Skip Navigation

Antiviral Drug Therapy of Filovirus Infections: S-Adenosylhomocysteine Hydrolase Inhibitors Inhibit Ebola Virus In Vitro and in a Lethal Mouse Model

  1. John Huggins,
  2. Zhen-Xi Zhang* and
  3. Mike Bray
  1. Virology Division, US Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick, Maryland
  1. Reprints or correspondence: Dr. John Huggins, Virology Division, USAM-RIID, 1425 Porter Road, Fort Detrick, MD 21702-5011 (huggins{at}ncifcrf.gov).

  • ↵Present affiliation: OraVax Inc., Cambridge, Massachusetts.

 
Next Section

Abstract

Ebola (subtype Zaire) viral replication was inhibited in vitro by a series of nine nucleoside analogue inhibitors of S-adenosylhomocysteine hydrolase, an important target for antiviral drug development. Adult BALB/c mice lethally infected with mouse-adapted Ebola virus die 5–7 days after infection. Treatment initiated on day 0 or 1 resulted in dose-dependent protection, with mortality completely prevented at doses ⩾0.7 mg/kg every 8 h. There was significant protection (90%) when treatment was begun on day 2, at which time, the liver had an average titer of 3 × 105 pfu/g virus and the spleen had 2 × 106 pfu/g. Treatment with 2.2 mg/kg initiated on day 3, when the liver had an average titer of 2 × 107 pfu/g virus and the spleen had 2 × 108 pfu/g, resulted in 40% survival. As reported here, Carbocyclic 3-deazaadenosine is the first compound demonstrated to cure animals from this otherwise lethal Ebola virus infection.

The emerging pathogens Marburg virus and Ebola (EBO) virus cause very severe hemorrhagic fevers and mortality as high as 90%. As stated by Piot et al. [1], “The evolution of disease often seems inexorable and invariable.” First recognized in 1976, EBO outbreaks in Africa have been associated with high rates of morbidity and mortality. Almost nothing is known about the natural histories of these viruses, despite the field investigations (designed to identify risk factors and reservoirs) that were done after each major outbreak. There are no specific treatments for Marburg and EBO viral hemorrhagic fevers. Ribavirin, an antiviral drug used to treat several other hemorrhagic fevers, has no in vitro effect on Marburg and EBO viruses, failed to protect in multiple primate studies, and is unlikely to have any clinical value to human patients [2]. Human convalescent plasma containing antibodies has been used for treatment in the past, despite the lack of coherent clinical or experimental data regarding its use. Equine IgG with high-titer neutralizing antibodies to EBO virus protected guinea pigs and baboons but failed to protect rhesus monkeys infected with an EBO virus isolated from the 1995 Zaire outbreak [3]. Similarly, human interferon was used in 1 patient [4] despite negative experimental evidence of efficacy in a primate model. The development of effective therapy must be a top priority in preparing to deal with subsequent outbreaks.

Antiviral therapy offers the possibility of reducing morbidity, mortality, and transmission by prophylaxis of high-risk contacts. Discovery of an active antiviral compound by random screening requires either screening tens of thousands of compounds, which is impractical in a maximum-containment biosafety level 4 (BSL-4) environment, or considerable luck. Selection of compounds with a higher probability of antiviral activity against filoviruses can be inferred from the activity of the compound against other viruses with similar molecular targets. Molecular targets for drug intervention are steps central to viral replication, whose inhibition selectively blocks viral replication, either because normal cells lack the equivalent target or the cellular analogue of the target is not inhibited within the range of concentrations that inhibits the viral target. The availability of the full-length sequence of EBO [5] has allowed the identification of homologous sequences in other viruses and the determination of potential shared molecular targets. Filoviruses share similar molecular organization with rhabdo- and paramyxoviruses but are closer to the family paramyxoviridae, as determined on the basis of several of our observations. Several targets appear to be functionally identical between EBO and Marburg viruses, making it more likely that one drug would inhibit both.

S-adenosylhomocysteine (SAH) hydrolase (EC 3.3.1.1) is a cell-encoded enzyme that is an important intracellular target for antiviral drug development. Inhibitors of the cellular enzyme indirectly inhibit transmethylation reactions by a feedback mechanism. Several transmethylation reactions involved in viral replication are potential targets, including the viral enzyme (guanine-7-)methyltransferase, which transfers a methyl group from S-adenosylmethionine to the 7 position of guanosine residue of 5′ capped viral mRNA, a required step for replication of all viruses that cannot initiate mRNA synthesis by the so-called cap-stealing process. Several characterized SAH inhibitors inhibit some DNA viruses (e.g., poxviruses) and negative-strand RNA viruses, including filoviruses (Marburg and EBO) [6], rhabdoviruses (rabies, vesicular stomatitis virus), and paramyxoviruses (parainfluenza, respiratory syncytial virus) [7].

