[Strategies for bioradical detection in pharmacology].
Explore the source record for details and available documents.
Biomedical subjects
Publications and source records attributed to E Okabe.
Explore the source record for details and available documents.
In recent years, reactive oxygen species have been implicated in the pathogenesis of a wide variety of disorders. Although the existence of reactive oxygen intermediates in drug metabolism can be inferred from end product analysis or from the effects of antioxidants or enzymes such as superoxide dismutase, only the technique of electron spin resonance (ESR) allows the direct detection of these highly reactive species. However, some free radical species cannot be detected by ESR due to their extremely short half-lives, which result in low steady-state concentrations of the radicals or to short radical relaxation times, which lead to a very broad line. These facts made recent development of spin-trapping and chemiluminescence techniques are widely used to detect free radicals. The goal of this paper is to introduce the various assays available for measurement of reactive oxygen species in biological models. This paper will focus on two topics: (1) the spin-trapping/ESR technique in vitro and vivo and (2) the chemiluminescence-optical biosensor application of this technique, a very sensitive method that has the advantage of being able to provide continuous, online, nondestructive monitoring of reactive oxygen species.
We investigated the role of reactive oxygen intermediates generated from photoactivation of xanthene dye rose bengal on skeletal sarcoplasmic reticulum (SR) function, which plays a major role in the regulation of intracellular Ca++ and thereby in the generation of force. We used SR microsomes of canine masseter muscle as a model system in which to explore the effect of oxidation by determining oxalate-supported Ca++ uptake, Ca++, Mg++-adenosine triphosphatase (Ca++-ATPase) activity and Ca++ permeability of the SR vesicles. Skeletal SR vesicles exposed to rose bengal (50 nM) illuminated at 560 nm resulted in significant inhibition of Ca++ uptake velocity and Ca++-ATPase activity and in stimulation of Ca++ permeability. The observed effect afforded by illuminated rose bengal was dependent on intensity of light. Most reactive oxygen species scavengers tested had no protective effect; histidine (a powerful quenching agent for singlet oxygen), however, significantly protected the effect of illuminated rose bengal on Ca++ uptake velocity and Ca++-ATPase activity. The illumination of rose bengal also caused histidine-inhibitable loss of total sulfhydryl groups of SR. The increased Ca++ permeability elicited by illuminated rose bengal was blunted by a cocktail of histidine-catalase, but not by histidine alone. Generation of reactive oxygen species (singlet oxygen, superoxide and hydroxyl radical) from photoactivation of rose bengal was studied by electron spin resonance spectroscopy by use of the spin trap 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) and 2,2,6,6-tetramethylpiperidine (TEMP). We found that illumination of rose bengal formed a 1:2:2:1 quartet, characteristic of the hydroxyl radical-DMPO spin adduct, which was effectively blunted by hydroxyl radical scavenger, dimethyl sulfoxide, and by superoxide scavenger, superoxide dismutase. The results of electron spin resonance study also showed that singlet oxygen was produced by photoactivation of rose bengal was detected as singlet oxygen-TEMP product (TEMPO); 2,2,6,6-tetramethylpiperidine-N-oxyl). The formation of TEMPO signal was strongly inhibited by histidine. Similarly, we could detect hydrogen peroxide production from illuminated rose bengal. It is suggested that photoactivation of rose bengal generated singlet oxygen, superoxide, hydrogen peroxide and hydroxyl radical, and the data obtained from the present study indicate that singlet oxygen, rather than superoxide, hydrogen peroxide and hydroxyl radical, to be the active agent in the Ca++ transport system of SR; the observed effect of singlet oxygen may be due to sulfhydryl group oxidation. Our results are also consistent with the view that singlet oxygen does not appear to be an exclusive species that increases Ca++ permeability of SR vesicles, but the increased Ca++ permeability may be caused in part by hydrogen peroxide as well as singlet oxygen.
