Effects of Acyclovir and Its Metabolites on Purine Nucleoside Phosphorylase*

Acyclovir (9-(2-hydroxyethoxymethyl)guanine), the clinically useful antiherpetic agent, is an “acyclic” analogue of 2’-deoxyguanosine. Purine nucleoside phos- phorylase partially purified from human erythrocytes did not catalyze detectable phosphorolysis of this drug or any of its metabolites (<0.07% of the rate with Guo). However, these compounds were competitive inhibitors of this enzyme with Ino as the variable substrate. Acyclovir per se was a relatively weak inhibitor. Its Ki value (91 p ~ ) was much greater than that for its 8- hydroxy metabolite (Ki = 4.7 p ~ ) but less than that for its carboxylic acid metabolite (9-carboxymethoxy- methylguanine) (K: = 960 f i ~ ) . The phosphorylated metabolites of acyclovir were more potent inhibitors than were their guanine nucleotide counterparts. At a phosphate concentration of 50 mM, the apparent Ki values for the mono- (120 p ~ ) , di- (0.51 p ~ ) , and tri (43 pM)-phosphate esters of acyclovir were 1 4 l/lm, and ‘/2e those for dGMP, dGDP, and dGTP, respectively. The concentration of phosphate not the Ki value of acyclovir those and Decreasing

The phosphorylated metabolites of acyclovir were more potent inhibitors than were their guanine nucleotide counterparts. At a phosphate concentration of 50 mM, the apparent Ki values for the mono-(120 p~) , di-(0.51 p~) , and tri (43 pM)-phosphate esters of acyclovir were 1 4 l/lm, and '/2e those for dGMP, dGDP, and dGTP, respectively. The concentration of phosphate did not markedly affect the Ki value of acyclovir but dramatically affected those of its phosphorylated metabolites and their nucleotide counterparts. Decreasing phosphate to a physiological concentration (1 mM) decreased the apparent Ki values for the mono-, di-, and triphosphate esters of acyclovir to 6.6, 0.0087, and 0.31 p~, respectively. Inhibition of the enzyme by acyclovir diphosphate was also influenced by pH. This metabolite of acyclovir is the most potent inhibitor of purine nucleoside phosphorylase reported to date. It has some features of a "multisubstrate" analogue inhibitor.
Acyclovir' is a clinically useful antiherpetic agent (1,2). Its structural formula and those of its metabolites have been described previously (3). Because of the structural resemblance of the 9-substituent to naturally occurring pentosyl moieties, acyclovir is referred to as an "acyclic" nucleoside analogue. A small percentage of the drug is oxidized to 8hydroxyacyclovir and to 9-carboxymethoxymethylguanine in human subjects and in experimental animals (4)(5)(6). In herpesinfected cells and to a much lesser extent in some uninfected cells, acyclovir is anabolized to its mono-, di-, and triphosphate derivatives (2, 7-10). Inhibitory effects of the triphosphate metabolite on viral DNA replication appear to be the major factor in the antiherpetic activity (2,11,12). Chain termination also appears to be involved in this inhibition (12).
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ''advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
In this study, acyclovir and its metabolites were tested as substrates and inhibitors of purine nucleoside phosphorylase (EC 2.4.2.1) from human erythrocytes. This endeavor was prompted by the structural similarities between acyclovir and natural substrates of this enzyme. Kinetic constants were determined for acyclic compounds and compared with those for their naturally occurring counterparts. Some of the metabolites of acyclovir were found to be much more potent inhibitors than were their natural counterparts.
EXPERIMENTAL PROCEDURES*

RESULTS
The partially purified purine nucleoside phosphorylase from human erythrocytes did not catalyze detectable phosphorolysis of acyclovir, or any of its metabolites, even at concentrations as high as 1 mM (Table I). As is characteristic of this enzyme (161, double reciprocal plots of initial velocity uersw concentrations of Ino, Guo, or dGuo were concave downward at substrate concentrations above 0.1 mM. For this reason, the kinetic constants listed in Table I were (Table I) were similar to reported values (16-18).
