Annals of the Rheumatic Diseases, 1980, 39, 103-117

Heberden oration 1979

Human aberrations of purine metabolism and theirsignificance for rheumatologyJ. EDWIN SEEGMILLER

University of California, San Diego, La Jolla, California 92093, USA

It is an honour and privilege to be invited to addressyou this day. Returning to the city of my grand-father's birth is always a great personal pleasure.This relationship is symbolic also of the closescientific and cultural heritage that we in Americaowe to this land. The origin ofmany ofour importantmedical and scientific concepts today stem from thegerminal ideas and observations of British scientistsand physicians of past ages who have contributedsubstantially to building the scientific base of theclinical medicine of today.Not the least of these contributions has come from

detailed study of patients afflicted with very specifictypes of arthritis. Although it was the Swedishchemist, Scheele, who first isolated uric acid from aconcretion of the urinary tract in 1776, it was theBritish chemist, Wollaston, a nephew of WilliamHeberden, who first demonstrated in 1798 thepresence of the same substance in a gouty tophuswhich he is purported to have removed from hisown ear. The subsequent demonstration of hyper-uricaemia in patients with gouty arthritis in 1848 byAlfred Baring Garrod provided the impetus for muchmore detailed study of the chemical factors causingdisease, which has proliferated remarkably in theensuing 132 years and continues today at an evenmore accelerated pace.' 2From detailed studies of patients with another

heritable form of arthritis, alcaptonuria withochronosis, Archibald Garrod carried his father'sconcepts to a more fundamental basis with for-mulation of a revolutionary new concept to accountfor the accumulation of large amounts of homo-gentistic acid in these alcaptonuric patients. In hisCroonian lectures in 1908 he proposed that theaccumulation and excretion of homogentistic acidin this disease is a direct result of a hereditarydeficiency of the enzyme for its further processingwithin the body.3 This concept of one gene beingresponsible for one enzyme was thus formulated over30 years before its more detailed investigation

by the microbiologists Beadle and Tatum.4 Fromthese studies he also formulated the concept ofinborn errors of metabolism and thus laid thegroundwork for modern-day developments of ourconcept of hereditary diseases. The younger Garrod5classified gouty arthritis as one of the inborn errorsof metabolism, but unlike other such disorders nosingle enzyme defect has been found to accountfor the disease. However, the absence of the enzymeuricase in man and the higher apes, coupled with aremarkably inefficient mechanism for renal excretionof uric acid, results in a mean plasma concentrationof urate in adult males of 5c 1 mg/dl-a value notfar from the theoretical limit of solubility of uratein serum of 6 4 mg/dl. The lower mean serum urateof women of 4 - mg/dl, until after the menopause,may well account for the far lower incidence of goutin women. The supersaturation of all hyperuricaemicplasma sets the stage for the deposition of needle-shaped crystals of monosodium urate monohydratein and about the joints and in the parenchyma of thekidney which account for the major pathology inthis disease, as proposed by the elder Garrod over acentury ago.2 Although his concepts languishedfor over half a century, new evidence of the past 2decades has validated his major proposals.6-9 Onlyin the past decade and a half have specific enzymedefects in purine metabolism responsible for thisdisease been identified.Today I will review some of the circumstances

involved in the discovery of these enzyme defectsand also review the recent work in which otherspecific inborn errors of purine metabolism havebeen found to be associated with deficiency of theimmune function and the new insight into possibleregulatory mechanisms of the immune system thatmay well be affected by these enzyme defects. I wantto acknowledge the major contributions of themany young investigators who have worked with mein these studies. I also hope to enlist your supportin a systematic approach to the classification of our

103

104 Seegmiller

gouty patients that will greatly facilitate, not onlyour rational management of their problems byselection of the most appropriate therapeutic agents,but will also allow us to work together in theirfurther biochemical and genetic characterisation.Since it is a lifetime of treatment to which we arecommitting the patient, the extra time it takes forclassification seems warranted and can also beconveniently performed while awaiting for thepatient to recover from an acute attack of goutyarthritis and before institution of treatment witha drug known to lower the serum urate concentration.

Classification of gouty patients

The genetic heterogeneity of the gouty populationis readily revealed by the simple test I am proposingwe use on all our gouty patients-the determinationof the 24-hour excretion of uric acid for each of 3days after the patient has been equilibrated on adiet virtually free of purines.10 The mean value foundin normal adult males is 425 mg/24 hours (Fig. 1)with a spread that places the upper range of normalat 600 mg per day. The population of gouty patientsshowed a mean value only slightly higher but thespread was much larger with around one-quarterof the gouty population showing production ofexcessive quantities of uric acid in the 24-hourperiod. As might be expected, the patients who hadsubstantial renal damage excreted smaller amounts ofuric acid than was found in the normal, indicatingthe basis for the well-known accleration in rate ofdevelopment of pathological changes of gout whenrenal impairment supervenes.

8

Knowledge of the classification of the goutypatient in regard to the degree of production ofuric acid can also be of practical help to the clinicianin the selection of the most appropriate drugs formanagement of the particular patient's disease. Weknow that allopurinol, in addition to its primaryaction in blocking the enzyme, xanthine oxidaseresponsible for the synthesis of uric acid, also hasa secondary beneficial effect in shutting off excessivepurine synthesis." It therefore becomes the drug ofchoice for patients who have evidence ofan excessiverate of purine synthesis, uric acid lithiasis orimpaired renal function.Although some of the patients who show a uric

acid excretion in the normal range may have modestdegrees of overproduction of uric acid that isobscured by increased extrarenal disposal intotophi or by intestinal uricolysis, a substantialportion of such patients show a diminished abilityof the kidney to excrete uric acid as the cause of theirhyperuricaemia.8 1012 Therefore, uricosuric drugscorrect the primary abnormality and provide arational therapeutic approach for controlling theirdisease.We have recently been using a total of 6 days on

the purine-free diet with collection of urine duringthe last 3 days as an abbreviated procedure moreacceptable to the patients. Even this degree of co-operation is often difficult to elicit. For detectionof patients with the most extreme degrees of purineoverproduction, a simple ratio of uric acid tocreatinine in the morning urine suffices. However,it is not nearly as discriminating as the 24-hourcollection when properly performed.13 A further

.7

C_ Non-op*oceos 34Tophocos S

RenMlMR rnolrnnt

w

0

cr 3 _ X - 111-

200 OR 250 300 350 400 450 500 550 600 650 700 750 800 OR

UNDER AVERAGE 24 HOUR URINARY URIC OVER

B ACID EXCRETION mg./24 HR.

Fig. 1 Uric acid excretion bynormal and goutypatients afterequilibration for 5 days on a dietvirtuallyfree ofpurines. FromSeegmiller et al.10

Human aberrations ofpurine metabolism and their significance for rheumatology 105

modification has been proposed in which the mgof uric acid excreted per ml of glomerular filtrateis obtained by dividing the above ratio by theserum creatinine.14The origin and disposition of uric acid in the

normal and gouty patients is shown in Fig. 2.Purines which give rise to uric acid can originatefrom purines in the diet or from those synthesisedendogenously in all nucleated cells. From eithersource the purines can be incorporated into cellularnucleotides such as ATP and nucleic acid. Themetabolic turnover of these purine-containing com-pounds gives rise eventually to uric acid. Aroundtwo-thirds to three-quarters of the uric acid formedeach day is excreted unchanged in the urine, whilethe remainder is excreted in intestinal secretionswhere it is degraded by the action of uricolyticenzymes of the gut bacteria.9 1012 15The hyperuricaemia required for development of

gouty arthritis results from an expansion of theurate pool. In some patients a genetically determinedexcessive rate of production of uric acid is respon-sible for the hyperuricaemia. In others this excessiveproduction arises from an enhanced cellular pro-liferation such as is found in myeloproliferative dis-orders with an increased turnover of nucleic acid,while in others, and this may well be the majorityof patients, a diminished renal excretion of uric acidis responsible for hyperuricaemia.'0 In some casesit has been traced to an associated lactic acidosis withcompetition of lactic acid or other organic acids

A. NORMAL

is PLONE-IIYNTHESISJ`1IB80Y KUR E_I TISSUEITIDITARYI NUCLE0T=DEINCLEIC ACSIDq-URINe-ETPURNESI ~ ADS

FRIC ACIt \ INTESTINALI URICOLYSIS

RENAL EXCRETION

C. GOUT

Associated with myeloproliferative disease

URICOLYSIS

RENAL EXCRETION

B.