Sevoflurane is well known to cause depression of cardiovascular function, but detailed information on its actions on the contractility and reactivity of blood vessels is lacking. We have assessed therefore the direct effect of this anaesthetic on the functional reactivity of isolated rabbit mesenteric artery ring preparations. We found that contractions of endothelium intact rings induced by noradrenaline and phenylephrine were significantly attenuated by 4% sevoflurane; the observation that the maximal tension generation decreased without a significant reduction in pD2 is consistent with the view that receptor dysfunction was not involved. The effect of sevoflurane was not affected by NG-monomethyl-L-arginine. Sevoflurane 4% also produced attenuation of noradrenaline-induced contractions of endothelium denuded ring preparations. The contractions of endothelium denuded ring preparations produced by noradrenaline in Ca(2+)-free media in the presence of K+ were not affected by 4% sevoflurane, but sevoflurane depressed external Ca(2+)-dependent contractions. When vasodilators (acetylcholine and nitroglycerin) were added to the bathing media in the presence of 2% sevoflurane, the endothelium-dependent relaxation produced by acetylcholine, but not the endothelium-independent relaxation produced by nitroglycerin, was attenuated; superoxide dismutase inhibited the effect of sevoflurane on endothelium-dependent relaxation. These results are consistent with the view that sevoflurane inhibits alpha adrenoceptor-mediated contractions of isolated rabbit mesenteric artery ring preparations; this effect may be caused by reduced Ca2+ influx, as estimated from the effect on external Ca(2+)-dependent contractions, but is unlikely to be caused by reduced Ca2+ release from the sarcoplasmic reticulum of vascular smooth muscle, as estimated from noradrenaline-induced contractions in Ca(2+)-free bathing media.(ABSTRACT TRUNCATED AT 250 WORDS)
To understand the effect of oxygen free radicals on Ca(2+)-ATPase, we used sarcoplasmic reticulum (SR) microsomes of canine masseter muscle as a model system in which to explore the effects of oxidation on a biological membrane, and we investigated the effect of hydroxyl radicals (.OH) generated from Fenton's reagent (H2O2/FeSO4). H2O2 (10 mM) alone had no effect on Ca(2+)-ATPase activity; in the presence of FeSO4 (0.2 mM), H2O2 inhibited the enzyme activity. Oxygen free radical species generated from H2O2/FeSO4 under the conditions employed in the Ca(2+)-ATPase assay were verified by highly sensitive electron spin resonance spectroscopy and the spin-trap 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) in the absence of SR vesicles; the 1:2:2:1 quartet (AN = A beta H = 1.49 mT), characteristic of the DMPO-OH spin adduct, was observed. The Ca(2+)-ATPase activity was inversely correlated with the calculated signal intensity of DMPO-OH, which is indicative of the amount of .OH radical generated. The effect of Fenton's reagent was effectively inhibited by catalase, dimethylsulfoxide, and dimethylthiourea; the effect was also inhibited by sulfhydryl (SH) group reducing agents, cysteine and dithiothreitol. The SH group modifying agents, p-chloromercuric benzoate and 5,5'-dithiobis(2-nitrobenzoic acid) depressed Ca(2+)-ATPase activity; the effects of the SH group modifying agents used were potentiated in the presence of Fenton's reagent. It is suggested that .OH radical-induced oxidant injury may be caused primarily by modification of the key SH group(s) on the ATPase molecule of masseter muscle SR vesicles.
The nature of neurogenic relaxation was investigated in ring preparations of canine lingual artery. In all experiments, the preparations were previously treated with guanethidine (5 x 10(-6) M) to block neurogenic constrictor responses. In the presence of norepinephrine (10(-5) M) to induce tone, electrical stimulation (10 V, 4 to 16 Hz, for 45 sec) produced relaxation of the rings in an endothelium-independent fashion. The relaxant response in endothelium-denuded rings was not changed by propranolol (10(-5) M), and atropine (10(-5) M) did not affect the relaxation elicited by electrical stimulation in endothelium-intact rings. NG-monomethyl-L-arginine (10(-4) M) or NG-nitro-L-arginine methyl ester (10(-4) M), a nitric oxide (NO) synthase inhibitor, had no effect on the electrical stimulation-induced relaxation of endothelium-denuded rings. Human calcitonin gene-related peptide (CGRP)-(8 - 37) (2 x 10(-8) M), a CGRP1-receptor antagonist, inhibited neurogenic relaxation of endothelium-denuded rings; substance P (10(-6) M) failed to mimic the observed effect of electrical stimulation. The demonstrated effect of electrical stimulation was inhibited by glibenclamide (10(-5) M), but not tetraethylammonium (2 x 10(-4) M); glibenclamide abolished the relaxation in response to exogenous CGRP or the ATP-sensitive K+ channel opener cromakalim (10(-6) M) in endothelium-denuded rings. Moreover, tetrodotoxin (3.13 x 10(-6) M) inhibited the relaxation of endothelium-denuded rings induced by electrical stimulation. The relaxation was selectively inhibited when endogenous CGRP had been depleted from perivascular nerves by capsaicin (10(-6) M). These results suggest that CGRP, but not NO, released from non-adrenergic non-cholinergic nerves by electrical stimulation produces relaxation of canine lingual artery that is mediated by activation of CGRP1 receptors.
