Short Communication Lack of Efficacy of Aurintricarboxylic Acid and Ethacrynic Acid Against Vaccinia Virus Respiratory Infections in Mice

Antiviral Chemistry & Chemotherapy 2010 20:201–205 (doi: 10.3851/IMP1480)

Short communication Lack of efficacy of aurintricarboxylic acid and ethacrynic acid against vaccinia virus respiratory infections in mice

Donald F Smee1*, Brett L Hurst1 and Min-Hui Wong1

1Institute for Antiviral Research, Department of Animal, Dairy and Veterinary Sciences, Utah State University, Logan, UT, USA

*Corresponding author e-mail: [email protected]

Background: Aurintricarboxylic acid (ATA) and ethacrynic 50% cytotoxicity at 84–173 µM, giving low (1.3–4.2) acid (ECA) have been reported to exhibit antiviral activity selectivity index values. Preliminary toxicity tests in against vaccinia virus infections in cell culture by inhib- uninfected mice indicated that ATA and ECA were both iting early and late gene transcription, respectively. The overtly toxic at 100 mg/kg/day. No protection from purpose of this work was to determine if these inhibitors mortality was afforded by treatment of vaccinia virus would effectively treat vaccinia virus infections in mice, infections with ATA or ECA, but 100% survival was which has not previously been studied. achieved in the cidofovir group. ATA- and ECA-treated Methods: ECA was investigated by cell culture plaque reduc- mice died significantly sooner than placebo-treated tion assay for the inhibition of cowpox and vaccinia virus animals, indicating that these compounds exacerbated infections to clarify issues regarding its potency and selec- the infection. tivity. Mice infected intranasally with vaccinia virus were Conclusions: Both ATA and ECA lack antiviral potency treated by intraperitoneal route twice daily for 5 days with and selectivity in cell culture. The compounds were ATA (10 and 30 mg/kg/day) and ECA (15 and 30 mg/kg/day) ineffective in treating mice at intraperitoneal doses or once daily for 2 days with cidofovir (100 mg/kg/day). of ≤30 mg/kg/day. These compounds do not appear to Results: ECA caused 50% inhibition of virus plaque have potential for the treatment of poxvirus infections formation at 20–79 µM in four cultured cell lines, with in vivo.

Introduction

Aurintricarboxylic acid (ATA) and ethacrynic acid 2.5–32 were based upon 50% reductions in virus titre (ECA) have been reported to inhibit vaccinia virus (at 5.4–19.3 µM) rather than on 90% effective con- replication in cell culture [1,2]. ATA inhibits an early centration (EC90 ) values [2]. Both compounds have transcriptional event of vaccinia virus replication by modes of action that differ from nucleotide analogues targeting cellular and viral factors [1]. The compound such as cidofovir that inhibits viral DNA synthesis blocks the phosphorylation of extracellular signal- [3] or the non-nucleoside ST-246 that prevents virus regulated kinase 1/2 (a cellular enzyme) and the phos- assembly [4]. phatase activity of the viral AH1L enzyme. ATA caused The authors of the reports of ATA and ECA inhibition a 90% reduction in virus titre at concentrations of indicate that these molecules might be drug candidates 25–200 µg/ml (60–475 µM), depending upon the cell or provide new directions in pursuing antipoxvirus com- line used, and toxicity was not evident at 500 µg/ml pounds that could be useful as drugs [1,2]. Thus, the (1,185 µM) [1]. ECA was shown to inhibit late viral compounds appear to merit further study, particularly gene expression, with no effect on other virus targets, as no data of their activities in animal models have been such as virus entry into the cell, early gene expression reported. Because of the need for antipoxvirus agents, or viral DNA synthesis [2]. ECA caused a 90% reduc- owing to bioterrorism concerns surrounding the delib- tion in vaccinia virus yield at concentrations of 19–61 erate release of smallpox or monkeypox viruses into µM, depending upon cell type. Cytotoxicity was evi- human populations [5,6], studies to identify new agents dent at 35–290 µM. Selectivity index (SI) values of that have potential for treating humans are warranted.

