ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Aug. 1986, p. 215-219 Vol. 30, No. 2 0066-4804/86/080215-05$02.00/0 Copyright © 1986, American Society for Microbiology In Vitro Activity of against Herpes Simplex Virus JOHN C. POTTAGE, JR., 2* HAROLD A. KESSLER,"2 JAMES M. GOODRICH,"2 ROBERT CHASE' CONSTANCE A. BENSON,' KATHI KAPELL,' AND STUART LEVIN' Section ofInfectious Disease, Department ofInternal Medicine,' and Department ofImmunology/Microbiology,2 Rush- Presbyterian-St. Luke's Medical Center, Chicago, Illinois 60612 Received 23 January 1986/Accepted 13 May 1986

The effects of ketoconazole alone and in combination with acyclovir and adenine arabinoside upon the replication of herpes simplex virus types 1 and 2 (HSV-1 and -2) were investigated by using a yield reduction assay. Ketoconazole demonstrated antiviral activity against HSV-1 and -2 and synergistic antiviral activity when it was combined with acyclovir. Combinations of ketoconazole with adenine arabinoside resulted in either interference or indifference. The effects of ketoconazole upon the protein synthesis of HSV-2-infected cells were also determined in an effort to define the mechanism of action for the antiviral activity of ketoconazole. There was no reduction of' HSV proteins when compared with acyclovir. These findings suggest that further investigations of the use of ketoconazole for the treatment of HSV are warranted.

Recently, several reports have described a novel approach 107 PFU/ml were grown in monolayers of Vero cells and to antiviral chemotherapy with the ' agent stored at -70°C. . Amphotericin B, a polyene , ex- Cells. Vero and human lung (HL) cells (derived from an erts its antifungal effect by binding to sterol components in HL carcinoma) were used in all assays. Vero cells were the fungal cell membrane, which results in increased perme- grown in RPMI 1640 supplemented with 10% inactivated ability of the membrane (11). By using the water-soluble fetal bovine serum, 1% glutamine, 100 U of penicillin per'ml, ester of amphotericin B, Jordan and Seet demonstrated and 50 jig of streptomycin per ml; and HL cells were grown antiviral effects against enveloped viruses in plaque reduc- in Hanks balanced salt solution, basal medium Eagle im- tion assays (5). Similarly, Kessler et'al. have shown that proved, supplemented with 10% inactivated fetal bovine treatment of hepatitis B virus with increasing doses of serum, 1% glutamine, 100 U of penicillin per ml, and 50 ,ug amphotericin B caused disruption of the viral particles (6). of streptomycin per ml. Maintenance medium for both cell Extending these observations, Malewicz et al. investigated lines was supplemented with 1% inactivated fetal bovine the combined effects of amphotericin B and acyclovir upon serum in place of 10% inactivated fetal bovine serum. the replication of pseudorabies virus. Low doses of Drugs. Ketoconazole was' provided kindly by Janssen amphotericin B in combination with acyclovir produced a Pharmaceutical, New Brunswick, N.J. Acyclovir wvas pro- synergistic