INTRODUCTION

Urate oxidase or uricase (urate: oxygen oxidoreductase, EC1.7.3.3), an enzyme that catalyzes the oxidation of uric acid toallantoin, occupies a pivotal position in the chain of enzymesresponsible for the metabolism of purines (Keilin, 1958).Degradation of purines to uric acid is common to all speciesbut the degradation of uric acid, however, varies from speciesto species. For example, bacteria and some marine inverte-brates degrade purines to uric acid, which is then oxidized toallantoin by urate oxidase (Keilin, 1958). Hydrolysis ofallantoin by allantoinase (EC 3.5.2.5) results in the formationof allantoic acid for further metabolism by allantoicase (EC3.5.3.4) to yield urea. The enzyme urease (EC 3.5.1.5) thenconverts urea to ammonia and carbon dioxide. Most mammals,with the exception of human and hominoid primates, containurate oxidase in their liver (Keilin, 1958; Freidman et al., 1985)and excrete allantoin as the end product of purine metabolism,as these animals do not contain allantoinase and allantoicase.

The loss of urate oxidase activity in humans and hominoidprimates, such as chimpanzee, gorilla and orangutan, due tononsense mutations in the urate oxidase gene, results in theexcretion of uric acid (Wu et al., 1989; Yeldandi et al., 1990;1991). In amphibia and fish liver, uric acid is degraded all theway to urea because all three enzymes, urate oxidase, allan-toinase and allantoicase, are present (Takada and Noguchi,1983; Noguchi et al., 1986; Hayashi et al., 1989).

In the rat, and in most other mammals that display urateoxidase activity and metabolize uric acid to allantoin, thisenzyme is associated specifically with the crystalloid corepresent within the peroxisomes in hepatic parenchymal cells(De Duve and Baudhuin, 1966; Shnitka, 1966; Tsukada et al.,1966; Lata et al., 1977; Tolbert, 1981; Usuda et al., 1988a,b;Alvares et al., 1992). Allantoin generated within the peroxi-some in mammalian hepatocytes is excreted in urine, but theprecise mechanism of allantoin transport out of the peroxi-some remains unclear. In amphibia and fish liver, thediscovery of the presence of all three enzymes of the uric acid

1073

On the basis of differential and density gradient centrifu-gation studies, the site of the uric acid degrading enzymes,urate oxidase and allantoinase, in amphibia was previouslyassigned to the hepatic peroxisomes. Using specific anti-bodies against frog urate oxidase and allantoinase, we haveundertaken an immunocytochemical study of the localiza-tion of these two proteins in frog liver and kidney, anddemonstrate that whereas urate oxidase is present in per-oxisomes, allantoinase is localized in mitochondria. Urateoxidase and allantoinase were detected by immunoblotanalysis in both frog liver and kidney. The subcellularlocalization of these two enzymes was ascertained byProtein A-gold immunocytochemical staining of LowicrylK4M-embedded tissue. Peroxisomes in frog liver parenchy-mal cells and kidney proximal tubular epitheliumcontained a semi-dense subcrystalloid core, which was

found to be the exclusive site of urate oxidase localization.Allantoinase was detected within mitochondria, but not inperoxisomes of hepatocytes or proximal tubular epithe-lium. No allantoinase was detected in the mitochondria ofnonhepatic parenchymal cells in liver and of the cells liningthe distal convoluted tubules of the kidney. These resultsdemonstrate that, unlike rat kidney peroxisomes whichlack urate oxidase, peroxisomes of frog kidney contain thisenzyme. Contrary to previous assumptions, these studiesalso clearly establish that urate oxidase and allantoinase,the first two enzymes involved in uric acid degradation, arelocalized in different subcellular organelles in frog liverand kidney.

