The Influence of Ribose 5-Phosphate Availability on Purine Synthesis of Cultured Human Lymphoblasts and Mitogen-stimulated Lymphocytes*

The intracelluar ribose 5-phosphate concentration was found to be an important determinant of rates of de novo purine synthesis. When ribose 5-phosphate production was reduced in cultured human lymphoblasts by glucose starvation, the intracellular phos-phoribosylpyrophosphateconcentration and rates of de nouo purine synthesis decreased. Inosinate-guany- 1ate:pyrophosphate phosphoribosyltransferase (HPR transferase)-deficient cells were relatively more re- sistant to glucose starvation. To minimize the effect of purine nucleotide feedback inhibition on the de novo pathway, cells were treated with inhibitors of IMP dehydrogenase and adenylosuccinate synthetase. In normal lymphoblasts, purine synthesis was stimulated only at glucose concentrations greater than 100 pM while in HPR transferase-deficient lymphoblasts, stimulation occurred even in the absence of glucose. The differences between the normal and HPR trans-ferase-deficient cells were lost when ribose reutiliza- tion from endogenous nucleotide breakdown was im-paired in the HPR transferase-deficient cells by incu- bation with 2‘-deoxyinosine. Endogenous ribose reutilization for purine synthesis is, therefore, important when either glucose availability is limited or synthesis is stimulated. In the absence of glucose, exogenous purine nucleotides restored the intracellular concentrations of ribose 5-phosphate, phosphoribosylpyro-

The Influence of Ribose 5-Phosphate Availability on Purine Synthesis of Cultured Human Lymphoblasts and Mitogen-stimulated Lymphocytes* (Received for publication, August 8, 1983) Renate B. Pilzz, Randall C. Willis, and Gerry R. Boss4 With the technical assistance of Soha D. Idriss The intracelluar ribose 5-phosphate concentration was found to be an important determinant of rates of de novo purine synthesis. When ribose 5-phosphate production was reduced in cultured human lymphoblasts by glucose starvation, the intracellular phosphoribosylpyrophosphateconcentration and rates of de nouo purine synthesis decreased. Inosinate-guany-1ate:pyrophosphate phosphoribosyltransferase (HPR transferase)-deficient cells were relatively more resistant to glucose starvation. To minimize the effect of purine nucleotide feedback inhibition on the de novo pathway, cells were treated with inhibitors of IMP dehydrogenase and adenylosuccinate synthetase. In normal lymphoblasts, purine synthesis was stimulated only at glucose concentrations greater than 100 pM while in HPR transferase-deficient lymphoblasts, stimulation occurred even in the absence of glucose. The differences between the normal and HPR transferase-deficient cells were lost when ribose reutilization from endogenous nucleotide breakdown was impaired in the HPR transferase-deficient cells by incubation with 2'-deoxyinosine. Endogenous ribose reutilization for purine synthesis is, therefore, important when either glucose availability is limited or synthesis is stimulated. In the absence of glucose, exogenous purine nucleotides restored the intracellular concentrations of ribose 5-phosphate, phosphoribosylpyrophosphate, and purine nucleotides to almost 100% and rates of purine synthesis to 50-75% of those at 10 mM glucose. When ribose 5-phosphate production was increased in peripheral blood lymphocytes by phytohemagglutinin activation, the intracellular phosphoribosylpyrophosphate concentration and rates of de nouo purine synthesis increased.
PP-Rib-P' plays a critical role in the regulation of de nouo Service Grants GM 17702, AM 13622, HD 10847, and 2S07 RR05665-* This work was supported in part by United States Public Health 15 administered through the Academic Senate, University of California, San Diego, the Kroc Foundation, and the Clayton Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ Supported by the Deutsche Forschungsgemeinschaft.
Rib-5-P, the precursor of PP-Rib-P, has also been proposed to be an important factor controlling the rate of de nouo purine synthesis, but this has never been established (1)(2)(3)13). That increased Rib-5-P availability leads to purine overproduction is suggested by three genetic disorders which cause hyperuricemia in man. The first is a mutant PP-Rib-P synthetase (ATP:a-~-ribose-5-phosphate pyrophosphotransferase, EC 2.7.6.1) with increased affinity for Rib-5-P; this effectively increases the kinetic availability of Rib-5-P (14). The other two are glucose 6-phosphatase deficiency (15) and increased activity of glutathione reductase (16); both mutations are thought to increase Rib-5-P generation by increasing the flux through the pentose phosphate pathway (3) (Fig. 1). The experimental decrease of the NADPH/NADP ratio induced by methylene blue or other electron acceptors significantly increases the intracellular concentrations of Rib-5-P and PP-Rib-P and stimulates de nouo purine synthesis (6)(7)(8).