How can inhibition of SAH hydrolase, a cellular enzyme involved in the metabolic pathway of many intracellular methylation reactions, impart any specificity toward viral replication that would make it a selective antiviral target? One possible explanation, borne out by experimental data, is that qualitative differences exist between the sensitivity to feedback inhibition by the reaction product SAH of viral methyltransferases and that of the host cell—encoded enzyme (reviewed in [8]). This allows for inhibition at concentrations that do not interfere with cellular methylation. The methyltransferase reaction involves transfer of the methyl group from S-adenosylmethionine to an acceptor molecule, resulting in SAH as a reaction product. This compound is a feedback inhibitor of the enzyme and must be removed by the cellular enzyme SAH hydrolase in order for the enzyme to continue effective methylation. Thus, inhibition of SAH hydrolase by a drug effectively shuts down methylation and any steps in viral replication that are dependent upon methylation.

Previous SectionNext Section

Materials and Methods

Biologic containment

Infectious material and animals were handled in maximum-containment BSL-4 facilities at the US Army Medical Research Institute of Infectious Diseases (USAMRIID). Laboratory personnel wore positive-pressure protective suits (ILC; Dover, Frederica, DE) equipped with HEPA filters and supplied with umbilical-fed air.

Virus and cells

The 1976 Mayinga strain of EBO (subtype Zaire [EBO-Z]) virus was passaged once in Vero cells or three times intracerebrally in suckling mice and then once in Vero cells [9]. EBO-Z viruses (1976 and 1995 strains) that had been passaged twice in Vero cells were provided by Peter Jahrling (USAMRIID). The viruses were amplified in Vero E6 cells, and the supernatant was collected to produce stocks of EBO-Z '76 Vp2, EBO-Z '76 Mp3 Vp2, and EBO-Z '95 Vp3. The derivation and biologic properties of “mouse-adapted virus,” a doubly plaque-purified, ninth mouse—passage virus derived from the 1976 strain of EBO-Z and designated EBO-Z '76 Mp3 Vp2 Mp9 GH, is described elsewhere in this supplement [10]. The E6 line of Vero African green monkey kidney cells, Vero C1008 (ATCC CRL 1586); DBS-FRhL (ATCC CL 160); LLC-MK2 (ATCC CLL 7); MRC-5 (ATCC CCL 171); SW13 (ATCC CLL 105); Vero 76 (ATCC CRL 1586); and BALB/3T3 clone A31 (ATCC CCL 163) were propagated in Eagle MEM with Earle's salts, nonessential amino acids, 10% fetal bovine serum (FBS), glutamine, penicillin, and streptomycin; the same medium with only 2% FBS was used as replacement medium (subcomplete medium) after cell infection.

Compounds

Carbocyclic adenosine (aristeromycin; Ca-Ado) was purchased from Sigma (St. Louis). Carbocyclic 3-deazaadenosine (3-deazaaristeromycin; Ca-c3 Ado) [11] and (−)-9-[trans-2′,trans - 3′ - dihydroxy - 4′ - (methyl) - cyclopent - 4′ - enyl] - 3 - deazaadenine (Me-Ca-c3 Ado) were obtained from John Montgomery and John Secrist (Southern Research Institute, Birmingham, AL). Neplanocin A [(−)-9-[trans-2′,trans-3′-dihydroxy-4′-hydroxymethyl-cyclopent-4′-enyl]-adenine] (Npc A) was obtained from Roland K. Robins (Brigham Young University, Provo, UT). 3-deazaneplanocin A [(−)-9-[trans-2′,trans-3′-dihydroxy-4′-hydroxymethyl-cyclopent-4′-enyl]-3-deazaadenine, c3-Npc A] was obtained from John S. Driscoll (National Cancer Institute, Bethesda, MD). 9-[trans-2′,trans-3′-dihydroxycyclopent-4′-enyl]-adenine (DHCpAdo) and 9-[trans-2′,trans-3′-dihydroxycyclopent-4-enyl]-3-deazaadenine (DHCp-c3Ado) were obtained from Ronald T. Borchardt (University of Kansas, Lawrence, KS). 4′,5′-didehydro-5′-deoxy-5′-fluoroadenosine (DDFA) was obtained from James McCarthy (Merrell Dow Research Institute, Cincinnati). Ribavirin (β-d-ribofuranosyl-1,2,4 triazole-3-carboximide) was obtained from ICN Pharmaceutical (Covena, CA). Guanosine-2′,3′-dialdehyde (2′-O-[(R)-formyl(guanin-9-yl)methyl]-(R) glyceraldehyde) and adenosine-2′,3′-dialdehyde (2′-O-[(R)-formyl(adenin-9-yl)methyl]-(R) glyceraldehyde) were obtained from John P. Neenan (Rochester Institute of Technology, Rochester NY). All compounds were submitted to our laboratory for evaluation in the Filovirus Antiviral Testing Program.

