Metabolism of ganciclovir and cidofovir in cells infected with drug-resistant and wild-type strains of murine cytomegalovirus

ELSEVIER

QIAntiviral Research

Antiviral Research 35 (1997) 83 90

Metabolism of ganciclovir and cidofovir in cells infected with drug-resistant and wild-type strains of murine cytomegalovirus K e v i n M. Okleberry a, Reed P. Warren

b Donald F.

S m e e a,*

a Institute for Antiviral Research, Utah Sate University, Logan, Utah USA b Department of Biology, Utah State University, Logan, Utah USA

Received 12 November 1996; accepted 3 March 1997

Abstract Murine cytomegalovirus (MCMV) has been used extensively as an animal model for human cytomegalovirus (HCMV). Understanding drug resistance and its treatment in MCMV may lead to more effective treatments of HCMV disease. Most ganciclovir-resistant HCMV clinical isolates exhibit a decreased capacity to induce ganciclovir phosphorylation (to its biologically active form) in infected cells. Using an M C M V strain resistant to both ganciclovir and cidofovir, the intracellular metabolism of these drugs was studied to determine if M C M V resistance correlates with decreases in drug phosphorylation. The wild-type (WT) M C M V used for comparison was inhibited in plaque reduction assays, by ganciclovir and cidofovir by 50% at 5.1 and 0.24 p M, respectively; the resistant strain was inhibited at 72 and 2.7/~M, respectively. In uninfected, WT, or resistant virus-infected cells, the extent of metabolism of 10/zM ganciclovir or 1 /zM cidofovir to intracellular triphosphorylated species was similar. Phosphorylation and catabolism (following drug removal) rates over time were also similar. Intracellular levels of ganciclovir triphosphate and cidofovir diphosphate increased less than two-fold with increasing multiplicity of virus infection. Because few differences in drug phosphorylation between WT and resistant virus-infected cells were found, virus resistance to ganciclovir and cidofovir apparently is not linked to altered drug phosphorylation. Since the viral D N A polymerase is the antiviral target for these compounds, the resistant M C M V is most likely a D N A polymerase mutant. © 1997 Elsevier Science B.V. Keywords: Antiviral; Ganciclovir; Cidofovir; Drug resistance; Phosphorylation; Murine cytomegalovirus

1. Introduction * Corresponding author. Present address: Virology Division, US Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick, MD 21702-5011, USA.

T h e c y t o m e g a l o v i r u s e s are u b i q u i t o u s agents t h a t c o m m o n l y infect m a n y species o f a n i m a l s

0166-3542/97/$17 00 © 1997 Elsevier Science B.V. All rights reserved. PII S01 66-3 542(97)00013-2

84

K.M. Okleberry et al./Antiviral Research 35 (1997) 83 90

and man (Alford and Britt, 1990). In recent years infections due to human cytomegalovirus (HCMV) have grown in significance largely due to the AIDS epidemic. HCMV disease in immunocompromised patients is being treated with the drugs ganciclovir (Laskin et al., 1987), foscarnet (Walmsley et al., 1988), and more recently cidofovir (Polis et al., 1995). Strains of HCMV have been isolated which are resistant to one or more of these compounds and to other antiviral substances under clinical or preclinical development (Biron, 1991; Sullivan and Coen, 1991; Tatarowicz et al., 1992; Sullivan et al., 1993). Because of the problem of drug resistance, a need exists to identify ways of preventing or greatly delaying the emergence of drug-resistant HCMV. HCMV possesses a unique enzyme (kinase) capable of phosphorylating ganciclovir to ganciclovir monophosphate in infected cells (Littler et al., 1992; Sullivan et al., 1992). Ganciclovir monophosphate is then readily converted by cellular enzymes to ganciclovir triphosphate (Smee et al., 1985). Ganciclovir triphosphate is the active form of the drug which inhibits the viral DNA polymerase (Freitas et al., 1985). The majority of the clinical H C M V isolates resistant to ganciclovir have mutations in this viral kinase (Biron, 1991), so that infected cells produce low intracellular concentrations of ganciclovir triphosphate, levels that are inadequate for virus inhibition. Mutations of the viral DNA polymerase can also render the virus less susceptible to inhibition by antiviral drugs (Tatarowicz et al., 1992; Sullivan et al., 1993; Baldanti et al., 1995, 1996). Our efforts have focused on studying murine cytomegalovirus (MCMV) as a model for HCMV. Because MCMV causes severe disease in susceptible mice (Kern, 1991), the treatment of infections in vivo caused by wild type (WT) and drug resistant viruses can be studied. Recently, we reported the development of three drug-resistant strains of MCMV with different sensitivities to a number of anti-cytomegalovirus compounds (Smee et al., 1995). One such virus (designated GCV-r strain), was cross-resistant to both ganciclovir and cidofovir. The reason this virus is less susceptible to inhibition by these drugs compared to WT virus is being investigated.

