Comparative Activities of Lipid Esters of Cidofovir and Cyclic Cidofovir against Replication of Herpesviruses In Vitro

Abstract

Cidofovir (CDV) is an effective therapy for certain human cytomegalovirus (HCMV) infections in immunocompromised patients that are resistant to other antiviral drugs, but the compound is not active orally. To improve oral bioavailability, a series of lipid analogs of CDV and cyclic CDV (cCDV), including hexadecyloxypropyl-CDV and -cCDV and octadecyloxyethyl-CDV and -cCDV, were synthesized and found to have multiple-log-unit enhanced activity against HCMV in vitro. On the basis of the activity observed with these analogs, additional lipid esters were synthesized and evaluated for their activity against herpes simplex virus (HSV) types 1 and 2, human cytomegalovirus, murine cytomegalovirus, varicella-zoster virus (VZV), Epstein-Barr virus (EBV), human herpesvirus 6 (HHV-6), and HHV-8. Using several different in vitro assays, concentrations of drug as low as 0.001 μM reduced herpesvirus replication by 50% (EC₅₀) with the CDV analogs, whereas the cCDV compounds were generally less active. In most of the assays performed, the EC₅₀ values of the lipid esters were at least 100-fold lower than the EC₅₀ values for unmodified CDV or cCDV. The lipid analogs were also active against isolates that were resistant to CDV, ganciclovir, or foscarnet. These results indicate that the lipid ester analogs are considerably more active than CDV itself against HSV, VZV, CMV, EBV, HHV-6, and HHV-8 in vitro, suggesting that they may have potential for the treatment of infections caused by a variety of herpesviruses.

Herpesviruses cause a number of clinical manifestations in humans, including genital herpes, herpes zoster, neonatal herpes, cytomegalovirus (CMV) retinitis, infectious mononucleosis, exanthema subitum, and Kaposi's sarcoma, and those most affected are immunocompromised patients (6, 12). CMV disease is a particularly important problem in patients with organ transplants and represents a major cause of morbidity and mortality (14, 33, 36).

Currently, antiviral agents licensed for use in treatment of infections caused by herpesviruses include acyclovir (ACV), cidofovir (CDV), and ganciclovir (GCV) (18). More recent compounds that have been approved for therapy include penciclovir (PCV), valacyclovir (prodrug for ACV), famciclovir (prodrug for PCV) (2), formivirsen (11), and foscarnet (PFA) (10). Treatment with nucleoside analogs against herpes simplex virus (HSV), varicella-zoster virus (VZV), and CMV infections has been established for many years; however, problems of toxicity, emergence of drug resistance, and lack of oral bioavailability underscore the need to develop new and improved antiviral agents for treatment of these and other herpesvirus infections, such as Epstein-Barr virus (EBV), human herpesvirus 6 (HHV-6), and HHV-8.

To obtain drug activity when the drug was given orally, lipid conjugates for ACV, GCV, and PCV were prepared; these conjugates had been reported to be effective against drug-resistant herpes simplex viruses and experimental infections caused by human CMV (HCMV) (15, 16, 17, 24, 25). In addition, studies with hepatitis B virus (HBV) suggest that an alkoxyalkyl-phosphate-ACV analog had significant activity in vitro and in vivo against woodchuck hepatitis virus (15, 16). The hexadecyloxypropyl-phospho-ACV (HDP-P-ACV) compound also exhibited little or no toxicity (17). Similarly, conjugates of GCV and PCV (HDP-P-GCV and HDP-P-PCV) have been synthesized and have been found to have good activity against HCMV, HSV type 1 (HSV-1), and HSV-2 in vitro and in vivo (17).

CDV is an acyclic nucleoside phosphonate analog that has demonstrated good activity in vitro and in vivo against a number of infections caused by herpesviruses (8, 9) and is licensed for treatment of CMV retinitis in patients with AIDS. It has prolonged activity in cells due to its long intracellular half-life. It is slowly and poorly absorbed into cells and is not bioavailable when administered orally. In addition, continuous use of this compound causes nephrotoxicity in patients treated for CMV infections (18). To circumvent these problems, several ether lipid ester conjugates of CDV have been synthesized that have enhanced activity against HSV, CMV, and orthopoxviruses in vitro and are active when given orally in vivo (3, 4, 19, 22, 31). These conjugates were designed to resemble natural lipids that are readily absorbed intact from the small intestine and distributed to tissues via plasma or lymph (29). The conjugates appear to facilitate the uptake of CDV into cells where it is likely cleaved and then phosphorylated to the active CDV-diphosphate (1).

We have reported previously that hexadecyloxypropyl cidofovir (HDP-CDV), octadecyloxypropyl cidofovir (ODE-CDV), oleyloxyethyl cidofovir (OLE-CDV), and oleyloxypropyl cidofovir (OLP-CDV) may be good candidates for new antiviral agents (3, 20, 21, 22, 23, 31). On the basis of the activities of these compounds, additional analogs of CDV and cyclic CDV (cCDV) were synthesized. These analogs of CDV include the following: alkyl esters, such as octyl (O)-, dodecyl (DD)-, tetracosyl (TC)-, hexadecyl (HD)-, eicosyl (EC)-, and docosyl (DC)-CDV; additional alkoxyalkyl esters, such as dodecyloxypropyl- (DDP)-, tetradecyloxypropyl- (TDP)-, octadecyloxypropyl (ODP)-, and eicosyloxypropyl- (ECP)-CDV; and the glycerol ether lipid 1-octadecyl-2-benzylglycero (ODBG)-CDV. The purpose of the present study was to evaluate the activities of these additional compounds against replication of HSV-1, HSV-2, VZV, CMV, EBV, HHV-6A, HHV-6B, and HHV-8.

MATERIALS AND METHODS

Antiviral compounds.

CDV was kindly provided by Gilead, Inc., Foster City, CA; ACV and GCV were purchased from the University of Alabama Hospital Pharmacy, and PFA was purchased from Sigma (St. Louis, MO). The CDV and cCDV analogs were provided by Karl Hostetler, University of California at San Diego, La Jolla, CA, through the Antiviral Substance Program, National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH), Bethesda, MD. Their structure, synthesis, purification, and characterization have been published previously (3, 21, 23, 38).

The names of the CDV and cCDV analog compounds, their abbreviations, the linker used, and the number of carbon atoms in the alkyl chain are summarized in Table 1.

TABLE 1.

