CELLULAR IMMUNOLOGY 86, 5 lo-5 17 (1984)

The Mechanism of Inhibition and “Reversal” of Mitogen-Induced Lymphocyte Activation in a Model of Adenosine

Deaminase Deficiency’

DANIEL ALBERT: HARRYG.BLUESTEIN, LINDATHOMPSON, ANDJ.E. SEEGMILLER

Department of Medicine, University of California Medical Center, 225 Dickinson Street, San Diego, California 92103

Received August II, 1983; accepted February 21, 1984

The biochemical mechanism of lymphocyte dysfunction with adenosine deaminase deficiency has been investigated using cultured phytohemagglutinin stimulated normal peripheral blood lymphocytes and the adenosine deaminase (ADA) inhibitor 2’deoxycoformycin. The addition of deoxyadenosine to ADA-inhibited (but not to uninhibited) cells generated increased dATP pools (up to 50-fold greater than controls) and depressed the mitogen response. dATP Accumulation was accompanied by depletion of the other three deoxynucleoside triphospbate (dNTP) pools (dTTP, dCTP, and dGTP). Suppression of the mitogen response could be prevented (“reversed”) to 90% of control levels by the addition of deoxynucleoside precursors for the depleted dNTPs at the initiation of mitogen stimulation. “Reversal” restored the dTTP and possibly the dGTP pools. Thus the mechanism of toxicity in this model appears to be inhibition of ribonucleotide reductase by massive accumulation of dATP, resulting in starvation for the other three deoxy- ribonucleoside triphospbates. “Reversibility” of this toxicity by providing sources for the missing three deoxynucleoside triphosphates argues for ribonucleotide reductase inhibition rather than other mechanisms of deoxyadenosine toxicity in this model.

INTRODUCTION

The coexistence of severe combined immunodeficiency with deficiency of the en- zyme adenosine deaminase (adenosine aminohydrolase., EC 3.5.4.4) was discovered by Eloise Giblett in 1972 (1). Subsequent studies have established a causal relationship between the enzyme deficiency and the immunodeficiency state. On the basis of a number of studies, including measurements on the patients themselves and on cell culture systems, several different mechanisms for the lymphocyte dysfunction have been proposed. These include adenosine-mediated pyrimidine starvation (2); aden- osine-mediated elevation of cyclic AMP (3); adenosine-mediated Sadenosylhomo- cysteine accumulation with resulting inhibition of methylation reactions (4); and last, dATP accumulation with resulting inhibition of ribonucleotide reductase (5).

’ Portions of this work were presented at the 44th annual meeting of the American Rheumatism Association and published as an abstract in Arthritis Rheum. 23,647, 1980. This work was supported in part by USPHS Research Grant Al-10931 and Training Grant AM-14916.

* To whom reprint requests should be sent at the current address: Box 404, University of Chicago Medical Center, 950 East 59th St., Chicago, Ill. 60637.

510

OOOS-8749184 $3.00 Copyright Q 1984 by Academic Fwss, Inc. All rights of reproduction in any form reserved.

ADENOSINE DEAMINASE DEFICIENCY 511

We have investigated the biochemical mechanism of lymphocyte dysfunction in a model system of adenosine deaminase deficiency using cultured phytohemagglutinin- stimulated normal peripheral blood lymphocytes and the adenosine deaminase in- hibitor 2’deoxycoformycin. With this model we have explored the suppression of mitogen stimulation by adenosine and deoxyadenosine. We have also investigated the phenomenon of “reversal” of deoxyadenosine toxicity by other deoxyribonucle- osides. Finally, we measured deoxyribonucleoside triphosphate pools during suppres- sion of lymphocyte activation by deoxyadenosine and “reversal” by other deoxy- nucleosides. This model and these experiments were designed to test the hypothesis that deoxyadenosine toxicity in adenosine deaminase-inhibited lymphocytes is me- diated by inhibition of ribonucleotide reductase.