Virus plaque assay

Specimens were diluted serially in sub-complete medium, adsorbed onto confluent Vero E6 cells in 6- or 12-well plates, incubated for 1 h at 37°C, and covered with an agarose overlay (as modified from reference [12]). A 1:5000 dilution of neutral red in buffered saline solution was added 6 days later, and plaques were counted the following day.

Drug-screening antiviral-activity assay based on MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide)

Vero E6 cells were plated in 96-well microtiter plates in complete medium and used within 72 h of reaching confluency. If required, compounds were first dissolved in DMSO or ethanol to give no more than a 1% final concentration of solvent. Compounds were dissolved and diluted in subcomplete growth medium.

Two-fold dilutions of compound (50 µL) were added and incubated for 1 h. Four rows of the microtiter plate were infected with EBO-Z viruses (MOI, 0.01 pfu/cell) in 50 µL, and four rows were sham-infected with subcomplete medium (50 µL) as controls for drug toxicity. The plate was incubated for 6–7 days in a CO2 incubator at 37°C until cytopathic effect (CPE) reached 3+ in control wells.

Viable cells were assayed by using an MTT-based procedure that measures the conversion of colorless MTT tetrazolium to colored MTT formazan by mitochondrial enzymes in the electron transport chain [13]. Previously filtered MTT (0.83 mg/mL final concentration in subcomplete medium) was then added to the plate and incubated for 2 h at 37°C. Medium was then removed, and the insoluble reaction product was solubilized by adding 100 µL of DMSO. The plate was mixed on an orbital shaker until the reaction product was completely dissolved. The colored MTT formazan reaction product was measured at 570 nm with a Vmax ELISA reader (Molecular Devices, Menlo Park, CA), and data were captured on a computer for manipulation and analysis. The resulting dose-response curves were fitted to a four-parameter logit curve by use of a computer program (SOFTmax; Molecular De- vices) to determine the IC50, the TC50 (50% toxic concentration), and the therapeutic index, which was calculated as TC50/IC50. Antiviral protection curves were corrected for toxicity, using a previously described toxicity curve [13].

Drug-screening antiviral-activity assay based on ELISA

Plates (96-well) of Vero E6 cells (or FRL-103, LLC-MK2, MRC-5, SW-13, or Vero cells) were prepared, treated with drug, and infected as above except that separate plates were used for infection and toxicity assays. Both plates were incubated for 5 days in a CO2 incubator at 37°C until CPE in control wells reached only 1+ to 2+ and the monolayers were still intact. Toxicity was measured in uninfected treated plates exactly as described above.

The virus-infected cells were fixed by adding 500 µL of 10% neutral buffered formalin for 20 min at room temperature and then inverted on absorbent paper and air dried. Plates were decontaminated by irradiation with 5 mR of gamma irradiation. EBO antigen was detected by ELISA essentially as described in [14], utilizing a mouse monoclonal antibody directed against the viral glycoprotein as a detector. The resulting dose-response curves were analyzed as above.

Drug-screening antiviral-activity assay based on plaque assay

The plaque assay described above was modified as follows. Vero E6 cells were plated in 6-well plates in complete medium and used within 72 h of confluency. If required, compounds were first dissolved in DMSO or ethanol to give no more than a 1% final concentration. Compounds were dissolved in 2X overlay medium without agarose and diluted 1:4 in the same medium, and 50 pfu/well EBO virus was adsorbed for 1 h at 37°C. Appropriate 2X dilutions of compound in 2X overlay medium were mixed with agarose and added to triplicate infected wells and one uninfected well serving as a drug toxicity control. The rest of the assay was done as described. Toxicity was scored visually with a phase contrast microscope, and CPE was scored on a scale of 1+ to 4+ [15].

