Proc. Nati Acad. Sci. USA Vol. 79, pp. 3848-3852, June 1982 Medical Sciences Possible metabolic basis for the different immunodeficient states associated with genetic deficiencies of adenosine deaminase and purine nucleoside phosphorylase (deoxyadenosine/deoxyguanosine/adenosylhomocysteine/lymphocytes) DENNIS A. CARSON, DONALD B. WASSON, ELLEN LAKOW, AND NAOYUKI KAMATANI Department of Clinical Research, Scripps Clinic and Research Foundation, 10666 North Torrey Pines Road, La Jolla, California 92037 Communicated by Ernest Beutler, March 3, 1982 ABSTRACT An inherited deficiency of adenosine deaminase mechanism. Thus, mouse T lymphoblasts with a mutant form (Ado deaminase; adenosine aminohydrolase, EC 3.5.4.4) causes ofthe enzyme insensitive to inhibition by dATP and dGTP are severe combined immunodeficiency disease in humans. A similar also resistant to deoxyadenosine and deoxyguanosine toxicity deficiency in purine nucleoside phosphorylase (Puo phosphoryl- (14, 15). ase; purine-nucleoside:orthophosphate ribosyltransferase, EC The above hypotheses, although offering a cogent explana- 2.4.2.1) engenders a selective cellular immune deficit. To eluci- tion for lymphocyte-specific toxicity in Ado deaminase and Puo date the possible metabolic basis for the contrasting immunologic phenotypes, we compared the toxicity toward mature resting hu- phosphorylase deficiencies, do not explain the immunologic man lymphocytes ofthe Ado deaminase substrates deoxyadenosine differences between the two syndromes. Ado deaminase-defi- and adenosine and the Puo phosphorylase substrate deoxyguano- cient children usually suffer from classical severe combined im- sine. When Ado deaminase was inhibited, micromolar concentra- munodeficiency disease, with absent cellular and humoral im- tions of deoxyadenosine progressively killed nondividing helper mune functions (16). On the contrary, most children with Puo and suppressor-cytotoxic T cells, but not B cells. The toxicity re- phosphorylase deficiency have a selective cellular immune def- quired phosphorylation, with subsequent dATP formation. The icit with normal immunoglobulin levels (2, 17). deoxyadenosine analogs 2-chlorodeoxyadenosine, 2-fluorodeoxy- Most in vitro studies concerning the mechanisms of deoxy- adenosine, and adenine arabinonucleoside also killed resting T adenosine, adenosine, and deoxyguanosine toxicity have uti- cells. Cell death was unrelated to inhibition of adenosylhomocys- lized rapidly dividinglymphocytes, including both cell lines and teinase (EC 3.3.1. 1) but was preceded by a gradual decline in ATP peripheral blood lymphocytes stimulated with mitogens. How- levels. As much as 1 mM deoxyguanosine did not impair resting ever, the majority of the long-lived T lymphocytes in human lymphocyte viability, despite the synthesis of dGTP. The combi- peripheral blood are actually in a "resting" state (18). In many nation of200 ,M adenosine plus 500 ,uM homocysteine thiolactone but not all systems, these cells can provide helper function killed dividing lymphocytes but had no discernible toxic effect to- without cell division in the induction ofa humoral immune re- ward resting T cells, which accumulated adenosylhomocysteine sponse (19-21). over a 4-hr period but thereafter excreted the nucleoside into the Kredich and Hershfield suggested that increased concentra- culture medium. The different clinical syndromes associated with tions of adenosine or deoxyadenosine in Ado deaminase-defi- genetic deficiencies ofAdo deaminase and Puophosphorylase may cient patients might inhibit biologically important transmeth- be explained by the ability of dATP to kill mature resting T lym- ylation reactions (22, 23). When its normal routes ofcatabolism phocytes by depleting ATP levels. are lacking, adenosine will combine readily with homocysteine in a reversible reaction catalyzed by adenosylhomocysteinase Inherited deficiencies of two enzymes of purine metabolism, (AdoHcyase; EC 3.3.1.1). The thermodynamically favored adenosine deaminase (Ado deaminase; adenosine aminohydro- product, adenosylhomocysteine (AdoHcy), is a potent natural lase, EC 3.5.4.4) and purine nucleoside phosphorylase (Puo inhibitor of transmethylation reactions (22). Additionally, de- phosphorylase; purine-nucleoside:orthophosphate ribosyl- not a substrate for transferase, EC 2.4.2.1), produce immunodeficiency disease in oxyadenosine itself, although AdoHcyase, human beings (1-3). Previous experiments have attempted to inactivates the enzyme and inhibits the hydrolysis of endoge- determine the reasons for lymphocyte-specific toxicity in both nously formed AdoHcy (23). Because some transmethylation Ado deaminase and Puo phosphorylase deficiency (4-12). In probably occurs throughout the cell cycle, inhibition ofAdoHcy