Inhibition constants for acyclovir and its metabolites and for some of their natural counterparts were determined at 50 mM phosphate (Table I). The Ki value for acyclovir (91 p~) was more similar to the KA values for dGuo (65 p~) and Guo (46 p~) than to the Ki value for Gua (4.6 p~) . 9-Carboxymethoxymethylguanosine (K! = 960 p M ) was a less potent inhibitor of this enzyme than was acyclovir, whereas 8-hydroxyacyclovir (Ki = 4.7 g M ) was much more potent. The phosphorylated metabolites of acyclovir were more potent inhibitors than were their nucleotide counterparts. Surprisingly, acyclovir diphosphate was the most potent inhibitor tested. The apparent K , value at 50 mM phosphate for this inhibitor (0.51 p~) was %ZOO that for dGDP. Acyclovir triphosphate (K! = 43 p~) was a more potent inhibitor than was acyclovir monophosphate (K! = 120 pM). Of the naturally occurring guanine nucleotides examined, 2'-deoxyguanosine-3'-phosphate was the most potent inhibitor (K! = 390 g~) and guanosine-2'-phosphate was the least potent inhibitor ( K ! = 3200 p~) .
The 2'-deoxyribonucleotides were more potent In addition to desalted purine nucleoside phosphorylase, these assay mixtures also contained 100 mM Tris.
* KL values were determined from a minimum of six initial velocity measurements at substrate concentrations ranging from 10 to 100 p~.
Vvalues are expressed as a percentage of the Vfor Guo phosphorolysis (2.6 units/mg protein). When no rate of phosphorolysis of a compound (0.1 mM) could be detected, the lower limit of detectability was calculated from the amount of enzyme used and the Ac value for the reaction. When 1.0 mM compound was used (parenthetical values), the reactions were monitored in cuvettes having a 1 mM path length in order to avoid spectral artifacts resulting from high absorbance values (13).
KI values were determined as described under "Experimental Procedures" with Ino as the variable substrate. A minimum of six concentrations of Ino, ranging from 10 to 100 pM, were used. The kinetic data for all inhibitors with Ino as the variable substrate were statistically consisWht with competitive inhibition (14). e The Kk for Ino determined at 1 mM phosphate was 41 f 3 @M. 'The actual value obtained, 30 f 2.1 phi, was corrected for the 0.5% acyclovir diphosphate present in the acyclovir triphosphate. This contamination was quantitated by high performance liquid chromatography (3). The equation used for the correction was derived from the velocity equation for pure competitive inhibition by two different exclusive inhibitors (15). The value calculated for the standard error assumes no error in the determination of the acyclovir diphosphate contamination. This is therefore a minimal standard error.
8The actual value obtained, 0.26 f 0.010 p~, was corrected for the 0.5% acyclovir diphosphate present in the acyclovir triphosphate as described in Footnote f. inhibitors than their corresponding ribonucleotides.
The above kinetic constants were determined at 50 mM phosphate. This is a customary concentration for studies with this enzyme since phosphate is used both as substrate and buffer. However, the range of physiological phosphate concentrations is considerably lower (i.e. around 1 mM for human erythrocytes) (19). A decrease in the phosphate concentration from 50 to 12.5 mM had no detectable effect on the inhibition by acyclovir. In contrast, the degree of inhibition by the phosphorylated metabolites of acyclovir was found to be a function of the phosphate concentration. The change from 50 to 12.5 mM phosphate increased the per cent inhibition by these compounds approximately Since phosphate had such a marked effect on the inhibition by these phosphate esters, their apparent Ki values and those for some of their ' The addition of 37.5 mM potassium sulfate or 100 mM potassium chloride to the assays containing 12.5 mM phosphate had no effect on the per cent inhibition by acyclovir diphosphate. This indicated that it was phosphate rather than potassium ion or ionic strength which was responsible for this effect. natural counterparts were determined at 1 mM phosphate, a concentration in the physiological range (Table I). For acyclovir and 8-hydroxyacyclovir, the Ki values determined at l mM phosphate and 50 mM phosphate were not significantly different. However, for the ionized compounds, the apparent Ki values determined at 1 mM phosphate were less than the values determined at 50 mM phosphate. For the phosphate esters, the magnitude of this decrease in apparent K t value appeared to be related to both, the glycosyl moiety (i.e. 2-hydroxyethoxymethyl > 2'-deoxyribosyl > ribosyl) and the number of phosphate moieties ( i e . tri->di->mono-). At 1 mM phosphate, the apparent Ki value (8.7 nM) for the most potent inhibitor, acyclovir diphosphate, was l/4,mm and ' / ropoo that for dGDP and GDP, respectively.