Yield-inhibition antiviral assays

Vero E6 cells in 6-well plates were infected with 100 pfu of EBO-Z and incubated in the presence of increasing amounts of Ca-c3Ado or DDFA for 4 days. The virus titers in the supernatants were then determined by plaque assay as described above.

Drug evaluation: EBO mouse model

Groups of 6-week-old BALB/c mice (Charles River Breeding Laboratory, Wilmington, MA) were housed in filtertop microisolator cages and given commercial mouse chow and water ad libitum. Mice were inoculated intraperitoneally (ip) with 1 pfu (30 LD50 was determined in adult mice) of mouse-adapted EBO-Z. Toxicity groups were sham-infected with 0.1 mL of diluent. Ca-c3 Ado was diluted in PBS at concentrations that allowed the appropriate dose to be delivered ip in 0.1 mL. Care was taken to maintain the accuracy of the Ca-c3 Ado dosing interval at 8 h throughout the study. Mice were weighed daily, and deaths were recorded at 8-h intervals. Animals that survived 45 days were rechallenged with 100 pfu of the same virus to determine any protective immunity. Details of the model are published in a companion paper [10].

Previous SectionNext Section

Results

Antiviral activity of SAH hydrolase inhibitors

An antiviral screening program previously identified anti-EBO activity of a known inhibitor of SAH hydrolase [6]. Inhibition of EBO-Z replication in Vero E6 cells by a series of known SAH inhibitors is shown in figure 1. Permissive Vero E6 monolayers were infected with EBO virus at a low MOI such that if viral replication was inhibited by the compound, there would not be significant spread of the virus to other cells in the monolayer and the ELISA-based viral antigen-detection system would not detect increased viral antigen. Ca-c3 Ado and c3-Npc A showed inhibitory values against the EBO-Z viruses (1976 and 1995 strains), EBO (subtype Sudan), and Marburg (Musoke strain) virus in cell lines of primate (FRhL, LLC-MK2, MRC-5, SW13, Vero 76, Vero E6) and mouse (BALB/3T3 clone A31) origin (data not shown).

Figure 1
View larger version:
Figure 1

Inhibition of Ebola virus replication in Vero E6 cells by series of S-adenosylhomocysteine hydrolase inhibitors. Compounds were assayed by microtiter format ELISA/MTT as described in Materials and Methods; values are average of at least duplicate determinations. AVS = antiviral screening identification no.; R1 and R2 = substituent groups at R1 and R2 sugar moiety; TC50 = 50% toxic concentration; TI = therapeutic index; NA = not active. * Not toxic at maximum concentration tested.

IC50 was not dependent on the assay format or EBO strain, as shown in table 1, wherein three assay formats (ELISA, MTT, and plaque reduction) were compared for 2 virus strains. Npc A and its structural analogues, c3-Npc A, DHCp-Ado, and DHCp-c3 Ado, had equivalent antiviral inhibitory activity (IC50 = 2–17 µM) and were among the most active SAH hydrolase inhibitors found on a molar basis. Npc A, based on a cyclopentonyl (Cp) sugar, was quite toxic (TC50 = 60 µM) to cells. The analogue c3-Npc A, lacking a nitrogen at the 3 position of the adenosine ring and not phosphorylated by adenosine kinase, was nontoxic to cells. Because of the lack of a 5′ -OH on the sugar, the analogues HDCpAdo and DHCp-c3 Ado could not be metabolized to form the toxic nucleotide triphosphate derivative and, therefore, were also nontoxic (TC50 >1700 µM).

For compounds based on the carbocyclic (Ca) sugar, Ca-Ado (phosphorylated) was toxic, but its derivatives Ca-c3 Ado and Me-Ca-c3 Ado (not phosphorylated) were nontoxic. DDFA, the only suicide inhibitor (irreversible inactivator) of the enzyme among the compounds tested, was the most effective at completely eliminating viral replication (see below and figure 2). However, the 3.5-log reduction seen with Ca-c3 Ado was adequate to treat animals with considerable virus burdens (see below). The possible advantage of complete inhibition of viral replication in disease treatment has not been evaluated to date. Adenosine dialdehyde, a known SAH hydrolase inhibitor, was effective, while the structurally related guanosine dialdehyde, which is not an inhibitor of the enzyme, was not active. Ribavirin was also inactive, as reported previously [16].