We chose the following experimental approaches to explore the influence of Rib-5-P availability on the rate of de m u 0 purine synthesis in cultured human Iymphoblasts: 1) starvation for glucose to limit the Rib-5-P generation through the pentose phosphate pathway; 2) incubation with purine nucleosides to serve as alternative Rib-5-P sources; and 3) inhibition of purine nucleoside phosphorylase (purine nucleoside:orthophosphate ribosyltransferase, EC 2.4.2.1) to prevent endogenous ribose reutilization from intracellular nucleotide breakdown (Fig. 1). The above studies were performed with de nouo purine synthesis under normal feedback control by purine nucleotides and with feedback inhibition released by a combination of drugs that decrease the formation of adenylates and guanylates from IMP. The latter condition allowed us to investigate the Rib-5-P requirement for maximally stimulated de mu0 purine synthesis. We also used mitogenactivated peripheral blood lymphocytes as a model for stimulated purine synthesis and measured the specific activities of PP-Rib-P synthetase and amidophosphoribosyltransferase, and the intracellular concentrations of Rib-5-P, PP-Rib-P, purine nucleotides, and inorganic phosphate. We found that the intracellular Rib-5-P concentration can be rate-determining for de nouo purine synthesis.

Effect of Glucose Starvation on de Novo Purine Synthesis in
Cultured Lymphoblasts-When normal or HPR transferasedeficient lymphoblasts were transferred to glucose-deficient media, rates of de nouo purine synthesis decreased by 75% within the first 15 min. After 2 h of starvation, rates of de novo purine synthesis stabilized in the normal and HPR transferase-deficient lymphoblasts at 7 f 2 and 14 -t 3%, respectively, of the control values measured in the presence of 10 mM glucose. Addition of 10 mM glucose to either of the cell types at any time during a 6-h glucose starvation resulted in a rapid (within 15 min) and a complete restoration of de novo purine synthesis. Maximal rates of de novo purine synthesis occurred at 10 mM glucose in both cell types, but halfmaximal rates of de novo purine synthesis occurred at greater than 500 pM glucose in the normal cells and 100 PM glucose Effect of Mycophenolic Acid, Alanosine, and Hadacidin on de Novo Purine Synthesis during Glucose Starvation in Cultured Lymphoblasts-To eliminate the control of de novo purine synthesis by purine nucleotide feedback inhibition (3-5 ) , we incubated lymphoblasts with the combination of mycophenolic acid, alanosine, and hadacidin. We previously demonstrated (27) and further show below that these drugs decrease the intracellular concentration of purine nucleotides and cause a stimulation of de novo purine synthesis. In addition to release of feedback inhibition, the drugs also increase the reutilization of the ribose moiety of purine nucleotides (see below). In the normal lymphoblasts, de nouo purine synthesis was stimulated by the drugs only at glucose concentrations greater than 100 pM (Fig. 2, closed circles).
In the HPR transferase-deficient lymphoblasts, on the other hand, de rwvo purine synthesis was stimulated significantly ( p < 0.05, Student's t test) by the drugs at all concentrations of glucose and even in its absence (Fig. 3, closed circles). In both cell types, the degree of stimulation increased as the glucose concentration increased; stimulation was maximal at 10 mM glucose in the normal lymphoblasts and at 1 mM glucose in the HPR transferase-deficient lymphoblasts.