such as ketoacids in diabetic acidosis or in maplesyrup urine disease for transport mechanisms.'6In type 1 glycogen storage disease, with its glucose-6-phosphatase deficiency, both a lactic acidosisand an increased rate of purine synthesis are found."7

Lesch-Nyhan disease with severe deficiency of hypo-xanthine-guanine phosphoribosyl transferase (HPRT)

Clinical Presentation. The most extreme example ofexcessive uric acid production has been found in thepaediatric practice rather than in the rheumatologyclinic. A haematuria and uric acid crystalluria firstbrought a young boy with cerebral palsy to theattention of a medical student, Michael Lesch, andDr W. L. Nyhan at Johns Hopkins Hospital.'8Both the patient and his brother had a severe neuro-logical disorder consisting of choreoathetosis,spasticity, mental retardation, and a compulsivebiting away of lips and tongue and amputation ofthe fingers by biting. Both children excreted in thedaily urine over 6 times the normal quantity of uricacid. The affected children differed from most goutypatients in failing to diminish the rate of theirpurine synthesis in response to azathioprine, a 6-mercaptopurine derivative. Furthermore, their cellsin culture failed to diminish the rate of purinesynthesis when 6-mercaptopurine was added to theculture. We eventually traced this aberrant pharma-cological response to a gross deficiency of the enzymehypoxanthine guanine phosphoribosyl transferase

GOUTWith overproduction of uric acid

Fig. 2 Causes ofhyperuricaemia.From Seegmiller et al.100

ISNTEI UIN TISSUEIIPURINESI ,\DETA EQTQSUG.EIQCAD-SJ

IUPH-l uff_4CINTESTINAL

I URICOLYSISRENAL EXCRETION

D. GOUTWith diminished renal excretion of uric acid

IPURINE1ISYNTHIESISI\BD UIE|I TSU

INTESTINALT URICOLYSIS

RENAL EXCRETION

106 Seegmiller

(HPRT) normally present in all tissues of the bodyand required for converting 6-mercaptopurine to itsmetabolically active nucleotide form'9 20 (Fig. 3).Up until this time the HPRT enzyme had been

assigned only the role of a salvage system for thethrifty reutilisation of purine bases by convertingthem to purine ribonucleotides at a cost of one ATPmolecule as compared to at least 6 ATPs required forsynthesis of the purine nucleotide de novo. Severeneurological dsiease and excessive uric acid pro-duction of the affected children was the first indi-cation of an important role for this enzyme in bothneurological function and in the control of the rate ofpurine synthesis.

X-linked inheritance. As more families were studiedit was soon realised that this disorder affects onlymales and the pattern ofinheritance with transmission

only through their mothers, suggested an X-linkedpattern of inheritance. In order to demonstrate moreconclusively the X-linked inheritance we made use ofa theory first proposed by Dr Mary Lyon2' here inEngland, to account for the presence of Barr bodiesin female somatic cells. She proposed their formationto result from the random inactivation of one of thetwo X chromosomes in female somatic cells atan early stage of zygote development. A consequenceof such a random inactivation would be the presenceof both normal and mutant phenotypes in cells ofwomen heterozygous for an X-linked disorder,reflecting the inactivation of the X chromosomecarrying either the mutant or the normal gene. Thedevelopment of an analytical system for detectingthe presence or absence of HPRT enzyme in singlecells permitted the demonstration of these two cell

RIBOSE 5 - P + ATP

5 - Phosphoribosyl - 1 - pyrophosphate (PRPP) + Glutamine

5 Phosphoribosyl - 1 - amine N

V FormateA \

/ 1 sFormylglycinamide ribonucleotide (F

_ -

Guanylic Acid no-Inosinic Acid

Hypoxanthine GuanineGuanosine Phosphoribosyltransferase Inosine * Ad(

Guanine Hypoxanthine

Xanthined

Uric Acid

FGR)

-o Adenylic Acid

lenosine ® p

Adenine

Xanthine oxidase

Fig. 3 Known enzyme defects in human purine metabolism. (1) Increased phosphoribosylpyrophosphate synthetaseactivity in patients with overproduction of uric acid and gout. (2) Gross deficiency ofhypoxanthine guaninephosphoribosyltransferase in children with Lesch-Nyhan disease andpartial deficiency of the same enzyme in patientswith overproduction of uric acid and gout. (3) Adenine phosphoribosyltransferase deficiency in patients with kidneystones composed of2-8 dioxyadenine that are often confused with uric acid stones. (4) Xanthine oxidase deficiency inpatients with xanthinuria who are at increased risk for xanthine calculi of the urinary tract and in occasional patientsmyalgia from xanthine crystals in the muscle. (5) Adenosine deaminase deficiency associated with severe combinedimmunodeficiency disease. (6) Purine nucleoside phosphorylase deficiency associated with isolated defect in T cells.(7) Purine S'-nucleotidase activity is low in lymphocytes ofpatients with agammaglobulinaemia that may be secondaryto loss ofB cells. (8) Adenosine kinase deficiency has so far been developed only in the human lymphoblast cell lines.Its counterpart in patients is yet to be identified.

NucleicAcids

NucleicAcids

A

I

Human aberrations ofpurine metabolism and their significance for rheumatology 107

populations predicted by Dr Lyon's hypothesis. Thedeficient enzyme is required for the first step in theincorporation of tritiated hypoxanthine into nucleicacid in cultured fibroblasts. Upon overlaying thewashed cells with photographic emulsion, normalcells showed abundant3H incorporation, while cellsfrom affected children showed virtually no incor-poration. As predicted by the Lyon hypothesis, cellsfrom the mother of an affected child showed popu-lations of both normal and mutant cells, thus con-firming very effectively the location of the gene forcoding the HPRT enzyme on the X chromosome.22The 2 populations are also detectable by enzymeassay of hair follicles.23

Prenatal diagnosis. This same technique, which weapplied to amniotic cells in culture, demonstratedfor the first time the usefulness of biochemistry indetermining genotype for an X-linked disease.24We have since monitored 21 pregnancies at risk forthis disorder and have identified seven affectedfetuses, each sufficiently early in gestation that theparents' desire to terminate the pregnancy could bemet.2526 At the present time prenatal diagnosis istheoretically possible for well over 50 serioushereditary metabolic diseases, thus introducing anew and dynamic role to genetic counselling, whichpermits individuals known to be at risk for producingchildren with certain specific serious hereditarydefects to produce the normal children which theydesire.

Screening test. Now that the Lesch-Nyhan diseaseis preventable through monitoring of pregnancies,the identification of affected children becomesimportant in order that their female relatives carryingthe gene can be identified and informed of the needfor monitoring of pregnancies. We have devised ascreening test in which the automated determinationof uric acid and creatinine in liquid urine samplesor urine samples preserved by drying on filter paperis used to identify those patients in neurology clinics,genetics clinics, or cerebral palsy clinics in whomthe more specific enzyme assay on erythrocytesshould be performed.'327 Female relatives who arecarrying this gene can then be identified and alertedto the need to have their pregnancies monitored ifthey are to avoid producing affected children.

Neurological disorder. The mechanism by whichthe enzyme deficit in brain tissue leads to the severeneurological disease and compulsive behaviouris not well understood.20 In glioma and neuro-blastoma cell lines deficient in HPRT, monoamineoxidase activity is significantly diminished, sug-gesting the possibility of an imbalance also beingcreated in the neurotransmitter system of the brainby this enzyme deficit.28 29 Hydroxytryptophan hasbeen reported by Mizuno and Yugari3O to ameliorate

the self-mutilating behaviour in children in Japan,but Frith et al.31 in this country and Ciaranello el al.32in the United States were unable to confirm thisobservation. However, a combination of hydroxy-tryptophan and carbidopa (MK486) produced atransient cessation of compulsive self-mutilationin some patients which could not be sustained even byincreased doses of the medications33 (Nyhan,personal communication). A role for uric acidseems to be ruled out by the normal concentrationsof uric acid found in the cerebrospinal fluid. How-ever, hypoxanthine and xanthine were elevated in thecerebrospinal fluid and the methylated xanthines,caffeine, and theophylline have been shown toinduce an analogous compulsive self-mutilation inrats and rabbits.20 3

Gout with partial deficiency of HPRT

Clinical features. Understanding the enzyme defectof Lesch-Nyhan disease has substantially extendedour understanding of other clinical disorders asso-ciated with excessive production of uric acid.34Less severe deficiencies of the same enzyme havebeen found in members of certain families withgout.35 36 Three gouty brothers, each showing a 3-to 4-fold increase over normal in daily uric acidproduction, had around 1% of normal HPRTenzyme activity in their erythrocyte lysates and nodetectable neurological disease. In another familywith a different mutation in the same enzyme show-ing around 0 5% of normal activity with guanineas a substrate both brothers showed an attenuatedneurological disease consisting of a mild spasticquadriplegia in one young man and mental retar-dation and kidney stones in his 10-year-old brother.Additional neurological disorders have been foundin other families, including seizures, dysarthria,mild motor difficulties, and mental retardation.20 36 37Obviously, the clinical neurological signs wouldnot permit such divergent clinical presentations to beclassified together. Only the HPRT enzyme defi-ciency shared by these affected patients, has permittedthe relationship to be established. As might beexpected, all of the affected patients showed anexcessive production of uric acid, a high incidenceof renal calculi, and a progressive renal damageoften beginning in childhood, with development ofgouty arthritis at an earlier age than most goutypatients.

