Metabolism of Cyclopropavir and Ganciclovir in Human Cytomegalovirus-Infected Cells Brian G. Gentry,a John C. Drachb Department of Pharmaceutical, Biomedical and Administrative Sciences, College of Pharmacy and Health Sciences, Drake University, Des Moines, Iowa, USAa; Department of Biologic and Materials Sciences, School of Dentistry, University of Michigan, Ann Arbor, Michigan, USAb
H
uman cytomegalovirus (HCMV) is a widespread pathogen infecting between 40% and 80% of the population worldwide (1). Although individuals with competent immune systems rarely manifest any symptoms, HCMV infections can result in severe interstitial pneumonia, encephalitis, and gastroenteritis in immunocompromised individuals (2). HCMV is also the most common congenital infection in the United States and results in over 4,000 cases of severe mental disabilities, hearing, and/or vision loss in infants each year (3, 4). Drugs currently approved by the FDA for the treatment of systemic HCMV infections are ganciclovir (GCV; Fig. 1) and its oral prodrug valganciclovir, foscarnet (PFA), and cidofovir (5–8). The mechanism of action for each of these drugs involves inhibition of the viral DNA polymerase, resulting in inhibition of HCMV DNA synthesis and viral replication (5). However, long-term chemotherapy for HCMV is generally required due to recurrence of infection upon cessation of treatment. As such, the selection of strains with decreased drug susceptibility is common (6, 9–11). Because adverse effects occur with a high rate of incidence (up to 30% of patients) (12), there is a need for new compounds with a greater therapeutic index for the treatment of systemic HCMV infection. We have previously demonstrated that cyclopropavir (CPV; Fig. 1), a methylenecyclopropane guanosine nucleoside analog, is approximately 10-fold more active in vitro (50% effective concentration [EC50] ⫽ 0.46 M) than GCV (EC50 ⫽ 4.1 M) without any observed increase in cytotoxicity (13). In addition, CPV is also active against several HCMV strains that are resistant to GCV or PFA (14). Further experimentation in vivo with CPV demonstrated a 2 to 5 log10 reduction in titers of murine cytomegalovirus, resulting in reduced mortality in severe combined immunodeficient (SCID) mice and reduced viral replication in human fetal tissue implanted in SCID mice infected with HCMV (15). Toxi-
April 2014 Volume 58 Number 4
cology studies performed in vivo demonstrated few to no adverse effects at therapeutic concentrations, making CPV a good clinical candidate for the treatment of systemic HCMV infections (16). We and others have previously established that the mechanism of action of CPV is similar to that of GCV, namely, phosphorylation to a monophosphate (MP) by the viral pUL97 protein kinase (17–19), additional phosphorylation to a triphosphate (TP) by an endogenous cellular kinase (20), and viral DNA synthesis inhibition resulting in inhibition of viral replication (14, 21). Although enzymatic conversion of CPV to a triphosphate by the pUL97 viral protein kinase and an endogenous cellular kinase has been established (17, 20), this conversion has not been observed in virusinfected cells. Therefore, the goal of this study was a comparison of the metabolism of CPV and the current standard for HCMV chemotherapy, GCV, in HCMV-infected cells. MATERIALS AND METHODS Viral strain and chemicals. HCMV strain Towne was kindly provided by M. F. Stinski, University of Iowa. GCV was kindly provided by Hoffman La Roche (Palo Alto, CA). Cyclopropavir [(Z)-9-{[2,2-bis-(hydroxymethyl)cyclopropylidene]-methyl}guanine; CPV], along with its monophosphates, diphosphates (DP), and triphosphates (TP), was kindly provided by Jiri Zemlicka (Karmanos Cancer Center, Wayne State University, Detroit, MI) (13, 22). 8-[3H]GCV (19 Ci/mmol) and 8-[3H]CPV (0.3
Received 23 October 2013 Returned for modification 2 December 2013 Accepted 31 January 2014 Published ahead of print 10 February 2014 Address correspondence to Brian G. Gentry, [email protected]. Copyright © 2014, American Society for Microbiology. All Rights Reserved. doi:10.1128/AAC.02311-13
Antimicrobial Agents and Chemotherapy
p. 2329 –2333
aac.asm.org
2329
Downloaded from http://aac.asm.org/ on March 8, 2015 by ST ANDREWS UNIV
Human cytomegalovirus (HCMV) is a widespread pathogen that can cause severe disease in immunologically immature and immunocompromised patients. The current standard of therapy for the treatment of HCMV infections is ganciclovir (GCV). However, high incidence rates of adverse effects are prevalent and limit the use of this drug. Cyclopropavir (CPV) is 10-fold more effective against HCMV in vitro than GCV (50% effective concentrations [EC50s] ⴝ 0.46 and 4.1 M, respectively) without any observed increase in cytotoxicity (S. Zhou, J. M. Breitenbach, K. Z. Borysko, J. C. Drach, E. R. Kern, E. Gullen, Y. C. Cheng, and J. Zemlicka, J. Med. Chem. 47:566 –575, 2004, doi:10.1021/jm030316s). We have previously determined that the viral protein kinase pUL97 and endogenous cellular kinases are responsible for the conversion of CPV to a triphosphate (TP), the active compound responsible for inhibiting viral DNA synthesis and viral replication. However, this conversion has not been observed in HCMVinfected cells. To that end, we subjected HCMV-infected cells to equivalently effective concentrations (⬃5 times the EC50) of either CPV or GCV and observed a time-dependent increase in triphosphate levels for both compounds (CPV-TP ⴝ 121 ⴞ 11 pmol/106 cells; GCV-TP ⴝ 43.7 ⴞ 0.4 pmol/106 cells). A longer half-life was observed for GCV-TP (48.2 ⴞ 5.7 h) than for CPV-TP (23.8 ⴞ 5.1 h). The area under the curve for CPV-TP produced from incubation with 2.5 M CPV was 8,680 ⴞ 930 pmol · h/106 cells, approximately 2-fold greater than the area under the curve for GCV-TP of 4,520 ⴞ 420 pmol · h/106 cells produced from incubation with 25 M GCV. We therefore conclude that the exposure of HCMV-infected cells to CPV-TP is greater than that of GCV-TP under these experimental conditions.
Gentry and Drach
FIG 1 Structures of cyclopropavir (CPV) and ganciclovir (GCV).
2330
aac.asm.org
FIG 2 Biosynthesis of CPV-TP and GCV-TP in HCMV-infected HFF cells.
Uninfected and HCMV-infected HFF cells were incubated with 2.5 M CPV or 25 M GCV (used as a positive control). Samples were taken at designated times following the addition of drug and assayed for CPV-TP or GCV-TP by HPLC. The values represent the mean ⫾ standard deviation from at least two experiments. Biosynthesis of CPV-TP in HCMV-infected cells is significantly different from the biosynthesis of GCV-TP in HCMV-infected HFF cells (*, P ⬍ 0.01) and from the biosynthesis of CPV-TP in uninfected HFF cells (‡, P ⬍ 0.01).
software [version 7.0]; Beckman Coulter, Inc., Indianapolis, IN). Before injection, each sample was centrifuged at 14,000 rpm for 10 min to remove any remaining particulate matter. Samples were loaded onto a 5-m-pore-size Hypersil strong anion exchange column (Thermo Scientific, Waltham, MA) (250 by 4.6 mm) at a flow rate of 1.0 ml/min. Baseline separation of GCV and its phosphorylated derivatives was achieved by elution with 10 mM ammonium phosphate (pH 3.0) and 500 mM ammonium phosphate (pH 3.0) (10 mM ammonium phosphate isocratic conditions for 12 min followed by 25% gradient of 500 mM ammonium phosphate over 24 min). One-minute fractions were collected and analyzed and tritium levels quantified by liquid scintillation spectrometry using a Beckman LS 6500 scintillation counter (Beckman Coulter, Inc., Indianapolis, IN). Concentrations of GCV-TP were calculated as described above for CPV-TP. Data analysis. Upon collection of data and calculation of triphosphate concentrations, results were graphed and analyzed using Prism (version 5.0; GraphPad Software, San Diego, CA) to determine standard deviations, linear regressions, statistical significance (Student’s t test), and areas under the curve.
