Inhibition of Cap (m7GpppXm)-Dependent Endonuclease ...€¦ · ANTIMICROBIALAGENTSANDCHEMOTHERAPY, Dec. 1994, p. 2827-2837 0066-4804/94/$04.00+0 Copyright X 1994, American Society

ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Dec. 1994, p. 2827-28370066-4804/94/$04.00+0Copyright X 1994, American Society for Microbiology

Inhibition of Cap (m7GpppXm)-Dependent Endonucleaseof Influenza Virus by 4-Substituted 2,4-Dioxobutanoic

Acid CompoundsJ. TOMASSINI,'* H. SELNICK,3t M. E. DAVIES,2 M. E. ARMSTRONG,2 J. BALDWIN,3t M. BOURGEOIS,3

J. HASTINGS,' D. HAZUDA,' J. LEWIS,2 W. McCLEMENTS,2 G. PONTICELLO,3 E. RADZILOWSKI,3G. SMITH,3 A. TEBBEN,3 AND A. WOLFE'

Departments ofAntiviral Research,' Virus and Cell Biology,2 and Medicinal Chemistry,yMerck Research Laboratories, West Point, Pennsylvania 19486-0004

Received 15 June 1994/Returned for modification 2 August 1994/Accepted 3 October 1994

Synthesis of influenza virus mRNA is primed by capped and methylated (cap 1, m7GpppXm) RNAs whichthe virus derives by endonucleolytic cleavage from RNA polymerase II transcripts in host cells. The conservednature of the endonucleolytic processing provides a unique target for the development of antiviral agents forinfluenza viruses. A series of 4-substituted 2,4-dioxobutanoic acid compounds has been identified as selectiveinhibitors of this activity in both influenza A and B viruses. These inhibitors exhibited 50%o inhibitoryconcentrations in the range of 0.2 to 29.0 ,uM for cap-dependent influenza virus transcription and had no effecton the activity of other viral and cellular polymerases when tested at 100- to 500-fold higher concentrations.The compounds did not inhibit the initiation or elongation of influenza virus mRNA synthesis but specificallyinhibited the cleavage of capped RNAs by the influenza virus endonuclease and were not inhibitory to theactivities of other nucleases. Additionally, the compounds specifically inhibited replication of influenza A andB viruses in cell culture with potencies comparable to the 50% inhibitory concentrations obtained fortranscription.

Influenza A and B viruses cause an acute respiratory infec-tion in humans which can progress to severe pulmonaryinfection, resulting in considerable mortality and morbidity inelderly and other high-risk individuals. Because currentlyavailable vaccines are not completely protective against influ-enza virus infection, there is interest in the development ofanti-influenza virus therapeutic agents. Presently, the onlylicensed antiviral agents for influenza virus are amantadine andrimantadine, which are of limited utility because of theirineffectiveness against influenza B viruses and because theygenerate transmissible resistant mutants (14).

Influenza virus is a negative-strand RNA virus with a

segmented genome. Synthesis of influenza virus mRNA occursin the nucleus of infected host cells and is catalyzed by a virallyencoded polymerase complex consisting of three polymeraseproteins, PB1, PB2, and PA (21). Influenza virus transcriptionis initiated by a novel mechanism in which capped andmethylated RNAs, acquired from RNA polymerase II tran-scripts of host cells, are used to prime the synthesis of mRNA.The primers are derived by a virion-encoded endonucleasewhich cleaves the capped cellular RNAs 13 nucleotides (nt)from the 5' end, yielding a 13-nt capped RNA with a 3' OHterminus. The nucleotide complementary to the penultimatebase in the 3' end of the template viral RNA is then addedonto the 3' end of the primer, and the elongation of mRNAchains then proceeds (28).

Primary transcription of influenza virus is an attractiveantiviral target for several reasons. The transcriptase is essen-tial for virus replication in mammalian cells (27). It is highlyconserved in influenza A and B viruses (40), and cap-depen-

* Corresponding author.t Corresponding author, chemistry.t Present address: Pharmacopeia, Princeton, NJ 08540.

dent cleavage appears to be a unique property of influenzavirus, having no known cellular counterpart. Only Bunyavi-ruses, also negative-strand, segmented RNA viruses, have beenshown to possess a similar cap-dependent endonuclease activ-ity. Bunyavirus transcription differs, however, in that Bunyavi-rus mRNAs are synthesized in the cytoplasm, and therefore,transcription is primed with a stable pool of capped RNAtranscripts (31).Although our understanding of the function and interaction

of the polymerase proteins in the transcriptase complex islimited, the activities associated with the PB1 and PB2 proteinshave been described on the basis of genetic and biochemicalstudies. Polymerase protein PB1 is required for the incorpo-ration of the first nucleotide onto the capped RNA primersand is also involved in the elongation of mRNA chains (5, 17,29). The PB2 protein binds to the capped primer and may alsobe responsible for the endonuclease activity (5, 25, 37). Severalpolymerase motifs have been identified in the PB1 genes ofinfluenza A and B viruses, and these motifs are conservedamong RNA-dependent polymerases of viral and retroid ori-gin, including two motifs which are highly conserved amongDNA and RNA polymerases of procaryotes and eucaryotes (9,19). It has also been suggested that homology exists betweenthe primary sequence of eucaryotic cap-binding proteins andthe influenza virus cap binding protein, PB2; however, thedegree of conservation is far less extensive (8). Therefore,although the homology shared by PB1 and other polymerasesindicates that elongation of mRNA may not be an activityspecific to influenza virus, the endonucleolytic processing ofcapped RNA primers appears to be a unique viral function.

Cap-dependent transcription of influenza virus can be per-formed in vitro (4). This is a multistep process that includes thebinding and endonucleolytic cleavage of capped RNA primers,the initiation of RNA polymerization, and the subsequent

2827

Vol. 38, No. 12

on October 19, 2017 by guest

http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/

ANTIMICROB. AGENTS CHEMOTHER.

elongation of mRNA complementary to the viral templates.Using such an assay, we identified a series of compounds,4-substituted 2,4-dioxobutanoic acids, which are inhibitors ofthe cap-dependent endonuclease activity of influenza virus.The specific antiviral activities of these inhibitors are de-scribed.

MATERIALS AND METHODS

Enzymes and reagents. Guanylyltransferase was purchasedfrom Bethesda Research Laboratories (Gaithersburg, Md.),S-adenosyl methionine (SAM) was from Boehringer Mann-heim (Indianapolis, Ind.), RNases U2 and Ti were fromPharmacia (Piscataway, N.J.), and bovine pancreatic A RNasewas from Sigma (St. Louis, Mo.). Nucleoside-2'-methyltrans-ferase was purified from vaccinia virus virions which werekindly supplied by R. Bablanian (Health Science Center, StateUniversity of New York, Brooklyn) by the method of Barbosaand Moss (2).Chemical synthesis of influenza virus transcriptase inhibi-

