Delayed low-dose oral administration of 4'-fluorouridine inhibits pathogenic arenaviruses in animal models of lethal disease

Abstract

Development of broad-spectrum antiviral therapies is critical for outbreak and pandemic preparedness against emerging and reemerging viruses. Viruses inducing hemorrhagic fevers cause high morbidity and mortality in humans and are associated with several recent international outbreaks, but approved therapies for treating most of these pathogens are lacking. Here, we show that 4'-fluorouridine (4'-FlU; EIDD-2749), an orally available ribonucleoside analog, has antiviral activity against multiple hemorrhagic fever viruses in cell culture, including Nipah virus, Crimean-Congo hemorrhagic fever virus, orthohantaviruses, and arenaviruses. We performed preclinical in vivo evaluation of oral 4'-FlU against two arenaviruses, Old World Lassa virus (LASV) and New World Junín virus (JUNV), in guinea pig models of lethal disease. 4'-FlU demonstrated both advantageous pharmacokinetic characteristics and high efficacy in both of these lethal disease guinea pig models. Additional experiments supported protection of the infected animals even when 4'-FlU delivery was reduced to a low dose of 0.5 milligram per kilogram. To demonstrate clinical utility, 4'-FlU treatment was evaluated when initiated late in the course of infection (12 or 9 days after infection for LASV and JUNV, respectively). Delayed treatment resulted in rapid resolution of clinical signs, demonstrating an extended window for therapeutic intervention. These data support the use of 4'-FlU as a potent and efficacious treatment against highly pathogenic arenaviruses of public health concern with a virus inhibition profile suggesting broad-spectrum utility as an orally available antiviral drug against a wide variety of viral pathogens.

Fig. 1.. 4′-FlU inhibits pathogenic arenaviruses in cell culture.

(A) Huh7 cells were treated with the ribonucleoside analog 4′-FlU and, 1 to 2 hours later, were infected with LASV-ZsG at an MOI of 0.125. ZsG fluorescence was measured at 3 days dpi, and fluorescence relative to DMSO-only treated cells (mock control) was calculated (green). Cell viability (blue) relative to DMSO-only treated cells (mock control) was calculated in parallel. (B) Huh7 cells were treated with 4′-FlU and, 2 hours later, were infected with LASV-ZsG at an MOI of 0.1. After 1 hour, the inoculum was replaced with a complete medium containing 4′-FlU. Supernatants were harvested at 3 dpi, and the viral titer of LASV-ZsG was measured (green). Dotted line indicates the LOD; asterisk indicates no virus detected. (C to E) HeLa cells were treated with 4′-FlU and, 1 to 2 hours later, were infected with wild-type LASV or JUNV at an MOI of 0.1. The medium was removed at 3 dpi, and the cells were fixed and stained to visualize HeLa cell nuclei (blue) or infected cells (green). The number of cells infected with LASV (C) or JUNV (D) decreased with increasing 4′-FlU concentrations. For (A) to (D), symbols represent the mean and error bars indicate the SD from four biological replicates. A representative of three independent experiments is shown; data are presented as % of DMSO-only controls. (E) Representative photomicrographs of LASV-infected HeLa cells (green) treated with increasing concentrations of 4′-FlU (HeLa cell nuclei, blue). Scale bars, 500 μm. (F to H) Plasma concentrations of 4′-FlU in guinea pig plasma after intravenous (iv) or oral dosing. (F) Plasma concentrations of 4′-FlU after a single oral dose treatment (3 to 50 mg/kg) or single intravenous treatment (5 mg/kg). (G) Plasma concentrations of 4′-FlU after a single oral dose or seven oral doses QD, all at 3 mg/kg. (H) Plasma concentrations of 4′-FlU after a single oral dose, four oral doses once QOD, or seven oral doses QD, all at 10 mg/kg. Data are shown as the mean with error bars representing SDs (n = 3 per group; n = 6 when males and females were tested).

Fig. 2.. Therapeutic efficacy of 4′-FlU in guinea pigs infected with LASV.

