Biochemical Basis for Differential Deoxyadenosine Toxicity To

Home , Deoxyadenosine, Deoxyadenosine kinase

Proc. Natl. Acad. Sci. USA Vol. 76, No. 5, pp. 2434-2437, May 1979 Medical Sciences Biochemical basis for differential deoxyadenosine toxicity to T and B lymphoblasts: Role for 5'-nucleotidase (deoxyadenosine kinase/deoxyadenylate kinase/immunodeficiency) ROBERT L. WORTMANN, BEVERLY S. MITCHELL, N. LAWRENCE EDWARDS, AND IRVING H. Fox Human Purine Research Center, Departments of Internal Medicine and Biological Chemistry, Clinical Research Center, University of Michigan Medical Center, Ann Arbor, Michigan 48109 Communicated by James B. Wyngaarden, March 7, 1979

ABSTRACT Deoxyadenosine metabolism was investigated tained from Calbiochem. Erythro-9-[3-(2-hydroxynonyl)]- in cultured human cells to elucidate the biochemical basis for adenine (EHNA) was a gift from G. B. Elion of Burroughs the sensitivity of T lymphoblasts and the resistance of B lym- Wellcome (Research Triangle Park, NC). Horse serum was phoblasts to deoxyadenosine toxicity. T lymphoblasts have a 20- to 45-fold greater capacity to synthesize deoxyadenosine nu- obtained from Flow Laboratories (Rockville, MD), and Eagle's cleotides than B lymphoblasts at deoxyadenosine concentrations minimal essential medium was purchased from GIBCO. From of 50-300 ,uM. During the synthesis of dATP, T lymphoblasts Amersham/Searle, [U-14C]deoxyadenosine (505 mCi/mmol), accumulate large quantities of dADP, whereas B lymphoblasts [8-14C]hypoxanthine (52.5 mCi/mmol), and [U-14C]deoxy- do not accumu ate dADP. Enzymes affecting deoxyadenosine adenosine monophosphate (574 mCi/mmol) were purchased; nucleotide synthesis were assayed in these cells. No substantial and, from ICN, [8-14C]adenosine monophosphate (34.4 mCi/ differences were evident in activities of deoxyadenosine kinase (ATP: deoxyadenosine 5'-phosphotransferase, EC 2.7.1.76) or mmol) was purchased (1 Ci = 3.7 X 1010 becquerels). All other deoxyadenylate kinase [ATP.(d)AMP phosphotransferase, EC reagents were of the highest quality commercially available. 2.7.4.111. The activity of 5'-nucleotidase (5'-ribonucleotide Cell Lines. The MOLT-4 (T cell lymphoblast) cell line was phosphohydrolase, EC 3.1.3.5) was increased 44-fold for AMP obtained from HEM Research (Rockville, MD). The cells were and 7-fold for dAMP in B lymphoblasts. A model for the regu- negative for surface immunoglobulins, contained 23% T-ro- lation of deoxyadenosine nucleotide synthesis by 5'-nucleotidase setting cells, and were originally derived from a patient with activity is proposed on the basis of the observations. acute lymphoblastic leukemia. The MGL-8 (B-cell lympho- Deoxyadenosine and adenosine concentrations are increased blasts) cell line was derived from a normal individual and was in individuals with adenosine deaminase deficiency and severe a gift from J. Epstein, Johns Hopkins University. These cells combined immunodeficiency disease (1, 2). The accumulation were positive for surface immunoglobulins (8-20%) and in- of deoxyadenosine leads to the increased concentrations of corporated [3H]leucine into protein that migrated as heavy and dATP and dADP in erythrocytes, peripheral blood lympho- light chains on a sodium dodecyl sulfate gel after immu- cytes, and bone marrow cells of these patients (2-4). The ele- noprecipitation. vated levels of deoxyadenosine nucleotides are believed to Deoxyadenosine Metabolism. T and B lymphoblasts were provide the biochemical basis for the immune dysfunction removed from flasks in which they were grown in continuous observed in the enzyme deficiency state. culture. Cells were washed twice in normal saline and 100 mM Although the precise mechanism for the immune dysfunction Tris-HCl, pH 7.4. Cells were