Figure 2
View larger version:
Figure 2

Reduction in virus yield of Vero E6 cells infected with Ebola (subtype Zaire; 1976 strain) virus in presence ofS-adenosylhomocysteine hydrolase inhibitors. Ca-c3 Ado = carbocyclic 3-deazaadenosine; DDFA = 4′,5′-didehydro-5′-deoxy-5′-fluoroadenosine.

Quantitative determination of reduction in virus yield

The reduction in virus yield was determined at the time of maximum viral replication for a graded series of concentrations of Ca-c3 Ado and compared with DDFA (figure 2). In a dose-dependent manner, Ca-c3 Ado (2 µg/mL, 7.6 µM) reduced viral replication by 3 logs. There was no further inhibition, even with a 100-fold increase in drug concentration. DDFA completely inhibited viral replication at ⩾32 µg/mL (119 µM). Most other SAH hydrolase inhibitors tested could not completely inhibit viral replication (data not shown).

Evaluation of prophylactic administration of Ca-c3 Ado in mice

Previous work in a lethal SCID mouse model for EBO infection demonstrated that both Ca-c3 Ado and c3Npc A could inhibit viral replication when administered every 12 h, but the lack of a functional immune system in these animals precluded a realistic evaluation of the compounds (data not shown). Pharmacokinetic and distribution studies [17, 18] in adult BALB/c mice showed that Ca-c3 Ado A is less rapidly eliminated in tissues than c3-Npc A. Although the IC50 for Ca-c3 Ado was 15-fold higher than for c3-Npc A, the area under the curve, an indication of the amount of compound available, was 100-fold better. Ca-c3 Ado had a slightly longer serum half-life than c3-Npc A (23 vs. 13 min) [17]. Adult BALB/c mice infected with 30 LD50 of mouse-adapted EBO were treated prophylactically with Ca-c3 Ado every 8 h, beginning 24 h before infection, with doses ranging from 0.03 to 20 mg/kg. A dose-dependent increase in mean time to death was seen as doses increased from 0.08 to 0.3 mg/kg (table 2). Mice were completely protected by doses of ⩾0.7 mg/kg. Drug was well tolerated at the highest dose tested (20 mg/kg), which was 28 times the minimum protective dose. Some weight loss occurred during compound dosing (days 01 to 8), but at doses <2.2 mg/kg, animals returned to baseline weights during the subsequent 5 days. At and above this dose, a less dramatic weight recovery was seen, but all animals remained healthy throughout the study period. Infected, treated animals that survived were rechallenged with 100 pfu (3000 LD50) of the same virus to obtain positive evidence that they were infected in the experiment. All animals were protected, including animals in the higher-dose groups that never showed signs suggesting viral infection (i.e., weight loss greater than toxicity controls).

Figure 3
View larger version:
Figure 3

Therapeutic window for initiation of therapy of EBO-Z—infected BALB/c mice treated with carbocyclic 3-deazaadenosine (Ca-c3 Ado) and extent of viral production at time of initiation of treatment. Stippled columns = % of mice surviving when treated with Ca-c3 Ado administered every 8 h at dose that provided maximum survival; ◊ = mean viremia (pfu/mL); ○ = mean liver titer (pfu/g tissue); △ = mean spleen titer (pfu/g tissue). Virus titers were mean for 4 animals (2 at each time point in 2 separate experiments). Virus titers were determined as described in [10].

Table 1
View larger version:
Table 1

Comparison of IC50 (µM) and 50% toxic dose (µM) for various drug compounds by various antiviral assay formats.

Table 2
View larger version:
Table 2

Effect of treatment with carbocyclic 3-deazaadenosine (Ca-c3 Ado) on survival of 6-week- old BALB/c mice infected with 30 LD50 of Ebola virus.