MATERIALS AND METHODS

PHA-induced lymphocyte proliferation. Normal human mononuclear cells obtained by Ficoll-Hypaque density gradient (6) centrifugation of heparinized venous blood from volunteers were cultured in 200 &well, round-bottom microtiter plates (Linbro Scientific, New Haven, Conn.) at a density of lo6 cells/ml in RPM1 1640 medium (Grand Island Biological Co., Grand Island, N.Y.), supplemented with 0.1% fetal calf serum (Irvine Scientific, Irvine, Calif.) and 4.0% albumin, as previously described (7). Phytohemagglutinin ( 11.1 units/mg (Wellcome Reagents Ltd., Beckenham, En- gland) at 1 &ml) and 2’-deoxycoformycin (Pentostatin, Parke, Davis & Co., Detroit, Mich.; 1 PM) as well as deoxyribonucleosides (Sigma Chemical Co., St. Louis, MO.) at the concentrations indicated were added at time of initiation. [3H]Leucine (New England Nuclear Corp., Boston, Mass.; 5 1.6 Ci/mmol, 1 &i/well) was added for a 4-hr pulse prior to termination of the cultures at 48 hr with a Mash harvester.

Nucleotide profiles. For nucleotide analysis, cells were cultured in lOO-ml batches in 490-cm* roller bottles (Coming Glass Works, Coming, N.Y.), extracted in 0.4 N perchloric acid, and neutralized with Alamine-Freon (8). Deoxynucleotide profiles were obtained by the DNA polymerase assay (9) using Escherichia coli DNA poly- merase (P-L Biochemicals, Inc., Milwaukee, Wise.), the deoxynucleotide polymers poly(dI-dC) for dCTP and dGTP assays or poly(dA-dT) for dATP and dTTP assays (Miles Laboratories, Elkhart, Ind.) and [3H]dGTP, 10.0 Ci/mmol; [3H]dCTP, 25.23 CQmmol; [3H]dATP, 13.18 Ci/mmol; and [3H]dTTP, 17.3 Ci/mmol (New England Nuclear Corp.). The reaction mixture for the polymerase assay included the following in a volume of 200 ~1: 50 m&f Tris-HCl buffer, pH 8.3; E. coli polymerase, 1 .O unit; poly(dI-dC) or poly(dA-dT), 0.05 A 260 unit; 5 mM MgCl, for dCTP and dGTP or 50 m&f MgC12 for dATP and dTTP; 1.5 PCi [3H]dNTP; O-50 pmol (O-25 a) of limiting substrate; and 1.50 mmol excess substrate.

Reactions were run for 25 min at 37°C in microtiter plate wells and were terminated by addition of ice-cold 10% trichloroacetic acid and cooling in an ice water bath. Reaction mixtures were then aspirated onto glass-fiber filter paper, washed with ice- cold 5% trichloroacetic acid-l% sodium pyrophosphate (about 20 ml/reaction mixture) followed by ethanol, then dried and counted in 10 ml Liquifluor (New England Nuclear Corp.) on a Beckman LS-230 liquid scintillation counter (Fullerton, Calif.). Four concentrations of standards and 2 to 5 dilutions of the unknown were run in order to assure linear incorporation of radioactivity into polymer.

512 ALBERT ET AL.

\ D-,- ---y

-0

I

I IO 50 CONCENTRATION ( j,iM)

l%. 1. The inhibition of mitogen-stimulated human lymphocytes by adenosine and deoxyadenosine. PHA-stimulated (1 &ml) normal human PBLs treated with 1 N 2’deoxycoformycin and various con- centrations of either adenosine (AR) or 2’deoxyadenosine (AdR). Response was measured by [3H]leucine for a 4-hr pulse prior to harvest at 48 hr and calculated as 96 PHA response = 3H cpm with AdR or AR/ 3H cpm without AdR or AR.

RESULTS ’

Adenosine versus deoxyadenosine inhibition. Both adenosine (AR)3 and deoxy- adenosine (AdR) inhibit the mitogen stimulation of normal human lymphocytes if adenosine deaminase is inhibited. However, deoxyadenosine is more potent than adenosine, requiring one order of magnitude less substrate for a 50% inhibition of [3H]leucine uptake (Fig. 1).

Deoxyadenosine triphosphate accumulation. dATP concentrations are about lOO- fold greater in the red blood cells of ADA-deficient patients than controls (10, 11) and similar data have been found in mononuclear cells of ADA-deficient patients (12). We also found elevated levels of dATP in ADA-inhibited, PHA-stimulated lymphocytes (Table 1). Increased pools of dATP are found only in ADA-inhibited lymphocytes and not in PHA-stimulated cells (with or without deoxyadenosine) that are not exposed to 2’deoxycoformycin.

