Inhibitory effect of E-5-(2-bromovinyl)-1-beta-D-arabinofuranosyluracil on herpes simplex virus replication and DNA synthesis.

JOURNAL July 43,No. 0022-538X/82/070332-05$02.00/0

Inhibitory

Effect of

E-5-(2-Bromovinyl)-l-Ip-D-Arabinofuranosyluracil

on

Herpes Simplex

Virus

Replication

and

DNA

Synthesis

J. DESCAMPS,1 R. K.SEHGAL,1 E. DECLERCQ,2ANDH. S.ALLAUDEEN1*

DepartmentofPharmacology, YaleUniversitySchool of Medicine, New Haven, Connecticut06S10,1and RegaInstitute,Katholieke Universiteit Leuven, B-3000 Leuven,

Belgium2

Received23October1981/Accepted2March1982

The effect of E-5-(2-bromovinyl)-1-13-D-arabinofuranosyluracil (BVaraU) on

herpessimplex virus (HSV) replicationwasexamined and compared with thatof

E-5-(2-bromovinyl)-2'-deoxyuridine

(BVdUrd). The

50%

inhibitory dose against

HSVtype1 (HSV-1)was0.1 ,ug/mlcomparedwith 0.008,ug/mlfor BVdUrd; the

antimetabolic 50% inhibitory dose of BVaraU ranged from 20 to 95 ,ug/ml. The

addition of 50 gxg of BVaraUperml toHSV-1-infected Vero cells decreasedthe

synthesis of viral and cellular DNA by 37 and 28%, respectively. The

5'-triphosphate(BVaraUTP) competed withdTTPin DNAsynthesis by the

herpes-viral and cellular DNA polymerases; the apparent K, values of HSV-1 DNA

polymerase, DNApolymerase a,and DNApolymerase 1Bwere0.14, 0.32, and5

,uM, respectively. Thus, BVaraUwas aless effective

antiherpesvirus

agentthan

BVdUrd; unlike BVdUrd, it did not appear to be

internally

incorporated

into

replicatingDNA in virus-infected cells.

Among thenewly developed analogs,

E-5-(2-bromovinyl)-2'-deoxyuridine (BVdUrd)

emerged as one of the most promising agents

against herpes simplex virus type 1 (HSV-1)

replicationboth in cell culture and in

experimen-tal animals (5, 6). More recently, anarabinose analog of BVdUrd,

E-5-(2-bromovinyl)-1-13-D-arabinofuranosyluracil (BVaraU), was

synthe-sized by Sakata et al. (12) and at the Rega

Institute, Leuven, Belgium. Inits initial in vitro

trials, it showed promising selective antiviral

action. (E. De Clercq, R. Busson, L. Colla, J.

Descamps, J. Balzarini, and H. Vanderhaeghe,

Program Abstr. Int. Congr. Chemother., 12th,

Florence, Italy, abstr. no. 316, 1981). To

eluci-date the mode of selective antiherpesviral action of BVaraU, we investigated its antiviral and

antimetabolic propertiesandexamineditseffect onDNAreplication in virus-infectedcells. The

effect of chemically synthesized 5'-triphosphate

on the activities of purified viral and cellular DNApolymerasesis alsodescribedin this

com-munication.

The antiviral assays were carried out with

primaryrabbitkidney(PRK) cellcultures

infect-ed with HSV (4). The results are expressed as

ID50(50%inhibitorydose), the concentrationof

compound required to reduce viral cytopatho-genic effect (CPE) by 50% when virus-infected control PRK cells reached complete CPE. The data presented are anaverage of at least three separate experiments. The ID50 value for

BVaraUagainst three strains ofHSV-1was0.1

,ug/ml compared with 0.007 to 0.009 p.g/ml for

BVdUrd (Table 1). BVaraU was less effective against HSV-2 replication;theID50rangedfrom

