Activities of Several Classes of Acyclic Nucleoside Phosphonates against Camelpox Virus Replication in Different Cell Culture Models

Abstract

Camelpox virus (CMLV) is the closest known virus to variola virus. Here we report on the anti-CMLV activities of several acyclic nucleoside phosphonates (ANPs) related to cidofovir [(S)-1-(3-hydroxy-2-phosphonomethoxypropyl)cytosine (HPMPC; Vistide)] against two CMLV strains, CML1 and CML14. Cytopathic effect (CPE) reduction assays performed with human embryonic lung fibroblast monolayers revealed the selectivities of the first two classes of ANPs (cHPMPA, HPMPDAP, and HPMPO-DAPy) and of the hexadecyloxyethyl ester of 1-{[(5S)-2-hydroxy-2-oxido-1,4,2-dioxaphosphinan-5-yl]methyl}-5-azacytosine (HDE-cHPMP-5-azaC), belonging to the newly synthesized ANPs, which are HPMP derivatives containing a 5-azacytosine moiety. The inhibitory activities of ANPs against both strains were also confirmed with primary human keratinocyte (PHK) monolayers, despite the higher toxicity of those molecules on growing PHKs. Virus yield assays confirmed the anti-CML1 and anti-CML14 efficacies of the compounds selected for the highest potencies in CPE reduction experiments. Ex vivo studies were performed with a 3-dimensional model of human skin, i.e., organotypic epithelial raft cultures of PHKs. It was ascertained by histological evaluation, as well as by virus yield assays, that CMLV replicated in the human skin equivalent. HPMPC and the newly synthesized ANPs proved to be effective at protecting the epithelial cells against CMLV-induced CPE. Moreover, in contrast to the toxicity on PHK monolayers, signs of toxicity in the differentiated epithelium were seen only at high ANP concentrations. Our results demonstrate that compounds belonging to the newly synthesized ANPs, in addition to cidofovir, represent promising candidates for the treatment of poxvirus infections.

Camelpox virus (CMLV), the etiologic agent of camelpox, belongs to the family Poxviridae, genus Orthopoxvirus. Camelpox occurs naturally in Old World camels including Camelus dromedarius and Camelus bactrianus (43). This viral infectious disease has been reported throughout areas of Africa, north of the Equator, the Middle East, and Asia, and outbreaks of CMLV infection in camelids can be responsible for severe economic losses in these countries (22, 27). Camelids may become infected via small abrasions of the skin, by aerosol infection of the respiratory tract, or by arthropod bites. Clinically, two distinct forms can be distinguished: the severe generalized form (mostly among young animals) and the milder localized form (mostly in old camels). In both forms, initial multiplication of the virus occurs at the site of entry. In systemic disease, further viral multiplication in the draining lymph nodes is followed by a primary viremia and viral replication in organs and tissues (43). This results in a second viremia and subsequent infection of the skin. Eruptions over the entire body are found in generalized forms, in which the mortality rate can reach 28% (26). In the milder forms, pustules appear on the nostrils and eyelids, and on the oral and nasal mucosae. Camelpox is most probably not a zoonosis. Skin eruptions of camel herdsmen for whom CMLV was clinically highly suspected but not confirmed microbiologically have been reported (26).

In 2002, a comparative study of the CMLV and variola virus (VARV) genomes showed that of all poxviruses, CMLV is the orthopoxvirus most closely related to VARV, the causative agent of smallpox (23). These two viruses not only share colinearity in their genomes but also have similar abilities to induce high morbidity and mortality in a single host species (20). Although smallpox has been eradicated, there are concerns about the potential use of VARV in bioterrorism. Moreover, most of the world's population has become susceptible to any potential infection with poxviruses since the end of the smallpox vaccination campaign in 1978 (6). Thus, human health may be threatened by the emergence or reemergence of orthopoxviruses such as VARV and monkeypox virus (18).

Several orthopoxviruses have been extensively studied as surrogate models of VARV. This research has been based on molecular biology, genomics, diagnostics, vaccines, and antiviral drugs. The cytosine derivative cidofovir [(S)-1-(3-hydroxy-2-phosphonomethoxypropyl)cytosine (HPMPC; Vistide)], belonging to the first class of acyclic nucleoside phosphonates (ANPs), has been licensed for the treatment of cytomegalovirus retinitis in AIDS patients and is currently recommended by the U.S. Centers for Disease Control and Prevention for the treatment of severe adverse effects following smallpox vaccination (9, 14). Recently, vaccinia-immune globulins, cidofovir, and ST-246 have been used to treat a child suffering from eczema vaccinatum (31). Cidofovir is available for intravenous use as a solution, and the nephrotoxicity that may be associated with its use can be prevented by coadministration of probenecid (13). Cidofovir is a potent antiviral agent that has been shown to be active against poxviruses in cell cultures and in animal models (13). It improved the survival of mice infected with lethal doses of cowpox virus or vaccinia virus (8, 36, 38) and of monkeys with lethal respiratory monkeypox virus infections (25, 42).

Recently, two new classes of ANPs have been described: 6-[2-(phosphonomethoxy)alkoxy]-2,4-diaminopyrimidine (DAPy) derivatives, which are considered to be the second class of ANPs, and a third class of ANPs, including HPMP derivatives with a 5-azacytosine moiety (4, 16, 24, 28; M. Krečmerová, A. Holý, R. Pohl, M. Masojídková, G. Andrei, L. Naesens, J. Neyts, J. Balzarini, E. De Clercq, and R. Snoeck, submitted for publication).

In this study we evaluated the activities of more than 40 compounds selected from the three classes of ANPs against the replication of CMLV strains Iran (33) and Dubai (32) in cell monolayers, as well as in 3-dimensional (3-D) epithelial raft cultures. We used both CMLV strains, isolated from two independent outbreaks, as surrogate models with which to identify compounds potentially active against infections caused by VARV and other poxviruses.

MATERIALS AND METHODS

Cells.

