THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1991 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 266, No. 1, Issue of January 5, pp. 238-244,1991 Printed in U. S. A.

Herpes Simplex Virus-specified DNA Polymerase Is the Target for the Antiviral Action of 9-(2-Phosphonylmethoxyethyl)adenine*

(Received for publication, June 25, 1990)

Scott A. Foster$, Jaroslav Cerny, and Yung-chi Chengt From the Department of Pharmacology, Yale University School of Medicine, New Haven, Connecticut 06510

9-(2-Phosphonylmethoxyethyl)adenine (PMEA) is a new antiviral compound with activity against herpes simplex virus (HSV) and retroviruses including human immunodeficiency virus. Although it has been sug- gested that the anti-HSV action of PMEA is through inhibition of the viral DNA polymerase via the diphos- phorylated metabolite of PMEA (PMEApp), no conclu- sive evidence for this has been presented. We report that in cross-resistance studies, a PMEA-resistant HSV variant (PMEA’- 1) was resistant to phosphonoformic acid, a compound which directly inhibits the HSV DNA polymerase. In addition, phosphonoformic acid-resist- ant HSV variants with defined drug resistance muta- tions within the HSV DNA polymerase gene were re- sistant to PMEA. Furthermore, the HSV DNA polym- erase purified from PMEA” 1 was resistant to PMEApp in comparison with the enzyme from the parental vi- rus. Moreover, PMEA inhibited HSV DNA synthesis in cell culture. These results provide strong evidence that HSV DNA polymerase is the major target for the anti- viral action of PMEA. Further studies showed that HSV DNA polymerase incorporated PMEApp into DNA in vitro, while the HSV polymerase-associated 3’-5’ exonuclease was able to remove the incorporated PMEA. Thus, the inhibition of HSV DNA polymerase by PMEApp appears to involve chain termination after its incorporation into DNA.

Acquired immunodeficiency syndrome (AIDS)’ has focused attention on the need for new and efficacious compounds that not only inhibit HIV, the etiological agent of AIDS, but also that are effective against the many opportunistic infections associated with AIDS. One such opportunistic infection is herpes simplex virus. Although many individuals are infected

*This work was supported by National Cancer Institute Grant CA-44358. 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.

$ This work represents partial fulfillment of the requirements for the degree of Doctor of Philosophy from the University of North Carolina at Chapel Hill.

§To whom correspondence and reprint requests should be ad- dressed Dept. of Pharmacology, Yale University School of Medicine, 333 Cedar St., New Haven, CT 06510.

The abbreviations used are: AIDS, acquired immunodeficiency syndrome; acyclovir, 9-(2-hydroxyethoxymethyl)guanine; ACGTP, acyclovir triphosphate; araT, 1-(P-o-arabinofuranosy1)thymine; HIV, human immunodeficiency virus; HPMPA, (S)-9-(3-hydroxy-2-phos- phonylmethoxypropy1)adenine; HPMPApp, HPMPA diphosphate; HSV, herpes simplex virus; IDW, dose of compound required to inhibit virus replication by 90%; PFA, phosphonoformic acid; PMEA, 9-(2- phosphonylmethoxyethy1)adenine; PMEAp, PMEA monophosphate; PMEApp, PMEA diphosphate; SDS, sodium dodecyl sulfate.

with herpes simplex virus, the disease is manifested to a much greater extent in AIDS patients.

Several new compounds have been synthesized by Holy and Rosenberg (1-3) that show potent antiviral activity against a wide spectrum of viruses. Fig. lA shows the structure of one of these new compounds, PMEA. PMEA is an adenine nu- cleotide analog that contains an acyclic phosphonylmethox- yethyl moiety in place of the phosphorylated sugar (3). The phosphonate group is designed to be resistant to phosphatases (1,3) and allows the drug to be metabolized, at least i n vitro, to mono- and diphosphate forms in the absence of viral thymidine kinase (4). The diphosphoryl form of PMEA, which is analogous to a nucleotide triphosphate, is an inhibitor of HSV-specified DNA polymerase i n uitro (5). Since the anti- viral activity of PMEA does not depend on activation by viral thymidine kinase, it is effective against viruses that lack thymidine kinase (6). This is in contrast to other compounds, such as acyclovir, which require phosphorylation by viral thymidine kinase to achieve antiviral activity (7).

PMEA was reported to be active against several herpes viruses including HSV-1 and HSV-2 in cell culture (6,8) and in mice (8) and Epstein-Barr virus in the virus-producing P3HR-1 and latently infected Raji cell models (9). In addition, PMEA has been reported to be effective against several retro- viruses including HIV in cell culture systems (6, lo), Moloney murine sarcoma virus in cell culture (10) and in mice (11), and the LP-BM5 retrovirus in a murine acquired immuno- deficiency syndrome model (12). Therefore, PMEA may have potential for use as an antiviral against HSV, Epstein-Barr virus, or HIV.

