INFECTION AND IMMUNITY, Dec. 1975, p. 1237-1241Copyright (© 1975 American Society for Microbiology

Vol. 12, No. 6Printed in U.S.A.

Antigenic Independence of Some Microbial Urate OxidasesD. A. FITZPATRICK* AND K. F. McGEENEY

Department of Medicine and Therapeutics, University College, Dublin 4, Ireland

Received for publication 19 June 1975

Antisera were prepared to urate oxidase derived from three microbial species,a yeast (Candida utilis), a mold (Aspergillus flavus), and a bacterium(Bacillus fastidiosus). The antisera inhibited enzyme activity to a limitedextent. Cross-reaetion studies with preparations of the enzyme from these andother species indicated that the microbial enzyme exhibits a high degree ofantigenic independence. This appeared to be particularly true of the bacteriastudied.

The control of hyperuricemia by the adminis-tration of urate oxidase (urate:02 oxidoreduc-tase, EC 1.7.3.3), first carried out by Oppenhei-mer (8), has found new impetus (1, 2, 5, 9, 10)due to the recent availability of highly purifiedpreparations of the enzyme from microbialsources (6, 7). Repeated administration of theenzyme has resulted in antibody production (5,10), which reduces the efficacy of the treatment.A study of the immunology of urate oxidase wasundertaken to determine whether the enzyme,from a number of microbial species, exhibitedantigenic uniformity. Antisera to purified ur-ate oxidase from a yeast, a mold, and a bacte-rium were produced, and the interaction of theantisera and the enzymes was studied. In addi-tion, a number of selected species were cul-tured, and extracts of urate oxidase from thesewere tested for cross-reaction with the threeantisera. A low level of structural homologywas observed.

MATERIALS AND METHODSProduction of antisera and immunodiffusion

technique. The antigens used were urate oxidasederived from Aspergillus flavus, Bacillus fastidio-sus, and Candida utilis. The C. utilis enzyme wasSigma type IV urate oxidase. P. Laboureur donatedthe A. flavus enzyme and L. Mahler donated the B.fastidiosus enzyme. Guinea pigs were immunizedsubcutaneously using a solution containing 5.0 IU ofenzyme emulsified with an equal volume of Freundcomplete adjuvant. Each animal received at leastthree injections at 10-day intervals before bleeding.Antisera were studied using the micro-Ouchterlonytechnique, and undiluted antiserum was used in allcross-reaction studies. The amount of antigen neces-sary to form a visible precipitin line with the homolo-gous antiserum was determined empirically. In gen-eral, a concentration of approximately 0.8 IU/ml wasfound to yield a satisfactory precipitin line. All im-munodiffusion patterns were stained with enzymestain whether or not a precipitin line was visible.

The use of specific enzyme stain permits cross-reac-tions to be evaluated with greater validity.Enzyme stain, assay, and precipitation. The so-

lution used for staining the precipitin pattern wasthat developed by Graham and Karnovsky (4) for thehistological detection of urate oxidase. In all cases,5,6,7,8,-tetrahydro-l-naphthylamine was added tothe solution. After immunodiffusion 1 or 2 drops ofstain was placed over the precipitin pattern to forma bolus and was incubated either at 37 C for 1 h or atroom temperature (20 C) overnight. Staining allowsthe specific detection of urate oxidase complexes,thus increasing the sensitivity of the technique. En-zyme activity was assayed in a Unicam S.P. 800spectrophotometer with a thermostatted cell holderat 293 nm. The substrate was 0.0001 M uric acid in0.02 M borate buffer, pH 9.0.