RESULTS
Metabolism of CPV and GCV to their respective triphosphates in HCMV-infected HFF cells. We have hypothesized that the mechanism of action of CPV involves phosphorylation to a triphosphate, the active compound that inhibits the viral DNA polymerase (14, 17, 20). However, the production of CPV-TP in HCMV-infected cells has not been demonstrated. Therefore, to test for the biosynthesis of CPV-TP in vitro, HFF cells infected with the Towne strain of HCMV (MOI, ⬃5) were incubated with either 2.5 M CPV or 25 M GCV (positive control) and cell extracts were analyzed for the presence of nucleoside analog triphosphates (Fig. 2 and Table 1). HCMV-infected cells incubated with 2.5 M CPV demonstrated a time-dependent increase in CPV-TP levels resulting in a maximum of 121 ⫾ 11 pmol/106 cells at 120 h. HCMV-infected cells incubated with 25 M GCV demonstrated a similar time-dependent increase in triphosphate levels. However, the maximum quantity of GCV-TP (43.7 ⫾ 0.4
Antimicrobial Agents and Chemotherapy
Downloaded from http://aac.asm.org/ on March 8, 2015 by ST ANDREWS UNIV
Ci/mmol) were purchased from Moravek Biochemicals (Brea, CA) and provided through the courtesy of Microbiotix, Inc. (Worchester, MA). Cell culture procedures. Human foreskin fibroblasts (HFF) were grown in minimal essential medium with Earle’s salts and 10% fetal bovine serum. They were grown at 37°C in a humidified atmosphere of 3% to 5% CO2 and 97% to 95% air and were regularly passaged at 1:2 dilutions using conventional procedures with 0.05% trypsin– 0.02% EDTA– HEPES-buffered saline (23). Nucleoside analog triphosphate biosynthesis. HFF cells were seeded at 250,000 cells per well in a 6-well cluster dish and infected the following day at a multiplicity of infection (MOI) of ⬃5 PFU per cell. At 2 h postinfection, 2.5 M [3H]CPV or 25 M GCV (both [3H]GCV [⬃2%] and unlabeled GCV) was added to the cells. At designated times, cells were removed from plates using trypsin (0.05% with 0.02% EDTA), counted, and lysed using water, and proteins were precipitated using perchloric acid (final concentration, 0.4 N). Samples were neutralized using 10 N potassium hydroxide and stored at ⫺20°C until analysis by high-pressure liquid chromatography (HPLC). Nucleoside analog triphosphate half-lives. HFF cells were seeded at 250,000 cells per well in a 6-well cluster dish and infected the following day at a MOI of ⬃5 PFU per cell. At 2 h postinfection, 2.5 M [3H]CPV or 25 M GCV (both [3H]GCV [⬃2%] and unlabeled GCV) was added to the cells. Following 5 days of drug incubation, media containing drug were removed and replaced with fresh media without drug. Samples were collected at the time of and at designated times following replacement of media and processed as described above. Reverse-phase HPLC. CPV and its phosphorylated derivatives (CPVMP, CPV-DP, and CPV-TP) were separated by reverse-phase HPLC (System Gold Programmable Solvent Module 125 and System Gold Programmable Detector Module 166 controlled by 32 Karat software [version 7.0]; Beckman Coulter, Inc., Indianapolis, IN). Before injection, each sample was centrifuged at 14,000 rpm for 10 min to remove any remaining particulate matter. Samples were loaded onto a 10-m-pore-size Alphabond C18 reverse-phase column (Alltech, Deerfield, IL) (300 by 3.9 mm) at a flow rate of 1.0 ml/min. Baseline separation of CPV and its phosphorylated derivatives was achieved by elution with 150 mM ammonium phosphate (pH 3.0) and 100% methanol (linear gradient of 20% methanol over 30 min followed by 10 min of isocratic conditions [80% ammonium phosphate and 20% methanol]). One-minute fractions were collected and analyzed and tritium levels quantified by liquid scintillation spectrometry using a Beckman LS 6500 scintillation counter (Beckman Coulter, Inc., Indianapolis, IN). Concentrations of CPV-TP were calculated on the basis of the amount of label in the HPLC effluent fractions corresponding to the known position of CPV-TP plus the specific activity of [3H]CPV. Final concentrations were adjusted for differences in cell confluence and infection. Strong anion exchange HPLC. GCV and its phosphorylated derivatives (GCV-MP, GCV-DP, and GCV-TP) were separated by strong anion exchange HPLC (System Gold Programmable Solvent Module 125 and System Gold Programmable Detector Module 166 controlled by 32 Karat
Metabolism of CPV and GCV in HCMV-Infected Cells
TABLE 1 Comparison of CPV-TP and GCV-TP
Compound
Peak triphosphate level (pmol/106 cells)a
Half-life (t1/2) (h)a
Area under the curve (pmol · h/106 cells)b
CPV-TP GCV-TP
121 ⫾ 11 43.7 ⫾ 0.4
23.8 ⫾ 5.1 48.2 ⫾ 5.7
8,680 ⫾ 930 4,520 ⫾ 420
Values represent mean ⫾ standard deviation from at least two experiments. Values were calculated as a result of combining the data from the triphosphate biosynthesis and half-life studies.
a b
FIG 4 Total exposure of HCMV-infected HFF cells to CPV-TP and GCV-TP. Data used to generate Fig. 1 and 2 were combined to calculate areas under the curve as a means to compare the two active compounds. The area under the curve of CPV-TP (8,680 ⫾ 930 pmol · h/106 cells) is significantly greater than the area under the curve of GCV-TP (4,520 ⫾ 420 pmol · h/106 cells) (P ⬍ 0.05) even though cells were incubated with 10 times less CPV. The values represent the mean ⫾ standard deviation from at least two experiments.
2-fold-longer half-life (48.2 ⫾ 5.7 h) than CPV-TP (23.8 ⫾ 5.1 h). Thus, while the accumulation of GCV-TP was not as large as that of CPV-TP, GCV-TP appears to have persisted longer than CPVTP. In fact, even though the dose of GCV was 10 times greater than that of CPV, the level of GCV-TP (10.2 ⫾ 1.3 pmol/106 cells) at 96 h post-wash out was statistically greater than that of CPV-TP (6.5 ⫾ 1.0 pmol/106 cells) (P ⬍ 0.05), although this difference is less than 2-fold. Total exposure of CPV-TP and GCV-TP to HCMV-infected HFF cells. Since the administration of CPV resulted in greater biosynthesis of triphosphate when compared to GCV but with a shorter half-life, combining the data into a single plot and measuring areas under the curve was used to determine which combination of properties resulted in the greatest exposure of HCMVinfected cells to active compound (Fig. 4 and Table 1). The results demonstrate that, under the conditions used in these experiments, the area under the curve for CPV-TP (8,680 ⫾ 930 pmol · h/106 cells) is approximately 2-fold greater than the area under the curve for GCV-TP (4,520 ⫾ 420 pmol · h/106 cells) even though cells were exposed to 10 times more GCV than CPV. We therefore conclude that the exposure of HCMV-infected cells to CPV-TP is greater than that to GCV-TP under equivalently effective concentration (EC) conditions—approximately five times the EC50 for both drugs (13). DISCUSSION
FIG 3 Depletion of CPV-TP and GCV-TP in HCMV-infected HFF cells. HFF cells infected with HCMV were incubated with 2.5 M CPV or 25 M GCV. After 5 days, media containing drug were removed and replaced with fresh media without drug. Samples were taken at designated times following drug removal and assayed by HPLC for CPV-TP or GCV-TP. Both compounds demonstrated first-order kinetics and half-lives were calculated from linear regression lines. The half-life of CPV-TP in HCMV-infected cells (23.8 ⫾ 5.1 h) is significantly different from the half-life of GCV-TP in HCMV-infected HFF cells (48.2 ⫾ 5.7 h) (P ⬍ 0.05). The values represent the mean ⫾ standard deviation from at least two experiments.