Because the diphosphorylated metabolite of PMEA, PMEApp, is a potent inhibitor of HSV DNA polymerase i n uitro, it was suggested by Merta et al. (5), that HSV DNA polymerase is a possible target for the antiviral activity of PMEA. However, no conclusive evidence for this has been presented.

In view of the unique features of PMEA including its spectrum of activity and structure, the mechanism of the antiviral action of PMEA is of considerable interest. The present study was aimed at elucidating the target for the anti- HSV-1 action of PMEA. To this end, a PMEA-resistant HSV- 1 strain denoted PMEA”1 was examined. PMEA”1 was cross- resistant to PFA, a compound which directly inhibits HSV DNA polymerase. In addition, PFA-resistant mutants with defined drug resistance mutations within the HSV DNA polymerase gene were cross-resistant to PMEA. Moreover, studies of the purified DNA polymerase from PMEA”1 con- firmed that the enzyme was resistant to PMEApp and PFA. Furthermore, PMEA inhibited HSV DNA synthesis in in- fected cells. These results provide strong evidence that the target for the antiviral action of PMEA is the HSV DNA polymerase. Additional experiments showed that PMEApp

238

HSV DNA Polymerase Is the Antiviral Target of PMEA 239

A.

0

R - 0 4 4

0'

R: H = PMEA Po, = PMEAp (Po,), = PMEApp

B. 120 I I

100 200 300

PMEA (vY)

FIG. 1. Structure and anti-HSV activity of PMEA. A, struc- ture of PMEA and phosphorylated derivatives. B, inhibition of herpes simplex virus yield by PMEA. HeLa S3 cells were infected with HSV- 1 (KOS) (0) or HSV-2 (333) (0), and the virus yield was determined at 24 h postinfection.

was incorporated into and removed from DNA by HSV DNA polymerase in uitro.

MATERIALS AND METHODS

Compounds-PMEA, PMEAp, PMEApp, HPMPA, HPMPApp, and ACGTP were generously provided by A. Holy and I. Rosenberg from the Institute of Organic Chemistry and Biochemistry, Czecho- slovak Academy of Sciences, Prague, Czechoslovakia. Acyclovir was a gift from Thomas Spector, Wellcome Research Laboratories, Re- search Triangle Park, NC. PFA was purchased from Sigma. Aphidi- Colin was obtained from the National Cancer Institute.

Cells and Viruses-HeLa S3 (ATCC CCL 2.2) and Vero (ATCC CCL 80) cells, originally from the American Type Culture Collection, Rockville, MD, were maintained in RPMI 1640 medium supple- mented with 5% fetal bovine serum and 100 pg/ml kanamycin in a 37 "C incubator with 5% CO,. The viruses used included HSV-1 strain KOS and HSV-2 strain 333. PMEA', a gift from V. Vonka, Dept. of Experimental Virology, Institute of Sera and Vaccines, 101 03 Praha 10 (Czechoslovakia), was isolated by stepwise selection of KOS in rabbit embryo fibroblast cells as described (13). For this study, PMEA' was subcloned and denoted PMEA"1. The isolation of PFAr-l, PFAr-2, and PFA"5 has been described previously (14). PFA"2 was a plaque-purified version of the originally described virus. Stocks of the viruses were prepared by low multiplicity passage in Vero cells as previously described (15). Virus yield assays were per- formed essentially as described (16).

DNA Probes"pSG18 plasmid DNA (17) containing the EcoRI F fragment from HSV-1 DNA was a gift from Donald M. Coen, Harvard Medical School. The plasmid DNA was used to transform Escherichia coli from which a stock of pSG18 was prepared. The EcoRI F fragment was then purified from the plasmid by standard techniques (18). The 0.88-kilobase mitochondrial DNA probe was from this laboratory (19). The probes were labeled with [ C Y - ~ ~ P I ~ C T P (6000 Ci/mmol, Amersham) by the random primers method (20). The labeling kit (Boehringer Mannheim) was used as recommended by the manufac- turer.

Purification of HSV DNA Polymerase-HSV DNA polymerase was purified from HSV-infected Hela S3 cells by sequential DEAE-, phospho-, and native DNA-cellulose chromatography as previously described (14).

Preparation of Crude Extracts-Cells were harvested by scraping, rinsed with phosphate-buffered saline, and resuspended in extraction buffer which contained 0.3 M sodium phosphate, pH 7.5,10% glycerol, 1 mM EDTA, 2 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride. The samples were frozen and thawed three times followed by sonication and pelleting of the cell debris at 12,000 X g for 5 min. The resulting supernatant fraction was used for electrophoresis and DNA polymerase activity analysis.