These studies include the determination of compound MK2 and MA-104 cells) foetal bovine serum (FBS). efficacy in animals [7,8]. A549 cells required Ham’s F12K medium with 0.22% From a review of the published literature [1,2], it was sodium bicarbonate and 10% FBS. The medium for all our impression that ATA would be too weak in potency antiviral assays was MEM with bicarbonate, 2% FBS, in vitro to show efficacyin vivo. Although ECA is more and 50 µg/ml of gentamicin. potent than ATA, we were concerned that ECA might be poorly selective (that is, the 50% antiviral and cyto- Plaque reduction assays toxic activities might not be far enough apart). Thus, Sensitivities of cowpox and vaccinia viruses to ECA we performed some cell culture analyses with ECA to were determined in 12-well microplates of cells [9,10]. confirm our hypothesis. We felt that the antiviral activ- The cells were infected with about 75–100 plaque ity of ATA was sufficiently documented not to pursue forming units (PFU) of virus per well, the virus was further in vitro studies with that material. Next, we per- adsorbed for 1 h, then twofold dilutions of antiviral formed animal experiments using intranasally infected compounds were applied for 3 days. Cells were fixed mice treated with ATA and ECA and made comparisons and stained in 5% buffered formalin containing 0.2% with cidofovir treatment. The lack of utility of ATA and crystal violet for 15 min. The aspirated plates were ECA in treating vaccinia virus infections in mice was rinsed with water and plaques were counted. Concen- clearly demonstrated. trations of compounds reducing plaque numbers by

50% (50% effective concentration [EC50] values) were Methods determined by plotting the percentage of plaques rela- tive to untreated cultures versus inhibitor concentra- Antiviral compounds tion on semi-log10 paper. ATA and ECA were purchased from Sigma (St. Louis, MO, USA). Cidofovir was kindly provided by M Hitch- Cytotoxicity assays cock of Gilead Sciences (Foster City, CA, USA). The ECA was overtly toxic to uninfected cells at particular acids were dissolved in 2% sodium bicarbonate solu- concentrations. Percent cytotoxicity was quantified by tions. ECA was further diluted into cell culture medium staining cell monolayers in 12-well plates with crystal for studies conducted in vitro. Prior to discovering a violet after 3 days of ECA treatment. After 15 min, suitable solvent for the compounds (ECA being the first the cells were rinsed of excess dye and the plates were one examined), animal experiments were performed allowed to dry. The dye was eluted from the cells with ECA suspended in saline for treatment of mice. with ethanol and quantified spectrophotometrically ATA was dissolved in 2% sodium bicarbonate buffer at 590 nm [11]. Absorbance values were converted to for mouse experiments. Cidofovir, a positive control, percentages of the untreated control. The 50% cyto- was dissolved in saline. The placebo for animal studies toxic concentration (CC50 value) was determined as was saline for the evaluation of ECA or 2% sodium described above. bicarbonate for the study of ATA. Mouse experiments Viruses and cells Specific pathogen-free BALB/c mice weighing 14–15 g Vaccinia virus IHD and WR strains were purchased were obtained from Charles River Labs (Wilmington, from the American Type Culture Collection (ATCC, MA, USA). Mice were infected intranasally with 50 µl of Manassas, VA, USA). Cowpox virus (Brighton strain) vaccinia (IHD strain) virus (approximately 1×105 PFU/ was obtained from the US Army Medical Research mouse) following anaesthesia with ketamine (100 mg/ Institute of Infectious Diseases (Fort Detrick, Fre- kg, by intraperitoneal [ip] injection). Treatments with derick, MD, USA). The origin of the cowpox virus compounds were given by ip injection starting 24 h was the Centers for Disease Control and Prevention after virus exposure. The treatment schedules and doses (Atlanta, GA, USA). The viruses were initially prop- of antiviral compounds that were used for the experi- agated in African green monkey kidney (Vero) cells ments were based upon our previous work [12]. The (ATCC). Higher virus titres for later use were obtained animals were kept for 21 days to record deaths and to by propagating the viruses in a second line of African monitor body weight. green monkey kidney (MA-104) cells, purchased from BioWhittaker (Walkersville, MD, USA). Other cells Statistical analyses used for in vitro assays included human lung carci- Survival curves of data from mouse studies were noma (A549) and rhesus monkey kidney (LLC-MK2) assessed for all groups by the Mantel–Cox log rank test. cells. The three monkey cell lines were propagated in Because statistical significance among groups was seen minimal essential medium (MEM) containing 0.22% (P<0.001), pairwise comparisons were made using the sodium bicarbonate and 5% (Vero cells) or 10% (LLC- Gehan–Breslow–Wilcoxon test. Evaluations were made