Apparently genetic heterogeneity is present evenin defects involving the same gene, since differentfamilies carry different types of mutations. However,in general the severity of the clinical manifestationsare proportional to the severity of the enzyme defect,although exceptions are known.20 38 Any patient

108 Seegmiller

carrying the HPRT deficiency shows an aberrantresponse to allopurinol in which the deficit in uricacid production is made up by hypoxanthine andxanthine without the decrease in the total purinesynthesis encountered in other gouty patients.39

Mechanism of excessive purine synthesis in HPRTdeficiency

The mechanism by which the deficiency of HPRTleads to an excessive rate of purine synthesis is nowfairly well defined. Since the salvage activity of theenzyme produces purine nucleotides directly fromfree purine bases, its gross deficiency could con-ceivably result in a diminished intracellular con-centration of purine nucleotides leading to a releaseof the normal feedback inhibition (Fig. 3). Althoughthis is an attractive hypothesis, we have found noevidence to support it. The intracellular concentrationof purine nucleotides in fibroblasts cultured frompatients was the same as in those of normal fibro-blasts.40 Lymphoblasts carrying this enzyme deficitalso showed no increase in adenine nucleotides,although an increased concentration of pyrimidinenucleotides was found.4' 42 An alternative hypo-thesis was suggested by the fact that the presumedrate-limiting amidotransferase shares with the HPRTenzyme a common substrate, phosphoribosyl-l-pyrophosphate (PP-ribose-P). Therefore, the enzymedeficiency conceivably could increase the intra-cellular concentration of this substance. This, indeed,proved to be the case. HPRT-deficient fibroblastsshowed 3- to 4-times greater than normal con-centration of PP-ribose-P and in erythrocytes ofaffected children the PP-ribose-P was increased asmuch as 10-fold above normal.2040

Gout with increased phosphoribosyl pyrophosphatesynthetase (PP-ribose-P)

Additional evidence in support of the importantrole of PP-ribose-P in determining the rate of purinesynthesis came from studies of patients with goutyarthritis who had an excessive rate of purine syn-thesis de novo but normal HPRT activity. Theirfibroblasts in culture also showed an increasedconcentration of PP-ribose-P. 40 43 A differentmutation was identified by Dr Becker while a post-doctoral fellow in our laboratory to account forthis increased PP-ribose-P concentration.44-46 Thismutation was most unusual in that it consisted of anincreased activity, rather than a decreased activity,of the enzyme phosphorybosylpyrophosphate (PP-ribose-P) synthetase. It differed from a mutationdescribed by Sperling, et al.47 and Zoref48 in a goutypatient in that this high activity was apparent at all

concentrations of inorganic phosphate with normalresponse to feedback inhibitors. Subsequent workby Dr Becker's laboratory has identified additionalfamilies, each with a different type of mutation inthis same enzyme but all with a high activity. Achild originally reported to show autistic behavior49was now found instead to be deaf and detailedexamination showed both feedback resistant PP-ribose-P synthetase, and a high specific activity ofthis enzyme.50 The presence of 2 different mutations,each leading to an accumulation of PP-ribose-Pand an excessive rate of purine synthesis, providesstrong evidence of the important role of the intra-cellular concentration of this substance as a deter-minant of the rate of purine synthesis de novo (Fig.3).

Adjacent location of HPRT and PRPP synthetase onthe X chromosome

Pedigrees of families carrying the gene for increasedPP-ribose-P synthetase were compatible with anX-linked inheritance. Two cell populations showingnormal and mutant enzyme have been found.5152In collaborative studies with Dr Goss at Oxford,using human-rodent hybrid cells, the genes forHPRT and PP-ribose-P synthetase were located neareach other on the long arm of the X chromosome.'02This is the first example of genes for 2 successiveenzyme reactions being found on the same mam-malian chromosome.

Inosinic dehydrogenase deficiency as a potentialenzyme defect

An increased purine synthesis accompanying a grossimpairment of inosinic dehydrogenase activity hasso far been studied only in cultured human cellsexposed to pharmacological agents known toincrease serum and urinary uric acid when admini-stered in vivo. 2-ethyl amino 1, 3, 4-thiadiazole, anantitumour drug, when given to patients produced amarked increase in both serum urate concentrationand in the excretion of urinary uric acid.53 Its clinicaladministration greatly enhanced the incorporation ofisotopically labelled glycine into urinary acid uric.'5Studies in vitro by Nelson et al.54 55 demonstrated amarked inhibition of the enzyme, inosinic dehydro-genase, through formation ofan inhibitory thiadiazolanalog of NAD. Ribavirin, an antiviral drug, alsoinhibits inosinic dehydrogenase and in culturedhuman lymphoblasts produces a marked increasein excretion of hypoxanthine into the medium.56It thus becomes a model for an additional her-editary defect that remains to be diagnosed amongour gouty patients who are producing excessive

Human aberrations ofpurine metabolism and their significance for rheumatology 109

quantities of uric acid. We can supply shippingvials with culture media for sending blood frompotential candidates to our laboratory for study.

Other enzyme defects of purine metabolism

Adenine phosphoribosyltransferase deficiency. Thisdeficit was first identified in heterozygote state byKelley et al.103 from our routine assay of this enzymeas a control in our HPRT studies. The homozygotesfor this disorder excrete large amounts of adeninein the urine along with 2-8-dioxyadenine, which isvery sparingly soluble and forms calculi in theurinary tract as the presenting symptom that areeasily confused with uric acid 57-59 (Fig. 3).

Adenosine kinase deficiency. This mutation hasso far been described only in cultured humanlymphoblasts from a mutation derived in vitro.60No enhanced rate of purine synthesis was found.Its clinical presentation remains to be described(Fig. 3).

Xanthinuria. A marked hypouricaemia with thereplacement of uric acid in urine by xanthine andhypoxanthine was first described by Dent andPhilpot6l in a child who passed a xanthine calculusin the urine. The gross deficiency of xanthine oxidase,the enzyme responsible for uric acid formation, wassubsequently demonstrated by Ayvazian et al.62and Watts et al.63 64 (Fig. 3). Patients are usuallydetected by finding a profound hypouricaemia inroutine tests. Around one-third of the known caseshave developed xanthine calculi of the urinary tract.9Three patients who had myalgias showed depositsof xanthine crystals in frozen sections of the musclebiopsies and one such patient was found in arheumatology clinic.6566

T-CELLT-CELL EFFECTOR

BURSALEQUIVALENT

B-CELL

Defects of purine metabolism associated with immuno-deficiency disease

In the past 3 decades remarkable progress has beenmade in our understanding of the role of variouscomponents of the immune system. The originand role of T and B cells has been defined (Fig. 4)and many phenomena regulating the response inthis 'black box' have been identified.67 The recentbiochemical identification of specific enzyme defectsassociated with immune dysfunction has now pro-vided a crack in the black box to let us glimpsesome of the biochemical processes involved inregulation of the immune response.Enzyme defects. Gross deficiency of each of 3

sequential enzymes involved in the catabolic degrada-tion of purine nucleotides have been reported inpatients with various types of defects of the immunesystem (Fig. 5). The first enzyme defect identifiedwas a gross deficiency ofadenosine deaminase (ADA)noted in red cells of 2 children with severe combinedimmunodeficiency disease by Dr Giblett and associ-ates in 1972.68

All patients with ADA deficiency show a uniformsevere impairment of T-cell function and varyingdegrees of impairment of B-cell function, with asevere lymphopenia. Some 3 years later she reportedthe first patient with a gross deficiency of purinenucleoside-phosphorylase (PNP) associated with anisolated defect in T-cell function.69 As might beexpected, this patient excreted large amounts of theconventional substrates for the missing enzymesinosine and guanosine; but quite unexpected was therelatively large amount of the corresponding deoxy-nucleosides, deoxyinosine, and deoxyguanosine,amounting to around one-third of total nucleosides. '

