Unique distribution profiles of glutathione S-transferases in regions of kidney, ureter, and bladder of rabbit.

BACKGROUND: Glutathione S-transferases detoxify a broad range of exogenous compounds, but are important also in the metabolism of endogenous compounds. Physiologically relevant substrates are the endoperoxide and hydroperoxide metabolites of arachidonic acid that play important roles in many tissues including the kidney. EXPERIMENTAL DESIGN: We used immunohistochemical and immunoblotting techniques in a systematic study of renal localization of four rabbit enzymes that represent three major mammalian cytosolic glutathione S-transferase classes, alpha, pi, and mu. RESULTS: The two alpha-class enzymes (rbGST alpha I, rbGST alpha II) were distributed discretely in kidney, ureter, and bladder, while pi and mu were widely distributed in the renal system. Immunohistochemical localization in paraffin sections with antibodies specific for rbGST alpha I or rbGST alpha II demonstrated that no compartment of the renal system contained both enzymes. Collecting ducts of the inner medulla and all epithelial cells of the kidney pelvis, ureter, and bladder contained rbGST alpha I. All cells lining proximal tubules contained rbGST alpha II. No other compartment of the renal system exhibited immunoreactivity with anti-rbGST alpha II. Antibody specific for pi reacted with cells lining nephrons, ureter, and bladder and with endothelial cells throughout the renal system. Localization of pi was most prominent in the collecting ducts of medulla and in the epithelial cells lining the kidney pelvis, ureter, and bladder. As anti-mu did not react in tissue sections, distribution of mu was determined by immunoblotting. Immunoblots of cytosolic preparations from whole kidney, cortex, medulla, and epithelia of ureter, bladder, and kidney pelvis were prepared and tested with each of the 4 antibodies. This second localization method confirmed the distribution data from tissue sections for rbGST alpha I, rb GST alpha II, and pi; also, it demonstrated that the staining observed in tissue was specifically for each enzyme. mu was detected in all the renal cytosolic preparations except those from the epithelium of the kidney pelvis. CONCLUSIONS: The discrete renal distribution of rbGST alpha I and rbGST alpha II and their distinct catalytic activities with prostaglandin substrates suggest important roles for these enzymes in prostaglandin-dependent renal functions.

The major alpha-class glutathione S-transferases of rabbit lung and liver. Primary sequences, expression, and regulation.

The complete primary structures of two distinct rabbit alpha-class glutathione S-transferase (GST) subunits, rbGST alpha I and rbGST alpha II, have been derived from cDNA sequences. Clones encoding rbGST alpha I were isolated from both hepatic and pulmonary cDNA libraries, whereas clones encoding rbGST alpha II were isolated only from the hepatic library. Immunochemical and peptide sequence data confirmed that rbGST alpha I corresponds to the 27-kDa alpha-class subunit purified from rabbit lung (Serabjit-Singh, C. J., and Bend, J. R. (1988) Arch. Bioch. Biophys. 267, 184-194). Expression of rbGST alpha II in liver but not in lung and expression of rbGST alpha I in both liver and lung was substantiated by Northern and immunochemical analyses. rbGST alpha I and rbGST alpha II are composed of 223 and 221 amino acids, respectively, and are 78% identical in amino acid sequence. Compared to published GST sequences, both proteins are most closely related to the human Ha subunit (greater than 80% identity). On the basis of sequence comparison and Northern and Southern analyses, we conclude that rbGST alpha I and rbGST alpha II are products of different genes that are independently regulated. Further, the regulatory elements of the alpha-class GST genes may be significantly different in the rabbit as compared to the rat, as evidenced by the lack of induction by phenobarbital of rabbit hepatic or pulmonary alpha-class GST subunits, enzymatic activity, or mRNA. This tissue- and species-dependent expression of the predominant class of cytosolic GST implies unique functions for each isozyme and may contribute to the differential susceptibility of tissues and animals to toxicants.

The mechanisms of inactivation of sulfite oxidase by periodate and arsenite.

The inactivation of sulfite oxidase, a molybdoenzyme containing the Mo cofactor, by arsenite and periodate was investigated. In contrast to ferricyanide (Gardlik, S., and Rajagopalan, K.V. (1991) J. Biol. Chem. 266, 4889-4895), neither of these reagents causes oxidation of the pterin ring of the Mo cofactor. Instead, inactivation by these reagents appears to involve attack on sulfhydryl groups at the active site of the enzyme. The inactivation of sulfite oxidase by arsenite was shown to be dependent on the presence of O2 and on the enzymatic oxidation of arsenite to arsenate. The inactivation was preventable by the presence of sulfite, or by the use of cytochrome c as the electron acceptor instead of O2. It is concluded that inactivation by arsenite is the result of arsenite displacement of Mo during enzymatic oxidation of arsenite to arsenate, when Mo transiently breaks its bond to protein or molybdopterin sulfhydryl(s) in order to provide a site for transfer of electrons to O2. Data indicate that arsenite is properly oriented to displace Mo only once every 20,800 turnovers, thus accounting for the slow rate of inactivation by this reagent. Inactivation of sulfite oxidase by periodate is believed to occur as the result of direct attack of periodate on the thiolate ligands of Mo, either those of the protein and/or molybdopterin, leading to Mo loss. Treatment of enzyme with even low levels of periodate resulted in loss of Mo and both sulfite:cytochrome c and sulfite:O2 activities. Molybdopterin of periodate-inactivated enzyme retained the ability to reconstitute nitrate reductase apoprotein in nit-1 extracts and the ability to reduce dichlorophenolindophenol, indicating that the pterin ring had not been oxidized.

