Clinical Science (1982) 63,429-43s
Purine synthesis de novo in lymphocytes from patients with gout
R . B . G O R D O N , A. C . C O U N S I L M A N , S . M . C . C R O S S A N D B . T. EMMERSON The University of Queensland Department of Medicine at Princess Alexandra Hospital, Brisbane 4102, Queensland, Australia
(Received 4 January124 May 1982; accepted 2 June 1982)
1. Variables that affect the measurement of purine synthesis de novo in human lymphocytes were studied and a reliable method of measurement of purine synthetic activity in these cells was established. 2. Purine synthesis de novo was measured as the rate of incorporation of [14Clformate into u-Nformylglycinamide ribonucleotide when further steps in the biosynthetic pathway had been blocked by azaserine. Incubation was carried out in a synthetic medium with a high phosphate concentration (25 mmol/l). 3. Purine synthesis de novo was measured in lymphocytes obtained on several occasions both from control subjects and from patients with gout, particularly those who tended to overproduce urate as suggested by high values of urinary urate. 4. Lymphocytes obtained from each individual on different occasions showed considerable variations in purine biosynthetic activity. This variation was such that there was no difference between the mean values obtained for the gouty subjects and the control subjects. 5. No correlation was obtained between the mean purine synthetic activity de novo in lymphocytes and either the serum urate concentration or the 24 h urinary urate excretion on a purine-free diet. 6. Apart from those with recognized enzyme mutations, no subgroup of the gouty population has been demonstrated in whom isolated lymphoCorrespondence: Professor B. T. Emmerson, Department of Medicine, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, Queensland 4 102, Australia. 0143-S221/82/110429-07%01.S0
cytes demonstrate an intrinsic abnormality of purine synthesis de novo. Key words: gout, lymphocytes, purine synthesis. Introduction
The relative importance of abnormalities of production and excretion of urate in the aetiology of the hyperuricaemia of gout has never been clearly established. In the first half of this century, hyperuricaemia was mostly attributed to underexcretion of urate, that is, impairment of renal excretion of urate in the presence of otherwise normal renal function 111. However, with the demonstration that some patients with gout incorporated increased amounts of glycine into urate, the importance of excessive endogenous production of urate (over-production) in the aetiology of hyperuricaemia also came to be appreciated [21. After a period in which it was argued that a disorder of either excretion or production was the cause, it became recognized that over-production and under-excretion of urate could both be important factors in the aetiology of the hyperuricaemia of gout. At that stage, a group of patients with gout was defined who demonstrated an impairment of renal excretion of urate 13, 41 and another group was defined who demonstrated abnormalities of urate production, either by the excessive incorporation of labelled precursors into urate or by studies of the miscible urate pool and its turnover [ 5 , 61. Some patients were shown to have evidence of both overproduction and under-excretion simultaneously. However, with the recognition of mutations of
0 1982 The Biochemical Society and the Medical Research Society
R . B. Gordon et al.
hypoxanthine/guanine phosphoribosyltransferase (EC 184.108.40.206) and ribose phosphate pyrophosphokinase (EC 220.127.116.11) [7, 8, 91, the importance of over-production of urate as a mechanism of hyperuricaemia has been emphasized, and this has led to the consideration of the possibility that other more subtle mutations might be more widespread in the community and be contributing to less severe degrees of urate over-production and hence to hyperuricaemia and gout. However, problems have existed in defining the presence of urate over-production in patients without undertaking the rather cumbersome radioisotope studies to demonstrate either an excessive incorporation of glycine into urate or an increased rate of turnover of the urate pool. In 22 normal males on a purine-free diet, the mean urinary urate excretion was found to be less than 3.4 mmoV24 h [61. Accordingly, those patients with gout who excreted more than this value were regarded as demonstrating urate over-excretion, and it was deduced that they probably suffered from urate over-production as well. With the appreciation that both over-production of urate and underexcretion of urate may coexist, it was recognized that the 24 h urinary urate excretion on a purinefree diet might be less than 3.4 mmo1/24 h even in the presence of moderate degrees of urate overproduction. The need arose therefore for the development of a simple index of urate production in subjects with gout without the need to undertake the complicated pool and turnover or glycineincorporation studies. The possibility that abnormal purine biosynthesis might be demonstrable in single cells caused Kamoun et al. (101 to study purine biosynthesis de nouo in lymphocytes from patients with gout. This study revealed considerable overlap in measures of purine bio-
synthesis between normal subjects and those with gout, although the average was higher in the gouty subjects than in the controls. As might be expected from the depletion of PRPP by allopurinol, purine biosynthesis de nouo was reduced during therapy with allopurinol. However, these studies did not follow the purine biosynthetic activity in lymphocytes from the one individual at different times, nor did they provide any assessment of whether those gouty subjects with higher purine synthesis de novo in lymphocytes tended to be those who produced excessive urate. The present study was therefore designed to follow purine synthesis de nouo in lymphocytes from gouty patients who had never been treated with allopurinol and in whom the 24 h urinary urate excretion after 5 days on a purine-free diet was known, thereby providing, with the serum urate, an indication of the balance between urate production and excretion in that individual.