Human embryonic lung (HEL) fibroblasts (HEL-299; ATCC CCL-137) were cultured in minimal essential medium with Earle's salts (Gibco, Invitrogen Corporation, United Kingdom) supplemented with 10% fetal calf serum (FCS), 1% l-glutamine, 1% nonessential amino acids, 1% sodium pyruvate, and 1% antibiotic/antimycotic at 37°C under a 5% CO₂ atmosphere. HEL cells were used for cytotoxicity and antiviral assays.

Primary human keratinocytes (PHKs) were isolated from neonatal foreskins. Tissue fragments were incubated with trypsin-EDTA for 1 h at 37°C. The growth medium was serum-free keratinocyte medium (keratinocyte-SFM) (Gibco, Invitrogen Corporation, United Kingdom) supplemented with 0.5 μg of hydrocortisone per ml, 10 ng of epidermal growth factor per ml, 2 mmol of l-glutamine per liter, 10 mmol of HEPES per liter, 1 mmol of sodium pyruvate per liter, 1 × 10⁻¹⁰ mol of cholera toxin per liter, 5 μg of insulin per ml, 5 μg of human transferrin per ml, and 15 × 10⁻⁴ mg of 3,3′,5-triiodo-l-thyronine per ml. PHKs were cultured at 37°C under a 5% CO₂ atmosphere and were used both for cytotoxicity and antiviral assays in monolayers and for organotypic raft cultures.

Swiss 3T3 J2 fibroblasts, used for organotypic raft cultures, were grown in Dulbecco's modified Eagle's medium supplemented with 10% FCS, 2 mmol l-glutamine per liter, 10 mmol HEPES per liter, and 1 mmol sodium pyruvate per liter.

Viruses.

The following viral strains were used: CMLV strain Iran (CML1) and strain Dubai (CML14). Both were kindly provided by H. Meyer (Bundeswehr Institute of Microbiology, Germany) (32, 33).

Compounds.

A list of the compounds that were evaluated for their activities against both CMLV strains is presented in Table 1. Compounds were synthesized at the Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic, Prague.

TABLE 1.

Compounds tested against CMLV

Compound no.	Compound name	Abbreviation
1a	(S)-1-[3-Hydroxy-2-(phosphonomethoxy)propyl]cytosine	HPMPC
2a	Cyclic (S)-HPMPC	cHPMPC
1b	(S)-9-[3-Hydroxy-2-(phosphonomethoxy)propyl]adenine	HPMPA
2b	Cyclic (S)-HPMPA	cHPMPA
1c	(S)-9-[3-Hydroxy-2-(phosphonomethoxy)propyl]-2,6-diaminopurine	HPMPDAP
1d	(S)-9-[3-Hydroxy-2-(phosphonomethoxy)propyl]-3-deazaadenine	3-Deaza-HPMPA
3a	9-[2-(Phosphonomethoxy)ethyl]adenine	PMEA
3b	9-[2-(Phosphonomethoxy)ethyl]-2,6-diaminopurine	PMEDAP
3c	9-[2-(Phosphonomethoxy)ethyl]-2-amino-6-chloropurine
3d	9-[2-(Phosphonomethoxy)ethyl]guanine
3e	9-[2-(Phosphonomethoxy)ethyl]-6-dimethylaminopurine
3f	9-[2-(Phosphonomethoxy)ethyl]-2-amino-6-methylpurine
4	N-(3-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}amino)propanoic acid
5	N-(4-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}amino)butanoic acid
6	N-(RS)-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}iminodiacetic acid
7	N-(RS)-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}serine
8	N-(RS)-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}aspartic acid
9	N-(RS)-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}phenylalanine
10	N-(RS)-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}leucine
11a	N-(RS)-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}threonine
11b	N-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}-(2R,3S)-threonine (l)
11c	N-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}-(2S,3R)-threonine (d)
12	N-(RS)-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}aminomethyl-sulfonic acid
13	N-(RS)-2-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}aminoethyl-sulfonic acid
14	N-(RS)-2-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}aminoethane-phosphonic acid
15	N-(RS)-2-{2-Amino-9-[2-(phosphonomethoxy)ethyl]purin-6-yl}aminoethoxy-methanephosphonic acid
16a	2,4-Diamino-6-[2-(phosphonomethoxy)ethoxy]pyrimidine	PMEO-DAPy
16b	5-Bromo-2,4-diamino-6-[2-(phosphonomethoxy)ethoxy]pyrimidine
17	(R)-{2,4-Diamino-3-hydroxy-6-[2-(phosphonomethoxy)propoxy]}pyrimidine	HPMPO-DAPy
18	Cyclic (R)-{2,4-Diamino-3-hydroxy-6-[2-(phosphonomethoxy)propoxy]}pyrimidine
19a	9-(S)-2-(Phosphonylmethoxypropyl)adenine	(S)-PMPA
19b	9-(S)-(3-Fluoro-2-phosphonylmethoxypropyl)adenine	(S)-FPMPA
20	9-(R)-(3-Fluoro-2-phosphonylmethoxypropyl)adenine	(R)-FPMPA
21a	1-(S)-[3-Hydroxy-2-(phosphonomethoxy)propyl]-5-azacytosine	HPMP-5-azaC
21b	1-{[(5S)-2-Hydroxy-2-oxido-1,4,2-dioxaphosphinan-5-yl]methyl}-5-azacytosine	cHPMP-5-azaC
21c	Hexadecyloxyethyl ester of 1-{[(5S)-2-hydroxy-2-oxido-1,4,2-dioxaphosphinan-5-yl]methyl}-5-azacytosine	HDE-cHPMP-5-azaC
21d	Pivaloyloxymethyl ester of 1-{[(5S)-2-hydroxy-2-oxido-1,4,2-dioxaphosphinan-5-yl]methyl}-5-azacytosine (cyclic POM-HPMP-azaC) and bis(pivaloyloxymethyl) ester of 1-(S)-[3-hydroxy-2-(phosphonomethoxy)-propyl]-5-azacytosine	POM-cHPMP-5-azaC
21e	Octadecyl ester of 1-{[(5S)-2-hydroxy-2-oxido-1,4,2-dioxaphosphinan-5-yl]methyl}-5-azacytosine	Octadecyl-cHPMP- 5-azaC

Antiviral assays.

