Broad-spectrum antiviral activity of Virazole: 1-beta-D-ribofuranosyl-1,2,4-triazole-3-carboxamide.

Virazole is a synthetic nucleoside active in tissue culture against at least 16 DNA and RNA viruses. Applied topically, it inhibits herpetic keratitis in rabbits and tail lesions induced by herpes, vaccinia, and vesicular stomatitis viruses in mice. Injected intraperitoneally into mice, it inhibits splenomegaly and hepatomegaly induced by Friend leukemia virus and respiratory infections caused by influenza A(O), A(2), and B viruses and parainfluenza 1 virus. infections is also effective.

Early-initiated zidovudine therapy prevents disease but not low levels of persistent retrovirus in mice.

An F1 hybrid mouse strain containing the Rfv-3r/s genotype was inoculated with Friend virus complex (FV) and treated with zidovudine (ZDV) intraperitoneally three times daily for 20 days beginning as early as 10 min after initial viral exposure. This strain of mice develops FV-specific neutralizing antibodies that aid in reducing viremia and splenic virus titers but do not prevent splenomegaly and eventual FV-associated death. The virally exposed mice treated with ZDV did not develop splenomegaly or have detectable viremia after the last drug treatment. On day 21, a single animal had demonstrable virus in the spleen as determined by a focal immunoenzyme assay; 57% had detectable virus at 5 weeks, but non displayed splenic virus after 35 weeks. None of the animals died after the 35-week holding period, compared to 38% dying in placebo-treated mice. To detect low levels of the virus, or potentially latent virus, splenocytes were cocultivated with a cell line known to readily propagate FV, and the cells were subsequently passaged four times to amplify replication of the virus. After amplification, a significant increase was seen in the number of mice testing positive for virus. Thus, ZDV treatment initiated early after virus exposure was effective in preventing FV-induced splenomegaly and death, but did not prevent low levels of persistent retrovirus in the mice.

HIV-1 LTR activation model: evaluation of various agents in skin of transgenic mice.

Mice containing the HIV-1 long terminal repeat (LTR) regulating the expression of firefly luciferase reporter gene were investigated for their use as a model for activation of the LTR. As a limited test of this model, a number of different factors were screened for their ability to affect reporter gene activities in the skin. Reporter gene levels were increased in the skin by topical treatment of dimethylsulfoxide, retinoic acid, phorbol ester, ultraviolet light, and hydrogen peroxide, all of which have previously been shown to cause increased HIV production in cultured human cells. Topically applied arachidonic acid, histamine, ethanol, acetone, and methanol did not increase reporter gene activities. A lack of published reports on activation of HIV-1 in human cells by these agents suggests that they do not activate viral expression in human cells, which corroborates with the findings of this report. Minor forms of skin wounding and intraperitoneally administered psoralen plus ultraviolet light also increased reporter gene activities in skin. Control and test treatments could be performed on the same mouse and repetitive samples could be obtained from each treatment area. These transgenic mice might be useful as predictive models for regulation of the LTR in epidermal or dendritic cells.

Effects of zidovudine on Friend virus complex infection in Rfv-3r/s genotype-containing mice used as a model for HIV infection.

Infection with the Friend virus complex (FV) in (B10.A x A/WySn)F1 mice containing the Rfv-3r/s genotype results in several disease manifestations analogous to those seen in patients with acquired immune deficiency syndrome, predominantly high levels of specific antibody and low levels of infectious virus with eventual retroviral disease-induced death of the host. Other immunologic manifestations of FV infection in this murine host included inhibition of percent total T, T-helper, and T-suppressor/cytotoxic cells of total splenic lymphocytes and phytohemagglutinin-induced response of spleen cells. Interleukin-1 production was not affected but the numbers of splenic B cells were increased by the infection. 3'-Azido-3'-deoxythymidine (zidovudine, AZT) administered (a) intraperitoneally three times daily for 24 days beginning 4 h after virus inoculation in doses of 60 to 480 mg/kg/day, (b) in drinking water for 22 days beginning 4 h after virus inoculation in doses of 22 to 216 mg/kg/day, or (c) in drinking water for 29 days beginning 6 days after virus inoculation in doses of 22 to 216 mg/kg/day markedly inhibited FV-induced disease. In the mice receiving early-initiated AZT therapy, FV-induced splenomegaly and hematocrit values were inhibited and infectious centers in the spleen and FV titers in the plasma were reduced to below detectable levels at the higher AZT dosage levels. The percent of total T cells in splenic lymphocytes was increased in the infected, AZT-treated mice. In the intraperitoneal experiment, FV disease-induced death was prevented by treatment with all doses of AZT. Neutralizing antibody to FV was significantly reduced in all AZT-treated groups. Toxicologic manifestations of these AZT treatments included splenic enlargement and reduced hematocrit, although all treated, uninfected mice survived the treatments, gained weight, and displayed no significant effects on enumeration of T and B cells.