Incubation with homologous antigen was carriedout at room temperature, 20 C. To 0.8 ml of buffer(0.02 M borate, pH 9.0) was added 0.1 ml of antise-rum and 0.1 ml of the antigen solution (2 IU/ml).The reaction was performed in a 1-ml cuvette (lightpath, 20 mm), and the rate of precipitation (turbid-ity) was followed in a Unicam S.P. 800 spectropho-tometer at 500 nm. The extinction rose rapidly in thefirst 30 min and thereafter it rose only very slowly,indicating that the bulk of the precipitation tookplace quickly. The solutions were allowed to standovernight and were then assayed for enzyme activ-ity and compared to controls (normal serum).

Culture of organisms. Neurospora crassa (CMI53238ii), Schizophyllum commune (IMI 93108), Peni-cillium expansum (CMI 39761), Schizosaccharo-myces octosporus (CMI 140427), and Aspergillus nid-ulans (local culture) were grown in a liquid medium(2%, wt/vol; Oxoid [London]) malt extract contain-ing two tablets of Oxoid nutrient broth no. 2 per 100ml). Bacillus megaterium (IMD 15; Industrial Micro-biology Department, University College, Dublin)was grown on a minimal salt medium containing 1%(wt/vol) sucrose as energy source, since addition ofurate to nutrient broth did not lead to production ofdetectable levels of the enzyme in this organism. Allmedia were adjusted to a concentration of 0.5 g ofuric acid per liter, the pH being maintained con-stant during this adjustment with sodium hydrox-

1237

on April 23, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

1238 FITZPATRICK AND McGEENEY

ide. The media were thus supplemented, becausethe levels of enzyme in the microorganisms were lowwhen grown in the absence of urate. All the orga-nisms were grown in 4-oz (ca. 190-ml) culture flasks.

Harvesting and extraction of organisms. Themolds were harvested by filtration, assisted bycompression of the mycelium with a spatula. Theorganisms were broken by shaking in a Mickle disin-tegrator (2 g of glass beads; type 100-500 S 13M Co.,St. Paul. Minn.] and 10 ml of 0.1 M Na2CO3, pH 10.0,per g of compressed mycelium) for two 10-min pe-riods. The yeast and bacteria were collected by cen-trifuging, suspended in the same buffer, and dis-rupted in an X press (A.B. Biox, model X25). Celldebris was removed by centrifuging (37,000 x g for30 min), and the supernatants were assayed for en-zyme activity. To obtain samples of enzyme of suffi-cient concentration (see above) for immunodiffusion,the A. nidulans extract was concentrated 10-fold inan ultrafiltration cell (Amicon N.V., Holland) with aUM-10 membrane. The other extracts were fraction-ated by ammonium sulfate precipitation, and the 25to 657 pellet was dissolved in 0.1 M Na2CO:.

RESULTSIncubation studies indicated that the antigen

retained enzymatic activity after precipitationwith homologous antiserum. The degree of inhi-bition observed depended on the specific systemstudied, but was in all cases less than 2057( andthus allowed specific staining of the precipitinlines.The microbial urate oxidase exhibited a con-

siderable degree of antigenic independence.When each of the three antigens used in produc-

B

tion of the antisera was tested for cross-reactionwith the heterologous antisera, no precipitinlines were formed. Serial dilution of antigenhas shown that staining can reveal the pres-ence of a precipitin reaction in the absence ofvisible precipitin lines. Such a situation islikely to arise when either the antigen concen-tration is low or a heterologous antigen sharinga limited number of determinants is being stud-ied. Incubation of the slides with the enzymestain did not, however, reveal any evidence ofweak reactions (Fig. 1). As the three homolo-gous antigens exhibited complete antigenic in-dependence, crude preparations of urate oxi-dase from a number of other species were inves-tigated.Antiserum to the A. flavus enzyme was

tested by immunodiffusion against urate oxi-dase from four other molds, A. nidulans, N.crassa, P. expansum, and S. commune. Moldswere preferentially chosen, as it is the A. flavusurate oxidase which has been used in the treat-ment of hyperuricemia and gout. Both the A.nidulans and P. expansurn extracts cross-re-acted, but no precipitin line was visible afterreaction with the other two fungal enzymes.The antiserum also failed to react with theother bacterial and yeast enzymes studied. AsA. nidulans is the more closely related speciesfrom which the two cross-reacting enzymeswere derived, the nature of the intrageneric re-action was studied. The reaction was one of par-tial identity only (Fig. 2), indicating that even

