key: cord-0770016-1ln02cyq
authors: Balzarini, Jan; Keyaerts, Els; Vijgen, Leen; Vandermeer, Frank; Stevens, Miguel; De Clercq, Erik; Egberink, Herman; Van Ranst, Marc
title: Pyridine N-oxide derivatives are inhibitory to the human SARS and feline infectious peritonitis coronavirus in cell culture
date: 2005-12-29
journal: J Antimicrob Chemother
DOI: 10.1093/jac/dki481
sha: fc0b145cd796b7f52c798192d1f206568c74fe00
doc_id: 770016
cord_uid: 1ln02cyq

Objectives: Evaluation of a wide variety of pyridine N-oxide derivatives on their inhibitory activity against feline coronavirus (FIPV strain) and human SARS-CoV (Frankfurt strain-1) in cell culture. Methods: FIPV and SARS-CoV were exposed to confluent Crandel feline kidney (CRFK) and simian kidney (Vero) cell cultures in the presence of serial concentrations of the test compounds. The anti-cytopathic activity of the pyridine N-oxide derivatives was monitored by spectrophotometric analysis. Results and conclusions: A wide variety of pyridine N-oxide derivatives have been found to be inhibitory against feline coronavirus (FIPV strain) and human SARS-CoV (Frankfurt strain-1) in CRFK and simian kidney (Vero) cell cultures, respectively. The oxide part on the pyridine moiety proved indispensable for anti-coronavirus activity. The potency and virus specificity of the pyridine N-oxide derivatives varied depending the nature and specific location of substituents (i.e. alkyl, halogeno, nitro, etc.) on the different parts of the molecule. The most selective compounds were active in the higher microgram per litre range, being non-toxic at 50–100 mg/L. There was a poor structure-antiviral activity relationship (SAR) for the pyridine N-oxide derivatives against Fe-CoV and SARS-CoV. One of the most active and selective compounds was shown to inhibit Fe-CoV replication at the transcriptional level.

The discovery of a novel human coronavirus (H-CoV) as the cause of the newly recognized severe acute respiratory syndrome (SARS) 1-4 provides a new challenge to the medical community to keep control on this disease. Although human coronaviruses cause up to 30% of colds, they rarely cause a lower respiratory tract disease, and never have a devastating effect as seen with the SARS-CoV. 5 In contrast, animal coronaviruses are known to cause devastating epizootics of respiratory or enteric diseases in livestock and poultry. 5 In fact, coronaviruses have been isolated from avian, porcine, feline, murine, bovine and canine species. 5 All known coronaviruses are categorized in three serologically well-defined and unrelated groups. The SARS coronavirus is clearly new to the human population and its RNA genome differs substantially from sequences of all known coronaviruses. 6, 7 The natural host of the SARS-CoV is most likely a Civet cat, whose virus had acquired the ability to infect humans. Interestingly, SARS-CoV can be readily isolated and grown in monkey kidney Vero cell cultures. 8 Vaccines are available for some animal coronaviruses. However, although vaccination with live, attenuated virus is effective against porcine epidemic diarrhoea virus and avian infectious bronchitis virus, recombination of the genome of vaccine strains with wild-type coronavirus is a potential risk when applied in humans. 5 Moreover, some vaccines against feline coronaviruses have been proven to enhance disease when vaccinated animals were exposed to wild-type virus, and thus antibody enhancement of disease is a potential risk of SARS-CoV vaccines in humans. 5 Therefore, it is prudent to develop safe and effective drugs against SARS-CoV as quickly as possible in case a novel wide-spread outbreak would occur. The development of effective drugs against SARS-CoV may also provide new strategies for the prevention or treatment of other coronavirus diseases in animals or humans. Indeed, there are no approved drugs with proven efficacy against coronaviruses. Ribavirin has been given to SARS patients but its efficacy is unclear. Several other compounds have been recently mentioned as potential drug leads against SARS-CoV. 9 In this study, we evaluated a variety of 192 compounds that all belonged to the class of the pyridine N-oxide derivatives ( Figure 1 ) against both SARS-CoV and the type II strain of feline infectious peritonitis virus (FIPV). FIPV virus causes a severe disease in cats characterized by vasculitis and disseminated pyogranulomatous lesions in various tissues and organs. The pyridine N-oxide compounds have previously been demonstrated to be inhibitory against the human immunodeficiency virus (HIV) in cell culture. [16] [17] [18] [19] Several of the pyridine N-oxide derivatives have been demonstrated to act at a post-integrational event in the replication cycle of HIV, i.e. HIV gene expression. 18 Prolonged exposure of one of the prototype compounds (JPL-32) to DBA/2 and SCID mice demonstrated lack of acute toxicity. Moreover, a preliminary efficacy experiment showed protective activity against HIV-induced destruction of CD4-positive human T-lymphocytes in SCID mice. 19 Our aim in the present study was to reveal whether pyridine N-oxide derivatives are endowed with inhibitory activity against SARS-CoV and FIPV, and whether the antiviral potencies of the pyridine N-oxide derivatives previously reported for HIV could be correlated with their antiviral activity against SARS-CoV and FIPV.