15to70 ,ug/mlcomparedwith 1 to1.5,ug/mlfor

BVdUrd. Neither BVaraU nor BVdUrd was

active against an HSV-1 TK- (B2006) mutant, indicating that both nucleoside analogs needto

be phosphorylated by the virus-induced

thymi-dine kinase to exert their antiviral activity. BVaraUwasnotactiveagainst vaccinia virusor

vesicularstomatitisvirus. The antimetabolic

ac-tivity ofBVaraUwas monitored by the

inhibi-tion ofdThdordUrdincorporation intoDNAof

PRKcells (4);[1',2'-3H]2'-deoxyuridine

(specif-icactivity, 42 Ci/mmol), and

[methyl-3H]thymi-dine(specific activity 12 Ci/mmol) were used as

thelabeled DNA precursors. BVaraU inhibited

cellular DNA synthesis only at doses much higher than that required to inhibit the viral-induced CPE (Table 1).

To confirm that the inhibitory effect of BVaraUonCPE isdue toits abilitytoinhibit the

virusmultiplication, we tested its effect on the replication of HSV-1 (KOS) in PRK cells. Both BVaraU and BVdUrd inhibited the viral multi-plication, albeit to adifferent extent. For exam-ple,at100,ug/ml after 24 h, BVaraU reduced the virus replicationby 3.9 log'0, whereas BVdUrd reduced it by 5.9 log10;after 48 h,BVaraUat 100

,ug/ml reduced virus replication by 3.4 log10, whereas BVdUrd reduced it by 5.4

logl0.

332

on November 10, 2019 by guest

http://jvi.asm.org/

[image:2.504.60.256.94.277.2]

NOTES 333

TABLE 1. Antiviral and antimetabolic activities of BVaraU and BVdUrd in PRK cell culturesa

Antiviralor-metabolicactivity D qgMb

BVaraU BVdUrd Virus

HSV-1 (KOS) 0.1 0.007

HSV-1 (F) 0.1 0.007

HSV-1 (Mac Intyre) 0.1 0.009

HSV-1TK-(B2006) >200 100

HSV-2 (G) 15 1

HSV-2(Lyons) 20 1

HSV-2 (1%) 70 1.5

Vaccinia >200 7

Vesicular stomatitis virus >200 >200

Antimetabolicactivityc

[methyl-3H]dThd 20 70

incorporation

[1',2'-3H]dUrd 95 20

incorporation

a PRK cells were inoculated with 100 cell culture

infectingdoses of virus for 1 h at 37°C and exposed to various concentrations of the test compound (in Eagle minimal essential medium plus 3% fetal bovine se-rum). CPE was recorded daily. In virus-infected but untreated cell culture controls, the viral CPE was generally complete after3daysfollowing viral inocula-tion.

bResultsareexpressedasID50,i.e., the concentra-tionrequired to inhibit viral CPE or incorporation of labeled precursors by 50%compared with untreated

controls;0.1p.g/mlisequivalentto0.29and 0.3 I1M of BVaraU and BVdUrd,respectively.

c[1'2'-3H]2'-deoxyuridine (specific activity, 42 Ci/ mmol)and

[methyl-3H]thymidine

(specific activity,12

Ci/mmol)wereusedasthelabeled DNA precursors in

experiments todeterminetheantimetabolicactivity.

The effectof the 5'-triphosphate(BVaraUTP)

onthe utilization ofdTTPbyviral and cellular

DNA polymerases was determined. The

BVaraUwasconvertedtoits5'-triphosphateby

procedures described previously (3). The DNA

polymerases used in these experiments were

purified by successive chromatography, using

phosphocellulose

and DNA-cellulose columns.

These enzymes were free of

cross-contamina-tion;forexample,HSV-1 DNApolymerasewas

free of DNA polymerase a and vice versa. In

their

properties

suchaselutionpositionsin

ion-exchangecolumns,primer-template preference, effectsofmono-anddivalentcations,sensitivity

to

N-ethylmaleimide

and aphidicolin, these

en-zymesresembledtherespectiveviral and

cellu-lar DNA polymerases. The polymerase assays were

performed

under conditions optimum for

theindividualenzymes with the concentration of

[3H]dTTP

maintainedattwo tothree times the

Km value for the respective polymerase (1 LM

for HSV-1 DNA polymerase, 10 ,uM for DNA

polymerasea, and30 ,uM for DNApolymerase

,B and Epstein-Barrvirus [EBV] DNA

polymer-ase). HSV-1 DNA polymerase was the most

sensitive among the polymerases tested; DNA

polymerase a was the next sensitive enzyme.

