Inhibition of Purified Human and Herpes Simplex Virus-induced DNA Polymerases by 9-(2-Hydroxyethoxymethyl)guanine Triphosphate EFFECTS ON PRIMER-TEMPLATE FUNCTION*

The inhibition of highly purified herpes simplex virus (HSV)-induced and host cell DNA polymerases by the triphosphate form of 9-(2-hydroxyethoxymethyl)gua-nine (acyclovir; acycloguanosine) was examined. Acyclovir triphosphate (acyclo-GTP) competitively in- hibited the incorporation of dGMP into DNA, catalyzed by HSV DNA polymerase; apparent K,,, and Ki values of dGTP and acyclo-GTP were 0.15 PM and 0.003 p ~ , respectively. HeLa DNA polymerase a was also com- petitively inhibited; K,,, and Ki values of dGTP and acyclo-GTP were 1.2 and 0.18 p ~ , respectively. In contrast, HeLa DNA polymerase p was insensitive to the analogue. The “limited” DNA synthesis observed when dGTP was omitted from HSV or a DNA polymerase reactions was inhibited by acyclo-GTP in a con- centration-dependent manner. Prior incubation of activated DNA, acyclo-GTP, and DNA polymerase (a or HSV) resulted in a marked decrease in the utilization of the primer-template in subsequent DNA polymerase reactions.

The inhibition of highly purified herpes simplex virus (HSV)-induced and host cell DNA polymerases by the triphosphate form of 9-(2-hydroxyethoxymethyl)guanine (acyclovir; acycloguanosine) was examined. Acyclovir triphosphate (acyclo-GTP) competitively inhibited the incorporation of dGMP into DNA, catalyzed by HSV DNA polymerase; apparent K,,, and Ki values of dGTP and acyclo-GTP were 0.15 PM and 0.003 p~, respectively. HeLa DNA polymerase a was also competitively inhibited; K,,, and Ki values of dGTP and acyclo-GTP were 1. 2 and 0.18 p~, respectively. In contrast, HeLa DNA polymerase p was insensitive to the analogue. The "limited" DNA synthesis observed when dGTP was omitted from HSV or a DNA polymerase reactions was inhibited by acyclo-GTP in a concentration-dependent manner. Prior incubation of activated DNA, acyclo-GTP, and DNA polymerase (a or HSV) resulted in a marked decrease in the utilization of the primer-template in subsequent DNA polymerase reactions. This decreased ability of preincubated primer-templates to support DNA synthesis was dependent on acyclo-GTP, enzyme concentration, and the time of prior incubation. Acyclo-GMP-terminated DNA was found to inhibit HSV DNA polymerase-catalyzed DNA synthesis. Kinetic experiments with variable concentrations of activated DNA and fixed concentrations of acyclo-GMP-terminated DNA revealed a noncompetitive inhibition of HSV-1 DNA polymerase. The apparent K,,, of 3'-hydroxyl termini was 1.1 X lo" M, the Kii and Kk of acyclo-GMP termini in activated DNA were 8.8 X lo-* M and 2.1 X lo-' M, respectively. Finally, l4Clabeled acyclo-GMP residues incorporated into activated DNA by HSV-1 DNA polymerase could not be excised by the polymerase-associated 3',5'-exonuclease activity.
The nucleoside analogue, 9-(2-hydroxyethoxymethyl)guanine, is a potent and selective inhibitor of herpes simplex virus replication in uiuo and in vitro (1,2,4). The fist step in * This work was supported in part by Research Project Grant CH29C from the American Cancer Society. 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.
fj This work was submitted in partial fulfillment for the degree of Doctor of Philosophy to the faculty of the graduate school, State University of New York at Buffalo, Program of Pharmacology, Reswell Park Memorial Institute Division.
1 Scholar of the American Leukemia Society. To whom reprint requests should be addressed. the conversion of acyclovir' to the active metabolite, acyclo-GTP, is catalyzed exclusively by the HSV-induced deoxythymidine kinase (1,3). Acyclo-GTP competitively inhibited the incorporation of dGMP into DNA catalyzed by partially purifled HSV DNA polymerase and host a-polymerase (1,4). AC~C~O['~C)GMP was shown to be incorporated into DNA by HSV DNA polymerase and, at a slower rate, by a polymerase, in vitro (4).
Acyclo-GMP incorporated into primer termini would prohibit subsequent primer elongation owing to its lack of an extendable 3'-hydroxyl moiety. The suggestion has been made (4) that the inhibition of HSV DNA polymerization by acyclo-GTP might be analogous to the inhibition of Escherichia coli DNA polymerase I by 2',3'-dideoxythymidine 5'-triphosphate (5). That is, HSV DNA polymerase inhibition might be due to an effect concomitant with or subsequent to acyclo-GMP incorporation. To date, an examination of the primer-template effects of acyclo-GTP, using highly purified DNA polymerases, has not been reported.