- LYMPHOKINES

Fig. 4 The origin of T and Bcells. From Waldmann andBroder.101

ANTIBODIESIgMIgGIgAIgDIgE

110 Seegmiller

ENZYME DEFECT

NH2

NNDNYN-RCBODE-P R CLEO/DASE

OR

DEOXYADENYLIC ACID

NH2

N N\ JOfNOSINEN N-RIBOSE D#NS

ADENOSINE ORDEOXYADENOSINE

NH2

N NN>RIBOSE

ADENOSINE ORDEOXYADENOSINE

OH

iII> + NH3i N- RIBOSE

INOSINE ORDEOXYINOSINE

OH OH

PUR/NE NCXLIVE

BSE

+ pi P1N0SPORMLASEO N>+ RIBOSE-IP

INOSINE OR HYPOXANTHINE ORDEOXYINOSINE DEOXYRIBOSE-I-P

Fig. 5 Enzyme defects ofpurine nucleotide catabolismassociated with various immunodeficiency diseases.From Seegmiller et al.83

On a molar basis the total daily purine excreted wasquite comparable to the excessive rate of purinesexcreted by patients with Lesch-Nyhan disease.Since such children are incapable of making hypo-xanthine (or consequently uric acid) they show a

profound hypouricaemia. Both of these disorders aredefinitely genetic in origin.More recently Johnson et al.71 and Webster et al.72

have reported a markedly decreased activity ofecto-purine-5'-nucleotidase, the first enzyme in thissequence in lymphocytes of patients with adult onsetagammaglobulinaemia and a similar reduction inenzyme activity has been reported by Edwards et al.73in lymphocytes of patients with X-linked agamma-globulinaemia. However, the enzyme ecto-5'-nucleotidase is most abundant in normal mature Bcells and is virtually absent from these cells at birth,suggesting that it may also be a marker for B-cellmaturation, since it is also low in activity in manypatients with chronic lymphocytic leukaemia.7476The low activities reported in peripheral lymphocytesmay therefore reflect merely the reduced number offunctioning B cells present in these hypogamma-globulinaemic states. Against this view are reportsof low activity in T cells in patients with X-linkedagammaglobulinaemia.7 78

Mechanisms of immunodeficiency

The concept of a severe impairment of the immunesystem being the consequence of a single enzymedeficit was quite foreign to the thinking of mostimmunologists. One ofthe proposals made to accountfor it was that the genetic locus for ADA mightreside in close proximity to the locus for immuneresponse genes and that both were lost with a

IMMUNE SYSTEM genetic deletion of these 2 regions of the chromo-DEFECT Such deletions, however, quite and

over 2 dozen families have now been described inB-CELL which this association has been found. Furthermore,

the gene for adenosine deaminase has now beenlocated on chromosome 20 and that for the immuneresponse genes is chromosome 6, thus making the

T-o4 B-Cellconcept quite untenable.Enzyme distribution. Knowledge of the distribution

of the enzyme adenosine deaminase shown in Table1 might well have allowed us to anticipate the con-

sequences of its deficiency. The enzyme is mostT-CELL active in thymus, the site of origin of T cells, and

next highest in spleen, lymph nodes, and otherlymphoid organs.79 80

Metabolic consequences of ADA deficiency

The unexpectedly large amounts of deoxynucleosidesproduced by children with PNP deficiency providesevidence that the degradative pathway for purineribonucleotides shown in Fig. 5 serves also fordegradative metabolism of deoxynucleosides as well.Although one might expect a substantial accumu-

lation of both adenosine and deoxyadenosine in

Table 1 Enzyme activities in human tissues

Tissue Deoxy- Adenosine Purineadenosine deaminase* nucleosidekinase* phosphorylase*

Thymus 0.78 282-8 23.3Spleen 0.20 12-4 54.0Brain 0.14 5*0 10-3Kidney 0.07 1-8 100.0Liver 0.07 1.1 36-2Lung 0.06 0.8 38.0Small intestine 0.08 14.2 63.9Heart 0.08 2-1 32-2Peripheral lymphocytes 0.32 20.7 114.7Peripheral granulocytes 0-05 11.9 121.4

Human tissues were obtained from one baby who died during partu-rition, and peripheral lymphocytes and granulocytes were isolatedfrom the blood of a normal adult.Activities are expressed as nmoles of product per minute per milli-gram of protein at a substrate concentration of 300 pM. From Carsonet al.80

METABOLIC FATES OF ADENOSINE AND DEOXYADENOSINE

S-ADENOSYL HOMOCYSTEINE

ADENOSINE Ade?ne Kinase Km = 2 Wz ADENYLIC ACID

-INOS~~~~NOINE

ADA iZ-D.EDEX EYINOSINEDEOXYADENOSINE ,

~DDEOXYADENYLIC ACIDKm = 400 uM

Fig. 6 Alternate pathways of metabolism for adenosineand deoxyadenosine. From Seegmiller et al.83

Human aberrations ofpurine metabolism and their significance for rheumatology 111

ADA-deficient patients, only deoxyadenosine ac-cumulates to any large extent and is excreted in theurine in increased amounts.81 82 Potential pathwaysfor further metabolism of adenosine and deoxy-adenosine are shown in Fig. 6. The failure of thesechildren to excrete any appreciable amounts ofadenosine could be related to either the high affinityof adenosine for its kinase or to use of an alternativepathway (not shown) for breakdown of adenylic acidby the enzyme adenylic acid deaminase.83

Adenosine deaminase deficiency

Effect ofadenosine and deoxyadenosine on cellgrowth.The availability of a number of inhibitors of ADA,

luM lOumMMolarity of Inhibitor

100QM

Fig. 7 Effect of adenosine and deoxyadenosine on thegrowth ofa human leukaemic T-cell line and B-cellline. All cultures contain 5 F±M EHNA. From Carsonet al.99

the most potent of which is deoxycoformcin with aKi of 2 5 x 1O-12M and erythro-9-(2 hydroxyl-3nony) adenine (EHNA) which has a Ki of 1I3 x10-9, has allowed the study of the metabolic con-sequences of impaired ADA function in a varietyof cell types. Adenosine, at relatively low concen-trations, is lethal for cultured mammalian cells oflymphoid origin where it produced a gross deficiencyof pyrimidine nucleotides which was reversed byuridine, suggesting that this was the main actionof adenosine toxicity.84 A similar type of adenosinetoxicity in human lymphoblast lines treated withEHNA was reversed only in part by uridine. Thefailure of uridine to restore either mitogen res-ponsiveness of lymphocytes of ADA-deficientchildren when studied in vitro or immune functionwhen given in vivo argues strongly against thismechanism being a valid model of the disease.8587As shown in Fig. 7, deoxyadenosine inhibited pro-liferation of lymphoblasts at lower concentrationsthan did adenosine, and a T-cell line was far moresensitive to this inhibition than was a B-cell line.As shown in Table 2 the inhibition of the T-cellline was accompanied by a marked accumulationof deoxyATP. This greater accumulation in the

Table 2 Effect of deoxyadenosine on deoxy-ATP andATP concentrations in T and B cell linesCell line Deoxy- EHNA ATP Deoxy-ATP

adenosine (PM) (pmoles/106 (pmoles/106(PM) cells) cells)

T cells 2100+78 (SEM) <50T cells 20 5 1900±165 607+207B cells 2753+1054 <50B cells 20 5 2075±582 <50

From Carson et al.99

Fig. 8 Breakdown ofdATP byT- and B-cell lines. The T-celllines (. left) and the B-cell line(o right) at a density of5 x 106cellsfml were incubatedfor threehours with 1- 0 mMdeoxyadenosine withoutdeoxycoformycin and then washedandplaced in fresh medium. Atvarious time points thereafter thecells were washed and extractedand deoxyATP concentration weredetermined by DNA preliminaryassays. From Carson et al. ProcNatl Acad Sci USA 1979; 76:2430-3.