However, DNA polymerase M and EBV DNA

polymerase were relatively insensitive. At 1

FM, BVaraUTPinhibited the activities of HSV-1 DNApolymerase, DNApolymerase a, DNA

polymerase

1,

and EBV DNA polymerase by

60, 48, 15, and 10%, respectively; at 10 FM, it

inhibitedtheiractivitiesby 90, 82, 34, and 23%,

respectively.

Todeterminethe nature ofBVaraUTP inhibi-tion, further experiments were performedwith

increasing concentrations of dTTP. Results

show that the inhibition was competitive with

dTTP(8); the mode ofinhibition appears to be

thesameforall ofthe enzymes examined(Fig.

1).TheKmvalues for dTTP and the

Ki

value for

BVaraUTP are depicted in Table 2. Although

theKivalues forHSV-1 DNApolymerase(0.14

,uM BVaraUTP) and DNA polymerase a (0.32

TABLE 2. Kinetic analysis of BVaraUTP

inhibitiona

Km Kj

Enzyme (dTTP) (BVaraUTP) K/K, (>M) j(M)

HSV-1 DNA polymerase 0.35 0.14 2.5 Human DNApolymerase a 5.16 0.32 16.12

Human DNApolymerase,1 14.8 52.5

aDNApolymeraseaactivitywasassayed ina50-,ul

reaction mixture consisting of 50 mM Tris-hydrochlo-ride (pH 8.0), 2 mMdithiothreitol, 8mMMgCl2,100 FM each of dATP, dCTP, and dGTP, and various concentrations of[3H]dTTP (530cpm/pmol),10 ,ugof activated calf thymus DNA, 10 to 20

"g

of bovine serumalbumin,5tol1o0 glycerol,and enzyme. Incu-bationwas at37°Cfor 30 min.Acid-insoluble

radioac-tivitywascollectedon anitrocellulose filter(Gelman

orMillipore,0.45 ,um)washedseveral times with 5%

trichloroacetic acid containing 2 mM sodium

pyro-phosphate,washedoncewith70%oethanol,dried,and measured inaliquid scintillationcounter. DNA

poly-merase ,B activity was assayed under similar

condi-tions except thatapH 9.0Tris-hydrochloride buffer, 50 FM ofnonradioactive triphosphates and 40 mM

KCl were used. The reaction mixture for assaying HSV-1 DNA polymerase activity contained 50 mM

Tris-hydrochloride (pH 8.3), 2 mM dithiothreitol, 4

mMMgCI2, 10

FzM

eachofdATP,dCTP,anddGTP, andvarious concentrations of[3H]dTTP(4,300cpm/ pmol), 5 ,ug ofactivatedcalfthymus DNA, 50 mM

ammoniumsulfate,10 FLgof bovineserumalbumin, 5

to10%7glycerol, and enzyme. Other conditionswere

similartothosedescribed above formeasuringDNA polymeraseaactivity.EBV DNApolymeraseactivity

wasassayedundersimilarconditions for DNA

poly-meraseaexcept that 4mMMgCl2and 100 mMKCI

wereused. VOL.43, 1982

on November 10, 2019 by guest

http://jvi.asm.org/

[image:2.504.267.458.330.413.2]

334 NOTES

,uM BVaraUTP) are almost similar, the Km/Ki

ratio forDNApolymeraseawassix timesmore than thatfor the HSV-1 DNApolymerase.This

implies that BVaraUTP is less selective than

BVdUTP. Since the submission of this

manu-script, Ruth and Cheng reported a Kivalue of

0.013 ,uM for BVaraUTP of HSV-1 DNA

poly-merase(11).

DNA polymerases

require

all four natural

deoxyribonucleoside triphosphates

for

optimum

activity. However, a limited DNA

synthesis

occurs whenoneor two

triphosphate

substrates

are omitted from the reaction mixture.