April 2014 Volume 58 Number 4
The formation of CPV-TP has been assumed to be an essential element for CPV to inhibit viral replication and elicit an antiviral effect (14, 17, 18, 20). Since our previous experiments demonstrated that CPV is a better substrate for the pUL97 viral protein kinase than GCV and that the initial phosphorylation to a monophosphate catalyzed by this enzyme is the rate-limiting step in the formation of triphosphate (17), we hypothesize that the production of CPV-TP would be greater than that of GCV-TP under equivalent concentrations. Consistent with this hypothesis, the conversion of CPV to CPV-TP occurred to a greater extent than that of GCV to GCV-TP despite being administered at a lower concentration (2.5 M CPV versus 25 M GCV, equivalently effective concentrations [approximately five times the EC50]). aac.asm.org 2331
Downloaded from http://aac.asm.org/ on March 8, 2015 by ST ANDREWS UNIV
pmol) was 2.5-fold lower and occurred earlier (96 h) than that of CPV-TP even though the concentration of GCV with which HCMV-infected cells were incubated was 10-fold greater than that of CPV. No mono- or diphosphates of CPV or GCV were detected, indicating that the rate-limiting step in the biosynthesis of both triphosphates is the initial phosphorylation step. Uninfected HFF cells incubated with 2.5 M CPV demonstrated no measurable level of CPV-TP (limit of detection, ⬃0.6 pmol/106 cells), indicating that the presence of virus, and, more specifically, of the viral protein kinase pUL97 (17), is necessary for the conversion of CPV to CPV-TP. Half-life of CPV-TP and GCV-TP in HCMV-infected HFF cells. Enzymatic conversion of CPV to CPV-TP, while necessary for the drug to elicit an antiviral effect, is not the sole determinant of efficacy. Drug half-life, or the length of time that the virus is exposed to the active compound, is another major determinant of efficacy. Therefore, we measured the half-life of CPV-TP and GCV-TP in HCMV-infected cells following 5 days of exposure to 2.5 M CPV and 25 M GCV, respectively (Fig. 3 and Table 1). The results for both compounds demonstrated first-order kinetics, and the half-lives were calculated from their respective linear regression lines. Although triphosphate levels for both compounds persisted through 96 h, GCV-TP had an approximately
Gentry and Drach
2332
aac.asm.org
antiviral response (13). We therefore hypothesize that the superior efficacy of CPV when compared to GCV does not come from better efficacy of their respective active compound (CPV-TP versus GCV-TP) but stems from the fact that CPV is a better substrate for the viral protein kinase pUL97 and thus is phosphorylated to a monophosphate (rate-limiting step) to a much greater extent than GCV. This would result in a greater production of CPV-TP than GCV-TP under equivalent concentration conditions. Consistent with this hypothesis, we have previously determined that the HCMV protein kinase pUL97 phosphorylates CPV to CPV-MP at a rate 45 times greater than that of GCV to GCV-MP at equivalent concentrations (17). Our current results demonstrate that at equivalently effective concentrations of CPV and GCV, HCMV-infected cells are exposed to CPV-TP to a greater extent than GCV-TP, the active compounds that elicit an antiviral response. In addition, the greater efficacy of CPV observed in vitro and in vivo without any increase in toxicity (13, 16) and the ability to achieve therapeutic concentrations in vivo without prodrug modification (7, 16) are two reasons why CPV appears to be superior to GCV for the treatment of HCMV disease. ACKNOWLEDGMENTS We thank Andrew Samann and James Simmer for assistance with the use of James Simmer’s HPLC system and liquid scintillation spectrometer system. We especially thank the inventor of CPV, Jiri Zemlicka, for his interest, encouragement, and support. We also thank James Sacco for his help with the half-life data analysis. This work was supported by the gift of tritiated CPV and GCV from Microbiotix, Inc. This work was supported by funds from the University of Michigan and Drake University.
REFERENCES 1. Ho M, 2008. The history of cytomegalovirus and its diseases. Med. Microbiol. Immunol. 197:65–73. http://dx.doi.org/10.1007/s00430-007-0066-x. 2. Landolfo S, Gariglio M, Gribaudo G, Lembo D. 2003. The human cytomegalovirus. Pharmacol. Ther. 98:269 –297. http://dx.doi.org/10 .1016/S0163-7258(03)00034-2. 3. Fortunato EA, Spector DH. 1999. Regulation of human cytomegalovirus gene expression. Adv. Virus Res. 54:61–128. http://dx.doi.org/10.1016 /S0065-3527(08)60366-8. 4. Manicklal S, Emery VC, Lazzarotto T, Boppana SB, Gupta RK. 2013. The “silent” global burden of congenital cytomegalovirus. Clin. Microbiol. Rev. 26:86 –102. http://dx.doi.org/10.1128/CMR.00062-12. 5. Andrei G, De Clercq E, Snoeck R. 2009. Drug targets in cytomegalovirus infection. Infect. Disord. Drug Targets 9:201–222. http://dx.doi.org/10 .2174/187152609787847758. 6. Biron KK. 2006. Antiviral drugs for cytomegalovirus diseases. Antiviral Res. 71:154 –163. http://dx.doi.org/10.1016/j.antiviral.2006.05.002. 7. Cocohoba JM, McNickoll IR. 2002. Valganciclovir: an advance in cytomegalovirus therapeutics. Ann. Pharmacother. 36:1075–1079. http://dx .doi.org/10.1345/aph.1A393. 8. Balfour HH. 1999. Antiviral drugs. N. Engl. J. Med. 340:1255–1268. http: //dx.doi.org/10.1056/NEJM199904223401608. 9. Baldanti F, Lurain N, Gerna G. 2004. Clinical and biologic aspects of human cytomegalovirus resistance to antiviral drugs. Hum. Immunol. 65:403– 409. http://dx.doi.org/10.1016/j.humimm.2004.02.007. 10. Chou S, Lurain NS, Thompson KD, Miner RC, Drew WL. 2003. Viral DNA polymerase mutations associated with drug resistance in human cytomegalovirus. J. Infect. Dis. 188:32–39. http://dx.doi.org/10.1086 /375743. 11. Erice A. 1999. Resistance of human cytomegalovirus to antiviral drugs. Clin. Microbiol. Rev. 12:286 –297. 12. Morris DJ. 1994. Adverse effects and drug interactions of clinical importance with antiviral drugs. Drug Saf. 10:281–291. http://dx.doi.org/10 .2165/00002018-199410040-00002.