The antiviral activities of the compounds against each of the strains listed above were evaluated with HEL cells and PHKs. Cells were cultured in 96-well microtiter plates in their respective growth media. Confluent monolayers were infected with each strain at a multiplicity of infection (MOI) of approximately 0.01 for 2 h at 37°C under a 5% CO₂ atmosphere. The medium used for HEL cells to allow CMLV infection and growth was minimal essential medium with Earle's salts containing 2% FCS. For PHKs, the medium used was a mixture of equal parts of keratinocyte-SFM and Dulbecco-F-12 medium. Dulbecco-F-12 medium is a mixture of one part Ham F-12 medium and two parts Dulbecco's modified Eagle's medium, supplemented with 10% FCS, 2 mmol l-glutamine per liter, 10 mmol HEPES per liter, 1 mmol sodium pyruvate per liter, and 5 ml of 100× antibiotic-antimycotic per liter (Gibco, Invitrogen Corporation). At 2 h postinfection, residual virus was removed and immediately replaced with the medium containing serial dilutions of the test compounds (in duplicate). After 6 days, both HEL cells and PHKs were fixed with ethanol and stained with Giemsa solution. Viral cytopathic effect (CPE) was recorded, and the 50% effective concentration (EC₅₀) was defined as the concentration of a compound required to reduce the viral CPE by 50%. The EC₅₀s of the compounds tested against each strain were calculated as the means from three independent experiments.

Cytotoxicity assays.

The toxicities of the compounds for cells were evaluated based on inhibition of cell growth. Cells were seeded into 96-well microtiter plates at 3.5 × 10³ cells in 100 μl of medium per well for HEL cells and at 5 × 10³ cells in 100 μl of medium per well for PHKs by using their respective media, described above. After 1 day (for HEL cells) or 3 days (for PHKs) of incubation, 100 μl of medium containing serial dilutions of the test compounds was added (in duplicate). Following 4 days of incubation in the presence of the compounds, the cells were trypsinized, and the number of cells was determined by automatic counting with a model Z1 Coulter counter, using a lower limit size of 8 μm (Analis, Namur, Belgium). The cytotoxic effects of the compounds were expressed as the 50% cytostatic concentration (CC₅₀), defined as the concentration required to reduce cell growth by 50%. Alternatively, the cytotoxicities of the test compounds were expressed as minimum cytotoxic concentrations (MCC), defined as the concentrations that caused a microscopically detectable alteration of cell morphology.

Virus yield assays.

The effects of several dilutions of compounds selected from the three classes of ANPs (HPMPC, HPMPA, cHPMPC, cHPMPA, HPMPDAP, HPMPO-DAPy, HPMP-5-azaC, cHPMP-5-azaC, and HDE-cHPMP-5-azaC) (Table 1) were evaluated in both HEL cells and PHKs infected with CMLV. Both cell types were grown in six-well microtiter plates, and the confluent monolayers were infected with each virus at an MOI of 0.01 in their respective culture media, as described above. After 2 h of incubation at 37°C under a 5% CO₂ atmosphere, residual virus was removed, each well was washed once with 2 ml of sterile phosphate-buffered saline (PBS) per well, and the different concentrations of the test compounds were added. A virus control (untreated infected cells) and a cell control (uninfected, untreated cells) were included at each time point. The supernatants and the infected cell monolayers were harvested 6 days postinfection and frozen at −20°C. Samples were thawed, and the virus yield was determined by virus titration in 96-well microtiter plates of HEL cells. Titrations were performed with serial 10-fold dilutions of the samples, and the viral titer was expressed in log₁₀ PFU per milliliter. The EC₉₉ (99% effective concentration; the compound concentration required to reduce the viral titer by 99%) was then calculated for each compound.

Organotypic epithelial raft cultures.

For the preparation of epidermal equivalents, a collagen matrix solution was made with collagen mixed on ice with 10-fold-concentrated Ham F-12 medium (Gibco, Invitrogen Corporation), 10-fold-concentrated reconstitution buffer (22% NaHCO₃, 2% NaOH, and 47.6% HEPES), 4 mg/ml collagen type I (Becton Dickinson, Pharmingen, San Diego, CA), and 1.0 × 10⁶ Swiss 3T3 J2 fibroblasts per 0.4 ml. Nine hundred microliters of the collagen matrix solution was poured onto 24-well plates and was solidified at 37°C for 2 h. After equilibration of the gel with 1 ml of Dulbecco-F-12 growth medium overnight at 37°C, 2.0 × 10⁵ PHKs were seeded on top of the gels and were maintained submerged for 24 to 48 h in a 1:1 mixture containing keratinocyte-SFM and raft medium. The raft medium was made from Dulbecco-F-12 medium supplemented with 0.5 μg hydrocortisone per ml, 10 ng epidermal growth factor per ml, 10⁻¹⁰ mol cholera toxin per liter, 5 μg insulin per ml, 5 μg human transferrin per ml, and 15 × 10⁻⁴ mg 3,3′,5-triiodo-l-thyronine per ml. The collagen rafts were raised and placed on stainless steel grids at the interface between the air and the liquid culture medium. The medium was replaced every 2 days. The epithelial cells were then allowed to stratify for 5 days, and at this time, the cultures were infected with 100 μl of CML1 (1,000 PFU/100 μl), which was placed on top of the rafts. To test the effects of the selected molecules (HPMPC, cHPMPC, HPMPDAP, HPMPO-DAPy, HPMP-5-azaC, cHPMP-5-azaC, and HDE-cHPMP5-azaC) (Table 1) on the replication of strain CML1, the medium containing different concentrations of a compound was added at the time of infection. In the same assay, controls were included to verify the normal differentiated epithelium (uninfected rafts) and viral replication (untreated infected rafts). At day 7 postinfection (i.e., at 12 days after lifting), the first set of rafts was fixed in 10% buffered formalin and embedded in paraffin. Sections of about 4 to 6 μm thickness were stained with hematoxylin-eosin for histological examination. The second set of rafts was immerged in 3 ml of sterile PBS. Samples were frozen at −20°C. Rafts were thawed and then centrifuged at 1,800 rpm for 10 min, and the supernatants containing the released virus were collected and used for the quantification of infectious virus. Titration of infectious virus was performed in HEL cells with 10-fold dilutions of the samples, and viral titers were expressed as log₁₀ PFU per raft.