CDV and cCDV analogs used in this study

Compound	Abbreviation	No. of atoms in ester chain	No. of double bonds
Cidofovir	CDV
Cidofovir analogs
Propanediol linkers
Dodecyloxypropyl	DDP-CDV	16	0
Tetradecyloxypropyl	TDP-CDV	18	0
Hexadecyloxypropyl	HDP-CDV	20	0
Octadecyloxypropyl	ODP-CDV	22	0
Oleyloxypropyl	OLP-CDV	22	1
Eicosyloxypropyl	ECP-CDV	24	0
Ethanediol linkers
Octadecyloxyethyl	ODE-CDV	21	0
Oleyloxyethyl	OLE-CDV	21	1
Glycerol linker
1-O-Octadecyl-2-O-benzyl-glyceryl	ODBG-CDV	NAa	NA
No linker
Octyl	O-CDV	8	0
Dodecyl	DD-CDV	12	0
Hexadecyl	HD-CDV	16	0
Eicosyl	EC-CDV	20	0
Docosyl	DC-CDV	22	0
Tetracosyl	TC-CDV	24	0
Cyclic cidofovir	cCDV
Cyclic cidofovir analogs
Propanediol linkers
Dodecyloxypropyl	DDP-cCDV	16	0
Hexadecyloxypropyl	HDP-cCDV	20	0
Octadecyloxypropyl	ODP-cCDV	22	0
Oleyloxypropyl	OLE-cCDV	22	1
Ethanediol linkers
Octadecyloxyethyl	ODE-cCDV	21	0
Oleyloxyethyl	OLE-cCDV	21	1
Glycerol linker
1-O-Octadecyl-2-O-benzyl-glyceryl	ODBG-cCDV	NA	NA
No linker
Hexadecyl	HD-cCDV	16	0
Eicosyl	EC-cCDV	20	0
Docosyl	DC-cCDV	22	0
Tetracosyl	TC-cCDV	24	0

^aNA, not applicable.

Cell culture and virus strains.

Human foreskin fibroblast (HFF) cells were prepared as primary cultures and used in assays for HSV, CMV, and VZV. The cells were propagated in minimal essential medium (MEM) with 10% fetal bovine serum (FBS), l-glutamine, penicillin, and gentamicin. Strains of HSV-1, HSV-2, HCMV, and VZV were obtained from sources reported previously (3, 34) and propagated in HFF cells using standard virological methods.

All lymphocyte cell lines were grown in RPMI 1640 with 10% FBS, l-glutamine, and antibiotics and split 1:5 twice a week as described previously (26, 39). Daudi and P3HR-1 cells latently infected with EBV were obtained from the American Type Culture Collection (Manassas, VA). EBV stocks were prepared from P3HR-1 cells as described previously (39). HHV-6A strain GS and HSB-2 cells were obtained through the NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH. HHV-6B strain Z29 and Molt-3 cells were obtained from Scott Schmid at the Centers for Disease Control and Prevention, Atlanta, GA. HHV-6A and HHV-6B were propagated by cocultivating infected cells with uninfected HSB-2 and Molt-3 cells, respectively. Infection was monitored by an immunofluorescence assay (IFA) using a monoclonal antibody to p41/38 (ABI, Columbia, MD), and infected cells were frozen in liquid nitrogen when 50% of the cells were infected. HHV-8-infected BCBL-1 cells were obtained from the NIH AIDS Research and Reference Reagent Program. Lytic virus replication was induced with phorbol 12-myristate 13-acetate (PMA) at a concentration of 100 ng/ml.

Antiviral assays. (i) Plaque reduction assay in HFF cells.

HFF cells were plated in six-well plates and incubated at 37°C. When the cell layer reached confluence, the medium was aspirated from the wells, 0.2 ml of virus was added to each well (three wells) to yield 20 to 30 plaques per well, and 0.2 ml of medium was added to each well to test for drug toxicity. The plates were then incubated for 1 h with shaking every 15 minutes after which drug concentrations ranging from 100 μg/ml to 0.03 μg/ml were added to duplicate wells. After the appropriate incubation times, cell monolayers were stained with 1% crystal violet in 20% methanol for HSV-1 and -2 or with a 1% neutral red solution for HCMV and VZV. The stain was aspirated, the wells were washed with phosphate-buffered saline (PBS), and plaques were counted using a stereomicroscope. The concentration of drug that reduced plaque formation by 50% (EC₅₀) was determined by comparing drug-treated cultures with untreated cultures and calculated using MacSynergy II software (26, 30, 39).

(ii) IFA.

Daudi cells were fixed on slides with acetone at room temperature, rinsed with PBS, incubated with the monoclonal antibody to EBV and with viral capsid antigen (VCA) (Chemicon, Temecula, CA) and then with fluorescein isothiocyanate (FITC)-labeled immunoglobulin G (IgG) and IgM (Jackson ImmunoResearch, West Grove, PA) at 37°C for 1-h intervals. Cells were counterstained with 0.1% Evans blue dye (Fisher, Fair Lawn, NJ) in PBS and mounted using 50% glycerol in PBS. Stained cells were enumerated with a Nikon fluorescence microscope (Nikon, Melville, NY). Five fields were scanned, 100 cells per field were counted, and the percentage of positive cells was calculated at each drug concentration. Drug efficacy was determined by plotting drug concentration versus the percentage of positive cells and then interpolating the concentration of drug required to inhibit 50% of virus replication using MacSynergy II software (26, 30, 39).

(iii) ELISA.

An enzyme-linked immunosorbent assay (ELISA) was performed on cells fixed with 5% acetic acid in ethanol, rinsed with PBS, and incubated with a monoclonal antibody to EBV VCA (Chemicon) and then with horseradish peroxidase-labeled goat anti-mouse IgG1 (Southern Biotechnology, Birmingham, AL). The antibodies were diluted with 1% BSA in PBS. The colorimetric reaction was initiated by addition of O-phenylenediamine dihydrochloride in citrate buffer (pH 5.0) and hydrogen peroxide, and the reaction was stopped by the addition of 3 N sulfuric acid. The EC₅₀ value for each drug was interpolated from the plot of drug concentration versus the average optical density at 492 nm as described previously (26, 39).

(iv) In situ DNA hybridization assay.

The Simply Sensitive Horseradish Peroxidase-AEC In Situ detection system (Enzo Diagnostics, Farmingdale, NY) was used to monitor EBV DNA synthesis in the presence of antiviral compounds and was performed according to the manufacturer's instructions. The EC₅₀ value for each drug was calculated by previously described methods (26, 39).

(v) Flow cytometric assays.

Flow cytometry was used to evaluate the activity of the antiviral compounds against HHV-8 by methods described previously (26, 39). Samples were fixed in 2% paraformaldehyde in PBS and analyzed using a Becton-Dickinson FacsCalibur instrument (Becton-Dickinson, Franklin Lakes, N.J.). The WinMDI 2.7 program was used to analyze the data.

(vi) HHV-6 (6A or 6B) DNA hybridization assay.