Using

thisphenomenon,wetested whether BVaraUTP

can substitute dTTP in aDNA

polymerization

reaction. BVaraUTP was added to the HSV-1

DNApolymerase reactionmixture

containing

all

triphosphates except

dTTP,

and the reaction

wasfollowed for180min

(Fig. 2).

The addition

ofBVaraUTP instead of dTTP did notenhance the

polymerization

reaction. There was no in-crease in DNA

synthesis

even when excess

dTTPwasaddedatthe endof 120 min.

Howev-er, theadditionofdTTPtothereaction mixture

containing onlythe three nucleotides at120min

increasedDNA

synthesis

as

expected.

The

addi-tion ofnew enzyme and dTTP to the reaction mixture containing the analog did not enhance

DNAsynthesis;theaddition of excesstemplate

anddTTP increasedDNAsynthesis only

slight-ly. It appears that BVaraUTP was not an alter-nate substrate for DNA polymerase and that it did not getinternally incorporatedinto replicat-ing DNA (Fig. 2). A similarpatternwasobtained for DNA polymerase a. Preincubation of the enzyme in the presence of different concen-trations ofBVaraUTP for 18 h at 40C did not alter the enzyme reaction (data not shown). These results areconsistent with the view that BVaraUTP may be a chain terminator for the

polymerizationreaction.

To study theeffect of BVaraU on DNA syn-thesis in virus-infected cells, Vero cells were

infectedwithHSV-1 and grown in the presence

of BVaraU; the viral and cellular DNAs were

separatedby isopycnic CsCl centrifugation (2).

FIG. 1. DNA polymerases a and a were purified

from leukocytesofapatient withacutemyelogenous leukemia. EBV DNA polymerasewaspurified froma

virus-producing cell line, P3HR-1,originally derived fromapatientwith Burkitt'slymphoma. The isolation proceduresweredescribed previously(1).TheHSV-1 DNApolymerasewaspurifiedfromKBcellsinfected with HSV-1 strain HF.Growth conditions and

isola-tion procedures of HSV-1-infected KB cells were

described by Shipman et al. (13). (A) Effect of BVaraUTP on HSV-1 DNA polymerase reaction in thepresenceof different concentrationsof[3H]dTTP with activatedDNA template.Symbols: 0,no

inhibi-tor, 0,1 ,LM BVaraUTP;and0,4,uM BVaraUTP.

(B) Effect ofBVaraUTPonDNApolymerasea

reac-tion in the presence of different concentrations of [3H]dTTP withactivated DNAtemplate. Otherassay

conditions wereoptimal for the DNA polymerase a

activity. Symbols: *,noinhibitor; A,0.5,M

BVar-aUTP;0,1p,MBVaraUTP;and0,4,uMBVaraUTP.

(C) Effect ofBVaraUTPonDNApolymerase

reac-tion in the presence of different concentrations of

[3H]dTTP withactivatedDNA template. Otherassay

conditions wereoptimal for the DNA polymerase ,B activity. Symbols: 0, no inhibitor; 0, 4 p,M

BVar-aUTP; and0,20P,MBVaraUTP.

A

B

C

I/dTTP. .M'

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.504.70.229.64.549.2]

NOTES 335

DNA (2). Recently, Pelling et al. (10) demon-f strated that another arabinose analog,

9-P-D-arabinofuranosyladenine (araA), is not an

abso-lute chain terminator and is internally incorporated into viral and cellular DNAs.

The results presented in this communication demonstrate that BVaraU is approximately 10 times less active than BVdUrd against HSV-1 replication in PRK cells. However, recently Machida et al., usingdifferent HSV-1 strains and humanembryonic lung fibroblast cells, observed nodifference between BVaraU and BVdUrd in their abilities to inhibit replication of HSV-1

(0.032 ,ug/ml) and HSV-2 (100 ,ug/ml) (9). The

180 reason for differences in values of the anti-herpesvirus activity between the two

labora-FIG. 2. Time course of BVaraUTP inhibition of

HSV-1 DNA polymerase reaction. Activated DNA

wasusedastemplateand theconditionswereoptimal

for HSV-1 DNA polymerase. Symbols: 0, all four

nucleotidespresent; 0,dATP,dCTP,and [3H]dGTP

present;*,dATP, dCTP,and [3H]dGTP,plus 40,uM BVaraUTP present. At 120 min, 40 MMdTTP was

addedto thereaction mixturecontaining eitherthree

nucleotides (U) or three nucleotides plus 40 FM BVaraUTP(O) andthe reactionwasstoppedat30and

60min afterthe addition.