Key words: urate oxidase, allantoinase, purine metabolism,peroxisome, mitochondrion

SUMMARY

Uric acid degrading enzymes, urate oxidase and allantoinase, are associated

with different subcellular organelles in frog liver and kidney

Nobuteru Usuda 1,*, Sueko Hayashi 1,†, Satoko Fujiwara 2, Tomoo Noguchi 2, Tetsuji Nagata 3,M. Sambasiva Rao 1, Keith Alvares 1, Janardan K. Reddy 1 and Anjana V. Yeldandi 1,‡

1Department of Pathology, Northwestern University Medical School, Chicago, Illinois 60611, USA2Department of Biochemistry, Kyushu Dental College, Kokura, Kitakyushu 803, Japan3Department of Anatomy and Cell Biology, Shinshu University School of Medicine, Matsumoto 390, Japan

*Present address: Department of Anatomy and Cell Biology, Shinshu University School of Medicine, Matsumoto 390, Japan†Present address: Kyushu Dental College, Kokura, Kitakyushu 803, Japan‡Author for correspondence

Journal of Cell Science 107, 1073-1081 (1994)Printed in Great Britain © The Company of Biologists Limited 1994

1074

degradation pathway led to studies on their subcellular distri-bution (Scott et al., 1969; Visentin and Allen, 1969).Following differential and density gradient centrifugation pro-cedures, it was reported that urate oxidase, allantoinase andallantoicase are associated with the peroxisome (Visentin andAllen, 1969). The subcellular compartmentalization of thesethree enzymes within hepatic peroxisomes thus apparentlyserved to optimize the metabolic degradation of uric acid tourea (Hayashi et al., 1989). Recent subcellular fractionationstudies of liver demonstrated that marine fishes, such assardine and mackerel, contain allantoinase both in the perox-isome and cytosol, whereas in fresh water fishes (e.g carp,bass) this enzyme is found only in the cytosol (Fujiwara et al.,1989).

In view of the current interest in the evolutionary loss of per-oxisomal enzymes responsible for the degradation of uric acid(Wu et al., 1989; Yeldandi et al., 1990; 1991), the postulatedrole of uric acid as a potent biological antioxidant (Ames et al.,1981), and the possible implications of the generation of theoxidant H2O2 as a result of uric acid metabolism by urateoxidase (De Duve and Baudhuin, 1966), it appeared necessaryto confirm visually the peroxisomal localization of urateoxidase, allantoinase and allantoicase in the frog liver byimmunocytochemical localization. Our objective in the presentstudy was to localize urate oxidase and allantoinase, the firsttwo enzymes responsible for uric acid degradation in the liverand kidney, using Protein A-gold immunocytochemical local-ization. The results show that urate oxidase is localized exclu-sively to the subcrystalloid core present within the peroxisomesof both parenchymal cells of liver and the proximal tubularepithelium of kidney. Allantoinase, however, was notdetectable within the peroxisomes in these cells, but waslocalized to the mitochondrion.

MATERIALS AND METHODS

AntibodiesUrate oxidase and allantoinase were purified from the liver of adult bullfrogs (Rana catesbeiana) and used to raise polyclonal monospecific anti-bodies in rabbits, as previously described (Fujiwara et al., 1987; Noguchiet al., 1986). Antibodies against rat liver catalase were generated andcharacterized as described elsewhere (Reddy and Kumar, 1979).

Preparation of tissue for Lowicryl K 4M embeddingAdult bull frogs and tadpoles (stages 51 to 60) were used in this study.The developmental stages were based on the criteria outlined by Taylorand Kollros (1946). For electron microscopic immunohistochemistry,pieces of liver and kidney from four frogs and three tadpoles were fixedby immersion in 4% paraformaldehyde/0.1% glutaraldehyde in 0.1 Msodium phosphate, pH 7.4, for 16 hours (Usuda et al., 1988b). Afterrinsing in 0.1 M sodium phosphate buffer (pH 7.4) containing 0.15 MNaCl and 0.1 M lysine for 1 hour, tissues were dehydrated in a gradedseries of ethanol and embedded in Lowicryl K4M at −20°C. Post-osmi-fication of the tissues was omitted. Ultrathin sections (0.1 µm) werecut on a Dupont-Sorvall MT2B ultramicrotome and mounted on nickelgrids with Formvar membrane. For light microscopic immunohisto-chemistry, tissues from three adult frogs and five tadpoles were fixedin 70% ethanol or paraformaldehyde at 4°C, and embedded in paraffin(Usuda et al., 1991). Sections, 5 µm thick, were subjected to stainingafter deparaffinization and rehydration.