Transgenic mice as a chemotherapeutic model for hepatitis B virus infection.

Many transgenic mice carrying either portions of the hepatitis B virus (HBV) genome, or the complete genome, have been developed as models because HBV does not infect any other organisms besides humans and chimpanzees to cause a productive infection and disease. Some of these models have been useful in evaluating chemotherapeutic agents such as interferon-alpha, interleukins-2 and -12, other cytokines and nucleoside analogues for efficacy against HBV. A recently developed transgenic mouse (Guidotti et al., Journal of Virology 69:6158-6169.) which supports the replication of high levels of infectious HBV, provides the opportunity to evaluate the effect of antiviral drugs on various portions of the HBV life cycle in the whole animal. Evaluation of lamivudine, zidovudine and interferon-alpha B/D (IFN-alpha) in this HBV transgenic mouse model are described. Lamivudine and IFN-alpha were highly efficacious in reducing serum HBV DNA. As might be predicted, zidovudine was not efficacious. IFN-alpha was more effective in reducing virus titres in male mice as compared to female mice. This gender difference was not due to lower ability of female mice to express the virus. One anticipates that as this high level HBV transgenic-expressing mouse becomes more fully developed as a chemotherapeutic model, questions about the efficacy of different agents, routes of administration, synergy of antiviral combinations and novel drug therapies will be answered.

Synthesis and antiviral and antimicrobial activity of certain 1-beta-D-ribofuranosyl-4,5-disubstituted imidazoles.

Starting with AICA ribonucleoside the following nucleosides were prepared. Methyl 5-amino-1-(2,3,5-tri-O-acetyl-beta-D-ribofuranosyl)imidazole-4-carboxylate (5) was converted into methyl 5-chloro-1-(2,3,5-tri-O-acetyl-beta-D-ribofuranosyl)imidazole-4-carboxylate (6) via diazotization in the presence of cuprous chloride. Similarly, 5-amino-1-(2,3,5-tri-O-acetyl-beta-D-ribofuanosyl)imidazole-4-carbonitrile (9) was converted into 5-chloro-, 5-bromo-, and 5-iodo-1-(2,3,5-tri-O-acetyl-beta-D-ribofuranosyl)imidazole-4-carbonitrile derivatives. These 5-halogenated imidazole nucleosides were treated with several nucleophiles such as ammonia, hydroxylamine, and hydrogen sulfide to provide, respectively, 5-haloimidazole-4-carboxamide, 5-haloimidazole-4-carboxamidoxime, and 5-haloimidazole-4-thiocarboxamide ribonucleosides. 5-Chloro- or 5-bromo-1-(2,3,5-tri-O-acetyl-beta-D-ribofuranosyl)imidazole-4-carbonitrile was treated with potassium hydrosulfide to yield 5-mercapto-1-beta-D-ribofuranosylimidazole-4-thiocarboxamide (16). The catalytic reduction of 5-chloro- or 5-bromo-1-beta-D-ribofuranosylimidazole-4-carboxamidoxime provided 1-beta-D-ribofuranosylimidazole-4-carboxamidines as their hydrochloride and hydrobromide salts, respectively. These nucleosides were tested for in vitro antiviral, antifungal, and antibacterial activity. The 5-halo analogues of 1-beta-D-ribofuranosylimidazole-4-carboxamide showed significant antiviral activity whereas compound 16 was found inhibitory to fungi.