Like E. COZZ DNA polymerase I and several other prokaryotic DNA polymerases, HSV DNA polymerase possesses a 3',5'-exonuclease activity (6)(7)(8) which may remove mismatched or fraudulent nucleotides incorporated into DNA. A previous report showed that acyclovir treatment of cells, which contained the HSV-specified deoxythymidine kinase, caused an accumulation of very short nascent DNA fragments (9). These low molecular weight fragments, which contained a~yclo-[~H]GMP residues at their 3'-termini, could not be chased into high molecular weight material even after the drug was removed (9). The potential ability of the HSV DNA polymerase 3',5'-exonuclease activity to remove 3"terminal acyclo-GMP residues is important mechanistically and pharmacologically.
In the present communication, a further investigation of the inhibition by acyclo-GTP of highly purified HSV and HeLa cell DNA polymerases is presented. The primer-template effects of the drug were examined and the question of the ability of the HSV DNA polymerase-associated 3',5'-exonuclease activity to remove 3'4erminal acyclo-GMP residues was addressed.

MATERIALS AND METHODS
All chemicals used were of reagent grade or better. Deoxynucleoside triphosphates and calf thymus DNA were purchased from Sigma.
HeLa S3 cells were the source of DNA polymerases a and p. All steps in the purification were performed at 4 "C and all buffers contained, in addition to the indicated components, 2 mM dithiothreitol, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and 20% (v/ v) glycerol. Log phase cells were extracted by sonic disruption in 300 mM KPO+ pH 8.0, and kept on ice for 3 h. Cellular debris was removed by centrifugation and the supernatant was passed through a DEAEcellulose column, equilibrated with the extraction buffer, to remove nucleic acids. The column flow through was dialyzed against 50 mM KPO,, pH 8.0, and applied to a second DEAE-cellulose column equilibrated with the dialysis buffer. The column was washed with 50 mM KPO4, pH 8.0, then eluted with 200 mM KP04, pH 8.0. 8-Polymerase activity was recovered in the column flow through and wash and a polymerase activity was eluted with 200 mM KPO4. Subsequent purification of a-polymerase through phosphocellulose, hydroxylapatite, and denatured DNA cellulose column chromatographic steps was performed essentially according to the method of Fisher and Korn (10). Further purification of P-polymerase on phosphocellulose, hydroxylapatite, and native DNA cellulose columns was done according to Stalker et al. (11). The specific activities of DNA polymerases a and were 8 X lo3 and 1.4 X lo4 units/mg, respectively.
No endo-or exonuclease activity could be detected under DNA polymerase reaction conditions (8).
Enzyme Assays-DNA polymerase assays were performed according to the published procedures (7). One unit of DNA polymerase activity is defmed as the amount of enzyme catalyzing the incorporation of 1 nmol of dTMP/h at 37 "C under the following reaction conditions.
Standard DNA polymerase / 3 reactions were identical with a polymerase except that KC1 was present at 0.1 M. HSV 3',5"exonuclease activity was determined as previously described (8).
DNA Substrates-Activated calf thymus DNA was prepared by the method of Baril et al. (12). The concentration of "usable" 3'hydroxyl termini was determined by measuring the extent of HSV-1 DNA polymerase-catalyzed DNA synthesis in the presence of [3H] dTTP alone or [3H]dTTP plus dGTP and dCTP (10, 14). Acyclo-GMP-terminated activated DNA was prepared in a 0.5-ml reaction mixture containing: 50 mM Tris-C1, pH 8.0; 4 mM MgC12; 0.5 mM dithiothreitol; 170 pg of activated calf thymus DNA: 2 p~ acyclo-GTP; and 0.8 unit of HSV-1 DNA polymerase. After 30 min of incubation at 37 "C another 0.8 unit of polymerase was added and incubation was continued for 30 min longer. The reaction was stopped and enzyme was denatured by heating at 65 "C for 10 min. After the reaction was allowed to cool slowly to 25 "C, the DNA was separated from free nucleotides by Sephadex G-25 column chromatography. The DNA, present in the excluded volume fractions, was precipitated in 3 volumes of absolute ethanol at -20 "C and finally resuspended in To determine the fraction of usable 3'-hydroxyl termini which were subsequently occupied by acyclo-GMP residues, an identical parallel reaction was performed in which 2 PM a~yclo-['~C]GTP(117 cpm/ pmol) was present. The product DNA was worked up exactly as described above then placed in 10 ml of aqueous scintillation fluid (New England Nuclear 947) and counted. Calculation of picomoles of a~yclo-['~C]GMP incorporated revealed that approximately 10% of the usable 3'-hydroxyl termini had been occupied by acyclo-GMP residues.