Electron microscopy (EM).

Confluent 75-cm² flasks of HEL cells were infected with CML1 at an MOI of 0.06 and incubated at 37°C. At 30 h postinfection, the cells were first washed with trypsin and then trypsinized. The cells were collected by centrifugation and fixed with 2.5% glutaraldehyde in 0.1 mol/liter PBS at 4°C overnight. After 1 h of postfixation in 1% osmium tetroxide-0.1 mol/liter PBS at 4°C, the sample was dehydrated in a graded series of alcohol and embedded in epoxy resin. Ultrathin sections (thickness, 50 to 60 nm) were cut, stained with uranyl acetate and lead citrate, and examined at 50 kV using a Zeiss EM 900 electron microscope. Images were recorded digitally with a Jenoptik Progress C14 camera system operated using Image-Pro Express software.

RESULTS

EM examination.

CMLV-infected HEL cells were fixed at 30 h postinfection and processed for EM analysis. Many infected and lysed cells were found. As shown in Fig. 1, virus particles were within the cell (surrounding the nucleus), including viral crescents, immature virions, and intracellular mature virus. We could find only a few intracellular enveloped virus particles (19).

FIG. 1.

Visualization of CML1 virus under EM. CMLV-infected HEL cells were fixed at 30 h postinfection and processed for EM analysis showing characteristic accumulation of viral crescents (C), immature virions (IV), and intracellular mature virus (IMV). N, nucleus. Bars, 1.50 μm.

Antiviral and cytotoxicity assays.

The antiviral activities of the three classes of ANPs in HEL cell and PHK monolayers against both strains of CMLV are shown in Table 2. EC₅₀s obtained in HEL cell monolayers were consistently in the same range as those obtained in PHK monolayers. Nevertheless, selectivity indices (SI; calculated as CC₅₀/EC₅₀ ratios) were lower for PHKs than for HEL cells, due to lower CC₅₀s on growing PHK monolayers. In contrast, PHKs growing in rafts demonstrated lower cytotoxicity for a given molecule and consequently higher selectivity. The MCC required to alter both HEL cell and PHK morphology were undetectable on confluent cell monolayers, i.e., 2-D culture, at the highest concentrations tested. SI based on MCC are invariably higher than those calculated with CC₅₀s.

TABLE 2.