Antimicrobial Agents and Chemotherapy
Downloaded from http://aac.asm.org/ on March 8, 2015 by ST ANDREWS UNIV
Our results demonstrated a significant difference between the half-lives of CPV-TP (23.8 ⫾ 5.1 h) and GCV-TP (48.2 ⫾ 5.7 h) (Fig. 3 and Table 1). In these experiments, GCV-TP appeared to reach equilibrium within the HCMV-infected cell (a point at which the rate of dephosphorylation is equal to the rate of phosphorylation) at 48 h post-drug wash out. The half-life of GCV-TP before this state of equilibrium occurred was approximately 27 h and is not statistically different from that observed for CPV-TP. It also appears that this state of nucleoside analog triphosphate equilibrium between the two compounds is common; the levels of CPV-TP toward the end of the half-life experiment (84 h postwash out) also appear to have reached an equilibrium state. We hypothesize that if the duration of the experiment had been longer than 96 h post-wash out, we would have also observed a persistent, stable level of CPV-TP. In contrast to our results, a previous study by Biron et al. reported that the half-life of GCV-TP in HCMV-infected cells was approximately 12 h (24), a significant variance from what we have observed (48.2 h). There is, however, a significant difference between the two studies that can partially account for the variance in results. Biron et al. conducted their experiments using a MOI of 0.5 PFU per cell, which would result in an HCMV infection rate of approximately 50% (24). We conducted these experiments using a MOI of ⬎5 PFU per cell, which would result in nearly 100% of the cells being infected with HCMV. This difference in MOI not only would result in different intracellular concentrations of GCV-TP but also could affect half-life. In fact, Gentry et al. previously demonstrated that a culture in which only 50% of the cells were able to produce GCV-TP resulted in about a 2-fold reduction in half-life compared to a culture in which 100% of the population of cells could produce GCV-TP (25). This reduction in half-life was hypothesized to be the result of two cells dephosphorylating GCV-TP for every cell that could produce GCV-TP in comparison to a culture in which all cells produce and dephosphorylate GCVTP. The presumed mechanism by which this occurs is transfer of phosphorylated GCV metabolites from virus-infected cells to uninfected cells through gap junctions, intracellular communication channels capable of direct transfer of small molecules from the cytoplasm of one cell to that of another (26). In addition, if you remove the latter time points from our experiment in which the level of GCV-TP was stable (48 h postdrug removal), the calculated half-life would be 27 h, or approximately 2-fold greater than what was previously observed. Therefore, this combination of factors could account for the variance between the GCV-TP half-life reported here and that reported by Biron et al. By performing these experiments at equivalently effective concentrations, we are able to speculate about the relative efficacy for each compound acting at the enzymatic and whole-cell levels. Regarding the action of the triphosphates against HCMV DNA polymerase and assuming a cell volume of ⬃5 pl (27), our determination of 121 pmol CPV-TP/106 cells and 43.7 pmol GCV-TP/106 cells would translate to intracellular concentrations of 5.3 M and 1.2 M, respectively. For a whole-cell comparison, the area under the curve for CPV-TP (8,680 ⫾ 930 pmol · h/106 cells) is approximately 2 times greater than the area under the curve for GCV-TP (4,520 ⫾ 420 pmol · h/106 cells). The inference is that it requires two to four times the amount of CPV-TP compared to GCV-TP to elicit the same antiviral effect. However, it has been previously reported that CPV (EC50 ⫽ 0.46 M) is more efficacious than GCV (EC50 ⫽ 4.1 M) since it requires less drug to elicit the same
Metabolism of CPV and GCV in HCMV-Infected Cells
April 2014 Volume 58 Number 4
20. Gentry BG, Gentry SN, Jackson TL, Zemlicka J, Drach JC. 2011. Phosphorylation of antiviral and endogenous nucleotides to di- and triphosphates by guanosine monophosphate kinase. Biochem. Pharmacol. 81:43– 49. http://dx.doi.org/10.1016/j.bcp.2010.09.005. 21. Chou S, Marousek G, Bowlin TL. 2012. Cyclopropavir susceptibility of cytomegalovirus DNA polymerase mutants selected after antiviral drug exposure. Antimicrob. Agents Chemother. 56:197–201. http://dx.doi.org /10.1128/AAC.05559-11. 22. Li C, Gentry BG, Drach JC, Zemlicka J. 2009. Synthesis and enantioselectivity of cyclopropavir phosphates for cellular GMP kinase. Nucleosides Nucleotides Nucleic Acids 28:795– 808. http://dx.doi.org /10.1080/15257770903172720. 23. Shipman C, Jr. 1969. Evaluation of 4-(2-hydroxyethyl)-1piperazineethanesulfonic acid (HEPES) as a tissue culture buffer. Proc. Soc. Exp. Biol. Med. 130:305–310. http://dx.doi.org/10.3181/00379727 -130-33543. 24. Biron KK, Stanat SC, Sorrell JB, Fyfe JA, Keller PM, Lambe CU, Nelson DJ. 1985. Metabolic activation of the nucleoside analog 9-[(2-hydroxy-1(hydroxymethyl)ethoxy]methyl)guanine in human diploid fibroblasts infected with human cytomegalovirus. Proc. Natl. Acad. Sci. U. S. A. 82: 2473–2477. http://dx.doi.org/10.1073/pnas.82.8.2473. 25. Gentry BG, Im M, Boucher PD, Ruch RJ, Shewach DS. 2005. GCV phosphates are transferred between HeLa cells despite lack of bystander cytotoxicity. Gene Ther. 12:1033–1041. http://dx.doi.org/10.1038/sj.gt .3302487. 26. Evans WH, Martin PE. 2002. Gap junctions: structure and function (Review). Mol. Membr. Biol. 19:121–136. http://dx.doi.org/10.1080 /09687680210139839. 27. Mastrocola T, Lambert IH, Kramhøft B, Rugolo M, Hoffmann EK. 1993. Volume regulation in human fibroblasts: role of Ca2⫹ and 5-lipoxygenase products in the activation of the Cl- efflux. J. Membr. Biol. 136:55– 62.
aac.asm.org 2333
Downloaded from http://aac.asm.org/ on March 8, 2015 by ST ANDREWS UNIV
13. Zhou S, Breitenbach JM, Borysko KZ, Drach JC, Kern ER, Gullen E, Cheng YC, Zemlicka J. 2004. Synthesis and antiviral activity of (Z)and (E)-2,2-[bis(hydroxymethyl)cyclopropylidene]methylpurines and -pyrimidines: second-generation methylenecyclopropane analogues of nucleosides. J. Med. Chem. 47:566 –575. http://dx.doi.org/10 .1021/jm030316s. 14. Kern ER, Kushner NL, Hartline CB, Williams-Aziz SL, Harden EA, Zhou S, Zemlicka J, Prichard MN. 2005. In vitro activity and mechanism of action of methylenecyclopropane analogs of nucleosides against herpesvirus replication. Antimicrob. Agents Chemother. 49:1039 –1045. http: //dx.doi.org/10.1128/AAC.49.3.1039-1045.2005. 15. Kern ER, Bidanset DJ, Harline CB, Yan Z, Zemlicka J, Quenelle DC. 2004. Oral activity of a methylenecyclopropane analog, cyclopropavir, in animal models for cytomegalovirus infections. Antimicrob. Agents Chemother. 48:4745– 4753. http://dx.doi.org/10.1128/AAC.48.12.4745-4753 .2004. 16. Bowlin TL, Brooks JL, Zemlicka J. 2009. Preclinical pharmacokinetic, toxicokinetic and toxicology results for cyclopropavir, a promising new agent for the treatment of beta- and gamma-herpesviruses. Antiviral Res. 82:A46 –A47. http://dx.doi.org/10.1016/j.antiviral.2009.02.104. 17. Gentry BG, Kamil JP, Coen DM, Zemlicka J, Drach JC. 2010. Stereoselective phosphorylation of cyclopropavir by pUL97 and competitive inhibition by maribavir. Antimicrob. Agents Chemother. 54:3093–3098. http://dx.doi.org/10.1128/AAC.00468-10. 18. Gentry BG, Vollmer LE, Hall ED, Borysko KZ, Zemlicka J, Kamil JP, Drach JC. 2013. Resistance of human cytomegalovirus to cyclopropavir maps to a base pair deletion in the open reading frame of UL97. Antimicrob. Agents Chemother. 57:4343– 4348. http://dx.doi.org/10.1128/AAC .00214-13. 19. Chou S, Bowlin TL. 2011. Cytomegalovirus UL97 mutations affecting cyclopropavir and ganciclovir susceptibility. Antimicrob. Agents Chemother. 55:382–384. http://dx.doi.org/10.1128/AAC.01259-10.