(A) Schematic of experimental design. Strain 13/N guinea pigs (n = 5 per group) were infected subcutaneously (sc) with 10⁴ TCID₅₀ of LASV strain Josiah and treated either QD or QOD with 4′-FlU (5 mg/kg) between 7 and 20 days dpi. Vehicle-treated control animals received vehicle only QD between 7 and 20 dpi. Animals were monitored until the end of the study at 35 dpi or until animals were euthanized because of reaching clinical end points determined by the clinical score. (B) Weight change (relative to baseline taken at −1 dpi), body temperature, and clinical score were recorded each day for each animal. Individual animals are shown, with the black solid line indicating the group daily mean, and the gray dashed line (treated groups only) representing the vehicle-only daily mean. Clinical signs were scored from 0 to 12, with severity depicted by increased intensity of red color. Animals scoring ≥ 12 were humanely euthanized; any animal that succumbed to disease before euthanasia was allocated a score of 12. Gray boxes in clinical score plots indicate the end of monitoring/scoring because of euthanasia/death. Green, QD treatment group; blue, QOD treatment group; orange, vehicle-only control group; light blue shading indicates treatment window. Significance was calculated by two-way ANOVA: *P ≤ 0.05; ***P ≤ 0.001. Also presented are combined graphs showing the mean weight change and mean body temperature for all groups. (C) Kaplan-Meier survival curves show the percent survival for each group. Survival significance was calculated by log-rank Mantel-Cox test: ***P ≤ 0.001. (D and F) RT-qPCR was used to detect LASV viral RNA in select tissues and swab samples taken from mucosal surfaces of infected guinea pigs. (E and G) Infectious LASV was isolated from select tissues and swab samples taken from mucosal surfaces of infected guinea pigs, and viral titer was measured (TCID₅₀/g or TCID₅₀/ml for tissue samples or swabs, respectively). Individual animals are represented as dots, and error bars represent the SD relative to the mean. Significance was calculated by unpaired t test, controlled for false discovery rate by two-stage step-up (Benjamini, Krieger, and Yekutieli method): *P ≤ 0.05; **P ≤ 0.01. ND, not determined; Mam., mammary gland; Conj., conjunctiva.

Fig. 3.. Therapeutic efficacy of 4′-FlU in a guinea pig model of early LASV infection.

(A) Schematic of experimental design. Strain 13/N guinea pigs (n = 5 per group) were infected subcutaneously with 10⁴ TCID₅₀ of LASV strain Josiah and treated either QD or QOD with 4′-FlU (5 mg/kg) between 7 and 15 days dpi. Vehicle-treated control animals received vehicle only QD between 7 and 15 dpi. Animals were monitored until the end of the study at 16 dpi. (B) Weight change (relative to baseline taken at −1 dpi), body temperature, and clinical score were recorded each day for each animal. Individual animals are shown, with a black solid line indicating the group daily mean, and a gray dashed line (treated groups only) representing the vehicle-only daily mean. Clinical signs were scored from 0 to 12, with severity depicted by increased intensity of red color. Animals scoring ≥ 12 were humanely euthanized; any animal that succumbed to disease before euthanasia was allocated a score of 12. Green, QD treatment group; blue, QOD treatment group; orange, vehicle-only control group; light blue shading indicates treatment window. Significance was calculated by two-way ANOVA: *P ≤ 0.05. Also presented are combined graphs showing the mean weight change and mean body temperature for all groups. (C and D) RT-qPCR was used to detect LASV viral RNA in tissues and swab samples taken from mucosal surfaces of infected guinea pigs. (E and F) Infectious LASV was isolated from tissues and mucosal swab samples of infected guinea pigs, and viral titer was measured (TCID₅₀/g or TCID₅₀/ml for tissues and swabs, respectively). Individual animals are represented as dots, and error bars represent the SD relative to the mean. Significance was calculated by unpaired t test, controlled for false discovery rate by two-stage step-up (Benjamini, Krieger, and Yekutieli method): *P ≤ 0.05; **P ≤ 0.01. Mam., mammary gland; Conj., conjunctiva.

Fig. 4.. Therapeutic efficacy of 4′-FlU against JUNV infection in guinea pigs.