suspended in Eagle's minimal is unclear, several studies in vitro have attempted to determine essential medium containing 10% dialyzed horse serum, 1.2 mM the underlying molecular pathology. The addition of deoxy- potassium phosphate at pH 7.4, 25 mM Tris-HCl at pH 7.4, and adenosine reduces the response of peripheral blood lymphocytes 0 or 5.0 ,uM EHNA. Cell counts were performed on the cell to mitogen stimulation when adenosine deaminase is inhibited suspensions. (1, 5). The combination of deoxyadenosine and adenosine Aliquots (50 ,l) of cell suspension (2.0-6.0 X 106 cells per ml) deaminase inhibition is also cytotoxic to T lymphoblasts but not were incubated at 37°C for 20-35 min, removed and placed B lymphoblasts (6, 7). Deoxyadenosine-mediated cytotoxicity on ice, and immediately mixed with substrate solution (50,ul) in T lymphoblasts is accompanied by increased concentrations that contained Eagle's minimal essential medium. The final of dATP. concentrations were as follows: 1.2 mM potassium phosphate Deoxyadenosine metabolism was investigated in cultured at pH 7.4, 20 ,uM adenine, 25 mM Tris-HCl at pH 7.4, 0 or 5.0 human cells to elucidate the biochemical basis for the sensitivity ,uM EHNA, and 20-400 jiM radiolabeled deoxyadenosine di- of T lymphoblasts and the resistance of B lymphoblasts to luted with nonisotopic deoxyadenosine. The mixture was in- deoxyadenosine toxicity. These studies have revealed a major cubated at 37°C. The reaction was stopped by heat inactivation difference in the capacity of T lymphoblasts to accumulate at 85°C for 2 min and cooled to 40C. After centrifugation at deoxyadenosine nucleotides compared to B lymphoblasts, 1500 X g for 10 min, 20 pA of the supernatant solution was ap- possibly because of a difference in activity of 5'-nucleotidase plied to two sheets of Whatman 3 MM chromatography paper previously spotted with 20 Al of 1 mg/ml nonradioactive dATP, (5'-ribonucleotide phosphohydrolase, EC 3.1.3.5). dADP, dAMP, dIMP, adenine, deoxyinosine, deoxyadenosine, MATERIALS AND METHODS and hypoxanthine. Deoxynucleotides were separated by high-voltage electrophoresis in 50 mM sodium citrate at pH 3.9 Reagents. Deoxyadenosine, deoxyinosine, ATP, dATP, and deoxynucleosides and bases, in 50 mM sodium borate at pH dADP, dAMP, dIMP, Tris base, dithiothreitol, and EDTA were 8.9. Compounds were visualized with ultraviolet light and cut purchased from Sigma. Adenine and hypoxanthine were ob- out, and their radioactivities were measured in a liquid scintillation spectrometer system. The publication costs of this article were defrayed in part by page reaction mixture to block charge payment. This article must therefore be hereby marked "ad- Adenine was added to the hypo- vertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact. Abbreviation: EHNA, erythro-9-13-(2-hydroxynonyl)ladenine. 2434 Downloaded by guest on September 29, 2021 Medical Sciences: Wortmann et al. Proc. Natl. Acad. Sci. USA 76 (1979) 2435 xanthine phosphoribosyltransferase by depleting the intracel- labeled dADP and dATP formed from radiolabeled dAMP. lular phosphoribosyl pyrophosphate supply through its WUtili- Diluted cell extracts (25 ,gl) were mixed with a substrate solution zation in the adenine phosphoribosyltransferase reaction. To (75 ,l) providing final concentrations of 1.0 mM ATP, 5.0mM ensure that hypoxanthine formed from the degradation of la- magnesium chloride, 100 mM potassium chloride, 50 mM beled deoxyadenosine was not being salvaged by hypoxanthine TrisIHCI at pH 7.4, and 0.1 mM [U-14C]dAMP. The mixture phosphoribosyltransferase and contaminating the deoxynu- was incubated at 370C and the reaction was stopped by the cleotide pool, radiolabeled hypoxanthine was substituted for addition of 0.2 M EDTA and heating