An adequate blood supply to the dental pulp is essential to the health of the tooth. A recent concept is that repeated stimulation of sensitive teeth may induce pulpal changes; this could occur through induction of neurogenic inflammation and alteration of pulpal blood flow. One possibility is that production of oxygen-derived free radicals at sites of inflammation contributes to alterations in local blood flow. The first target of free radicals, generated in several pathological processes, is the vascular system (essentially the endothelium). Although the exact mechanism by which free radicals induce changes in vascular conductance is still uncertain, they may act directly on vascular smooth muscle or modify vascular tone by interacting with the production and/or biological activity of endogenous vasoactive mediators. Recent data indicate that the oxygen-derived, free radical-generating system can decrease pulpal blood flow in the dog via endothelial dysfunction when applied locally in deep dentinal cavities. In addition to the part played by oxygen-derived free radicals, the measurement of pulpal blood flow and the effects of endogenous vasoactive substances on flow are discussed.
To understand the contractile response to histamine, drug-induced tension developments were measured in intact and staphylococcal alpha-toxin-treated permeabilized smooth muscle preparations of rabbit lingual artery. Histamine produced an endothelium-independent contraction; the observed effect was antagonized by diphenhydramine and was attenuated by nifedipine. Histamine produced only transient contraction in the Ca(2+)-free bathing media. In permeabilized preparations, histamine-induced contraction was abolished by ryanodine (in the presence of caffeine) or prior repeated application of caffeine; however, contraction produced by IP3 was still observed under the same experimental condition as that of the histamine study. Histamine- or caffeine-induced contraction was abolished in the permeabilized preparations by prior repeated application of IP3; caffeine or IP3 produced contraction after repeated application of histamine. Ryanodine in the presence of histamine was ineffective on caffeine-induced contraction, suggesting that histamine by itself may not have the ability to act on caffeine-sensitive Ca2+ release channels. Neomycin and H-7 completely abolished the histamine-induced contraction. Hence, it is suggested that histamine can contract the lingual artery via H1-receptor-coupled phospholipase C activation, and the contraction consists of a phasic response evoked by Ca2+ release from the IP3- and caffeine-sensitive Ca2+ storage site and the tonic response generated by the voltage-dependent influx of Ca2+. It is also postulated that IP3 produced by histamine may be able to produce Ca2+ release from only a part of the intracellular Ca2+ stores in the smooth muscle of rabbit lingual artery.
A 19-year-old male developed nephrotic syndrome during the course of chronic osteomyelitis complicating a traumatic suppurative arthritis of the knee. Renal biopsy revealed severe mesangial proliferative glomerulonephritis, and immunofluorescent microscopy demonstrated the presence of IgA. Nephrotic syndrome remitted during the treatment for chronic osteomyelitis, suggesting a close association of the two conditions.
In order to clarify the abnormalities of the coagulation and fibrinolysis system in patients with various renal diseases, we produced a new monoclonal antibody for FDP (fibrin/fibrinogen degradation product) D-dimer (D-D E72). We also established a new highly sensitive method of enzyme-linked immunosorbent assay (ELISA) for urinary FDP D-dimer using this monoclonal antibody. The urine from 110, patients with various renal diseases was investigated for the FDP D-dimer. The results are summarized as follows: 1) Urinary FDP D-dimer in normal subjects was 0.69 +/- 0.60 ng/ml. 2) The level of urinary FDP D-dimer in patients with primary nephrotic syndrome and in patients with chronic renal failure was significantly higher than that of normal subjects, whereas the urinary FDP D-dimer levels in patients with diabetes mellitus were higher than those of normal subjects. 3) In the CGN and NS groups there was a tendency for an increase in the level of urinary FDP D-dimer in more active forms of the disease. 4) A significant correlation between urinary FDP D-dimer and urinary protein in the CGN and NS groups was demonstrated. 5) In all of the renal diseases investigated in this study, the ratio of urinary FDP D-dimer to total FDP was less than 4%.