Immunocytochemical labelingFor immunocytochemical localization, thin sections were incubatedon drops of 0.5 M Tris-HCl, pH 7.5, containing 0.05% Triton X-100,5 mg/ml bovine serum albumin and 0.15 M NaCl for 1 hour. Subse-quently, the sections were transferred onto drops of antibody solution(1:1000 dilution of antiserum in 0.05 M Tris-HCl, pH 7.5, containing0.05% Triton X-100, 5 mg of bovine serum albumin and 0.15 MNaCl) and incubated for 4 hours. They were then washed severaltimes with 0.05 M Tris-HCl buffer containing 0.1% Triton X-100 and0.15 M NaCl. Protein A-gold labeling, using 15 nm size gold particles,(EY Laboratories) was carried out as previously described (Bendayanand Reddy, 1982; Bendayan et al., 1983). After counterstaining withuranyl acetate and lead citrate, the sections were examined in a JEOL100 CEX electron microscope. A minimum of 20 randomly selectedelectron micrographs were obtained from each animal. To demon-strate the immunocytochemical specificity, control sections werereacted with nonimmune serum.

Immunoperoxidase stainingFor the demonstration of allantoinase, 3- to 4-µm thick ethanol- orparaformaldehyde-fixed, paraffin-embedded sections were stainedusing the avidin-biotin-peroxidase procedure (Usuda et al., 1991).Briefly, the deparaffinized sections were incubated in normal goatserum (1:50 dilution) for 20 minutes and then treated with anti-allan-toinase antibodies (1:300 dilution) for 2 hours followed by bio-tinylated goat anti-rabbit IgG (1:100 dilution) for 1 hour, and rabbitperoxidase anti-peroxidase complex (1:100 dilution) for 1 hour.Antibody solutions were made with 50 mM Tris-HCl, pH 7.5, con-taining 150 mM NaCl and normal goat serum. Sections were thenincubated in the medium for peroxidase which contained 0.05%3,3′-diaminobenzidine tetrahydrochloride and 0.01% H2O2 in 0.05 MTris-HCl, pH 7.2, for 10 minutes, and counterstained with 1%methyl green.

Subcellular fractionationFrogs were fasted overnight and their liver perfused in situ with0.25 M sucrose under light ether anesthesia. The liver and kidneyswere removed, finely chopped and homogenized in 5 vols of 0.25M sucrose/10 mM Tris-HCl, pH 7.5, using a Potter-Elvehjem tissuegrinder. The homogenate was centrifuged at 960 g for 10 minutesto remove nuclei and unbroken cells. The supernatant was carefullydecanted and centrifuged at 15,000 g for 20 minutes to separatecrude mitochondrial fraction (De Duve et al., 1955). The crudemitochondrial fraction was suspended in 20 ml of 0.25 Msucrose/10 mM Tris-HCl and washed twice to remove adsorbedcytosol. The homogenates, and the mitochondrial and 100,000 gsupernatant fractions, were processed for immunoblotting asdescribed below. Samples of the crude mitochondrial fractions werealso fixed in 4% paraformaldehyde/0.1% glutaraldehyde andembedded in Lowicryl K4M for immunocytochemical use, asdecribed above.

Gel electrophoresis and immunoblottingHomogenates of liver and kidney were prepared from three adult frogsand six tadpoles, and the proteins separated by SDS-PAGE using 10%polyacrylamide gels (Laemmli (1970). In addition, crude mitochon-drial and 100,000 g supernatant fractions were also subjected to SDS-PAGE. Immunoblotting with frog anti-urate oxidase, frog anti-allan-toinase, rat anti-catalase and rat anti-urate oxidase antibodies was thenperformed (Towbin et al., 1979). The antigen and antibody complexeswere identified by immunoperoxidase staining using 4-chloro-1-naphthol as reagent. The immunoblotting of homogenates wasrepeated at least four times.