Materials and methods Patients studied Subjects studied were selected from new patients referred to a Gout Clinic because of the severity of their clinical gout. The diagnostic criteria used were those of the Gout Classification Criteria Subcommittee of the American Rheumatism Association [ 1I]. When the acute attack had subsided, and before any specific treatment for the hyperuricaemia (including allopurinol or uricosuric agents), the 2-day mean urinary excretion of urate was determined after the patient had been on an essentially purine-free diet for 5 days. Patients for lymphocyte studies were then chosen whose urate excretion exceeded the normal range or lay near its upper limit
TABLE1. Details ofpatients studied Low purine diet. t Reference range: 0.24-0.42 for males; 0.18-0.36 for females. No.
I 2 3 4 5 6 7 8 9 10 11 12 13
34 37 54 39 33 35 49 46 41 31 57 32 40
M M M
M M M F M M M
M M M
Mean urinary urate excretion. (mmo1/24 h)
115 114 93 110 110 83 77 83 85 98 84 113 72
185 I80 171 179 190 173 I66 183 188 178 171 182 I58
2.36 2.32 2.04 2.26 2.37 1.97 1.85 2.04 2.12 2.15 1.96 2.32 1.73
152 160 144 155 138 126 138
0.40 0.6 1 0.54 0.6 1 0.69 0.6 I 0.49 0.56 0.47 0.57 0.55 0.54 0.59
3.2 5.1 3.5 1.7 3.2 3.4 2.5 3.3 2.7 3.5 3.7 1.3 3.7
137 130 154 129
Lymphocyte purine synthesis in gout
(Table l), but two patients were included who were primary under-excretors of urate (urine urate less than 2 mmo1/24 h). All had normal renal function. Surface area and percentage of desirable weight (weight associated with the lowest mortality by the Metropolitan Life Insurance Company; Geigy Scientific Tables, 1971) were calculated from height and weight (Table 1). Blood samples for the preparation of lymphocytes were obtained on at least three occasions during a period of 3 weeks, but occasionally samples were collected over a longer period of time. On two occasions when treatment with allopurinol could not be delayed, only two specimens were studied. Lymphocytes from no more than two patients and two controls were studied in a batch, and each batch included lymphocytes from both patients and controls.
Preparation of lymphocytes For this, 20 ml of blood was collected in preservative-free heparin (200 units; Weddel Pharmaceuticals, London) and then diluted with 20 ml of NaCl solution (154 mmol/l: saline). The diluted blood was layered on 15 ml of FicollPaque (Pharmacia Fine Chemicals) in two 27 mm x 100 mm plastic tubes and centrifuged at 400 g for 45 min. Cells at the plasma/Ficoll interface were collected with a Pasteur pipette and washed in incubation buffer (pH 7.4) consisting of K,HPO, (25 mmol/l), NaCl (100 mmol/l), Hepes (20 mmol/l), glucose ( 5 . 5 mmol/l), glycine (4 mmol/l), NaHCO, (10 mmol/l) and bovine serum albumin (004%). This buffer had a measured osmolality of 295 mosmol/kg. The cells were suspended in 15 ml of incubation buffer, carefully layered over 10 ml of sucrose (584 mmol/l) in a plastic tube and centrifuged at 140 g for 25 min. The cell pellet was washed twice with incubation buffer and the cells were recovered by centrifugation at 140 g for 10 min. Cells were counted (Coulter FN) and resuspended in incubation buffer at a cell density of 5 x lo66 x 106/ml. The viability of lymphocytes as determined by Trypan Blue exclusion was 98.4 f 0.56% (mean & SEM, n = 15). With this procedure, the ratio of platelets to lymphocytes was 0.37 & 0.05 (mean 2 SEM, n = 15). Although defibrination techniques gave slightly less platelet contamination, they gave a much lower yield of lymphocytes.