The purpose of this study was to investigate the effect of the oxidants hypochlorous acid (HOCl) and hydrogen peroxide (H2O2) on the vulnerability of the myocardium to reperfusion-induced arrhythmias following global ischemia. After a 15 min equilibration period with or without experimental intervention, isolated perfused rat hearts in the Langendorff mode were made globally ischemic for 5 min by cross-clamping the aortic line. No dysrhythmias were evoked upon reperfusion at the 5 min global ischemia time period. HOCl or H2O2 were added to the perfusate 5 min into the equilibration period with a total exposure of 10 min. Global ischemia was then induced for 5 min followed by 10 min of reperfusion. A dose-response curve for HOCl (50-200 microM) indicated the development of idioventricular rhythms, in a concentration-dependent way. Furthermore, coronary flow of the hearts exposed to 100 and 200 microM HOCl, at 5 min post-reperfusion, was decreased; methionine (10 microM to 1 mM), an accepted scavenger for HOCl, prevented the responses to 200 microM HOCl, in a concentration-dependent manner. All hearts exposed to 200 microM H2O2 developed ventricular dysrhythmias during the reperfusion period. Coronary flow increased after 5 min of exposure to 200 microM H2O2 and remained elevated during reperfusion. It is concluded that toxic oxygen derived products are capable of increasing the susceptibility of the myocardium to reperfusion induced arrhythmias, and that although the electrical responses to exposure to those two oxidants were similar, the effects on the vasculature were not the same.
To understand the direct involvement of hydroxyl radical (.OH) in the modification of functional reactivity in isolated rabbit lingual artery ring preparations, this study was undertaken to examine the effect of .OH generated from dihydroxy fumarate (DHF) plus Fe(3+)-ADP or from H2O2 plus FeSO4. When vasodilators (acetylcholine and nitroglycerin) were given after the .OH-generating system was removed from the organ chamber, the earlier .OH exposure produced an attenuation of the ring relaxation induced by acetylcholine but not that by nitroglycerin. Moreover, the earlier .OH exposure attenuated caffeine-induced contraction and depressed the phasic response, but potently enhanced the tonic response of norepinephrine-induced contraction. Both the enhanced tonic response of KCl-induced contraction produced by earlier .OH exposure and norepinephrine-induced contraction was inhibited by nisoldipine. These results are consistent with the view that .OH radicals can potentiate the voltage-dependent influx of Ca. It is also postulated that .OH may damage sarcoplasmic reticulum (SR) function in the smooth muscle cells, thus reducing Ca release from the SR (this may be reflected by the attenuation of the phasic response), and may selectively attenuate endothelium-dependent relaxation as opposed to endothelium-independent relaxation.
To determine whether sevoflurane alters endothelium-mediated vasodilation of vascular smooth muscle, isolated ring preparations of canine mesenteric arteries were suspended for isometric tension recordings in modified Krebs-Ringer bicarbonate solution at 37 degrees C. Following contraction with norepinephrine, cumulative concentration-response curves were generated using endothelium-dependent vasodilators (acetylcholine, bradykinin, and calcium ionophore A23187) or nitroglycerin. The relaxation produced by acetylcholine, bradykinin, or A23187 was impaired by sevoflurane (2.3 and 4.6 vol%); sevoflurane did not affect relaxation caused by nitroglycerin, which, in these vessels, acts by an endothelium-independent mechanism. Under the same experimental conditions as those used for the concentration-response relationship, electron spin resonance spin-trapping with 5,5-dimethyl-1-pyrroline N-oxide verified generation of hydroxyl radical from the sevoflurane-delivered bathing media; the generation of hydroxyl radical was inhibited by superoxide dismutase, a scavenger of superoxide anion radical, or by the powerful iron chelator deferoxamine. Furthermore, sevoflurane-induced impairment of the relaxation caused by the endothelium-dependent vasodilators used was significantly decreased by superoxide dismutase. These results indicate that superoxide anion radical and/or closely related species of oxygen free radicals, possibly hydroxyl radical, are involved in the observed effect of sevoflurane. We propose that sevoflurane selectively impairs endothelium-dependent relaxation in canine mesenteric arteries by an oxygen free radical mechanism, mainly due to inactivation of endothelium-derived relaxing factor.
