Enhanced broad spectrum in vitro antiviral efficacy of 3-F-4-MeO-Bn, 3-CN, and 4-CN derivatives of lipid remdesivir nucleoside monophosphate prodrugs

Broad spectrum oral antivirals are urgently needed for the early treatment of many RNA viruses of clinical concern. We previously described the synthesis of 1-O-octadecyl-2-O-benzyl-glycero-3-phospho-RVn (V2043), an orally bioavailable lipid prodrug of remdesivir nucleoside (RVn, GS-441524) with broad spectrum antiviral activity against viruses with pandemic potential. Here we compared the relative activity of V2043 with new RVn lipid prodrugs containing sn-1 alkyl ether or sn-2 glycerol modifications. We found that 3-F-4-MeO-Bn, 3-CN-Bn, and 4-CN-Bn sn-2 glycerol modifications improved antiviral activity compared to V2043 when tested in vitro against clinically important RNA viruses from 5 virus families. These results support the continued development of V2043 and sn-2 glycerol modified RVn lipid prodrugs for the treatment of a broad range of RNA viruses for which there are limited therapies.


Introduction
Remdesivir (RDV; GS-5734) is an adenosine analog prodrug that binds and inhibits viral RNA-dependent RNA polymerase and exhibits broad-spectrum antiviral activity in vitro and in vivo.When administered to nonhospitalized COVID-19 infected patients, RDV reduced hospitalization or death by 87% in individuals at high risk for progression to severe disease (Gottlieb et al., 2022).
Here, we investigate the in vitro antiviral activity of V2043 and various analogs, including 3-F-4-MeO-Bn, 3-CN-Bn, or 4-CN-Bn sn-2 glycerol modifications, against members of the Flaviviridae, Filoviridae, Pneumoviridae, Paramyxoviridae, and Coronaviridae families.We show that 3-F-4-MeO-Bn, 3-CN-Bn, and 4-CN-Bn sn-2 glycerol modifications improve the antiviral efficacy of RVn lipid prodrugs against most RNA viruses tested.Collectively, this work demonstrates the broad-spectrum in vitro activity of RVn lipid prodrugs and identifies specific sn-2 glycerol modifications that improve broad spectrum antiviral potency.

Compounds
Compounds were synthesized at the University of California, San Diego, Nanosyn, Inc. (Santa Clara, CA and J-Star Research, South Plainfield, NJ.Detailed synthesis methods and characterization are provided in Carlin et al. JMedChem 2023.

Differentiation of monocyte derived dendritic cell (moDCs) from primary human monocytes
Blood was drawn from healthy human donors under IRB #181624.Blood on top of Histopaque was centrifuged at 400g for 45 min at 4 • C without acceleration.The buffy coat was isolated and washed with PBS containing 0.02% w/v EDTA.To lyse remaining red blood cells, the cell pellet was resuspended in molecular grade water.Subsequently, 10X PBS was added to a final concentration of 1X PBS prior to filtering cells with a 70 μm filter.Cells were centrifuged at 300×g for 10 min at 4 • C and washed once with PBS containing 0.02% w/v EDTA.Monocytes were isolated following the Pan Monocyte Isolation Kit (Miltenyi Biotec) instructions.Cells were plated at a seeding density ranging from 1.0 × 10 6 to 1.7 × 10 6 cells/mL in 6 well plates in complete DC media.Complete DC media consists of RPMI 1640 (Gibco) containing 10% FBS, 1% penicillin/streptomycin, 1% HEPES, 100 ng/mL recombinant human granulocyte-macrophage colony stimulating factor, and 100 ng/ mL recombinant human interleukin 4. Cells were incubated for a total of 9 days at 37 • C; media was changed on day 4 and 8, and cell differentiation was checked on day 7 or 8. Compounds and virus were added on day 8.

Drug treatments
For antiviral assays, compounds were added to cells at 0.013, 0.041, 0.123, 0.371, 1.11, 3.33, 10, and 30 μM immediately prior to addition of virus and were present for the duration of infection.For cell toxicity assays, compounds were added to cells at 1. 23, 3.70, 11.11, 33.33, and   100 μM.moDCs and Huh7.5 cells were incubated with compounds or controls for 24 and 48 h, respectively.For antiviral assays in TIME and HSAEC1-KT cells, compounds were added to cells at 0.0023, 0.0068, 0.021, 0.062, 0.19, 0.56, 1.67, and 5 μM 1 h prior to addition of virus and were present for the duration of infection.For RSV antiviral assays, compounds were added to cells at 0.0025, 0.01, 0.04, 0.16, 0.64, and 2.5 μM immediately prior to addition of virus and were present for the duration of infection.For HCoV 229E antiviral assays, compounds were added to cells at 0.0025, 0.01, 0.04, 0.16, 0.64, 2.5, and 10 μM immediately prior to addition of virus and were present for the duration of infection.For HeLa, HEp-2 and MRC-5 cell toxicity assays, compounds were added to cells at 0.0125, 0.05, 0.20, 0.80, 3.20, 12.5 and 50 μM.