;Assay ofpurine synthesis de nouo Purine synthesis de novo was measured by the incorporation of [ 14C]formate (0.5 mmol/l,
20-25 pCi/pmol) into a-N-formylglycinamide ribonucleotide (FGAR) in the presence of azaserine (0.3 mmol/l), which inhibits further metabolism of FGAR. The rate of FGAR synthesis was dependent on preincubation time in the presence of azaserine, periods of preincubation exceeding 30 min resulting in a rapid decline in FGAR synthesis. Lymphocytes were incubated at a density of 6 x lo6 cells/ml in the incubation buffer in an atmosphere of 5% CO, in air in a shaking water bath (100 oscillations/min) at 37OC. After a preincubation period of 30 min, and the addition of azaserine at 15 min, glutamine ( 2 mmol/l) and the [14Clformatewere added and the tubes regassed with 5% C O , and incubated for 60 min. At the end of the incubation, the cells were washed twice with 5 ml of ice-cold wash buffer, consisting of NaCl (130 mmol/l), Hepes (40 mmol/l, pH 7.4) and bovine serum albumin (0.4%). The cell pellet was resuspended in 1 ml of deionized water and an extract made by adding 100 ,ul of HCIO, (10 mol/l). The acid-soluble supernatant was neutralized by the addition of Tris formate (1 mol/l; pH 7.0) and KOH (10 mol/l). After removal of KCIO,, the neutral supernatant (1000 pl) was applied to a column ( 5 mm x 70 mm) of anion-exchange resin (AG-1 x 8, 200-400 mesh, Bio-Rad). Columns were washed with 20 ml of formic acid (0.5 molll) and the FGAR was eluted with 10 ml of formic acid (4 mol/l) [ 121. Fractions (2 ml) were collected and counted for radioactivity in Triton/toluene scintillant in a Packard Tri-Carb 2660 liquidscintillation counter. All assays were done in duplicate or triplicate. The rate of [ 14C]formate incorporation into FGAR was dependent on the incubation medium used, and was nine times greater at a phosphate concentration of 25 mmol/l than at physiological phosphate concentration. Incubation in the tissue-culture medium RPMI- I640 decreased the [ 14C]formate incorporation by 30%. Incubation in the presence of undialysed autologous serum or plasma inhibited formate incorporation by 70-90%. The validity of this FGAR isolation technique was tested by using authentic a-N-f~rmyl['~C]glycinamide ribonucleotide kindly supplied by Professor Peter Rowe (Royal Alexandra Hospital for Children, Sydney). Recovery of this authentic [I4C]FGAR added to the HCIO, (10 mol/l) extract was 95%. Samples of the formic acid (4 mol/l) eluate from lymphocyte assays were freeze-dried, and the residue was dissolved in deionized water and subjected to high-voltage paper electrophoresis (3000 V for 30 min) in sodium borate (50 mmol/l)/KCl (20 mmol/l)/ EDTA (1 mmol/l) buffer (pH 8.9). Greater than
R . B . Gordon el al.
90% of the applied radioactivity was recovered in a spot which migrated in an identical manner to the authentic [I4C]FGAR. Blank assays containing [l4C1formate were included in each batch of assays; these gave negligible radioactivity in the formic acid (4 mol/l) eluate. Purine synthesis was also assessed by the incorporation of [I4C]formate into total acidsoluble purine compounds. Incubations were carried out as described above, but in the absence of azaserine. The HCIO, cell extract was hydrolysed at 100°C for 60 min to convert all purine nucleotides and nucleosides into the corresponding bases, and these purine bases were isolated as the silver salts and the radioactivities determined 1131. By addition of 500 pl of AMP ( 5 mmol/l) to each assay tube before the hydrolysis step, the recovery of purine compounds during the isolation of the purine bases was estimated spectrophotometrically, and recoveries of 75-85% were obtained. The radioactivity in the isolated silver-purines was corrected for these recoveries.