Enzyme Assays-The HSV DNA polymerase standard reaction mixture contained 50 mM Tris-HC1, pH 8.0, 5 mM MgC12, 0.5 mM dithiothreitol, 0.5 mg/ml heat-inactivated bovine serum albumin, 150 mM KC1, 120 pg/ml DNase I-activated calf thymus DNA (211, 0.1 mM concentration each of dTTP, dCTP, and dGTP, and 5 p M [3H] dATP (ICN) at 2 Ci/mmol in a volume of 50 pl. For analysis of crude extracts, column fractions, and the determination of enzyme units, the KC1 was at 200 mM and the labeled substrate was [3H]dTTP (ICN). Following the reaction at 37 'C for 15-30 min, 40-p1 samples were spotted onto GF/A disks (Whatman) for the determination of trichloroacetic acid-insoluble radioactivity. One unit is defined as the incorporation of 1 pmol of dTMP per min into acid-insoluble mate- rial.

The assay for HSV DNA polymerase-associated 3'-5' exonuclease was performed as described previously (22). The reaction mixture contained the same components as the HSV DNA polymerase assay in a volume of 0.2 ml except that the deoxynucleotides were omitted and the activated DNA was replaced with 10 pg/ml of 3' terminally labeled activated calf thymus DNA (6 X lo3 cpm/pg). The labeled DNA was prepared essentially as described (23) with [3H]dTTP as the labeled nucleotide.

Protein concentrations were determined by using the Bio-Rad protein assay (Bio-Rad).

Polyacrylamide Gel Electrophoresis and Western Blotting-SDS- polyacrylamide gel electrophoresis was carried out in 7.5%, 0.75-mm slab gels by using the buffer system previously described (24). Follow- ing electrophoresis of the crude cell extracts, the proteins were transferred to nitrocellulose by diffusion blotting (25). The HSV- specific proteins were detected by immunostaining with rabbit anti- HSV-1 (Dako, Santa Barbara, CA) and goat anti-rabbit alkaline phosphatase conjugate (Promega, Madison, WI) as the primary and secondary antibodies, respectively. The staining procedure was essen- tially as recommended by Promega except that 2% nonfat dried milk with 0.01% thimerosal in phosphate-buffered saline was used as the blocking solution. Color development was achieved by using the nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate system as recommended by Promega.

In Situ Lysis Agarose Gel Electrophoresis-In situ lysis agarose gel electrophoresis was a modification of a previous procedure (26). The body of the gel was 1% agarose while the gel surrounding the wells was 0.5% agarose and contained 2% SDS and 1 mg/ml proteinase K (Boehringer Mannheim). The gel and electrophoresis buffer con- tained 0.5 X TBE (90 mM Tris-borate, 2 mM EDTA, pH 7.8). The gel was placed into a CHEF-DR I1 pulsed field electrophoresis chamber (Bio-Rad) with the electrophoresis buffer. Approximately 1.5 X lo6 cells, which were suspended in TBE with 15% Ficoll, 0.8 mg/ml RNase A, and 0.01% bromphenol blue were loaded into the wells. After 30 min, electrophoresis was started at 15 V. Two h later, the voltage was increased to 150 V. The buffer was circulated through a water bath to maintain approximately 16 "C in the electrophoresis chamber. The total run time was 18 h with an A B switching ramp of 60 to 90 s. Following electrophoresis, the DNA was stained with ethidium bromide and transferred to a Hybond-N nylon membrane (Amersham) by using a vacuum blot apparatus (American Bionetics, Hayward, CA) according to the manufacturer's instructions. The blot was probed for HSV DNA with EcoRI F fragment by standard techniques (18). After autoradiography with Kodak X-Omat-RP film (Sigma), the bands were excised from the blot for scintillation count- ing, and the remaining sections of the blot were probed again with the mitochondrial DNA probe as a control for even loading. For the virus-infected cells, the samples were loaded evenly as judged from the ethidium bromide staining and mitochondrial DNA hybridization.

Primer Extension/Exonuclease Analysis-Primer extension analy- sis was performed essentially as previously described (27). The M13 template-primer was prepared by annealing a 5'-32P-labeled M13 17- base primer (27) to single-stranded M13mp19 DNA. The M13mp19 bacteriophage was propagated and the DNA was purified by standard methods (18). The reaction mixture contained the same components as the HSV DNA polymerase assay except that the activated DNA was replaced by 25 pg/ml labeled M13 template-primer. dATP and

240 HSV DNA Polymerase Is the Antiviral Target of PMEA PMEApp were added or omitted as indicated. The reaction volume was 25 pl. After a 30-min incubation a t 37 "C, the reaction was terminated by the addition of 12.5 p1 of 50 mM EDTA, 0.025% bromphenol blue in deionized formamide; the samples were placed in boiling water for 3 min and then placed on ice. Five microliters were loaded per lane on a 15% polyacrylamide/urea sequencing gel. After electrophoresis, autoradiography was performed with Kodak X- Omat-RP film.

PMEA-terminated primer was generated by running a 50-pl reac- tion in which PMEApp was added a t 50 PM in place of dATP. Ten units of Klenow enzyme (Boehringer Mannheim) was used as the DNA polymerase. This resulted in the extension of the original 17- base primer by 2 nucleotides with PMEA as the terminal nucleotide. The terminated reaction was run on a preparative sequencing gel, and the PMEA-terminated primer was excised and purified as de- scribed (28). The resulting primer was reannealed to single-stranded M13mp19 DNA and used in exonuclease assays. The exonuclease reaction mixture was the same as the primer extension reaction except that the deoxynucleotides were omitted and the M13 template-primer was replaced by MlB/PMEA-terminated primer. The reaction prod- ucts were analyzed as in the primer extension assay.