Activities and toxicities of compounds tested against CMLV^a

Compound		Activity in HEL cells					Activity in PHK cells
No.	Abbreviation	EC50 (μM)		CC50 (μM)	SI		EC50 (μM)		CC50 (μM)	SI
		CML1	CML14		CML1	CML14	CML1	CML14		CML1	CML14
1a	HPMPC	8.2 ± 3.7	6.6 ± 1.9	192 ± 78	23	29	5.3 ± 2.6	8.8 ± 4.7	7 ± 9	1	1
2a	cHPMPC	6.1 ± 3.1	8.8 ± 0.6	335 ± 222	55	38	7.2 ± 3.6	9.1 ± 5.1	17 ± 19	2	2
1b	HPMPA	0.5 ± 0.4	0.5 ± 0.3	32 ± 14	66	60	1.0 ± 1.2	1.4 ± 1.4	1 ± 1	1	1
2b	cHPMPA	0.2 ± 0.0	0.1 ± 0.1	63 ± 7	393	441	1.6 ± 1.9	3.1 ± 3.8	3 ± 2	2	1
1c	HPMPDAP	0.2 ± 0.2	0.6 ± 0.8	45 ± 8	249	70	0.8 ± 0.9	1.2 ± 1.0	2 ± 1	2	1
1d	3-Deaza-HPMPA	1.5 ± 0.8	1.7 ± 0.7	30 ± 15	20	18	3.1 ± 4.9	2.1 ± 1.9	0.3 ± 0.0	0.1	0.1
3a	PMEA	>183	>183	211 ± 28	<1	<1	162 ± 30	>183	76 ± 83	0.5	<0.4
3b	PMEDAP	>173	>173	100 ± 25	<1	<1	93 ± 48	63.7 ± 9.7	8 ± 6	0.1	0.1
3c		6.9 ± 3.3	9.3 ± 5.5	8 ± 6	1	1	7.9 ± 0.0	15.9 ± 12.9	1 ± 1	0.2	0.1
3d		3.3 ± 3.2	5.5 ± 4.7	1.0 ± 0.8	0.3	0.2	2.9 ± 1.6	4.4 ± 3.3	0.3 ± 0.1	0.1	0.1
3e		29.4 ± 19.3	24.6 ± 15.4	22 ± 17	1	1	8.5 ± 4.8	10.9 ± 2.9	1 ± 1	0.1	0.1
3f		>174	>174	>174	1	1	>174	>174	>70	0.4	0.4
4		>139	>139	294 ± 14	<2	<2	>139	>139	258 ± 30	<2	<2
5		>134	>134	282 ± 47	<2	<2	>134	>134	>134	1	1
6		>124	>124	>124	1	1	>124	>124	>124	1	1
7		54.5 ± 23.6	55.7 ± 26.6	3 ± 0.1	0.1	0.1	81.8 ± 9.4	93.4 ± 39.9	2 ± 0	0.03	0.03
8		48.3 ± 25.9	46.0 ± 20.6	20 ± 21	0.4	0.4	>124	>124	16 ± 16	<0.1	<0.1
9		>115	>115	101 ± 54	<1	<1	>115	>115	27 ± 2	<0.2	<0.2
10		53.0 ± 5.7	44.3 ± 9.3	12 ± 10	0.2	0.3	77.8 ± 24.6	87.4 ± 37.3	8 ± 7	0.1	0.1
11a		44.1 ± 32.0	29.0 ± 1.3	9 ± 9	0.2	0.3	47.7 ± 6.1	71.5 ± 19.8	2 ± 1	0.04	0.03
11b		63.0 ± 20.4	71.6 ± 8.5	56 ± 51	1	1	>128	>128	>16	<0.1	<0.1
11c		112 ± 27	112 ± 27	39 ± 31	0.3	0.3	>128	>128	19 ± 21	<0.2	<0.2
12		1.9 ± 0.8	1.9 ± 1.4	2 ± 1	1	1	3.7 ± 3.0	4.9 ± 3.3	0.1 ± 0.0	0.03	0.03
13		>126	>126	258 ± 39	<2	<2	>126	>126	102 ± 12	<1	<1
14		>127	>127	43 ± 85	<0.3	<0.3	>127	>127	59 ± 66	<0.5	<0.5
15		>118	>118	ND	ND	ND	>118	>118	ND	ND	ND
16a	PMEO-DAPy	>189	>189	66 ± 50	<0.3	<0.3	49.2 ± 34.1	53.7 ± 38.8	4 ± 2	0.1	0.1
16b		>156	>156	52 ± 52	<0.3	<0.3	>156	>156	ND	ND	ND
17	HPMPO-DAPy	1.6 ± 0.5	2.0 ± 0.6	162 ± 127	99	82	3.9 ± 1.6	4.6 ± 2.6	5 ± 3	1	1
18		5.2 ± 1.5	7.6 ± 3.4	147 ± 32	28	19	16.1 ± 9.4	18.4 ± 13.4	43 ± 31	3	2
19a	(S)-PMPA	>176	>176	>176	1	1	>176	>176	>176	1	1
19b	(S)-FPMPA	>164	>164	539 ± 84	<3	<3	>164	>164	151 ± 71	<1	<1
20	(R)-FPMPA	>164	>164	>164	1	1	>164	>164	>164	1	1
21a	HPMP-5-azaC	12.3 ± 3.2	12.0 ± 1.9	550 ± 74	45	46	14.8 ± 3.3	9.7 ± 3.1	71 ± 45	5	7
21b	cHPMP-5-azaC	4.3 ± 0.7	3.0 ± 0.4	143 ± 13	33	48	18.0 ± 1.8	14.8 ± 4.4	34 ± 3	2	2
21c	HDE-cHPMP-5-azaC	0.02 ± 0.00	0.02 ± 0.00	5 ± 5	250	250	0.6 ± 0.2	0.3 ± 0.1	1 ± 1	2	4
21d	POM-cHPMP-5-azaC	8.0 ± 0.6	6.5 ± 1.4	79 ± 0	10	12	12 ± 0.0	10.9 ± 1.4	54 ± 38	5	5
21e	Octadecyl-cHPMP-5-azaC	8.9 ± 1.1	5.9 ± 0.3	>97	11	17	10 ± 0.0	8.5 ± 1.8	6 ± 1	1	1

^aThe EC₅₀ of each compound represents the mean ± standard deviation of the EC₅₀s from at least three independent experiments, while the CC₅₀ of each compound represents the mean ± standard deviation of the CC₅₀s from at least two independent experiments. ND, not determined.

Among the first class of ANPs, HPMPA, cHPMPA, HPMPDAP, and 3-deaza-HPMPA exhibited mean EC₅₀s in HEL fibroblasts ranging from 0.1 to 1.7 μM, 5- to 80-fold lower than those obtained with HPMPC (mean EC₅₀s, 8.2 and 6.6 μM against CML1 and CML14, respectively, in HEL cells). Similar EC₅₀s were obtained for PHKs (Table 2). In HEL cells, cHPMPA and HPMPDAP proved to be highly selective, with SI of >100. The cyclic counterpart of HPMPC, cHPMPC, showed antiviral activity comparable to that of the parent compound both in HEL cells and in PHKs (mean EC₅₀s ranged from 6.1 to 9.1 μM).

As expected, none of the phosphonomethoxyethyl (PME) derivatives (compounds 3a to 3f) revealed any anti-CMLV selectivity in HEL cells (SI, <1). Most of the ANP derivatives (Table 2) numbered as compounds 4 to 15 displayed very poor antiviral activities in both cell lines, except for compound 12 (mean EC₅₀s ranged from 1.9 to 4.9 μM in HEL cells and PHKs). In addition, compounds 4 to 15 had SI of <2.

Among compounds belonging to the second class of ANPs (compounds 16a, 17, and 18), HPMPO-DAPy (compound 17) was able to inhibit CMLV replication in both cell lines, with mean EC₅₀s ranging from 1.6 to 4.6 μM. These results are comparable to those found with this compound's alkyl purine counterpart, HPMPDAP. HPMPO-DAPy proved to be highly selective (SI, 99 and 82 in HEL cells against CML1 and CML14, respectively). The cyclic form of HPMPO-DAPy (compound 18) also showed antiviral efficacy similar to that of HPMPC. Compound 16a (PMEO-DAPy), like PMEA, showed no antiviral activity.