To determine the latest time that Ca-c3 Ado treatment could be successfully initiated, we infected mice as above, and treatment at three dose levels was initiated on days 0, 1, 2, or 3 relative to virus challenge (see below; figure 3). Complete protection occurred when treatment was initiated immediately and after 1 day of infection. Mice were significantly protected (90%) when treatment was begun on day 2. Other experiments showed that the liver had an average titer of 3 × 105 pfu/g virus and the spleen had 2 × 106 pfu/g. While controls died within a mean of 6.7 days, the only treated animal that died did so on day 15. Treatment (2.2 mg/kg) for 10 mice was initiated on day 3, when the liver had an average titer of 2 × 107 pfu/g virus and the spleen had 2 × 108 pfu/g virus; 4 of the 10 mice survived. Toxicity, as judged by weight loss greater than that for untreated controls, was seen at the 20 mg/kg dose when treatment was begun on days 1–3. Toxicity was also seen with a 6.7-mg/kg dose when treatment was begun on day 3 (table 3), which required a shorter treatment schedule. Because substantial infection of the liver occurred by this time, the lack of protection may have been due to altered compound metabolism or pharmacokinetics caused by reduced drug elimination, leading to toxic levels. We did not test lower doses of compound, so we do not know if this was the maximum protection possible.

Table 3
View larger version:
Table 3

Effect of treatment with carbocyclic 3-deazaadenosine (Ca-c3 Ado) on survival of 6-week-old BALB/C mice infected with 30 LD50 of Ebola virus.

When treatment was initiated on day 3, higher doses of drug resulted in a significant increase in mean time to death. None of the animals had signs of disease at the end of the treatment period; however, they all died 5–6 days after cessation of treatment. Two treated moribund animals were killed 5 days after the end of treatment; both had virus titers in the liver that were 2 logs lower than those expected of terminal animals, suggesting that viral replication initially may have been partially controlled but resumed after treatment was discontinued. Treated survivors were rechallenged with 100 pfu (3000 LD50) of virus and, once again, all animals were protected, demonstrating that they were infected.

Previous SectionNext Section

Discussion

The task of developing filovirus chemotherapy is complicated by the constraints imposed in conducting research under BLS-4 conditions, the limited information available on the mechanism of viral replication, and the lack of closely related viruses of lower pathogenicity that could serve as surrogates for drug studies. The recognition that filoviruses share similar gene organization with paramyxoviruses provided a rationale for selecting compounds for initial testing. Inhibition of EBO virus replication by several compounds known to inhibit SAH hydrolase led us to investigate a series of known inhibitors, nine of which we report here. All SAH hydrolase inhibitors tested inhibited viral replication, and where their inhibition constant was known [19], it correlated with the EBO IC50 (data not shown). The specificity of inhibition was shown by the lack of antiviral activity of guanosine dialdehyde (not an inhibitor of SAH hydrolase) compared with that of adenosine dialdehyde. SAH hydrolase has been recognized as an important target for antiviral drug development for some time [8].

Antiviral activity of Ca-c3 Ado in vitro and in vivo was in the range reported for other viruses, including respiratory syncytial virus and parainfluenza virus [7, 20]. Initial evaluation with dosing either once or twice per day demonstrated moderate antiviral activity [20]. Studies with c3-Npc A in a SCID mouse model showed that while dosing every 12 h only reduced virus titers by 1–2 logs, dosing every 8 h reduced the titer by >6 logs (unpublished results). Ca-c3 Ado was initially evaluated in a SCID mouse model, in which it significantly slowed the rate of viral replication and increased the mean time to death (unpublished observations); however, this model was far from ideal for drug evaluation.

Our efforts to adapt EBO virus to produce lethal disease in immunocompetent adult mice by serial passage is published elsewhere in this supplement [10]. Adult BALB/c mice infected by ip inoculation with 30 LD50 of mouse-adapted EBO virus become ill within 3–4 days and uniformly die on days 5–7. The virus infects cells of the mononuclear phagocytic system, hepatocytes, and endothelial cells. Pathologic changes in the liver and spleen resembled those seen in EBO-Z infection of humans and nonhuman primates. Viral replication is first detectable on day 2 and leads to a rapid rise of virus titers in serum and tissues, reaching near maximum titers by day 4. This pathology resembles that observed in primates dying of EBO infection [10]. While this is clearly a severe model of viral infection, it appears appropriate as a model for the human disease.

Our initial concern was that the model would present a formidable challenge for antiviral intervention. The design of initial prophylactic studies had the benefit of appropriate pharmacokinetic studies that established the requirement for dosing at least every 8 h. The treatment duration was empirically selected as 1.5 times the expected mean time to death, and this parameter had not been evaluated to determine if this is optimal. Studies to determine the therapeutic window for intervention showed that treatment could be initiated quite late in the disease process on day 3, when 107–108 pfu/g virus was present in organs. These animals were symptomatic when treatment was initiated and recovered slowly over the next 2 weeks.