reduction in viral replication as shown with a vided kindly by Burroughs Wellcome Co., Research Trian- plaque reduction assay. The exact mechanism for this syn- gle Park, N.C., and adenine arabinoside was obtained from ergistic effect was not elucidated (9). Sigma Chemical Co., St. Louis, Mo. In view of this promising work, we investigated the Yield reduction assay. Vero and HL cells were grown to possibility of antiviral activity for a different antifungal 90% confluency in six-well disposable trays (Costar, agent, ketoconazole. Ketoconazole is an com- Cambridge, Mass.). Growth medium was removed, and the pound which' exerts its antifungal activity through inhibition cells were washed three times with phosphate-buffered sa- of lanosterol demethylation. This blocks the synthesis of line, pH 7.2. Virus in maintenance medium (0.5 ml) was , the major sterol component of the fungal cell added at a multiplicity of of 0.5. After 1 h of membrane (2). In mammalian cells, ketoconazole also inhib- adsorption at 37TC, the inoculum was removed, and mainte- its lanosterol demethylation, with a subsequent decrease in nance medium with various concentrations and combina- the biosynthesis of cholesterol, the major sterol component tions of ketoconazole, acyclovir, and adenine arabinoside of mammalian cell membranes. In addition, ketoconazole was added. Control cells were inoculated with virus and interferes with cellular fatty acid and phospholipid biosyn- treated with maintenance medium without drugs. The plates thesis (2). In this study, we examined the effects of ketocon- were incubated at' 37°C in 5% CO2 for 24 h, after which they alone and in combination with acyclovir and adenine were frozen at -70°C until titration by plaque assay. arabinoside upon herpes simplex virus types 1 and 2 (HSV-1 Virus plaque assay. HL cells grown to confluence in sterile and -2) replication with a yield reduction assay. plastic six-well plates (Costar) were inoculated with logl0 dilutions of supematants from the yield reduction assay in MATERIALS AND METHODS triplicate. The plates were incubated at 390C in 5% CO2 for 1 h with intermittent rocking at 15-min intervals. After Viruses. HSV-1 (F strain) and -2 (G strain) were obtained adsorption, 5 ml of carboxymethyl cellulose overlay was from the American Type Culture Collection, Rockville, Md. added to each well, and the plates were incubated for an An acyclovir-resistant HSV-1 thymidine kinase-deficient additional 72 h at 37°C in 5% CO2. Monolayers were then (TK-) mutant (VL 8738) was provided kindly by Sandra fixed for 2 h in a solution of ethanol, glacial , and Nusinoff-Lehrman (Burroughs Wellcome Co., Research Tri- Formalin (6:2:1) and stained with 1% crystal violet. Plaques angle Park, N.C.). Virus stocks with titers of approximately were quantitated with a dissecting stereomicroscope. Preincubation of virus in the presence of ketoconazole. * Corresponding author. HSV-2 (G strain) was incubated with ketoconazole in tripli- 215 216 POTTAGE ET AL. ANTIMICROB. AGENTS CHEMOTHER.