Effect of Alternative Ribose Sources on de Novo Purine Synthesis in Cultured Lymphoblasts-Since lymphoblasts lack ribokinase and are not readily permeable to ribose phosphates, we used purine nucleosides to function as alternative ribose phosphate source^.^ Inosine, guanosine, and their analogs are readily transported inside the cell and are then phosphohydrolyzed by purine nucleoside phosphorylase. This phosphorolysis yields the purine base and Rib-1-P which is converted to Rib-5-P by phosphoribomutase (Fig. 1). We used 8-aminoguanosine as the alternative ribose source for normal lymphoblasts. A disadvantage of 8-aminoguanosine is that this nucleoside is a relatively poor substrate of purine nucleoside phosphorylase (33). However, the advan-  tage of this compound is that the product, &aminoguanine, is not a substrate for HPR transferase and, therefore, cannot lead to the production of a nucleotide inhibitor of PP-Rib-P synthetase or amidophosphoribosyltransferase (34). In the absence of glucose, 50 phi 8-aminoguanosine increased de nouo purine synthesis of normal lymphoblasts almost &fold (compare bars a and b in the absence of glucose, inset, Fig. 2) and provided sufficient ribose to the lymphoblasts so that they further increased their rates of de novo purine synthesis in response to mycophenolic acid, alanosine, and hadacidin (compare bars c and d, inset, Fig. 2). In the presence of 100 p M glucose, 8-aminoguanosine increased de novo purine synthesis significantly only in the presence of mycophenolic acid, alanosine, and hadacidin (bar d at 100 FM glucose, inset, Fig.  2). At 10 mM glucose, the addition of 8-aminoguanosine had no significant effect.
HPR transferase-deficient lymphoblasts are unable to salvage hypoxanthine and, therefore, afford the opportunity to provide Rib-5-P to the cell from the preferred purine nucleoside phosphorylase substrate, inosine, without producing purine nucleotides (Fig. 4). At concentrations below 30 p~, inosine increased de nouo purine synthesis more than did glucose in the absence or presence of mycophenolic acid, alanosine, and hadacidin. The maximal rate of de novo purine synthesis was achieved with 30 p M to 1 mM inosine in the absence of the drugs and was 55% of the rate achieved with 1-10 mM glucose. However, in the presence of the drugs, the maximal rate of de novo purine synthesis was achieved with 1-3 mM inosine and was about 70% of the rate achieved with 1-10 mM glucose. These differences between glucose and inosine were still observed when the experimental protocol was modified and glucose and inosine were simultaneously added to parallel cultures.
Effect of Purine Nucleoside Phosphorylase Inhibition on de Novo Purine Synthesis in Cultured Lymphoblasts-Inhibition of purine nucleoside phosphorylase activity should allow one to determine the contribution of endogenous ribose reutilization to the Rib-5-P and PP-Rib-P available for de novo purine synthesis (Fig. 1). We used 2'-deoxyinosine to inhibit the phosphohydrolysis of inosine by purine nucleoside phosphorylase in HPR transferase-deficient lymphoblasts. While more potent purine nucleoside phosphorylase inhibitors are available, they all can either serve to some degree as a ribose source or function as inhibitors, of de novo purine synthesis (33). 2"Deoxyinosine is hydrolyzed by purine nucleoside phosphorylase, but the resulting 2'-deoxy-Rib-l-P cannot serve as a precursor of PP-Rib-P and as before, the HPR transferase-deficient cells cannot salvage the hypoxanthine.  In either the absence of glucose or in the presence of 100 PM glucose, [14C]formate incorporation was significantly reduced by 100 p~ 2'-deoxyinosine (compare bars a and b in the absence of glucose and at 100 pM glucose, inset, Fig. 3). This suppression was due neither to 2'-deoxy-Rib-1-P toxicity nor to nucleotide inhibition via residual HPR transferase activity, since it could be completely reversed by the addition of 100 p~ inosine. In the presence of 10 mM glucose, 2'deoxyinosine had no effect. When de novo purine synthesis of HPR transferase-deficient lymphoblasts was stimulated by mycophenolic acid, alanosine, and hadacidin, the addition of Z'-deoxyinosine significantly decreased the rate of [I4C]formate incorporation irrespective of the glucose concentration (compare bars c and d at all three glucose concentrations, inset, Fig. 3). The most pronounced effect of 2'-deoxyinosine was seen in the absence of glucose where it prevented any stimulation of de novo purine synthesis by the drugs. These results indicate that the contribution of endogenous ribose reutilization to the Rib-5-P available for de nova purine synthesis is significant when 1) the supply of Rib-5-P from the pentose phosphate pathway is reduced because of glucose starvation, or 2) de novo purine synthesis is stimulated by mycophenolic acid, alanosine, and hadacidin.