tors. The dioxobutanoic acids were all prepared from thecorresponding methyl ketones by reaction with dimethyl ox-alate and sodium hydride and then hydrolysis of the interme-diate methyl dioxobutanoic ester (11, 38). Compounds 4 to 8(Tables 1 and 2) were prepared from commercially availablemethyl ketones by the method described above. The synthesisof compound 3 (L-735,882; 4-[N-benzyl-3-(4-chlorobenzyl)pi-peridin-3-yl]-2,4-dioxobutanoic acid) was accomplished bytreatment of ethyl-N-(tert-butyloxycarbonyl)-3-piperidine car-boxylate with lithium bis(trimethylsilyl amide) in tetrahydrofu-ran and then 4-chlorobenzyl chloride, which gave rise to ethyl-3-(4-chlorobenzyl)-N-(tert-butyloxycarbonyl)-3-piperidine car-boxylate. Hydrolysis of the ester (10 N NaOH) gave thecarboxylic acid, which was converted to a methyl ketone by thefollowing sequence: (i) acid chloride formation with oxalylchloride, (ii) N,O-dimethylhydroxamide formation with N,O-dimethylhydroxyl amine, and (iii) reaction with methylmagne-sium bromide at 60°C. The tert-butyloxycarbonyl protectinggroup was removed with gaseous hydrochloric acid to give thepiperidine amine, 3-acetyl-3-(4-chlorobenzyl), which was thenalkylated with benzyl chloride to form 3-acetyl-N-benzyl-3-(4-chlorobenzyl)piperidine. Treatment of this methyl ketone withdimethyl oxalate and sodium hydride produced the corre-sponding methyl dioxobutanoic ester, which was converted tocompound 3 by acid hydrolysis. Compound 1 was prepared bya similar route except that the 3-acetyl-3-(4-chlorobenzyl)pi-peridine was sulfonylated with phenylmethylsulfonyl chloridebefore the oxalation sequence. Compound 2 was prepared by asimilar route except that preparation was started with ethyl-N-(tert-butyloxycarbonyl)-4-piperidine carboxylate and alkylationof the piperidine nitrogen with bromomethyl cyclohexanebefore oxalation. Compound 10 was prepared by treatment ofthe corresponding methyl ester with ammonia in methanol.Compounds 11 to 15 were commercially available. Compound16 was prepared from methyl phenyl sulfone by treatment withsodium hydride and dimethyl oxalate and then hydrolysis withaqueous sodium hydroxide. Compound 17 was prepared by themethod of Horne et al. (16) from 2-formyl benzofuran by theHorner-Emmons reaction. Compound 18 was prepared bytreatment of 2-methylbenzothiazole with sodium hydride anddimethyl oxalate and then hydrolysis with aqueous sodiumhydroxide (1, 39). Compounds 19 and 20 were prepared bysimilar procedures starting from the appropriate 2-methylheterocycles. The nuclear magnetic resonance spectra of allcompounds synthesized are consistent with their assignedstructures. All compounds gave satisfactory combustion anal-

ysis results and were resuspended in 100% dimethyl sulfoxide(Aldrich, Milwaukee, Wis.) for testing.

Cells and viruses. Baby hamster kidney (BHK-21) andMadin-Darby canine kidney (MDCK) cells were obtained fromthe American Type Culture Collection. Influenza viruses (AlJapan/305/57, A/Port Chalmers/21/73, and B/HK/5/72), murineencephalitis virus (MEV) Theiler's, and New Castle diseasevirus (NDV) New Jersey-Roakin (1946) were also purchasedfrom the American Type Culture Collection. Vesicular stoma-titis virus (VSV) Mudd Summers was obtained from J. Condra(Merck). Influenza virus A/Pr/8/34 was kindly supplied by P.Palese (Mt. Sinai School of Medicine, New York, N.Y.),A/Hong Kong/8/68 was supplied by A. Friedman (Merck), andA/WSN was supplied by R. Krug (Rutgers University). LaCross virus (LACV) was generously provided by N. Nathan-son, University of Pennsylvania.

Purification of virus and polymerase cores. The influenzavirus strains used in transcription reactions were grown in11-day-old embryonated eggs as described by Barrett andInglis (3). Virus containing allantoic fluid was aspirated fromeggs and was clarified by centrifugation at 10,000 x g for 20min at 4°C. Virions were purified from the fluid by ultracen-trifugation at 50,000 x g for 2.25 h at 4°C. Virus wasresuspended in 10 mM Tris-Cl (pH 7.4)-O.1 M NaCl-1 mMEDTA at a concentration of 1 mg of viral protein per ml andwas stored at -70°C. Influenza virus polymerase cores werepurified essentially as described by Honda et al. (15).

Preparation of ALMV substrates. Alfalfa mosaic virus(ALMV) segment 4 RNA containing a cap 0 structure waspurchased from J. Bol, Leiden, The Netherlands. The 5' cap ofALMV RNA was methylated with nucleoside-2'-O-methyl-transferase in a reaction mixture containing 25 mM Tris-Cl(pH 7.5), 2 mM dithiothreitol (DTT), 2 mM MgCl2, 100 ,uMS-adenosylmethionine, 2.3 ,uM ALMV RNA, and 100 RI ofpurified methylase per ml for 35 min at 37°C. Radiolabeled cap1 ALMV substrate was prepared by recapping ALMV RNAfrom which the cap had been removed by oxidation and thenP-elimination of the temrinal m7G according to Frankel-Conrat and Steinschneider (12) by incorporation of[32P]GTP onto the 5' end of ALMV RNA in a reaction mix-ture containing 25 mM Tris-Cl (pH 7.5), 2 mM MgCl2, 30,uM GTP, 10 ,uM [oc-32P]GTP, 500 ,uM SAM, 1.6 U of RNasinper ,ul, 0.35 ,uM decapped ALMV RNA, 0.15 U of guanylyl-transferase per ,ul, and 0.1 ,ul of methyltransferase per ,ul for 45min at 37°C. Synthetic substrates were prepared by cappinggel-purified RNA transcripts as described above; the tran-scripts were synthesized by the T7 duplex promoter method ofMilligan et al. (24).

In vitro influenza virus transcriptase assays. In vitro influ-enza virus transcriptase assays with detergent-disrupted virionsor purified polymerase cores were performed by a modificationof a previously described procedure (4). Test compounds wereincubated for 60 min at 31°C with 20 ng of purified influenzavirus per RI in a final reaction mixture containing 100 mMTris-HCl (pH 7.8), 0.25% Triton N-101, 100 mM KCl, 5 mMMgCl2, 1 mM DTT, 2 ng of tRNA per ,u, 100 ,uM ATP, 50 ,uM(each) CTP and GTP, 1 jiM UTP, 0.3 ,iM [35S]UTP, and 7 nMsubstrate ALMV capped primer of 880 nt or 200 jiM adenyl-yl(3'-5')guanosine (ApG). Transcriptions performed with 2 ngof purified polymerase cores per RI were similar, except that 90mM HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonicacid; pH 7.3) replaced Tris-HCl, the Triton N-101 concentra-tion was reduced to 0.05%, and KCl was reduced to 80 mM.The reaction product was quantitated by trichloroacetic acid(TCA) precipitation on glass fiber filters; this was followed byliquid scintillation counting. For 13-, 22-, and 70-nt primed

2828 TOMASSINI ET AL.


http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/

INHIBITION OF INFLUENZA VIRUS 2829

TABLE 1. Transcriptase inhibition by selected compoundsa

Coapound Structure ICso (M)b

0 OH

OH

=0

C1

0.19 + 0.04

0.33 + 0.02

1.1 + 0.23, L-735,882

O OH

R

R = PhR = H

OH

O

0

OH

Me 0

OH

0

3.7 + 0.221.3 + 0.8

29.4 + 1.0

56.7 + 1.5

307.0 + 4.6

a Compounds were tested in an influenza virus in vitro transcription assay primed with cap 1 ALMV as described in Materials and Methods.b IC50s are geometric means ± standard errors of at least three determinations.