TABLE 1. Yield reduction assay of HSV with ketoconazole, acyclovir, and adenine arabinoside in HL cells HSV-1 (F strain) HSV-2 (G strain) HSV-1 (VL 8737)a Drug Concn (~Lg/ml) Titerb % of control Titer % of control Titer % of control Control 0 4.1 x 106 100 1.5 x 106 100 8.5 x i0S 100

Ketoconazole 1 3.6 x 106 87 1.1 X 106 79 7.9 X 105 93 S 1.3 x 106 33 4.2 x 105 28 5.4 x 105 63 10 2.4 x 105 5 7.5 x 104 5 4.7 x 105 56

Acyclovir 1 1.4 X 106 33 6.4 X 105 43 7.0 x 105 83 5 1.0 X 105 2.5 3.0 x 104 2.3 5.3 x 105 62

Adenine arabinoside 1 1.2 x 106 29 1.5 X 105 10 2.2 X 105 26 5 1.9 x 10 4 1.5 x 104 1 3.2 x 104 3 a Acyclovir-resistant thymidine kinase-negative mutant. b Virus titer, PFU/ml. cate under various conditions before plaque assay. The virus titer produced in the absence of these two drugs. The variables included time (0, 20, 40, or 60 min), temperature (4, calculated or expected yield of control (Yc) is the product of 25, or 37°C), and ketoconazole concentrations (0, 5, or 10 YA and YB. If YAB is equal to Yc, the interaction of the drugs ,ug/ml). The reaction mixture consisted of growth media, in combination is additive. If YAB is less than Yc, then the HSV-2 (5 x 102 PFU/ml), and the appropriate concentration interaction is synergistic. If YAB is greater than the most of ketoconazole. The final volume of the reaction mixture effective agent alone, the interaction is interference. If YAB is was 2 ml. After the appropriate incubation period, the viral greater than the least effective agent alone, the interaction is titer of the reaction mixture was determined by plaque assay antagonism. When the inhibition is not greater than the most and compared with proper control mixtures. effective drug alone, the interaction is defined as indifference Protein synthesis studies. The effect of ketoconazole upon (14). total protein synthesis by HSV-infected cells was deter- Drug toxicity assay. Approximately 5 x 105 cells per ml of mined by incorporation of [35S]methionine into trichloroace- Vero and HL cells in growth medium were inoculated into tic acid-precipitable material. Confluent Vero cells were 25-cm2 tissue culture flasks. The cells were incubated for 24 infected with HSV-2 (G strain) at a multiplicity of infection h at 37°C in 5% C02, after which the growth medium was of 5.0. After 1 h of adsorption, fresh medium with and replaced with fresh growth medium with various concentra- without ketoconazole (10 ,ug/ml) was added. At various tions of ketoconazole, acyclovir, and adenine arabinoside times after infection (3, 5, 9, 12, 15, and 18 h), the monolay- alone and in combination. Controls were grown in growth ers were pulsed with 20 p.Ci of [35S]methionine for 2 h. The medium without drugs. At 24-h intervals for 72 h, the medium was removed, and the monolayers were dissolved in monolayers from three flasks were trypsinized, and viable polyacrylamide (PAGE) solubilization cells were quantitated by the trypan blue exclusion method. buffer. Solubilized monolayers were then boiled for 4 min and precipitated with 5% trichloroacetic acid. The precipi- RESULTS tates were collected on membrane filters (Millipore Corp.) Treatment of HSV-1 and -2 with ketoconazole in HL cells (0.45-,um pore size), washed, and counted in a liquid scintil- resulted in a dose-dependent reduction in viral titer (Table lation counter. Results were expressed as a percentage of 1). The greatest effect occurred at a ketoconazole concen- uninfected controls. tration of 10 p.g/ml. Of note, the antiviral effect of ketocon- PAGE. Vero cells were grown in 24-well disposable trays azole was similar for both HSV-1 and -2. The HSV-1 TK- (Costar) until confluent (2.2 x 10- cells per tray). The cells strain (VL 8737) was also tested against ketoconazole (Table were infected with HSV-2 at a multiplicity ofinfection of5.0. 1). Although there was a dose-dependent reduction in viral The monolayers were pulsed with 20 ,uCi of [35S]methionine titer, the reduction was of a lesser degree than that of the per ml for 2-h periods at 3, 6, 9, 12, 15, and 18 h postinfec- HSV-1 (F strain). The reasons for the decreased susceptibil- tion. After pulse labeling, the medium was removed, and the ity of the TK- mutant are not known and will require further monolayers were washed and solubilized as described evaluation. Similar results were obtained in Vero cells (data above. Sodium dodecyl sulfate-PAGE was performed by the not shown). method of Laemmli (7). The final concentration of All combinations and concentrations of ketoconazole and acrylamide in the separating gel was 8.5%. The gels were acyclovir demonstrated synergy against both HSV-1 and -2. fixed, stained with Coomassie blue, dried, and autoradio- This was particularly obvious at high concentrations of graphed at -70°C. ketoconazole and acyclovir (Table 2). The antiviral activity Definition of drug interactions. The definitions of drug of ketoconazole alone and in combination with acyclovir was interactions are as described by Schinazi et al. (14). Briefly, similarly demonstrated in Vero cells. In contrast, combina- the yield of control for drug A (YA) or drug B (YB) is the virus tions of ketoconazole and adenine arabinoside demonstrated titer produced in the presence of drug, divided by the virus either interference, indifference, or in one case, antagonism titer produced in the absence ofdrug. The yield ofcontrol for (Table 2). These effects occurred in both cell lines tested. the combination of two drugs (YAB) is the virus titer pro- Pretreatment of the HSV-2 strain with various concentra- duced in the presence of these two drugs, divided by the tions of ketoconazole for periods up to 60 min at the various VOL. 30, 1986 ANTIVIRAL EFFECTS OF KETOCONAZOLE 217