(A) Schematic of experimental design. Hartley guinea pigs (n = 5 or 6 per group) were infected intraperitoneally (ip) with 10⁴ TCID₅₀ of JUNV strain Romero and treated QOD with 4′-FlU (5, 1.5, 0.5, or 0.15 mg/kg) between 7 and 19 days dpi. Mock-treated control animals received vehicle only QOD between 7 and 19 dpi. Animals were monitored until the end of the study at 28 dpi. (B) Weight change (relative to baseline taken at 0 dpi), body temperature, and clinical score were recorded each day for each animal; clinical signs (scored from 0 to 7) were recorded daily, with severity depicted by increased intensity of red color. Gray boxes in clinical score plots indicate the end of monitoring/scoring because of euthanasia/death. Individual animals are shown, with a black solid line indicating the group daily mean, and a gray dashed line (treated groups only) representing the vehicle-only daily mean. Blue, 5 mg/kg; green, 1.5 mg/kg; purple, 0.5 mg/kg; teal, 0.15 mg/kg; orange, vehicle only; light blue shading indicates treatment window. Significance compared with vehicle-only control animals was calculated by two-way ANOVA: *P ≤ 0.05. Also presented are combined graphs showing the mean weight change and mean body temperature for all groups. For the Kaplan-Meier survival plots, significance was calculated by log-rank Mantel-Cox test: **P ≤ 0.01; ***P ≤ 0.001. (C) Schematic of experimental design. Hartley guinea pigs (n = 6 per group) were infected intraperitoneally with 10⁴ TCID₅₀ of JUNV strain Romero and treated QOD with 4′-FlU (1.5 or 0.15 mg/kg) between 7 and 11 dpi. Mock-treated control animals received vehicle only QOD between 7 and 11 dpi. Animals were monitored until the end of the study at 12 dpi. (D) Weight change (relative to baseline taken at 0 dpi), body temperature, and clinical scores were recorded each day for each animal; clinical signs (scored from 0 to 7) were recorded daily, with severity depicted by increased intensity of red color. Individual animals are shown, with a black solid line indicating the group daily mean, and a gray dashed line (treated groups only) representing the vehicle-only daily mean. Green, 1.5 mg/kg; purple, 0.15 mg/kg; orange, vehicle-only; light blue shading indicates treatment window. Significance was calculated by two-way ANOVA: *P ≤ 0.05. Also presented are combined graphs showing the mean weight change and mean body temperature for all groups. (E) Infectious JUNV was measured in select tissues and sera of infected guinea pigs, and viral titers were measured (TCID₅₀/g or ml). Individual animals are represented as dots, and error bars represent the SD relative to the mean. Significance was calculated by unpaired t test, controlled for false discovery rate by two-stage step-up (Benjamini, Krieger, and Yekutieli method): **P ≤ 0.01.

Fig. 5.. Efficacy of reduced dosing and delayed initiation of 4′-FlU treatment against LASV infection in guinea pigs.

(A) Schematic of experimental design. Strain 13/N guinea pigs (n = 6 per group) were infected subcutaneously with 10⁴ TCID₅₀ of LASV strain Josiah and treated QOD with 4′-FlU (1.5, 0.5, or 0.15 mg/kg) between 7 and 15 days dpi. Mock-treated control animals received vehicle only QOD between 7 and 15 dpi. Purple, 1.5 mg/kg QOD; teal, 0.5 mg/kg QOD; brown, 0.15 mg/kg QOD; orange, vehicle only. (B and C) Weight changes (relative to baseline taken at −1 dpi), body temperature, and clinical scores for guinea pigs treated between 7 and 15 dpi were recorded each day for each animal. (D) Schematic of experimental design. Strain 13/N guinea pigs (n = 6 per group) were infected subcutaneously with 10⁴ TCID₅₀ of LASV strain Josiah and treated with 4′-FlU (5 mg/kg) between 12 and 15 dpi either 5 mg/kg QD (green) or 5 mg/kg QOD (blue). (E) Weight changes (relative to baseline taken at −1 dpi), body temperature, and clinical scores for guinea pigs treated between 12 and 15 dpi were recorded each day for each animal. For all graphs, individual animals are represented, with a black solid line indicating the group daily mean, and a gray dashed line (treated groups only) representing the vehicle-only daily mean; light blue shading indicates treatment window. Clinical signs (scored from 0 to 12) were recorded daily, with severity depicted by increased intensity of red color. Animals scoring ≥ 12 were humanely euthanized; any animal that succumbed to disease before euthanasia was allocated a score of 12. Significance of treated compared with vehicle-only control animals was calculated by two-way ANOVA: *P ≤ 0.05; n.s., not significant. (F and G) RT-qPCR was used to detect LASV viral RNA in select tissues and swab samples taken from mucosal surfaces of guinea pigs. (H and I) Infectious LASV was isolated from select tissues and swab samples, and viral titers were measured (TCID₅₀/g or TCID₅₀/ml for tissues or swabs, respectively). Individual animals are represented as dots, and error bars represent the SDs relative to the mean. Significance was calculated by unpaired t test, controlled for false discovery rate by two-stage step-up (Benjamini, Krieger, and Yekutieli method): *P ≤ 0.05; **P ≤ 0.01. Mam., mammary gland; Conj., conjunctiva.