transcription assays, test compounds were tested as describedabove with purified polymerase cores and 10 nM syntheticALMV capped primers for 40 min. The 50% inhibitoryconcentrations (IC50s) were determined from the geometricmeans of separate determinations analyzed by four-parameterfit function and plotted as percent inhibition versus concentra-tion by the method of DeLean et al. (10). Geometric meansrather than arithmetic means were calculated because the dataare geometrically scaled (i.e., log-scaled inhibitor concentra-tions). The Michaelis-Menten constant (Kin) and maximumvelocity (Vm.) were determined by fitting the data from the

initial velocity versus substrate concentration curves to theMichaelis-Menten equation (velocity form; hyperbolic).Other polymerase assays. For VSV transcription, com-

pounds were incubated for 60 min at 31°C in a final reactionmixture containing 50 mM Tris-HCl (pH 8.0), 100 mM NaCl,4 mM DTT, 0.05% Triton N-100, 5 mM MgCl2, 2 ng of tRNAper ,ul, 100 ,uM ATP, 50 ,uM (each) CTP and GTP, 1 puM UTP,0.3 ,uM [35S]UTP, and 20 ng of VSV per ,ul. The reactionproduct was quantitated by TCA precipitation onto glass fiberfilters as described above; this was followed by liquid scintilla-tion counting. For the HeLa cell RNA polymerase II assay,

1

2

45

6

7

8

VOL. 38, 1994


http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/


TABLE 2. Effect of side chain modifications on transcriptase inhibition

Structure

O OH

N ~ONe-0

O OH

o)WNNHz2

0

O X

OCOHX = OHX = NH2X = H

O H

A0

0

HOH0

OH

11

I 0}

IC50 (gm)a

>500

>500

>500>500>500

>500

>500

>500

x y

0sCC \>+OH 0

S

NMe0

CHNNN

>500>500>500_100

a IC50s were determined in the influenza virus in vitro transcription assay (see footnotes a and b of Table 1).

compounds were incubated in a runoff transcription assay by aprocedure similar to that described by Manley (22). HeLa cellextract (0.25 pA), made from whole-cell lysates kindly providedby P. Huang (Merck) and 40 nM pD5 template containing theadenovirus major late promoter (pD5 plasmid from R.LaFemina, Merck), were reacted in a final reaction mixturecontaining 15 mM Tris-HCl (pH 7.9), 7.0 mM MgCl2, 32 mM(NH4)2SO4, 0.2 mM EDTA, 1.3 mM DTI, 500 ,uM (each)ATP, CTP, and GTP, and 0.5 ,uM [32P]UTP for 60 min at 30°C.Following ethanol precipitation, the reaction products wereelectrophoresed on 8% polyacrylamide gels containing 7 Murea and were quantitated by direct radioanalytic imaging on aphosphorimager (Molecular Dynamics, Sunnyvale, Calif.). Hu-man immunodeficiency virus (HIV) reverse transcriptase and

HeLa cell DNA polymerase ot assays were performed asdescribed by Goldman et al. (13) and Copeland and Wang (7),respectively.

Influenza virus cleavage assay. The compounds were testedin a final reaction mixture containing 50 mM Tris-HCl (pH7.8), 100mM KCl, 5 mM MgCl2, 1 mM DTT, 1 ng of tRNA per,u, 0.2 U of RNasin per ,u, 5.5 nM 32P-radiolabeled ALMV,and 2.5 ng of purified polymerase cores per RI for 30 min at310C. The reaction was stopped by adding an equal volume of95% formamide buffer, and then the reaction products wereelectrophoresed on 12% polyacrylamide gels containing 7 Murea; this was followed by quantitation on the phosphorimager.

Other nuclease assays. Radiolabeled ALMV substrate (5.5nM), prepared as described above, was incubated with 0.02 U

Coound

9

10

111213

14

15

16

17181920



http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/


of RNase Ti per pl or 0.04 U of RNase U2 per ml in 10 mMsodium citrate (pH 5.0)-0.5 mM EDTA-7 M urea at 56°C for20 min. RNase A at 4 ng/,ul was incubated with 5.5 nMradiolabeled ALMV substrate in 10 mM sodium citrate (pH3.5)-0.5 mM EDTA-7 M urea at 30°C for 20 min. The reactionproducts were electrophoresed on 12% polyacrylamide gelscontaining 7 M urea and were quantitated by phosphorimaging. HIV RNase H activity was assayed as described byGoldman et al. (13). The EcoRI assay was performed by theprocedure of Jeltsch et al. (18).

In vitro cell culture replication assays. Subconfluent mono-layers of BHK-21 cells grown in high-glucose (4.5 g/liter)Dulbecco's modified eagle medium (DMEM) supplementedwith 10% heat-inactivated fetal bovine serum (FBS)-2 mML-glutamine-50 U of penicillin per ml-50 ,ug of streptomycinper ml-25 mM HEPES in 48-well tissue culture clusters(Costar) were washed and fed with DMEM containing 20 ,uMmethionine. Test compounds were added at various concen-trations to a final solvent concentration of 0.5%, and themixture was incubated for 1 h at 37°C with gentle rocking.Plates were chilled on wet ice for 5 min prior to the addition oftest virus at a multiplicity of infection of 1.0 PFU per cell andwere then incubated for 1 h at 4°C to synchronize the infection.Plates were transferred to 37°C, and at various times postin-fection (2.5 h for A/WSN and VSV; 4.5 h for A/Pr/8/34, LACV,and MEV; 5.5 h for B/HK/5/72; 6 h for NDV) cultures werelabeled for 1 h with 50 ,uCi of [35S]methionine per ml. Labeledcultures were washed, harvested with Tricine sample buffer(Novex, San Diego, Calif.) containing 40 mM DTT, anddisrupted by freeze-thaw lysis and sonication. Lysates wereresolved by sodium dodecyl sulfate (SDS)-polyacrylamide gelelectrophoresis on 10% Tricine gels (Novex) and were pro-cessed for fluorography with En3Hance (DuPont NEN, Bos-ton, Mass.).The radioactivities of the virus-specific bands (previously

identified on the basis of immunological reactivity and infec-tivity time course studies) were quantitated with a radioana-lytic device (Ambis, San Diego, Calif.). To compensate forvariation in sample preparation and loading, the counts werenormalized by determining the total counts in the lanes andwere expressed as a ratio of viral to total counts (R). A controlratio (Rc) was calculated from the solvent-treated infectedcontrol by the same method. The background ratio (Rb) wasobtained by determining the counts in the mock-infectedcontrol at the lane position corresponding to where viral bandsmigrated in the adjacent control lane divided by the totalcounts in the lanes. The antiviral activity was then calculated bythe following formula: 1 - [(R, - Rb)I(RC - Rb)] X 100. IC50swere extrapolated from the resulting sample concentrationversus percentage activity plot.

Influenza virus plaque inhibition and virus yield assays.The influenza virus plaque inhibition assay was performed onMDCK cells by the previously published procedure of Tobitaet al. (36) with the following modifications. MDCK cells,trypsinized with 0.05% trypsin-0.53 mM EDTA (GIBCO),were seeded 2 to 3 days prior to use into 12-well tissue cultureplates in high-glucose (4.5 g/liter) DMEM (GIBCO) supple-mented with 10% heat-inactivated FBS-2 mM L-glutamine-25mM HEPES-50 U of penicillin per ml-50 pLg of streptomycinper ml. Confluent monolayers of cells were washed withDMEM (minus serum), infected with 50 PFU of A/Pr/8/34 per0.2 ml of DMEM containing 1% bovine serum albumin (BSA;35% solution; ICN Biomedicals, Costa Mesa, Calif.), andadsorbed for 30 minat 37°C. The innoculum was aspirated andthe cells were overlaid with 1 ml (per well) of test compounddiluted in 2x minimal essential medium (MEM) overlay (2x