RESULTS

Sensitiuity of HSV to PMEA-virus yield experiments in HeLa S3 cells showed that both HSV-1 and HSV-2 were inhibited by PMEA with IDso values of 80 and 150 p ~ , respectively (Fig. 1B). These values are somewhat higher than those reported previously (8). This discrepancy may be due to different methodology or cell lines used.

Cross-resistance Studies-In virus yield assays, PMEK-1 was found to be highly resistant to PMEA, PFA, and acyclovir (Table I). Cross-resistance of PMEA' to acyclovir has been reported previously (13). Because PFA and acyclovir exert their antiviral action on the HSV DNA polymerase, these results suggest that PMEA resistance may be associated with alterations in the DNA polymerase. Consistent with this hypothesis, three independent PFA-resistant variants of HSV-1 (KOS), PFA'-1, PFA"2, and PFA"5 (14) were also resistant to PMEA (Table I). Drug resistance in each virus is associated with specific nucleotide substitutions within the HSV DNA polymerase gene (29). Thus, these results suggest that HSV DNA polymerase is the likely target for the antiviral action of PMEA.

In addition, Table I shows that PMEA"1 was collaterally sensitive to aphidicolin. A previous report indicated that a common phenotype for HSV-1 mutants selected for resistance to PFA was collateral sensitivity to aphidicolin (30). Thus, it is possible that collateral sensitivity to aphidicolin may be a common phenotype for PMEA-resistant mutants.

A previous report indicated that PMEA' was more sensitive to HPMPA than was KOS (13). This is somewhat surprising since HPMPA differs from PMEA only by having a methoxy group on the alkyl side chain (3). However, our results with PMEA'-1 confirmed that PMEA resistance was associated with HPMPA hypersensitivity (Table I).

TABLE I ID, values as determined by virus yield assay

ID, Virus

PMEA PFA Acyclovir HPMPA Aphidicolin

" KOS 80 160 14 50 4 PMEA"1 >2000 3000 380 14 0.1 PFA"1 520 900 100" ND" hs' PFA"2 >lo00 >lo00 100" ND PFA"5 370 830 78" ND

hs' h sc

'From Ref. 16. I, ND, not determined.

hs, hypersensitive. From Ref. 30, plaque reduction assay.

Effect of PMEA on HSV DNA and Protein Synthesis- Since the cross-resistance study suggested that HSV DNA polymerase was the target for the antiviral action of PMEA, HSV DNA synthesis in HSV-infected HeLa S3 cells treated for 8 h with different concentrations of PMEA was examined to see whether HSV DNA synthesis was blocked. For the HSV DNA analysis, whole HSV-1 (K0S)-infected cells were harvested and subjected to in situ lysis agarose gel electro- phoresis as described under "Materials and Methods." After electrophoresis, the DNA was transferred to a nylon mem- brane and probed with HSV-1 EcoRI F fragment from plasmid pSG18. There was a dose-dependent decrease in the content of HSV DNA, indicating that PMEA interfered with HSV DNA synthesis (Fig. 2 A ) . The decrease in the HSV DNA content correlated with the decrease in virus yield (Fig. 2B).

HSV protein synthesis as determined by Western blot analysis with anti-HSV-1 antibody, was affected very little under the same conditions used for the DNA synthesis exper- iment (Fig. 3A). While most of the proteins were not affected by the PMEA treatment, some of the protein bands appear to be reduced by the PMEA treatment. These bands most likely represent late proteins whose expression is decreased by blocking DNA synthesis (31). In addition, the activity of HSV DNA polymerase recovered in crude extracts from virus- infected cells was not changed appreciably by the PMEA treatment (Fig. 3B). Thus, while PMEA inhibited HSV DNA synthesis, the induction of HSV proteins as a whole was not significantly impaired. These results are consistent with the hypothesis that PMEA exerts its antiviral action by inhibiting the HSV DNA polymerase.

Analysis of Purified HSV DNA Polymerase-Since the pre- vious results indicated that the HSV DNA polymerase was the likely target for the antiviral action of PMEA, the purified DNA polymerases induced by KOS and PMEA'-1 were tested for sensitivity to the various inhibitors.

Fig. 4A shows that DNA polymerase from KOS was sensi- tive to inhibition by PMEApp, whereas the PMEA'-l enzyme

A, M I 2 3 4 5

a m "

B.