Effect of Glucose Starvation on the Concentration of Purine Nucleotides in Cultured Lymphoblasts-When normal lymphoblasts were incubated in the glucose-deficient media for 2% h, the intracellular GTP, ATP, and ADP concentrations decreased by 43, 40, and 30%, respectively, as compared to control values measured in the presence of 10 mM glucose; the GDP concentration remained unchanged ( Table 1). We were unable to detect major changes, greater than 2-fold, in the intracellular concentrations of AMP, GMP, and IMP during glucose starvation (data not shown). Results obtained with HPR transferase-deficient lymphoblasts were virtually identical. When inosine was added to glucose-starved HPR transferase-deficient cells the intracellular concentrations of purine nucleotides were similar to those at 10 mM glucose (data not shown).
Effect of Mycophenolic Acid, Alanosine, and Hadacidin on the Concentration of Purine Nucleotides in Cultured Lymphoblasts-When normal lymphoblasts were incubated for 2% h in the presence of 10 mM glucose with mycophenolic acid, alanosine, and hadacidin, the intracellular concentration of purine nucleotides decreased (Table I). Compared to control cultures without the drugs, the intracellular ATP and ADP concentrations decreased by about 21 and 18%, respectively, while the intracellular GTP and GDP concentrations decreased by about 74 and 44%, respectively. When the drugs were added to glucose-starved lymphoblasts, they did not decrease the concentration of adenylates any further, while GTP and GDP decreased by 28 and 17%, as compared to glucose-starved cultures without the drugs. When the drugs were added to cells in 0.01, 0.1, and 1 mM glucose, intermediate changes between those in the absence of glucose and 10 mM glucose were seen in the intracellular concentrations of purine nucleotides. Similar results were obtained in the HPR transferase-deficient lymphoblasts.

Effect of Glucose and Purine Nucleosides on Rib-5-P and PP-Rib-P Concentrations and [I4C]Adenine Incorporation in
Cultured Lymphoblasts-In the absence of exogenously added glucose, the intracellular Rib-5-P concentration was 70% of that at 500 p M glucose and 42% of that at 10 mM glucose (the difference between no added glucose and 10 mM glucose is statistically significant at the 0.05 value, Student's t test). Likewise, the [14C]adenine uptake, which as a measure of PP-Rib-P availability includes the cellular Rib-5-P content, significantly increased from the no added glucose condition more than 4-fold at 500 p~ glucose and more than 8-fold at 10 mM glucose.
8-Aminoguanosine at 50 p~ increased the PP-Rib-P concentration of normal, glucose-starved lymphoblasts to a value similar to that observed with 10 mM glucose. The Rib-5-P concentration in the presence of 50 p M 8-aminoguanosine was about 50% and the [l4C]adenine uptake about 75% of that measured in the presence of 10 mM glucose. As shown above, 8-aminoguanosine stimulated the rate of de nouo purine synthesis significantly, but only to 55% of the rate observed with 10 mM glucose.
In the presence of 10 mM glucose, the intracellular PP-Rib-P concentration in HPR transferase-deficient lymphoblasts was 2.1-fold higher than in normal lymphoblasts (Tables IT  and 111). In the absence of glucose, this difference was only 1.2-fold. The [l4C]adenine incorporation of HPR transferasedeficient cells was about twice that of normal cells in the presence or absence of glucose. The Rib-5-P concentration of HPR transferase-deficient lymphoblasts was only slightly higher than that of normal lymphoblasts at all glucose concentrations.
When 100 WM inosine was added to glucose-starved HPR transferase-deficient lymphoblasts, the Rib-5-P and PP-Rib-P concentrations were similar to those observed at 10 mM glucose, while the [14C]adenine incorporation was slightly lower. Inosine restored the rate of de nouo purine synthesis to only 55% of the rate observed with 10 mM glucose. The purine nucleoside phosphorylase inhibitor 2"deoxyinosine did not reduce the low Rib-5-P and PP-Rib-P concentrations of the glucose-starved HPR transferase-deficient lymphoblasts noticeably, but decreased the [14C]adenine incorporation. This latter value correlated with the decreased rate of de nouo purine synthesis caused by 2'-deoxyinosine.