MEM [GIBCO], 2x Basal Medium Eagle [BME] vitamins[GIBCO], 50 mM HEPES, 100 U of penicillin per ml, 100 ,ugof streptomycin per ml, and 50 Na-benzoyl-L-arginine ethylester [BAEE] units of twice-crystallized, virus- and mycoplas-ma-free [TRLVMF] trypsin [Worthington, Freehold, N.J.] perml) and mixed 1:1 with melted 1.5% Seakem ME agarose(FMC, Rockland, Maine). At 24 h postinfection, 1.0 ml ofagarose-2x MEM overlay containing 0.003% GIBCO neutralred (prepared by dilution of neutral red stock 1:50 in 2x MEMoverlay and then mixing 1:1 with 1.5% melted agarose imme-diately prior to overlaying) was added. Incubation at 37°C wascontinued and the plaques were counted on the next day (ca.42 h postinfection).The virus yield assay was done in 12-well plates simulta-

neously with the plaque inhibition assay or in 96-well platesin parallel with the [3-(4,5-dimethlythiazol-2-yl)-5-(3-carboxy-methoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium inner salt)](MTS) colorimetric assay (see below) on MDCK cells thatwere plated and washed as described above. For the assay in12-well plates, compound was diluted in 2x MEM (as de-scribed above) and mixed 1:1 with H20, and then 1 ml per wellwas overlaid following virus adsorption as described above. Forthe 96-well plate format, 50 ,l (per well) of compound dilutedinto DMEM plus 1% BSA at the indicated concentrations wasadded to cells for 30 min at 37°C prior to the addition of 5 RIof virus at 10 PFU per pl. Following virus adsorption for 30min at 37°C, the virus innoculum was removed and 100 pi (perwell) of compound diluted in 1x MEM overlay was added. At24 h postinfection, the supernatants from either 12- or 96-wellplates were harvested and titrated in a plaque assay similar tothe plaque inhibition assay, but in the absence of compound.

Cytotoxicity assays. For the [3H]uridine uptake assay,BHK-21 cells were grown as described above for the in vitrocell culture replication assay. Cultures were washed and refedwith serum-free DMEM, prior to the addition of compound ordimethyl sulfoxide to a final solvent concentration of 0.5% totriplicate wells, and were then incubated as described above forthe replication assay. [3H]uridine was added to a final concen-tration of 20 ,uCi/ml at the time of virus infection, and thecultures were labelled for 1 h at 37°C. Monolayers were washedwith phosphate-buffered saline (pH 7.2) and then harvestedwith 300 ptl of 4% SDS-40mM DTT per 0.5-cm2 well and weredisrupted by freeze-thaw lysis. Lysates were heated at 95°C forS min, and a 20-,ul aliquot was precipitated in 100 ,ul of 10 mMEDTA containing 100 ,ug of tRNA per ml and 10% TCA.Precipitates were collected by vacuum onto filter disks andwere counted with scintillant for 1 min.The effect of a compound upon cell proliferation was

measured with a nonradioactive cell proliferation assay kit,Cell Titer 96 AQueous (Promega, Madison, Wis.), which is aquantitative MTS colorimetric assay. Compound or solventwas added in duplicate to MDCK cell monolayers grown in96-well plates as described above for the virus yield assay for a24-h period or was added to cells plated at 50% confluency ingrowth medium (DMEM plus 10% FBS) for a 48-h period. Atthe indicated times, 20 ,ul of MTS at 333 p,g/ml in 25 mMphenazine methosulfate was added to wells containing 100 RIof medium, and the plates were then incubated at 37°C in 5%CO2 for 45 min and the A490 was read by using a MolecularDynamics (UV max) plate reader. Cell viability is expressed asthe percentage of control optical density for compound-treatedcells relative to the optical density of solvent-treated controlcells. Cell counts for solvent-treated controls were determinedin replicate with a hemocytometer at 0, 24, and 48 h in eachexperiment. The assay was linear (correlation coefficient,0.9942) for the number of MDCK cells and A490.

VOL. 38, 1994


http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/


RESULTS

Inhibition of influenza virus transcriptase activity. Com-pounds of the 4-substituted 2,4-dioxobutanoic acid series wereinitially identified as inhibitors of influenza virus cap-depen-dent transcription by random screening of the Merck samplerepository by using an influenza A virus in vitro transcriptaseassay primed with capped ALMV RNA (880 nt). The com-

pounds, cyclic or multicyclic structures containing a 2,4-di-oxobutanoic acid side chain, exhibited IC50s ranging from 0.2to 29.0 ,uM (Table 1). Testing of several structurally relatedanalogs revealed that the 2,4-dioxobutanoic side chain wasessential for the inhibitory activity, although compounds withthe side chain alone were significantly less potent than com-pounds substituted with lipophilic groups (compare com-pounds 7 and 8 with compounds 1 to 6; Table 1). Modificationof the side chain further showed that the carboxylic acid anddiketo groups were all required for activity (Table 2). Derivat-ization of the carboxylic acids of compounds 4 and 5 to amethyl ester or an amide (compounds 9 and 10, respectively)abolished activity, as did substitution of the ax-carbonyl groupof compound 5 with an alcohol, amine, methylene, or olefingroup (compounds 11 to 14, respectively) and the -y-carbonylwith methylene, sulfone, furan, thiazole, or imidazole groups(compounds 15 to 19, respectively). Although complete loss ofactivity was not observed with an oxazole group substitution atthe -y-carbonyl position, compound potency was reduced ap-proximately 30-fold (compound 20).

Substitutions were made in the lipophilic portion of themolecule, and it was found that placement of alkyl groupsproximal to the side chain resulted in improved potency. Manydioxobutanoic acid compounds with saturated cycloalkyl sub-stitutions were synthesized, and in particular, a class of piper-idine substituted 2,4-dioxobutanoic acids was found to be themost potent; examples of these compounds appear in Table 1(compounds 1 to 3; IC50s, 0.2 to 1.1 ,uM). Further structuralmodification of the piperidine nitrogen by benzyl and alkylsubstitutions yielded compounds which had antiviral activity incell culture. Two of the more active of these were compounds2 and 3 (L-735,882).

It should be noted that compound 4 and various mono- anddiacidic molecules with lipophilic substituents have previouslybeen described to be competitive inhibitors of porcine glyco-late oxidase (GAO) (30, 38). However, several reported inhib-itors of GAO had no effect on influenza virus transcriptionwhen the inhibitors were tested at up to 100-fold above theGAO IC50s (data not shown). The alkyl-substituted dioxobu-tanoic acids (compounds 1 to 3) were not tested in a GAOassay.

Specificities of transcriptase inhibitors. Several of the mostpotent compounds developed in the dioxobutanoic acid series(see above) were chosen for characterization, including com-pound 3, L-735,882, a benzyl-substituted piperidine, whichexhibited antiviral activity in cell culture (see below) and thusis described further. The inhibitory profile obtained forL-735,882 against several enzymes with similar activities(which was analogous to the inhibitory specificities elicited byall other derivatives of the series tested in the assays; data notshown) demonstrated that these inhibitors were highly selec-tive for the influenza yirus transcriptase. The compound was

specific for the influenza virus polymerase and had no effect inother polymerase assays including VSV transcription, humanimmunodeficiency virus (HIV) reverse transcriptase, T7 phage,HeLa cell RNA polymerase II, and HeLa cell DNA poly-merase ao when tested up to 100- to 500-fold above the IC50(Table 3).

TABLE 3. Biochemical specificity of L-735,882a

Polymerase or nuclease IC50 (>M)

PolymeraseALMV (880-nt)-primed influenza A

transcription ....... ...................... 1.1 (÷0.2)bApG-primed influenza A transcription ................... >1,000.O"bVSV transcription ............................. >500.0bHIV reverse transcriptase ............................. >300.0fT7 Phage RNA polymerase ............................. >100.0lHeLa RNA polymerase II ............................. >500.0bHeLa DNA polymerase a ................... .......... >100.0l

NucleaseInfluenza endonuclease.............................................HIV RNase H............................................................U. sphaerogena RNase U2........................................A. oryzae RNase Ti...................................................Bovine pancreatic RNase A.....................................EcoRI restriction endonuclease...............................