Time (minsJ

a-

co3-

ac1=

E4

112 Seegmiller

T-cell line was associated with a far less rapiddisappearance of deoxyATP upon removing deoxy-adenosine from the media than was seen in B-cellline (Fig. 8) suggesting a major role for degradativeenzymes in accounting for this difference in response.In broken cell preparations, shown in Table 3,deoxyATP, deoxyGTP and thymidine triphosphatewere all degraded far more slowly by T-cell thanB-cell lysates. This slower rate of degradation wascorrelated also with a substantially lower activityof ecto-purine-5'-nucleotidase in the T-cell line(Table 4).The addition of deoxycoformycin and deoxy-

adenosine to phytohaemagglutinin-stimulated lym-phocytes from normal peripheral blood also results ina substantial elevation in deoxyATP concentrationunder conditions where 14C leucine incorporationis grossly inhibited. As shown in Fig. 9, this inhibitionwas substantially overcome by addition of variousother single deoxynucleosides and a combinationof either thymidine or deoxycytidine with deoxy-guanosine completely overcame the inhibition.88

Since normal T cells from peripheral blood alsoshowed a substantially lower activity of ecto-purine-5'-nucleotidase than B cells, as shown inFig. 10, in patients with ADA deficiency, a similar

Table 3 Deoxyribonucleoside triphosphate catabolismin broken cells

Cell lines Substrate

dATP dGTP dTTP

T 11-1±7-6 10-0±5.2 4-0±1-2B 102-3±32 67-3±18 84-9±12

Three leukaemic T cell lines and 3 B cell lines were lysed by freezingand thawing and incubated with 300 jM tritiated dATP, dGTP, ordTTP. After 30 min for the B cell lines or 120 min for the T cell lines,nucleotides were separated from nucleosides and bases by thin-layerchromatography. Activities are expressed as pmol of product (nucleo-side + base)/min per 106 cells ± SD (n -I method). From Carsonet al. Proc Natl Acad Sci 1979; 76: 2430-3.

Table 4 5'-Nucleotidase activity in T and B cell lines

Cell lines Whole cells Broken cells

-AMP- +AMP- -AMP- +AMP-(CH2)P (CH2)P (CH2)P (CH2)P

T 3-07±1-4 3-07±1-4 46±8-5 42±15B 387±319 14±7-0 330±225 62±15

To three T cell and 3 B cell lines at densities of 1 5-3 *0 x 106 cellsper ml in a buffer containing 8 f6 mM magnesium chloride and 50 mMTris-HCI at pH 6-9 was added 215 jM (8-14C) inosine monophos-phate (specific activity 6 mCi/mmol, Amersham/Searle) in a finalvolume of 35 jl, either with or without 2 - 86 mM AMP-(CH2)P. After15-60 min at 370C the reaction was terminated by the addition of 5 Alof8M perchloric acid, and the products inosine and hypoxanthine wereseparated from inosine monophosphate by thin-layer chromatography.Activities are expressed as pmol of product (inosine + hypoxanthine)min per 106 cells ± SD for the 3 T and the 3 B cell lines. FromCarson et al. Proc Natl Acad Sci USA 1979; 76: 2430-3.

110T

100+

o 80U

0a 70CTt-

oERi 50600

0

03 30.0-i

* 20

10.

1c0eL

m..

,0:

ID

.x

5

.......

@:.EC_:.+

,1@.,,._

o,.c.,.'o.@.n

i..I,).

-.,iS,,

.........

.........

..:-.. .B:,. ......, ",:

, N {,,

X>

', ,Bi.

3..:j,:

All Nucleosides Added to 10;LM Final Concentration

Fig. 9 Inhibition ofPHA-induced proliferation ofhuman lymphoblast by 10 tim deoxyadenosine and 1 FiMdeoxycoformycin with reversal by addition ofdeoxynucleosides (10 M). From Bluestein et al.88

greater sensitivity of T cells to growth inhibitionand a greater accumulation of deoxyATP might beexpected.80 Both erythrocytes and lymphocytes ofADA-deficient children show marked increases indeoxyATP which decrease after therapy of thepatients with irradiated erythrocytes.85 89-91

Mechanism of immunosuppression of T cells andexcessive purine synthesis in PNP deficiency

A similar mechanism of inhibition of T-cell functionby accumulation of deoxyGTP has been proposedfor patients with PNP deficiency as a result ofaccumulation of deoxyguanosine in this disease.89 90At the present time the accumulation of deoxyGTPhas been shown only in erythrocytes of affectedchildren.The common denominator to account for the

excessive rate of purine synthesis in children withPNP deficiency and HPRT deficiency is the failureof children in both these diseases to reutilise hypo-xanthine. Children with HPRT deficiency are missingthe enzyme for utilising hypoxanthine while thosewith PNP deficiency are unable to make hypoxan-thine. Studies of human cells studied in vitro show

o.

Human aberrations ofpurine metabolism and their significance for rheumatology 113

(a

co

._

0)

C-CA)

a)

O.- -a.'

o

. E2 cn

=-5-Z

UcJ

Ua,

:a

coa)

CL

CD,a-0

PBM Rosetting Non- PBM Rosetting Non- PBM Rosetting Non-Cells rosetting Cells rosetting Cells rosetting

Cells Cells Cells

that the rate ofsynthesis ofpurines de novo in normalcells can be accelerated to values quite comparableto that seen in cells from children with Lesch-Nyhandisease by removing all hypoxanthine from themedium.92

Proposed mechanism of suppression of T-cell pro-liferation in ADA and PNP deficiency

A mechanism by which deoxyATP accumulationcan arrest cellular proliferation is provided by thevery well documented allosteric inhibition of allactivity of the enzyme ribonucleoside diphosphatereductase by deoxyATP 93 94 (Fig. 11). This enzymeconstitutes the only known route of deoxyribo-nucleotide synthesis from ribonucleotides. Con-sequently its inhibition could result in cessation of

Fig. 10 Ecto-S'-nucleotidaseactivity in lymphocytesfromnormal subjects and patients withX-linked agammaglobulinaemiaand their female relatives. FromThompson et al.76

DNA synthesis from an inadequate supply of thedeoxyribonucleoside triphosphate substrates. Theability of mixtures of deoxynucleosides to overcomethe toxicity of deoxyadenosine could readily beexplained by their providing a source for the for-mation of the other deoxynucleotides that are inshort supply. An alternative explanation could bean inhibition of deoxyATP formation due to theother deoxynucleosides competing with deoxy-adenosine in the deoxynucleoside kinase or sub-sequent phosphorylation reactions.An analogous mechanism could well be operating

for the immunodeficiency seen in patients with PNPdeficiency. DeoxyGTP is also a known allostericinhibitor of ribonucleoside diphosphate reductaseand it specifically inhibits the formation ofdeoxyCTPand TTP.9394 Deoxyguanosine toxicity for mouse

AdenosineDeoxyadenosine

CDP

GDP

ADP

ADAY, InosineDeoxyinosineGuanosine

Deoxyguanosine

dATP dGTP cRibonucleotide / dcReductase c

Hx+R-1-PPNP Hx+dR-1lP

Gu+R-lPGu+dR-1-P Fig. 11 Postulated mechanism

of inhibition of lymphocytet- TTP proliferation in genetic deficiency

ofADAorPNP.dCDP - dCTPdGTP - dGTP -.- DNAdADP -w dATP

- - I

114 Seegmiller

T-lymphoma cells has been reversed by the additionof deoxycytidine (and hypoxanthine) to the culturemedium.95The reversal of deoxyadenosine or deoxyguanosine

toxicity in cultured cell systems by the addition ofother deoxynucleosides suggests the possibility thatthese deoxynucleosides, particularly deoxycytidine,may be useful in treatment of patients with ADAdeficiency or PNP deficiency. A few clinical attemptshave been made to accomplish such a reversal,but no remarkable therapeutic benefit was produced.Success will depend on whether or not the patienthas an adequate population of viable stem cells thatcan respond to differentiation into immunocompetentcells and whether appropriate routes of admini-stration for the deoxynucleosides are used to deliverthem to the stem cells at sufficiently high concen-tration. Here in England, as well as in the UnitedStates, cautious evaluations are under way to applythis new knowledge to the clinical control of T-cellleukemia by use of deoxycoformycin.9698

Ecto-5'-nucleotidase activity in X-linked agamma-globulinaemia

In a recent study Drs Thompson and Boss76 in ourlaboratory separated T cells from B cells in peri-pheral blood by rosetting them with neuraminidase-treated sheep erythrocytes. As shown in Fig. 10,normal nonrosetting (B) cells, inagreement withRoweet al.,74 showed at least 3 times the activity of normalrosetting (T) cells. Monocytes and null cells showednegligible activity. In 5 patients with X-linked agam-maglobulinaemia the nonrosetting (B) cells showedless than 5% of normal activity and failed to showany B cells with surface immunoglobulin. Of specialinterest was the finding of only 56% of normalactivity in their rosetting (T) cells. However, thesignificance of this finding is not clear at this time.A summary of the known enzyme defects in

purine metabolism is shown in Fig. 3.