Enantiomeric synthesis of D- and L-cyclopentenyl nucleosides and their antiviral activity against HIV and West Nile virus.

Enantiomeric synthesis of D- and L-cyclopentenyl nucleosides and their antiviral activity against HIV and West Nile virus are described. The key intermediate (-)- and (+)-cyclopentenyl alcohols (7 and 15) were prepared from D-gamma-ribonolactone and D-ribose, respectively. Coupling of 7 with appropriately blocked purine and pyrimidine bases via the Mitsunobu reaction followed by deprotection afforded the target L-(+)-cyclopentenyl nucleosides (24-28, 31, 33, and 36). D-(-)-Cyclopentenyl nucleosides (1, 40, 43, and 52-56) were also prepared by a similar procedure for L-isomers from 15. The synthesized compounds were evaluated for their antiviral activity against two RNA viruses: HIV and West Nile virus. Among the synthesized D-(-)-nucleosides, adenine (1, neplanocin A), cytosine (55, CPE-C), and 5-fluorocytosine (56) analogues exhibited moderate to potent anti-HIV activity (EC(50) 0.1, 0.06, and 5.34 microM, respectively) with significant cytotoxicity in PBM, Vero, and CEM cells. Also, cytosine (55) and 5-fluorocytosine (56) analogues exhibited the most potent anti-West Nile virus activity (EC(50) 0.2-3.0 and 15-20 microM, respectively). Among L-(+)-nucleosides, only the cytosine (27) analogue exhibited weak anti-HIV activity (EC(50) 58.9 microM).

Synthesis and antiviral acticity of some phosphates of the broad-spectrum antiviral nucleoside, 1-beta-D-ribofuranosyl-1,2,4-triazole-3-carboxamide (ribavirin).

1-beta-D-Ribofuranosyl-1,2,4-triazole-3-carboxamide 5'-phosphate (2) was prepared and converted into the following derivatives: the 5'-phosphoramidate 3, the 5'-diphosphate 4, the 5'-triphosphate 5, and the cyclic 3',5'-phosphate 6. The cyclic 2',3'-phosphate 7 was prepared from the parent nucleoside, 1-beta-D-ribofuranosyl-1,2,4-triazole-3-carboxamide (1), and was opened to the 2'(3')-phosphate 8. These compounds were found to exhibit significant antiviral activity against several viruses in cell culture. Ribavirin 5'-phosphate (2) was shown to be effective when tested against lethal infections in mice caused by influenza A2, influenza B, and murine hepatitis viruses.

Synthesis and antitumor and antiviral activities of 1-beta-D-arabinofuranosyl-2-amino-1,4(2H)-iminopyrimidine and its derivatives.

1-beta-D-Arabinofuranosyl-2-amino-1,4(2H)-imino-5-fluoropyrimidine (10), 1-beta-D-arabinofuranosyl-2-amino-1,4(2H)-imino-5-fluoropyrimidine 3'-phosphate (9), and 1-beta-D-arabinofuranosyl-2-amino-1,4(2H)-imino-5-chloropyrimidine (11) have been synthesized from 2,2'-anhydro-1-beta-D-arabinofuranosyl-5-fluorocytosine (5), 2,2'-anhydro-1-beta-D-arabinofuranosyl-5-fluorocytosine 3'-phosphate (4), and 2,2'-anhydro-1-beta-D-arabinofuranosyl-5-chlorocytosine (6), respectively. 2,2'-Anhydro-1-beta-D-arabinofuranosylcytosine 3'-phosphate (7), 1-beta-D-arabinofuranosyl-2-amino-1,4-(2H)-iminopyrimidine (13), 1-beta-D-arabinofuranosyl-2-amino-1,3(2H)-iminopyrimidine 3'-phosphate (12), and compounds 4, 5, and 9 showed significant in vitro activity against a number of DNA viruses. Compounds 7 and 12 were also effective in vivo against type 1 herpes simplex virus. Compounds 7, 12, and 13 were extremely effective in the treatment of mice bearing leukemia L1210.