The structures of the compounds have been reported previously. 16 

The SARS-CoV (Frankfurt-1 strain) was kindly provided by Prof. Dr H. F. Rabenau (Johann Wolfgang Goethe University, Frankfurt, Germany). Vero E6 cells were propagated in minimal essential medium (MEM; Gibco Life Technologies, Rockville, MD, USA) supplemented with 10% fetal calf serum (FCS; Integro, Zaandam, The Netherlands), 2 mM L-glutamine (Gibco) and 1.4% sodium bicarbonate (Gibco Life Technologies, Rockville, MD, USA). Virus-infected cells were maintained at 37 C in a 5% CO 2 atmosphere in MEM supplemented with 2% FCS. The Fe-CoV (FIPV strain 1146) was originally isolated by McKeirnan et al. 20 and propagated from Crandel feline kidney (CRFK) cells maintained in RPMI-1640 medium (Gibco) and supplemented with 10% fetal bovine serum (Harlan Sera-Lab Ltd, Loughborough, UK), 2 mM L-glutamine (Gibco) and 0.075% sodium bicarbonate (Gibco). Virus-infected cells were maintained at 37 C in RPMI-1640 medium supplemented with 2% FCS.

Antiviral activity and cytotoxicity measurements were based on the viability of Vero cells that had been infected or mock-infected with 100 CCID 50 (50% cell culture infective dose) of the SARS-CoV or CRFK cells infected or mock-infected with 100 CCID 50 FIPV in the presence of various concentrations of the test compounds. The virus-drug mixture was not removed after the adsorption phase of the virus infection. The compounds were present throughout the whole time period of the experiment. Three days (SARS-CoV) or four days (Fe-CoV) after the infection, the number of viable cells was quantified by a tetrazolium-based colorimetric method as previously described for HIV by Pauwels et al. 21 The medium was aspirated and replaced by a solution of MTT [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulphophenyl)-2H-tetrazonium]/PMS (phenazinemethosulphate) according to the manufacturer's instructions (Promega Corporation, Madison, WI, USA). The results are given as the mean (-SD) of at least 2-3 independent experiments. The cytotoxic concentration was determined as the concentration of the compound that reduced cell viability by 50% (CC 50 or 50% cytotoxic concentration), and the antivirally effective concentration was determined as the compound concentration that prevented the viral cytopathic effect by 50% of the control value (EC 50 or 50% effective concentration). Thus, both the CC 50 and the EC 50 values were calculated from the absorbance values, and the EC 50 s were obtained after deduction of the absorbance value of the control cultures from the absorbance value of the virus-infected cultures.

To reveal at what stage in the infection cycle the compounds are inhibitory, a time-of-addition (TOA) experiment was carried out in which CRFK cell cultures were infected with a high dose ( ‡1000 CCID 50 ) of FIPV. The pyridine N-oxide 45 was used at 50 mg/L. At different time points post-infection (i.e. 0, 1, 2.5, 5, 7, 14, 25, 34, 49 and 72 h), the drug was added to the infected cell cultures. After 72 h, full cytopathic activity was noticed for the control cultures. Optical reading of the MTT-exposed cell cultures was then performed to quantify the protective effect of the test compounds added at the different time points against virus-induced cytopathicity.