As part of this effort, we examined the differences and similarities in phosphorylation of ganciclovir and cidofovir in uninfected cells and in cells infected with WT and resistant MCMVs. The results indicate that the phosphorylation of each drug is not appreciably different in cells infected with either virus. Thus, the mode of drug resistance of the GCV-r virus is not due to a decreased capability of infected cells to phosphorylate ganciclovir or cidofovir.

2. Materials and methods

2. I. Antiviral compounds

Ganciclovir (GCV) (Roche Laboratories, Nutley, N J) was purchased from a local pharmacy. Cidofovir (HPMPC) was obtained from Norbert Bischofberger, Gilead Sciences, Foster City, CA. [3H]ganciclovir and [3H]cidofovir were purchased from Moravek, Brea, CA. Ganciclovir was labeled on the guanine methyl group, while cidofovir was labeled on the hydrogen atom of the C5 position of cytosine. Constant amounts of radioactive compounds (20/~Ci/ml) were added to varying concentrations of non-radioactive ganciclovir or cidofovir for each reaction condition. 2.2. Virus and cells

The WT Smith strain of MCMV was obtained from the American Type Culture Collection (ATCC), Rockville, MD. The plaque purified ganciclovir/cidofovir-resistant MCMV (referred to previously as GCV-r strain) was developed by propagating the WT virus in increasing concentrations of ganciclovir (Smee et al., 1995). Plaque purification was accomplished by recovering the GCV-r strain at 65 days from salivary glands of severe combined immunodeficient (SCID) mice which had been treated with ganciclovir (50 mg/ kg per day) for the first month. The recovered virus was diluted to extinction and a single isolate was selected and expanded into a virus pool. C127I cells, derived from mouse mammary tumors, were obtained from ATCC. The cells were grown in Dulbecco's high glucose medium with 5

K.M. Okleberry et al./ Antiviral Research 35 (1997) 83-90

mM HEPES buffer, 10% fetal bovine serum (FBS) (Hyclone, Logan, UT), and 0.2% sodium bicarbonate. The cells were used to prepare large pools of virus and to conduct the antiviral and metabolism experiments. During virus infections, the FBS concentration in the culture medium was reduced to 2%. 2.3. Plaque reduction and virus yield assays

Sensitivities of the WT and plaque purified drug-resistant MCMVs to ganciclovir and cidofovir were evaluated by the published procedure (Smee et al., 1995). Twelve-well plates of C127I cells were infected with about 100 plaques per well, treated with drug, and incubated for 7 (WT virus) or 10 (resistant virus, which replicates slower than WT virus in culture) days prior to fixing, staining, and counting the plaques. Results, expressed as 50% effective drug concentrations (ECs0 values), represent means from three independent experiments. To determine the effects of ganciclovir and cidofovir on virus yield (which was conducted in parallel with one of the drug metabolism studies), cells in 12-well plates were infected with MCMV at a multiplicity of infection of one virus per cell. After virus adsorption (1-1.5 h) the cells were treated continuously with 10 pM ganciclovir or 1 p M cidofovir. Each day a portion of the infected cells were frozen, thawed, and sonicated for 1 min. Subsequently, the supernatant (representing both intracellular and extracelluar virus produced during the infection) was titrated by plaque assay on new monolayers of cells. After 7 or 10 days, the cells were fixed, stained, and plaques counted. 2.4. Determination o f drug metabolism in cells

The metabolism of ganciclovir and cidofovir to phosphorylated derivatives was studied in uninfected cells, and in cells infected with WT and drug-resistant viruses. Many parameters were studied, including different drug concentrations, varying numbers of infecting viruses per cell (multiplicity of virus infection), different times