Effect of Mycophenolic Acid, Alanosine, and Hadacidin on Rib-5-P and PP-Rib-P Concentrations and [14C)Adenine Incorporation i n Cultured Lymphoblasts-When
normal lymphoblasts were incubated in the absence of any added ribose source, mycophenolic acid, alanosine, and hadacidin only marginally increased the intracellular Rib-5-P and PP-Rib-P concentrations and the [l4C]adenine incorporation (Table 11). These results correlated with the lack of stimulation of de nouo purine synthesis by the drugs under these conditions. When HPR transferase-deficient lymphoblasts were incubated under the same conditions, the drugs caused a significant increase in the Rib-5-P and PP-Rib-P concentrations and [14C]adenine incorporation; 2'-deoxyinosine prevented this increase (Table 111). Again, these results correlated with the rates of de novo purine synthesis seen in glucose-starved HPR transferase-deficient lymphoblasts: mycophenolic acid, alanosine, and hadacidin caused a 2.4-fold stimulation which was prevented by 2'-deoxyinosine.
When normal and HPR transferase-deficient lymphoblasts were then incubated at glucose concentrations that allowed half-maximal rates of de nouo purine synthesis, mycophenolic acid, alanosine, and hadacidin increased the Rib-5-P and PP-Rib-P concentrations and ['*C]adenine incorporation to values that were markedly above the values measured with 10 mM glucose in the absence of the drugs. Finally, when lymphoblasts were incubated at 10 mM glucose, the drugs caused a more than 4-fold increase in the Rib-5-P and PP-Rib-P concentrations and [l4C]adenine incorporation in both cell types. However, the stimulation of de novo purine synthesis under these conditions was only 2-2.5-fold.

DISCUSSION
In this study, we demonstrated that the intracellular concentration of Rib-5-P, the precursor of PP-Rib-P, can be limiting for the rate of de nouo purine synthesis. When the intracellular Rib-5-P concentration was decreased in cultured lymphoblasts by glucose starvation, the intracellular PP-Rib-P concentration and rates of de novo purine synthesis decreased. Moreover, when the intracellular Rib-5-P concentration was increased in peripheral blood lymphocytes by mitogen activation, the intracellular PP-Rib-P concentration and rates of de novo purine synthesis increased. This latter increase of purine synthesis could not be attributed to an increase in the specific activities of either PP-Rib-P synthetase or amidophosphoribosyltransferase or to changes in the intracellular concentration of phosphate or purine nucleotides.
When HPR transferase-deficient lymphoblasts were compared to normal lymphoblasts, the deficient cells showed increased resistance to glucose starvation: they maintained higher rates of de novo synthesis in the absence of glucose and required less glucose for half-maximal rates. The difference between the two cell types may be explained by increased PP-Rib-P availability in the HPR transferase-deficient cells which is documented by an increased intracellular PP-Rib-P concentration and increased [l4C]adenine incorporation. This increased PP-Rib-P availability in the HPR transferase-deficient cells is because of ribose salvage from endogenous nucleotide breakdown in the absence of hypoxanthine salvage, i.e. unilateral ribose reutilization (42,45). The contribution of this ribose reutilization to Rib-5-P availability for de novo purine synthesis was studied in the HPR transferase-deficient cells by treating them with X'-deoxyinosine, an alternative substrate for purine nucleoside phosphorylase that does not yield Rib-5-P (33). We found that the reutilization of ribose from endogenous nucleotide breakdown becomes a significant factor for the maintenance of de novo purine synthesis at low glucose concentrations (5100 @hi), whereas in the presence of higher glucose concentrations, ribose reutilization does not seem to be necessary for normal rates of de novo purine synthesis.
We determined the degree to which purine nucleosides can replace glucose as a ribose source for de novo purine synthesis.