N. Usuda and others

1075Urate oxidase and allantoinase localization

RESULTS

Localization of urate oxidase and catalase in theliver and kidneyIn earlier studies, using immunoblotting and immunocyto-chemical approaches, we showed that polyclonal antibodiesgenerated against rat liver urate oxidase recognized urateoxidase present in bovine, murine, canine and feline liver(Usuda et al., 1988a). In the present study, immunoblotanalyses of frog liver and kidney homogenates with anti-raturate oxidase antibodies failed to show cross-reactivity withthe frog enzyme (data not shown). Likewise, antibodiesgenerated against frog liver urate oxidase did not cross-reactwith rat hepatic urate oxidase (data not shown), but these froganti-urate oxidase antibodies recognized a single polypeptidein the liver and kidney of adult frogs and tadpoles (Fig. 1). Thesubunit molecular mass of urate oxidase in the liver and kidneyof frogs and tadpoles is approximately Mr 32,500, which issmaller than that of the rat urate oxidase subunit (Mr 35,000)(Alvares et al., 1992). Polyclonal antibodies raised against ratliver catalase reacted with frog liver and kidney catalase whenthe homogenates were analyzed by immunoblotting (Fig. 2),indicating that catalase is highly conserved. The molecularmass of catalase subunit in the rat and the frog is the same (Mr60,000).

Peroxisomes are few in number in the hepatic parenchy-mal cells and kidney proximal tubular epithelium of frog andtadpole. A small proportion of peroxisomes from these twocell types in the adult frog have a dense subcrystalloid corein the matrix; this core is not as prominent as the crystalloidcore of mammalian liver peroxisomes (Shnitka, 1966, Lata etal., 1977; Usuda et al., 1988a). In the adult frog hepatic andrenal peroxisomes, urate oxidase is localized to this subcrys-talloid core (Fig. 3A and C), whereas catalase is distributedthroughout the matrix of all peroxisomes (Fig. 3B and D). Inthe tadpole, urate oxidase-containing peroxisomes with thesubcrystalloid core are detected in the liver and kidney duringdevelopmental stages 51 through 60 that were examined in

the present study (Fig. 4). The immunogold labelingdepicting urate oxidase localization is clearly restricted to thecore-like densities in the peroxisomal matrix (Fig. 4A and C).Catalase is present in all peroxisomes in both liver and kidneycells.

Localization of allantoinase in the liver and kidneyImmunoblot analysis revealed the presence of allantoinase inkidney and liver homogenates obtained from adult frogs andtadpoles (Fig. 5). The molecular mass of allantoinase subunitis approximately Mr 54,000 (Noguchi et al., 1986). The anti-bodies used in this study are highly specific for allantoinase;these antibodies inhibited allantoinase activity on immunoti-tration and precipitated an Mr 54,000 subunit protein (Noguchiet al., 1986). The cellular distribution of allantoinase in frogliver and kidney was investigated by immunoperoxidasestaining (Fig. 6). Hepatocytes showed a strongly positivereaction in the cytoplasm (Fig. 6A); no staining was seen inthe sinusoidal cells and bile duct epithelium. In the frog kidney(Fig. 6B), allantoinase appeared to be specifically localized tothe cytoplasm of cells lining the proximal convoluted tubules.No immunostaining was detected in the glomeruli and distalrenal tubules.

At the electron microscopic level, allantoinase was localizedpredominantly over the mitochondria of liver parenchymalcells (Fig. 7A and C). Labeling over peroxisomes was notdetected, indicating that allantoinase is not associated with thisorganelle in the frog or tadpole liver. All mitochondria in thehepatic parenchymal cells showed immunolabeling for allan-toinase. We evaluated over 300 hepatocyte mitochondria fromseveral different electron micrographs and found labeling overall of them. Few gold particles were detected over the endo-plasmic reticulum which, in some electron micrographs,appeared slightly more prominent than the overall background.In the adult and tadpole kidney, allantoinase was found also inthe mitochondria of cells lining the proximal tubules, but notin the peroxisome (Fig. 7B and D). In the kidney, the mito-chondria present in distal tubules contained no allantoinase. It

Fig. 1. Immunoblot analysis of adult frog liver (lane 1) and kidney(lane 2), and tadpole liver (lane 3) and kidney (lane 4) homogenates(100 µg protein per lane) for urate oxidase, using antibodiesgenerated against frog urate oxidase. Lane M, molecular massstandards (×10−3). Urate oxidase is approximately Mr 32,500.