using Prism software (GraphPad Software, San Diego, ECA group. All other groups of ATA- and ECA-treated CA, USA). mice showed increases in body weight. Based on the results of these experiments, it was decided to test each Results compound for antiviral activity at a maximum dose of 30 mg/kg/day. Cell culture experiments Antiviral studies were conducted in mice infected The variability in antiviral potency and SI values with vaccinia virus (IHD strain). This infection evolves reported previously for ECA [2] prompted us to carry slower than infections with the WR strain of vaccinia, out our own investigation into the antiviral activity of and antiviral treatment is more effective [12]. Thus, the compound, using one human and three monkey any positive benefit of treatment with ATA or ECA cell lines. Table 1 shows the results of plaque reduction would be evident. Mice infected intranasally were experiments. Cowpox virus was inhibited at 40–79 µM treated with the acids or with cidofovir starting 1 day and vaccinia virus was inhibited at 20–46 µM. Toxic- after infection (Figure 1). Treatment with ATA did not ity was observed in the cells at 84–173 µM, yielding SI prevent death. The 30 mg/kg/day dose accelerated the values of 1.3–4.2. Thus, ECA was a moderately potent time to death compared with placebo. Cidofovir treat- and poorly selective inhibitor of orthopoxviruses. ment for 2 days was 100% protective. Treatment with ECA also failed to prevent death and treatment with Mouse studies both doses shortened the time to death. Cidofovir was Prior to conducting antiviral experiments in mice, it was completely protective. important to determine the maximum tolerated doses of ATA and ECA in uninfected animals (Table 2). The com- Discussion pounds were evaluated by ip treatment route at doses of 3–100 mg/kg/day. Both ATA and ECA were overtly In these experiments, two compounds purported to toxic at 100 mg/kg/day, with rapid death occurring in have potential as drug candidates or lead compounds the ECA group. Mice treated with ATA for 5 days did for new directions in the development of drugs effective not die until after the last treatment was given. Weight against poxvirus infections were evaluated for antiviral loss without death was observed in the 30 mg/kg/day activity. ECA was found to be poorly selective in cell

Table 1. Antiviral activity and cytotoxicity of ECA in four cell lines

Cowpox virus Vaccinia virus (WR strain) a b c b c Cell line CC50 EC50 SI EC50 SI

A549 173 79 2.2 46 3.8

LLC-MK2 121 61 2.0 40 3.0 MA-104 87 66 1.3 39 2.2 Vero 84 40 2.1 20 4.2

a b CC50, 50% cytotoxic concentration (µM) assessed in confluent cell monolayers. EC50, 50% effective (virus inhibitory) concentration (µM) determined by plaque c reduction assays. SI, selectivity index (CC50/EC50). All values are means for two independent experiments. ECA, ethacrynic acid.

Table 2. Toxicities of ATA and ECA in uninfected mice

Compounda Doseb, mg/kg/day Survivors/total Mean host weight changec, g

ATA 100 0/6 (6.5 ±1.5)d -0.9 30 5/5 +0.8 10 5/5 +0.5 3 5/5 +0.6 Placebo 1 – 5/5 +0.8 ECA 100 0/5 (1.6 ±0.5) – e 30 5/5 -1.4 10 5/5 -0.1 3 5/5 -0.1 Placebo 2 0 5/5 +0.5 aThe compounds were evaluated independently in separate experiments. bDoses were given twice daily 12 h apart for 5 days. cBody weights were determined 1 day after the last treatment. dMean day of death ±s d . eMice were all dead prior to weighing. ATA, aurintricarboxylic acid; ECA, ethacrynic acid.