1. The effect of E. coli lipopolysaccharide (LPS) on sympathetic neuro-effector transmission was studied in the rabbit mesenteric artery. The experiments were performed on artery rings isolated 5 or 20 h after intravenous treatment with LPS or saline as well as on artery rings isolated from non-treated rabbits (for assessment of the effect of in vitro preincubation with LPS). In most experiments, neural elements in the arteries were stimulated electrically (10 V, 2 ms, 1-32 Hz). 2. Preincubation with LPS (10 micrograms ml-1) for 5 or 20 h had no effect on the contraction responses of endothelium-intact artery rings to electrical stimulation. In contrast, in vivo intravenous pretreatment with LPS (10 micrograms) led to an inhibition of the contraction; LPS elicited this effect when injected 20 h, but not 5 h, before the experiment. The effect of LPS was eliminated in artery rings isolated from animals receiving an inhibitor of protein synthesis (actinomycin D or cycloheximide) before treatment with LPS. LPS (injected 20 h before the experiment) had no effect on the concentration-response curves for exogenous noradrenaline and tyramine in endothelium-intact artery rings. 3. The inhibition of electrically induced contractions produced by LPS treatment in endothelium-intact artery rings was attenuated by atropine and yohimbine, but not by phentolamine. Yohimbine plus atropine restored the depressed contraction to the normal level. Clonidine and acetylcholine mimicked the effect of LPS in endothelium-intact artery rings isolated from saline-treated animals. 4. When steady-state contractions were induced by 5 min of stimulation at 16 Hz, acetylcholine or clonidine reduced the contraction in endothelium-denuded artery rings from both saline-treated rabbits and animals receiving LPS 20 h before the experiment. The reduction produced by acetylcholine or clonidine of the contraction in artery rings from LPS-treated rabbits was significantly greater than in artery rings from saline-treated animals.5. These results suggest that treatment of rabbits with LPS inhibits noradrenaline release from sympathetic nerve endings via increased sensitivity of both prejunctional inhibitory muscarinic receptors and x2-adrenoceptors in mesenteric arteries. They also suggest that the effect of LPS is independent of endothelial cells but linked to protein synthesis.
To understand the direct involvement of free radicals causing reduction in endothelium-dependent relaxation of isolated canine coronary ring preparations, this study was undertaken to examine the effect of free radicals generated from dihydroxy fumarate (DHF) plus Fe(3+)-ADP or from H2O2 plus FeSO4. The vasodilators (acetylcholine, bradykinin, A23187, and nitroglycerin) were given after DHF/Fe(3+)-ADP or H2O2/FeSO4 was removed from the organ chamber. The earlier DHF/Fe(3+)-ADP exposure produced an attenuation of the relaxation of the rings induced by acetylcholine, bradykinin, or A23187 but not of the relaxation induced by nitroglycerin. The observed effect of previous DHF/Fe(3+)-ADP exposure was significantly protected in the vessels isolated from the dogs treated with alpha-tocopherol. In the experiments for assessing the effect of various scavengers, 1O2 scavenger histidine or iron chelator deferoxamine effectively protected the attenuation induced by DHF/Fe(3+)-ADP exposure of the relaxation elicited by acetylcholine; superoxide dismutase (SOD), catalase, or dimethyl sulfoxide (DMSO) had no effect on this system. Furthermore, the relaxation elicited by acetylcholine, but not nitroglycerin, was significantly attenuated by the earlier exposure to .OH generated by Fenton's reagent (H2O2+FeSO4); the attenuation was significantly protected by DMSO. These results are consistent with the view that .OH, 1O2, and/or iron-dependent reactive species selectively damage endothelium-dependent relaxation as opposed to endothelium-independent relaxation in endothelium-intact coronary ring preparations. It is also postulated that lipid peroxidation may be responsible for this effect.