Fig. 2. Immunoblot analysis of homogenates obtained from adultfrog kidney (lane 1) and frog liver (lane 2) for catalase, using anti-ratcatalase antibodies. Rat liver homogenate is used in lane 3 forcomparison. The antibodies raised against rat catalase cross-reactwith the frog catalase. The catalase subunit molecular mass(Mr60,000) is similar in the rat and the frog. Lane M, molecular massstandards (Mr ×10−3).

1076 N. Usuda and others

Fig. 3. Immunocytochemical localization of urate oxidase and catalase in Lowicryl-embedded adult frog liver and kidney by the Protein A-gold method. Urate oxidase was localized using frog anti-urate oxidase antibodies (A and C), and catalase by rat anti-catalase antibodies (Band D) in the adult frog liver (A and B) and kidney (C and D). Note that urate oxidase is localized within the subcrystalloid nucleoid in theperoxisome (A and C), whereas catalase is localized diffusely over the peroxisomal matrix (B and D). P, peroxisome; M, mitochondrion.Bars, 20 µm.

1077Urate oxidase and allantoinase localization

Fig. 4. Immunocytochemical localization of urate oxidase and catalase in Lowicryl-embedded tadpole liver and kidney by the Protein A-goldmethod. Urate oxidase was localized using antibodies raised against frog liver urate oxidase and catalase by antibodies generated against ratliver catalase. (A) Tadpole liver urate oxidase and (B) tadpole liver catalase. (C) Tadpole kidney urate oxidase and (D) tadpole kidney catalase.Urate oxidase labeling is restricted to the subcrystalloid core of the peroxisome (P). M, mitochondrion. Bars, 20 µm.

1078

appears that urate oxidase and allantoinase are co-localized inthe same renal epithelial cells, albeit in different subcellularorganelles.

To confirm further the distribution of allantoinase in themitochondria, crude mitochondrial fractions from liver andkidney were processed for immunocytochemical localization.Allantoinase was found by Protein-A gold staining to belocalized in mitochondria isolated from rat liver (Fig. 8A andB), as was observed in tissue sections stained in situ. In mito-chondrial fractions isolated from the kidney, allantoinase wasdetected only in some mitochondria and not in others (Fig. 8C).This was expected, since allantoinase distribution is restrictedto the mitochondria in specific segments of the renal tubulesand not to those in other segments.

Immunoblot analysis of the crude mitochondrial fractionsthat were washed free of contaminating cytosolic proteins alsoprovided strong evidence for the presence of allantoinase inthis particulate fraction (data not shown). It should be notedthat the crude mitochondrial fractions were invariably con-taminated with a few peroxisomes and accordingly theimmunoblot results of these subcellular fractionations do notunequivocally prove that allantoinase is associated with mito-chondria. Nevertheless, the immunocytochemical localizationstudies on these crude mitochondrial fractions convincinglydemonstrate that allantoinase is in the mitochondria.

A trace amount of immunostaining was detected in the

N. Usuda and others

Fig. 5. Immunoblot analysis of adult frog liver (lane 1) and kidney(lane 2), and tadpole liver (lane 3) and kidney (lane 4) homogenates(100 µg protein per lane) for allantoinase, using antibodies againstfrog allantoinase. Lane M, molecular mass standards (×10−3).

Fig. 6. Immunoperoxidase localization of allantoinase in liver (A) and kidney (B) using antibodies against frog liver allantoinase. In liver,allantoinase is localized in the cytoplasm of hepatocytes and the proximal tubular epithelium. No staining is observed in the distal tubules andthe glomerulus in kidney. (A) ×240 and (B) ×400

1079Urate oxidase and allantoinase localization

100,000 g supernatant of liver, possibly due to leakage duringhomogenization. Nuclear-encoded mitochondrial proteins aresometimes detected outside the mitochondria in the cytosolic

pool, reflecting that a fraction of the newly synthesized proteinis in the process of translocating from the site of synthesis tothe target organelle (Grivell, 1988). Abundant mitochondrial

Fig. 7. Immunocytochemical localization of allantoinase in Lowicryl-embedded tissues by the Protein A-gold method, using frog liver anti-allantoinase antibodies. Adult frog liver (A) and tadpole liver (C), and adult frog kidney (B) and tadpole kidney (D). Allantoinase is localizedwithin the mitochondria (M), adjacent to the cristae. No allantoinase labeling is seen in peroxisomes (P) in liver and kidney. Subcrystalloidurate oxidase core is seen in peroxisomes in (B) and (D). Bars, 20 µm.

