Extracellular superoxide dismutase , nitric oxide , and central nervous system 02 toxicity

Although reactive 02 species appear to participate in central nervous system (CNS) 02 toxicity, the exact roles of different reactive 02 species are undetermined. To study the contribution of extracellular superoxide anion (02 ) to CNS 02 toxicity we constructed transgenic mice overexpressing human extracellular superoxide dismutase (ECSOD; superoxide:superoxide oxidoreductase, EC 1.15.1.1) in. the brain. Remarkably, when exposed to 6 atm (1 atm = 101.3 kPA) of hyperbaric oxygen for 25 min, transgenic mice demonstrated higher mortality (83%) than nontransgenic littermates (33%; P < 0.017). Pretreatment with diethyldithiocarbamate, which inhibits both ECSOD and Cu/Zn superoxide dismutase (Cu/Zn SOD) activity, increased resistance to CNS 02 toxicity, in terms of both survival (100% in triasgenics and 93% in nontransgenics) and resistance to seizures (4-fold increase in seizure latency in both transgenic and nontransgenic mice; P < 0.05). Thus, apparendly protects against CNS 02 toxicity. We hypothesized that °2 decreased toxicity by inactivating nitric oxide (NO'). To test this, we inhibited NO' synthase (EC 1.14.23) with N"-nitro-L-arginine to determine whether NO contributes to enhanced CNS 02 toxicity in transgenic mice. N"-nitro-L-arginine protected both transgenic and nontransgenic mice against CNS 02 toxicity (100% survival and a 4-fold delay in time to first seizure; P < 0.05), as well as abolishing the difference in sensitivity to CNS 02 toxicity between transgenic and nontransgenic mice. These results implicate NO' as an important mediator in CNS 02 toxicity and suggest that ECSOD increases CNS 02 toxicity by inhibiting 02-mediated inactivation of NO'. Several important antioxidant enzymes, such as catalase (hydrogen-peroxide:hydrogen-peroxide oxidoreductase, EC 1.11.1.6), glutathione peroxidase (glutathione:hydrogenperoxide oxidoreductase, EC 1.11.1.9), and the superoxide dismutases (SOD; superoxide:superoxide oxidoreductase EC 1.15.1.1) are known to exist within cells. However, extracellular fluids and the extracellular matrix contain only small amounts of these enzymes. It has been argued that the major extracellular antioxidants are proteins that sequester transition metals so that they cannot act as catalysts to form the highly reactive hydroxyl radical (OH*) (1). Other extracellular antioxidants include radical scavengers and inhibitors of lipid peroxidation, such as ascorbic acid, uric acid, and a-tocopherol (1). The relative lack of extracellular antioxidant enzymes may reflect the possible function of extracellular reactive oxygen species as bioeffector molecules (1); however, this may also lead to greater susceptibility to extracellular oxidant stresses. The enzyme extracellular superoxide dismutase (ECSOD) exists only at low concentrations in extracellular fluids and is not thought to function as a bulk scavenger of O2 (1). The physiologic role of ECSOD in vivo remains a mystery. ECSOD is a tetrameric Cu/Zn-containing glycoprotein with a subunit molecular weight of 30,000 (2, 3). Biochemical data suggest that ECSOD binds to heparan sulfate proteoglycans on endothelial cells, where it has been speculated to serve as a "protective coat" (4, 5). Endothelial cells secrete both O(reviewed in ref. 6) and endothelium-derived relaxing factor, putatively identified as nitric oxide (NO') (7). In addition to vasoregulation, NO functions to regulate neurotransmission (8), although it can be toxic to neurons in some situations (9). Because 0° is known to inactivate NO'induced vasorelaxation (10-14), one possible function for ECSOD may be to protect NO released from cells from O2j-mediated inactivation. In support of the potential importance of ECSOD acting as a "protector" of NO, a recent report has shown that blood pressure in spontaneously hypertensive rats could be lowered by giving them an intravenous injection of a SOD construct containing a high affinity to heparan sulfate on endothelial cells (15). In addition to inactivating NO'-induced vasorelaxation, the reaction of °2 with NO' also produces a potentially toxic intermediate in the form of the peroxynitrite anion (ONOO-) (16-19). While ONOOhas recently been shown to be capable of stimulating cGMP formation, it is a much less potent stimulator than NO' itself (20). Hyperbaric oxygen exposures equal to or greater than 2 atm (1 atm = 101.3 kPa) are well known to be toxic to the central nervous system (CNS), leading to generalized convulsions