A variety of substituted pyridine N-oxide derivatives have been evaluated in which a sulphone (SO 2 ), a sulphoxide (SO) or a thio ether (S) links the unsubstituted pyridine N-oxide group through a CH 2 to the phenyl moiety ( Table 1 ). The anti-coronavirus activities ranged between 0.29 and >100 mg/L for FIPV and between 4.2 and >100 mg/L for SARS-CoV depending on the nature and the locations of the substituents in the phenyl moiety. As a rule, when antiviral activity was recorded, the anti-FIPV activity was always more pronounced than the anti-SARS-CoV activity. Compounds that were weakly inhibitory against FIPV did not show detectable antiviral activity against SARS-CoV. Generally, sulphide and sulphoxide derivatives showed poor, if any, antiviral activity against both coronaviruses except for 27 and 35 that were inhibitory to Fe-CoV (EC 50 : <10 mg/L), but not to SARS-CoV (EC 50 : >100 mg/L). The position of one or more alkyl or alkoxy groups on the phenyl moiety does not play a marked role in the eventual anti-coronavirus potency of the test compounds (compare compounds 4-35). This was generally also observed for the halogen-, nitro-and cyano-substituted compounds. The most potent antiviral activity was noted for the trichloro (44), pentachloro (45), methyl/tetrachloro (46) and nitro (52) derivatives Figure 1 . Basic structure of the pyridine N-oxide derivatives. The 'n' annotation indicates that the carbon that is linked to the aryl, R 2 and pyridinesulphone moieties can be extended to a longer alkyl chain. Table 1 . Antiviral activity of pyridine N-oxide derivatives modified in the phenyl moiety However, it should be noticed that these compounds proved markedly cytotoxic to both CRFK and Vero cell cultures, and therefore potential anti-FIPV activity may have been masked by the cellular toxicity or, alternatively, the observed anti-SARS-CoV activity has to be interpreted as a toxic rather than a specific antiviral effect. For all other antivirally active compounds-as already noted for the unsubstituted series of compounds discussed above-anti-FIPV activity was more pronounced than the anti-SARS-CoV activity. Compound 72 represents the only exception where anti-SARS-CoV activity was noticed (EC 50 : 13 mg/L) without a trace of anti-FIPV activity. Intriguingly, this compound was toxic for CRFK cells (CC 50 : 1.3 mg/L) but not for Vero cells (CC 50 : >100 mg/L).

While the 116and 122-sulphide derivatives were markedly more active against FIPV than SARS-CoV, the dihalogen-substituted pyridine N-oxide derivative 120 was 4-fold more effective against SARS-CoV than FIPV (Table 3) . Also, 117 and 118 showed antiviral activity against SARS-CoV but not FIPV. Intriguingly, the active compounds in this series of pyridine N-oxide derivatives contained exclusively a sulphide moiety whereas the sulphones 119 and 121 were entirely devoid of antiviral activity. An opposite trend of (in)activity with respect to sulphones versus sulphides and sulphoxides was previously found for the first and second series of compounds depicted in Tables 1 and 2 . Anti-coronavirus activity of pyridine N-oxide derivatives modified at the pyridine oxide and phenyl and Z part of the molecule Among the pyridine N-oxide derivatives that contain alkyl, alkoxy, halogen or nitro substituents on the pyridine moiety in addition to substituents on the phenyl and the bridge between the pyridine thioether and the phenyl, a few compounds were endowed with a pronounced anti-FIPV selectivity (Table 4 ). Indeed, 155 had an EC 50 of 0.49 mg/L against FIPV being inactive against SARS-CoV. This compound was not cytotoxic at 100 mg/L, and therefore represents the most selective anti-FIPV compound among all pyridine N-oxide derivatives tested. The location of the chloro substituent on the pyridine moiety seems crucial, since moving the chloro from Y 1 to the Y 2 , Y 3 or Y 4 position resulted in completely inactive compounds (158, 161, 165) . Besides 155, also 132, 133, 142 and 146 can be regarded as highly selective anti-Fe-CoV compounds. As observed before, none of the compounds was more inhibitory against SARS-CoV than against Fe-CoV. Compounds 157 and 159 were the most interesting compounds that had comparable (potent) activity against both SARS-CoV and FIPV (EC 50 : 1.7-4.2 mg/L), being poorly cytotoxic (157; CC 50 : 62 mg/L for CRFK and >100 mg/L for Vero cell cultures) to moderately cytotoxic (159; CC 50 : 13 mg/L for CRFK and 71 mg/L for Vero cell cultures) (Table 4) .