1080

proteins are known to be detected in the cytosolic fractions andthe pool varies under different physiological conditions(Grivell, 1988).

DISCUSSION

The degradation of adenine and guanine, the purine moietiesof nucleic acid, results in the formation of uric acid in allanimals with the exception of leeches, fresh water mussels andspiders, where the degradation of these purines does not gobeyond the stage of hypoxanthine or xanthine (Keilin, 1958).Uric acid is an important biological molecule which exerts anantioxidant action and is postulated to protect against cancerand other disorders, including ageing, caused by oxygen freeradicals (Ames et al., 1981). The peroxisomal localization ofurate oxidase, the first of the three enzymes responsible for thedegradation of uric acid to urea (urate oxidase, allantoinase andallantoicase), and the observation that the urate oxidase genein humans and hominoid primates is mutated, have generatedconsiderable interest in the biological and evolutionary impli-cations of uric acid and the enzymes responsible for its metab-olism (Wu et al., 1989; Yeldandi et al., 1990; 1991). Informa-tion about genes encoding urate oxidase, allantoinase andallantoicase from several species, as well as the precise sub-cellular localization of these enzymes and their regulation atthe molecular level, will be essential to an understanding of thebiological role of uric acid in development and differentiation.

In the present study, we have demonstrated by immunoblot-ting the presence of urate oxidase and allantoinase in both liverand kidney. The presence of these two enzymes in the frogkidney is demonstrated for the first time. In addition, using thehighly specific Protein A-gold immunocytochemicaltechnique, we have localized the two uric acid degradingenzymes, urate oxidase and allantoinase, to peroxisomes andmitochondria respectively, both in the liver and kidney of adult

frogs and tadpoles. Urate oxidase is exclusively localized tothe subcrystalloid core of peroxisomes in the liver and kidneyof the adult frog and tadpole, whereas allantoinase is presentpredominantly in proximity to mitochondrial cristae.

The localization of urate oxidase to the hepatocyte peroxi-somes of liver is consistent with the distribution pattern of thisenzyme in most mammals. However, unlike in most mammals,we show that urate oxidase is expressed in the peroxisomes ofproximal convoluted tubules of the frog and tadpole kidney.This distribution pattern of urate oxidase is similar to that notedin the bovine liver and kidney (Zaar et al., 1986; Usuda et al.,1988b). The reasons for this particular hepatic and renal dis-tribution of urate oxidase remain to be elucidated. Neverthe-less, it could be speculated that this pattern is probably relatedto increased production of uric acid from nucleotides and otherdietary sources in the frog, which may not be efficientlydegraded by the liver urate oxidase. Alternatively, hepatic urateoxidase may not be readily accessible to uric acid foroxidation; as a result renal urate oxidase could participate inthe further degradation of uric acid prior to excretion. Hence,it is important to determine the relative proportions of urateoxidase in amphibian liver and kidney. It is interesting to notethat in the dalmatian dog, abnormalities of uric acid handlingresult in hyperuriciemia (Yu et al., 1971). Similar abnormali-ties of uric acid handling by the liver may be responsible forthe dual organ expression of this gene in the amphibia.

The localization of allantoinase to the mitochondria of theliver of frog is inconsistent with its previously reported local-ization in hepatic peroxisomes (Visentin and Allen, 1969). Thisinconsistency can be explained by the fact that previous studieson the localization of allantoinase consisted of density gradientand differential centrifugations, which cannot fully separateperoxisomes from mitochondria (Scott et al., 1969; Visentinand Allen, 1969). The present study provides direct visualevidence for the presence of allantoinase within the mitochon-dria, and not in the peroxisomes as was previously thought.

N. Usuda and others

Fig. 8. Immunocytochemical localization of allantoinase in isolated mitochondria obtained from subcellular fractions of liver (A and B) andkidney (C) from adult frogs embedded in Lowicryl. In the liver mitochondrial fractions, allantoinase is distinctly visualized in themitochondrial cristae (A,B). Not all mitochondria in the kidney fractions (C) show allantoinase because of heterogeneous distribution of thisenzyme in different segments of the renal tubule. Bars, 20 µm.