affected patients, T lymphocytes may selectively phosphorylate hydrolysis could impair the survival of both resting and prolif- either deoxyadenosine or deoxyguanosine, which are excreted erating lymphocytes. by other tissues unable to phosphorylate the nucleosides (4). The goals of the present investigations were (i) to compare The T lymphocytes progressively sequester dATP and dGTP, in normal, resting human T and B lymphocytes the toxicity and because in this cell type deoxynucleoside-phosphorylating ac- metabolism of deoxyadenosine, adenosine, deoxyguanosine, tivity exceeds dATP- and dGTP-degrading activity (4, 9-12). and related analogs, (ii) to ascertain the relative roles of nu- dATP and dGTP are established potent inhibitors of mam- cleoside phosphorylation and AdoHcy accumulation in mediat- malian ribonucleoside-diphosphate reductase (EC 1.17.4.1) ing any observed toxic effects, and (iii) thereby to identify the (13). The toxicity ofdeoxyadenosine and deoxyguanosine toward metabolic basis for the contrasting phenotypes associated with dividing, immature T lymphocytes may depend upon this genetic deficiencies ofAdo deaminase and Puo phosphorylase. The publication costs ofthis article were defrayed in part by page charge Abbreviations: Ado deaminase, adenosine deaminase; Puo phospho- payment. This article must therefore be hereby marked "advertise- rylase, purine nucleoside phosphorylase; AdoHcy, adenosylhomocys- ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact. teine; AdoHcyase, adenosylhomocysteinase. 3848 Downloaded by guest on September 29, 2021 Medical Sciences: Carson et al. Proc. Natl. Acad. Sci. USA 79 (1982) 3849 MATERIALS AND METHODS RESULTS Lymphocyte Culture. Peripheral blood mononuclear cells Deoxyadenosine Toxicity. Resting human lymphocytes in- were isolated from heparinized blood of normal volunteers by cubated at 370C in RPMI 1640 medium supplemented with 20% Ficoll/Hypaque sedimentation as described previously (24). autologous plasma remained 90% viable for at least 4 days. After being washed three times in RPMI 1640 medium (Flow When the tight-binding Ado deaminase inhibitor deoxycofor- Laboratories, McLean, VA) the cells were resuspended at a mycin was added to the same cultures at 0.1 uM, viability was density of 106 per ml in RPMI 1640 medium supplemented with still 80% after 4 days. Under these conditions, the addition of 20% autologous plasma, penicillin at 100 units/ml, strepto- increasing concentrations of deoxyadenosine to the medium mycin at 100 ,ug/ml, and 2 mM glutamine. One-milliliter ali- produced a dose- and time-dependent loss ofcell viability, with- recovered cells. One micromo- quots were dispersed in 12 x 75 mm plastic tubes (Falcon 2058) out altering the total number of 40%, and 10 deoxy- or After various time periods, lar deoxyadenosine killed an average of kLM with without added nucleosides. adenosine killed 60% of the cells at 4 days (Fig. 1). were removed and the percentage of cells nonadherent cells Nondividing T lymphocytes bearing the Leu-2a and Leu-3a excluding 0.1% trypan blue was determined in a hemocytometer. antigens were sensitive to deoxyadenosine toxicity. Thus, a 4- Peripheral blood mononuclear cells were depleted of T lym- day incubation ofunfractionated peripheral blood mononuclear phocytes before culturing by rosetting with neuraminidase- cells with concentrations of deoxyadenosine that killed 60% of treated sheep erythrocytes, followed by centrifugation through the cells did not alter the percentage of viable lymphocytes Ficoll/Hypaque (25). Approximately 70% of the cells at the in- binding fluoresceinated monoclonal anti-Leu-3a (55-60%) or terphase stained with fluoresceinated F(ab')2 polyvalent goat anti-Leu-2a (22-32%). The T-cell-depleted, B-lymphocyte-en- anti-human Ig (26). riched cell population was markedly more resistant to deoxy- The effects of nucleosides on phytohemagglutinin-induced adenosine toxicity than unfractionated lymphocytes were (Fig. lymphocyte proliferation were determined as described (24). 1). T Cell Phenotyping. At the initiation of culture, and at in- The toxicity of deoxyadenosine toward resting human pe- tervals thereafter, the peripheral blood lymphocytes were ripheral blood lymphocytes was preceded by a slow but pro- stained with directly fluoresceinated monoclonal antibodies gressive accumulation ofdATP over at least a 2-day period (Fig. anti-Leu-2a (Becton Dickinson, Sunnyvale, CA), which detects 2). Accompanying the increase was a fall in intracellular ATP suppressor and cytotoxic T cells, and anti-Leu-3a, which detects levels, which anteceded cell death by 1-2 days. Human eryth- helperT cells (26, 27). The cells were counterstainedwith ethid- rocytes incubated with deoxyadenosine also accumulated