Synthesis and antiviral activity of certain 5'-monophosphates of 9-D-arabinofuranosyladenine and 9-D-arabinofuranosylhypoxanthine.

A number of 5'-phosphates of 9-D-arabinofuranosyladenine and 9-D-arabinofuranosylhypoxanthine were prepared and tested against a variety of DNA viruses in tissue culture. The syntheses of the antiviral agent 9-beta-D-arabinofuranosylhypoxanthine 5'-monophosphate (6) and a series of related nucleotides, 9-beta-D-arabinofuranosyladenine 5'-O-methylphosphate (3), 9-beta-D-arabinofuranosylhypoxanthine 5'-O-methylphosphate (7), 9-beta-D-arabinofuranosylhypaxanthine cyclic 3',5'-phosphate (13), and 9-alpha-D-arabinofuranosylhypoxanthine 5'-monophosphate (17), are described. The concepts underlying the development of these antiviral agents are discussed. Comparison of the anti-DNA viral activity is made with 9-beta-D-arabinofuranosyladenine (ara-A). Reproducible antiviral activity against three DNA viruses in vitro at nontoxic dosage levels is demonstrated by 3,6, and other related nucleotides.

Synthesis and anti-DNA -virus activity of the 5'-monophosphate and the cyclic 3',5'-monophosphate of 9-(beta-D-xylofuranosyl) guanine.

9-(beta-TD-xylofuranosyl)guanine (xylo-G) was converted chemically to the 9-(beta-D-xylofuranosyl)guanine 5'-monophosphate (xylo-GMP) and 9-(beta-D-xylofuranosyl)guanine cyclic 3',5'-monophosphate (c-xylo-GMP). These compounds were tested against a variety of DNA viruses in tissue culture in parallel with 9-(beta-D-arabinofuranosyl)adenine (ara-A). This evaluation revealed that xylo-G, xylo-GMP, and c-xylo-GMP were all moderately active but less effective than ara-A. When the four compounds were administered intracerebrally as a treatment for herpes virus, type 1 induced encephalitis in mice, c-xylo-GMP exhibited superior activity to that shown by the other three. When administered intraperitoneally, c-xylo-GMP was found to have a therapeutic index of about 4, which is less than that for ara-A (approximately 30) in the same system.

Synthesis and antiviral activity of certain thiazole C-nucleosides.

A general reaction of glycosyl cyanides with liquid hydrogen sulfide in the presence of 4-dimethylaminopyridine to provide the corresponding glycosylthiocarboxamides is described. These glycosylthiocarboxamides were utilized as the precursors for the synthesis of 2-D-ribofuranosylthiazole-4-carboxamide and 2-beta-D-ribofuranosylthiazole-5-carboxamide (23). The structural modification of 2-beta-D-ribofuranosylthiazole-4-carboxamide (12) into 2-(2,3,5-tri-O-acetyl-beta-D-ribofuranosyl)thiazole-4-carboxamide (15), 2-beta-D-ribofuranosylthiazole-4-thiocarboxamide (17), and 2-(5-deoxy-beta-D-ribofuranosyl)thiazole-4-carboxamide (19) is also described. These thiazole nucleosides were tested for in vitro activity against type 1 herpes virus, type 3 parainfluenza virus, and type 13 rhinovirus and an in vivo experiment was run against parainfluenza virus. They were also evaluated as potential inhibitors of purine nucleotide biosynthesis. It was shown that the compounds (12 and 15) which possessed the most significant antiviral activity were also active inhibitors (40-70%) of guanine nucleotide biosynthesis.

In vitro and in vivo assay systems for study of influenza virus inhibitors.