A total of 23 compounds that lacked the oxygen at the N-atom of the pyridine moiety were evaluated for antiviral activity ( Table 5) . As a rule, lack of the oxide moiety proved detrimental for anti-SARS-CoV and anti-FIPV activity. Indeed, virtually none of the test compounds was antivirally active at subtoxic concentrations. Table 3 . Antiviral activity of pyridine N-oxide derivatives containing an ethylene bridge between the thioether and the phenyl moiety Table 4 . Antiviral activity of pyridine N-oxide derivatives modified in the pyridine oxide and phenyl and Z part of the molecule Table 5 . Antiviral activity of pyridine derivatives

EC 50 a (mg/L)

Only 173 (FIPV), 184 (SARS-CoV) and 185 (SARS-CoV) showed antiviral activity, although their EC 50 s were rather close to their CC 50 s.

The pyridine N-oxide derivatives represent a unique class of antivirals. Several members have previously been found to be active against HIV-1 and/or HIV-2, and human cytomegalovirus (HCMV), but not against several other DNA viruses including herpes simplex virus type 1 and type 2, varicella-zoster virus and vaccinia virus or against RNA viruses including vesicular stomatitis virus, reovirus-1, polio virus, Coxsackie virus B4, Semliki forest virus and parainfluenza virus. 16 Surprisingly, the pyridine N-oxides were now also found inhibitory against coronaviruses, in particular against the feline coronavirus type II strain of FIPV and the SARS coronavirus strain Frankfurt-1. Such an antiviral selectivity spectrum is rather unusual. Interestingly, as previously noted between HIV-1, HIV-2 and HCMV, there is no close correlation for the antiviral activity of the pyridine N-oxides against HIV-1 on the one hand, and FIPV and SARS-CoV on the other hand (r = 0.29 and 0.47, respectively) (compare the antiviral activities of the test compounds against HIV-1 with those found against the coronaviruses given in Tables 1-5) . Also, there is no correlation for the EC 50 s between FIPV and SARS-CoV in cell culture (r = 0.06). Indeed, there were a number of compounds that proved exclusively inhibitory to FIPV and not to SARS-CoV. Among these, only 63 and 116 were not active against HIV-1. The other compounds that discriminated between FIPV and SARS-CoV showed also activity against HIV-1. The pyridine N-oxide derivatives have a peculiar mechanism of antiviral action. It has previously been shown that pyridine N-oxide derivatives such as 11, 17, 45 and 160 act at a step in the HIV-1 and HIV-2 replication cycle that follows proviral integration, i.e. at the HIV gene expression level. 18 In this respect, the pyridine N-oxide derivatives inhibit binding of nuclear NF-kB to DNA in the intact cell system and regeneration of IkBa after TNF-a stimulation. 18, 19 Thus, targeting of a cellular protein in the transactivation process is the most likely explanation for the anti-HIV/HCMV activity of the pyridine N-oxide derivatives. The molecular target for the anticoronavirus activity is still unclear. Coronaviruses replicate in the cytoplasm, and not in the nucleus. However, a TOA experiment in which the pyridine N-oxide derivative 45 was added at different time points after Fe-CoV infection of CRFK cell cultures revealed that addition of the compound to the virus-infected cells could be substantially delayed before losing its antiviral potential (Figure 2) . A similar TOA experiment against HIV in human lymphocyte CEM cell cultures revealed also a post-integrational event as a target of 45 in the replication cycle of HIV. 18 Thus, compound 45 inhibits virus replication at an event that is clearly located after viral entry and acts most likely during the transcription process of the virus replication cycle. However, not only the differences in antiviral activity but also cytotoxicity depending on the virus type (HIV, Fe-CoV, SARS-CoV) or cell type (CEM, CRFK, Vero) is indicative for a rather specific interaction with a cellular and/or a viral factor to explain the rather unpredictable antiviral and cytostatic properties of the pyridine N-oxide derivatives. In this respect, while the oxide form of the compounds seems to be crucial to maintain anti-coronavirus activity, this requirement was much less stringent to keep anti-HIV-1 activity (compare anti-HIV and anti-coronavirus activity for the reduced pyridine N-oxides in Table 5 ).