85

of incubation of drug and virus on cells, and effect of drug removal on nucleotide catabolism. The parameters applicable to each experiment are indicated in the table and figures. Confluent monolayers of C127I cells (infected or uninfected) in 25 cm 2 flasks were radiolabeled with drug. Following incubation at 37°C, the medium was removed and the cells were rapidly rinsed once with drug-free medium. Following aspiration of the medium, the monolayers were treated with 3.5% perchloric acid for 5 min at 4°C, which caused the release of nucleosides and nucleotides from cells. The acidic fraction was collected and neutralized with 1 N KOH/0.2 M imidazole (1,3-diazo-2,4-cyclopentadiene). The precipitate which formed was eliminated by low speed (600 × g) centrifugation, and the supernatant was frozen at - 7 0 ° C prior to nucleotide analysis. Strong anion exchange (SAX) high pressure liquid chromatography (HPLC) was employed to separate phosphorylated derivatives of ganciclovir and cidofovir using our previously described method (Smee et al., 1985). This entailed using a 10 mm x 25 cm SAX column (Whatman, Clifton, N J). Elution of ganciclovir nucleotides was carried out using a linear gradient of 0.01 to 1 M NHaHPO 4 buffer (pH 3.5) from 0 to 35 min, followed by the 1 M buffer being run for an additional 5 min. Elution of cidofovir nucleotides was done using the same buffer but with a 45 min gradient followed by 5 min at 1 M buffer. One minute fractions were collected into scintillation vials for radioactivity counting. Counts of radioactivity in each peak were converted to pmol/106 cells, having first determined cell numbers in untreated control flasks using a hemacytometer. The monophosphate, diphosphate, and triphosphate metabolites of ganciclovir were eluted from the column at 8, 23 and 39 min, respectively. Three metabolites of cidofovir, the phosphate-choline derivative, monophosphate, and diphosphate, were eluted at 11, 20 and 33 min, respectively. These elution times were similar to those reported by us and others (Smee et al., 1985; Ho et al., 1992).

K.M. Okleberry et al./Antiviral Research 35 (1997) 83 90

86

Table 1 Metabolites of G C V and H P M P C formed after 24 h in uninfected cells, or in cells infected with W T and drug-resistant strains of MCMV Metabolite

GCV-P a GCV-PP GCV-PPP b HPMPC-P-choline HPMPC-P HPMPC-PP b

Metabolite (pmol/106 cells) formed in Uninfected cells

W T virus infected cells

Resistant virus infected cells

0.019 0.066 0.122 0.093 0.018 0.027

0.037 0.110 0.154 0.108 0.013 0.036

0.221 0.166 0.147 0.108 0.014 0.032

+ 0.012" + 0.013 _+ 0.023 + 0.022 _+ 0.006 +_ 0.016

+ 0.012 + 0.047 _+ 0.004 _+ 0.035 _+ 0.002 _+ 0.008

+ 0.030 __+0.099 _+ 0.011 + 0.046 + 0.001 _+ 0.009

The extracellular drug concentrations were: GCV, 10 /iM or H P M P C , 1 /~M. The multiplicity of virus infection was one virus per cell. alndicates phosphate group on drug. bThe antivirally active form of the compound. * Mean _+ S.D. for three independent assays.

3. Results

3.1. In vitro activities of drugs against W T and resistant M C M V In plaque reduction assays, the W T and resistant strains of M C M V were inhibited by 50% at ganciclovir concentrations of 5.1 and 72 #M, respectively. Mean ECs0 values for cidofovir were 0.24 and 2.7 /~M, respectively, against these viruses. Thus, the compounds were at least tenfold more potent against the W T virus than the resistant MCMV. The ECs0 values against the WT and plaque purified resistant virus were similar to previously reported values (Smee et al., 1995), and confirm that the plaque purification procedure selected for a virus still resistant to the two drugs. 3.2. Drug metabolism in infected and uninfected

cells Phosphorylation of ganciclovir and cidofovir was initially studied to establish what the differences in drug metabolism were in cells infected with W T and resistant MCMVs following a 24 h incubation period. The concentrations of ganciclovir (10 /2M) and cidofovir (1 /LM) used were ones that inhibited the replication of the WT virus