Evaluation of potential influenza virus inhibitors may utilize multiple steps. First would be to determine if the viral target (e.g. influenza virus neuraminidase) being focused upon will be inhibited in the appropriate assay. Standard in vitro antiviral assays, used next in antiviral evaluations, may utilize inhibition of viral plaques, viral cytopathic effect (CPE), and viral hemagglutinin or other protein, with inhibition of viral yield used in follow-up evaluations. The CPE can be determined visually and by dye uptake. Animal models used for study of potential influenza virus inhibitors include the ferret, the laboratory mouse, and the chicken, with a variety of parameters used to indicate the severity of the infection and its inhibition by therapy. Multiple parameters are recommended in any in vivo antiviral evaluation. The ferret and the mouse infection models have been useful in studying the development of drug resistance and the relative virulence of drug-resistant viruses. The influenza mouse model has also been of value for the evaluation of immunomodulating effects of test compounds and for the study of the utility of antiviral drugs for use against influenza virus infections in the immunocompromised host. In considering the use of any animal model, species differences in drug pharmacology and metabolism must be taken into account. This review has described the systems which have been used most frequently by antiviral investigators, using, as examples, recent studies with the clinically approved influenza virus neuraminidase inhibitors oseltamivir and zanamivir.

Intranasal treatment of cowpox virus respiratory infections in mice with cidofovir.

Orthopoxvirus infections in mice have been effectively treated with cidofovir, a clinically approved drug given by intravenous infusion to treat cytomegalovirus infections. In a bioterrorist scenario it would be technically difficult to give this drug to a large number of exposed individuals. New treatment approaches are being sought, which include giving cidofovir by alternative routes or designing oral prodrugs of cidofovir. In this report, intranasal cidofovir was investigated as a treatment of pulmonary cowpox virus infections in BALB/c mice. Ninety to 100% of animals given a single intranasal drug treatment (10, 20 or 40 mg/kg) 24 h after virus challenge survived the infection, whereas all placebo-treated mice died. Doses of 2.5 and 5 mg/kg resulted in 60 and 80% survival, respectively. Single treatments of 20 and 40 mg/kg could be given up to 3 days after virus inoculation and still be 80-90% protective. A single 40 mg/kg treatment of infected mice given 1 or 2 days after infection also resulted in statistically significant decreases in virus titer in lungs and nose/sinus compared to the placebo group. Drug efficacy was found to be contingent upon treatment volume. A 10 mg/kg intranasal dose given 24 h after virus challenge was 100 and 50% effective in volumes of 40 and 20 microl, respectively. The same dose in 5 and 10 microl volumes caused no decrease in mortality. The results of these studies establish the utility of cidofovir treatment of poxvirus infections in mice by intranasal route. The data suggest the possibility that aerosol delivery of cidofovir to human lungs may be a viable alternative to intravenous dosing.

Combination of antiviral immunotoxin and ganciclovir or cidofovir for the treatment of murine cytomegalovirus infections.

The effects of two anti-murine cytomegalovirus (MCMV) immunotoxins used in combination with ganciclovir (GCV) or cidofovir (HPMPC) against MCMV were determined in vitro and in mice. The inhibitors were added to cell cultures 24 or 48 h after MCMV adsorption so as to not affect the initial infection rate. The immunotoxins (0.63, 1.25 and 2.5 micrograms/ml) combined with GCV (1.25, 2.5 and 5 microM) or HPMPC (0.03, 0.06 and 0.12 microM) caused synergistic inhibition of virus yield in C127I cells at most of the combinations tested. No toxic effect on cell growth in culture was observed at these immunotoxin/drug combinations. The effects of immunotoxin and GCV treatment were studied further in MCMV-infected severe combined immunodeficient (SCID) mice. Immunotoxin (1 mg/kg per day) given by intraperitoneal (i.p.) injection on days 1, 4 and 7 of the infection did not extend the mean day to death compared with the placebo group. Once daily i.p. treatment with GCV (50 mg/kg per day) for days starting at 24 h after virus inoculation extended survival time almost 11 days. The combination of immunotoxin plus GCV was better than GCV alone, extending the mean day to death an additional 2 to 3 days, which is suggestive of a synergistic effect.

Utilization of transgenic mice replicating high levels of hepatitis B virus for antiviral evaluation of lamivudine.