The new HPMP derivatives containing a 5-azacytosine moiety, HPMP-5-azaC and cHPMP-5-azaC, displayed CMLV inhibitory potencies similar to those of the parent compounds, HPMPC and cHPMPC, respectively. HPMP-5-azaC showed EC₅₀s of 12.3 ± 3.2 μM and 14.8 ± 3.3 μM for CML1 in HEL fibroblasts and PHKs, respectively, and the SI of HPMP-5-azaC were similar to those of the parent compound in HEL cell and PHK monolayers (Table 2). The EC₅₀s of cHPMP-5-azaC against CML1 were 4.3 ± 0.7 μM and 18.0 ± 1.8 μM in HEL cell and PHK monolayers, respectively, and its selectivity was comparable to that of cHPMPC in both HEL fibroblasts and PHKs. Among all HPMP derivatives tested containing a 5-azacytosine moiety, HDE-cHPMP-5-azaC, the alkoxylalkyl ester prodrug of cHPMP-5-azaC, clearly showed the highest antiviral potency and selectivity against both CMLV strains. The EC₅₀s obtained for HDE-cHPMP-5-azaC against strains CML1 and CML14 were, respectively, 0.02 μM in HEL cells and 0.6 ± 0.2 μM and 0.3 ± 0.1 μM in PHKs. The SI obtained in HEL cells were five- to eightfold higher than those seen with the parent molecule, cHPMPC. Two other prodrugs of cHPMP-5-azaC (POM-cHPMP-5-azaC and octadecyl-cHPMP-5-azaC) demonstrated antiviral efficacy, with average EC₅₀s of 6.5 μM and 10.9 μM against CML14 in HEL fibroblasts and PHKs, respectively. In HEL cells, the SI of these prodrugs were of the same order of magnitude as that for their derivative, cHPMP-5-azaC.

Virus yield assays.

The antiviral efficacies of 10 compounds belonging to the three classes of ANPs were then investigated against both strains of CMLV in a virus yield reduction assay performed in HEL cell and PHK monolayers. Virus yields were determined at 6 days postinfection. The compounds that showed the highest potency and selectivity in the CPE reduction assay—HPMPC, cHPMPC, HPMPA, cHPMPA, HPMPDAP, 3-deaza-HPMPA, HPMPO-DAPy, HPMP-5-azaC, cHPMP-5-azaC, and HDE-cHPMP-5-azaC—were selected for further evaluation in the virus yield assays. A concentration-dependent inhibitory effect for each compound on CMLV yield was observed when different concentrations of each drug were added (Fig. 2).

FIG. 2.

Virus yield reduction assays in HEL fibroblasts and PHKs. The antiviral activities of several concentrations of the selected compounds against CMLV replication were determined. Virus yield, evaluated at 6 days postinfection, was measured by virus titration, and viral titers are expressed in log₁₀ PFU per milliliter. EC₉₉s, expressed in micromolar concentrations, are given in parentheses.

HPMPC and its cyclic counterpart, cHPMPC, exhibited similar antiviral efficacies against CMLV yield (EC₉₉s, 58 μM and 16 μM for HPMPC and 64 μM and 59 μM for cHPMPC against CML1 in HEL fibroblasts and PHKs, respectively). The higher potencies of HPMPA, cHPMPA, and HPMPDAP compared to that of HPMPC were also confirmed by the virus yield assay. The EC₉₉s against CML14 were 1 μM and 2 μM for HPMPA, 2 μM and 6 μM for cHPMPA, and 1 and 2 μM for HPMPDAP in HEL cells and PHKs, respectively. 3-Deaza-HPMPA also inhibited CML1 and CML14 replication by 99% with values in the range from 5 to 13 μM in both cell lines.

Among the DAPy derivatives, HPMPO-DAPy exhibited similar antiviral activities against both strains in both cell types, with an average EC₉₉ of 9.5 μM.

The prodrug HDE-cHPMP-5-azaC emerged as the most potent compound against both viral strains in HEL cells, affording complete inhibition of viral replication at 0.2 μg/ml (0.4 μM) (EC₉₉s, 0.1 μM) (Fig. 2). In PHKs, this compound abolished viral replication at 2 μg/ml (4 μM), and the resulting EC₉₉s were 1 and 3 μM against CML1 and CML14, respectively (Fig. 2). HPMP-5-azaC and cHPMP-5-azaC displayed, in both cell types, antiviral efficacies against CML1 and CML14 similar to those of their parent compounds, HPMPC and cHPMPC, and entirely suppressed viral replication at a concentration of 20 μg/ml (73 μM).

Organotypic epithelial raft cultures.

We evaluated the inhibitory effects of HPMPC, cHPMPC, HPMPDAP, and HPMPO-DAPy and of the three newly synthesized HPMP derivatives with a 5-azaC moiety (HPMP-5-azaC, cHPMP-5-azaC, and HDE-cHPMP-5-azaC) against CML1 replication in 3-D epithelial raft cultures.

As shown in Fig. 3, histological images obtained from uninfected, untreated rafts revealed a fully differentiated epithelium containing all the epidermal layers. CML1 proved to be able to replicate efficiently in this ex vivo model, which can be considered equivalent to human skin (Fig. 3). Histopathological images showed cytoplasmic swelling and ballooning of the keratinocytes comparable to that described for skin biopsy specimens from animals with camelpox disease (43). The differentiated epithelium was completely damaged at 7 days postinfection. CPEs in the 3-D culture were correlated with virus yield, confirming the replication of CML1 in this culture system (Fig. 3 and 4).

FIG. 3.

Histological images from rafts. Shown are epithelial raft cultures infected with strain CML1 (1,000 PFU/raft) after 7 days of differentiation and treated with various concentrations (0.5, 2, and 5 μg/ml) of either HPMP-5-azaC, cHPMP-5-azaC, or HDE-cHPMP-5-azaC at the time of infection. Magnification for all panels, ×40.

FIG. 4.

Activities of HPMPO-DAPy, HPMP-5-azaC, cHPMP-5-azaC, and HDE-cHPMP-5-azaC as well as their parent compounds against CML1 replication in organotypic epithelial raft cultures of PHKs. Epithelial raft cultures were infected with strain CML1 (1,000 PFU/raft) after 7 days of differentiation and were treated with various concentrations (0.2, 0.5, 2, 5, and 20 μg/ml) of the ANPs at the time of infection. Raft cultures were collected and then frozen, and viral titers were determined in HEL cells and expressed as the number of PFU per raft on a logarithmic scale.