1.8 (±O.1)d>100.0 d

>200.Od>200.Od>200.Od>500.Oc

" Compound L-735,882 was tested in biochemical assays at concentrations inthe range of 100 to 500 ,uM as described in Materials and Methods.

b IC50 is the geometric mean (± standard error) of minimally three determi-nations unless otherwise noted; the standard error for all other values was 0.

c Single determination.d The IC50s are averages of two determinations unless otherwise noted.

Compound L-735,882 specifically inhibited cap-dependentinfluenza virus transcription and did not inhibit transcriptionwhen primed cap independently with the dinucleotide ApG atconcentrations of compound up to 1,000-fold greater than theIC50 for ALMV-primed transcription (Table 3). Furthermore,L-735,882 showed activity in a cap-dependent primed tran-scription comparable to those against several other influenza Aand B virus polymerases (Table 4). Additionally, inhibition oftranscription was similar with either detergent-disrupted viri-ons or purified polymerase cores as the enzyme source (datanot shown).

Inhibition of influenza virus cleavage. The observation thatthe 4-substituted dioxobutanoic acids inhibited cap-dependentbut not cap-independent priming of influenza virus transcrip-tion raised the possibility that these compounds affected theendonucleolytic processing of capped primer substrate. There-fore, inhibitor L-735,882 was tested in a direct influenza viruscleavage assay by using an ALMV RNA substrate radiolabeledby incorporation of [32P]GTP into the 5' cap. The compoundinhibited cleavage of the 32P-end-labeled ALMV (Fig. 1 andTable 3) with an IC50 of 1.8 ,uM, which was comparable to theIC50 obtained in the transcription assay (1.1 ,uM). The IC50sfor transcription and cleavage were also similar to those ofother derivatives of this series. The inhibitor L-735,882 had noeffect upon other nucleases including HIV RNase H, RNaseU2 (Ustilago sphaerogena), RNase Ti (Aspergillus oryzae),

TABLE 4. Inhibition of influenza A and B virustranscription with L-735,882a

Virus strain IC50 (PWM)bA/PR/8/34 (HlNl) .............. .............................. 1.10A/Japan/305/57 (H2N2) ............................................ 0.25A/Port Chalmers/1/73 (H3N2) ............................................ 0.50A/Hong Kong/8/68 (H3N2) ............................................ 0.62B/Hong Kong/5/72 .............. .............................. 0.85

a Compound L-735,882 was tested by in vitro transcription with variousinfluenza viruses primed with cap 1 ALMV as described in Materials andMethods.

b Single determination for all viruses tested within the same assay.



http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/


100

80

C

0

-0ol

60

40

20

0v

-200.01 0.1 1 10 100 1000

PM 735,822FIG. 1. Inhibition of influenza virus endonuclease activity by L-735,882. (Inset) Cleavage of 32P-labeled ALMV substrate by influenza virus

polymerase as described in Materials and Methods in the presence of L-735,882 at 0, 333.0, 111.0, 37.0, 12.0, 4.0, 1.4, 0.5, 0.15, and 0.05 ,uM (lanes1 to 10, respectively).

bovine pancreatic RNase A, and EcoRI restriction endonucle-ase (Table 3) when tested at concentrations 100- to 500-foldgreater than the IC50 for cleavage. Specific inhibition of theinfluenza virus endonuclease was also demonstrated by otherderivatives of the series in these assays (data not shown).

Uncoupling of cleavage and elongation. Compound L-735,882was also tested in a transcriptase assay in which endonucleaseand elongation reactions were uncoupled. Capped RNA sub-strates of 13, 22, and 70 nt derived from the sequence of the 5'end of the ALMV RNA were synthesized by T7 runofftranscription by the duplex promoter method of Milligan et al.(24). The 13-nt substrate represented the authentic endonu-cleolytic cleavage product of the 880-nt ALMV substrate andwas not cleaved by influenza virus endonuclease when radio-labeled at the 5' end, while the two substrates of 22 and 70 ntcontained the cleavage site present in ALMV RNA and werecleaved by influenza virus endonuclease. These syntheticRNAs were used as primers in influenza virus transcription.Only transcription reactions primed with the 22- and 70-ntsubstrates, which undergo endonucleolytic processing, wereinhibited by L-735,882, with IC50s similar to those obtained fortranscription primed with the 880-nt ALMV substrate (Fig. 2).Furthermore, L-735,882 had no effect on the addition ofradiolabeled GTP onto the 3' OH terminus of the 13-nt primer(showing that the compounds did not inhibit the initiation ofelongation), but did inhibit the addition of [32P]GTP onto the22-, 70-, and 880-nt capped substrates which undergo endonu-cleolytic processing (data not shown). Similar activity profilesin these reactions were demonstrated for several derivatives inthe series. In contrast, several lignin and polyphenolic com-pounds (identified as nonspecific influenza virus transcriptaseinhibitors in a random screening effort) and the PP1 analogphosphonoformic acid (PFA) inhibited transcription primed

with all four capped RNA substrates but had no effect oncleavage (unpublished data).

Substrate competition. Compound L-735,882 was not com-petitive with respect to the cap 1 substrate ALMV RNA (880nt), as shown in Fig. 3. Increasing amounts of the substrate (upto 1 mM) did not alter the degree of inhibition caused byL-735,882. The Km of the substrate was relatively unchangedfrom 6.1 nM for the uninhibited reaction to 9.7 nM for theinhibited reaction, while Vm. was substantially decreased from2.4 to 0.71 pmol of product in 60 min. This was in contrast to

100

80

z

0

m

zI

60

40

20

0

-200.01 0.1 1

L-735,882 [gM]10 100

FIG. 2. Inhibition of in vitro transcription primed with cappedALMV RNA substrates (13, 22, 70, and 880 nt) by using purifiedpolymerase cores as the enzyme source.

Iiiiil iiiiiil iI I iiil l

VOL. 38, 1994


http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/


TABLE 5. Effect of L-735,882 in influenza virus yieldand plaque inhibition assays

L-735,882 Virus yield assay' Plaque inhibition assaybconcn(>M) Virus titer/ml % Control % Control Titer/plaquec

0 1.25 X 108 100 100 1.1 X 10625 2.00 x 107 16 88 2.0 x 10550 2.70 X 105 0.22 53 4.0 x 103100 4.40 x 103 0.0035 34 1.0 x 102

a virus titers of cell supernatants were determined at 24 h postinfection by0 : t .....counting the plaques on MDCK cells as described in Materials and Methods.

Virus titer and percent control were averaged from duplicate samples; variationaveraged 18%.