Conclusion

I have reviewed today for you the valuable insightinto a variety of biological processes that have beenprovided by human mutations in purine metabolism,many of them in patients with rheumatic diEorders.These mutations provide a dissecting probe andspotlight into biology for unravelling the complexinterrelations involved in biological systems. Detailedstudies of patients with alcaptonuria led Garrodto formulate the gene-enzyme hypothesis, which inturn provided the key for understanding a myriadother hereditary diseases. Once the abnormal geneproduct has been identified substantially more insight

into biological process of medical interest can beobtained. In many cases, as we have seen today forLesch-Nyhan disease, what was previously a curiousor even tragic clinical problem quite beyond ourcapability of understanding now becomes a fas-cinating 'experiment of nature' which with properstudy is capable of revealing a rational sequence bywhich the abnormal gene product gives rise to theclinical expression and disease, and frequentlyrational approaches to therapy. The identificationof enzyme defects associated with immunodeficiencydiseases have already provided us with valuableinformation on biochemical differences in T and Bcells, which could be used as the basis for develop-ment of more specific pharmacological agents forfine tuning correction of aberrations of the immunesystem which we now realise are responsible for awide range of diseases, including many of ourrheumatic diseases.

But the work has only begun. The enzyme defectsidentified so far in patients with gouty arthritis arebut a start, since they account for less than 5 % ofpatients with clinical gout seen in most clinics.However, with this knowledge we have identifiedin cell culture other defects in purine metabolismthat should be found as enzyme deficits in the goutypopulation. The abnormal gene products responsiblefor the hereditary tendency for development ofmany other common rheumatic diseases, such asdegenerative joint disease and chondrocalcinosis, andthe mechanism by which they produce pathologyremain to be identified. A similar need exists foridentification of additional genetic factors res-ponsible for susceptibility to other of our commondiseases of familial nature, including cardiovasculardisease, hypertension, neuromuscular diseases, andmental illness.

Perhaps the greatest promise of this new know-ledge of genetics is the potential it holds for earlydetection of specific disease-prone states and for theformulation of a rational therapeutic intervention.This knowledge, in turn, can shift our emphasis inthe future practice of medicine from merely awaitingthe clinical presentation of advanced pathology to awhole new orientation of preventive medicine inwhich we identify disease susceptibility at birthor even before birth and then use a lifetime ofpreventive approaches to its control. For this visionof medical practice of the future to become a realitywe must provide more opportunities and incentivefor our most gifted and inspired younger physiciansto become immersed in research endeavours at bothbasic and clinical levels. The potential savings can begreat in cost of medical care, in effective use ofmedical manpower and most important in the healthof our citizens.

Human aberrations ofpurine metabolism and their significance for rheumatology 115

The work was supported in part by National Institutes ofHealth grants GM 17702 and AM 13622 and a grant fromthe Kroc Foundation.

References

Garrod A B. Observations on certain pathologicalconditions of the blood and urine in gout, rheumatismand Bright's disease. Trans Med-Chir Soc (London) 1848;31: 83-98.

2 Garrod A B. The Nature and Treatment of Gout andRheumatic Gout. London: Walton and Maberly, 1859.

3 Garrod A E. The Croonian lectures on inborn errors ofmetabolism. Lecture II, Alkaptonuria. Lancet 1908;2: 73.

4 Beadle G W, Tatum E L. Genetic control of biochemicalreactions in Neurospora. Proc Natl Acad Sci USA 1941;27: 499.Garrod A E. Inborn Errors of Metabolism. London,Oxford. 1923.

6 McCarty D J. Phagocytosis of urate crystals in goutysynovial fluid. Arthritis Rheum 1961; 4: 425-6.

7 McCarty D J Jr. Mechanisms of the crystal depositiondiseases-gout and pseudogout. Ann Intern Med 1973;78: 767-71.

8 Seegmiller J E, Howell R R, Malawita S E. Inflammatoryreaction of sodium urate: Its possible relationship togenesis of acute gouty arthritis. JAMA 1962; 180:469-75.

9 Seegmiller J E. Diseases of purine and pyrimidinemetabolism. In: Bondy P K, Rosenberg L E, eds.Metabolic Control and Disease, 8th ed. Philadelphia:Saunders, 1980: 777-937.

10Seegmiller J E, Grayzel A I, Laster L, Liddle L. Uric acidproduction in gout. J Clin Invest 1961; 40: 1304-14.Rundles R W, Wyngaarden J B, Hitchings G H, ElionG B, Silberman H R. Effects of a xanthine oxidaseinhibitor on thiopurine metabolism, hyperuricemia, andgout. Trans Assoc Am Phys 1963; 76: 126-40.

12 Benedict J D, Forsham P H, Roche M, Soloway S,Stetten D W Jr. The effect of salicylates and adreno-corticotropic hormone upon the miscible pool of uricacid in gout.J Clin Invest 1950; 29: 1104-11.

13Kaufman J M, Greene M L, Seegmiller J E. Urine uricacid to creatinine ratio-A screening test for inheriteddisorders of purine metabolism. J Pediat 1968; 73:583-92.

14 Simkin P A, Hoover P L, Paxson C S, Wilson W F. Uricacid excretion: quantitative assessment from gout,midmorning serum and urine samples. Ann Intern Med1979; 91: 44-7.

15 Seegmiller J E, Grayzel A I, Liddle L, Wyngaarden J B.The effect of 2-ethyl-amino-1, 3, 4-thiadiazole on theincorporation of glycine into urinary purines and uricacid in man. Metabolism 1963; 12: 507-15.

16 Schulman J D, Lustberg T J, Kennedy J L, Museles M,Seegmiller J E. A new variant of maple syrup urinedisease (branched chain ketoaciduria): Clinical andbiochemical evaluation. Am J Med 1970; 49: 118-24.

17 Alepa F P, Howell R R, Klinenberg J R, Seegmiller J E.Relationships between glycogen storage disease andtophaceous gout. Am J Med 1967; 42: 58-66.

18 Lesch M, Nyhan W L. A familial disorder of uric acidmetablism and central nervous system function. Am JMed 1964; 36: 561-70.

19 Seegmiller J E, Rosenbloom F M, Kelley W N. Enzymedefect associated with a sex-linked human neurologicaldisorder and excessive purine synthesis. Science 1967;155: 1682-4.

20 Seegmiller J E. Inherited deficiency of hypoxanthine-guanine phosphoribosyltransferase in X-linked uricaciduria (the Lesch-Nyhan syndrome and its variants).In: Harris H, Hirschhorn K, eds. Advances in HumanGenetics. Plenum Publishing Corp., New York; Plenum,1976; 6: 75-163.

21 Lyon M F. Gene action in X chromosome of the mouse(Mus musculus L). Nature 1961; 190: 372-3.

22 Rosenbloom F M, Kelley W N, Henderson J F, Seeg-miller J E. Lyon hypothesis and X-linked disease. Lancet1967; 2: 305-6.

23 Gartler S M, Scott R C, Goldstein J L, Campbell B,Sparkes R. Lesch-Nyhan syndrome: Rapid detection ofheterozygotes by use of hair follicles. Science 1971; 172:572-4.

24 Fujimoto F Y, Seegmiller J E, Uhlendorf B W, JacobsonC B. Biochemical diagnosis of an X-linked disease inutero. Lancet 1968; 2: 511-2.

25 Boyle J A, Raivio K 0, Astrin K H, et al. Lesch-Nyhansyndrome: preventive control by prenatal diagnosis.Science 1970; 169: 688-9

26 Van Heeswijk P J, Blank C H, Seegmiller J E, JacobsonC B. Preventive control of the Lesch-Nyhan syndrome.Obstet Gynecol 1972; 40: 109-13.

27 McInnes R, Lamm P, Clow C L, Scriver C R. A filterpaper sampling method for the uric acid: creatinineratio in urine. Normal values in the newborn. Pediatrics1972; 49: 80-4.

28 Skaper S D, Seegmiller J E. Hypoxanthine-guaninephosphoribosyltransferase mutant glioma cells: Dimin-ished monoamine oxidase activity. Science 1976; 194:1171-3.

29 Breakefield X 0, Castiglione C M, Edelstein S B. Mono-amine oxidase activity decreased in cells lacking hypo-xanthine phosphoribosyltransferase activity. Science192: 1976; 1018-20.