Uninfected HSB-2 or Molt-3 cells were added to 96-well plates at a concentration of 1 × 10⁴ cells/well. Infection was initiated by adding HHV-6A-infected HSB-2 cells or HHV-6B-infected Molt-3 cells at a ratio of approximately 1 infected cell for every 10 uninfected HSB-2 cells or Molt-3 cells, respectively. Drugs were serially diluted and added to triplicate wells. Drug-free medium was added to the wells that contained the cell and virus controls. The plates were incubated for 7 days at 37°C. After incubation, 100 μl of 3× denaturation buffer (4.5 M NaCl, 1.2 M NaOH) was added to each well and 50 μl was added to the wells of a Biodot apparatus (Bio-Rad, Hercules, CA) containing an Immobilon nylon membrane (Millipore, Bedford, MA). A 50-μl aliquot of denatured DNA from the infected cells was aspirated though the membrane, and then 50 μl of denaturation buffer and 50 μl of PBS were added. The membrane was allowed to air dry, equilibrated in 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-0.1% sodium dodecyl sulfate (SDS), and prehybridized in DIG Easy Hyb (Roche Applied Science, Indianapolis, IN) for 30 minutes at 37°C. A specific HHV-6 digoxigenin (DIG)-labeled probe was prepared by PCR (Roche Applied Science) using primers 5′-CCT TGA TCA TTC GAC CGT TT-3′ and 5′-TGG GAT TGG GAT TAG AGC TG-3′ to amplify a segment of ORF2 (coordinates 37820 to 38418 in X83413). The denatured probe was allowed to hybridize to the viral DNA on the filter overnight at 37°C. The membrane was washed once at 37°C with 2× SSC-0.1% SDS, three times with 0.2× SSC-0.1% SDS, and once with 0.1× SSC-0.1% SDS. Detection of specifically bound DIG probe was performed with anti-DIG antibody using the manufacturer's protocol (Roche). An image of the film was captured digitally and quantified with Quantity One software (Bio-Rad), and EC₅₀ values were calculated as previously described (30).

Cytotoxicity assays with an adherent cell line (HFF). (i) Neutral red uptake cytotoxicity assay.

Cells were seeded into 96-well tissue culture plates at 2.5 × 10⁴ cells/well. After 24 h of incubation, the medium was replaced with MEM containing 2% FBS, and drug was added to the first row and then diluted serially fivefold from 100 to 0.03 μg/ml. The plates were incubated for 7 days, and the cells were stained with neutral red and incubated for 1 h. The plates were washed and shaken for 15 min, and the optical density was read at 540 nm (23). The concentration of drug that reduced cell viability by 50% (CC₅₀) was calculated as previously described (30).

(ii) Cell proliferation assay.

Cells were seeded into six-well plates at a concentration of 2.5 × 10⁴ cells/well. After 24 h, the medium was aspirated, and drug that had been serially diluted 1:5 was added. The plates were incubated for 72 h at 37°C, the cells were trypsinized and counted using a Beckman Coulter Counter, and 50% inhibition of cell proliferation (IC₅₀) values were calculated as previously described (30).

Cytotoxicity assays with a nonadherent cell line (Daudi). (i) MTS tetrazolium cytotoxicity assay.

Serial fivefold dilutions of drug starting at 50 μg/ml were prepared in medium and added to 10⁶ cells. Controls were prepared by incubating 10⁶ cells in drug-free medium. After a 3-day incubation, 200 μl of a cell suspension was transferred to a 96-well plate in duplicate. Twenty microliters of 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt (MTS) was added, and the plate was wrapped in foil and incubated at 37°C for 4 h. The quantity of formazan product was measured at 490 nm in a microplate reader and was directly proportional to the number of living cells in culture. The drug concentration was plotted against the optical density of each sample, and CC₅₀ values were calculated as described above.

(ii) Cell proliferation assay.

Serial fivefold dilutions of drug starting at 50 μg/ml were prepared in medium and added to 10⁶ cells. Controls were prepared by incubating 10⁶ cells in drug-free medium. After an incubation period of 3 to 4 days, the number of cells for each sample was determined by using a Coulter Counter. The drug concentration was plotted against the total concentration of cells for each sample, and IC₅₀ values were calculated as described above.

RESULTS

Activity against HSV-1, HSV-2, and VZV.

The activities of the compounds were evaluated against HSV-1, HSV-2, and VZV in plaque reduction assays, and their activities are summarized in Table 2. All of the isolates tested were significantly more susceptible to most of the analogs compared to the parent compound (CDV or cCDV), except for the shorter alkyl chains, O-, DD-, and TC-CDV. There were few differences between HSV-1 and HSV-2 strains. Many of the CDV analogs were 100- to 1,000-fold more active than the parent CDV was, and the HDP-, ODP-, OLE-, and OLP-CDV derivatives were the most active for all HSV strains tested. For the cCDV analogs, similar patterns of activity were obtained; however, the effective levels were generally less than those seen for CDV analogs. With VZV, EC₅₀ values of 0.0001 to 0.08 μM were seen with the various CDV analogs, which were about 10- to 100-fold more active than CDV was. The greatest increases in activity were observed for HDP-, ODP-, ODE-, and ODBG-CDV analogs. The cCDV analogs were not tested for VZV activity. In general, analogs with greatly improved activity against strains of HSV also had the most improved activity against VZV, particularly ODE-CDV.

TABLE 2.

Activities of ether lipid ester analogs against HSV-1, HSV-2, and VZV replication in HFF cells^a