A recently developed transgenic mouse strain which expresses high levels of hepatitis B virus (HBV) was studied as a model for evaluation of potential chemotherapeutic agents. Lamivudine ([-]2'-deoxy-3'-thiacytidine), known to reduce hepatitis B viremia in human patients, and zidovudine (3'-azido-3'-deoxythymidine), previously shown to be ineffective for HBV infections in man, were used in parallel in this transgenic animal model. Orally administered lamivudine at dosages of 100, 50, and 25 mg/kg per day given once a day for 21 days significantly decreased serum and liver HBV DNA titers in a dose-responsive manner. Zidovudine (approximately 22 mg/kg per day) administered in the drinking water for 21 days was not effective in reducing these HBV parameters as compared to transgenic placebo-treated controls. The serum HBV DNA titers rebounded to high levels 1 week after cessation of lamivudine treatment. Male and female mice responded in a similar manner to these therapies. The results using this transgenic mouse model were similar to what would be predicted from treatment of HBV-infected human patients with lamivudine and zidovudine, and indicate these mice may be useful as a small animal chemotherapeutic model for study of potential HBV inhibitors.

Prophylactic and therapeutic activities of 7-thia-8-oxoguanosine against Punta Toro virus infections in mice.

The biological response modifier 7-thia-8-oxoguanosine was evaluated in mice against the hepatotropic Adames strain of Punta Toro virus. When administered intraperitoneally in divided doses, significant protection from death was conferred at doses of 50 and 100 mg/kg/day given 24 and 17 h pre-virus inoculation, 25-100 mg/kg/day administered 4 h pre- and 3 h post-virus challenge, and 12.5 to 100 mg/kg/day administered 24 and 31 h after virus inoculation. These doses preventing death reduced liver icterus scores, serum alanine aminotransferase and aspartate aminotransferase levels, and liver and serum virus titers relative to placebo controls. Full daily doses administered at 24 h were somewhat less protective to mice than divided daily doses starting at the same time. The initiation of treatment could be delayed as late as 36 h after virus inoculation, resulting in complete protection from mortality at 100 mg/kg/day. This prevention of death occurred despite the acute nature of the infection which resulted in deaths by 96 h in the placebo-treated controls. These results show that 7-thia-8-oxoguanosine has both prophylactic and therapeutic potential as an anti-Phlebovirus agent. Interferon induction appears to be the reason for antiviral activity in this model, since up to 10,000 units of interferon/ml were induced in mice 1 h after treatment with 100 mg 7-thia-8-oxoguanosine per kg, and antibody to interferon alpha/beta administered shortly after treatment with the nucleoside negated the antiviral effect.

Effects of ribamidine, a 3-carboxamidine derivative of ribavirin, on experimentally induced Phlebovirus infections.

Ribamidine (1-beta-D-ribofuranosyl-1,2,4-triazole-3-carboxamidine) was inhibitory in rhesus monkey kidney (LLC-MK2 derivative) cells to Adames and Balliet strains of Punta Toro virus (PTV), a Phlebovirus related to Rift Valley fever and sandfly fever viruses. The 50% effective dose was 8 and 12 micrograms/ml against each respective virus strain; the 50% cytotoxic dose was 320 micrograms/ml, giving selectivity indices of 40 and 27 against each virus strain. The virus ratings were 1.2 and 1.0, respectively. In radiolabel uptake studies, ribamidine had a moderate effect on [3H]leucine uptake at dosages down to 1 microgram/ml, but [3H]thymidine, [32P], and [3H]uridine were inhibited at high (100-1000 micrograms/ml) doses only. Subcutaneous (s.c.) and oral treatments of Adames PTV-infected mice were equally highly effective, as evidenced particularly by 100% survivors. Reduced hepatic icterus, serum oxalic acid transaminase, serum glutamic pyruvic acid transaminase, and recoverable virus titers from livers and sera of infected mice were also seen as a result of ribamidine treatment. Twice daily treatment for 5 days could be started as late as 72 h post-virus inoculation (p.v.i.) with significant inhibition of PTV infection seen. Single s.c. treatments administered as late as 48 h p.v.i. were similarly effective. Using the chronic therapy schedule, the maximum tolerated dose was 1000 mg/kg/day and the minimum effective dose was 31.3 to 62.5 mg/kg/day. Using single treatment, a maximum tolerated dose was greater than 1000 mg/kg, and the minimum effective dose was 125 mg/kg. Ribamidine s.c. treatment of mice infected intracerebrally with the Balliet strain of PTV resulted in a moderate infection-inhibitory effect, seen especially by reduced virus titers in the brains of the infected, treated mice.