but were sub-inhibitory to the resistant virus. With regard to intracellular formation of ganciclovir triphosphate, which is the antivirally active form of the drug, approximately the same amount of this metabolite was formed in cells infected with the W T and resistant viruses (Table 1). Uninfected cells produce slightly less triphosphate, although not significantly less based upon the variability associated with the data. Cells infected with the resistant virus had more ganciclovir monophosphate than WT-infected and uninfected cells. In cells treated with cidofovir, relatively the same amounts of the phosphate-choline derivative, monophosphate, and diphosphate (the antivirally active metabolite) were produced in uninfected cells and in cells infected with WT and resistant MCMVs. In a follow-up experiment using the same study design as in Table 1, ganciclovir was tested at 3, 10 and 30 #M; cidofovir was evaluated at 0.3, 1 and 3/~M. At 24 h the extent of phosphorylation of ganciclovir to its triphosphate form and cidofovir to cidofovir diphosphate was linearly doseresponsive under all conditions tested (data not shown). There was no apparent difference in levels of ganciclovir triphosphate or cidofovir diphosphate produced in WT versus resistant virus-infected cells at these extracelluar concentrations.

K.M. Okleberry et al./Antiviral Research 35 (1997) 83-90

3.3. Rates of drug metabolism and catabolism The extent of ganciclovir triphosphate and cidofovir diphosphate formation was studied at multiple times over a 24 h period. Catabolism o f these nucleotides following drug removal was followed for an additional 24 h in the same experiment. The purpose of these studies was to determine if differences exist in metabolic and catabolic rates in cells infected with W T and resistant viruses. As depicted in Fig. 1, ganciclovir triphosphate forma-

0.16a

A

0.14;

o.12: ~= 0.10o.o6-

o od

"

0.040.02 0.0:

' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 , 1 , 1 , 1

87

tion was approximately linear in infected cells through 24 h, whereas uninfected cells phosphorylated ganciclovir through 12 h. The rate of ganciclovir triphosphate formation during the linear part of the curves was estimated at 0.006 pmol/h. This rate was the same for infected (WT and resistant virus) and uninfected cells. The formation of cidofovir diphosphate was nearly linear between 6 and 12 h, with an approximate rate of formation of 0.002 pmol/h. This rate was similar for infected (WT and resistant virus) and uninfected cells. The lag time between 0 and 6 h possibly reflects the time required for drug transport into the cell, since compounds such as cidofovir that contain a phosphate group do not penetrate cells readily. Dephosphorylation of ganciclovir triphosphate was obvious after 24 h following removal of the drug (Fig. 1). The rates of dephosphorylation were approximately the same in uninfected cells, and in cells infected with W T and resistant viruses. The half-life of ganciclovir triphosphate was estimated to be approximately 10 h. The rate of dephosphorylation of cidofovir diphosphate was also nearly the same in uninfected and W T and resistant virus-infected cells, with the half-life being about 24 h.

0.035 0.030-

B

3.4. Effect of multiplicity of virus infection (MOI) on drug metabolism

~) 0.0250.020.~ 0.015~" 0.010o. 0.0050.0:

0 '

'1

'1's'2b'i4'2

'd2'3

'4'0'4

'4

Hours After Infection Fig. 1. Metabolism of GCV to GCV triphosphate (GCVppp) (a) and cidofovir (HPMPC) to HPMPC diphosphate (HPMPCpp) (b) from 0 to 24 h, and catabolism of these nucleotides from 24 to 48 h following drug removal. Cells were incubated with 10 #M GCV or 1 #M HPMPC for periods up to 24 h followed by removal of drug from the culture medium. The multiplicity of virus infection was one virus per cell. Symbols: wild-type virus-infected cells (e); resistant virus-infected cells ( I ) ; and uninfected cells (&).

Presumably, if drug phosphorylation depended upon a virus-encoded enzyme, then increasing the multiplicity of virus infection would increase intracellular nucleotide formation. This effect should be more pronounced in W T virus-infected cells than in cells resistant to the virus. To investigate this, the metabolism of ganciclovir and cidofovir was studied in cells infected with MOIs of 0.1-3 viruses per cell (Fig. 2). Ganciclovir triphosphate formation increased slightly, particularly at 3 MOI. A trend toward a gradual increase in cidofovir diphosphate formation occurred as MOI increased. In either case the effect of increasing MOI on extent of phosphorylation was not dramatic, and there was very little difference between phosphorylation in cells infected with W T versus resistant virus.