Oxidation of molybdopterin in sulfite oxidase by ferricyanide. Effect on electron transfer activities.

The attenuation of the sulfite:cytochrome c activity of sulfite oxidase upon treatment with ferricyanide was demonstrated to be the result of oxidation of the pterin ring of the molybdenum cofactor in the enzyme. Oxidation of molybdopterin (MPT) was detected in several ways. Ferricyanide treatment not only abolished the ability of sulfite oxidase to serve as a source of MPT to reconstitute the aponitrate reductase in extracts of the Neurospora crassa mutant nit-1 but also eliminated the ability of sulfite oxidase to reduce dichlorobenzenoneindophenol after anaerobic denaturation. Additionally, the absorption spectrum of anaerobically denatured ferricyanide-treated molybdenum fragment of rat liver sulfite oxidase was typical of fully oxidized pterins. Ferricyanide treatment had no effect on the protein of sulfite oxidase or on the sulfhydryl-containing side chain of MPT. Quantitation of the ferricyanide reaction showed that 2 mol of ferricyanide were reduced per mol of MPT oxidized, yielding a fully oxidized pterin. These results corroborate the previously reported conclusion that the native state of reduction of MPT in sulfite oxidase is at the dihydro level (Gardlik, S., and Rajagopalan, K.V. (1990) J. Biol. Chem. 265, 13047-13054). As a result of oxidation of the pterin ring, the affinity of MPT for molybdenum is decreased, leading to eventual loss of molybdenum. Because the loss of molybdenum is slow, a population of sulfite oxidase molecules can exist in which molybdenum is complexed to oxidized MPT. These molecules retain sulfite:O2 activity, a function apparently dependent solely on the molybdenum-thiolate complex, yet have greatly decreased sulfite:cytochrome c activity, a function requiring heme as well as the molybdenum center of holoenzyme. These observations suggest that the pterin ring of MPT participates in enzyme function, possibly in electron transfer, directly in catalysis, or by controlling the oxidation/reduction potential of molybdenum.

The state of reduction of molybdopterin in xanthine oxidase and sulfite oxidase.

Methods have been devised to examine the spectral properties and state of reduction of the pterin ring of molybdopterin (MPT) in milk xanthine oxidase and the Mo-containing domain of rat liver sulfite oxidase. The absorption spectrum of the native pterin was visualized by difference spectroscopy of each protein, denatured anaerobically in 6 M guanidine hydrochloride (GdnHCl), versus a sample containing the respective apoprotein and other necessary components. The state of reduction of MPT was also probed using 2,6-dichlorobenzenoneindophenol (DCIP) to measure reducing equivalents/MPT, after anaerobic denaturation of the protein in GdnHCl in the presence or absence of Hg2+. In the case of xanthine oxidase the data indicate that the terminal sulfide ligand of Mo causes the reduction of a native dihydro form of MPT to the tetrahydro level. This reduction does not occur if Hg2+ is added prior to denaturation of the protein. Based on its observed behavior, the native MPT in the Mo cofactor of xanthine oxidase is postulated to exist as a quinonoid dihydropterin. Quantitation of DCIP reduction by MPT of Mo fragment of sulfite oxidase showed a two-electron oxidation of MPT, even when the Mo fragment was denatured in the presence of Hg2+ to prevent internal reduction reactions due to sulfhydryls or sulfide. Difference spectra of DCIP-treated versus untreated Mo fragment showed that MPT had been fully oxidized. These data indicate that the native MPT in sulfite oxidase must be a dihydro isomer different from that in xanthine oxidase.

A molybdopterin-free form of xanthine oxidase.

A previously unidentified fraction lacking xanthine:O2 activity has been isolated during affinity chromatography of bovine milk xanthine oxidase preparations on Sepharose 4B/folate gel. Unlike active, desulfo, or demolybdo forms of xanthine oxidase, this form, which typically comprises about 5% of an unfractionated enzyme solution, passes through the affinity column without binding to it, and is thus easily separated from the other species. The absorption spectrum of this fraction is very similar to that of the active form, but has a 7% lower extinction at 450 nm. Analysis of the fraction has shown that it is a dimer of normal size, but that it does not contain molybdenum or molybdopterin (MPT). The "MPT-free" xanthine oxidase contains 90-96% of the Fe found in active xanthine oxidase, and 100% of the expected sulfide. EPR and absorption difference spectroscopy indicate that the MPT-free fraction is missing approximately half of its Fe/S I centers. The presence of a new EPR signal suggests that an altered Fe/S center may account for the nearly normal Fe and sulfide content. Microwave power saturation parameters for the Fe/S II and Fe/S I centers in the MPT-free fraction are normal, with P1/2 equal to 1000 and 60 mW, respectively. The new EPR signal shows intermediate saturation behavior with a P1/2 = 200 mW. The circular dichroism spectrum of the MPT-free fraction shows distinct differences from that of active enzyme. The NADH:methylene blue activity of the MPT-free fraction is the same as that of active xanthine oxidase which exhibits xanthine:O2 activity, but NADH:cytochrome c and NADH:DCIP activities are diminished by 54 and 37%, respectively.