3 0 Mizuno T, Yugari Y. Prophylactic effect of L-5-hydroxy-tryptophan on self-mutilation in the Lesch-Nyhansyndrome. Neuropadiatrie 1975; 6: 13-23.

31 Frith C D, Johnstone E C, Joseph M H, Powell R J,Watts R W E. Double-blind clinical trial of 5-hydroxy-tryptophan in a case of Lesch-Nyhan syndrome. J NeurolNeurosurg Psychiat 1976; 39: 656-62.

32 Ciaranello R D, Anders T F, Barchas J D, Berger P A,Cann H M. The use of 5-hydroxytryptophan in a childwith Lesch-Nyhan syndrome. Child Psychiatry Hum Dev1976; 7: 127-33.

33 Nyhan W L. Behavior in the Lesch-Nyhan syndrome. JAutism Child Schizoph 1976; 6: 235-52.

34 Becker M A, Seegmiller J E. Genetic aspects of gout. InAnnual Review of Medicine. Annual Reviews, 1974: 25:15-28.

35 Kelley W N, Rosenbloom F M, Henderson J F, Seeg-miller J E. A specific enzyme defect in gout associatedwith overproduction of uric acid. Proc Natl Acad Sci USA1967; 57: 1735-9.

36 Kelley W N, Greene M L, Rosenbloom R M, HendersonJ F, Seegmiller J E. Hypoxanthine-guanine phosphoribo-syltransferase deficiency in gout. A review. Ann InternMed 1969; 70: 155-206.

3 7 Greene M L. Clinical features of patients with the 'partial'deficiency of the X-linked uric aciduria enzyme. ArchIntern Med 1972; 130: 193-198.

38 Emmerson B T, Thompson L. The spectrum of hypo-xanthine-guanine phosphoribosyltransferase deficiency.Q J Med 1973; 42: 423-40.

3 9 Kelley W N, Rosenbloom F M, Miller J, Seegmiller J E.An enzymatic basis for variation in response to allo-purinol. New Engl J Med 1968; 278: 287-93.

116 Seegmiller

40 Rosenbloom F M, Henderson J F, Caldwell I C, KelleyW N, Seegmiller J E. Biochemical bases of acceleratedpurine biosynthesis de novo in human fibroblasts lackinghypoxanthine-guanine phosphoribosyltransferase. J BiolChem 1968; 243: 1166-73.

41 Brenton D P, Astrin K H, Cruikshank M K, SeegmillerJ E. Measurement of free nucleotides in cultured humanlymphoid cells using high-pressure liquid chromato-graphy. Biochem Med 1977; 17: 231-47.

42 Nuki G, Astrin K, Brenton D, Cruikshank M, Lever J,Seegmiller J E. Purine and pyrmidine nucleotide con-centrations in cells with decreased hypoxanthine-guaninephosphoribosyltransferase (HGPRT) activity. In: Ad-vances in Experimental Medicine and Biology, Vol. 76A.(Series: Purine Metabolism in Man-II: Regulation ofPathways and Enzyme Defects). Muller, Kaiser, Seeg-miller eds. New York: Plenum Press, 1977: 326-39.

43 Becker M A. Patterns of phosphoribosylpyrophosphateand ribose-5-phosphate concentration and generationin fibroblasts from patients with gout and purine over-production. J Clin Invest 1976; 57: 308-18.

44 Becker M A, Kostel P J, Meyer L J, Seegmiller J E.Human phosphoribosylpyrophosphate synthetase: In-creased enzyme specific activity in a family with gout andexcessive purine synthesis. Proc Nat Acad Sci USA 1973;70; 2749-52.

45 Becker M A, Meyer L J, Wood A W, Seegmiller J E.Purine overproduction in man associated with increasedphosphoribosylpyrophosphate synthetase activity. Science1973; 179: 1123-6.

4 6 Becker M A, Meyer L J, Seegmiller J E. Gout with purineoverproduction due to increased phosphoribosylpyro-phosphate synthetase activity. Am J Med 1973; 55: 232-42.

47 Sperling 0, Eilam G, Persky-Brosh S, de Vries A.Accelerated erythrocyte-5-phosphoribosyl-l-pyrophos-phate synthesis. A familial abnormality associated withexcessive uric acid production and gout. Biochem Med1972; 6: 310-316.

48 Zoref E, de Vries A, Sperling 0. Mutant feedback-resistant phosphoribosylpyrophosphate synthetase asso-ciated with purine overproduction and gout. Phosphoribo-sylpyrophosphate and purine metabolism in culturedfibroblasts. J Clin Invest 1975; 56: 1093-9.

49 Nyhan W L, James J A, Teberg A J, Sweetman L,Nelson L G. A new disorder of purine metabolism withbehavioral manifestations. J Pediat 1969; 74: 20-7.

50 Becker M A, Raivio K 0, Bakay B, Adams W B, NyhanW L. Superactive phosphoribosylpyrophosphate (PRPP)synthetase with altered regulatory and catalytic functions.3rd International Symposium on Purine Metabolism inMan, Madred, Spain, 11-15 June 1979.

51 Zoref E, de Vries A, Sperling 0. (1977). Evidence for X-linkage of phosphoribosylpyrophosphate synthetase inman. Studies with cultured fibroblasts from a goutyfamily with mutant feedback-resistant enzyme. HumHered 1977; 27: 73-80.

52 Yen R C K, Adams B, Lazar C, Becker M A. Evidencefor X-linkage of human phosphoribosylpyrophosphatesynthetase. Proc Natl Acad Sci USA 1978; 75: 482-4.

53 Krakoff I H, Balis M E. Studies on the uricogenic effectof 2-substituted thiadiazoles in man. J Clin Invest 1959;38: 907.

54 Nelson J A, Rose L M, Bennett L L Jr. Effects of 2-amino-l, 3, 4-thiadiazole on ribonucleotide pools ofleukemia L1210 cells. Cancer Res 1976; 36: 1375-8.

55 Nelson J A, Rose L M, Bennett L L Jr. Mechanism ofaction of 2-amino-1, 3, 4-thiadiazole (NSC 4728).Cancer Res 1977; 37: 182-7.

56 Willis R C, Carson D A, Seegmiller J E. Adenosine

kinase initiates the major route of ribavirin activation in acultured human cell line. Proc Nail Acad Sci USA 1978;75: 3042-4.

57 Cartier M P, Hamet M. A new metabolic disease: Thecomplete deficit of adenine phosphoribosyltransferaseand lithiasis of 2, 8-dihydroxyadenine. CR Acad Sci(Paris) 1974; 279: 883-6.

58 Simmonds H A, Van Acker K J, Cameron J S, SneddenW. The identification of 2, 8-dihydroxyadenine, a newcomponent of urinary stones. Biochem J 1976; 157: 485-7.

69 Barratt T M, Simmonds H A, Cameron J S, et al. Com-plete deficiency of adenine phosphoribosyltransferase:A third case presenting as renal stones in a young child.Arch Dis Child 1979; 54: 25-31.

60 Hershfield M S, Spector E B, Seegmiller J E. Purinesynthesis and excretion in mutants of the WI-L2 humanlymphoblastoid line deficient in adenosine kinase (AK)and adenine phosphoribosyltransferase (APRT). InMuller Kaiser Seegmiller, eds. Advances in ExperimentalMedicine and Biology, Vol. 76A. (Series: Purine Meta-bolism in Man-II: Regulation of Pathways and EnzymeDefects). New York: Plenum Press, 1977: 303-13.

61 Dent C E, Philpot G R. Xanthinuria, an inborn erroror deviation) of metabolism. Lancet 1954; 1: 182-5.

62 Ayvazian J H. Xanthinuria and hemochromatosis.N EnglJ Med 1964; 270: 18-22.

63 Watts R W E, Engelman K, Klinenberg J R, SeegmillerJ E. Sjoerdsman A. The enzyme defect in xanthinuria.Nature 1964; 201: 395-6.

64 Engelman K, Watts RW E, Klinenberg J R, Sjoerdsma A,Seegmiller J E. Clinical, physiological and biochemicalstudies of a patient with xanthinuria and pheochromo-cytoma. Am J Med 1964; 37: 839-61.

65 Chalmers R S, Watts R W E, Bitensky L, Chayen J.Microscopic studies on crystals in skeletal muscle fromtwo cases of xanthinuria. J Pathol 1969; 99: 45.

66 Isaacs H, Hefferon J J A, Berman L, Badenhorst M,Pickering A. Xanthine hypoxanthine and muscle pain.Histochemical and biochemical observations. S Afr MedJ 1975; 49; 1035-8.