at 850C for 2 min. Product deoxyadenosine in the above system. Studies using 20 ,uM hy- formation was determined as described above with the ex- poxanthine demonstrated that less than 5% of the deoxynu- ception that high-voltage electrophoresis was performed in 50 cleotides could have originated from ribonucleotides. mM sodium citrate at pH 3.9. The reaction was linear with Viability of cells was checked at the end of each experiment protein concentration and time up to 45 min. Initial velocity in tubes not heated at 85°C by the nigrosin dye exclusion studies were performed with 12.5-100 MiM [U-14C]dAMP by method (8). Viability ranged from 94% to 98% in all experi- using similar methods. ments. Intact cell 5'-nucleotidase activity was quantitated by mea- Enzyme Assays. Membrane-depleted extracts of T and B suring the conversion of radioisotopic AMP or radioisotopik lymphoblasts were prepared for enzyme assays. Washed cells dAMP to radiolabeled nucleoside products as described by were frozen and thawed three times and centrifuged at 28,000 Edwards et al. (9). Lymphoblasts from continuous cell lines X g for 30 min at 4°C. The supernatant solutions were stored were harvested and suspended in Hanks' solution to 3.0-6.0 X at -70'C until use. Prior to assay the extracts were dialyzed in 106 cells per ml for assay. Aliquots (25 ,l) of cell suspension were normal saline with 1 mM Tris-HCI at pH 7.4 and 0.1 mM incubated in a total volume of 100 ,l with 40 mM Tris-HCI at EDTA for 6-14 hr. pH 7.5, 4.0 mM magnesium chloride, and 0.2 mM [8-14C]AMP Deoxyadenosine kinase activity (ATP:deoxyadenosine 5'- or [ U-14C]dAMP. The assay was linear with cell number and phosphotransferase, EC 2.7.1.76) was determined by the time up to 45 min. Initial velocity studies were performed with amount of radiolabeled deoxyadenosine nucleotides formed 5-50 ,M by using a similar method. from radiolabeled deoxyadenosine. Diluted cell extracts (50,ul) Concentrations. Proteins were measured by the method of were mixed with a substrate solution (50 ,l) to final concen- Lowry et al. (10) with bovine serum albumin as a standard. In trations of 5 mM ATP, 2 mM dithiothreitol, 40MuM EHNA, 200 the use of diluted radioisotopes, the concentration was deter- mM Tris-HCI at pH 7.4,20 mM magnesium chloride, and 0.5 mined by calculation from an absorbance measured at the ab- mM [U-14Cldeoxyadenosine. After 30-min incubation, the sorption maximum. reaction was stopped by heating at 85°C for 2 min. A portion (20 Ml) of the incubation mixture was spotted on Whatman 3 RESULTS MM chromatography paper, previously spotted with nonradioactive deoxyadenosine, deoxyinosine, dAMP, dADP, and Deoxyadenosine metabolism dATP. After 30 min of electrophoresis in 50 mM sodium borate Deoxyadenosine metabolism was investigated in an in vitro at pH 8.9 at 4000 V and 250 mA, deoxynucleotide spots were model of adenosine deaminase deficiency by using cultured localized by ultraviolet light and cut out, and radioactivities human T and B lymphoblasts. Adenosine deaminase activity were measured in a Packard liquid scintillation spectrometer was blocked by the addition of 5 MM EHNA. Deoxyadenosine system. The reaction was linear with protein concentration and nucleotides accumulated in both cell types and were synthesized with time up to 60 min. Initial velocity studies were performed by the phosphorylation'of deoxyadenosine. with concentrations of 0.1-2.0 mM [U-14C]deoxyadenosine by The most rapid synthesis of deoxyadenosine nucleotides was using similar methods. observed in the T lymphoblasts. Using 58 MuM deoxyadenosine, Deoxyadenylate kinase activity [ATP:(d)AMP phospho- T Iymphoblasts synthesized 396 pmol of deoxyadenosine nu- transferase, EC 2.7.4.11 was assessed by the amount of radio-