- I ; * 40

5 2 0 2

0 0 20 40 60 80 100 120

HSV DNA (%I FIG. 2. Inhibit ion of HSV-1 (KOS) DNA synthesis in HeLa

53 cells by PMEA. Cells were infected with 3 plaque-forming units/ cell of virus and treated with PMEA for 8 h, then harvested for HSV DNA analysis. A, an autoradiogram of a DNA blot hybridized with an HSV-1 DNA probe. Lanes 1-5, cells were treated with 0, 10, 25, 75, or 250 PM PMEA, respectively; M , mock-infected cells not treated with PMEA. R, the correlation between HSV-1 DNA content and virus yield of cells treated with PMEA. The HSV DNA data are from the experiment shown in A after cutting the bands from the blot and counting in a scintillation counter and the virus yield data are from Fig. 1B.

HSV DNA Polymerase Is the Antiviral Target of PMEA 241

A . k D o M I 2 3 4 5

B.

* a - 1 BO

128 T

Y 1 2 3 4 5

FIG. 3. Induction of HSV proteins in HeLa S3 cells treated with PMEA. Cells were infected with 3 plaque-forming units/cell of HSV-1 (KOS) and treated with PMEA for 8 h. Crude extracts then were prepared and used for SDS-PAGE with 50 pg of protein per lane followed by transfer to a nitrocellulose filter and immunostaining ( A ) or HSV DNA polymerase assays ( R ) . HSV DNA polymerase activity was normalized for the protein content of the extracts. The bars indicate standard deviation for triplicate assays. Lanes 1-5, cells were treated with 0, 10, 25, 75, or 250 p~ PMEA, respectively; M, mock- infected cells not treated with PMEA.

was resistant. Similarly, the PMEA'-1 DNA polymerase was resistant to PFA and ACGTP by a t least 10-fold in compari- son to KOS (Fig. 4, B and C). Although the PMEA'-1 virus was more sensitive to HPMPA, the purified PMEA"1 DNA polymerase was somewhat resistant to HPMPApp (Table 11). PMEA itself and PMEAp had little or no effect on the HSV DNA polymerase with at least 85-90% of activity a t 10 PM concentration of either compound (data not shown).

In view of the resistance of PMEA'-1 DNA polymerase to PMEApp, a more detailed kinetic study was performed. Ki- netic measurements indicated that PMEApp inhibition was competitive with dATP. This is consistent with another report (5). The K,,, of the PMEA"1 enzyme for dATP was increased approximately 5.6-fold along with a 97-fold increase in the Ki for PMEApp in comparison to KOS (Table 11). The net result was that the K,/Ki ratio for the KOS DNA polymerase was approximately 17-fold higher than for PMEA"1 DNA polym- erase. Thus, the enzyme data correlate well with the cell culture data and further support the idea that the HSV DNA polymerase is the target for PMEA via the diphosphorylated metabolite of PMEA.

In addition to HSV DNA polymerase activity, PMEApp, PFA, and ACGTP were tested for their ability to inhibit the polymerase-associated 3'-5' exonuclease activity. As with the DNA polymerase, 3'-5' exonuclease was inhibited by PMEApp, PFA, and ACGTP with the PMEA'-l enzyme being resistant in each case compared to the parental KOS enzyme (Fig. 5). Again, PMEA itself and PMEAp did not inhibit the exonuclease at concentrations up to 10 pM (data not shown). Also, it was noted that PMEA'-1 DNA polymerase had a 3- to 4-fold higher ratio of exonuclease to polymerase activity than the KOS enzyme (data not shown).

Incorporation of PMEApp into DNA-Since the above re- sults indicated that HSV DNA polymerase was the target for the antiviral action of PMEA, it was of interest to determine the mode of interaction between PMEApp and HSV DNA polymerase. Specifically, primer extension analysis was used

1

P 0 10 20 30 40 50

> v) I

C. I

0 1 2 3 4 5 6

Inhibitor concentration (pM)

FIG. 4. Inhibition of HSV DNA polymerase by PMEApp, PFA, and ACGTP. HSV DNA polymerase was purified from HSV- 1 (KOS) (0) or PMEA'-1-infected HeLa S3 cells (0) and tested for sensitivity to PMEApp ( A ) , PFA (R), or ACGTP ( C ) . The bars indicate standard deviation for triplicate assays. For ACGTP only, the dGTP was a t 5 p ~ , otherwise the standard reaction mixture was used.

TABLE I1 Kinetic constants for purified HSV DNA polymerases

Virus dATP K,,," PMEApp K: KJK, HPMPApp K? BM B M B M

KOS 0.23 f 0.02 0.022 ? 0.013 10.5 0.02 PMEA' 1.30 2 0.05 2.13 f 1.14 0.61 0.35 Kinetics constants were determined from Lineweaver-Burk plots

and replots of slopells uersus [I]. Each value represents the mean of three determinations f S.D.