Compound	HSV-1 EC50 (μM) for strain:			HSV-2 EC50 (μM) for strain:			VZV EC50 (μM) for strain Ellen
	E-377	F	HL-3	MS	G	SR
ACV	2.5 ± 1.3	2.8 ± 1.2	3.8 ± 0.4	4.4 ± 2.8	4.4 ± 0.4	4.4 ± 0.6	3.6 ± 2
CDV	5.5 ± 6.0	5.7 ± 7.4	3.0 ± 1.5	5.1 ± 5.4	5.3 ± 3.0	6.5 ± 1.9	0.5 ± 0.7
DDP-CDV	0.5 ± 0.6 (11)	0.5 ± 0.2 (11)	0.4 ± 0.4 (8)	0.3 ± 0.3 (17)	0.9 ± 0.007 (6)	0.7 ± 0.3 (9)	0.05 ± 0.03 (10)
TDP-CDV	0.08 ± 0.05 (69)	NT	NT	0.08 ± 0.07 (64)	NT	NT	0.01 ± 0.007 (50)
HDP-CDV	0.06 ± 0.03 (91)	0.009 ± 0.005 (633)	0.01 ± 0.01 (230)	0.08 ± 0.07 (64)	0.01 ± 0.01 (530)	0.02 ± 0.02 (325)	0.0004 ± 0.0003 (12,500)
ODP-CDV	0.05 ± 0.05 (110)	0.2 ± 0.2 (29)	0.05 ± 0.05 (60)	0.02 ± 0.03 (255)	0.1 ± 0 (53)	0.1 ± 0.09 (65)	0.005 ± 0 (100)
OLP-CDV	0.02 ± 0.01 (183)	0.08 ± 0.07 (71)	0.04 ± 0.01 (75)	0.007 ± 0.001 (729)	0.09 ± 0.02 (59)	0.1 ± 0.1 (65)	0.02 ± 0.01 (25)
ECP-CDV	1.2 ± 1.4 (5)	NT	NT	0.53 ± 0.22 (10)	NT	NT	0.07 ± 0.04 (7)
ODE-CDV	0.02 ± 0.01 (275)	0.002 ± 0 (2,850)	0.008 ± 0.01 (395)	0.03 ± 0 (170)	0.008 ± 0.001 (663)	0.007 ± 0.01 (929)	0.0001 ± 0.0001 (5,000)
OLE-CDV	0.03 ± 0 (275)	0.03 ± 0.03 (190)	0.05 ± 0.01 (60)	0.005 ± 0.005 (1,020)	0.06 ± 0.05 (88)	0.07 ± 0.004 (93)	0.02 ± 0.01 (25)
ODBG-CDV	0.04 ± 0.03 (138)	NT	NT	0.01 ± 0 (510)	NT	NT	0.005 ± 0.001 (100)
O-CDV	>10 ± 0 (<1)	NT	NT	>10 ± 0 (<1)	NT	NT	>1 ± 0 (<5)
DD-CDV	>10 ± 0 (<1)	NT	NT	>10 ± 0 (<1)	NT	NT	0.9 ± 0.01 (<1)
HD-CDV	0.3 ± 0.4 (18)	0.3 ± 0.2 (19)	0.3 ± 0.3 (10)	0.2 ± 0.2 (26)	0.6 ± 0 (9)	0.5 ± 0.25 (13)	0.03 ± 0.01 (17)
EC-CDV	0.1 ± 0.1 (39)	NT	NT	0.3 ± 0.03 (17)	NT	NT	0.01 ± 0.01 (38)
DC-CDV	2.8 ± 1.5 (2)	NT	NT	4.7 ± 5.2 (1)	NT	NT	0.03 ± 0.02 (17)
TC-CDV	>7.1 ± 4.1 (<1)	NT	NT	>6.3 ± 5.2 (<1)	NT	NT	0.08 ± 0.07 (6)
cCDV	11.9 ± 4.6	13.6 ± 8.6	9.9 ± 5.1	17.4 ± 0.2	14 ± 5.1	18.3 ± 12.3	3.6
DDP-cCDV	0.5 ± 0.4 (24)	0.8 ± 0.2 (17)	0.6 ± 0.03 (17)	0.6 ± 0.3 (29)	1.4 ± 0.8 (10)	1.1 ± 0.6 (16)	0.006 (600)
HDP-cCDV	0.6 ± 0.5 (20)	0.1 ± 0.1 (136)	0.2 ± 0.03 (50)	0.2 ± 0.2 (85)	0.1 ± 0.04 (140)	0.5 ± 0.5 (37)	NT
ODP-cCDV	0.3 ± 0.02 (40)	0.2 ± 0.04 (68)	0.2 ± 0.1 (49)	0.2 ± 0.1 (87)	0.1 ± 0.1 (140)	0.3 ± 0.1 (61)	NT
OLP-cCDV	0.07 ± 0.02 (170)	0.2 ± 0.1 (68)	0.2 ± 0.1 (50)	0.08 ± 0.01 (218)	0.5 ± 0.4 (28)	0.2 ± 0.02 (92)	NT
ODE-cCDV	0.4 ± 0.2 (30)	0.09 ± 0.07 (151)	0.2 ± 0.03 (50)	0.2 ± 0.1 (87)	0.5 ± 0 (28)	0.2 ± 0.2 (92)	NT
OLE-cCDV	0.03 ± 0.01 (397)	0.04 ± 0.01 (348)	0.06 ± 0.01 (165)	0.02 ± 0.01 (870)	0.05 ± 0.01 (280)	0.06 ± 0.02 (305)	NT
ODBG-cCDV	0.4 ± 0 (30)	0.5 ± 0.4 (27)	0.2 ± 0 (50)	0.3 ± 0.1 (58)	0.4 ± 0.3 (35)	0.3 ± 0 (61)	NT
HD-cCDV	2.9 ± 1.3 (4)	3.3 ± 0.3 (4)	3.0 ± 1.1 (3)	4.5 ± 5 (4)	5.3 ± 2.7 (3)	6.3 ± 4.9 (3)	0.1 (36)
EC-cCDV	>6.7 ± 4.6 (2)	NT	NT	6.1 ± 5.5 (29)	NT	NT	NT
DC-cCDV	>10 ± 0 (<1)	NT	NT	>10 ± 0 (<1)	NT	NT	NT
TC-cCDV	>10 ± 0 (<1)	NT	NT	>10 ± 0 (<1)	NT	NT	NT

^aValues represent the means±standard deviations of two assays. The increase in activity(fold increase) over that of the CDV or cCDV control is shown in parentheses. NT, not tested.

Activity against CMV.

The activities of three laboratory isolates (AD169, Towne, and Davis) and four clinical isolates (Toledo, Coffman, C8805/37-1-1, and CR9209/1-4-4) of HCMV were also evaluated by plaque reduction assays, and the results are summarized in Table 3. All of the compounds, except the short-chain alkyl esters, O-CDV and OD-CDV, demonstrated significantly greater activity against CMV than GCV did. Most of the analogs tested were 100- to 2,000-fold more active than the parent, CDV, against laboratory and clinical isolates with EC₅₀ values ranging from 0.001 to 0.3 μM. As seen with the alphaherpesviruses, TDP-, ODP-, HDP-, ODE-, OLE-, OLP-, and ODBG-CDV were among the compounds with the best activity. Similar patterns of activity were observed with cCDV analogs. A panel of four GCV-resistant (759^r D100, GDG^r P53, which also has resistance to CDV, 1117^r, C8914-6) and two PFA-resistant (VR4760^r, VR4955^r) HCMV isolates were also tested, and as expected, all the mutants were susceptible to CDV and cCDV. All of the analogs had enhanced susceptibility with EC₅₀ values from 0.001 to 0.1 μM in comparison to EC₅₀ values of 0.3 to 16 μM for CDV and cCDV (Table 4). The magnitude of the enhanced activity seen with both sets of analogs against GCV- and PFA-resistant isolates was similar to that observed against wild-type strains.

TABLE 3.

Activities of ether lipid ester analogs against clinical and laboratory isolates of HCMV in HFF cells

