidation of desferal to a nitroxide free radical. FEBS Lett 222246, 1987 35. Beckman JS, Beckman TW, Chen J, Marshall PA, Freeman B A Apparent hydroxyl radical production by peroxynitrite: Implications for endothelial cell injury from nitric oxide and superoxide. Proc Natl Acad Sci USA 87:1620,1990 36. Root RK, Cohen MS: The microbicidal mechanisms of human neutrophils and eosinophils. Rev Infect Dis 3:565, 1981 37. Diguiseppi J, Fridovich I: Ethylene from 2-keto-4-thiomethyl butyric acid: The Haber-Weiss reaction. Arch Biochem Biophys 295:323,1980 38. Britigan BE, Pou S, Rosen GM, Lilleg DM, Buettner G R Hydroxyl radical is not a product of the reaction of xanthine oxidase and xanthine. The confounding problem of adventitious iron bound to xanthine oxidase. J Biol Chem 265:17533,1990 39. Rosen H, Klebanoff SJ: Bactericidal activity of a superoxide generating system: A model for the polymorphonuclear leukocyte. J Exp Med 149:27,1978 40. Babior BM, Cumutte JT, Kipnes RS: Biological defense mechanisms. Evidence for the participation of superoxide in bacterial killing by xanthine oxidase. J Lab Clin Med 85:235,1975 41. Samuni A, Carmichael AJ, Russo A, Mitchell JB, Riesz P On the spin trapping and ESR detection of oxygen-derived radicals generated inside cells. Proc Natl Acad Sci USA 83:7593, 1986 42. Britigan BE, Coffman TJ, Adelberg DR, Cohen MS: Mononuclear phagocytes have the potential for sustained hydroxyl radical production: Use of spin trapping techniques to investigate mononuclear phagocyte free radical production. J Exp Med 168: 2367,1988 43. Pou S, Rosen GM, Britigan BE, Cohen MS: Intracellular spin trapping of 0,-centered radicals generated by human neutrophils. Biochim Biophys Acta 991:459, 1990 44. Rosen H, Klebanoff SJ: Hydroxyl radical generation by 707 polymorphonuclear leukocytes measured by electron spin resonance spectroscopy. J Clin Invest 64:1725,1979 45. Peterson MW, Stone P, Shasby DM: Cationic neutrophil proteins increase transcandothelial albumin movement across a cultured endothelium. J Appl Physiol621521,1987 46. Rodell TC, Cheronis JC, Ohnemus CL, Piermattel DJ, Repine JE: Xanthine oxidase mediates elastase-induced injury to isolated lungs and endothelium. J Appl Physiol67:2159, 1987 47. Matsubara T, Ziff M: Superoxide anion release by human endothelial cells: Synergism between a phorbol ester and a calcium ionophore. J Cell Physiol 127:207,1986 48. Rosen GM, Hassett DJ, Yankaskas JR, Cohen MS: Detection of free radicals as a consequence of dog tracheal epithelial cellular xenobiotic metabolism. Xenobiotica 19:635,1989 49. Kohen R, Chevion M: Cytoplasmic membrane is the target organelle for transition metal mediated damage induced by paraquat in Escherichia coli. Biochemistry 27:2597, 1988 50. Korbashi P, Kohen R, Katzhendler J, Chevion M: Iron mediates paraquat toxicity in Escherichia coli. J Biol Chem 261: 12472,1986 51. Sandy MS, Moldeus P, Ross D, Smith M T Cytotoxicity of the redox cycling compound diquat in isolated hepatocytes: Involvement of hydrogen peroxide and transition metals. Arch Biochem Biophys 259:29,1987 52. Bagley AC, Krall J, Lynch RE: Superoxide mediates the toxicity of paraquat for Chinese hamster ovary cells. Proc Natl Acad Sci USA 83:3189,1986 53. Krall J, Bagley AC, Mullenbach GT, Hallewell RA, Lynch RE: Superoxide mediates the toxicity of paraquat for cultured mammalian cells. J Biol Chem 263:1910,1988 54. Britigan BE, Coffman TJ, Buettner GR: Spin trap evidence for the lack of significant hydroxyl radical production during the respiration burst of human phagocytes using a spin adduct resistant to superoxide mediated destruction. J Biol Chem 265:2650, 1990 55. Britigan BE, Rosen GM, Thompson BY, Chai Y, Cohen MS: Stimulated neutrophils limit iron-catalyzed hydroxyl radical formation as detected by spin trapping techniques. J Biol Chem 261: 17026,1986 56. Winterbourn CC: Myeloperoxidase as an effective inhibitor of hydroxyl radical production: Implications for the oxidative reactions of neutrophils. J Clin Invest 78:545, 1986 57. Britigan BE, Hassett DJ, Rosen GM, Hamill DR, Cohen MS: Neutrophil degranulation inhibits potential hydroxyl radical formation: Differential impact of myeloperoxidase and lactoferrin release on hydroxyl radical production by iron supplemented neutrophils assessed by spin trapping. Biochem J 264:447,1989 58. Pou S, Cohen MS, Britigan BE, Rosen GM: Spin trapping and human neutrophils: Limits of detection of hydroxyl radical. J Biol Chem 264:12299,1989 59. Rosen GM, Pou S, Britigan BE, Cohen MS: Application of spin traps to biological systems. Free Radic Res Commun 9:187, 1990 60. Smedley LA, Tonneson MG, Sandhaus RA, Haslett C, Guthrie LA, Johnston RB Jr, Henson PM, Worthen GS: Neutrophil-mediated injury to endothelial cells: Enhancement by endotoxin and essential role of neutrophil elastase. J Clin Invest 77:1233, 1986 61. Henson PM, Johnston RB Jr: Tissue injury in inflammation: Oxidants, proteinases cationic proteins. J Clin Invest 79:669,1987 62. Peterson MW, Gruenhaupt D, Shasby DM: Neutrophil cathepsin G increases calcium flux and inositol polyphosphate production in cultured endothelial cells. J Immunol 143:609, 1989 From www.bloodjournal.org by guest on January 21, 2015. For personal use only. 1992 79: 699-707 Insight into the nature and site of oxygen-centered free radical generation by endothelial cell monolayers using a novel spin trapping technique BE Britigan, TL Roeder and DM Shasby Updated information and services can be found at: http://www.bloodjournal.org/content/79/3/699.full.html Articles on similar topics can be found in the following Blood collections Information about reproducing this article in parts or in its entirety may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://www.bloodjournal.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://www.bloodjournal.org/site/subscriptions/index.xhtml Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.
© Copyright 2018