DISCUSSION
Neither acyclovir nor its metabolites were phosphorolyzed by purine nucleoside phosphorylase from human erythrocytes (Table I). This is consistent with the finding that there is no detectable cleavage of acyclovir in uioo (1,4, 5).
Acyclovir per se was a weak inhibitor of purine nucleoside phosphorylase. The marked increase in affinity for the enzyme resulting from the substitution of a hydroxy group at the 8-position of acyclovir is reminiscent of the increase in affinity resulting from the substitution of an amino group at the 8-position of Gua and Guo (20). On the other hand, substitution of a hydroxy group at the 8-position of Gua diminished the affinity for the enzyme (Table I).
The finding that the apparent Ki value for the phosphorylated inhibitors was a function of the concentration of phosphate (Fig. 4) has important implications when nucleotides or their analogues are examined as inhibitors of this enzyme. In a recent study, purine nucleotides were found to be weak inhibitors of the enzyme from human granulocytes (21). However, the weak inhibition reported for these nucleotides may reflect the high concentration of phosphate (100 mM) used in the assays. For example, the apparent Ki values for dGMP, dGDP, and dGTP at 50 mM phosphate were 5, 16, and 34 times those at 1 mM phosphate (Table I).
From the effects of phosphate on the inhibition by guanine nucleotides revealed in this study, it would appear that these compounds and their acyclic analogues possess binding determinants for both the guanine/guanosine and phosphate/ribose-1-phosphate-binding regions of the enzyme. This suggests that these phosphate esters can be considered as "multisubstrate" analogue inhibitors (22). The markedly increased affinity of the enzyme for the phosphorylated metabolites of acyclovir, as compared with their nucleotide counterparts, may be due, in part, to the flexibility of the acyclic side chain. This might allow the phosphate moiety to be positioned in closer proximity to the phosphate-binding site of the enzyme. Such flexibility is readily apparent from space-filling models.
The diphosphate ester of acyclovir was a more potent inhibitor than its mono-and triphosphate esters (Table I). An optimum distance between guanine and the terminal phosphate moiety in acyclovir diphosphate might account for its greater inhibitory properties. It is noteworthy that acyclovir diphosphate is the most potent inhibitor of purine nucleoside phosphorylase reported to date. In patients infused intravenously with the recommended dose of 5 mg/kg of acyclovir, the mean steady state peak plasma acyclovir level observed was 44 W M (23). The intraerythrocyte concentration was found to be nearly the same as the plasma concentration (4). This concentration is one-half the Ki value for purine nucleoside phosphorylase. It is therefore unlikely that this enzyme would be significantly inhibited at this concentration of acyclovir. However, in patients treated with the higher dose of 10 mg/kg acyclovir (intravenously), the steady state plasma level was 92 p~ (23), a concentration which might theoretically result in some inhibition.
The intracellular or plasma concentrations of the two oxidized metabolites of acyclovir have not been determined in humans. However, the concentration of these metabolites found in the urine of patients treated with acyclovir was much lower than the concentration of acyclovir (4). Therefore, it can be inferred that their plasma concentrations are much lower than that of acyclovir. Little inhibition of purine nucleoside phosphorylase by the carboxylic acid metabolite of acyclovir would be expected since its Ki value is greater than that of acyclovir. Even though the K, value for 8-hydroxyacyclovir was %8 that for acyclovir, it is unlikely that it would inhibit significantly since only trace amounts of this metabolite are found in human urine (4).
The levels of the phosphorylated metabolites of acyclovir in animal tissues have been determined in vivo (24). At a relatively high dose (50 mg/kg, subcutaneously), the peak levels of acyclovir diphosphate found in the three tissues examined (brain, 0.06 p M ; liver, 0.14 p M ; and kidney, 0.24 pM) were greater than its apparent Ki value at 1 mM phosphate (Table I). This suggests that some inhibition of the enzyme in vivo might occur. However, the dose of acyclovir used was 10-fold higher than the recommended intravenous dose for human use.
One of the prime motivating factors in the search for inhibitors of purine nucleoside phosphorylase has been the discovery that patients who are genetically deficient in this enzyme have impaired cellular, but not humoral, immunity (25). It is possible that an inhibitor of this enzyme that is effective in uiuo might be a useful selective immunosuppressive agent. In view of the inhibition of purine nucleoside phosphorylase by acyclovir and its metabolites reported here, it is relevant to note that acyclovir has been shown to have little or no effect on a variety of immune-related functions in vivo at doses 10 to 40 times that recommended for intravenous use (26).