1081Urate oxidase and allantoinase localization

Protein A-gold immunocytochemistry is a highly specifictechnique at the ultrastructural level and this method clearlyestablishes the unequivocal presence of allantoinase in themitochondrial cristae.

Our data herein raise several interesting questions as to therole of peroxisomes and mitochondria in uric acid degradationin amphibia. One issue of interest is to explore how theallantoin generated in the peroxisome moves to the mitochon-dria. The role of mitochondria in the metabolism of allantointo allantoic acid also remains to be investigated since the pos-sibility exists that allantoicase may also be found in mito-chondria. If allantoicase is localized in the peroxisome and orcytosol of liver as well as kidney, the allantoic acid producedin the mitochondria has to be transported or diffused out of thisorganelle. Additional cellular and molecular studies arenecessary to understand the physiological and evolutionarysignificance of the unique pathway of uric acid degradation inamphibians.

We thank V. Subbarao and Frank Erfurth for excellent technicalassistance. This work was supported by NIH R37 GM23750, and bya Veterans Administration Research Advisory Grant.

REFERENCES

Alvares, K., Widrow, R. J., Abu-Jawdeh, G. M., Schmidt, J. V., Yeldandi,A. V., Rao, M. S. and Reddy, J. K. (1992). Rat urate oxidase produced byrecombinant baculovirus expression: Formation of peroxisome crystalloidcore-like structures. Proc. Nat. Acad. Sci. USA89, 4908-4912.

Ames, B. N., Cathcart, R., Schwiers, E. and Hochstein, P. (1981). Uric acidprovides an antioxidant defence in humans against oxidant and radicalcaused aging and cancer: A hypothesis. Proc. Nat. Acad. Sci. USA78, 6858-6862.

Bendayan, M. and Reddy, J. K. (1982). Immunocytochemical localization ofcatalase and heat-labile enoyl-CoA hydratase in the livers of normal andperoxisome proliferator-treated rats. Lab. Invest. 47, 364-369.

Bendayan, M., Reddy, M. K., Hashimoto, T. and Reddy, J. K. (1983).Immunocytochemical localization of fatty acid metabolizing heat-stable andheat-labile enoyl-CoA hydratases in liver and renal cortex. J. Histochem.Cytochem. 31, 509-516.

De Duve, C. and Baudhuin, P. (1966). Peroxisomes (microbodies and relatedparticles). Physiol. Rev. 46, 323-357.

De Duve, C., Pressman, B. C., Gianetto, R., Wattiaux R. and Appelmans, F.(1955). Tissue fractionation studies: 6. Intracellular distribution patterns ofenzymes in rat liver tissue. Biochem. J. 60, 604-617.

Friedman, T. B., Polanco, G. E., Appold, J. C. and Mayle, J. E. (1985). Onthe loss of uricolytic activity during primate evolution-I. Silencing of urateoxidase in a hominoid ancestor. Comp. Biochem. Physiol. 81B, 653-659.

Fujiwara, S., Ohashi, H. and Noguchi, T. (1987). Comparison ofintraperoxisomal localization, form and properties of amphibian (Ranacatesbeiana) uricase with those of other animal uricases. Comp. Biochem.Physiol. 86B, 23-26.

Fujiwara, S., Hayashi, S., Noguchi, T., Hanada, N. and Takehara, T.(1989). Subcellular distribution of hepatic allantoinase varies among fishes.Comp. Biochem. Physiol. 93B, 213-215.

Grivell, L. A. (1988). Protein import into mitochondria. Inter. Rev. Cytol. 111,107-141.

Hayashi, S., Fujiwara, S. and Noguchi, T. (1989). Degradation of uric acid infish liver peroxisomes. Intraperoxisomal localization of hepatic allantoicase

and purification of its peroxisomal membrane-bound form. J. Biol. Chem.264, 3211-3215.

Keilin, J. (1959). The biological significance of uric acid and guanineexcretion. Biol. Rev. 34, 265-296.

Laemmli U. K. (1970). Cleavage of structural proteins during the assembly ofthe head of bacteriophage T4. Nature227, 680-685.

Lata, G. F., Mamrak, F., Block, P. and Baker. B. (1977). An electronmicroscopic and enzymic study of rat liver peroxisomal nuceloid core and itsassociation with urate oxidase. J. Supramol. Struc. 7, 419-434.