H
uman cytomegalovirus (HCMV) is a widespread pathogen infecting between 40% and 80% of the population worldwide (1). Although individuals with competent immune systems rarely manifest any symptoms, HCMV infections can result in severe interstitial pneumonia, encephalitis, and gastroenteritis in immunocompromised individuals (2). HCMV is also the most common congenital infection in the United States and results in over 4,000 cases of severe mental disabilities, hearing, and/or vision loss in infants each year (3, 4). Drugs currently approved by the FDA for the treatment of systemic HCMV infections are ganciclovir (GCV; Fig. 1) and its oral prodrug valganciclovir, foscarnet (PFA), and cidofovir (5–8). The mechanism of action for each of these drugs involves inhibition of the viral DNA polymerase, resulting in inhibition of HCMV DNA synthesis and viral replication (5). However, long-term chemotherapy for HCMV is generally required due to recurrence of infection upon cessation of treatment. As such, the selection of strains with decreased drug susceptibility is common (6, 9–11). Because adverse effects occur with a high rate of incidence (up to 30% of patients) (12), there is a need for new compounds with a greater therapeutic index for the treatment of systemic HCMV infection. We have previously demonstrated that cyclopropavir (CPV; Fig. 1), a methylenecyclopropane guanosine nucleoside analog, is approximately 10-fold more active in vitro (50% effective concentration [EC50] ⫽ 0.46 M) than GCV (EC50 ⫽ 4.1 M) without any observed increase in cytotoxicity (13). In addition, CPV is also active against several HCMV strains that are resistant to GCV or PFA (14). Further experimentation in vivo with CPV demonstrated a 2 to 5 log10 reduction in titers of murine cytomegalovirus, resulting in reduced mortality in severe combined immunodeficient (SCID) mice and reduced viral replication in human fetal tissue implanted in SCID mice infected with HCMV (15). Toxi-
April 2014 Volume 58 Number 4
cology studies performed in vivo demonstrated few to no adverse effects at therapeutic concentrations, making CPV a good clinical candidate for the treatment of systemic HCMV infections (16). We and others have previously established that the mechanism of action of CPV is similar to that of GCV, namely, phosphorylation to a monophosphate (MP) by the viral pUL97 protein kinase (17–19), additional phosphorylation to a triphosphate (TP) by an endogenous cellular kinase (20), and viral DNA synthesis inhibition resulting in inhibition of viral replication (14, 21). Although enzymatic conversion of CPV to a triphosphate by the pUL97 viral protein kinase and an endogenous cellular kinase has been established (17, 20), this conversion has not been observed in virusinfected cells. Therefore, the goal of this study was a comparison of the metabolism of CPV and the current standard for HCMV chemotherapy, GCV, in HCMV-infected cells. MATERIALS AND METHODS Viral strain and chemicals. HCMV strain Towne was kindly provided by M. F. Stinski, University of Iowa. GCV was kindly provided by Hoffman La Roche (Palo Alto, CA). Cyclopropavir [(Z)-9-{[2,2-bis-(hydroxymethyl)cyclopropylidene]-methyl}guanine; CPV], along with its monophosphates, diphosphates (DP), and triphosphates (TP), was kindly provided by Jiri Zemlicka (Karmanos Cancer Center, Wayne State University, Detroit, MI) (13, 22). 8-[3H]GCV (19 Ci/mmol) and 8-[3H]CPV (0.3
Received 23 October 2013 Returned for modification 2 December 2013 Accepted 31 January 2014 Published ahead of print 10 February 2014 Address correspondence to Brian G. Gentry, [email protected]. Copyright © 2014, American Society for Microbiology. All Rights Reserved. doi:10.1128/AAC.02311-13
Antimicrobial Agents and Chemotherapy
p. 2329 –2333
aac.asm.org
2329
Downloaded from http://aac.asm.org/ on March 8, 2015 by ST ANDREWS UNIV
Human cytomegalovirus (HCMV) is a widespread pathogen that can cause severe disease in immunologically immature and immunocompromised patients. The current standard of therapy for the treatment of HCMV infections is ganciclovir (GCV). However, high incidence rates of adverse effects are prevalent and limit the use of this drug. Cyclopropavir (CPV) is 10-fold more effective against HCMV in vitro than GCV (50% effective concentrations [EC50s] ⴝ 0.46 and 4.1 M, respectively) without any observed increase in cytotoxicity (S. Zhou, J. M. Breitenbach, K. Z. Borysko, J. C. Drach, E. R. Kern, E. Gullen, Y. C. Cheng, and J. Zemlicka, J. Med. Chem. 47:566 –575, 2004, doi:10.1021/jm030316s). We have previously determined that the viral protein kinase pUL97 and endogenous cellular kinases are responsible for the conversion of CPV to a triphosphate (TP), the active compound responsible for inhibiting viral DNA synthesis and viral replication. However, this conversion has not been observed in HCMVinfected cells. To that end, we subjected HCMV-infected cells to equivalently effective concentrations (⬃5 times the EC50) of either CPV or GCV and observed a time-dependent increase in triphosphate levels for both compounds (CPV-TP ⴝ 121 ⴞ 11 pmol/106 cells; GCV-TP ⴝ 43.7 ⴞ 0.4 pmol/106 cells). A longer half-life was observed for GCV-TP (48.2 ⴞ 5.7 h) than for CPV-TP (23.8 ⴞ 5.1 h). The area under the curve for CPV-TP produced from incubation with 2.5 M CPV was 8,680 ⴞ 930 pmol · h/106 cells, approximately 2-fold greater than the area under the curve for GCV-TP of 4,520 ⴞ 420 pmol · h/106 cells produced from incubation with 25 M GCV. We therefore conclude that the exposure of HCMV-infected cells to CPV-TP is greater than that of GCV-TP under these experimental conditions.
Gentry and Drach
FIG 1 Structures of cyclopropavir (CPV) and ganciclovir (GCV).
2330
aac.asm.org
FIG 2 Biosynthesis of CPV-TP and GCV-TP in HCMV-infected HFF cells.
Uninfected and HCMV-infected HFF cells were incubated with 2.5 M CPV or 25 M GCV (used as a positive control). Samples were taken at designated times following the addition of drug and assayed for CPV-TP or GCV-TP by HPLC. The values represent the mean ⫾ standard deviation from at least two experiments. Biosynthesis of CPV-TP in HCMV-infected cells is significantly different from the biosynthesis of GCV-TP in HCMV-infected HFF cells (*, P ⬍ 0.01) and from the biosynthesis of CPV-TP in uninfected HFF cells (‡, P ⬍ 0.01).
software [version 7.0]; Beckman Coulter, Inc., Indianapolis, IN). Before injection, each sample was centrifuged at 14,000 rpm for 10 min to remove any remaining particulate matter. Samples were loaded onto a 5-m-pore-size Hypersil strong anion exchange column (Thermo Scientific, Waltham, MA) (250 by 4.6 mm) at a flow rate of 1.0 ml/min. Baseline separation of GCV and its phosphorylated derivatives was achieved by elution with 10 mM ammonium phosphate (pH 3.0) and 500 mM ammonium phosphate (pH 3.0) (10 mM ammonium phosphate isocratic conditions for 12 min followed by 25% gradient of 500 mM ammonium phosphate over 24 min). One-minute fractions were collected and analyzed and tritium levels quantified by liquid scintillation spectrometry using a Beckman LS 6500 scintillation counter (Beckman Coulter, Inc., Indianapolis, IN). Concentrations of GCV-TP were calculated as described above for CPV-TP. Data analysis. Upon collection of data and calculation of triphosphate concentrations, results were graphed and analyzed using Prism (version 5.0; GraphPad Software, San Diego, CA) to determine standard deviations, linear regressions, statistical significance (Student’s t test), and areas under the curve.