similar measurements for purine synthesis de novo (see Fig. 1). The assay was linear with time up to 120 min (Fig. l a ) and was proportional to cell number between 1 x lo6 and 6 x lo6 (Fig. lb). The incorporation of [I4C1formate into FGAR was found to obey Michaelis-Menten
Reliability of measurement of purine synthesis de novo The ability of freshly isolated lymphocytes to carry out purine synthesis de nouo was assessed by both (a) the incorporation of [l4CIformate into FGAR in the presence of azaserine and (b) the incorporation of [ I4Clformate into total purine compounds isolated as the silver salts from the acid-soluble fraction. Both procedures gave
FIG.2. Purine synthesis in freshly isolated lymphocytes from control individuals ( n = 12) and patients with gout (n = 13). Bars represent means f 1 SD. 12c
60 Time (min)
4 6 x Cell no.
FIG.1. (a) Incorporation of [I4Clformate into FGAR (open symbols) and into total purines (filled-in symbols) as a function of time. Cell density was 6 x 106/ml. 0 and 0 ,Data from two control subjects. (b) Incorporation of [14Clformate into FGAR (open symbols) and into total purines (filled-in symbols) as a function of cell number at 60 min incubation. 0, 0 ,A , Data
obtained from three control subjects.
Lymphocyte purine synthesis in gout
kinetics, with a K, of 0.1 mmol/l. Assays were performed at a final formate concentration of 0.5 mmol/l. Blood from a control donor was subjected to parallel isolation by the technique outlined in the Materials and methods section, and both lymphocyte preparations were assayed for rate of purine synthesis. The results were 25.3 and 24.7 pmol h-I cells, indicating satisfactory agreement and a minimal effect of cell handling on the assay for purine synthesis.
medium containing purine bases decreased the subsequent measurement of purine synthesis, even though the incubation was carried out in medium free of purine bases. Heparin and colchicine had no effect on purine synthesis, and allopurinol (40 pmolll) resulted in 35% inhibition. Adrenaline in high concentration (90 pmol/l> caused 30% stimulation, which was not affected by the preceptor-blocker propranolol.
Otherfactors that might affectpurine synthesis in lymphocytes
Purine synthesis de novo in patients with gout and controls
Inclusion of hypoxanthine during the preincubation or assay periods at 2 pmol/l gave 40-50% inhibition of formate incorporation into FGAR. This was probably responsible for the inhibition produced by undialysed serum or plasma. Washing the cells with a synthetic
The mean value of purine synthesis in lymphocytes obtained from the 12 control subjects and the 13 patients with gout are shown in Fig. 2. The actual values on the different occasions of assay reveal considerable variation in purine synthetic rates in the one individual
TABLE2. Variationin purine synthesis when assayed on direrent occasions The assay nos. 1-4 represent assays performed on freshly isolated lymphocytes prepared on subsequent occasions. The actual occasions of lymphocyte preparation and assay were not separated by the same number of days for each control subject or patient. The values are for fL4C1FGAR synthesized, in pmol h-I cells. N.S.,Not significant. Assay no.
Controls 1 2 3 4 5 6 7 8 9 10 11 12 Mean (n = 12) Patients 1 2 3 4 5 6 7 8 9 10
II 12 13 Mean (n = 13)
20.6 24.5 28.6 21.3 15.8 36.9 20.0 14.4 17.6 19.1 19.1 18.8 21.4
26.3 26.5 21.8 16.4 25.0 24.3 24.5 18.4 15.0 13.8 21.0 18.3 20.9
27.3 25.2 17.6 23.3 27.6 28.1 17.6 22.0 26.9 13.0 19.5 13.1 21.8
24.8 35.9 36.6 44.0 26.0 21.3 37.3 13.3 31.3 21.2 34.2 19.9 19.0 28.1
24.7 21.9 26.6 28.6 22.7 20.8 34.5 17.5 31.7 27.1 28.3 20.2 6.4 23.9
25.7 27.9 20.7 29.4
16.2 24.1 23.1 21.6
15.2 22.1 26.4
12.8 17.4 18.9 11.2 13.6 20.7
Comparison of controls with patients (Mann-Whitney (I-test) U 41.5 48 61 N.S. P= 0.05 N.S.