HPMPC and cHPMPC treatment of the infected rafts resulted in full protection of the epithelium at 20 μg/ml (65 μM) and 50 μg/ml (164 μM), respectively (data not shown). This was confirmed by the virus yield assay performed on the rafts, giving EC₉₉s of 6 and 38 μM against CML1 for HPMPC and cHPMPC, respectively (Fig. 4). HPMPDAP completely inhibited viral replication in the rafts at a concentration of ≥2 μg/ml (6 μM) (Fig. 4). Its diaminopyrimidine counterpart, HPMPO-DAPy, showed a similar inhibitory profile and exhibited an EC₉₉ of 5 μM (Fig. 4). In the third class of ANPs, HPMP-5-azaC exhibited an EC₉₉ values of 16 μM compared to 6 μM for its parent compound, HPMPC (Fig. 4). Histopathological images showed the presence of CPEs at HPMP-5-azaC concentrations of 0.5 μg/ml (1.8 μM), 2 μg/ml (7 μM), and 5 μg/ml (18 μM) (Fig. 3). However, we observed cytotoxic effects on the epithelium with increasing HPMP-5-azaC concentrations starting at 5 μg/ml (Fig. 3). The cHPMP-5-azaC molecule did not show any cytotoxicity on the epithelium at the highest concentration tested (20 μg/ml, or 76 μM) and was able to completely protect the rafts against CML1 replication at 20 μg/ml, like its parent compound, cHPMPC (EC₉₉s, 67 μM for cHPMP-5-azaC versus 38 μM for cHPMPC) (Fig. 3 and 4). The prodrug HDE-cHPMP-5-azaC, which was one of the most promising drugs in monolayers, was able to completely suppress viral replication at the highest concentration assessed, 20 μg/ml (38 μM) (Fig. 4), while also showing signs of cytotoxic effects (Fig. 3). CPEs were still visually observed at a concentration of 5 μg/ml (9 μM) of HDE-cHPMP-5-azaC (Fig. 3). Interestingly, HDE-cHPMP-5-azaC behaved similarly in this 3-D model to its parent compound, cHPMPC (EC₉₉, 14 μM versus 38 μM for cHPMPC).

DISCUSSION

The threat posed by the potential use of VARV, the etiological agent of smallpox, as a biological weapon in warfare or bioterrorism has stimulated the search for antiviral agents. In particular, the development of new antiviral drugs that are orally available and exhibit high selectivity in their antipoxvirus activities is highly warranted. Antipoxvirus compounds that target cellular enzymes (IMP dehydrogenase inhibitors, such as ribavarin, as well as the tyrosine kinase inhibitor STI-571, also called imatinib mesylate, or Gleevec) or viral enzymes (including inhibitors of viral morphogenesis [TTP-6171], viral release [ST-246], and viral DNA synthesis [e.g., ANPs analogues including HPMPC]) have been described (39). The antiviral efficacies of those compounds have been investigated in different poxvirus models available as surrogate models for the study of VARV, including vaccinia, cowpox, monkeypox, and ectromelia viruses (39).

The ANPs have acquired a prominent therapeutic position in the treatment of papillomavirus, herpesvirus, adenovirus, and poxvirus infections (cidofovir; HPMPC), as well as in the treatment of chronic hepatitis B virus infections {adefovir; 9-[2-(phosphonomethoxy)ethyl]adenine (PMEA)} and of human immunodeficiency virus infections [tenofovir; 9-(R)-(2-phosphonomethoxypropyl)adenine (PMPA)] (15). Cidofovir has been found effective in the treatment of human papillomavirus-associated lesions in nonimmunocompromised individuals as well as in organ transplant recipients and AIDS patients (7). Recently, new classes of ANPs (PMPO-DAPy, PMEO-DAPy, HPMPO-DAPy) have been accredited with antiviral potencies and selectivities similar to those of cidofovir, adefovir, and/or tenofovir. Nevertheless, cidofovir shows poor oral bioavailability, and this problem could be circumvented by the synthesis of its prodrug forms, as has already been demonstrated for two other ANPs, PMEA and PMPA (11, 34, 35, 41). A series of alkoxyalkyl cidofovir esters, which are orally bioavailable, has been synthesized by esterification with long-chain alkoxyalkanols (hexadecyloxypropyl-cidofovir [HDP-CDV] and octadecyloxyethyl-cidofovir [ODE-CDV]) (10).

The antiviral potency of the cidofovir derivative HPMP-5-azaC has been demonstrated against several DNA viruses, including herpes simplex virus types 1 and 2, varicella-zoster virus, human cytomegalovirus, adenovirus, and poxvirus (28). The HPMP derivatives bearing the 5-azacytosine moiety also showed high activities and selectivities against murine polyomavirus and the polyomavirus simian virus 40 (30).

In this study we evaluated the antiviral efficacies of several ANPs against the replication of CMLV, which has been shown to be the closest known orthopoxvirus to VARV. The current knowledge of the replication cycle of CMLV is very limited. Upon EM analysis performed with HEL cells at 30 h post-CML1 infection, a majority of intracellular mature virus forms were seen, as well as many lysed cells, but only rare intracellular enveloped virus particles could be observed, a finding different from that of other related orthopoxviruses (19). These observations with HEL cells may suggest that CMLV behaves differently from vaccinia virus and cowpox virus in terms of kinetics and the amount of viral forms produced, and they highlight the usefulness of a CMLV model.

We determined the susceptibilities of two different CMLV clinical samples isolated from two independent outbreaks, which were responsible for generalized diseases of camels (32, 33). Of all the compounds tested, cHPMPA, HPMPDAP, HPMPO-DAPy, and HDE-HPMP-5-azaC emerged as the most selective molecules against both CMLV strains (CML1 and CML14) in HEL monolayers, with SI ranging from 70 to 441. Due to the high sensitivity of growing PHK monolayers to the compounds, no selectivity could be seen. These observations could potentially translate into toxicity in vivo. However, 3-D cultures of human epithelial cells can be considered more predictive of the in vivo situation, and in this model, there were no cytotoxic alterations of the differentiated PHKs. The EC₅₀s of compounds in the first class of ANPs against CMLV were similar to those observed in the HEL and PHK cell lines against two other, related orthopoxviruses, vaccinia virus (strains Lister, Lederle, and Copenhagen) and cowpox virus (strain Brighton) (29, 40). Similarly, the selective and potent activity of HPMPO-DAPy against CMLV is consistent with previous reports on vaccinia virus, cowpox virus, and orf virus in both human and ovine cell monolayers (12, 16). The antiviral efficacies of the compounds belonging to the first and second classes of ANPs against CMLV replication were confirmed by virus yield reduction assays, with EC₉₉s ranging from 3 μM for cHPMPA to 64 μM for cHPMPC against CML1 in HEL cells. We observed a similar inhibitory effect against CML14 in both cell lines.