The effect of intravenous injection of 10 micrograms of a lipopolysaccharide (endotoxin) extracted from E. coli to rabbits on the responses of isolated lingual arteries to des-Arg9-bradykinin (a specific kinins B1-receptor agonist) was studied. Endotoxin injection led to the appearance of the endothelium-independent contractile effect of des-Arg9-bradykinin in the arteries; endotoxin elicited this response when administered at 1, 5 or 20 hr before the experiment, in a time-interval dependent manner. The contractile response to des-Arg9-bradykinin of the arteries isolated from animals receiving endotoxin 20 hr before the experiment was attenuated by des-Arg9-[Leu8]-bradykinin (a specific inhibitor of kinins B1-receptor) or pretreatment of the animals with an inhibitor of protein synthesis (cycloheximide and actinomycin D). When compared with the effect of des-Arg9-bradykinin, bradykinin (a potent kinin B2-receptor, but weak B1-receptor stimulant) caused slight contraction of the arteries; however, this effect was not endotoxin-dependent and was not modified by des-Arg9-[Leu8]-bradykinin. Effect of in vitro preincubation with endotoxin of the arteries isolated from animals receiving saline 20 hr before the experiment was further studied. The preincubation (for 1 and 5 hr) with endotoxin of the arteries in the presence or absence of plasma had no effect on the sensitivity of the arteries to des-Arg9-bradykinin; the sensitivity was also unaffected in the presence or absence of endotoxin, thus suggesting that there is no interaction between endotoxin and some plasma-related factors with the appearance of the contraction in response to the kinin B1-receptor agonist in the arteries in vitro.(ABSTRACT TRUNCATED AT 250 WORDS)
The effect of in vitro exposure of sarcolemmal membrane (SL) vesicles to Gram-negative endotoxin lipopolysaccharides (LPS) was studied. LPS decreased the Na,K-ATPase activity of SL vesicles; this effect was inhibited by hydroxyl radical (.OH) scavengers such as dimethylthiourea and dimethyl sulfoxide, but not by superoxide dismutase, a scavenger of superoxide anion radicals or by the hydrogen peroxide scavenger catalase. ESR spin-trapping with 5,5-dimethyl-1-pyrroline N-oxide verified the generation of .OH from LPS itself under the conditions used; .OH generated from LPS was not affected by deferoxamine, a powerful iron chelator. The Na,K-ATPase activity was reduced by an .OH radical generating system consisting of dihydroxyfumarate and Fe3(+)-ADP. Furthermore, exposure of SL vesicles to LPS caused an increase in malondialdehyde formation. It can be concluded that LPS damages cardiac SL by an oxygen free radical mechanism by the generation of .OH, due to inhibition of Na,K-ATPase activity and peroxidation of lipids, and that the effect of LPS is not dependent on the presence of contaminating iron.
The effect of the plant alkaloid ryanodine on the cardiac sarcoplasmic reticulum (SR) function, which plays a major role in the regulation of intracellular calcium and thereby in the generation of force, was studied by determining oxalate-supported calcium uptake, steady-state calcium load, calcium permeability, intravesicular-free calcium and Ca,Mg-adenosine triphosphatase (ATPase) activity of "heavy" vesicles in the presence or absence of the oxygen-free radical-generating system. In vitro generation of oxygen-free radicals by xanthine oxidase (0.09 u/ml), acting on xanthine (25 microM) as a substrate, increased the permeability of the vesicles to calcium, determined by measuring net efflux of calcium after stopping pump-mediated fluxes, and decreased oxalate-supported calcium uptake and steady-state calcium load with no effect on Ca,Mg-ATPase activity. This effect of oxygen-free radicals was inhibited completely by superoxide dismutase, which eliminated completely superoxide anion radical production and caused an anticipated increase in hydrogen peroxide from the xanthine-xanthine oxidase reaction in our system. The xanthine-xanthine oxidase reaction decreased intravesicular-free calcium. The diminished level of intravesicular-free calcium, which was reflected by the decreased steady-state calcium load induced by oxygen-free radicals, was prevented by specific closure of the SR calcium release channel by ryanodine under established optimal conditions; under the same conditions, ryanodine also prevented superoxide dismutase-inhibitable reduction of calcium uptake induced by oxygen-free radicals in the presence or absence of oxalate. Ryanodine was without effect on Ca,Mg-ATPase activity by itself and had no effect on any of the changes in calcium permeability mediated by the generation of oxygen-free radicals under the experimental conditions used.(ABSTRACT TRUNCATED AT 250 WORDS)