RESULTS
Metabolism of CPV and GCV to their respective triphosphates in HCMV-infected HFF cells. We have hypothesized that the mechanism of action of CPV involves phosphorylation to a triphosphate, the active compound that inhibits the viral DNA polymerase (14, 17, 20). However, the production of CPV-TP in HCMV-infected cells has not been demonstrated. Therefore, to test for the biosynthesis of CPV-TP in vitro, HFF cells infected with the Towne strain of HCMV (MOI, ⬃5) were incubated with either 2.5 M CPV or 25 M GCV (positive control) and cell extracts were analyzed for the presence of nucleoside analog triphosphates (Fig. 2 and Table 1). HCMV-infected cells incubated with 2.5 M CPV demonstrated a time-dependent increase in CPV-TP levels resulting in a maximum of 121 ⫾ 11 pmol/106 cells at 120 h. HCMV-infected cells incubated with 25 M GCV demonstrated a similar time-dependent increase in triphosphate levels. However, the maximum quantity of GCV-TP (43.7 ⫾ 0.4
Antimicrobial Agents and Chemotherapy
Downloaded from http://aac.asm.org/ on March 8, 2015 by ST ANDREWS UNIV
Ci/mmol) were purchased from Moravek Biochemicals (Brea, CA) and provided through the courtesy of Microbiotix, Inc. (Worchester, MA). Cell culture procedures. Human foreskin fibroblasts (HFF) were grown in minimal essential medium with Earle’s salts and 10% fetal bovine serum. They were grown at 37°C in a humidified atmosphere of 3% to 5% CO2 and 97% to 95% air and were regularly passaged at 1:2 dilutions using conventional procedures with 0.05% trypsin– 0.02% EDTA– HEPES-buffered saline (23). Nucleoside analog triphosphate biosynthesis. HFF cells were seeded at 250,000 cells per well in a 6-well cluster dish and infected the following day at a multiplicity of infection (MOI) of ⬃5 PFU per cell. At 2 h postinfection, 2.5 M [3H]CPV or 25 M GCV (both [3H]GCV [⬃2%] and unlabeled GCV) was added to the cells. At designated times, cells were removed from plates using trypsin (0.05% with 0.02% EDTA), counted, and lysed using water, and proteins were precipitated using perchloric acid (final concentration, 0.4 N). Samples were neutralized using 10 N potassium hydroxide and stored at ⫺20°C until analysis by high-pressure liquid chromatography (HPLC). Nucleoside analog triphosphate half-lives. HFF cells were seeded at 250,000 cells per well in a 6-well cluster dish and infected the following day at a MOI of ⬃5 PFU per cell. At 2 h postinfection, 2.5 M [3H]CPV or 25 M GCV (both [3H]GCV [⬃2%] and unlabeled GCV) was added to the cells. Following 5 days of drug incubation, media containing drug were removed and replaced with fresh media without drug. Samples were collected at the time of and at designated times following replacement of media and processed as described above. Reverse-phase HPLC. CPV and its phosphorylated derivatives (CPVMP, CPV-DP, and CPV-TP) were separated by reverse-phase HPLC (System Gold Programmable Solvent Module 125 and System Gold Programmable Detector Module 166 controlled by 32 Karat software [version 7.0]; Beckman Coulter, Inc., Indianapolis, IN). Before injection, each sample was centrifuged at 14,000 rpm for 10 min to remove any remaining particulate matter. Samples were loaded onto a 10-m-pore-size Alphabond C18 reverse-phase column (Alltech, Deerfield, IL) (300 by 3.9 mm) at a flow rate of 1.0 ml/min. Baseline separation of CPV and its phosphorylated derivatives was achieved by elution with 150 mM ammonium phosphate (pH 3.0) and 100% methanol (linear gradient of 20% methanol over 30 min followed by 10 min of isocratic conditions [80% ammonium phosphate and 20% methanol]). One-minute fractions were collected and analyzed and tritium levels quantified by liquid scintillation spectrometry using a Beckman LS 6500 scintillation counter (Beckman Coulter, Inc., Indianapolis, IN). Concentrations of CPV-TP were calculated on the basis of the amount of label in the HPLC effluent fractions corresponding to the known position of CPV-TP plus the specific activity of [3H]CPV. Final concentrations were adjusted for differences in cell confluence and infection. Strong anion exchange HPLC. GCV and its phosphorylated derivatives (GCV-MP, GCV-DP, and GCV-TP) were separated by strong anion exchange HPLC (System Gold Programmable Solvent Module 125 and System Gold Programmable Detector Module 166 controlled by 32 Karat
Metabolism of CPV and GCV in HCMV-Infected Cells
TABLE 1 Comparison of CPV-TP and GCV-TP
Compound
Peak triphosphate level (pmol/106 cells)a
Half-life (t1/2) (h)a
Area under the curve (pmol · h/106 cells)b
CPV-TP GCV-TP
121 ⫾ 11 43.7 ⫾ 0.4
23.8 ⫾ 5.1 48.2 ⫾ 5.7
8,680 ⫾ 930 4,520 ⫾ 420
Values represent mean ⫾ standard deviation from at least two experiments. Values were calculated as a result of combining the data from the triphosphate biosynthesis and half-life studies.
a b
FIG 4 Total exposure of HCMV-infected HFF cells to CPV-TP and GCV-TP. Data used to generate Fig. 1 and 2 were combined to calculate areas under the curve as a means to compare the two active compounds. The area under the curve of CPV-TP (8,680 ⫾ 930 pmol · h/106 cells) is significantly greater than the area under the curve of GCV-TP (4,520 ⫾ 420 pmol · h/106 cells) (P ⬍ 0.05) even though cells were incubated with 10 times less CPV. The values represent the mean ⫾ standard deviation from at least two experiments.
2-fold-longer half-life (48.2 ⫾ 5.7 h) than CPV-TP (23.8 ⫾ 5.1 h). Thus, while the accumulation of GCV-TP was not as large as that of CPV-TP, GCV-TP appears to have persisted longer than CPVTP. In fact, even though the dose of GCV was 10 times greater than that of CPV, the level of GCV-TP (10.2 ⫾ 1.3 pmol/106 cells) at 96 h post-wash out was statistically greater than that of CPV-TP (6.5 ⫾ 1.0 pmol/106 cells) (P ⬍ 0.05), although this difference is less than 2-fold. Total exposure of CPV-TP and GCV-TP to HCMV-infected HFF cells. Since the administration of CPV resulted in greater biosynthesis of triphosphate when compared to GCV but with a shorter half-life, combining the data into a single plot and measuring areas under the curve was used to determine which combination of properties resulted in the greatest exposure of HCMVinfected cells to active compound (Fig. 4 and Table 1). The results demonstrate that, under the conditions used in these experiments, the area under the curve for CPV-TP (8,680 ⫾ 930 pmol · h/106 cells) is approximately 2-fold greater than the area under the curve for GCV-TP (4,520 ⫾ 420 pmol · h/106 cells) even though cells were exposed to 10 times more GCV than CPV. We therefore conclude that the exposure of HCMV-infected cells to CPV-TP is greater than that to GCV-TP under equivalently effective concentration (EC) conditions—approximately five times the EC50 for both drugs (13). DISCUSSION
FIG 3 Depletion of CPV-TP and GCV-TP in HCMV-infected HFF cells. HFF cells infected with HCMV were incubated with 2.5 M CPV or 25 M GCV. After 5 days, media containing drug were removed and replaced with fresh media without drug. Samples were taken at designated times following drug removal and assayed by HPLC for CPV-TP or GCV-TP. Both compounds demonstrated first-order kinetics and half-lives were calculated from linear regression lines. The half-life of CPV-TP in HCMV-infected cells (23.8 ⫾ 5.1 h) is significantly different from the half-life of GCV-TP in HCMV-infected HFF cells (48.2 ⫾ 5.7 h) (P ⬍ 0.05). The values represent the mean ⫾ standard deviation from at least two experiments.