Later times of initiation of therapy have not yet been investigated because increasing drug toxicity, observed when treatment was initiated on day 2 or 3, suggested that either sufficient organ damage occurred to limit drug dosage due to increased sensitivity or that drug pharmacokinetics had been altered because the rate of drug elimination had changed. Elimination of Ca-c3 Ado from tissues of healthy mice is rapid and correlates with the appearance of a metabolite in serum that has yet to be identified [17]. Significant liver necrosis was seen by days 2 to 3. The liver is involved in the clearance of many nucleoside analogues. Additional studies of drug levels in mice 2–4 days after infection will be required to design studies to determine how late in the infection process treatment can be initiated. Nevertheless, Ca-c3 Ado is the first compound reported to cure animals from this otherwise lethal infection.

SAH hydrolase is a highly conserved enzyme among species, and the virus IC50 of SAH hydrolase inhibitors does not differ significantly among cells derived from human, nonhuman primate, and murine tissues. Optimization of conditions for Ca-c3 Ado delivery dramatically improved antiviral activity, compared with published results of activity against other viruses treated under less than optimal conditions. Two members of this class of nucleoside analogues had similar short serum half-lives, which suggests the need to reinvestigate the in vivo activity of other members of this class. The therapeutic window found for Ca-c3 Ado in this EBO mouse model supports the potential of this compound for therapy of filovirus infection, as does the activity in nonhuman primates who were infected with respiratory syncytial virus and dosed every 12 h (Soike K, personnel communication). The availability of an EBO virus nonhuman primate model [3] will facilitate further evaluation of this compound.

Previous SectionNext Section

Acknowledgments

We thank Merhl Gibson, Debbie Kefauver, Nancy Jaax, Peter Jahrling, and Michelle Young for expert assistance and helpful discussions.

Previous SectionNext Section

Footnotes

  • The views, opinions, and/or findings contained herein are those of the authors and should not be construed as an official Department of the Army position, policy, or decision unless so designated by other documentation.

  • The investigators adhered to the “Guide for the Care and Use of Laboratory Animals,” prepared by the Committee on Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources, National Research Council (National Institutes of Health Publication No. 86-23, revised 1985) and used facilities fully accredited by the American Association for Accreditation of Laboratory Animal Care.

  • This work was performed while M. Bray and Z.-X. Zhang held National Research Council-USAMRIID Senior Research Associateships.