If dATP accumulation was inhibitory for stimulated lymphocytes by feedback inhibition of ribonucleotide reductase, five consequences should result. First, dATP should accumulate in suppressed cells and accumulation should be inversely pro- portional to their mitogenic response. Second, inhibited cells should show diminution of the other three deoxyribonucleoside triphosphate pools. Third, supplying deoxy- ribonucleoside precursors for the missing deoxyribonucleoside triphosphates might prevent the toxicity. Fourth, this “reversal” might correlate with an increase in the pool size of the respective deoxyribonucleoside triphosphate. Fifth, the accumulation of dATP should persist during “reversal.”

Deoxyribonucleoside triphosphate pools during deoxyadenosine inhibition. With increasing concentrations of deoxyadenosine, dATP accumulates lo- to 50-fold over control cells coincident with a fall in PHA responsiveness (Fig. 2). In a representative

3 Abbreviations used: ADA, adenosine deaminase; PBL, peripheral blood lymphocyte(s); MR, deoxy- adenosine; CdR, deoxycytidine; GdR, deoxyguanosine; TdR, thymidine; AR, adenosine; UR, uridine; HYU, hydroxyurea; PNP, purine-nucleoside phosphorylase.

ADENOSINE DEAMINASE DEFICIENCY 513

TABLE 1

Deoxyribonucleoside Triphosphate Pool

Concentration (pmol/ 106 cells)

Control suvvressed Reversed

dATP dTTP dGTP dCTP

20.7 1064.0(3085 + 977) 368.0 (1143 k 453) 43.2 14.8 (43 + 15) 46.6 (93 + 13) 13.3 6.2 (75 + 38) 12.5 (95 + 28) 61.1 39.6 (74 f 23) 18.0 (65 + 37)

[‘H]Leucine (cpm) 2438 390 (16 f 4) 1777 (84 + 14)

Notes. Normal human PBLs were cultured with PHA (1 &ml) and 2deoxycoformycin (1 WV) for 48 hr. Control cultures had no nucleosides added. Suppressed cultures had 10 &f AdR added. Reversed cultures had 10 &4 TdR, CdR, and GdR in addition to 10 N AdR added. Cells were cultured for 48 hr, then were harvested and extracted with 0.4 N perchloric acid followed by neutralization with Alamine- Freon. Deoxynucleotide concentrations were measured by the DNA polymerase assay. [3H]Leucine uptake was performed as described under Materials and Methods. Data shown are from a representative experiment. Means of four experiments + SEM in % of control are shown in parentheses.

experiment (Table l), the 50-fold rise in dATP in suppressed versus control cells occurred coincident with the predicted fall in the other three deoxynucleoside tri- phosphate levels. While this fall is substantial for dlTP and dCTP, the resting level of dGTP is so low that suppressed levels are difficult to document.

‘Reversal” of toxicity. In cultures suppressed over 90% by 10 pil4 deoxyadenosine, the addition of deoxynucleoside precursors for the missing deoxynucleoside triphos- phates resulted in maintenance of PHA responsiveness (Fig. 3), which achieved virtuahy 100% of stimulated controls when 10 &f concentrations of all three deoxynucleoside triphosphate concentrations were added. “Reversal” with all three deoxynucleoside

DEOXYAOENOSME (PM)

FIG. 2. Relationship between the intracethJlar accum t&ion of dATP and the inhibition of PHA stimulation in ADA-inhibited lymphocytes cultured with PHA (I &nl) and AdR at concentrations indicated for 48 hr. [3H]Leucine was added to an aliquot of the cultures for a 4-hr pulse prior to harvesting and the response to PHA calculated as in Fig. 1.

514 ALBERT ET AL.

100

0 GdR CdR TdR UR GdR GdR YR Cd -

R

C:R T:R T:R T:R

G+R

FIG. 3. The effect of deoxynucleoside reversal of AdR toxicity in ADA-inhibited lymphocytes on the response to mitogen stimulation. Normal human PBL (106/ml) were cultured with 2’deoxycoformycin ( 1 a, PHA ( 1 &nl), AdR ( 10 CrM) and various deoxynucleosides and m-k-line at 10 &f concentrations. After 48 hr incubation, the % PHA response was measured as described in Fig. 1.

precursors (CdR, GdR, and TdR) resulted in dNTP pools of dTTP and dGTP that rose to control levels; however, dCTP levels remained at suppressed levels or fell further. During “reversal”, dATP levels felh however, they were stilI elevated within the range of accumulation that generated maximal inhibition of the PHA response (Table 1 and Fig. 2).