Despite the pronounced cytostatic activity noticed in proliferating human CEM lymphocyte cell culture, pyridine N-oxide derivatives are often poorly cytotoxic in confluent CRFK and Vero cell cultures, resulting in pronounced selectivity indices for several compounds (exceeding two orders of magnitude). Moreover, compound 45, which represents one of the most inhibitory and selective anti-coronavirus compounds in this study (lacking any pronounced selectivity against HIV-1 in CEM cell cultures), has been administered to mice at 15 mg/kg/ day for 10 days (upon continuous release through Alzet pumps) or at 100 mg/kg/day intraperitoneally for 10 consecutive days. Under these experimental conditions, 45 did not result in any visible signs of toxicity or side effects in the drug-exposed animals. These observations may justify more in-depth studies on the pharmacokinetics of 45 and related compounds and on the potential of pyridine N-oxide derivatives as anti-coronavirus drugs in vivo.

In conclusion, the selective anti-coronavirus activity found for several pyridine N-oxide derivatives in cell culture may warrant further pre-clinical investigations to reveal the potential of this class of antiviral drugs as selective inhibitors of coronaviruses, including SARS-CoV infection in humans and Fe-CoV infection in cats. 90001) and by the European Research Project SARS-DTV (contract no. SP22-CT-2004).

The authors have no financial or any other conflict of interest regarding the contents of the investigations. 

Identification of a novel coronavirus in patients with severe acute respiratory syndrome

A novel coronavirus associated with severe acute respiratory syndrome

Newly discovered coronavirus as the primary cause of severe acute respiratory syndrome

Coronavirus as a possible cause of severe acute respiratory syndrome

SARS-associated coronavirus

Characterization of a novel coronavirus associated with severe acute respiratory syndrome

The genome sequence of the SARS-associated coronavirus

Proliferative growth of SARS coronavirus in Vero B6 cells

Treatment of severe acute respiratory syndrome with lopinavir/ritonavir: a multicentre retrospective matched cohort study

Role of interferons in the treatment of severe acute respiratory syndrome

Glycyrrhizin, an active component of liquorice roots, and replication of SARS-associated coronavirus

Inhibition of severe acute respiratory syndrome coronavirus replication by niclosamide

HIV protease inhibitor nelfinavir inhibits replication of SARS-associated coronavirus

Inhibition of SARS-coronavirus infection in vitro by S-nitroso-N-acetylpenicillamine, a nitric oxide donor compound

In vitro inhibition of severe acute respiratory syndrome coronavirus by chloroquine

Pyridine oxide derivatives: structure-activity relationship for inhibition of human immunodeficiency virus and cytomegalovirus replication in cell culture

Inhibition of human immunodeficiency virus by a new class of pyridine oxide derivatives

Novel human immunodeficiency virus (HIV) inhibitors that have a dual mode of anti-HIV action

Pyridine N-oxide derivatives: unusual anti-HIV compounds with multiple mechanisms of antiviral action

Isolation of feline coronaviruses from two cats with diverse disease manifestations

Rapid and automated tetrazolium-based colorimetric assay for the detection of anti-HIV compounds