Compound	EC50 (μM)a
	AD169b	Davis	Towne	Toledo	Coffman	C8805/37-1-1	C9209/1-4-4
GCV	3.8 ± 0.1	3.4 ± 0.59	3.2 ± 2.6	37.2 ± 51.3	5.5 ± 1.2	47.4 ± 1.1	59.3 ± 0.2
CDV	0.4 ± 0.1	0.5 ± 0.3	0.5 ± 0.4	3.8 ± 2.8	1.9 ± 1.3	0.8 ± 0.2	1.4 ± 0.06
DDP-CDV	0.01 ± 0.01 (40)	NT	0.02 (27)	0.02 ± 0.02 (190)	0.3 ± 0.3 (6)	0.03 (24)	NT
TDP-CDV	0.002 ± 0.002 (200)	0.001 ± 0 (510)	NT	0.003 (1,267)	0.008 ± 0.002 (238)	0.003 (267)	0.007 ± 0.007 (200)
HDP-CDV	0.001 ± 0.001 (400)	0.001 ± 0 (510)	0.001 ± 0 (540)	0.03 ± 0.01 (127)	0.001 ± 0 (1,900)	0.001 ± 0.001 (800)	0.004 ± 0.001 (350)
ODP-CDV	0.002 ± 0.001 (200)	0.001 ± 0 (510)	0.001 (540)	0.01 ± 0.01 (380)	0.01 ± 0.01 (158)	0.02 ± 0.02 (47)	0.004 ± 0.003 (350)
OLP-CDV	0.002 ± 0 (200)	0.002 ± 0.001 (255)	0.002 (270)	0.005 ± 0.005 (760)	0.02 ± 0.02 (95)	0.008 ± 0.002 (100)	0.01 ± 0.005 (140)
ECP-CDV	0.03 ± 0.03 (13)	0.02 ± 0.003 (26)	NT	0.05 (76)	0.09 ± 0 (21)	0.02 (33)	0.01 ± 0.007 (100)
ODE-CDV	0.001 ± 0 (400)	0.001 ± 0.0001 (510)	0.001 ± 0 (540)	0.03 ± 0.01 (127)	0.001 ± 0 (1,900)	0.001 ± 0.001 (800)	0.004 ± 0.007 (350)
OLE-CDV	0.001 ± 0.001 (400)	0.001 ± 0 (510)	0.002 (270)	0.003 ± 0.001 (1,267)	0.02 ± 0.002 (95)	0.02 ± 0.01 (53)	0.005 ± 0.003 (280)
ODBG-CDV	0.003 ± 0.001 (133)	0.003 ± 0.003 (170)	0.003 (180)	0.01 ± 0.01 (380)	0.05 ± 0.06 (38)	0.009 ± 0.01 (89)	0.003 ± 0.001 (467)
O-CDV	5.3 ± 6 (<1)	NT	NT	NT	NT	NT	NT
DD-CDV	0.3 ± 0.2 (<1)	NT	NT	NT	NT	NT	NT
HD-CDV	0.005 ± 0 (80)	0.005 ± 0.001 (102)	0.006 (90)	0.02 ± 0.01 (190)	0.02 ± 0 (95)	0.03 (30)	NT
EC-CDV	0.001 (400)	0.001 (510)	NT	0.005 (760)	0.03 (63)	0.03 ± 0.03 (24)	0.01 ± 0 (140)
DC-CDV	0.004 (100)	0.01 (51)	NT	0.007 (543)	0.2 (9)	0.06 ± 0.05 (16)	0.02 ± 0.01 (70)
TC-CDV	0.06 ± 0.03 (7)	0.2 (2)	NT	0.04 (95)	>1 (<1)	0.5 ± 0.03 (1.6)	0.08 ± 0.08 (18)
cCDV	0.9 ± 0.5	0.6 ± 0.2	0.5 ± 0.02	11.8 ± 11.8	1.9 ± 0.26	1.0 ± 0.2	2.0 ± 0.9
DDP-cCDV	0.007 ± 0.002 (129)	0.02 ± 0.001 (30)	NT	0.03 (393)	0.05 ± 0.01 (38)	NT	NT
HDP-cCDV	0.001 ± 0 (900)	0.001 ± 0 (600)	0.001 ± 0 (500)	0.06 ± 0.03 (197)	0.002 ± 0.001 (950)	0.001 ± 0 (1,000)	0.005 ± 0 (400)
ODP-cCDV	0.006 ± 0.001 (150)	0.003 ± 0.002 (200)	NT	0.007 (1,686)	0.009 ± 0.001 (211)	NT	NT
OLP-cCDV	0.003 ± 0.001 (300)	0.006 ± 0.0001 (100)	NT	0.001 (11,800)	0.003 ± 0.0002 (633)	NT	NT
ODE-cCDV	0.002 ± 0.001 (450)	0.003 ± 0.001 (300)	0.001 ± 0 (500)	0.006 ± 0.03 (1,967)	0.002 ± 0.001 (950)	0.001 ± 0 (1,000)	0.005 ± 0.001 (400)
OLE-cCDV	0.002 ± 0.001 (450)	0.006 ± 0.002 (100)	NT	0.001 (11,800)	0.001 ± 0 (1,900)	NT	NT
ODBG-cCDV	0.01 ± 0.004 (90)	0.01 ± 0.001 (60)	NT	0.006 (1,967)	0.02 ± 0.003 (95)	NT	NT
HD-cCDV	0.04 ± 0.02 (23)	0.03 ± 0.01 (20)	NT	0.1 (118)	0.2 ± 0.006 (10)	NT	NT
EC-cCDV	0.01 ± 0.005 (90)	NT	NT	NT	NT	NT	NT
DC-cCDV	0.04 ± 0.01 (22)	NT	NT	NT	NT	NT	NT
TC-cCDV	0.29 ± 0.04 (3)	NT	NT	NT	NT	NT	NT

^aValues represent the means±standard deviations of two assays. The increase in activity(fold increase) over that of the CDV or cCDV control is shown in parentheses. NT, not tested.

^bResults are essentially the same as those in reference 38.

TABLE 4.

Activities of ether lipid ester analogs against drug-resistant HCMV isolates in HFF cells

Compound	EC50 (μM)a
	759r D100	GDGr P53	1117r	C8914-6r	VR4760r	VR4955r
GCV	113 ± 70	42.7 ± 39	40.3 ± 5.1	134 ± 81
PFA					373 ± 70	613 ± 200
CDV	2.0 ± 0.56b	15.7 ± 14.1b	0.4 ± 0.07	3.2 ± 3.5	0.4 ± 0.3	0.3 ± 0.05
DDP-CDV	0.2 ± 0.2 (10)	0.2 ± 0.1 (80)	0.06 ± 0 (7)	0.2 ± 0.04 (21)	0.01 ± 0.007 (35)	0.009 ± 0.005 (37)
TDP-CDV	0.009 ± 0.01 (222)	0.004 ± 0.003 (4,000)	0.01 ± 0.01 (43)	0.01 ± 0.004 (320)	0.004 ± 0.001 (88)	0.005 ± 0.003 (66)
HDP-CDV	0.007 ± 0.001 (285)b	0.02 ± 0.001 (800)b	0.003 ± 0.001 (143)	0.003 ± 0.001 (1,067)b	0.0008 ± 0.00007 (438)	0.0009 ± 0 (367)
ODP-CDV	0.02 ± 0.004 (100)	0.01 ± 0.001 (1,600)	0.005 ± 0 (86)	0.01 ± 0.008 (320)	0.002 ± 0.001 (175)	0.002 ± 0.001 (165)
OLP-CDV	0.03 ± 0.01 (67)	0.1 ± 0.07 (160)	0.005 ± 0 (86)	0.06 ± 0.008 (53)	0.002 ± 0.0006 (175)	0.004 ± 0.003 (83)
ECP-CDV	0.2 ± 0.2 (10)	0.08 ± 0.07 (200)	0.07 ± 0.02 (6)	0.1 ± 0.04 (32)	0.02 ± 0.01 (18)	0.02 ± 0.003 (17)
ODE-CDV	0.006 ± 0.001 (333)b	0.02 ± 0.001 (800)b	0.005 ± 0.001 (86)	0.001 ± 0 (3,200)b	0.0009 ± 0.0002 (389)	0.0009 ± 0.00007 (367)
OLE-CDV	0.1 ± 0.07 (20)	0.07 ± 0.02 (224)	0.03 ± 0.03 (14)	0.008 ± 0.05 (400)	0.007 ± 0.007 (50)	0.01 ± 0.008 (33)
ODBG-CDV	0.02 ± 0.006 (100)	0.02 ± 0 (800)	0.006 ± 0.002 (72)	0.02 ± 0 (160)	0.003 ± 0.002 (117)	0.006 ± 0.002 (55)
HD-CDV	0.08 ± 0.01 (25)	0.08 ± 0.03 (200)	0.03 ± 0.0007 (14)	0.09 ± 0.06 (35)	0.007 ± 0.004 (50)	0.007 ± 0.0007 (47)
cCDV	2.5 ± 0.1b	15.9 ± 12.2b	1.1 ± 0.8	2.7 ± 2.0	0.3 ± 0.1	0.4 ± 0.03
DDP-cCDV	0.1 ± 0.03 (25)	0.06 ± 0.05 (265)	0.2 ± 0.1 (5.5)	0.08 ± 0.004 (34)	0.03 ± 0.002 (11)	0.01 ± 0.01 (40)
HDP-cCDV	0.02 ± 0.01 (125)b	0.06 ± 0.05 (265)b	NT	0.005 ± 0.001 (540)b	NT	NT
ODP-cCDV	0.01 ± 0.01 (250)	0.01 ± 0.001 (1,590)	0.02 ± 0.02 (55)	0.02 ± 0.01 (135)	0.005 ± 0 (64)	0.003 ± 0.002 (133)
OLP-cCDV	0.002 ± 0.001 (1,250)	0.002 ± 0.001 (7,950)	0.003 ± 0.001 (367)	0.004 ± 0.002 (675)	0.001 ± 0.0001 (320)	0.001 ± 0 (400)
ODE-cCDV	0.02 ± 0.004 (125)b	0.05 ± 0.03 (318)b	NT	0.005 ± 0.001 (540)b	NT	NT
ODBG-cCDV	0.02 ± 0.01 (125)	0.01 ± 0 (1,590)	0.01 ± 0.01 (110)	0.03 ± 0 (90)	0.003 ± 0.002 (107)	0.005 ± 0.001 (80)
HD-cCDV	0.1 ± 0.05 (13)	0.07 ± 0.02 (227)	0.2 ± 0.03 (5.5)	0.4 ± 0.3 (7)	0.04 ± 0.02 (8)	0.03 ± 0.004 (13)