DISCUSSION

A major effort has been made to elucidate the mechanism of lymphocyte dysfunction in ADA deficiency. Early studies focused on the recognized role of adenosine deaminase in conversion of adenosine to inosine (Fig. 4)-a necessary step in puke salvage pathway. Additional evidence underscoring the importance of this pathway was the discovery of immunodeficiency associated with the next enzyme in the pathway- purine-nucleoside phosphorylase (PNP) (13).

Early studies ascribed adenosine toxicity in several cell culture systems to pyrimidine

GMP

t

Uric Acid

dAPP _ dAOP e dATP

FIG. 4. The role of adenosine deaminase (ADA) and purine nucleotide phosphorylase (PNP) in purine metabolism. ADA converts adenosine or deoxyadenosine to inosine or deoxyinosine. PNP catalyxes the conversion of guanosine, deoxyguanosine, inosine, and deoxyinosine to their respective purine bases.

ADENOSINE DEAMINASE DEFICIENCY 515

starvation (2). However, further investigations have revealed several features of the adenosine deaminase reaction which suggested that a deoxyadenosine triphosphate accumulation was responsible for the lymphocyte dysfunction. First, deoxyadenosine is the preferred substrate for adenosine deaminase with K,,, of 7 /.LM vs 24-52 fl for adenosine. Second, adenosine kinase has a high affinity for adenosine (Km 2-4 PM), while deoxycytidine kinase, the enzyme responsible for phosphorylating AdR, has a low affinity for deoxyadenosine (Km > 100 &f) (14). Thus, the primary metabolic route of adenosine appears to be phosphorylation rather than deamination and vice versa for deoxyadenosine (Fig. 4). Furthermore, as suggested by Carson et al., de- oxyadenosine could be trapped intracellularly by kinase-mediated phosphorylation to the deoxyribonucleoside triphosphate form. Indeed, elevated levels of dATP were found in the RBC of children with ADA deficiency ( 10).

Previous investigation on the mechanism of action of ribonucleotide reductase suggests that the enzyme plays a crucial role in DNA synthesis. Ribonucleotide re- ductase is required for de ~OVCJ synthesis of deoxynucleotides needed for DNA synthesis. The complex allosteric effector mechanism of ribonucleotide reductase by product deoxyribonucleoside triphosphates includes inhibition of all four substrate conversions by dATP. This provides a ready explanation for dATP suppression of lymphocyte activation. dATP inhibition of ribonucleotide reductase would result in inhibition of DNA synthesis because of starvation for the other three deoxynucleoside triphosphates.

Our data support such a model. In PHA-stimulated normal human lymphocytes with adenosine deaminase inhibited by 2’-deoxycoformycin, we have demonstrated (1) the greater suppression of the mitogen response to phytohemagglutinin by de- oxyadenosine than adenosine (as originally described by Simmonds (15)) (2) the accumulation of dATP in suppressed cells, (3) the diminution of dTTP and dCTP and perhaps dGTP in the suppressed cells, (4) prevention of suppression by the addition of deoxycytidine, deoxyguanosine, and thymidine, (5) increased pools of dTTP and perhaps dGTP in adenosine deaminase-inhibited cells to which thymidine, deoxycytidine, and deoxyguanosine were added (“reversal”), and (6) the presence of inhibitory levels of dATP in both suppressed and “reversed’ cells.

While consistent with the hypothesis that inhibition of ribonucleotide reductase is the mechanism of deoxyadenosine toxicity in this model, these data do not establish its certainty. On average, we measured only a 43% depression in dTTP pools. This is less than the decline noted in dCTP pools in ADA-inhibited mouse T-lymphoma (S49) cells (16). However, this may be enough of a decline in dTTP availability to stop cell replication since pools are dynamic and only sufficient for a few minutes of active DNA synthesis. In addition, the most profound diminution in dTTP may have occurred before 48 hr when we measured pools.

Further evidence in support of the ribonucleotide reductase model comes from two other sources. In our laboratory, we have examined the toxicity of hydroxyurea (HYU)-an inhibitor of ribonucleotide reductase (17) in an hypoxanthine-guanine phosphoribosyltransferase-deficient lymphoblastoid cell line. We have found the tox- icity of 50 MM HYU is reversible when all four deoxynucleosides (AdR, CdR, GdR, TdR) are added at 10 a concentrations (unpublished results). This model differs from AdR toxicity because cells are arrested in S phase rather than G 1 ( 18). It does, however, support the concept of “reversal by supplying precursors for diminished deoxynucleoside triphosphates during ribonucleotide reductase inhibition.