88

K.M. Okleberry et al./Antiviral Research 35 (1997) 83 90

m 0.25-

4. Discussion

8

~. 0.20Q.

oft.

G.

0.15.

0.10. ¢1. -

o

i

0.05-

~. 0.0 E 0

011

0~3

1;0

3Lo

Multiplicity of Infection Fig. 2. Effect of multiplicity of virus infection on phosphorylation of GCV to GCVppp and HPMPC to HPMPC diphosphate (HPMPCpp). Infected cells were incubated with 10/tM GCV or 1 /LM HPMPC for 24 h. Symbols: solid symbols-nucleotides from cells infected with wild-type virus; open symbols--nucleotides from cells infected with drug-resistant virus; GCV metabolism (OC)); HPMPC metabolism (li D).

In these experiments the phosphorylation of ganciclovir and cidofovir were studied in uninfected, WT virus-infected and resistant MCMVinfected cells in order to determine if virus infection leads to a dramatic enhancement of phosphorylation and if drug resistance is caused by the lack of stimulation of phosphorylation, as is most often the case in ganciclovir resistance by HCMV (Biron, 1991). Until this report, no drug phosphorylation studies of ganciclovir and cidofovir in MCMV-infected cells have been published.

0,o.5 t " o't _.

~

A

o

/I" /

0"4 t 0.3-t

b5

,/1'

~

[-4

3

o . o ~

3.5. Correlation of virus production to drug phosphoryIation over 96 h The results up the this point have suggested that the virus infection moderately stimulates drug phosphorylation. To investigate this further, an experiment was designed in which phosphorylation and virus replication were simultaneously monitored daily for 96 h (Fig. 3). The replication of MCMV and the production of ganciclovir triphosphate were greater in cells infected with the resistant virus, particularly at 72 and 96 h. In uninfected cells, the level of ganciclovir triphosphate remained constant throughout the entire 96 h period, indicating drug metabolism was at a steady state. The results with cidofovir metabolism and virus production in cidofovirtreated cells were similar. Virus production and cidofovir diphosphate formation were greater in cells infected with the resistant virus. Phosphorylation in uninfected cells was at a steady state for the entire 96 h period.

=

0.25-] 4/

~.

¢

~ I-4 l _ . . l l -s.5 ~

B

.;o.'oi

I O/'' t a.

~

-=o'ot

o.o¢~----~

0

~-o

/./I //

II

~

4 4 24 48 72 Hours After Infection

-= = l''' L2

,,

,0

96

Fig. 3. Correlation of extent of metabolism of GCV to GCVppp (a) and HPMPC to HPMPCpp (b) with virus yield. Cells were incubated with 10/~M GCV or 1 /~M HPMPC for the times indicated• Virus titers were determined from separate samples run in parallel with those used for HPLC analysis. The multiplicity of virus infection was one virus per cell. Symbols: solid symbols amounts of nucleotides formed; open symbols--virus titers; wild-type virus-infected cells (gO); resistant virus-infected cells ( l i D ) ; uinfected cells

(AA).

K.M. Okleberry et al. /Antiviral Research 35 (1997) 83-90

The metabolism studies showed that there was no significant difference in the amount of active forms of the drugs (ganciclovir triphosphate and cidofovir diphosphate) produced between 2 and 24 h in cells infected with WT and resistant MCMVs. Throughout all of the experiments the uninfected cells normally had lower levels of drug nucleotides. The rate of nucleotide formation in infected cells was nearly the same as in uninfected cells. There appears to be no metabolic mechanism during the first 24 h of infection that would account for viral resistance to the drugs. Cells infected with the resistant virus had reproducibly higher levels of ganciclovir monophosphate at 24 h than did WT virus-infected cells. At 96 h there was more ganciclovir triphosphate or cidofovir diphosphate produced in the resistant virus-infected cells, which correlated with greater amounts of virus being produced. These differences in the extent of drug phosphorylation between WT and resistant virus-infected cells may relate to the effect of MCMV to stimulate host cell kinase activities (Freitas et al., 1985; Stinski, 1990). This stimulation presumably was greater in cells infected with the resistant virus, since that virus was replicating to a much greater extent than the WT virus at the drug concentrations tested. A similar observation was made regarding the dephosphorylation data. The rates of dephosphorylation of ganciclovir triphosphate in infected and uninfected cells were similar, as were the rates of cidofovir diphosphate catabolism. At the end of the dephosphorylation period, the amounts of remaining nucleotide inside the cells infected with either WT virus or resistant virus were similar. Thus, drug resistance cannot be explained by an increased ability of cells infected with the resistant virus to catabolize nucleotides. Changing the parameter of multiplicity of infection (MOI) did not appear to discriminate between cells infected with WT versus resistant virus. If the WT virus encodecl a unique drug phosphorylating enzyme and the resistant virus encoded a defective enzyme, major differences in