BVaraUinhibited thesynthesis ofbothviraland

cellular DNAs (Fig. 3). Forexample, the addi-tion of 50 ,ug of BVaraU per ml resulted in a

decrease in synthesis of both viral and cellular DNA by 37 and 28%, respectively. The inhibi-tionwasproportionaltotheconcentration of the nucleoside analogadded. The effect of the addi-tion of 50,ug ofBVdUrdperml isshown inFig.

3D. Wehave alsoexamined whether therewas anyqualitative change in the viral and cellular

DNAs synthesizedin the presenceof BVaraU.

Previously, it has been shown that growing virus-infected cells in the presence ofeither

5-iodo-2'-deoxyuridine (IdUrd)or

5'-amino-5-iodo-2'-deoxyuridine(AIdUrd)resultedin the

incorpo-ration of these analogs into DNA, which was

measured by an increase in densities of the

DNAs(7). Substitution of2%ormoreof

thymi-dinebyBVaraUinto DNA would have resulted

in an increase in densities ofviral and cellular

DNAs. Therewas no differencein densitiesof

either viral or cellular DNA isolated from the virus-infected cells grown in presence of

BVaraU (Fig. 3).This indicated that theanalog

waseithernotinternallyincorporatedor insuffi-ciently incorporated into replicating DNAs in virus-infected cells, which is consistent with

results shown inFig.2.However,virus-infected

cellsgrownin the presence of 50,ugof BVdUrd permlshowedanincreaseindensityofthe viral

DNA(from1.725to1.74g/ml)butnot in cellular

3

to

o . . , .