We have also examined a model of purine-nucleoside phosphorylase deficiency

516 ALBERT ET AL.

using PI-IA-stimulated normal human lymphocytes grown in the presence of guanosine and deoxyguanosine which results in suppression of mitogen stimulation and elevated levels of dGTP. Suppression can be prevented by deoxycytidine and thymidine (19) again implicating a central suppression role for ribonucleotide reductase inhibition. Last, we have demonstrated that DNA synthesis inhibition during mitogen stimulation will substantially diminish protein synthesis as measured by [3H]leucine at 48 hr after initiation of stimulation (20). Thus, we can link the inhibition of ribonucleotide reductase by dATP, resulting in DNA synthesis inhibition, to the diminished uptake of [3H]leucine.

An alternative explanation is suggested by the work of Uberti et al., (21). They noted that deoxyadenosine toxicity in ADA-inhibited mitogen-stimulated lymphocytes reduced protein synthesis during the first 24 hr after mitogen addition. This, coupled with the observation that deoxyadenosine causes Gl arrest of lymphoblasts (22) and that the activity of ribonucleotide reductase does not increase in mitogen stimulated lymphocytes until after 24 hr, suggests that deoxyadenosine toxicity may prevent ribonucleotide reductase synthesis and thus maintain low dNTP pools. Our data do not formally exclude this possibility. However, if this were true, we would expect normal pools of dATP in cells in which deoxyadenosine inhibition had been prevented by deoxynucleoside addition, rather than the elevated pools we observed in our model system.

While both adenosine and deoxyadenosine result in inhibition of methylation (4, 23), we find this mechanism of toxicity difficult to reconcile with reversibility in the first instance by uridine (24) and in the second by combinations of the three depleted deoxynucleosides in PI-IA-stimulated peripheral blood mononuclear cells.

In all likelihood, there are several mechanisms of toxicity even in mitogen-stimulated lymphocytes. For example, Ballow and Pantschenko (25) identified two different populations of mitogen responsive T cells, of which one is sensitive and one resistant to the inhibitory effects of 2’-deoxycoformycin. Other investigations (26,27) into the specific attributes of the immune response failure with adenosine deaminase inhibition are in progress. Resting lymphocytes are also susceptible to the toxic effects of adenosine deaminase inhibition. ATP depletion occurs in resting ADA-inhibited lymphocytes and erythrocytes although the mechanism for this is not clear (28, 29). These issues and the vast literature concerning adenosine deaminase deficiency has been the subject of several recent reviews (30-33).

The mechanism of toxicity in affected individuals is still a matter of dispute. Hutton et al., (34) were not able to document evidence of inhibition of either ribonucleotide reductase or methylation in their longitudinal study of a patient with ADA deficiency. However, they were unable to measure lymphocyte dNTP pools at the only time at which the individual had elevated erythrocyte dATP. We too, were unable to obtain lymphocytes from a child who presented with ADA deficiency and thus could not confirm our in vitro findings. After transfusion therapy, Hutton’s patient had normal to slightly elevated dATP pools in both RBC and lymphocytes; thus their inability to show depletion of the other dNTP pools is not unexpected.

It is of interest that the deoxycytidine, the most potent reversing nucleoside, did not result in increased dCTP pools. Perhaps the dCTP pool is labile through incor- poration of the deoxynucleotide into new DNA, or through interconversion of dCMP with dUMP which may be converted to dTTP and subsequently incorporated into new DNA as we have suggested in PNP “reversal.”

ADENOSINE DEAMINASE DEFICIENCY 517

“Reversal” of toxicity with deoxynucleosides suggests a potential therapeutic in- tervention with ADA-deficient children. Deoxycytidine has been shown to reverse the defect in PHA transformation of lymphocytes from two ADA-deficient children. This success had not been forthcoming in the children treated with deoxynucleoside- perhaps because of failure to achieve appropriate serum concentrations (28). These data support a renewed effort to achieve a therapeutic regimen of deoxynucleosides for patients with adenosine deaminasedeficient severe combined immunodeficiency- especially for those children unable to undergo bone marrow transplantation and for those resistant to chronic transfusion therapy (35).

REFERENCES

1. Giblett, E. R., Anderson, J. E., Cohen, F., Pollara, B., and Meuwissen, H. J., Lancet 2, 1067, 1972. 2. Ishii, K., and Green, H., J. Cell Sci. 13, 429, 1973. 3. Carson, D. A., Kaye, J., and Seegmiller, J. E., J. Clin. Invest. 57, 274, 1976.