B

t-~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~a .z

... ~ ~ ~ ~~~~~~~..

t..(

FIG. 1. Absence of reaction with heterologous antiserum. Center wells: 1, Antiserum to A. flavus urateoxidase; 2, antiserum to C. utilis urate oxidase; 3, antiserum to B. fastidiosus urate oxidase. Outer wells: A,A. flavus urate oxidase; B, C. utilis urate oxidase; C, B. fastidiosus urate oxidase.

-M.H.Rdl-

INFECT IMMUN

on April 23, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

ANTIGENIC INDEPENDENCE OF URATE OXIDASE

..

"A

::

I

2"'_ 4. .

I

4

\.._.

FIG. 2. Heterogeneity in cross-reaction within the genus Aspergillus. Center well: Antiserum to A. flavusurate oxidase. Outer wells: 1 and 3, A. flavus urate oxidase; 2 and 4, A. nidulans urate oxidase; 5, N. crassaurate oxidase.

at generic level the enzymes ofAspergillus spe-cies exhibit a considerable degree ofheterogene-ity. This heterogeneity was not observed to thesame extent in the case of the yeast enzyme. S.

octosporus urate oxidase cross-reacted with theC. utilis antiserum; the precipitin line was ofequal strength with that observed between theAspergillus antiserum and the A. nidulans ex-

...i .....

1239VOL. 12, 1975

on April 23, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

1240 FITZPATRICK AND McGEENEY

tract and stronger than that between the sameantiserum and the P. expansum extract, de-spite the lower concentration of the yeast en-zyme (0.2 IU/ml). The reaction between theyeast enzyme and the C. utilis antiserum wasone of partial identity, as spur formation wasobserved. C. utilis antiserum failed to cross-react with extracts of the other fungi and thebacterium except in the case of the N. crassaextract, with which it formed a weak line.Antiserum to the B. fastidiosus enzyme

failed to cross-react with an extract of urateoxidase from B. megaterium. Not surprisingly,in view of the absence of an intrageneric reac-tion, neither did it react with any of the otherpreparations of microbial urate oxidase stud-ied. The various interreactions are summarizedin Table 1.

DISCUSSIONThe development of an immune response

after prolonged administration of urate oxidaseis an undesirable side effect in this mode oftherapy of hyperuricemia. One way of extend-ing this form of treatment would be the sequen-tial use of enzyme preparations that are knownto be immunologically independent of one an-other. Since vertebrate urate oxidase is not im-munologically independent (3), the investiga-tion of immunological dependence of microbialurate oxidase was undertaken. The existence ofcomplete antigenic independence between theenzymes of the yeast, the mold, and bacteriumused in the production of antisera indicate thepossibility of using such sources of enzyme foruse in therapy, as they are commercially availa-ble. Although the Aspergillus enzyme ex-hibited an intrageneric reaction of partial iden-

TABLE 1. Interaction ofmicrobial urate oxidase withspecific antisera to three of the enzymes

Antiserum against:

Species C. B. fas-A. flavus utilis tidiosus

A. flavus + a

A. nidulans +P. expansum +S. commune - - -N. crassa - +C. utilis - +bS. octosporus - +B. fastidiosus - - +B. megaterium -

a +, Presence of precipitin line of complete orpartial identity; -, absence of precipitin line of com-plete or partial identity.

bHomologous reaction.

tity, no extrageneric reaction was observed.The spur formation observed in the reactionbetween A. nidulans extract and the A. flavusantiserum contradicts the observation of Royeret al. (10) that the A. flavus enzyme possessesonly one determinant group. The yeast enzymeappears to possess a similar degree of homologyto the Aspergillus enzyme in that it cross-reacts with other members of the class to whichit belongs, but this reaction is one of partialidentity. In addition, determinants on the yeastmolecule are not confined to the yeasts but alsooccur on molds, as evidenced by the ability ofantiserum to the C. utilis enzyme to react withthe enzyme from another ascomycete, N.crassa.