Calculated from an HPMPApp dose-response experiment with 5 pM dATP.

to determine whether HSV DNA polymerase was capable of incorporating PMEApp into DNA. The template for these reactions was single-stranded M13mp19 DNA, and the primer was a 5'-32P-labeled 17-base oligonucleotide with the sequence 5"GTAAAACGACGGCCAGT. Fig. 6A shows the result of this analysis. Lanes 1 and 2 are control reactions without PMEApp, showing the addition of nucleotides to the primer. The enzyme appears to add the first 11 nucleotides in a distributive fashion followed by a more processive polymeri- zation. In reactions with the same 17-nucleotide primer but with M13mp18 as the template, a similar, but not identical pattern of polymerization was observed (not shown). When a 25-nucleotide primer was used with a M13 clone containing a partial genomic DNA of human topoisomerase I as the tem- plate, the pattern of DNA polymerization in general was more processive (not shown). Thus, the pattern of DNA polymeri- zation seems to depend on the primer length, sequence of

242 HSV DNA Polymerase Is the Antiviral Target of PMEA

120 -1 A.

8 0 2 4 6 8 1 0 1 2 -

0 ' ' I b 0 20 40 60 80 100 120

I b C

60

40 t 2o t

0 ' ' I

0 2 4 6 8 1 0 1 2

Inhibitor concentration (I"

FIG. 5. Inhibition of HSV DNA polymerase-associated 3'- 5' exonuclease by PMEApp, PFA, and ACGTP. HSV DNA polymerase was purified from HSV-1 (KOS) (0) or PMEA'-1-infected HeLa S3 cells (0) and the associated 3'-5' exonuclease was tested for sensitivity to PMEApp ( A ) , PFA ( B ) , or ACGTP (C). The bars indicate standard deviation for triplicate assays.

template, and enzyme. Lane 3 is a misincorporation control without dATP or PMEApp. Lane 4 contains PMEApp in place of dATP and shows the incorporation of PMEApp a t the end of the DNA. The faster migration of PMEA-termi- nated product compared to the dAMP-terminated product (lane 1 ) is expected based on the structure of PMEA, which does not have a complete sugar moiety.

Because HSV DNA polymerase was capable of incorporat- ing PMEApp into DNA, it was of interest to determine whether the enzyme could remove the PMEA from the end of the DNA by the polymerase-associated 3'-5' exonuclease which is thought to act as a proofreading activity. This is an important consideration since PMEA, which lacks a 3' hy- droxyl group for the addition of the next nucleotide, would be an absolute chain terminator unless it is removed. To test this, 5'-32P-labeled PMEA-terminated primer was purified, annealed to single-stranded M13 DNA, and used as the sub- strate in exonuclease assays. The products of the reaction were analyzed on sequencing gels. As can be seen in Fig. 6B, the purified HSV DNA polymerase from either KOS or PMEN-1 was able to remove the PMEA from the end of the DNA. In additional experiments (not shown), the PMEA was removed from approximately 20% of the 19-nucleotide PMEA-terminated primer during 4 h of incubation with either KOS (1.6 units) or PMEA"1 (0.52 unit) polymerase, whereas the 3' dTMP was removed from approximately 85% of the 17-nucleotide primer under the same conditions as estimated from densitometric scans of the autoradiograms. Based on these results, it is possible that the PMEA-terminated primer is digested at a slower rate than the primer with a 3' dTMP or that the primer with a 3' PMEA exists in multiple forms,

8 1 2 3 4 5 6 7 6

19-

FIG. 6. A, Incorporation of PMEApp into DNA by HSV-1 (KOS) DNA polymerase. Each reaction contained dTTP, dGTP, and dCTP, with 0.16 (lane 1 ) or 1.6 units of polymerase (lanes 2-4). 1 and 2, control reactions with 5 p~ dATP. 3, dATP omitted. 4,5 pM PMEApp added in place of dATP. B, Removal of incorporated PMEA from DNA by HSV-1 (KOS) and PMEA"1 DNA polymerase-associated exonuclease. Nineteen-base PMEA-terminated primer annealed to single-stranded M13 DNA was used as the substrate for exonuclease assays. Lane I contained 1.6 units (KOS) of HSV DNA polymerase; lane 5 contained 0.52 unit (PMEA"1) of HSV DNA polymerase. Lanes 2-4 and 6-8 contained 1 : lO serial dilutions of KOS and PMEA" 1 enzyme, respectively.

at least one of which is resistant to degradation. A third possibility is that products from the degradation of the PMEA-terminated primer are inhibitory to the reaction, al- though as noted above, PMEA itself had little inhibitory activity toward the 3'-5' exonuclease activity with the ['HI dTTP-labeled calf thymus DNA substrate. We are currently investigating these possibilities. This experiment also illus- trates the increase in the exonuclease to polymerase ratio for the PMEA'-1 polymerase. Note that the DNA polymerase activity for the KOS enzyme was approximately 3-fold higher than that for the PMEA"1 enzyme, yet the exonuclease activity as reflected in the gel is very similar.

DISCUSSION

PMEA has recently been evaluated for its antiviral effects against several types of viruses. PMEA was active against herpes simplex virus, retroviruses including HIV, and Ep- stein-Barr virus. Thus, PMEA could be useful in AIDS treat- ment for both the underlying HIV infection and some of the common opportunistic infections (12). Whether PMEA will be effective and safe in the treatment of human disease remains to be determined.