u,600 - A B 120.-

500 100 E E L, 400 80 , o a -5300 60 C C C 200 40 co Cc C

X 20 x o 100 0 0 1 0 20 30 1 0 20 30 Time, min 0 FIG. 1. [U- 4C]Deoxyadenosine metabolism during adenosine 0 10 20 30 0 10 20 30 deaminase inhibition by EHNA (5 ,uM) in continuous T lymphoblast Time, min (0) and B lymphoblast (0) cell lines. (A) Quantities of total de- FIG. 2. Synthesis of deoxyadenosine nucleotides from [U-14C]- oxyadenosine nucleotides formed by the phosphorylation of deoxy- deoxyadenosine with adenosine deaminase inhibition by EHNA (5 adenosine; (B) concomitant substrate utilization. The major pathway ,uM) in T lymphoblasts (A) and B lymphoblasts (B). Note the dif- of deoxyadenosine metabolism is deamination to deoxyinosine despite ferent vertical axis for each panel. 0, dATP; 0, dADP; A, dAMP; A, inhibition of adenosine deaminase by EHNA. dIMP. Downloaded by guest on September 29, 2021 2436 Medical Sciences: Wortmann et al. Proc. Natl. Acad. Sci. USA 76 (1979) Table 1. Properties of cultured human lymphoblast enzymes Specific activity, nmol/hr per mg protein or per 106 cells* Apparent Michaelis constant, gM Enzyme activity T lymphoblasts B lymphoblasts T lymphoblasts B lymphoblasts Deoxyadenosine kinase 142 101 293 307 Deoxyadenylate kinase 702 665 333 454 5'-Nucleotidase dAMP substrate 1.3 9.15 160 160 AMP substrate 1.1 48.5 * Specific activities of the two kinases are per mg of protein; 5'-nucleotidase is per 106 cells.

cleotides per 106 cells in 5 min, which can be compared to 9 (i) Deoxyadenosine Kinase Activity. Deoxyadenosine kinase pmol per 106 cells for B lymphoblasts (Fig. 1). Other studies activity was 142 and 101 nmol/hr per mg of protein in the T using 50-300 MuM deoxyadenosine confirmed findings that T and B lymphoblasts, respectively. The apparent values of the lymphoblasts have a 20- to 45-fold greater capacity to synthesize Michaelis constants for deoxyadenosine were identical (Table deoxyadenosine nucleotides than B lymphoblasts do. 1). Thus, although there is a slight increase in deoxyadenosine The individual deoxynucleotides synthesized f. m deoxy- kinase activity in the T lymphoblasts compared to B lym- adenosine were also measured; they demonstrate another dif- phoblasts, it is not great enough to account for the increased ference between T and B lymphoblasts (Fig. 2). While low phosphorylation observed in the T cell line. levels of dAMP and dIMP accumulate in each cell line, there (ii) Deoxyadenylate Kinase Activity. Deoxyadenylate ki- is a striking difference in the rate of dADP formation. In the nase activity was quite similar in the two cell types, with values first 10 min of the incubations, with 58 and 300 ,uM deoxy- of 702 nmol/hr per mg of protein for T lymphoblasts and 668 adenosine, the concentration of dADP exceeded the dATP nmol/hr per mg of protein for B lymphoblasts. The values for concentration in T lymphoblasts. At this point the dADP levels the apparent Michaelis constants for dAMP were also similar reached a plateau, while dATP concentrations continued to for this enzyme (Table 1). increase linearly. In B lymphoblasts the dADP concentrations (iii) 5'-Nucleotidase Activity. Although the values for the were always substantially lower than the dATP concentra- apparent Michaelis constant for 5'-nucleotidase activity are tions. identical for the two cell lines with dAMP as the substrate, the activity of this enzyme is distinctly different in T and B lym- Enzymes of deoxyadenosine metabolism phoblasts (Table 1). Compared to T lymphoblasts, B lym- The observations of greater accumulation of deoxyadenosine phoblasts have an average of 7 times more 5'-nucleotidase ac- nucleotides in the T lymphoblasts compared to B lymphoblasts tivity when dAMP is the substrate and 44 times more activity could be explained by either an increased synthesis in the T cells when AMP is the substrate. This difference suggests that the or an increased degradation by dephosphorylation in the B cells low rate of deoxyadenosine nucleotide synthesis in B cells may (Fig. 3). Assessment of individual enzymes was undertaken to be related to the increased capacity to dephosphorylate explain the increased synthesis of dADP observed with T dAMP. lymphoblasts as compared to B lymphoblasts. The possibilities investigated were: (i) Increased deoxyadenosine kinase activity DISCUSSION in T cells compared to B cells, (ii) increased deoxyadenylate The differential susceptibility of T and B lymphoblasts to kinase activity in the T cell line, and (iii) increased 5'-nucleo- deoxyadenosine toxicity implies an important biochemical tidase activity in the B cell line. difference between the two cell types. T lymphoblasts accumulate dATP and are killed when cultured in deoxyadenosine dATP during adenosine deaminase inhibition (6). B lymphoblasts, however, are resistant to deoxyadenosine and do not accumulate * 41 dATP under the same conditions. The current studies demonstrate that T lymphoblasts can accumulate deoxyadenosine at dADP a rate 20- to 45-fold faster than B lymphoblasts (Fig. 1). The capacity to accumulate large amounts of dADP in T lym- 4 2 phoblasts appears to be another difference from the B lym- 3