Among the newly synthesized HPMP derivatives, HPMP-5-azaC, cHPMP-5-azaC, POM-cHPMP-5-azaC, and octadecyl-cHPMP-5-azaC showed antiviral activities against CMLV similar to those previously described against vaccinia and cowpox viruses in HEL cell monolayers (28). HDE-cHPMP-5-azaC exhibited the highest potency against CMLV replication in HEL cells and PHKs. Similar results were observed against several strains of vaccinia virus and against cowpox virus strain Brighton, with EC₅₀s in HEL cells ranging from 0.03 ± 0.02 μM to 0.2 ± 0.1 μM (M. Krečmerová et al., submitted). However, in the case of the parapoxvirus orf, HDE-cHPMP-5-azaC has been shown to inhibit viral replication with EC₅₀s 100-fold lower than those for vaccinia virus (EC₅₀ against orf virus, 0.002 ± 0.0007 μg/ml [our unpublished data]). The lack of some genes in orf virus that are highly conserved in other chordopoxviruses and are likely involved in nucleotide metabolism may explain its unusual sensitivity to the polymerase inhibitor (17). The inhibitory activity of HDE-cHPMP-5-azaC against CMLV replication was confirmed by a virus yield assay performed on fibroblast and keratinocyte monolayers, with EC₉₉s 130- to 640-fold lower than those of the parent compound, cHPMPC, in HEL cells.

Here we also report on the ability of CMLV to replicate in a 3-D epithelial model. Organotypic epithelial raft cultures are tissue culture systems that allow full differentiation of keratinocyte monolayers via culturing of the cells on collagen gels at the air-liquid interface (1). Poxviruses are epitheliotropic, and both vaccinia and cowpox viruses have already been shown to be able to infect raft cultures of human keratinocytes (29). Surprisingly, we were able to easily infect human keratinocytes with CMLV, even though this virus is responsible for a disease restricted to camels. It has already been shown that CMLV is able to infect HEL fibroblasts (5). In CMLV-infected rafts, we observed cytopathic changes identical to those described for epithelium isolated from infected camels (27, 43). The addition of different concentrations of the selected ANPs to the medium of infected rafts led to inhibition of viral growth. HPMPO-DAPy and its counterpart HPMPDAP were found to be the most potent molecules in this system. HDE-cHPMP-5-azaC appeared to be as active as its parent compound, cHPMPC, in the 3-D model, a result different from those obtained in monolayers. One of the hypotheses to explain the decrease in HDE-cHPMP-5-azaC activity could be the long hydrophobic “tail” [O(CH₂)₂-O-(CH₂)₁₅CH₃] of the molecule linked to the phosphonate group. Indeed, this hydrophobic tail might not facilitate the transfer of the molecule through the collagen bed and thus to the epithelial layer. Preliminary experiments performed on mice infected intranasally with vaccinia virus and treated orally with HDE-cHPMP-5-azaC have pointed to difficulties in achieving adequate concentrations of the compound (or its active metabolite) in the lungs (2). Taken together, our results support the usefulness of the 3-D model, compared to the monolayers, in predicting the in vivo efficacies of new antipoxvirus drugs.

Camelpox disease, well known in the Middle East for its economic consequences, is characterized by a high morbidity and a relatively high mortality rate in young camels and pregnant females (43). Currently, two vaccines against camelpox infections are commercially available. Wernery and Zachariah have shown that the live attenuated vaccine Ducapox gave protection to 1-year-old camels for at least 6 years (44). However, Pfeffer et al. have reported outbreaks in Dubai, where seven camels with clinical signs of CMLV infection were identified out of 2,000 camels vaccinated with the modified live CMLV vaccine CaPV-298-2 (32). It could not be determined whether those camels represented true vaccination failures. Postexposure treatment of camelpox infections has not yet been described. Treatment of severe cases consists of minimizing secondary infections by local application or parenteral administration of broad-spectrum antibiotics and vitamins (43). Our study suggested that treatment with antiviral agents could be another therapeutic approach to managing camelpox infections, particularly for young camels. We have confirmed the antiviral potency of cidofovir against CMLV, already described by Smee et al., and demonstrated the anti-CMLV potencies of several other ANPs among the different classes of ANPs (37).

Finally, the probability that CMLV infections occur in humans is higher for immunocompromised individuals, since this population has already been shown to be more susceptible to poxvirus infections than immunocompetent persons (13). The treatment of poxvirus infections with cidofovir or other ANPs should therefore be considered. Indeed, cidofovir (administered either topically or intravenously) has been found to be effective in the treatment of molluscum contagiosum in AIDS patients (13). Similarly, a 39-year-old renal transplant patient, under immunosuppression, was completely cured of a giant orf (ecthyma contagiosum) lesion after topical treatment with cidofovir (21). Herdsmen may represent another group of persons more frequently exposed to camelpox disease. Thus, as proposed by Azwai et al., immunological surveys for specific anti-CMLV antibodies could be helpful in determining a possible transmission of CMLV to humans (3). A study of herdsmen, using three methods including EM to identify the presence of CMLV, was conducted in the early 1980s, but none were found positive for CMLV (26).

In conclusion, our results demonstrate that, in addition to cidofovir, other ANPs, particularly HPMP-5-azaC, cHPMP-5-azaC, and HDE-cHPMP-5-azaC, represent promising candidates for treating poxvirus infections.