The mechanism of the antiviral action of PMEA has not been reported to date. A report by Vonka et al. (13) indicated

HSV DNA Polymerase Is the Antiviral Target of PMEA 243

that PMEA' was cross-resistant to acyclovir and araT. These results suggested the possibility that HSV DNA polymerase could be involved in PMEA resistance. However, since the activation of acyclovir and araT by viral thymidine kinase is required for antiviral activity, cross-resistance to these com- pounds is not necessarily due to HSV DNA polymerase alter- ations. Thus, PMEA"1 was tested for cross-resistance to PFA. A body of evidence indicates that the antiviral target of PFA is the HSV DNA polymerase (for example Refs. 14 and 32). Furthermore, PFA does not require metabolic activation to achieve antiviral activity. Therefore, alterations that affect PFA sensitivity are very likely to involve HSV DNA polym- erase itself. PMEA"1 was found to be cross-resistant to PFA, which provides evidence that HSV DNA polymerase is in- volved in PMEA resistance.

In addition to testing PMEA"1 for cross-resistance to PFA, several PFA-resistant mutants were tested for cross-resist- ance to PMEA. Each PFA-resistant mutant was cross-resist- ant to PMEA. These results provide additional evidence that HSV DNA polymerase is the likely target for PMEA since drug resistance for each PFA-resistant mutant has been mapped to specific nucleotide substitutions within the HSV DNA polymerase gene (29).

Since the cross-resistance data implicated HSV DNA po- lymerase in PMEA resistance, PMEA was tested for its ability to inhibit HSV DNA replication in cell culture. The DNA synthesis experiment demonstrated that HSV DNA synthesis is inhibited by PMEA and since HSV protein synthesis was largely unaltered, inhibition of DNA synthesis is likely to be a primary effect of PMEA. In addition, the decrease in HSV DNA correlated with the decrease in virus yield.

In order to determine whether the HSV DNA polymerase was indeed altered in PMEA"1, the DNA polymerase was purified from HSV-1 (KOS) and PMEK-1. A previous report indicated that PMEApp was a potent inhibitor of HSV DNA polymerase (5). When tested for sensitivity to PMEApp, the PMEA"1 enzyme was quite resistant compared to the KOS enzyme. The PMEA'-1 enzyme was also resistant to PFA and ACGTP, in agreement with the cell culture data and confirm- ing that the PMEN-1 DNA polymerase was indeed altered. These results provide an additional level of evidence that HSV DNA polymerase is the target for the antiviral action of PMEA.

In addition to cross-resistance to PFA, PMEAr-l showed collateral sensitivity to aphidicolin. Collateral sensitivity of PFA-resistant mutants to aphidicolin has been previously reported, and it was suggested that combination chemother- apy with PFA and aphidicolin could be considered in HSV infections (30). In view of the common characteristics of PFA- and PMEA-resistant HSV-1, the same reasoning may apply to the combination of PMEA and aphidicolin. However, a study of several independently derived PMEA-resistant mu- tants would be necessary to substantiate this point further.

It is important to note that our results do not exclude the possibility that other actions of PMEA may contribute to its antiviral activity. For example, PMEApp was reported to be an inhibitor of HSV ribonucleotide reductase (33). However, we conclude that HSV DNA polymerase is the primary target for PMEA.

Since these studies focused on HSV-1, the conclusions must be restricted to HSV-1. Thus, although the target of PMEA against HSV-2 is likely to be the HSV-2 polymerase, this remains to be confirmed.

In terms of the mechanism of PMEApp inhibition of HSV DNA polymerase, the primer extension analysis demonstrated that PMEApp is a substrate for HSV DNA polymerase in

uitro and is incorporated at the end of the DNA strand. Since PMEA does not have a 3' hydroxyl group from which the DNA chain could be elongated, the addition of PMEA to the DNA marks the termination of DNA strand elongation. How- ever, the exonuclease results indicate that the HSV DNA polymerase-associated 3 '6 ' exonuclease activity is able to remove incorporated PMEA. Thus, if the same processes occur inside the cell, PMEA would be continually misincor- porated and removed by the HSV DNA polymerase resulting in inhibition of HSV DNA synthesis. Whether incorporation and removal occurs in the DNA of uninfected cells is not known, although cellular polymerases p and to a lesser extent a appear to be less susceptible to inhibition by PMEApp (5).

Finally, it is noteworthy that PMEA"1 was more sensitive to HPMPA compared to KOS. This result is in agreement with Vonka et al. (13). Also, the difference in HPMPA sen- sitivity between KOS and PMEA"1 was not reflected in the sensitivity of the purified polymerases to HPMPApp. Indeed, the PMEA"1 DNA polymerase was somewhat resistant to HPMPApp. These results raise the possibility that HPMPA may have a different mechanism of action than PMEA. An independent study by Vonka et al. (13) led to a similar suggestion. We are currently investigating this possibility.