TABLE 2. Yield reduction assay of HSV with combinations of ketoconazole, acyclovir, and adenine arabinoside in HL cells HSV-1 (F strain) HSV-2 (G strain) Drug Concn (jig/mi) Yield Yield Observed % Titer Observed % Titeria of controlb Expectedc of control Expected Ketoconazole/acyclovir 5/1 3.9 x 105 gd 10 1.4 x 105 gd 12 10/1 4.6 x 104 le 1 7.5 x 103 0.5d 2 5/5 1.2 x 104 0.2d 0.8 1.5 x 103 0. ld 0.6 10/5 1.3 x 103 0.03ld 0.125 4.3 x 102 0.02d 0.1

Ketoconazole/adenine arabinoside 5/1 8.9 x 105 21f 9 4.6 x 105 30h 2.8 10/1 6.6 x 10W 159 1.45 2.0 x 105 13h 0.5 5/5 6.6 x 105 159 1 7.9 x 104 S 0.28 10/5 4.8 x 105 llh 0.2 4.8 x 104 39 0.05 a Viral titer, PFU per milliliter. b Yield observed is the titer of virus produced in presence of drug divided by titer of virus produced in absence of drug times 100. c Yield calculated for additive interaction is the product of observed yields of the control for each drug in combination. d Synergistic interaction. e Additive interaction. f Indifferent interaction. 8 Interference interaction. h Antagonistic interaction.

temperatures tested had no effect upon virus titer as com- We have previously shown that ketoconazole inhibits hepa- pared with controls. These data indicate that ketoconazole titis B surface antigen production by a transformed human does not have a direct effect upon HSV, in contrast to what hepatoma cell line which is chronically infected with hepa- has been demonstrated for the methyl ester of amphotericin titis B virus (13). B (5). In the present study we have shown that ketoconazole at Toxicity of acyclovir and adenine arabinoside toward a concentration of 10 ,ug/ml causes an approximate 1-log rapidly dividing Vero cells and HL cells was minimal at the reduction of HSV titer by an in vitro yield reduction assay. concentrations used in these experiments. Acyclovir at 5 A yield reduction assay was chosen for these experiments to jig/ml caused no reduction in either viable Vero cells or HL give a better estimate of the antiviral activity of ketocona- cells at 24, 48, or 72 h. Adenine arabinoside at 5 ,ug/ml zole. The yield reduction assay allows for the use of a greater caused a 0, 4, and 11% reduction in viable Vero cells and a multiplicity of infection of virus and evaluation of higher 0, 3, and 8% reduction in HL cells at 24, 48, and 72 h, concentrations of antiviral drugs in combination studies (14). respectively. Ketoconazole at a concentration of 10 pug/ml In addition, this assay evaluates the antiviral activity of the caused 1, 17, and 31% reduction in viable Vero cells at 24, various drugs within 24 h of inoculation, which eliminates 48, and 72 h, respectively. Similar treatment of rapidly some of the nonspecific cytotoxicity which can be seen with dividing HL cells yielded 0, 12, and 26% reduction in cell assays requiring 48 to 72 h of incubation. When tested in a viability by trypan blue exclusion. In control flasks, there standard plaque reduction assay, ketoconazole at a concen- were 1.2 x 106 Vero cells per flask and 2 x 105 HL cells per tration of 10 ,ug/ml causes a 3- to 4-log reduction in HSV titer flask at 72 h. Combinations of ketoconazole with acyclovir (J. Pottage, H. Kessler, J. Goodrich, T. Hurley, and S. and adenine arabinoside did not cause increased cytoxicity. Levin, Clin. Res. 31:736A, 1983). Of potential clinical sig- Ketoconazole at a concentration of 10 ,ug/ml had little nificance is the finding of a synergistic antiviral effect against effect upon protein synthesis in both uninfected and infected HSV-1 and -2 when ketoconazole is combined with Vero cells as determined by [35S]methionine incorporation. acyclovir. Like acyclovir, ketoconazole is an oral agent. The The effect of ketoconazole versus acyclovir upon the pro- levels of ketoconazole studied in these experiments are duction of individual HSV proteins by PAGE is shown in obtainable in the serum of humans. Whether the combina- Fig. 1. Again, ketoconazole at a concentration of 10 ,ug/ml tion of acyclovir and ketoconazole will be clinically effective showed little effect upon protein synthesis in infected Vero in HSV infections remains to be seen and will require cells. Acyclovir at a concentration of 10 ,ug/ml showed a additional in vivo animal studies. reduction in several late HSV proteins, in accordance with The mechanism of action of the antiviral activity of what has been previously reported (4). ketoconazole is unknown. Our experiments show that there seems to be no direct effect upon the virus particles by ketoconazole. In mammalian cells, ketoconazole interferes DISCUSSION with cholesterol biosynthesis by inhibiting lanosterol de- Ketoconazole is an imidazole antibiotic which has been methylation (2). It also interferes with phospholipid biosyn- used extensively for the treatment of fungal infections. thesis by the cell (2). There is a paucity of information Ketoconazole has also been found to have in vitro inhibitory regarding the role of biosynthesis in HSV replication. A activity for other microorganisms, including such as possible site of action for the antiviral activity of ketocona- (15), and the protozoans Plasmo- zole is interference with lipid metabolism resulting in the dium falciparum (12), Trypanosoma cruzi (10), and Leish- production of a defective lipid envelope for the virus, an mania tropica (1). The exact mechanism for the inhibitory action that should decrease virus infectivity. Partial support activity against these diverse organisms is not known, but is for this hypothesis comes from the observation that keto- thought to be related to alterations in their lipid metabolism. conazole does not demonstrate antiviral activity against the 218 POTTAGE ET AL. ANTIMICROB. AGENTS CHEMOTHER.