On a molar basis, low concentrations of 8-aminoguanosine in normal lymphoblasts and inosine in HPR transferase-deficient lymphoblasts were more efficient than glucose in increasing Rib-5-P and PP-Rib-P concentrations, as well as rates of de novo purine synthesis. However, higher concentrations of nucleosides were less effective and neither nucleoside could fully replace glucose; maximal rates of de muo purine synthesis achieved in the presence of 8-aminoguanosine or inosine were about 55% of those observed in the presence of 10 mM glucose. Comparably high intracellular Rib-5-P and PP-Rib-P concentrations were established in the presence of inosine and glucose. This eliminates the possibility that the rate of nucleoside transport or the conversion by purine nucleoside phosphorylase and phosphoribomutase might have been rate-limiting under these conditions. Inhibition of de novo purine synthesis by residual HPR transferase activity and consequent nucleotide formation is also excluded since the concentration of purine nucleotides in the presence of 1 mM inosine were the same as in the presence of 10 mM glucose4 and maximal, feedback-released de novo purine synthesis in the presence of mycophenolic acid, alanosine, and hadacidin with 1 mM inosine was still only 70% of that observed with 10 mM glucose in the presence of the drugs. We cannot exclude the possibility of some other unidentified side effect of high purine nucleoside concentrations on the de novo pathway. A more plausible explanation is that glucose metabolism provides a factor in addition to Rib-5-P that is necessary for maximal rates of de novo purine synthesis. The control of de novo purine synthesis is complex and feedback inhibition by purine nucleotides is a major determinant (3-5, 43). The intracellular concentrations of GTP, ATP, and ADP decreased during glucose starvation without major changes in the purine nucleoside monophosphate pools. A partial release of feedback inhibition on PP-Rib-P synthetase and amidophosphoribosyltransferase in glucose-starved cells is, therefore, possible (5,43). We felt it necessary to eliminate the control of de novo purine synthesis by feedback inhibition in order to fully evaluate the influence of Rib-5-P and PP-Rib-P availability on the rate of de novo purine synthesis. The combination of mycophenolic acid, alanosine, and hadacidin prevents the synthesis of guanylates and adenylates from IMP, thereby depleting the intracellular nucleotide pools and releasing feedback inhibition (22,24,27). We found that in normal lymphoblasts de novo purine synthesis could be stimulated by the drugs only at glucose concentrations greater than 100 PM, whereas in HPR transferasedeficient lymphoblasts, stimulation was observed even in the absence of glucose. The changes in the concentration of purine nucleotides during glucose starvation and incubation with the drugs were virtually identical in both cell types. Again, the difference between the two cell types can be explained by the increased PP-Rib-P availability in the HPR transferase-deficient cells because of unilateral ribose reutilization (42,45 In addition to release of feedback inhibition, mycophenolic acid, alanosine, and hadacidin increase ribose reutilization. Under normal conditions, cytoplasmic 5'-nucleotidase competes with IMP-dehydrogenase and adenylosuccinate synthetase for the substrate IMP (Fig. 1). Previous studies indicate that as much as 30-50% of newly synthesized IMP is subject to degradation and participates in the "inosinate cycle" (44,45). This cycle is composed of cytoplasmic 5'-nucleotidase, purine nucleoside phosphorylase, phosphoribomutase, PP-Rib-P synthetase, and HPR transferase (Fig. 1). In the presence of inhibitors of IMP-dehydrogenase and adenylosuccinate synthetase, the only possible fate of the accumulating IMP is degradation. Thus, the flux through the inosinate cycle is greatly enhanced. Hypoxanthine that is generated can be excreted into the media, whereas the ribose moiety is salvaged as ribose phosphate, retained within the cell, and reutilized. Indeed, in the presence of mycophenolic acid, alanosine, and hadacidin, greater than 90% of the total de novo synthesized purines were found in the culture media and the intracellular Rib-5-P and PP-Rib-P concentrations were markedly increased.
Because mycophenolic acid, alanosine, and hadacidin increase the flux through the inosinate cycle and increase Rib-5-P reutilization, de novo purine synthesis in the presence of the drugs may require less ribose from the pentose phosphate pathway than without the drugs. We should note that we observed ribose reutilization for purine synthesis, but we do not know how much ribose may also be used for synthesis of other compounds, for example, lactic acid. The importance of ribose reutilization for the stimulation of de novo purine synthesis by mycophenolic acid, alanosine, and hadacidin was shown in HPR transferase-deficient cells, since 2"deoxyinosine prevented the stimulation at glucose concentrations of 100 p~ or less. Again, 100 P M glucose seems to be a critical concentration for de novo purine synthesis. Below this concentration, de novo purine synthesis cannot be increased in response to released feedback inhibition unless additional Rib-5-P and PP-Rib-P are provided by endogenous ribose reutilization in the absence of hypoxanthine salvage (HPR transferase deficiency) or by an alternative exogenous ribose source. Even at 10 mM glucose, stimulation of de novo purine synthesis by the drugs was reduced in the presence of 2'deoxyinosine. The importance of ribose reutilization for the stimulation of de novo purine synthesis by mycophenolic acid, alanosine, and hadacidin also has been shown in studies with purine nucleoside phosphorylase-deficient lymphoblasts (46).