REFERENCES 1. Holy, A., and Rosenberg, I. (1987) Collect. Czech. Chem. Commun.

2. Rosenberg, I., and Holy, A. (1987) Collect Czech. Chem. Commun.

3. Holy, A., and Rosenberg, I. (1987) Collect. Czech. Chem. Commun.

4. Merta, A., Vesely, J., Votruba, I., Rosenberg, I., and Holy, A. (1990) Neoplusma (Bratisl.) 37, 111-120

5. Merta, A., Votruba, I., Rosenberg, I., Otmar, M., Hrebabecky, H., Bernaerts, R., and Holy, A. (1990) Antiviral Res. 13, 209-218

6. De Clercq, E., Sakuma, T., Baba, M., Pauwels, R., Balzarini, J., Rosenberg, I., and Holi, A. (1987) Antiviral Res. 8, 261-272

7. Coen, D. M., and Schaffer, P. A. (1980) Proc. Natl. Acad. Sci. CJ.

8. De Clercq, E., Holy, A., and Rosenberg, I. (1989) Antimicrob. Agents Chemother. 33, 185-191

9. Lin, J.-C., De Clercq, E., and Pagano, J. S. (1987) Antimicrob. Agents Chemother. 31, 1431-1433

10. Pauwels, R., Balzarini, J., Schols, D., Baba, M., Desmyter, J., Rosenberg, I., Holy, A., and De Clercq, E. (1988) Antimicrob. Agents Chemother. 32,1025-1030

11. Balzarini, J., Naesens, L., Herdewijn, P., Rosenberg, I., Holy, A., Pauwels, R., Baba, M., Johns, D. G., and De Clercq, E. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 332-336

12. Gangemi, J. D., Cozens, R. M., De Clercq, E., Balzarini, J., and Hochkeppel, H.-K. (1989) Antimicrob. Agents Chemother. 33,

13. Vonka, V., Anisimova, E., Cerny, J., Holy, A., Rosenberg, I., and

14. Derse, D., Bastow, K. F., and Cheng, Y. (1982) J. Biol. Chem.

15. Nutter, L. M., Grill, S. P., Dutschman, G. E., Sharma, R. A., Bobek, M., and Cheng, Y . 4 . (1987) Antimicrob. Agents Chem- other. 3 1, 369-374

16. Cheng, Y.-C., Huang, E.&, Lin, J.-C., Mar, E.-C., Pagano, J. S., Dutschman, G . E., and Grill, S . P. (1983) Proc. Natl. Acad. Sci.

17. Goldin, A. L., Sandri-Goldin, R. M., Levine, M., and Glorioso, J. C. (1981) J. Virol. 38, 50-58

18. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY

19. Chen, C.-H., and Cheng, Y.-C. (1989) J. Biol. Chem. 264,11934- 11937

20. Feinberg, A. P., and Vogelstein, B. (1983) Anal. Biochem. 132,

21. Baril, E., Mitchener, J., Lee, L., and Baril, B. (1977) Nucleic

52,2775-2791

52,2792-2800

52,2801-2809

S. A. 77, 2265-2269

1864-1868

Votruba, I. (1990) Antiviral Res., 14, 117-122

257, 10251-10260

U. S. A 80,2767-2770

6-13

Acids Res. 4, 2641-2653

244 HSV DNA Polymerase Is the Antiviral Target of PMEA

22. Hoffmann, P. J., and Cheng, Y. (1978) J. Biol. Chern. 253,3557- 28. Vazquez-Padua, M. A., Starnes, M. C., and Cheng, Y.-C. (1990)

23. Derse, D., and Cheng, Y.-C. (1981) J. Biol. Chern. 2 5 6 , 8525- 29. Gibbs, J. S., Chiou, H. C., Bastow, K. F., Cheng, Y.-C., and Coen,

24. Laemmli, U. K. (1970) Nature 227,680-685 30. Bastow, K. F., Derse, D. D., and Cheng, Y.-C. (1983) Antirnicrob. 25. Bowen, B., Steinberg, J., Laemmli, U. K., and Weintraub, H. Agents Chernother. 23,914-917

(1980) Nucleic Acids Res. 8, 1-20 31. Honess, R. W., and Roizman, B. (1974) J. Virol. 14,8-19 26. Gardella, T., Medveczky, P., Sairenji, T., and Mulder, C. (1984) 32. Cheng, Y.-C., Grill, S., Derse, D., Chen, J.-Y., Caradonna, S. J.,

J. Virol. 5 0 , 248-254 and Connor, K. (1981) Biochim. Biophys. Acta 652,90-98 27. Townsend, A. J., and Cheng, Y.-C. (1987) Mol Pharrnacol. 32 , 33. Cerny, J., Votruba, I., Vonka, V., Rosenberg, I., Otmar, M., and

3562 Cancer Commun. 2,55-62

8530 D. M. (1988) Proc. Natl. Acad. Sci. U. S. A. 85,6672-6676

330-339 Holy, A. (1990) Antiviral Res. 13, 253-264