Bacterial urate oxidase would appear to dif-fer substantially from the fungal enzyme inexhibiting an even greater degree of antigenicindependence, as evidenced by the failure of B.fastidiosus antiserum to react with enzymefrom another species of the same genus.

In general, it appears that considerable alter-ation in the structure of urate oxidase amongthe protists has occurred. If one regards Asper-gillus as belonging to the class Fungi imper-fecti, then no cross-reactions were observed be-tween enzymes from any one microbial classand antiserum to enzymes from another class.Furthermore, cross-reactions within classesand genera are of partial identity, indicatingthat even relatively closely related species stillshow heterogeneity of the enzyme molecule. Apossible explanation for this apparent lack ofhomology would be if the microbial enzymespossessed few determinant groups. In the caseof the urate oxidases so far studied, the molecu-lar weight was found to be in the region of100,000 or more (6, 7), so the enzyme couldreasonably be expected to possess a number ofdeterminant groups. It would thus appear thatmicrobial urate oxidase possesses a low degreeof immunological homology.

ACKNOWLEDGMENTS

D. A. F. acknowledges receipt of a Fellowship of theMedical Research Council of Ireland.We thank Oliver FitzGerald for help and encourage-

ment, and W. Fogarty and H. Fuller for supplying thecultures.

LITERATURE CITED

1. Brogard, J. M., D. Coumaros, J. Franckhauser, A.Stahl, and J. Stahl. 1972. Enzymatic uricolysis: astudy of the effect of a fungal urate-oxydase. Eur. J.Clin. Biol. Res. 17:890-895.

2. Dumas, R., J. Castel, and R. Jean. 1973. Urate oxidasein pediatrics. Pathol. Biol. 21:425-429.

3. Fitzpatrick, D. A., and K. F. McGeeney. 1975. Compara-tive immunology of vertebrate urate oxidase. Comp.Biochem. Physiol. B 51:37-39.

INFECT. IMMUN.

on April 23, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

ANTIGENIC INDEPENDENCE OF URATE OXIDASE 1241

4. Graham, R. C., and M. J. Karnovsky. 1965. The histo-chemical demonstration of uricase activity. J. Histo-chem. Cytochem. 13:448-453.

5. Kissel, P., M. Lamarche, and R. Royer. 1968. Modifica-tion of uricaemia and the excretion of uric acid nitro-gen by an enzyme of fungal origin. Nature (London)217:72-74.

6. Labourer, P., and C. Langlois. 1968. Urate oxydased'Aspergillus flavus. I. Obtention, purification, prop-erties. Bull. Soc. Chim. Biol. 50:811-825.

7. Mahler, J. L. 1970. A new bacterial uricase for uric aciddetermination. Anal. Biochem. 38:65-84.

8. Oppenheimer, E. H., and H. G. Kunkel. 1943. Furtherobservations on the lowering of blood uric acid byuricase injections. Bull. Johns Hopkins Hosp. 73:40-53.

9. Potaux, L., M. Aparicio, C. Maurel, E. Ruedas, and C.Martin-Dupont. 1975. Uricolytic therapy. Value ofurate oxydase in the treatment of hyperuricaemia.Nouv. Presse Med. 4:1109-1112.

10. Royer, R., J. Vindel, M. Lamarche, and P. Kissel. 1968.Modalaties of purine excretion during enzyme treat-ment ofgout and other hyperuricemic conditions withurate oxidase. Presse Med. 76:2325-2328.

VOL. 12, 1975

on April 23, 2021 by guest

http://iai.asm.org/

Dow

nloaded from