April 2014 Volume 58 Number 4
The formation of CPV-TP has been assumed to be an essential element for CPV to inhibit viral replication and elicit an antiviral effect (14, 17, 18, 20). Since our previous experiments demonstrated that CPV is a better substrate for the pUL97 viral protein kinase than GCV and that the initial phosphorylation to a monophosphate catalyzed by this enzyme is the rate-limiting step in the formation of triphosphate (17), we hypothesize that the production of CPV-TP would be greater than that of GCV-TP under equivalent concentrations. Consistent with this hypothesis, the conversion of CPV to CPV-TP occurred to a greater extent than that of GCV to GCV-TP despite being administered at a lower concentration (2.5 M CPV versus 25 M GCV, equivalently effective concentrations [approximately five times the EC50]). aac.asm.org 2331
Downloaded from http://aac.asm.org/ on March 8, 2015 by ST ANDREWS UNIV
pmol) was 2.5-fold lower and occurred earlier (96 h) than that of CPV-TP even though the concentration of GCV with which HCMV-infected cells were incubated was 10-fold greater than that of CPV. No mono- or diphosphates of CPV or GCV were detected, indicating that the rate-limiting step in the biosynthesis of both triphosphates is the initial phosphorylation step. Uninfected HFF cells incubated with 2.5 M CPV demonstrated no measurable level of CPV-TP (limit of detection, ⬃0.6 pmol/106 cells), indicating that the presence of virus, and, more specifically, of the viral protein kinase pUL97 (17), is necessary for the conversion of CPV to CPV-TP. Half-life of CPV-TP and GCV-TP in HCMV-infected HFF cells. Enzymatic conversion of CPV to CPV-TP, while necessary for the drug to elicit an antiviral effect, is not the sole determinant of efficacy. Drug half-life, or the length of time that the virus is exposed to the active compound, is another major determinant of efficacy. Therefore, we measured the half-life of CPV-TP and GCV-TP in HCMV-infected cells following 5 days of exposure to 2.5 M CPV and 25 M GCV, respectively (Fig. 3 and Table 1). The results for both compounds demonstrated first-order kinetics, and the half-lives were calculated from their respective linear regression lines. Although triphosphate levels for both compounds persisted through 96 h, GCV-TP had an approximately
Gentry and Drach
2332
aac.asm.org
antiviral response (13). We therefore hypothesize that the superior efficacy of CPV when compared to GCV does not come from better efficacy of their respective active compound (CPV-TP versus GCV-TP) but stems from the fact that CPV is a better substrate for the viral protein kinase pUL97 and thus is phosphorylated to a monophosphate (rate-limiting step) to a much greater extent than GCV. This would result in a greater production of CPV-TP than GCV-TP under equivalent concentration conditions. Consistent with this hypothesis, we have previously determined that the HCMV protein kinase pUL97 phosphorylates CPV to CPV-MP at a rate 45 times greater than that of GCV to GCV-MP at equivalent concentrations (17). Our current results demonstrate that at equivalently effective concentrations of CPV and GCV, HCMV-infected cells are exposed to CPV-TP to a greater extent than GCV-TP, the active compounds that elicit an antiviral response. In addition, the greater efficacy of CPV observed in vitro and in vivo without any increase in toxicity (13, 16) and the ability to achieve therapeutic concentrations in vivo without prodrug modification (7, 16) are two reasons why CPV appears to be superior to GCV for the treatment of HCMV disease. ACKNOWLEDGMENTS We thank Andrew Samann and James Simmer for assistance with the use of James Simmer’s HPLC system and liquid scintillation spectrometer system. We especially thank the inventor of CPV, Jiri Zemlicka, for his interest, encouragement, and support. We also thank James Sacco for his help with the half-life data analysis. This work was supported by the gift of tritiated CPV and GCV from Microbiotix, Inc. This work was supported by funds from the University of Michigan and Drake University.
REFERENCES 1. Ho M, 2008. The history of cytomegalovirus and its diseases. Med. Microbiol. Immunol. 197:65–73. http://dx.doi.org/10.1007/s00430-007-0066-x. 2. Landolfo S, Gariglio M, Gribaudo G, Lembo D. 2003. The human cytomegalovirus. Pharmacol. Ther. 98:269 –297. http://dx.doi.org/10 .1016/S0163-7258(03)00034-2. 3. Fortunato EA, Spector DH. 1999. Regulation of human cytomegalovirus gene expression. Adv. Virus Res. 54:61–128. http://dx.doi.org/10.1016 /S0065-3527(08)60366-8. 4. Manicklal S, Emery VC, Lazzarotto T, Boppana SB, Gupta RK. 2013. The “silent” global burden of congenital cytomegalovirus. Clin. Microbiol. Rev. 26:86 –102. http://dx.doi.org/10.1128/CMR.00062-12. 5. Andrei G, De Clercq E, Snoeck R. 2009. Drug targets in cytomegalovirus infection. Infect. Disord. Drug Targets 9:201–222. http://dx.doi.org/10 .2174/187152609787847758. 6. Biron KK. 2006. Antiviral drugs for cytomegalovirus diseases. Antiviral Res. 71:154 –163. http://dx.doi.org/10.1016/j.antiviral.2006.05.002. 7. Cocohoba JM, McNickoll IR. 2002. Valganciclovir: an advance in cytomegalovirus therapeutics. Ann. Pharmacother. 36:1075–1079. http://dx .doi.org/10.1345/aph.1A393. 8. Balfour HH. 1999. Antiviral drugs. N. Engl. J. Med. 340:1255–1268. http: //dx.doi.org/10.1056/NEJM199904223401608. 9. Baldanti F, Lurain N, Gerna G. 2004. Clinical and biologic aspects of human cytomegalovirus resistance to antiviral drugs. Hum. Immunol. 65:403– 409. http://dx.doi.org/10.1016/j.humimm.2004.02.007. 10. Chou S, Lurain NS, Thompson KD, Miner RC, Drew WL. 2003. Viral DNA polymerase mutations associated with drug resistance in human cytomegalovirus. J. Infect. Dis. 188:32–39. http://dx.doi.org/10.1086 /375743. 11. Erice A. 1999. Resistance of human cytomegalovirus to antiviral drugs. Clin. Microbiol. Rev. 12:286 –297. 12. Morris DJ. 1994. Adverse effects and drug interactions of clinical importance with antiviral drugs. Drug Saf. 10:281–291. http://dx.doi.org/10 .2165/00002018-199410040-00002.