FIG. 1. Sodium dodecyl sulfate-PAGE of [5S]methionine-labeled proteins from HSV-2-infected Vero cells in the presence of either no drug (c), ketoconazole (k) (10 ,ug/ml), or acyclovir (a) (5 ,ug/ml). Top line is the time (h) postinfection when the samples were pulsed labeled with [I5S]methionine. Molecular weights of protein standards are on the vertical axes on either side.

nonenveloped poliovirus in a similar yield reduction assay in approximately 1 in 10,000 persons (8). Further evaluation (unpublished data). We have also shown that the of ketoconazole will be required before it can be investigated antifungal agents, and R47,829, which are more for use in the treatment of human HSV infections. Impor- specific for inhibition offungal sterol biosynthesis and do not tantly, however, ketoconazole may prove to be a valuable interfere with cholesterol biosynthesis in mammalian cells, agent for the study of lipid biosynthesis and metabolism in have no antiviral activity against HSV-1 and -2 (J. Pottage, HSV-infected cells. Additionally, these observations may H. Kessler, J. Goodrich, K. Kapell, and S. Levin, Program provide the start for a new approach in the development of Abstr. 25th Intersci. Conf. Antimicrob. Agents Chemother., antiviral agents. abstr. no. 120, 1985). Alternatively, the antiviral effect of ketoconazole could be caused by a specific alteration in ACKNOWLEDGMENT DNA or protein synthesis. Ketoconazole has been demon- strated to inhibit DNA synthesis of stimulated lymphocytes We thank Julie Adams for secretarial support. (3). Although our studies with [35S]methionine incorporation and PAGE suggest that ketoconazole does not affect the LITERATURE CITED production of the total quantity of HSV proteins in infected 1. Berman, J. D. 1981. Activity of against Leishmania cells, studies of the effects of ketoconazole upon individual tropica in human macrophage cultures. Am. J. Trop. Med. Hyg. proteins will need to be performed. The mechanism for the 30:566-569. 2. Borgers, M. 1980. Mechanism of action of antifungal drugs, with synergistic antiviral effects when ketoconazole was com- Dis. bined with acyclovir and interference when combined with special reference to the imidazole derivatives. Rev. Infect. at this time. Further 2:520-534. adenine arabinoside is unexplained 3. Buttke, T. M., and S. W. Chapman. 1983. Inhibition by keto- studies will be required to determine the exact mechanism of conazole of mitogen-induced DNA synthesis and cholesterol action of the anti-HSV activity of ketoconazole. biosynthesis in lymphocytes. Antimicrob. Agents Chemother. Whether these in vitro observations regarding the antiviral 24:478-485. activity of ketoconazole will correlate with clinical efficacy 4. Furman, P. A., and P. V. McGuirt. 