89

phosphorylation should have been observed as MOI increased. The results showing a slight enhancement in drug phosphorylation with increasing MOI, and at 96 h in cells infected with resistant virus, are consistent with the known observation that host cell enzyme activities are stimulated by cytomegalovirus infection (Freitas et al., 1985; Stinski, 1990). We conclude that this is the reason for stimulated drug metabolism in MCMV-infected cells, and explains why uninfected cells produced the least amount of drug metabolites, particularly compared to cells infected with the resistant virus for 72 and 96 h. HCMV resistance to ganciclovir may result from decreased phosphorylation of the parent compound in infected cells (Biron, 1991), and/or reduced ability of the compound to inhibit the viral DNA polymerase (Sullivan and Coen, 1991; Tatarowicz et al., 1992; Sullivan et al., 1993). Drug phosphorylation was not reduced in cells infected with resistant MCMV relative to WT virus, indicating that the first mode of HCMV drug resistance does not apply to MCMV. The resistant virus used in these studies is also cross-resistant to foscarnet (Smee et al., 1995), a compound that is not metabolized (nor is metabolism required for antiviral activity) and inhibits viral DNA polymerases (Oberg, 1989). Burns and colleagues (Burns et al., 1982) developed MCMV strains resistant to acyclovir that were also cross-resistant to phosphonoacetic acid (closely related to foscarnet). They were able to establish that mutations in the viral DNA polymerase were responsible for drug resistance (Sandford et al., 1985). Since the viral DNA polymerase is the antiviral target for these compounds in their biologically active forms, the ganciclovir/cidofovir-resistant MCMV is most likely a DNA polymerase mutant.

Acknowledgements This work was supported by grant number 1 R21 AI39378 from the National Institute of Allergy and Infectious Diseases, NIH.

90

K.M. Okleberry et al./Antiviral Research 35 (1997) 83 90

References Alford, C.A. and Britt, W.J. 0990) Cytomegaloviruses. In: B.N. Fields, D.M. Knipe, M.S. Chanock, M.S. Hirsch, J.L. Melnick, T.P. Monath and B. Roizman (Eds.), Virology (2nd edition), pp. 1981-2010, Raven Press, New York. Baldanti, F., Underwood, M.R., Stanat, S.C., Biron, K.K., Chou, S., Sarasini, A., Silini, E. and Gerna, G. (1996) Single amino acid changes in the DNA polymerase confer foscarnet resistance and slow-growth phenotype, while mutations in the UL97-encoded phosphotransferase confer ganciclovir resistance in three double-resistant human cytomegalovirus strains recovered from patients with AIDS. J. Virol. 70, 1390-1395. Baldanti, F., Sarasini, A., Silini, E., Barbi, M., Lazzarin, A., Biron, K.K. and Gerna, G. (1995) Four dually resistant human cytomegalovirus strains from AIDS patients: single mutations in UL97 and UL54 open reading frames are responsible for ganciclovir- and foscarnet-specific resistance, respectively. Scand. J. Infect. Dis. (Suppl. 99), 103 104. Biron, K.K. (1991) Ganciclovir-resistant human cytomegalovirus clinical isolates; resistance mechanisms and in vitro susceptibility to antiviral agents. Transplant. Proc. 23, 162-167. Burns, W.H., Wingard, J.R., Sandford, G.R., Bender, W.J. and Saral, R. (1982) Acyclovir in mouse cytomegalovirus infections. Am. J. Med. 73 (Suppl. 1A), 118-124. Freitas, V.R., Smee, D.F., Chernow, M., Boehme, R. and Matthews, T.R. (1985) Activity of 9-(l,3-dihydroxy-2-propoxymethyl)guanine compared with that of acyclovir against human, monkey and rodent cytomegaloviruses. Antimicrob. Agents Chemother. 28, 240 245. Ho, H.-T., Woods, K.L., Bronson, J.J., De Boeck, H., Martin, J.C. and Hitchcock, M.J.M. (1992) Intracellular metabolism of the antiherpes agent (S)-l-[3-hydroxy-2(phosphonylmethoxy)propyl]cytosine. Mol. Pharmacol. 41, 197-202. Kern, E.R. (1991) Value of animal models to evaluate agents with potential activity against human cytomegalovirus. Transplant. Proc. 23 (Suppl. 3), 152-155. Laskin, O.L., Cederberg, D.M., Mills, J., Eron, L.J., Mildvan, D. and Spector, S.S. (1987) Ganciclovir for the treatment and suppression of serious infections caused by cytomegaloviruses. Am. J. Med. 83, 201 207. Littler, E., Stuart, A.D. and Chee, M.S. (1992) Human cytomegalovirus UL97 open reading frame encodes a protein that phosphorylates the antiviral nucleoside analog ganciclovir. Nature (London) 358, 160 162.