^aValues represent the means±standard deviations of two assays. The increase in activity(fold increase) over that of the CDV or cCDV control is shown in parentheses. NT, not tested.

^bData from reference 3.

Activity against HHV-6A and HHV-6B.

The activities of the analogs against HHV-6A and HHV-6B were determined using a DNA hybridization assay (Table 5). Both viruses were susceptible to CDV and cCDV, and many of the analogs had enhanced activities of 50- to 1,000-fold for these viruses. The magnitude of the enhancement over the activity of the parent compound, however, was generally less than that seen with some of the other herpesviruses, particularly CMV. There was almost complete agreement in the susceptibilities of the two viruses to the analogs, although the relative difference between the two was variable.

TABLE 5.

Activities of ether lipid ester analogs against HHV-6 replication in HSB-2 or Molt-3 cells

Compound	EC50 (μM)a
	HHV-6A	HHV-6B
CDV	2.7 ± 2.3	5.4 ± 4.6
DDP-CDV	0.6 ± 0.08 (5)	7.3 ± 3.9 (<1)
TDP-CDV	>25 ± 0 (<1)	>25 (<1)
HDP-CDV	0.003 ± 0 (900)	0.007 ± 0.005 (771)
ODP-CDV	0.07 ± 0.04 (39)	0.2 ± 0.3 (27)
OLP-CDV	0.2 ± 0.1 (14)	0.05 ± 0.05 (108)
ECP-CDV	2.9 ± 1.4 (<1)	14.9 ± 2.3 (<1)
ODE-CDV	0.003 ± 0 (900)	<0.003 ± 0 (>1,800)
OLE-CDV	0.6 ± 0.6 (5)	0.3 ± 0.4 (18)
ODBG-CDV	<0.008 ± 0 (338)	0.009 ± 0.001 (600)
O-CDV	NT	NT
DD-CDV	NT	NT
HD-CDV	0.2 ± 0.07 (14)	0.6 ± 0.9 (9)
EC-CDV	0.3 ± 0.06 (9)	0.6 ± 0.2 (9)
DC-CDV	1.2 ± 1.4 (2)	>27.1 ± 3 (<1)
TC-CDV	NT	NT
cCDV	1.6 ± 2.2	16 ± 1.3
DDP-cCDV	0.3 ± 0.1 (5)	2.3 ± 0.8 (7)
HDP-cCDV	0.1 ± 0.1 (16)	0.4 ± 0.4 (40)
ODP-cCDV	0.09 ± 0.01 (18)	0.2 ± 0.004 (80)
OLP-cCDV	<0.009 ± 0.001 (178)	<0.008 ± 0 (2,000)
ODE-cCDV	0.006 ± 0.003 (267)	0.1 ± 0.1 (160)
OLE-cCDV	<0.008 ± 0 (>200)	<0.008 ± 0 (2,000)
ODBG-cCDV	<0.008 ± 0 (>200)	0.06 ± 0.03 (267)
HD-cCDV	NT	NT
EC-cCDV	3.9 ± 1.2 (<1)	>24.4 ± 0.8 (<1)
DC-cCDV	20.2 ± 0.9 (<1)	13.4 ± 0.3 (1.2)
TC-cCDV	NT	NT

^aValues represent means±standard deviations of two assays. The fold increase in activity(fold increase) over that of the CDV or cCDV control is shown in parentheses. NT, not tested.

Activity against EBV and HHV-8.

In three different assay systems for EBV, CDV and cCDV had only minimal activity, so ACV was used as an additional positive control for these experiments (Table 6). In contrast, using IFA, ELISA, and in situ DNA hybridization assays, many of the CDV analogs were very active with EC₅₀ values ranging from <0.002 to 1.6 μM. The antiviral activities of the cCDV analogs were higher than those of the parent compound but less than those of the CDV analogs. Although there was variability in the results for the three assays, which measure different aspects of viral replication, the correlation among them was quite good. The same analogs that were the most active against the other viruses were also the most active against EBV. Although CDV and cCDV were not active against EBV in our assays, both were active against HHV-8. The lipid-containing analogs that were active against EBV also had increased activity to HHV-8. Additionally, with one exception, O-CDV, all the analogs tested that were not active in the ELISA against EBV were very active against HHV-8.

TABLE 6.