1983. Effect of acyclovir on in humans is unknown. There is a report of a patient with viral protein synthesis in cells infected with herpes simplex type infection who favorably responded 1. Antimicrob. Agents Chemother. 23:332-334. recurrent HSV-2 genital of The a female with a 4-year 5. Jordan, G. W., and E. C. Seet. 1978. Antiviral effects to ketoconazole (16). patient, Antimicrob. Agents Chemother. history of recurrent genital HSV infections (recurring every amphotericin B methyl ester. had no further 13:199-204. 2 weeks) and concomitant candida vaginitis 6. Kessler, H. A., J. Dixon, C. R. Howard, K. Tsiquaye, and A. J. recurrences of genital HSV over a 5-month follow-up period Zuckerman. 1981. Effects of amphotericin B on hepatitis B (16). Although this may be a coincidental effect, this case virus. Antimicrob. Agents Chemother. 20:826-833. report and our in vitro data suggest there may be a role for 7. Laemmli, U. K. 1970. Cleavage of structural proteins during the ketoconazole in the treatment of HSV infections. assembly of the head of bacteriophage T4. Nature (London). Although ketoconazole is a relatively safe antifungal 227:680-685. agent, it has been associated with significant hepatic toxicity 8. Lewis, J. H., H. J. Zimmerman, G. Benson, and K. G. Ishak. VOL. 30, 1986 ANTIVIRAL EFFECTS OF KETOCONAZOLE 219

1984. Hepatic injury associated with ketoconazole therapy: J. Infect. Dis. 144:372-375. analysis of 33 cases. Gastroenterology 86:503-513. 13. Pottage, J. C., Jr., and H. A. Kessler. 1985. Inhibition of in vitro 9. Malewicz, B., M. Momsen, and H. Jenkin. 1983. Combined HBsAg production by amphotericin B and ketoconazole. J. effect of acyclovir and amphotericin B on the replication of Med. Virol. 16:275-281. pseudorabies virus in BHK-21 cells. Antimicrob. Agents Che- 14. Schinazi, R. F., J. Peters, C. C. Williams, D. Chance, and A. J. mother. 23:119-124. Nahmias. 1982. Effect of combinations of acyclovir with 10. McCabe, R. E., E. G. Araujo, and J. S. Remington. 1983. vidarabine or its 5'-monophosphate on herpes simplex viruses in Ketoconazole protects against infection with Trypanosoma cell culture and in mice. Antimicrob. Agents Chemother. cruzi in a murine model. Am. J. Trop. Med. Hyg. 32:960- 22:499-507. 962. 15. Sud, I. J., and D. S. Feingold. 1982. Action of antifungal 11. Medoff, G., and G. A. Kobayashi. 1980. The polyenes, p. 3-33. imidazoles on Staphylococcus aureus. Antimicrob. Agents Che- In D. Speller (ed.), Antifungal chemotherapy. John Wiley and mother. 22:470-474. Sons, Ltd., London. 16. Tkach, J. R., and M. G. Rinaldi. 1983. Treatment of vaginal 12. Pfaler, M. A., and D. J. Krogstad. 1981. Imidazole and polyene with ketoconazole. Am. J. Obstet. Gynecol. 144: activity against chloroquine resistant Plasmodium falciparum. 122.