Previous Section
 

References

  1. 1.↵
    1. Piot P,
    2. Sureau P,
    3. Bremen G,
    4. et al.,
    5. Pattyn SR
    . Ebola virus haemorrhagic fever. Amsterdam: Elsevier/North-Holland Biomedical Press; 1978. Clinical aspects of Ebola virus infection in Yambuku area, Zaire, 1976; p. 7-14.
  2. 2.↵
    1. Huggins JW
    . Prospects for treatment of viral hemorrhagic fevers with ribavirin, a broad-spectrum antiviral drug. Rev Infect Dis 1989;11:S750-61.
    Medline
  3. 3.↵
    1. Jahrling PB,
    2. Geisbert J,
    3. Swearengen JR,
    4. et al
    . Passive immunization of Ebola virus-infected cynomolgus monkeys with immunoglobulin from hyperimmune horses. Arch Virol Suppl 1996;11:135-40.
    Medline
  4. 4.↵
    1. Emond RT,
    2. Evans B,
    3. Bowen ET,
    4. Lloyd G
    . A case of Ebola virus infection. Br Med J 1977;2:541-4.
    Abstract/FREE Full Text
  5. 5.↵
    1. Sanchez A,
    2. Kiley MP,
    3. Holloway BP,
    4. McCormick JB,
    5. Auperin DD
    . The nucleoprotein gene of Ebola virus: cloning sequencing, and in vitro expression. Virology 1989;170:81-91.
    CrossRefMedlineWeb of Science
  6. 6.↵
    1. Huggins JW,
    2. Zhang Z,
    3. Monath TI
    . Inhibition of Ebola virus replication in vitro and in a SCID mouse model by S-adenosylhomocysteine hydrolase inhibitors [abstract]. Antiviral Res 1991: suppl 1:122.
  7. 7.↵
    1. De Clercq E,
    2. Montgomery JA
    . Broad-spectrum antiviral activity of the carbocyclic analog of 3-deazaadenosine. Antiviral Res 1983;3:17-24.
    CrossRefMedlineWeb of Science
  8. 8.↵
    1. De Clercq E
    . S-adenosylhomocysteine hydrolase inhibitors as broad-spectrum antiviral agents. Biochem Pharmacol 1987;36:2567-75.
    CrossRefMedlineWeb of Science
  9. 9.↵
    1. Geisbert TW,
    2. Jahrling PB
    . Differentiation of filoviruses by electron microscopy. Virus Res 1995;39:129-150.
    CrossRefMedlineWeb of Science
  10. 10.↵
    1. Bray M,
    2. Davis K,
    3. Geisbert T,
    4. Schmaljohn C,
    5. Huggins J
    . A mouse model for evaluation of prophylaxis and therapy of Ebola hemorrhagic fever. J Infect Dis 1998;178:651-61. 1999; 179 (suppl 1):S248–58.
    Abstract/FREE Full Text
  11. 11.↵
    1. Montgomery JA,
    2. Clayton SJ,
    3. Thomas HJ,
    4. et al
    . Carbocyclic analogue of 3-deazaadenosine: a novel antiviral agent using S-adenosylhomocysteine hydrolase as a pharmacological target. J Med Chem 1982;25:626-9.
    CrossRefMedlineWeb of Science
  12. 12.↵
    1. Moe JB,
    2. Lambert RD,
    3. Lupton HW
    . Plaque assay for Ebola virus. J Clin Microbiol 1981;13:791-3.
    Abstract/FREE Full Text
  13. 13.↵
    1. Gabrielsen B,
    2. Monath TP,
    3. Huggins JW,
    4. et al
    . Antiviral (RNA) activity of selected Amaryllidaceae isoquinoline constituents and synthesis of related substances. J Nat Prod 1992;55:1569-81.
    CrossRefMedline
  14. 14.↵
    1. Horling J,
    2. Lundkvist A,
    3. Huggins J,
    4. Niklasson B
    . Antibodies to Puumala virus in humans determined by neutralization test. J Virol Methods 1992;39:139-47.
    CrossRefMedlineWeb of Science
  15. 15.↵
    1. Kirsi JJ,
    2. North JA,
    3. Mckernan PA,
    4. et al
    . Broad-spectrum antiviral activity of 2-β-d-ribofuranosyl selenazole 4 carboxamide, a new antiviral agent. Antimicrob Agents Chemother 1983;24:353-61.
    Abstract/FREE Full Text
  16. 16.↵
    1. Huggins JW,
    2. Galasso GJ,
    3. Whitley RJ,
    4. Merigan TC
    . Antiviral agents and viral diseases of man. New York: Raven Press; 1990. RNA viruses that cause hemorrhagic, encephalitic febrile disease; p. 691-726.
  17. 17.↵
    1. Coulombe RA Jr.,
    2. Huie JM,
    3. Sharma RP,
    4. Huggins JW
    . Pharmacokinetics of the antiviral agent carbocyclic 3-deazaadenosine. Drug Metab Dispos 1993;21:555-9.
    Abstract
  18. 18.↵
    1. Coulombe RA Jr.,
    2. Sharma RP,
    3. Huggins JW
    . Pharmacokinetics of the antiviral agent 3-deazaneplanocin A. Eur J Drug Metab Pharmacokinet 1995;20:197-202.
    MedlineWeb of Science
  19. 19.↵
    1. Guranowski A,
    2. Montgomery JA,
    3. Cantoni GL,
    4. Chiang PK
    . Adenosine analogues as substrates and inhibitors of S-adenosylhomocysteine hydrolase. Biochemistry 1981;20:110-5.
    CrossRefMedline
  20. 20.↵
    1. Wyde PR,
    2. Ambrose MW,
    3. Meyer HL,
    4. Zolinski CL,
    5. Gilbert BE
    . Evaluation of the toxicity and antiviral activity of carbocyclic 3-deazaadenosine against respiratory syncytial and parainfluenza type 3 viruses in tissue culture and in cotton rats. Antiviral Res 1990;14:215-25.
    CrossRefMedlineWeb of Science
| Table of Contents

This Article

  1. J Infect Dis. (1999) 179 (Supplement 1): S240-S247. doi: 10.1086/514316
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 205 (5)
  1. Journal of Infectious Diseases
  1. Alert me to new issues

Most