B.BVoraU I0jg/mI

C

ii 6

4

0~~~~~~~~~~ %

011 0

OI

09

OD 0

090--6 C. BVoroU 50#Lg/mI

a. 4 Ca

~~~~~~~~~~~~~00

2

o diOumi.0 i i

50Fg/mlo 144,u;and D. BVdUrd 50ag/mi

2 F.0

0 5 10 5 20 25

Fractions

FIG. 3. CsCl density gradient analysis ofcellular and viral DNA in HSV-1 (Cl101)-infectedVero cells

at a multiplicity of infection of four. After 1 h of contactat370C,residual viruswasremoved,and 4 ml ofgrowthmediumcontainingeithernonucleosideor

BVaraU was added. To each petri dish, 20 MCi of

[3H]deoxyguanosine (specific activity, 14.5 Ci/mmol)

wasadded andincubatedfor 24 h. (A) Cellular DNA

density,1.695g/ml;HSV-1 DNAdensity,1.725g/mlin

control;(B)BVaraU, 10iLg/mlor29 MLM;(C)BVaraU 50ILg/mlor144 MLM;and(D)BVdUrd50 pLg/mlor150

MLM.

0₀ 0

20

CL

0

I 0

lo

0.

x

3CL

90 Time (min)

VOL.43,1982

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.504.64.252.56.236.2] [image:4.504.268.458.258.556.2]

tories is not yetknown. HSV-1 DNA polymer-ase and cellular polymerase a are the most

sensitive to the BVaraUTPinhibitionamongthe

enzymestested. TheKivalues for BVaraUTPof

DNA polymerase specific to two strains of HSV-1, i.e., strain Cl 101 propagated in Vero cells and strain HF propagated in human KB

cells are similar (0.14 to 0.2 ,uM). The analog does not appearto get internally incorporated into replicating DNA. Our results show that

BVaraU wasapproximately100 times less

effec-tive ininhibiting the replication of HSV-2 than

HSV-1. Since HSV-2 markedly differs from

HSV-1 in its susceptibility to BVaraU, the dif-ferential sensitivity can be exploited in

distin-guishing the two types ofHSV.

This research wassupported byPublic Health Service grant CA28852-01 and AmericanCancerSocietygrant CH47-C.

LITERATURE CITED

1. Allaudeen, H. S., and J. R. Bertino. 1978. Isolation of herpesvirus-specific DNA polymerase from tissues of an American patient with Burkitt's lymphoma. Proc. Natl. Acad.Sci. U.S.A. 75:4504-4508.

2. Allaudeen, H. S., M. S. Chen, J. J.Lee, E. De Clercq, and W.H. Prusoff.1982. Incorporation of E-5-(2-halovinyl)-2'-deoxyuridines intodeoxyribonucleic acids of herpes sim-plex virus type-1 infected cells. J. Biol. Chem. 257:603-606.

3. Allaudeen,H. S., J. W. Kozarich, J. R. Bertino, and E. DeClercq. 1981.On the mechanism of selective inhibition ofherpesvirusreplication by E-5-(2-bromovinyl)-2' deox-yuridine. Proc.Nati.Acad. Sci.U.S.A. 78:2698-2702. 4. De Clercq, E., J. Descamps, G. F. Huang, and P. F.

Torrence. 1978. 5-Nitro-2'-deoxyuridine and 5-nitro-2'-deoxyuridine 5'-monophosphate: antiviral activity and

inhibition of thymidylate synthetase in vivo. Mol. Phar-macol. 14:422-430.

5. DeClercq, E., J.Descamps, P. De Somer, P. J. Barr, A. S. Jones, and R. T. Walker. 1979. (E)-5-(2-bromovinyl)-2'-deoxyuridine: a potent and selective antiherpes agent. Proc. Natl.Acad. Sci. U.S.A. 76:2947-2951.

6. De Clercq, E., J. Descamps, G.Verhelst, R. T. Walker, A.S. James, P. F. Torrence, and D.Shugar. 1980. Com-parative efficacy of antiherpes drugs against different strains of herpes simplex virus. J. Infect. Dis. 14:563-574. 7. Fisher, P. H., M. S.Chen, and W. H. Prusoff. 1980. The

incorporationof5-iodo-5'-amino-2',5'-dideoxyuridine and 5-iodo-2'-deoxyuridine into herpes simplex virus DNA. Biochim. Biophys. Acta 606:236-246.

8. Lineweaver, H., and 0. Burk. 1934. The determination of enzyme dissociation constants. J. Am. Chem. Soc. 56:658-666.

9. Machida, H., S. Sakata, A. Kuninaka, and H. Yoshino. 1981. Antiherpesviral and anticellular effects of 1-1-D-arabinofuranosyl-E-5-(2-halogenovinyl)uracils. Antimi-crob. AgentsChemother. 20:47-52.

10. PellIng, J. C., J. C. Drach, and C. Shipman, Jr. 1981. Internucleotide incorporation of arabinosyladenine into herpes simplex virus and mammalian cell DNA. Virology 109:323-335.

11. Ruth, J. L., and Y. C. Cheng. 1981.Nucleoside analogs with clinical potential in antivirus chemotherapy. The effect ofseveral thymidine and 2'-deoxycytidine analog 5'-triphosphatesonpurifiedhuman(a,>)and herpes sim-plexvirus (types 1, 2) DNA polymerases. Mol. Pharma-col. 20:415-422.

12. Sakata, S., S. Shibuya, H. Machida, H. Yoshino, K. Mirota, K. S. Senda, K. Ikeda, and Y. Mizuno. 1980. Synthesisand antiherpesviral activity of 5-C-substituted uracil nucleosides. Nucleic Acids Res. Symp. Ser. 8:539-542.

13. Shipman, C., Jr., S. M. Smith, R. M. Carlson, and J. C. Drach. 1976. Antiviral activity of arabinosyladenine and arabinosylhypoxanthinein herpessimplex virus-infected KB cells: selective inhibition of viral deoxyribonucleic acid synthesis in synchronized suspension cultures. Anti-microb.AgentsChemother. 9:120-127.

on November 10, 2019 by guest

http://jvi.asm.org/