24.7 25.4 20.8 20.3 22.6 28.9 20.7 18.3 22.3 15.3 19.9 16.7 21.3
3 .O 0.8 5.1 2.9 4.4 4.8 2.9 3.1 6.1 2.7 0.8 2.6
22.9 27.5 26.8 30.9 24.4 24.5 31.1 15.4 25.3 21.9 27.1 17.1 13.0 23.69
3.9 5.3 6.1 8. I 1.6 3.5 6.9 2. I 8.8 4.0 6.3 4.2 5.2
R . B. Gordon et al.
(Table 2). If the mean value of all measurements in the one individual is taken as the representative value for that individual, no significant difference is obtained between rates of purine synthesis in lymphocytes from patients and controls (MannWhitney U-test). There was no correlation between purine synthesis in lymphocytes and the urinary urate excretion for 24 h on a purine-free diet or the serum urate concentration. Discussion The present study confirms a previous report [ 101 that freshly isolated peripheral lymphocytes carry out purine synthesis de nouo. To lessen the extent of contamination of the lymphocyte fraction by platelets, an additional centrifugation step was included in the isolation procedure. This is important because platelets have been reported to carry out purine synthesis de nouo, although at a much lower rate than lymphocytes [141. Our preparations contained a mean of 2-7 times as many lymphocytes as platelets, in comparison with previous studies, in which there were 13 times as many platelets as lymphocytes [ 101. Lymphocytes isolated from the one individual on different occasions exhibited considerable variations in purine synthetic activity de nouo, even though incubations and assays were always performed under identical conditions. This applied in both control subjects and gouty patients. The reason for this variation on different occasions is not explained, although presumably it represents an intrinsic difference in the responses of the lymphocytes on the different occasions. Measurement of the rate of purine synthesis in lymphocytes is dependent on a variety of factors, the most significant being the concentration of phosphate. The considerable increase with increasing phosphate concentration probably reflects the availability of the rate-controlling metabolite, PRPP, the synthesis of which is activated by phosphate 115, 161. The inhibition by allopurinol and hypoxanthine is thought to be mediated by their conversion into their respective nucleotides by the purine salvage enzyme, hypoxanthinelguanine phosphoribosyltransferase. This reaction utilizes PRPP, which consequently decreases its availability for purine synthesis de nouo as measured by the incorporation of [ I4CIformate into purines. This competition between the synthetic pathway de novo and the salvage pathway for PRPP has been suggested as a basis for the regulation of purine synthesis in fibroblast and lymphoblast cultures [ 13, 171. In their lymphocyte isolation procedure,
Kamoun et af. [lo] included a wash step with Hanks medium 199. In our hands, a wash using this medium resulted in rates of purine synthesis in the subsequent assay which were low and quite variable. This probably resulted from the presence in medium 199 of all four purine bases, especially adenine (Gibco catalogue 1978179). Lymphocytes were not available from untreated patients with hypoxanthinelguanine phosphoribosyltransferase deficiency, and it was considered unethical to withdraw their allopurinol treatment to determine whether their lymphocytes would show increased 1l4Clformate incorporation into FGAR by the present technique. However, EB-virus-transformed lymphocytes (lymphoblasts) from such patients gave [ 14Clformate incorporations three times those of controls as measured by the present method of formate incorporation. No significant difference was observed between the rates of purine biosynthesis de nouo in lymphocytes from the control subjects and the patients with severe gout (Fig. 2 and Table 2). This finding differs from previous studies [lo], which reported that, although there was considerable overlap between normal and gouty groups, the mean rate of purine synthesis was higher in lymphocytes from the gouty patients than in controls. It is accordingly of interest to compare the significant differences obtained in these two groups on the first occasion of the present study when a significant difference was found (assay no. 1 of Table 2) with the fact that this difference became insignificant when the assay was repeated on several occasions in each subject. If the variability in the purine synthetic activity de nouo on different occasions had not been appreciated, a similar conclusion could have been drawn to that of previous reports [ 101. It should also be noted that no subgroup of the gouty patients could be defined who demonstrated a consistently increased rate of purine synthesis. The two individuals in whom the mean purine synthesis de nouo was above the highest value obtained in the normal group (no. 4 and no. 7) were not those who demonstrated the greatest degree of urate over-production (as judged by the mean urinary urate excretion on a low-purine diet: Table 1). Indeed one was included in the study because he was a primary under-excretor of urate. There was, moreover, no over-all correlation between purine synthesis de nouo and either the serum urate concentration or the 24 h urinary urate excretion, thereby suggesting that the measurement of purine synthesis in circulating lymphocytes is unlikely to provide a simple and useful index of urate over-production.