-- phoblasts (Fig. 2). dIMP dAMP In an attempt to understand the basis for the more rapid rate and the 5 5* *4 of deoxyadenosine nucleotide accumulation higher concentration of dADP observed in T lymphoblasts, enzymes 6 affecting the synthesis of these compounds were assayed (Table Deoxyinosine -f Deoxyadenosine 1). The data show that only 5'-nucleotidase activity differs A EHNA between the two cell types. No substantial variation was evident in activities of deoxyadenosine kinase or deoxyadenylate kinase. Previous observations (11) of a low Vmax and high apparent Michaelis constant for deoxyadenosine monophosphate Hypoxanthine deaminase make an alteration of this enzyme activity un- Fic. 3. Abbreviated diagram of deoxyadenosine metabolism. likely. Enzymes displayed are: 1, deoxyadenosine diphosphate kinase; 2, The interaction of 5'-nucleotidase activity and deoxyade- deoxyadenylate kinase; 3, deoxyadenosine monophosphate deami- nucleotide may be based upon the rapid de- nase; 4, deoxyadenosine kinase; 5, 5'-nucleotidase; 6, adenosine nosine synthesis deaminase (this enzyme step is inhibited by EHNA); 7, purine nu- phosphorylation and removal of dAMP when substantial 5'- cleoside phosphorylase. nucleotidase activity is present. Fig. 4 illustrates this proposed Downloaded by guest on September 29, 2021 Medical Sciences: Wortmann et al. Proc. Natl. Acad. Sci. USA 76 (1979) 2437

T cell / B cell

I Deoxyinosine * | Deoxyadenosine Deoxyadenosine I * Deoxyinosine

IF 4F dAMP dAMP "P 1t dADP dADP 41. 1t dATP dATP --city

FIG. 4. Model for differential toxicity of deoxyadenosine to cultured lymphoblasts during adenosine deaminase inhibition. In T lymphoblasts the small amount of 5'-nucleotidase leaves dAMP available for further conversion to dADP and dATP. In B lymphoblasts the-high activity of 5'-nucleotidase reconverts dAMP to deoxyadenosine, preventing its conversion to dADP and dATP. dATP is thought to be responsible for the cytotoxic effects.