Antiviral activities of nucleosides and nucleotides against wild-type and drug-resistant strains of murine cytomegalovirus.

Resistance of human cytomegalovirus to approved antiviral drugs is becoming a problem of increasing concern. In order to further study drug resistance in a related virus, strains of murine cytomegalovirus (MCMV) have been prepared in vitro by extensive adaptation of the virus to increasingly higher concentrations of either ganciclovir, foscarnet, or (S)-9-(3-hydroxy-2-[phosphonylmethoxy]propyl)cytosine (HPMPC). Plaque reduction 50% effective concentrations (EC50) for the above inhibitors increased 9-, 7-, and 23-fold, respectively (against the corresponding virus), compared to wild-type MCMV. Each virus was then evaluated against other known anti-MCMV agents to determine cross-resistance patterns. These compounds included 3-hydroxy-phosphonylmethoxypropyl derivatives of adenine (HPMPA) and guanine (HPMPG), 2-phosphonylmethoxyethyl derivatives of adenine (PMEA) and 2,6-diaminopurine (PMEDAP), cyclobutylguanine, acyclovir, and the methylene phosphonate derivatives of acyclovir (SR3722) and ganciclovir (SR3773). The ganciclovir-resistant MCMV was cross-resistant to foscarnet, HPMPA, HPMPC, HPMPG, SR3722, and SR3773. The foscarnet-resistant virus was also resistant to acyclovir, PMEA, PMEDAP, SR3722, and SR3773. The HPMPC-resistant MCMV was cross-resistant to HPMPA, HPMPG, and SR3773. Changes in susceptibility were from 3- to 22-fold relative to the wild-type virus. Virus yield reduction data correlated with the plaque assay results. Only cyclobutylguanine was approximately equally active against wild-type and the three drug-resistant MCMVs. The patterns of cross-resistance correlated with resistance seen in human cytomegalovirus strains expressing altered DNA polymerase function. The GCV-resistant and HPMPC-resistant viruses were markedly attenuated in their ability to kill severe combined immunodeficient mice.

Selective inhibition of cytomegaloviruses by 9-(3'-ethylphosphono-1'-hydroxymethyl-1'-propyloxy-methyl)g uanine.

9-(3'-ethylphosphono-1'-hydroxymethyl-1'-propyloxy-methyl)gu anine (SR 3727A) was significantly inhibitory to strain AD169 of human cytomegalovirus (HCMV) utilizing plaque reduction and inhibition of intra- and extracellular virus yield in MRC-5 cells. The 50% effective concentrations (EC50) ranged from 6-17 microM for three laboratory strains of HCMV, whereas the 50% cytotoxic doses were > 4200 microM as determined by viable cell assay and inhibition of radiolabeled precursors into DNA, RNA and protein. EC50 values against ganciclovir-sensitive clinical isolates ranged from 8-47 microM. Against two ganciclovir-resistant strains of HCMV, EC50 values of SR 3727A were 84 and 320 microM; against murine CMV (MCMV); 17 microM and against guinea pig CMV, 56 microM. SR 3727A was most effective when infected cells were treated 24 h or less after virus adsorption. BALB/c mice infected intraperitoneally (i.p.) with a lethal dose of MCMV were treated i.p. with 31.3, 62.5, 125, or 250 mg/kg/day of SR 3727 twice daily for 5 days beginning 4 h pre-virus inoculation. All doses were well tolerated; the 125 and 250 mg/kg/day doses significantly prevented death. In a second experiment, SR 3727 at 125 mg/kg/day markedly reduced titers of recoverable virus from spleens, kidneys, and salivary glands harvested at varying times after virus inoculation.