Activities of ether lipid ester analogs against the replication of EBV and HHV-8

Compound	EC50 (μM)a EBV by:			EC50 (μM)a against HHV-8 by FACSc
	IFA	ELISA	DNA assayb
ACV	6.2 ± 1.3	8.5 ± 1.9	6.2 ± 1.3	>100 ± 0
CDV	75.4 ± 39.3	>108 ± 70	65.6 ± 59.9	2.6 ± 3.8
DDP-CDV	0.9 ± 0.1 (84)	0.7 ± 0.2 (154)	0.4 ± 0.02 (164)	0.04 ± 0.002 (65)
TDP-CDV	NT	>50 (<2)	NT	<0.02 (>130)
HDP-CDV	0.02 ± 0.02 (3,770)	0.04 ± 0.01 (2,700)	0.03 ± 0.01 (2,186)	0.02 ± 0.002 (130)
ODP-CDV	1.6 ± 1.1 (47)	0.6 ± 0.4 (180)	>0.6 ± 0.6 (<109)	0.007 ± 0 (370)
OLP-CDV	0.9 ± 0.4 (84)	0.04 ± 0.01 (2,700)	0.4 ± 0.2 (164)	0.02 ± 0.01 (130)
ECP-CDV	NT	>50 (<2)	NT	0.1 ± 0.08 (26)
ODE-CDV	0.2 ± 0.13 (377)	0.02 ± 0.01 (5,400)	0.1 ± 0.08 (656)	0.03 ± 0.01 (87)
OLE-CDV	0.04 ± 0.01 (1,885)	<0.002 ± 0.001 (>54,000)	0.03 ± 0.03 (2,187)	0.02 ± 0.01 (130)
ODBG-CDV	0.8 ± 0.3 (94)	<0.002 ± 0.001 (>54,000)	0.5 ± 0.6 (131)	0.02 ± 0.003 (130)
O-CDV	NT	>50 (<2)	NT	>50 (<1)
DD-CDV	NT	>50 ± 0 (<2)	NT	2.7 ± 1.6 (<1)
HD-CDV	6.1 ± 8.3 (12)	>0.6 ± 0.5 (<180)	0.6 ± 0.3 (109)	0.03 ± 0.006 (87)
EC-CDV	NT	>10 (<11)	NT	0.05 (52)
DC-CDV	NT	4.1 ± 3.4 (26)	NT	0.3 (9)
TC-CDV	NT	>50 (<2)	NT	0.2 ± 0.01 (13)
cCDV	>133 ± 47	>125 ± 58	>166 ± 0	27 ± 0.3
DDP-cCDV	>1 ± 0	>1 ± 0	>0.8 ± 0.2 (<207)	0.04 ± 0.01 (675)
HDP-cCDV	3.8 ± 3.1 (>35)	>9.2 ± 1.1 (>14)	4.3 ± 2.4 (>39)	0.08 ± 0.03 (335)
ODP-cCDV	>1 ± 0	>0.7 ± 0.5 (<178)	>1 ± 0	0.03 ± 0.003 (900)
OLP-cCDV	>1 ± 0	0.3 ± 0.2 (>417)	0.4 ± 0.03 (>415)	0.006 ± 0 (4,500)
ODE-cCDV	6.9 ± 0.14 (>20)	8.8 ± 1.7 (>14)	5.7 ± 2.3 (>29)	0.09 ± 0.08 (300)
OLE-cCDV	>1 ± 0	0.2 ± 0.3 (>625)	0.06 ± 0.03 (>2,767)	0.007 ± 0 (3,857)
ODBG-cCDV	>1 ± 0	>1 ± 0	>1 ± 0	0.05 ± 0.01 (540)
HD-cCDV	>1 ± 0	>1 ± 0	>1 ± 0	0.2 ± 0.09 (135)
EC-cCDV	NT	1.9	NT	NT
DC-cCDV	NT	>50 ± 0	NT	NT
TC-cCDV	NT	>50 ± 0	NT	NT

^aValues represent means±standard deviations of two experiments. The increase in activity(fold increase) over that of the CDV or cCDV control is shown in parentheses. NT, not tested.

^bIn situ DNA hybridization assay.

^cFACS, fluorescence-activated cell sorting.

Toxicity of ether lipid ester analogs.

Toxicity was determined by assays that measured both cell killing (CC₅₀) and cell proliferation (IC₅₀). Although there was more toxicity seen with the analogs in both assays in comparison to the parent compounds (Table 7), their selectivity is substantially higher because the analogs were effective at lower concentrations. This increase in selectivity has been documented previously for CMV, adenovirus, and orthopoxviruses. Due to the large number of viruses used in this study, we have not presented selective index values for each virus, as they are similar to those published previously for other viruses (3, 13, 20, 21).

TABLE 7.

Toxicity in adherent and nonadherent cells^a

Compound	CC50 (μM)b		IC50 (μM)c
	Daudi	HFF	Daudi	HFF
ACV	>444	>444	>444	>444
CDV	>318	>100	234	183 ± 78
DDP-CDV	18.6	>100	0.4	>20
TDP-CDV	>20	>100 ± 0	>20	100
HDP-CDV	>34.3	34 ± 14	>34.3	3.0 ± 3.0
ODP-CDV	17.8	22 ± 11	16	3.0 ± 3.0
OLP-CDV	>32.8	114	>32.8	>20
ECP-CDV	>20	77 ± 12	8.1	80
ODE-CDV	>33.5	18 ± 5	>33.5	0.8 ± 0.3
OLE-CDV	31.1	33 ± 0	>33.6	11.4 ± 0.3
ODBG-CDV	16	>100	4.8	3.3 ± 0.9
O-CDV	>20	>100 ± 0	>20	>100
DD-CDV	>20	>100 ± 0	>20	>100
HD-CDV	12.2	90	18.4	7.6 ± 5.4
EC-CDV	>20	52 ± 19	0.53	53
DC-CDV	>20	>100	>20	>100
TC-CDV	>20	>100	2.4	>100
cCDV	>331	>331 ± 0	>329	113 ± 88
DDP-cCDV	>1 ± 0	61	>20	>20 ± 0
HDP-cCDV	>45	>73	>20	32.4 ± 19.8
ODP-cCDV	>1 ± 0	75	>20	22 ± 1.6
OLP-cCDV	>1 ± 0	46	19.7	14.7 ± 4.4
ODE-cCDV	>38	62 ± 36	>20	3.9 ± 0.7
OLE-cCDV	>1 ± 0	26	14.0	9.1 ± 5.4
ODBG-cCDV	>1 ± 0	>100	>20	>20
HD-cCDV	>1 ± 0	69	>20	>32 ± 9
EC-cCDV	>20	>100 ± 0	1.25	>100
DC-cCDV	>20	>100 ± 0	>20	>100
TC-cCDV	>20	>100 ± 0	>20	>100

^aSome values represent means ± standard deviations of two assays. The other values are from single determinations.

^bCC₅₀, concentration of drug that kills 50% of cells in stationary culture(HFF) or in suspension (Daudi).

^cIC₅₀, concentration of drug that inhibits cell growth by 50%.

DISCUSSION

The significant side effects resulting from long-term treatment of herpesvirus infections in the immunocompromised host with GCV, PFA, or CDV along with the emergence of drug resistance has demonstrated the need for new and improved drug therapy. Although CDV is highly active against most drug-resistant mutants, its lack of activity when given orally and the potential for producing nephrotoxicity have limited its usefulness. Hostetler and colleagues (16, 17) reported that ether lipid ester analogs of PCV, GCV, and PFA had significantly increased activity against HSV and CMV as well as HBV and woodchuck hepatitis virus (15, 16). These findings were subsequently extended to the synthesis of lipid conjugates for other antivirals, including CDV. It has been reported previously that the ether lipid esters of CDV, HDP-CDV, and ODE-CDV are highly active in tissue culture against herpesviruses, adenoviruses, and orthopoxviruses (1, 3, 13, 19, 21). They have enhanced activity against laboratory and clinical HCMV isolates in vitro as well as GCV- and PFA-resistant isolates (3).