67 Eisen H N. Immunology. An introduction to molecular andcellular principles of the immune responses. Hagerstown:Harper and Rowe, 1974: 475.

68 Giblett E R, Anderson J E, Cohen F, Pollara B, Meu-wissen H J. Adenosine-deaminase deficiency in twopatients with severely impaired cellular immunity.Lancet 1972; 2: 1067-9.

69 Giblett E R, Ammann A J, Wara D W, Sandman R,Diamond L K. Nucleoside-phosphorylase deficiencyin a child with severely defective T-cell immunity andnormal B-cell immunity. Lancet 1975; 1: 1010-3.

70 Cohen A, Doyle D, Martin D W Jr, Ammann A J.Abnormal purine metabolism and purine overproductionin a patient deficient in purine nucleoside phosphorylase.N EnglJ Med 1976; 295: 1449-54.

71 Johnson S M, Asherson G L, Watts RW E, North M E,Allsop J, Webster A D B. Lymphocyte-purine 5'-nucleotidase deficiency in primary hypogammaglobu-linaemia. Lancet 1977; 1: 168-70.

72 Webster A D B, Asherson G L. Identification andfunction of T cells in the peripheral blood of patientswith hypogammaglobulinemia. Clin Exp Immunol1974; 18: 499-504.

73 Edwards N L, Magilavy D B, Cassidy J T, Fox I H.Lymphocyte ecto-5'-nucleotidase deficiency in con-genital agammaglobulinemia. Clin Res 1978; 26: 513A.

74 Rowe M, DeGase C G, Platts-Mills T A E, AshersonG L, Webster A D B, Johnson S M. 5'-nucleotidase ofB and T lymphocytes isolated from human peripheralblood. Clin Exp Immunol 1979; 36: 97.

Human aberrations ofpurine metabolism and their significance for rheumatology 117

75 Boss G R, Thompson L F, Spiegelberg H L et al.

Lymphocyte ecto-5'-nucleotidase activity as a markerof B-cell maturation. Trans Assoc Am Phys 1980: inpress.

76 Thompson L F, Boss G R, Spiegelberg H L et al. Ecto-5'-nucleotidase activity in T and B lymphocytes fromnormal subjects and patients with congenital X-linkedagammaglobulinemia. J Immunol 1980; 123: 2475-8.

7 Edwards N L, Cassidy J T, Fox I H. Lymphocyte 5'-nucleotidase deficiency: clinical characteristics of theassociated hypogammaglobulinemia. J Clin Chem ClinBiochem 1979; 17: 404.

78 Edwards N L, Gelfaud E W, Burk L, Dosch H -M,Fox I H. Distribution of 5'-nucleotidase in humanlymphoid tissues. Proc Natl Acad Sci USA 1979; 76:3474-6.

79 Adams A, Harkness R A. Adenosine deaminase activityin thymus and other human tissues. Clin Exp Immunol1976; 26: 647-9.

80 Carson D A, Kaye J, Seegmiller J E. Lymphospecifictoxicity in adenosine deaminase deficiency and purinenucleoside phosphorylase deficiency: Possible role ofnucleoside kinase(s). Proc Natl Acad Sci USA 1977;74: 5677-81.

81 Simmonds H A, Panayi G S, Corrigal V. A role forpurine metabolism in the immune response: Adenosine-deaminase activity and deoxyadenosine catabolism.Lancet 1978; 1: 60-3.

82 Kuttesch J F, Schmalstieg F C, Nelson J A. Analysis ofadenosine and other adenine compounds in patientswith immunodeficiency diseases. J Liq Chromatogr1978; 1: 97-109.

83 Seegmiller J E, Thompson L, Bluestein H, Willis R,Matsumoto S, Carson D. Nucleotide and nucleosidemetabolism and lymphocyte function. In: Gelfand E W,Dosch H-M, eds. Biological Basis of Immunodeficiency.New York: Raven Press: 1980; 251-68.

84 Ishii K, Green H. Lethality of adenosine for culturedmammalian cells by interference with pyrmidine bio-synthesis. J Cell Sci 1973; 13: 429-39.

85 Polmar S H, Wetzler E M, Stern R C. Hirschhorn R.Restoration of in vitro lymphocyte responses withexogenous adenosine deaminase in a patient withsevere combined immunodeficiency. Lancet 1975; 2:743-6.

88 Polmar S H, Wetzler E M, Stern R C. Adenosinedeaminase deficiency: Enzyme replacement therapy andinvestigations of the biochemical basis of immuno-deficiency. In: Wamberg E, Guttler E, eds. InbornErrors of Immunity and Phagocytes. Lancaster: NTPPress, 1978.

87 Hirschhorn R. Defects of purine metabolism in immuno-deficiency diseases. In: Schwartz R S, ed. Progress inClinical Immunology. San Francisco: Grune andStratton, 1977: 67-83.

88 Bluestein H G, Willis R C, Thompson L F, MatsumotoS, Seegmiller J E. Accumulation of deoxyribonucleotidesas a possible mediator of immunosuppression in here-tary deficiency of adenosine deaminase. Trans AssocAm Phys 1978; 92: 394-402.

8 9 Cohen A, Gudas L J, Ammann A J, Staal G E J, MartinD W. Deoxyguanosine triphosphate as a possible toxicmetabolite in the immunodeficiency associated withpurine nucleoside phosphorylase deficiency. J Clin Invest1978; 61: 1405-9.

90 Cohen A, Hirschhorn R, Horowitz S D et al. Deoxy-adenosine triphosphate as a potentially toxic metabolitein adenosine deaminase deficiency. Proc Natl Acad SciUSA 1978; 75: 472-6.

91 Coleman M S, Donofrio J, Hutton J J, et al. Identi-fication and quantitation of adenine deoxynucleotidesin erythrocytes of a patient with adenosine deaminasedeficiency and severe combined immunodeficiency.J Biol Chem 1978; 253: 1619-26.

9 2 Hershfield M S, Seegmiller J E. Regulation of de novopurine synthesis in human lymphoblasts: Similar ratesof de novo synthesis during growth by normal cells andmutants deficient in hypoxanthine-guanine phospho-ribosyltransferase activity. J Biol Chem 1977; 252:6002-10.

93 Moore E C, Hurlbert R B. Regulation of mammaliandeoxyribonucleotide biosynthesis by nucleotides asactivators and inhibitors. J Biol Chem 1966; 241: 4802-9.

94 Reichard P. Control of deoxyribonucleotide synthesisin vitro and in vivo. Adv Enzyme Regul 1972; 10: 3-16.

9s Chan T-S. Deoxyguanosine toxicity on lymphoid cellsas a cause for immunosuppression in purine nucleosidephosphorylase deficiency. Cell 14 1978: 523-530.

96 Paine M R, Smyth J F, Harrap K R. Biochemicalconsequences of treatment with the adenosine deaminaseinhibitor 2'-deoxycoformycin. J Clin Chem Clin Biochem1979; 17: 434.

97 Yu A, Kung F, Bakay B, Nyhan W L. Preliminaryclinical trial of deoxycoformycin in human T-cellleukemia. J Clin Chem Clin Biochem 1979; 17: 451-452.

98 Mitchell B S, Koller C, Mejias E, Kelley W N. Inhi-bition of adenosine deaminase by 2'-deoxycoformycinin acute lymphoblastic leukemia. J Clin Chem ClinBiochem 1979; 17: 427-8.

99 Carson D A, Kaye J, Seegmiller J E. Differentialsensitivity of human leukemic T cell lines and B celllines to growth inhibition by deoxyadenosine. J Immunol1978; 121: 1726-31.

100 Seegmiller J E, Laster L, Howell R R. Biochemistry ofuric acid and its relation to gout. N Engl J Med 1963;268: 712-6.

101 Waldmann T A, Broder S, Schwartz R S, ed. Suppressorcells in the regulation of the immune response. In:Schwartz R S ed. Progress in Clinical Immunology. NewYork: Grune and Stratton, 1977; 3.

102 Becker M, Yen R C K, Itkin P, Goss S J, SeegmillerJ E, Bakay B. Regional localization of the gene forhuman phosphoribosylpyrophosphate synthetase onthe X-chromosome. Science 1979; 203: 1016-9.

103 Kelley W N, Levy R I, Rosenbloom F M, HendersonJ F, Seegmiller J E. Adenine phosphoribosyltransferasedeficiency: a previously undescribed genetic defect inman. J Clin Invest 1968; 47: 2281-9.