5 '_.,_,_ ,_ ,_,_.,______ bPlaque number was determined at 48 h postinfection as described in-200 0 200 400 600 800 1000 1200 Materials and Methods. Percent control was averaged from duplicate samples;

variation averaged 17%.nM ALMV c Virus titer obtained at 24 h postinfection in the virus yield assay per total

FIG. 3. Competition of capped dinucleotide m7GpppGm andL-735,882 with capped ALMV RNA substrate (880 nt) by in vitrotranscription with purified polymerase cores. A, uninhibited reaction;0, 250 P.M m7GpppGm; 0, 6 p.M L-735,882.

the inhibition of transcription by 250 puM cap dinucleotidem7GpppGm, a known inhibitor of cap-dependent transcription(20), in which the observed inhibition was overcome by in-creasing amounts of the ALMV substrate. At high concentra-tions of the ALMV RNA substrate, the reaction inhibited bythe cap dinucleotide reached a Vm, of 2.4 pmol of product in60 min, and the Km of the substrate was increased to 95 nM.Therefore, unlike the cap 1 analog, L-735,882 was not com-petitive with capped substrate. Related dioxobutanoic acidcompounds were also tested in mammalian and viral caprecognition assays and were not inhibitory to in vitro transla-tion of mRNA in rabbit reticulocytes and vaccinia virusmethylation of capped RNA when tested at concentrations ashigh as 1.5 mM. Additionally, the compounds were not com-petitive with nucleoside triphosphate substrates in transcrip-tion (data not shown).

Effect on viral replication in vitro. The antiviral activities of

AAntiviral Activity of L-735,882

number of plaques.

the compounds were evaluated in a cell culture assay forinfluenza viruses in which the effects of the compounds on asingle cycle of replication were quantitated by measurement ofviral protein synthesis. As shown in Fig. 4, L-735,882 inhibitedthe synthesis of 35S-radiolabeled influenza A and B virusproteins by 50% at concentrations of 6 and 2 ,uM, respectively,but had no effect on the synthesis of LACV (a Bunyavirus)proteins when it was tested up to 30 puM. Additionally, thecompound had no effect on the replication of VSV, NDV, orMEV when it was tested in a similar assay. L-735,882 at aconcentration of 30 p,M had no effect on cellular RNA orprotein synthesis during the time period of the assay (data notshown). Testing of other derivatives in the class demonstratedthat the IC50s in cell culture were comparable to the potenciesobtained for transcriptional activity. For example, compound 2had an IC50 of 0.3 ,uM for in vitro transcription (Table 1) andan IC50 of 1 ,uM in cell culture (data not shown).

Effect of compound in multicycle cell culture assays. Com-pound L-735,882 inhibited the multicycle replication of influ-enza A virus in MDCK cells, as demonstrated in both the

BLACV M PR8

C0 C) 0 0 jiMVirus Activity IC50 (PM)

+ 6.0

6.0

2.0

nda

ndb

ndb

ndb

FIG. 4. Antiviral activity of compound L-735,882 in a replication assay in BHK-21 cells infected with various viruses. Compound was added atthe concentrations indicated in the presence of [35S]methionine as described in Materials and Methods. Cells were lysed and electrophoresed onSDS-10% polyacrylamide gels. nd, not determined; ano antiviral activity at 100 puM; bno antiviral activity at 30 ,uM; arrow, virus-specific band (Ml).

2.5

2.0

1.5

1.0

0.5

0

0._

-JCwLL

A/WSN/33A/Pr/8/34B/HK/5/72VsVNDV

MEV

LACV

AA A

A

ANTiMICROB. AGENT'S CHEMOT14ER.

O.(

-O.'


http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/


04

0

woE-z0

80

60

40

20

0.01 0.1 1

iM L-735,882

FIG. 5. Effect of L-735,882 on virus yield andassays in MDCK cells. The compound was addedconcentrations to MDCK cells as described in Mate]for the virus yield and cell proliferation assays. 0, p4

control titer obtained in the virus yield assay from cc24 h postinfection with influenza virus AIPr/8/34 by r

on MDCK cells in duplicate. The percentage of ticontrol optical density obtained in the MTS cell proluninfected MDCK cells in growth medium at 24 h (Ifrom duplicate samples was measured. Percent va

15% for virus yield data, 4% for 24-h MTS data, anddata. Solvent-treated control cell monolayers contairper well at 0 h, 3.0 x 104 cells per well at 24 h, andwell at 48 h.

plaque reduction and virus yield assays, affectiand the number of plaques (Table 5). In the p

assay, the IC50 for the reduction of plaqL-735,882 was approximately 50 ,uM (Table creduction in plaque size was observed (daHowever, in the virus yield assay, this concenpound reduced virus titers by greater than27-fold on a per plaque basis (Table 5), and tIC50 was approximately 2 ,uM (Fig. 5). Notiattained in the plaque reduction assay (50 pu1potent than that attained in the virus yielddemonstrating the effect of the compoundreduction. The antiviral activity was observed icytoxicity, because testing of compound L-73'trations of up to 100 ,iM had no effect on theMDCK cells in growth medium when growthby the MTS colorimetric method during 24-periods (Fig. 5). At concentrations of up tcompound likewise had no effect on the growcells in the virus yield assay measured durinperiod of the assay.

DISCUSSIONThe transcriptase of influenza virus is a pot

the development of antiviral agents becausefor an essential and highly conserved catalyticvirus. In this report we described a novel classinhibitors, 4-substituted 2,4-dioxobutanoic aci4ically inhibit the endonuclease function of the

Others have reported on the inhibition of

transcriptase by PPi analogs (e.g., PFA) and heavy metalchelating compounds (26, 34) which were defined as inhibitorsof elongation. Both groups of compounds are thought toinhibit virus transcription by complexing with a zinc ion at theactive site of the polymerase, thereby preventing the binding ofnucleotide triphosphate or the release of Pi. These compoundsare inhibitory to other polymerases, presumably by the samemechanism of action. Similarly, polymerized polyprenoids(lignins) are elongation inhibitors of mRNA synthesis whichinhibit the growth of influenza virus and other viruses in cellculture (33). Anti-influenza virus activity has also been ob-served in vivo for the nucleoside analogs 2'-deoxy-2'-fluorori-bosides, whose modes of action are possibly directed atprimary and secondary transcription. However, these com-pounds also inhibit other viral and bacterial polymerase as well(35). The inhibition of influenza virus polymerases and otherpolymerases by these compounds may reflect the high degreeof sequence homology that exists between the elongation

10 100 subunit (PB1) of the influenza transcriptase complex and otherRNA and DNA polymerases.

cell proliferation The 4-substituted 2,4-dioxobutanoic acids, which are cycliccat the indicated or multicyclic structures containing a 2,4-dioxobutanoic siderials and Methods chain, were identified in a cap-dependent transcription assaytercentage of virus for influenza viruses. Structure-activity relationship studies:11 supernatants at revealed that the side chain was essential for inhibitory activity,neasuring plaques although lipophilic substitution was required for potency, withhe solvent-treated piperidine substituents yielding the most potent compounds.ilferation assay for Unlike the aforementioned transcriptase inhibitors, these com-E) and at 48 h (°) pounds were selective for the influenza virus transcriptase,%ations averaged having no effect on the activities of other viral and mammalian2% for 48-h MTS

ied 1.3 x 104 cells polymerases. Additionally, the inhibitory activity was con-

5.9 x 104 cells per served because the compounds inhibited the transcriptaseactivities of several strains of influenza A and B viruses withsimilar potencies, an observation consistent with the extent ofhomology (38 to 60%) present in the primary amino acid

ing both the size sequences of influenza A and B virus polymerases (40).)laque reduction The mode of action of the 4-substituted dioxobutanoic acidslue number by was determined in activity assays in which the steps of cap-5), and a sixfold dependent transcription were uncoupled. The dioxobutanoicta not shown). acids had no effect on transcription reactions primed in theitration of com- absence of capped RNA substrate with the dinucleotide ApG,400-fold or by which primes mRNA elongation by base pairing to the 3' endthe extrapolated of the viral RNA template (23), thereby indicating that the:e that the IC50 dioxobutanoic acids were not inhibitors of elongation butt)is 25-fold less instead affected the endonucleolytic processing of RNA and/orLassay (2 uM)l the initiation of mRNA synthesis. Direct evidence for theon plaque size inhibition of endonucleolytic activity was obtained in an assayn the absence of in which the compounds specifically inhibited the cleavage of5,882 at concen- radiolabeled capped RNA by the influenza virus endonucleaseproliferation of but had no effect on the activities of other nucleases.was quantitated The putative mode of endonuclease inhibition was con-and 48-h time firmed in transcription and initiation reactions primed with the