Noguchi, T., Fujiwara, S. and Hayashi, S. (1986). Evolution of allantoinaseand allantoicase involved in urate degradation in liver peroxisomes. J. Biol.Chem. 261, 4221-4223.

Reddy, J. K. and Kumar, N. S. (1979). Stimulation of catalase synthesis andincrease of carnitine acetyltransferase in the liver of intact female rats fedclofibrate. J. Biochem. 85, 847-856.

Scott. P. J., Visentin, L. P. and Allen, J. M. (1969). The enzymaticcharacteristics of peroxisomes of amphibian and avian liver and kidney. Ann.NY Acad. Sci. 168, 244-264.

Shnitka, T. K. (1966). Comparative ultrastructure of hepatic microbodies insome mammals and birds in relation to species differences in uricase activity.J. Ultrastruc. Res. 16, 598-625.

Takada, Y. and Noguchi, T. (1983). The degradation of urate in liverperoxisomes: Association of allantoinase with allantoicase in amphibianliver but not in fish and invertebrate liver. J. Biol. Chem. 258, 4762-4764.

Taylor, A. C. and Kollros, J. J. (1946). Stages in the normal development ofRana pipienslarvae. Anat. Rec. 94, 7-23.

Tolbert, N. E. (1981). Metabolic pathways in peroxisomes and glyoxisomes.Annu. Rev. Biochem. 50, 133-157.

Towbin, H., Staehelin, T. and Gordon, J. (1979). Electrophoretic transfer ofproteins from polyacrylamide gels to nitrocellulose sheets: procedure andsome applications. Proc. Nat. Acad. Sci. USA. 76, 4350-4354.

Tsukada, H., Mochizuki, Y. and Fujiwara, S. (1966). The nucleoids of ratliver cell microbodies: fine structure and enzymes.J. Cell Biol.28, 449-460.

Usuda, N., Kong, Y., Hagiwara, M., Uchida, C., Terasawa, M., Nagata, T.and Hikada, H. (1991). Differential localization of protein kinase Cisoenzymes in retinal neurons. J. Cell Biol. 112, 1241-1247.

Usuda, N., Reddy, M. K., Hashimoto, T., Rao, M. S. and Reddy, J. K.(1988a). Tissue specificity and species differences in the distribution of urateoxidase in peroxisomes. Lab. Invest. 58, 100-111.

Usuda, N., Usman, M. I., Reddy, M. K., Hashimoto, T., Reddy, J. K. andRao, M. S. (1988b). Immunocytochemical localization of urate oxidase,fatty acyl-CoA oxidase and catalase in bovine kidney peroxisomes. J.Histochem. Cytochem. 36, 253-258.

Visentin, L. and Allen, J. (1969). Allantoinase: Association with amphibianhepatic peroxisomes. Science 163, 1463-1464.

Wu, X., Lee, C. C., Muzny, D. M. and Caskey, C. T. (1989). Urate oxidase:Primary structure and evolutionary implications. Proc. Nat. Acad. Sci. USA.86, 9412-9416.

Yeldandi, A. V., Wang, X., Alvares, K., Kumar, S., Rao, M. S. and Reddy,J. K. (1990). Human urate oxidase gene : cloning and partial sequenceanalysis reveal a stop codon within the fifth exon. Biochem. Biophy. Res.Commun. 171, 641-646.

Yeldandi, A. V., Yeldandi, V., Kumar, S., Murthy, C. V. N., Wang, X.,Alvares, K., Rao, M. S. and Reddy, J. K. (1991). Molecular evolution ofthe urate oxidase-encoding gene in hominoid primates: Nonsense mutations.Gene109, 281-284.

Yu, T., Gutman, A. B., Berger, L. and Kaung, C. (1971). Low uricaseactivity in the Dalmatian dog simulated in mongrels given oxonic acid. Amer.J. Physiol. 220, 973-979.

Zaar, K., Volkl, A. and Fahimi, D. (1986). Isolation and characterization ofperoxisomes from the renal cortex of beef, sheep and cat. Eur. J. Cell Biol40,16-24.

(Received 19 October 1993 - Accepted 22 December 1993)