Fig. 6.. Efficacy of delayed 4′-FlU treatment against JUNV infection in guinea pigs.

(A) Schematic of experimental design. Hartley guinea pigs (n = 5 or 6 per group) were infected intraperitoneally with 10⁴ TCID₅₀ of JUNV strain Romero and treated with 4′-FlU (1.5 mg/kg) either 9 to 22, 10 to 23, 11 to 24, or 12 to 25 days dpi. Mock-treated control animals (n = 8) received vehicle only 9 to 22 dpi. For all treatment schedules, animals were treated QD for the first 2 days and then QOD for the remaining days. Animals were monitored until the end of the study at 28 dpi. (B) Weight changes (relative to baseline taken at 0 dpi), body temperature, and clinical scores were recorded each day for each animal; clinical signs (scored from 0 to 7) were recorded daily, with severity depicted by increased intensity of red color. Gray boxes indicate the end of monitoring/scoring because of euthanasia/death. Individual animals are shown, with a black solid line indicating the group daily mean, and a gray dashed line (treated groups only) representing the vehicle-only daily mean. Blue, 9 to 22 dpi; green, 10 to 23 dpi; purple, 11 to 24 dpi; brown, 12 to 25 dpi; orange, vehicle only 9 to 22 dpi; light blue shading indicates treatment window. Significance of weight and temperature differences compared with vehicle-only controls was calculated by two-way ANOVA: *P ≤ 0.05. Also presented are combined graphs showing the mean weight change and mean body temperature for all groups. *P ≤ 0.05. (C) Kaplan-Meier survival plots show the percent survival. Significance was calculated by log-rank Mantel-Cox test: **P ≤ 0.01; ***P ≤ 0.001; n.s., not significant.

Fig. 7.. Efficacy of 4′-FlU treatment with loading and maintenance doses against JUNV infection in guinea pigs.

(A) Schematic of experimental design. Hartley guinea pigs (n = 6 per group) were infected intraperitoneally with 10⁴ TCID₅₀ of JUNV strain Romero and then given a loading dose of 5 mg/kg at 9 dpi orally, followed by loading doses of 1.5 mg/kg at either 11 dpi only (two treatments total); 11 and 13 dpi (three treatments total); 11, 13, 15, and 17 dpi (five treatments total); or 11, 13, 15, 17, 19, and 21 dpi (seven treatments total). Mock-treated control animals received vehicle only QOD between 9 and 22 dpi. Animals were monitored until the end of the study at 35 dpi. (B) Weight changes (relative to baseline taken at 0 dpi), body temperatures, and clinical scores were recorded each day for each animal; clinical signs (scored from 0 to 7) were recorded daily, with severity depicted by increased intensity of red color. Gray boxes indicate the end of monitoring/scoring because of euthanasia/death. Individual animals are represented, with a black solid line indicating the group daily mean, and a gray dashed line (treated groups only) representing the vehicle-only daily mean. Blue, seven treatments; green, five treatments; purple, three treatments; brown, two treatments; orange, vehicle only up to seven treatments; light blue shading indicates treatment window; blue vertical dashed lines represent treatment days. Significance of weight and temperature differences compared with vehicle-only controls was calculated by two-way ANOVA: *P ≤ 0.05. Also presented are combined graphs showing the mean weight change and mean body temperature for all groups: *P ≤ 0.05. (C) Kaplan-Meier survival plots show the percent survival. Significance was calculated by log-rank Mantel-Cox test: ***P ≤ 0.001.