Oberg, B. (1989) Antiviral effects of phosphonoformate (PFA, Foscarnet sodium). Pharmaceut. Ther. 40, 213-285. Polis, M.A., Spooner, K.M., Baird, B.F., Manischewitz, J.F., Jaffe, H.S., Fisher, P.E., Falloon, J., Davey, R.T., Kovacs, J.A., Walker, R.E., Whitcup, S.M., Nussenblatt, R.B., Lane, H.C. and Masur, H. (1995) Anticytomegaloviral activity and safety of cidofovir in patients with human immunodeficiency virus infection and cytomegalovirus viruria. Antimicrob. Agents Chemother. 39, 882-886. Sandford, G.R., Wingard, J.R., Simons, J.W., Staal, S.P., Saral, R. and Burns, W.H. (1985) Genetic analysis of the susceptibility of mouse cytomegalovirus to acyclovir. J. Virol. 54, 104-113. Smee, D.F., Barnett, B.B., Sidwell, R.W., Reist, E.J. and Holy, A. (1995) Antiviral activities of nucleosides and nucleotides against wild-type and drug-resistant strains of murine cytomegalovirus. Antiviral Res. 26, 1 9. Smee, D.F., Boehme, R., Chernow, M., Binko, B.P. and Matthews, T.R. (1985) Intracellular metabolism and enzymatic phosphorylation of 9-(1,3-dihydroxy-2-propoxymethyl)guanine and acyclovir in herpes simplex virus-infected and uninfected cells. Biochem. Pharmacol. 34, 1049 1056. Stinski, M.F. (1990) Cytomegalovirus and its replication. In: B.N. Fields, D.M. Knipe, M.S. Chanock, M.S. Hirsch, J.L. Melnick, T.P. Monath and B. Roizman (Eds.), Virology (2nd edition), pp. 1959 1980, Raven Press, New York. Sullivan, V. and Coen, D.M. (1991) Isolation of foscarnet-resistant human cytomegalovirus patterns of resistance and sensitivity to other antiviral drugs. J. Infect. Dis. 164, 781 784. Sullivan, V., Biron, K.K., Talarico, C., Stanat, S.C., Davis, M., Pozzi, L.M. and Coen, D.M. (1993) A point mutation in the human cytomegalovirus DNA polymerase gene confers resistance to ganciclovir and phosphonylmethoxyalkyl derivatives. Antimicrob. Agents Chemother. 37, 19-25. Sullivan, V., Talarico, C.L., Stanat, S.C., Davis, M., Coen, D.M. and Biron, K.K. (1992) A protein kinase homologue controls phosphorylation of ganciclovir in human cytomegalovirus-infected cells. Nature (London) 358, 162 164. Tatarowicz, W.A., Lurain, N.S. and Thompson, K.D. (1992) A ganciclovir-resistant clinical isolate of human cytomegalovirus exhibiting cross-resistance to other DNA polymerase inhibitors. J. Infect. Dis. 166, 904 907. Walmsley, S.L., Chew, E., Fanning, M.M., Read, S.E., Vellend, H., Salit, I. and Rachlis, A. (1988) Treatment of cytomegalovirus retinitis with trisodium phosphonoformate hexahydrate (foscarnet). J. Infect. Dis. 157, 569-572.