Antiviral Chemistry & Chemotherapy 20.5 203 DF Smee et al.

Figure 1. Effects of ATA, ECA and cidofovir on survival of a vaccinia virus (IHD strain) respiratory infection in mice

100 a 100 a

80 80

60 60

40 40 Survival, % Survival, %

20 20

b ac 0 0 036912 15 18 21 036912 15 18 21 Time after virus exposure, days Time after virus exposure, days

ATA (30 mg/kg/day) Cidofovir (100 mg/kg/day) ECA (30 mg/kg/day) Cidofovir (100 mg/kg/day) ATA (10 mg/kg/day) Placebo ECA (15 mg/kg/day) Placebo

Treatments with aurintricarboxylic acid (ATA) and ethacrynic acid (ECA) were given twice daily for 5 days. Cidofovir was administered intraperitoneally once daily for 2 days. Treatments began 24 h after virus exposure. The compounds were evaluated in two separate experiments. Each compound-treated and placebo group had 10 and 20 mice, respectively. aP<0.001, bP<0.01, cP<0.05. culture infected with cowpox and vaccinia viruses. The report that the related cowpox virus is also inhibited, results confirmed those reported previously [2], if SI which is not surprising. ATA has been reported to values are calculated based on 90% (10-fold) reduction inhibit a number of different viruses, including dengue in virus titre rather than the published method (50% or and Wesselbron flaviviruses [13], HCV [14], HIV [15], twofold reduction in virus titre). We have found that influenza virus [16], severe acute respiratory syndrome

EC90 values for virus titre reduction correlate well with coronavirus [17], vaccinia virus [1] and vesicular sto- the EC50 value for plaque reduction (DFS and BLH, matitis virus [18]. The possibility exists that ATA could unpublished), which was the assay that we performed. inhibit some of these viruses in vivo wherein the anti- A recalculation of the previous data [2] based upon viral potency is considerably better (that is, lower 50%

EC90/IC50 (or CC50) gave SI values of <1–9.9, compared virus inhibitory concentrations). The studies presented with 1.3 and 4.2 reported here. Thus, our results are here establish the maximum tolerated dose of the com- consistent with the former results for the moderate pounds in mice and show that they are not effective potency and low selectivity of ECA against vaccinia against vaccinia virus infections. virus in cell culture. The low potency of ATA (60–475 µM) and the low Acknowledgements selectivity of ECA (<1–9.9) in cell culture were initial indicators that, despite their uniquely identified modes This work was supported by contracts NO1-AI-15435 of action, they might not exhibit antiviral activity and NO1-AI-30063 (awarded to Southern Research in vivo. Indeed, the compounds exhibited no protective Institute) from the Virology Branch, National Institute activity in mice infected intranasally with vaccinia (IHD of Allergy and Infectious Diseases, National Institutes strain) virus. In fact, mice treated with the compound of Health, Bethesda, MD, USA. The contents of this fared worse (that is, died sooner) than animals treated article do not necessarily reflect the position or policy of with placebo. ECA showed some toxicity in uninfected the government and no official endorsement should be mice at the 30 mg/kg/day dose by inducing weight loss, inferred. The investigators adhered to the ‘Guide for the and this may have contributed to early death. Cidofovir Care and Use of Laboratory Animals’, prepared by the was highly effective in this model, as was previously Committee on Care and Use of Laboratory Animals of reported [10]. the Institute of Laboratory Animal Resources, National To date, ECA has only been reported to be inhibitory Research Council and used facilities fully accredited by to poxvirus replication, specifically against vaccinia virus the Association for Assessment and Accreditation of [2], but not other related or unrelated viruses. Here, we Laboratory Animal Care International.