relationship. With only a small quantity of 5'-nucleotidase ac- The authors thank Jumana Judeh for her excellent typing of the tivity in the T lymphoblasts, dAMP is rapidly phosphorylated manuscript. This work is supported by U.S. Public Health Service to dADP Grants AM 19674 and SM 01 RR 42 and a grant from the American and dATP. The accumulation of these deoxynucleo- Heart Foundation and the Michigan Heart Association. B.S.M. is a tides leads to cytotoxicity. The B lymphoblasts, on the other recipient of a National Institutes of Health Clinical Investigator Award. hand, can only form small quantities of dADP and dATP be- N.L.E. is the recipient of a Clinical Associate Physician Award from cause of the highly active and competing dephosphorylation the General Clinical Research Center Branch of the National Institutes of dAMP catalyzed by 5'-nucleotidase. An important role for of Health. dephosphorylating activity has been previously observed in 1. Simmonds, H. A., Panayi, G. S. & Corrigali, V. (1978) Lancet i, 6-thiopurine-resistant sublines of sarcoma 180 cells, which have 60-63. a 100-fold increase in alkaline phosphatase compared to 6- 2. Donofrio, J., Colem~an, M. S., Hutton, J. J., Daoud, A., Lampkin, thiopurine-sensitive cell lines (12-14). The increased alkaline B. & Dyminski, J. (1978) J. Clin. Invest. 62, 884-887. phosphatase prevents the accumulation of the toxic phos- 3. Coleman, M. S., Donofrio, J., Hutton, J. J., Hahn, L., Daoud, A., Lampkin, B. & Dyminski, J. (1978) J. Biol. Chem. 253, 1619- phorylated derivatives of 6-thiopurine. 1626. If the human lymphoblast models are relevant to adenosine 4. Cohen, A., Hirschhorn, R., Horowitz, S. D., Rubinstein, A., Pol- deaminase deficiency and immune dysfunction in man, they mar, S. H., Hong, R. & Martin, D. W., Jr. (1978) Proc. Natl. Acad. Sci. USA 75, 472-476. suggest a major abnormality of T cell function in that disease. 5. Carson, D. A., Kaye, J. & Seegmiller, J. E. (1977) Proc. Natl. Acad. The immune disorder of adenosine deaminase deficiency does Sci. USA 74,5677-5681. encompass a spectrum from complete absence of both B and 6. Mitchell, B. S., Mejias, E., Daddona, P. E. & Kelley, W. N. (1978) T cell function to a pure T cell abnormality. In fact, the latter Proc. Natl. Acad. Sci. USA 75,5011-5014. 7. Ullman, B., Gudas, L. J., Cohen, A. & Martin, D. W., Jr. (1978) situation occurs in 10-15% of adenosine deaminase-deficient Cell 14, 365-375. patients (15). If deoxyadenosine nucleotide accumulation alone 8. Hudson, L. & Hay, F. (1976) Practical Immunology (Blackwell, is responsible for the immune dysfunction, these clinical ob- Oxford), pp. 30-32. 9. Edwards, N. L., Magilavy, D. B., Cassidy, J. T. & Fox, I. H. (1978) servations suggest that the primary disorder in adenosine Science 201, 628-630. deaminase deficiency involves T cell function and that the B 10. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. cell abnormalities are a secondary phenomenon. This infor- (1951) J. Biol. Chem. 193, 265-275. mation together with observations of increased concentrations 11. Yun, S. L. & Suelter, C. H. (1978) J. Biol. Chem. 253,404-408. of dATP and dADP in cells from adenosine deaminase-deficient 12. Wolpert, M. K., Damle, S. P., Brown, J. E., Sznycer, E., Agrawal, patients demonstrates similarities between the disease in man K. C. & Sartorelli, A. C. (1971) Can. Res. 31, 1620-1626. and the lymphoblast culture model. 13. Rosman, M., Lee, M. H., Creasey, W. A. & Sartorelli, A. C. (1974) Thus, the cell culture model may reveal a biochemical basis Can. Res. 34, 1952-1956. 14. Lee, M. H., Huang, Y. M. & Sartorelli, A. C. (1978) Can. Res. 38, for the immune dysfunction in adenosine deaminase deficiency. 2413-2418. The relevance of the model is dependent upon demonstrating 15. Hirschhorn, R. (1977) in Progress in Clinical Immunology, ed. in normal human tissues the 5'-nucleotidase differences ob- Schwartz, R. S. (Grune and Stratton, New York), Vol. 3, pp. served in T and B lymphoblasts. 67-83. Downloaded by guest on September 29, 2021