It has also been reported that the oral bioavailability of CDV in mice was significantly higher for HDP-CDV and ODE-CDV than for CDV (7). When the two ether lipid ester analogs were evaluated in animal models for efficacy against murine CMV, CMV, and poxvirus infections, they were as active when given orally as the parent, CDV, given parenterally (4, 5, 23, 31). In addition to its lack of oral bioavailability, another limitation to the use of CDV is its nephrotoxicity (35). It has been reported that oral administration of HDP-CDV or ODE-CDV to mice results in high concentrations of CDV in the liver and spleen but in considerably lower levels in the kidneys, which suggests that oral administration of the lipid compounds should result in significantly less nephrotoxicity (29).

In the present study, we evaluated a large number of the lipid ester analogs having a variable alkyl chain length that had a propanediol, ethanediol, or glycerol linker or no linker for their activity against HSV, CMV, VZV, EBV, HHV-6, and HHV-8. In previous studies carried out with CMV, adenoviruses, or orthopoxviruses, a structure-activity analysis indicated that the best activity was observed when the alkyoxyalkyl chain length was 18 to 22 atoms and included either an ethanediol or a propanediol linker (13, 20, 38). Straight alkyl chains were also optimally active if the chain length was 18 to 22 atoms (38). The structure-activity patterns were similar for all the herpesviruses used in the present study, as was reported previously for CMV, orthopoxviruses, and adenoviruses.

In cells infected with either HSV-1, HSV-2, or VZV, the most active analogs were HDP-, ODP-, ODE-, OLP-, and OLE-CDV with EC₅₀ values that were 10- to 1,000-fold greater than the value for CDV. For the cCDV analogs, a similar pattern was observed, but the increased activity was generally less than that seen with the CDV analogs. It has been reported previously that the AD169 strain of HCMV is also inhibited by concentrations of the analogs that were 1 to 3 log₁₀ units lower than the CDV concentration (38). In the present study, we tested two additional laboratory isolates of HCMV (Davis and Towne), four clinical isolates of HCMV (Toledo, Coffman, C8805/37-1-1, and C9209/1-4-4), and six isolates that are resistant to GCV or PFA. All the isolates had similar sensitivities to the analogs, and all isolates, including the GCV-resistant and PFA-resistant isolates, had significantly enhanced susceptibilities to the analogs than to the parent CDV. The cCDV analogs again had a pattern of susceptibility similar to that of the CDV analogs. This increase in activity is likely due to the increased uptake of the analogs compared to the parent CDV (1).

The other betaherpesviruses tested, HHV-6A and HHV-6B, were both susceptible to CDV and cCDV, and our results are consistent with previous reports (32, 37) for strains of HHV-6A. In our assay, the GS strain of HHV-6A was 5- to 20-fold as active (2.7 to 1.6 μM) for CDV and cCDV as results published previously (32, 37). The Z-29 strain of HHV-6B was inhibited by about 5 to 16 μM for the two parent compounds. In general, the susceptibilities of the two strains of HHV-6 were increased by 50- to 1,000-fold by the lipid ester modification of CDV or cCDV.

In our studies using three different assay methods, EBV was not susceptible to inhibition by CDV or cCDV. This is in contrast to the studies of Lin et al. (27) who reported that EBV was very susceptible to CDV, with an EC₅₀ of 0.03 μM, although in their assay CDV was left in the cultures for 14 days rather than the 3-day period in our assays and they used P3HR-1 cells rather than Daudi cells. Although EBV was not susceptible to CDV or cCDV in our assay system, the most active analogs, HDP-, ODP-, OLP-, and OLE-CDV and -cCDV, were very active in all three assay systems, suggesting that greater intracellular levels of drug obtained with the analogs were necessary for inhibition of EBV replication in Daudi cells. It has been reported previously (1) that the prototype, HDP-CDV, is taken up by cells much more readily than CDV and metabolized to CDV-diphosphate. This is also the case for ODE-CDV, OLE-CDV, and ODBG-CDV (K. Y. Hostetler and K. A. Aldern, unpublished data). Intracellular levels of CDV-diphosphate have been reported to be 100 times greater with HDP-CDV than that observed with CDV alone (1). Consistent with this interpretation, Daudi cells used for assay of EBV appeared to be much less sensitive to the toxicity of CDV than HFF cells. In general, the analogs exhibited a similar level of toxicity in both Daudi and HFF cells, suggesting that more drug was being taken up by Daudi cells from the lipid ester analogs. Another gammaherpesvirus, HHV-8, was susceptible to levels of 2.6 μM of CDV and 27 μM of cCDV compared to the EC₅₀ value of 6.3 μM for CDV that was reported by Neyts and De Clercq (28). While the EC₅₀ values for inhibition of HHV-8 replication by the lipid ester analogs were considerably reduced compared to the values for CDV and cCDV, the increase was generally less than seen for some of the other viruses, particularly CMV.

Not unexpectedly, a diverse pattern of activity was seen in the various subfamilies of herpesviruses, probably due to the fact that the life cycles of lymphoblastic herpesviruses, EBV, HHV-8, and HHV-6 are remarkably different than those of HSV, CMV, and VZV. Further differences with these viruses may be due to the fact they replicate in lymphocyte-derived cells rather than fibroblasts as do HSV, VZV, and CMV. As was reported previously (13, 20, 39), the analogs that were the most active were generally the most toxic both in cell killing (CC₅₀) and cell proliferation (IC₅₀) assays. Presumably this is due to the increased uptake of the drug into the cell because of the lipid conjugate and accelerated conversion to CDV-diphosphate (1). Nevertheless, due to the significantly greater antiviral activity of these analogs, they have very good selective indices (13, 20, 38).

As reported in previous studies, the ether lipid ester analogs of CDV, HDP-CDV, ODE-CDV, OLE-CDV, and OLP-CDV were generally the most active analogs (3, 20, 21, 22, 23, 31, 38) against all the herpesviruses tested. The fact that such greatly improved efficacy was observed against a number of different viruses, including the alpha, beta, and gamma subfamilies of herpesviruses as well as adenoviruses and orthopoxviruses, demonstrates the robustness of this strategy for improving the pharmacological properties of these analogs. Since HDP-CDV and ODE-CDV are in ongoing development and safety studies for orthopoxvirus infections (29), it is plausible to assume that one of these two compounds is the most likely candidate to be developed for use in herpesvirus, adenovirus, and other DNA virus infections.

The observations in this study with lipid ester analogs of CDV and cCDV against three subfamilies of herpesviruses suggest that these compounds could be used for the treatment of a variety of human herpesvirus infections, including those caused by the lymphotropic viruses, EBV, HHV-6, and HHV-8, which have few drug treatment options available. Although results from experiments in vitro and in vivo suggest that these compounds have the potential for use against a variety of DNA virus infections, additional safety and efficacy studies will be needed in order to determine the potential of CDV analogs for treatment of a variety of these infections.