Antimicrobial Agents and Chemotherapy
Downloaded from http://aac.asm.org/ on March 8, 2015 by ST ANDREWS UNIV
Our results demonstrated a significant difference between the half-lives of CPV-TP (23.8 ⫾ 5.1 h) and GCV-TP (48.2 ⫾ 5.7 h) (Fig. 3 and Table 1). In these experiments, GCV-TP appeared to reach equilibrium within the HCMV-infected cell (a point at which the rate of dephosphorylation is equal to the rate of phosphorylation) at 48 h post-drug wash out. The half-life of GCV-TP before this state of equilibrium occurred was approximately 27 h and is not statistically different from that observed for CPV-TP. It also appears that this state of nucleoside analog triphosphate equilibrium between the two compounds is common; the levels of CPV-TP toward the end of the half-life experiment (84 h postwash out) also appear to have reached an equilibrium state. We hypothesize that if the duration of the experiment had been longer than 96 h post-wash out, we would have also observed a persistent, stable level of CPV-TP. In contrast to our results, a previous study by Biron et al. reported that the half-life of GCV-TP in HCMV-infected cells was approximately 12 h (24), a significant variance from what we have observed (48.2 h). There is, however, a significant difference between the two studies that can partially account for the variance in results. Biron et al. conducted their experiments using a MOI of 0.5 PFU per cell, which would result in an HCMV infection rate of approximately 50% (24). We conducted these experiments using a MOI of ⬎5 PFU per cell, which would result in nearly 100% of the cells being infected with HCMV. This difference in MOI not only would result in different intracellular concentrations of GCV-TP but also could affect half-life. In fact, Gentry et al. previously demonstrated that a culture in which only 50% of the cells were able to produce GCV-TP resulted in about a 2-fold reduction in half-life compared to a culture in which 100% of the population of cells could produce GCV-TP (25). This reduction in half-life was hypothesized to be the result of two cells dephosphorylating GCV-TP for every cell that could produce GCV-TP in comparison to a culture in which all cells produce and dephosphorylate GCVTP. The presumed mechanism by which this occurs is transfer of phosphorylated GCV metabolites from virus-infected cells to uninfected cells through gap junctions, intracellular communication channels capable of direct transfer of small molecules from the cytoplasm of one cell to that of another (26). In addition, if you remove the latter time points from our experiment in which the level of GCV-TP was stable (48 h postdrug removal), the calculated half-life would be 27 h, or approximately 2-fold greater than what was previously observed. Therefore, this combination of factors could account for the variance between the GCV-TP half-life reported here and that reported by Biron et al. By performing these experiments at equivalently effective concentrations, we are able to speculate about the relative efficacy for each compound acting at the enzymatic and whole-cell levels. Regarding the action of the triphosphates against HCMV DNA polymerase and assuming a cell volume of ⬃5 pl (27), our determination of 121 pmol CPV-TP/106 cells and 43.7 pmol GCV-TP/106 cells would translate to intracellular concentrations of 5.3 M and 1.2 M, respectively. For a whole-cell comparison, the area under the curve for CPV-TP (8,680 ⫾ 930 pmol · h/106 cells) is approximately 2 times greater than the area under the curve for GCV-TP (4,520 ⫾ 420 pmol · h/106 cells). The inference is that it requires two to four times the amount of CPV-TP compared to GCV-TP to elicit the same antiviral effect. However, it has been previously reported that CPV (EC50 ⫽ 0.46 M) is more efficacious than GCV (EC50 ⫽ 4.1 M) since it requires less drug to elicit the same
Metabolism of CPV and GCV in HCMV-Infected Cells
April 2014 Volume 58 Number 4
20. Gentry BG, Gentry SN, Jackson TL, Zemlicka J, Drach JC. 2011. Phosphorylation of antiviral and endogenous nucleotides to di- and triphosphates by guanosine monophosphate kinase. Biochem. Pharmacol. 81:43– 49. http://dx.doi.org/10.1016/j.bcp.2010.09.005. 21. Chou S, Marousek G, Bowlin TL. 2012. Cyclopropavir susceptibility of cytomegalovirus DNA polymerase mutants selected after antiviral drug exposure. Antimicrob. Agents Chemother. 56:197–201. http://dx.doi.org /10.1128/AAC.05559-11. 22. Li C, Gentry BG, Drach JC, Zemlicka J. 2009. Synthesis and enantioselectivity of cyclopropavir phosphates for cellular GMP kinase. Nucleosides Nucleotides Nucleic Acids 28:795– 808. http://dx.doi.org /10.1080/15257770903172720. 23. Shipman C, Jr. 1969. Evaluation of 4-(2-hydroxyethyl)-1piperazineethanesulfonic acid (HEPES) as a tissue culture buffer. Proc. Soc. Exp. Biol. Med. 130:305–310. http://dx.doi.org/10.3181/00379727 -130-33543. 24. Biron KK, Stanat SC, Sorrell JB, Fyfe JA, Keller PM, Lambe CU, Nelson DJ. 1985. Metabolic activation of the nucleoside analog 9-[(2-hydroxy-1(hydroxymethyl)ethoxy]methyl)guanine in human diploid fibroblasts infected with human cytomegalovirus. Proc. Natl. Acad. Sci. U. S. A. 82: 2473–2477. http://dx.doi.org/10.1073/pnas.82.8.2473. 25. Gentry BG, Im M, Boucher PD, Ruch RJ, Shewach DS. 2005. GCV phosphates are transferred between HeLa cells despite lack of bystander cytotoxicity. Gene Ther. 12:1033–1041. http://dx.doi.org/10.1038/sj.gt .3302487. 26. Evans WH, Martin PE. 2002. Gap junctions: structure and function (Review). Mol. Membr. Biol. 19:121–136. http://dx.doi.org/10.1080 /09687680210139839. 27. Mastrocola T, Lambert IH, Kramhøft B, Rugolo M, Hoffmann EK. 1993. Volume regulation in human fibroblasts: role of Ca2⫹ and 5-lipoxygenase products in the activation of the Cl- efflux. J. Membr. Biol. 136:55– 62.
aac.asm.org 2333
Downloaded from http://aac.asm.org/ on March 8, 2015 by ST ANDREWS UNIV
13. Zhou S, Breitenbach JM, Borysko KZ, Drach JC, Kern ER, Gullen E, Cheng YC, Zemlicka J. 2004. Synthesis and antiviral activity of (Z)and (E)-2,2-[bis(hydroxymethyl)cyclopropylidene]methylpurines and -pyrimidines: second-generation methylenecyclopropane analogues of nucleosides. J. Med. Chem. 47:566 –575. http://dx.doi.org/10 .1021/jm030316s. 14. Kern ER, Kushner NL, Hartline CB, Williams-Aziz SL, Harden EA, Zhou S, Zemlicka J, Prichard MN. 2005. In vitro activity and mechanism of action of methylenecyclopropane analogs of nucleosides against herpesvirus replication. Antimicrob. Agents Chemother. 49:1039 –1045. http: //dx.doi.org/10.1128/AAC.49.3.1039-1045.2005. 15. Kern ER, Bidanset DJ, Harline CB, Yan Z, Zemlicka J, Quenelle DC. 2004. Oral activity of a methylenecyclopropane analog, cyclopropavir, in animal models for cytomegalovirus infections. Antimicrob. Agents Chemother. 48:4745– 4753. http://dx.doi.org/10.1128/AAC.48.12.4745-4753 .2004. 16. Bowlin TL, Brooks JL, Zemlicka J. 2009. Preclinical pharmacokinetic, toxicokinetic and toxicology results for cyclopropavir, a promising new agent for the treatment of beta- and gamma-herpesviruses. Antiviral Res. 82:A46 –A47. http://dx.doi.org/10.1016/j.antiviral.2009.02.104. 17. Gentry BG, Kamil JP, Coen DM, Zemlicka J, Drach JC. 2010. Stereoselective phosphorylation of cyclopropavir by pUL97 and competitive inhibition by maribavir. Antimicrob. Agents Chemother. 54:3093–3098. http://dx.doi.org/10.1128/AAC.00468-10. 18. Gentry BG, Vollmer LE, Hall ED, Borysko KZ, Zemlicka J, Kamil JP, Drach JC. 2013. Resistance of human cytomegalovirus to cyclopropavir maps to a base pair deletion in the open reading frame of UL97. Antimicrob. Agents Chemother. 57:4343– 4348. http://dx.doi.org/10.1128/AAC .00214-13. 19. Chou S, Bowlin TL. 2011. Cytomegalovirus UL97 mutations affecting cyclopropavir and ganciclovir susceptibility. Antimicrob. Agents Chemother. 55:382–384. http://dx.doi.org/10.1128/AAC.01259-10.
