Oral pharmacokinetics and efficacy of oral phospholipid remdesivir nucleoside prodrugs against SARS-CoV-2 in mice

Abstract

Oral broad-spectrum antivirals are urgently needed for the treatment of many emerging and contemporary RNA viruses. We previously synthesized 1-O-octadecyl-2-O-benzyl-sn-glyceryl-P-RVn (ODBG-P-RVn, V2043), a phospholipid prodrug of GS-441524 (remdesivir nucleoside, RVn), and demonstrated its in vivo efficacy in a SARS-CoV-2 mouse model. Structure-activity relationship studies focusing on the prodrug scaffold identified two modifications, 3-fluoro-4-methoxy-benzyl (V2053) and 4-cyano-benzyl (V2067), that significantly enhanced the in vitro broad-spectrum antiviral activity against multiple RNA viruses when compared to V2043. Here, we demonstrate that V2043, V2053, and V2067 are all orally bioavailable, well-tolerated, and achieve high sustained plasma levels after single oral daily dosing. All three phospholipid prodrugs are significantly more active than RVn in vitro and significantly reduce SARS-CoV-2 lung titers in prophylaxis and treatment mouse models of SARS-CoV-2 B.1.351 infection. On a molar basis, V2043 and V2067 are substantially more active than obeldesivir/GS-5245 and molnupiravir in vivo. Together, these data support the continued development of phospholipid RVn prodrugs for the treatment of SARS-CoV-2 and other RNA viruses of clinical concern.

Keywords: COVID-19; SARS-CoV-2; antiviral agent; broad spectrum antiviral; in vivo efficacy; lipid prodrug; mouse model; pharmacokinetics; remdesivir; remdesivir nucleoside.

Fig 1

Phospholipid RVn prodrugs bypass rate-limiting phosphorylation. Metabolic activation pathways of lipid RVn monophosphate prodrugs and isobutyric acid RVn prodrugs, such as GS-5245. (A) Orally administered lipid RVn-MP prodrugs are absorbed, distributed, and taken up into cells intact. Intracellularly, these compounds are directly converted to RVn-MP through phospholipase C cleavage and then metabolized to the active metabolite, RVn-TP, by nucleotide kinases. Bypass of the rate-limiting first phosphorylation step increases the efficiency of lipid RVn-MP prodrug metabolism to RVn-TP. (B) Orally administered isobutyryl ester RVn prodrugs are metabolized pre-systemically by carboxyesterases in the intestine and liver to RVn. Circulating RVn is taken up into cells where it is first metabolized to RVn-MP, and then to the active metabolite, RVn-TP by nucleoside kinases. (C) Conversion to the active metabolite, RVn-TP, is rate limited by the metabolism of RVn to RVn-MP.

Fig 2

In vivo plasma pharmacokinetics of oral phospholipid RVn prodrugs after single dose PO or IV administration in BALB/c mice: (A–C) Male 22–24 g BALB/c mice were given a single oral dose of (A) 20, (B) 60, or (C) 180 mg/kg in 0.1 M sodium carbonate/bicarbonate pH 9 and plasma was obtained at 0.5, 1, 2, 3, 6, 9, 12, and 24 h (In Vivo Technologies, Moffatt Field, CA). Dotted lines represent the in vitro EC₉₀ for SARS-CoV-2 infection in Calu-3 cells: V2043 = 0.320 μM, V2053 = 0.154 μM, and V2067 = 0.082 μM in Calu-3 (7). (D) Male 20–22 g BALB/c mice were given a single dose 7 mg/kg by slow IV infusion in Ringer’s Lactate pH 7.4 and plasma was obtained at 0.083, 0.16, 0.25, 0.5, 1, 3, 6, and 9 h (In Vivo Technologies, Moffatt Field, CA). Plasma samples were analyzed to determine prodrug levels (Aliri Bio, Colorado Springs, CO).

Fig 3

Oral RVn prodrugs maintain high serum levels after repeated dosing in mice and hamsters. Multiple-dose oral tolerability and pharmacokinetics in mice and Syrian hamsters: (A and B) 5-day oral tolerability study in BALB/c mice with pharmacokinetics obtained after the final dose. Male 20–22 g mice were given the indicated drugs at (A) 80 mg/kg or (B) 100 mg/kg PO daily in 0.1 M sodium carbonate/bicarbonate pH 9 for 5 days. Plasma was obtained after the last dose at 1, 3, 6, and 12 h (In Vivo Technologies, Moffatt Field, CA). (C) 7-day oral tolerability study in Syrian hamsters with pharmacokinetics obtained after the final dose. Male 95–130 g Syrian hamsters were given V2043 at 100 mg/kg PO daily in 0.1 M sodium carbonate/bicarbonate pH 9 for 7 days and plasma was obtained after the last dose at 1, 3, 6, and 12 h (Attentive Sciences, Stillwell, KS). All plasma samples were analyzed to determine prodrug levels (Aliri Bio, Colorado Springs, CO). Dotted lines represent the in vitro EC₉₀ for SARS-CoV-2 infection in Calu-3 cells: V2043 = 0.320 μM, V2053 = 0.154 μM, and V2067 = 0.082 μM in Calu-3 (7).

Fig 4

Prophylactic efficacy of oral phospholipid RVn prodrugs in a SARS-CoV-2 mouse model: (A) 8- to 10-week-old wild-type Balb/c mice (n = 5 per group/time point) were challenged with 1 × 10⁵ PFU of SARS-CoV-2 Beta variant (B.1.351). (B) Mice were treated with vehicle, remdesivir (GS-5734), EIDD-2801 (molnupiravir), GS-5245 (obeldesivir), or lipid RVn prodrugs V2043, V2053, or V2067 orally beginning 12 h before infection. (C and D) Lungs were harvested at (C) day 2 and (D) day 4 post-infection for viral titer quantification. n = 5 mice per group, mean ± SD is shown. Groups were compared using a mixed effect model with a comparison of each treatment group with vehicle controls using Dunnett’s multiple comparison test; *P < 0.05, **P < 0.01, and ***P < 0.001. Dotted lines represent the PFU limit of detection.

Fig 5

Therapeutic efficacy of oral phospholipid RVn prodrugs in a SARS-CoV-2 mouse model: (A) 8- to 10-week-old wild-type Balb/c mice (n = 5 per group/time point) were challenged with 1 × 10⁵ PFU of SARS-CoV-2 Beta variant (B.1.351). (B) Mice were treated with vehicle, remdesivir (GS-5734), EIDD-2801 (molnupiravir), GS-5245 (obeldesivir), or lipid RVn prodrugs V2043, V2053, or V2067 orally beginning 12 h after infection. (C and D) Lungs were harvested at (C) day 2 and (D) day 4 post-infection for viral titer quantification. n = 5 mice per group, mean ± SD is shown. Groups were compared using a mixed effect model with Geisser-Greenhouse correction with comparison of each treatment group with vehicle controls using Dunnett’s multiple comparison test; *P < 0.05, **P < 0.01, and ***P < 0.001. Dotted lines represent the PFU limit of detection.