Excision of a~yclo-['~C]GMP Residues from Activated DNA-Removal of acyclo-GMP residues in DNA by the HSV-1 DNA polymerase-associated 3',5'-exonuclease activity was analyzed indirectly by   (4), and may be due to the greater purity of the enzymes used here or the different reaction conditions employed. HeLa DNA polymerase a was also competitively inhibited by acyclo-GTP as shown in Fig.   1C; apparent K , and Ki values of dGTP and acyclo-GTP were 1.2 p~ and 0.18 p~, respectively. In contrast, HeLa p-polymerase was not inhibited by acyclo-GTP concentrations up to 50 p~ at 1.0 p~ dGTP. Interestingly, the pattern of inhibition of these DNA polymerases by acyclo-GTP was the reverse of that observed with 2',3'-dideoxy-GTP. The latter compound specifically inhibited HeLa /?-polymerase but it was only weakly inhibitory toward HSV DNA polymerase and HeLa a-polymerase,' in agreement with the results of Zimmermann et al. (15). While acyclo-GTP and ddGTP are similar in their lack of 2'-and 3"hydroxyl groups, the absence of the 2' and 3' carbon atoms in acyclo-GTP may allow a degree of flexibility which is prohibited by the rigid furanoside ring structure of ddGTP.
DNA polymerases catalyze a limited rate of DNA synthesis when a single deoxynucleoside triphosphate is omitted from the reaction (13,14). The relative rates of DNA synthesis with a limited versus complete complement of dNTPs has previously been used to assess the processivity of DNA polymerases (10, 14). In order to determine possible effects of acyclo-GTP beyond the competitive inhibition of dGMP incorporation, we examined the influence of the drug on DNA synthesis limited by the omission of dGTP. As shown in Fig. 2  fects of acyclo-GMP incorporation on primer-template function, activated DNA was incubated with DNA polymerase (HSV or CY) and acyclo-GTP for varied periods of time. The DNA polymerase activity was then destroyed and an aliquot of fresh DNA polymerase plus an excess of deoxynucleoside triphosphates were added and DNA synthesis was monitored. Fig. 3 shows that with increasing times of prior incubation in the presence of acyclo-GTP, the subsequent ability of the DNA to support a normal rate of synthesis progressively declined. The rate and extent of this decrease was dependent on the amounts of enzyme and acyclo-GTP present in the prior incubation.' HSV DNA synthesis was apparently more sensitive to these prior incubations than a-polymerase. The decreased rate of DNA synthesis on preincubated primertemplates could be explained by a significant decrease in the concentration of 3'-termini available or, more likely, to a high affinity of DNA polymerase for a limited number of acyclo-GMP termini. To test the latter possibility, acyclo-GMP-terminated DNA was synthesized, as described under "Materials and Methods." Ten per cent of the initial usable 3'termini in this activated DNA were occupied by acyclo-GMP residues, as determined by the amount of a~yclo-['~C]GMP incorporated in a parallel reaction. Fixed concentrations of this partially acyclo-GMP-terminated, "inhibitor," DNA preparation were added to HSV-1 DNA polymerase reactions with varying amounts of activated DNA. The data are presented in the Lineweaver-Burk plot shown in Fig. 4A. The abscissa is the inverse concentration of activated DNA added; therefore the lines where inhibitor DNA was present are curved, due to the presence of additional 3'-hydroxyl termini in this preparation. When the DNA concentrations in Fig. 4A were corrected to include the total concentration of usable primer-template present and replotted, Fig. 4B was obtained.
Unexpectedly, a noncompetitive inhibition pattern was reproducibly observed. This is possibly a reflection of the tight binding of DNA polymerase to acyclo-GMP termini. Alternatively, acyclo-GMP-terminated DNA may interact with HSV-1 DNA polymerase at a second DNA binding site. It should be noted that acyclo-GTP was not present in the inhibitor DNA preparation. The apparent K , of activated DNA was 5.2 p g / d , K,, and I C c i values of acyclo-GMP-terminated DNA, calculated from slope and intercept replots, were 0.1 pg/ml and 4.2 pg/ml, respectively. In terms of primertermini, the K,,, of usable 3'-hydroxyl termini and the Kia of acyclo-GMP termini were 1. has been shown to possess a 3',5'-exonuclease activity (6)(7)(8), which may serve a proof-reading role in removing mismatched or fraudulent nucleotides incorporated into DNA. We have recently shown that this 3',5'-exonuclease activity can remove 3'-terminal araAMP residues in DNA in vitro (8). It was of interest, in elucidating the mechanism of action of acyclovir, to determine whether this activity could also remove acyclo-GMP residues from DNA.    DNA preparation (synthesized as described under "Materials and Methods").
In A, the abscissa is the inverse concentration of activated DNA added to each reaction and does not include the usable primer-template present in the acyclo-GMP-terminated DNA sample. In B, the total concentration of usable activated DNA present in each reaction was calculated; that is, the activated DNA added plus 90% of the acyclo-GMP-terminated DNA present. Lines were drawn by the method of least squares analysis.
In incubation with the analogue decreased primer-template functionality strongly support this view. In addition, the kinetically derived affinity of HSV-1 DNA polymerase for acyclo-GMP termini in DNA was approximately 50 times greater than for 3'-hydroxyl termini. Thus, the DNA product of acyclo-GMP incorporation was a potent inhibitor of further polymerization. The apparent inability of the HSV-1 DNA polymerase 3',5'exonuclease activity to remove acyclo-GMP residues incorporated into DNA may contribute to the tight binding of the polymerase to these termini as well as to prolonging the inhibition of DNA polymerization.