10. Smee DF, Humphreys DE, Hurst BL, et al. Antiviral Disclosure statement activities and phosphorylation of 5-halo-2′-deoxyuridines and N-methanocarbathymidine in cells infected with The authors declare no competing interests. vaccinia virus. Antivir Chem Chemother 2008; 19:15–24. 11. Smee DF, Morrison AC, Barnard DL, et al. Comparison of colorimetric, fluorometric, and visual methods for References determining anti-influenza (H1N1 and H3N2) virus activities and toxicities of compounds. J Virol Methods 1. Myskiw C, Deschambault Y, Jefferies K, et al. 2002; 106:71–79. Aurintricarboxylic acid inhibits the early stage of vaccinia 12. Smee DF, Wong MH, Bailey KW, et al. Effects of four virus replication by targeting both cellular and viral factors. antiviral substances on lethal vaccinia virus (IHD strain) J Virol 2007; 81:3027–3032. respiratory infections in mice. Int J Antimicrob Agents 2. Spisakova M, Cizek Z, Melkova Z. Ethacrynic and alpha- 2004; 23:430–437. lipoic acids inhibit vaccinia virus late gene expression. Antiviral Res 2009; 81:156–165. 13. Milani M, Mastrangelo E, Bollati M, et al. Flaviviral methyltransferase/RNA interaction: structural basis for 3. Magee WC, Aldern KA, Hostetler KY, et al. Cidofovir and enzyme inhibition. Antiviral Res 2009; 83:28–34. (S)-9-[3-hydroxy-(2-phosphonomethoxy)propyl]adenine are highly effective inhibitors of vaccinia virus DNA polymerase 14. Chen Y, Bopda-Waffo A, Basu A, et al. Characterization when incorporated into the template strand. Antimicrob of aurintricarboxylic acid as a highly potent hepatitis C Agents Chemother 2008; 52:586–597 virus replicase inhibitor. Antivir Chem Chemother 2009; 20:19–36. 4. Yang G, Pevear DC, Davies MH, et al. An orally bioavailable antipoxvirus compound (ST-246) inhibits 15. Balzarini J, Mitsuya H, De Clercq E, et al. extracellular virus formation and protects mice from lethal Aurintricarboxylic acid and Evans Blue represent two orthopoxvirus challenge. J Virol 2005; 79:13139–13149. different classes of anionic compounds which selectively 5. Leggiadro RJ. Bioterrorism: a clinical reality. Pediatr Ann inhibit the cytopathogenicity of human T-cell lymphotropic 2007; 36:352–358. virus type III/lymphadenopathy-associated virus. Biochem Biophys Res Commun 1986; 136:64–71. 6. Strikas RA, Neff LJ, Rotz L, et al. US civilian smallpox preparedness and response program, 2003. Clin Infect Dis 16. Hung HC, Tseng CP, Yang JM, et al. Aurintricarboxylic 2008; 46 Suppl 3:S157–S167. acid inhibits influenza virus neuraminidase. Antiviral Res 2009; 81:123–131. 7. Smee DF, Sidwell RW. A review of compounds exhibiting anti-orthopoxvirus activity in animal models. Antiviral Res 17. He R, Adonov A, Traykova-Adonova M, et al. Potent and 2003; 57:41–52. selective inhibition of SARS coronavirus replication by aurintricarboxylic acid. Biochem Biophys Res Commun 8. Smee DF. Progress in the discovery of compounds 2004; 320:1199–1203 inhibiting orthopoxviruses in animal models. Antivir Chem Chemother 2008; 19:115–124. 18. Hunt DM, Wagner RR. Inhibition by aurintricarboxylic 9. Smee DF, Wandersee MK, Bailey KW, et al. Cell line acid and polyethylene sulfonate of RNA transcription of dependency for antiviral activity and in vivo efficacy of vesicular stomatitis virus. J Virol 1975; 16:1146–1153. N-methanocarbathymidine against orthopoxvirus infections in mice. Antiviral Res 2007; 73:69–77. Received 11 September 2009, accepted 14 October 2009

Antiviral Chemistry & Chemotherapy 20.5 205