:o 100 ,uM, the 13-nt cleavage product of substrate RNA, an intermediate inth of uninfected the transcription reaction which is directly elongated. Theg the 24-h time inhibitors had no effect in reactions primed with the 13-nt

substrate; only reactions primed with capped substrates whichrequired endonucleolytic processing prior to nucleotide addi-tion were inhibited. In contrast, elongation inhibitors such aslignins, polyphenolic compounds, and PFA inhibited influenza

tential target for virus transcription primed with both the 13-nt substrate andit is responsible the capped substrates which undergo endonucleolytic process-function of the ing, but they had no effect on cleavage (unpublished data).of transcriptase Thus, the dioxobutanoic acid compounds are a class of tran-ds, which specif- scriptase inhibitors which differ mechanistically from inhibitorstranscriptase. of elongation, perhaps by interaction with a form of the

f influenza virus polymerase complex involved in cap-dependent initiation that

VOL. 38, 1994


http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/


is distinct from the enzyme conformation during elongation. Inthis regard, rearrangements of the influenza virus polymerasecomplex have been demonstrated to occur during cap-depen-dent transcription by biochemical cross-linking and inhibitorstudies (5, 34).

It has been predicted that the cap-binding activity of poly-merase protein PB2 may be conserved among mammaliancap-binding proteins (8, 32), and thus, it is possible that aninhibitor of cap-dependent priming, which is competitive withthe binding of capped RNA substrate to the polymerase, wouldnot be influenza virus specific. The 4-substituted dioxobutanoicacid compounds were not competitive inhibitors of in vitrotranscription with respect to the cap 1 RNA substrate, indicat-ing that the compounds did not interact with the site to whichthe 5' capped RNA substrate binds to polymerase protein PB2.This was further consistent with the observations that thedioxobutanoic acids neither affected the priming of transcrip-tion with the 13-nt capped RNA substrate (above) nor inhib-ited assays involving cap 1 recognition (data not shown). Itshould be noted that Chung et al. (6) recently described apotent class of short, capped oligoribonucleotides which inhibitcap-dependent influenza virus transcription, presumably bycompetitively inhibiting the binding of capped RNA substrateto the polymerase. Although this observation demonstrated anapproach to the development of potent inhibitors of transcrip-tion initiation, the specificities of these inhibitors for influenzavirus cap-binding protein was not investigated.

In addition to the inhibitory activities observed for thedioxobutanoic acids in the in vitro enzyme assays, benzyl-substituted compounds inhibited the replication of both influ-enza A and B viruses in cell culture in the absence of apparentcytotoxic effects. The antiviral activity was specific for influenzaviruses because the compounds had no effect on the replicationof several other viruses including LACV, a Bunyavirus, afinding which correlated with the biochemical selectivities ofthe dioxobutanoic acids. Furthermore, the inhibitors wereactive in both single and multicycle replication assays, demon-strating the effectiveness of these compounds not only on aninitial cycle of influenza virus replication but on subsequentrounds as well.

In summary, this is the first report of a specific class ofinhibitors for the in vitro transcriptase and replicative activitiesof influenza A and B viruses. The mode of action of thesecompounds, inhibition of the cap-dependent endonucleaseactivity of the viral polymerase complex, differs from thepreviously reported activities of transcriptase inhibitors, whichinhibited the elongation of mRNA by influenza virus poly-merases and other polymerases. The selectivities of the inhib-itors described here demonstrate the unique and conservednature of the endonuclease of influenza virus transcriptase,thereby supporting the potential exploitation of this viralenzyme as a target of intervention in the replication ofinfluenza viruses. Although additional studies are required todefine precisely the mechanistic role of these inhibitors inendonuclease inhibition, the dioxobutanoic acids will be usefulin further probing the biochemical and biological properties ofthis antiviral target.

ACKNOWLEDGMENTSWe thank Emilio Emini (Antiviral Department, Merck) for helpful

discussions and critical reading of the manuscript, Dolores Wilson(Antiviral Department, Merck) for preparation of the manuscript, andJules Shafer (Biological Chemistry Department, Merck) for advice andinterest in our influenza transcription program. We are also grateful toJulie O'Brien (New Lead Pharmacology, Merck) for performing theHIV reverse transcriptase, HIV RNase H, and HeLa cell DNA

polymerase a assays and to Paula Giesa (formerly of Virus and CellBiology, Merck) for invaluable assistance in reagent preparation.

REFERENCES

1. Acheson, R. M., and W. R. Tully. 1968. Addition reactions ofheterocyclic compounds. XXXVI. Dialkyl acetylenedicarboxylateswith 2 substituted benzimidazoles, benzoxazoles and bezothia-zoles. J. Chem. Soc. C: 13:1623-1629.

2. Barbosa, E., and B. Moss. 1978. mRNA (nucleoside-2'-)-methyl-transferase from vaccinia virus. J. Biol. Chem. 253:7962-7697.

3. Barrett, T., and S. C. Inglis. 1985. Growth, purification andtitration of influenza viruses, p. 119-50. In B. W. S. Mahy (ed.),Virology: a practical approach. IRL Press, Oxford.

4. Bouloy, M., S. J. Plotch, and R. M. Krug. 1978. Globin mRNAs areprimers for the transcription of influenza viral RNA in vitro. Proc.Natl. Acad. Sci. USA 75:4886-4890.

5. Braam, J., I. Ulmanen, and R. Krug. 1983. Molecular model of aeukaryotic transcription complex: functions and movements ofinfluenza P protein during capped RNA-primed transcription. Cell34:609-618.

6. Chung, T. D. Y., C. Cianci, M. Hagen, B. Terry, J. T. Matthews, M.Krystal, and R. J. Colonno. 1994. Biochemical studies on cappedRNA primers identify a class of oligonucleotide inhibitors of theinfluenza virus RNA polymerase. Proc. Natl. Acad. Sci. USA 91:2372-2376.

7. Copeland, W. C., and T. S.-F. Wang. 1991. Catalytic subunit ofhuman DNA polymerase a overproduced from baculovirus-in-fected insect cells. J. Biol. Chem. 266:22739-22748.

8. De la Luna, S., C. Martinez, and J. Ortin. 1989. Molecular cloningand sequencing of influenza virus A/Victoria/3/75 polymerasegenes: sequence evolution and prediction of possible functionaldomains. Virus Res. 13:143-155.

9. Delarue, M., 0. Poch, N. Tordo, N. Moras, and P. Argos. 1990. Anattempt to unify the structure of polymerases. Protein Eng. 3:461-467.

10. DeLean, A., P. J. Munson, and D. Rodbard. 1978. Simultaneousanalysis of families of sigmoidal curves: application to bioassay,radioligand assay, and physiological dose-response curves. Am. J.Physiol. 235(2):E97-E102.

11. Emerson, D. W., R. L. Titus, and R. M. Gonzalez. 1991. Evidencefor ketene intermediates in the decarbonylation of 2,4-dioxoacidsand esters and 2-oxobutanedioic acid esters. J. Org. Chem. 56:5301-5307.

12. Frankel-Conrat, H., and A. Steinschneider. 1967. Stepwise degra-dation of RNA: periodate followed by aniline cleavage. MethodsEnzymol. 12B:243-246.