Lymphocyte purine synthesis in gout
We are grateful to the National Health and Medical Research Council of Australia for financial support. References 111 THANNHAUSER,S.J. (1956) The pathogenesis
of gout. Metabolism, 5,582-593. I21 WYNGAARDEN, J.B. (1957) Overproduction of uric acid as the cause of hyperuricemia in primary gout. Journal of Clinical Investigation, 36,1508-1575. I31 SEEGMILLER, J.E., GRAYZEL,A.I., HOWELL,R.R. & PLATO,C. (1962) The renal excretion of uric acid in gout. Journal of Clinical Investigation, 41, 1094-1098. [41 RIESELBACH, R.E., SORENSEN,L.B., SHELP,W.D. & STEELE, T.H. (1970) Diminished renal urate secretion per nephron as a basis for primary gout. Annals of Internal Medicine, 13,359-366. 151 WYNGAARDEN, J.B. (1960) In: The Metabolic Basis of Inherited Disease, p. 679. Ed. Stanbury, J.B., Wyngaarden, J.B. & Fredrickson, D.S. McGraw-Hill, New York. [61 SEEGMILLER, J.E., GRAYZEL,A.I., LASTER,L. & LIDDLE,L. (1961) Uric acid production in gout. Journal of Clinical Investigation, 40, 1304-13 14. [71 KELLEY,W.N., ROSENBLOOM, F.M., HENDERSON,J.F. & SEEGMILLER, J.E. (1967) A specific enzyme defect in gout associated with overproduction of uric acid. Proceedings of the National Academy ofsciences U S A . , 51,1735-1739. 181 SPERLING,0..PERSKY-BROSH, S., BOER,P. & DE VRIES. A. (1973) Human erythrocyte phosphoribosylpyrophosphate
synthetase mutationally altered in regulatory properties. Biochemical Medicine, 7,389-395. [91 BECKER,M.A., MEYER, L.J., WOOD, A.W. & SEEGMILLER, J.E. (1973) Purine overproduction in man associated with increased phosphoribosylpyrophosphate synthetase activity. Science, 179,1123-1 126. I101 KAMOUN,P., CHANARD, J., BRAMI,M. & FUNCK-BRENTANO, J.L. (1978) Purine synthesis de nouo by lymphocytes in gout. Clinical Science, 54,595-601. [ I l l WALLACE,S.L., ROBINSON,H., MASI, A.T., DECKER,J.L., MCCARTY, D.J. & Yu, T.-F. (1977) Preliminary criteria for the classification of the acute arthritis of primary gout. Arthritis and Rheumatism, 20,895-900. 1121 MARTIN,D.W. & OWENS,N.T. (1972) Repression and derepression of purine biosynthesis in mammalian hepatoma cells in culture. Journal of Biological Chemistry, 241, 54775485. [ 131 GORDON, R.B., THOMPSON, L., JOHNSON,L.A. & EMMERSON, B.T. (1979) Regulation of purine de novo synthesis in cultured human fibroblasts: the role of P-ribose-PP. Biochimica et Biophysica Acta, 562, 162-176. [I41 JERUSHALMY, Z., PATYA,M., BOER,P. & SPERLING, 0. (1980) De nouo synthesis of purine nucleotides in human blood platelets. Haemoslasis, 9,20-27. [IS] HERSHKO,A., RMm, A. & MAGER,J. (1969) Regulation of the synthesis of 5-phosphoribosyl- I-pyrophosphate in intact red blood cells and in cell-free preparations. Biochimica et Biophysica Acta, 184,64-76. [I61 Fox, I.H. & KELLEY,W.N. (1971) Human phosphoribosylpyrophosphate synthetase. Distribution, purification, and properties. Journal of Biological Chemistry, 246,5739-5748. 1171 HERSHFIELD,M.S. & SEEGMILLER, J.E. (1977) Regulation of de novo purine synthesis in human lymphoblasts. Journal of BioIogicar Chemistry. 252,6002-6010.