4. Kredich, N. M., and Martin, D. W., Jr., Ceil 12, 931, 1977. 5. Carson, D. A., Kaye, J., and seegmiller, J. E., Proc. Nati. Acad. Sci. USA 74, 5677, 1977. 6. Boyum, A., Scan. J. Clin. Lab. Invest. 21, Suppl. 97, 77, 1968. 7. Willis, R. C., and Seegmiier, J. E., Monogr. Hum. Genet. 10,92, 1978.

8. Khym, J. X., Clin. Chem. (Winston-Salem N.C.) 21, 1245, 1975.

9. Lowe, J. K., and GsindIey, G. B., Mol. Pharrnacol. 12, 177, 1976. 10. Cohen, A. R., H&horn, S. D., Horowitz, A., Rubinstein, S. H., Polmar, R., Hong, R., and Martin,

D. W., Jr., Proc. Natl. Acad. Sci. USA 75, 472, 1978. 11. Coleman, M. S., Donofrio, J., Hutton, J. J., and Hahn, L., J. Eiol. Chem. 253, 1619, 1978. 12. Donofrio, J., Coleman, M. S., Hutton, J. J., Daoud, A., Lampkin, B., and Dyminski, J., J. Clin. Invest.

62, 884, 1978. 13. Giblett, E. R., Amann, A. J., Sandman, R., Wara, D. W., and Diamond, L. K., Lancet 1, 1010, 1975. 14. Hex&field, M. S., Fetter, J. E., Small, W. C., Bagnara, A. S., Williams, S. R., Ullman, B., Martin,

D. W., Wasson, D. B., and Carson, D. A., J. Biol. Chem. 257,6380, 1982. 15. Simmonds, H. A., Panayi, G. S., and Cortigall, V., Lancet 1,60, 1978.

16. Ullman, B., Gudas, L. J., Cohen, A., and Martin, D. W., Jr., Cell 14, 365, 1978. 17. Young, D. W., and Hodes, S., Science 146, 1172, 1964. 18. Albert, D. A., and Bluestein, H. G., Znt. J. Immunopharmacol.. in press. 19. Albert, D. A., Bluestein, H. G., Willis, R. C., Nette, K., and Seegmiller, J. E., Arthritis Rheum. 24,

175, 1981. 20. Albert, D. A., and Bluestein, H. G., Fed. Proc. 39, 574, 1980.

21. Uberti, J., Lightbody, J. J., and Johnson, R. M., J. Zmmunol. 123, 189, 1979. 22. Fox, R. M., Kefford, R. F., Tripp, E. H., and Taylor, I. W., Cancer Res. 41, 5 141, 1981.

23. Hershfield, M. S., J. Biol. Chem. 254, 22, 1979. 24. Green, J., and Chan, T., Science 182, 836, 1973. 25. Ballow, M., and Pantschenko, A. G., Cell. Immunol. t&29, 1981. 26. Thuillier, L., Garreau, F., and Cattier, P., Cell. Immunol. 63, 81, 1981. 27. Greever, M. R., Siaw, M. F. E., Coleman, M. S., Whisler, R. L., and Eblcerzak, S. P., J. Immunol.

129, 365, 1982. 28. Carson, D. A., Wasson, D. B., Lakow, E., and Kamatani, N., Proc. Natl. Acad. Sci. USA 79, 3348,

1982. 29. Simmonds, H. A., Levinsky, R. J., Pet-ret-t, D., and Webster, D. R., Biochem. Pharmacol. 31, 947,

1982. 30. Thompson, L. T., and Seegmiller, J. E., Adv. Enzymol. Relat. Areas Mol. Biol. 51, 167, 1980.

31. Fox, I. H., and Kelley, W. N., Annu. Rev. B&hem. 47, 655, 1978. 32. Hirschom, R., and Martin, D. W., Jr., Springer Semin. Immunopathol. 1,299, 1978. 33. Martin, D. W., Jr,, and Gelfand, E. W., Annu. Rev. Biochem. 50, 845, 1981. 34. Hutton, J. J., Wiginton, D. A., Coleman, M. S., Fuller, S. A., Limouze, S., and Lampkin, B. C., J.

Clin. Invest. 68, 413, 1981. 35. Dyminski, J. W., et al., Clin. Immunol. Immunopathol. 14, 307, 1979.