13. Goldman, M. E., J. H. Nunberg, J. A. O'Brien, J. C. Quintero,W. A. Schleif, K. F. Freund, S. L. Gaul, W. S. Saari, J. S. Wai, J. M.Hoffman, P. S. Anderson, D. J. Hupe, E. A. Emini, and A. M.Stern. 1991. Pyridinone derivatives: specific human immunodefi-ciency virus type 1 reverse transcriptase inhibitors with antiviralactivity. Proc. Nati. Acad. Sci. USA 88:6863-6867.

14. Hayden, F. G., and A. J. Hay. 1992. Emergence and transmissionof influenza A viruses resistant to amantadine and rimantadine. JCurr. Top. Microbiol. Immunol. 176:119-130.

15. Honda, A., J. Mukaigawa, A. Yokoiyama, A. Kato, S. Ueda, K.Nagata, M. Krystal, D. P. Nayak, and A. Ishihama. 1990. Purifi-cation and molecular structure of RNA polymerase from influenzavirus A/PR8. J. Biochem. 107:624-628.

16. Horne, D., J. Gaudino, and W. J. Thompson. 1984. A convenientmethod for the synthesis of alpha keto esters from aldehydes.Tetrahedron Lett. 25:3529-3532.

17. Horisberger, M. 1982. Identification of a catalytic activity of thelarge basic P polypeptide of influenza. Virology 120:279-286.

18. Jeltsch, A., A. Fritz, J. Alves, H. Wolfes, and A. Pingoud. 1993. Afast and accurate enzyme-linked immunosorbent assay for thedetermination of the DNA cleavage activity of restriction endo-nucleases. Anal. Biochem. 213:234-240.

19. Kamer, G., and P. Argos. 1984. Primary structural comparison ofRNA-dependent polymerases from plant, animal and bacterialviruses. Nucleic Acids Res. 12:7269-7282.

20. Kawakami, K., K. Mizumoto, A. Ishihama, K. Shinozaki-yamagu-



http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/


chi, and K. Miura. 1985. Activation of influenza virus-associatedRNA polymerase by cap-1 structure (m7GpppNm). J. Biochem.97:655-661.

21. Krug, R. M., F. V. Alonso-Caplan, I. Julkunen, and M. G. Katze.1989. Expression and replication of the influenza virus genome, p.89-152. In R. M. Krug (ed.), The influenza viruses. Plenum Press,New York.

22. Manley, J. L. Transcription of eucaryotic genes in a whole cellextract, p. 71-88. In B. D. Hames and S. J. Higgins (ed.), Tran-scription and translation: a practical approach. IRL Press, Oxford.

23. McGeoch, D., and N. Kitron. 1975. Influenza virion RNA-depen-dent RNA polymerase: stimulation by guanosine and related com-pounds. J. Virol. 15:686-695.

24. Milligan, J. F., D. R Groebe, G. W. Witherell, and 0. C.Uhlenbeck 1987. Oligoribonucleotide synthesis using T7 RNApolymerase and synthetic DNA templates. Nucleic Acids Res. 15:8783-8789.

25. Nichol, S. T., C. R. Penn, and B. W. J. Mahy. 1981. Evidence forthe involvement of influenza A (Four Plaque Rostock) virusprotein P2 in ApG and mRNA-primed in vitro RNA synthesis. J.Gen Virol. 57:407-413.

26. Oxford, J. S., and D. D. Perrin. 1977. Influenza RNA transcriptaseinhibitors: studies in vitro and in vivo. Ann. N. Y. Acad. Sci. 284:613-623.

27. Palese, P., M. B. Ritchey, and J. L. Schulman. 1977. P1 and P3proteins of influenza virus are required for complementary RNAsynthesis. J. Virol. 21:1187-1195.

28. Plotch, S. J., M. Bouloy, I. Ulmanen, and R M. Krug. 1981. Aunique cap (m7GpppXm)-dependent influenza virion endonucle-ase cleaves capped RNAs to generate the primers that initiate viralRNA transcription. Cell 23:847-858.

29. Romanos, M. A., and A. J. Hay. 1984. Identification of theinfluenza virus transcriptase by affinity-labeling with pyridoxal5'-phosphate. Virology 132:110-117.

30. Rooney, C. S., W. C. Randall, K. B. Streeter, C. Ziegler, E. J.Cragoe, Jr., H. Schwam, S. R Michelson, H. W. R William, E.Eichler, D. E. Duggan, E. H. Ulm, and R M. Noll. 1983. Inhibitorsof glycolic acid oxidase. 3-Hydroxy-lH-pyrrole-2,5-dione deriva-tives. J. Med. Chem. 26:700-714.

31. Rossier, C. J., J. L. Patterson, and D. Kolakofsky. 1986. La Crossevirus small genome mRNA is made in the cytoplasm. J. Virol. 58:647-650.

32. Rychlik, W., L. L. Domier, P. R. Gardner, G. M. Hellmann, andR. E. Rhoads. 1987. Amino acid sequence of the mRNA cap-binding protein from human tissue. Proc. Natl. Acad. Sci. USA 84:945-949.

33. Sakagami, H., K. Nagata, A. Ishihama, T. Oh-hara, and Y.Kawazoe. 1990. Anti-influenza virus activity of synthetically poly-merized phenylpropenoids. Biochem. Biophys. Res. Commun.172:1267-1272.

34. Stridh, S., R. Daetma, and C. Scholtisselk 1987. Studies oninhibitors of influenza mRNA synthesis, p. 91-98. In B. W. S.Mahy and D. Kolakofsky (ed.), The biology of negative strandviruses. Elsevier Science Publisher, Amsterdam.

35. Tisdale, M., G. Appleyard, J. V. Tuttle, D. J. Nelson, S. Nusinof-Lehrman, W. Al Nakib, J. N. Stables, D. J. M. Purifoy, K. L.Powell, and G. Darby. 1993. Inhibition of influenza A and Bviruses by 2'-deoxy-2'-fluororibosides. Antiviral Chem. Chemo-ther. 4:281-287.

36. Tobita, K., A. Sugiura, C. Enomoto, and M. Furuyama. 1975.Plaque assay and primary isolation of influenza A viruses in anestablished line of canine kidney cells (MDCK) in the presence oftrypsin. Med. Microbiol. Immunol. 162:9-14.

37. Ulmanen, I., B. A. Broni, and R M. Krug. 1981. Role of two of theinfluenza virus core P proteins in recognizing cap 1 structures(m7GpppNm) on RNAs and in initiating viral RNA Transcription.Proc. Natl. Acad. Sci. USA 78:7355-7359.

38. Williams, H. W. R, E. Eichler, W. C. Randall, C. S. Rooney, E. J.Cragoe, K. B. Streeter, H. Schwam, S. R Michelson, A. Patchet,and D. Taub. 1983. Inhibitors of glycolic acid oxidase. 4-Substi-tuted 2,4 dioxobutanoic acid derivatives. J. Med. Chem. 26:1196-1200.

39. Wolfe, J. F., D. E. Portlock, and D. J. Feuerbach. 1974. Syntheticand mechanistic aspects of the sodium hydride promoted acylationof methylated aromatics. J. Org. Chem. 39:2006-2010.

40. Yamashita, M., M. Krystal, and P. Palese. 1989. Comparison ofthe three large polymerase proteins of influenza A, B, and Cviruses. Virology 171:458-466.

VOL. 38, 1994


http://aac.asm.org/

Dow

nloaded from

http://aac.asm.org/

Inhibition of Cap (m7GpppXm)-Dependent Endonuclease ...€¦ · ANTIMICROBIALAGENTSANDCHEMOTHERAPY, Dec. 1994, p. 2827-2837 0066-4804/94/$04.00+0 Copyright X 1994, American Society

Documents