and eventually to death (21). The mechanisms of this toxicity have been proposed to involve increased production of intracellular reactive oxygen species like °2 and H202. The role of extracellular °2 and NO' have not yet been studied. To begin addressing the roles of extracellular 0°2 ECSOD, and NO in CNS 02 toxicity, we constructed transgenic mice that express elevated levels ofECSOD in the brain. These mice were then exposed to hyperbaric oxygen to see what effect this had on resistance to CNS 02 toxicity. MATERIALS AND METHODS Materials. Concanavalin A-Sepharose, Nw-nitro-L-arginine, equine cytochrome c, and diethyldithiocarbamate were purchased from Sigma. The plasmid pMSG was purchased from Pharmacia LKB. pKS was purchased from Stratagene. The human 3-actin promoter was provided by Larry Kedes (University ofSouthern California, Los Angeles). Restriction endonucleases were purchased from New England Biolabs. Animals. (C57BL/6 x C3H)F1 and CD1 mice were purchased from Charles River Breeding Laboratories. Abbreviations: SOD, superoxide dismutase; CNS, central nervous system; ECSOD, extracellular SOD. tTo whom reprint requests should be addressed. 9715 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. §1734 solely to indicate this fact. Proc. Natl. Acad. Sci. USA 89 (1992) Construction of Human ECSOD Expression Vector. The ECSOD expression vector (Fig. 1) was constructed as follows: The entire human ECSOD cDNA fragment (22, 23) flanked by EcoRI restriction sites was converted and treated with mung bean nuclease to form blunt ends, ligated to Sal I linkers, digested with Sal I, and then inserted into the Sal I site of the human f3-actin expression vector pH/8APr-1. The EcoRI/HindIll fragment of the resultant plasmid containing the human P-actin promoter, intron, and ECSOD cDNA was isolated. In addition, the Hpa I site of simian virus 40 (SV40) at position 2666 in plasmid pMSG was converted to a HindIII site by linker ligation and the HindIII/Pst I fragment containing the polyadenylylation site of the SV40 early region was isolated. These two DNA fragments were then ligated to an EcoRI plus Pst I-digested pKS vector. The EcoRI/Xba I fragment containing the entire expression construct free of plasmid sequences was isolated and used to establish transgenic mice. All of the recombinant DNA procedures were done according to established methods (24). Development of Transgenic Mice. Purified DNA at 2.5 jig/ml in 5 mM Tris HCl, pH 7.4/0.1 mM EDTA was injected into the pronuclei of fertilized eggs isolated from mice [(C57BL/6 x C3H)F1 x (C57BL/6 x C3H)F1]. Mouse eggs surviving microinjection were then implanted into the oviducts of pseudopregnant foster mothers (CD1) following procedures described by Hogan et al. (25). Mice carrying the transgene were identified by Southern blot analysis of tail DNA probed with the entire human ECSOD cDNA. Transgenic founders were bred with (C57BL/6 x C3H)F1 to produce offspring for further studies. Northern Blot Analysis. Transgenic mice and nontransgenic littermates were killed with an overdose of pentobarbital. Tissues were quickly excised and frozen in liquid nitrogen. Total RNA was then isolated by the CsCl procedure as described (24). Twenty micrograms of total RNA from the tissues of transgenic mice and nontransgenic littermates and an RNA ladder was then denatured with glyoxal, electrophoresed through a 1.2% agarose gel, and blotted to nitrocellulose as described (24). The blots were then probed with the entire human ECSOD cDNA. Separation ofSOD Isoenzymes by Concanavalin A-Sepharose Chromatography. Tissues from three mice were weighed, combined, and homogenized in 10 vol of ice-cold 50 mM potassium phosphate (pH 7.4) with 0.3 M KBr/3 mM diethylenetriaminepentaacetic acid/0.5 mM phenylmethylsulfonyl fluoride. Separation of ECSOD from Cu/Zn SOD and Mn SOD was accomplished by passing tissue homogenates over a concanavalin A-Sepharose column as described (26). SOD Activity. ECSOD activity and total SOD activity (Cu/Zn SOD and Mn SOD) remaining after ECSOD extraction were measured by inhibition of cytochrome c reduction at pH 10 as described (27). Total protein was determined by the BCA protein assay (Pierce). Oxygen Exposures. Mice (7-8 weeks old) were exposed to hyperbaric oxygen five at a time in a small-animal chamber (Bethlehem, PA). After flushing the chamber with pure 02, compression to 50m (6 atm) was performed within 5 min. The 02 concentration in the chamber was monitored continuously with a Servomex 02 analyzer (model 572; Sybron, Norwood,