Viral infection
Huh7.5 cells were seeded at 18,000 cells per well in flat bottom 96 well plates 24 h prior to treatment.Cells were treated with compounds or controls at the indicated concentrations.Following drug treatment, Huh7.5 cells were infected with DENV2 UIS 353 at an MOI of 0.02 or ZIKV PRVABC59 at an MOI of 0.0110-0.0125.Compounds and virus were incubated on cells for 48 h.On day 8 of moDC differentiation, cells were seeded at 50,000 cells per well in 96 well round bottom plates.After compound or control treatment at the indicated concentrations, moDCs were infected with DENV2 UIS 353 at an MOI of 0.33 or ZIKV PRVABC59 at an MOI of 1 for 24 h.HeLa and HEp-2 cells were seeded at 10 5 cells per well and MRC-5 cells were seeded at 10 4 cells per well, respectively, in flat bottom 96 well plates 24 h prior to treatment.Cells were treated with compounds or controls at the indicated concentrations.Following drug treatment, HeLa, HEp-2, and MRC-5 cells were infected with RSV at an MOI of 0.01 or HCoV 229E at an MOI of 0.01.

Viruses
ZIKV PRVABC59 and DENV2 UIS 353 were acquired from the World Reference Center for Emerging Viruses and Arboviruses.ZIKV PRVABC59 and DENV2 UIS 353 were expanded on C6/36 Aedes albopictus mosquito cells.Virus was titrated on 70-90% confluent U2OS cells.U2OS cells were seeded in a 6 well plate at 300,000 to 400,000 cells per well, or in a 12 well plate at 100,000 cells per well and incubated overnight.5-fold serial dilutions of virus were diluted in U2OS media with 2% FBS and incubated on the cells for 1 h, rocking plates every 15 min.Cells were washed once with PBS and complete media was added to cells.After 24 h, cells were prepared for flow cytometry according to the BD Fixation/Permeabilization kit.Virus was stained using AF647-conjugated 4G2 mAb.Infection rate was determined following flow cytometry using the ACEA Novocyte flow cytometer and subsequent FlowJo analysis.The henipaviruses used in this study (NiV-B, rNiV-ZsG (NiV-M genotype), and HeV (Chua et al., 2000;Harcourt et al., 2005;Murray et al., 1995) were propagated and titered on Vero (ATCC, CCL-81) cells.The recombinant ebolavirus used in this study (Albariño et al., 2015) was propagated and titered on Huh7 cells (APATH, LLC).Prior to conducting antiviral assays, all viruses were titered on both TIME and HSAEC1-KT cells to preclude differences in cell type-specific infectivity.

Cell viability assay
Huh7.5 cells were plated at 10,000 cells per well in opaque 96 well plates and incubated overnight.Cells were treated at the indicated concentrations with compounds or controls for 48 h.On day 8 of differentiation, moDCs were seeded at 50,000 cells per well in opaque 96 well plates and treated with compounds or controls at the indicated concentrations for 24 h.HeLa and HEp-2 were plated at 10 5 cells per well and MRC-5 cells were plated at 10 4 cells per well in opaque 96 well plates and incubated overnight.Cells were treated at the indicated concentrations with compounds or controls for 5-7 days.Cell Titer Glo or Cell Titer Glo 2.0 reagent (Promega) were used according to the manufacturer's instructions to measure cell toxicity by ATP levels.Luminescence was recorded with the Veritas Microplate Luminometer (Turner BioSystems) or HD1 Synergy plate reader.CC 50 values were calculated using Prism 9 by normalizing cell viability to the DMSO controls.

Immunofluorescence of infected Huh7.5s
Huh7.5 cells were washed three times with PBS and fixed in 4% paraformaldehyde for 30 min at room temperature (RT).Cells were washed three times with PBS and permeabilized with 0.1% Triton X in a solution of 1% BSA for 30 min at RT. Cells were incubated with pan flavivirus mouse antibody, 4G2 in 1% BSA with 0.1% Triton X overnight at 4C.After three PBS washes, cells were stained Alexa fluor 594 goat anti-mouse IgG2a(y2a) (Invitrogen) and SYTOX Green nucleic acid stain (Invitrogen) for 1 h at RT. Cells were washed three times with PBS and imaged on the Incucyte S3 System (Sartorius).IC50 values were calculated using Prism 9 by normalizing infection rate (cells with AF594 signal/cells with Sytox Green signal) to the DMSO controls.

Flow cytometry of infected moDCs
moDCs were centrifuged for 10 min at 200g and washed once with PBS.Cells were stained with Zombie Violet Fixable Viability stain according to the kit instructions and incubated for 15 min at RT. moDCs were fixed and permeabilized according to the BD Fixation/Permeabilization kit, using AF647-conjugated 4G2 mAb to stain for ZIKV or DENV viral envelope protein.Infection rate was determined following flow cytometry using the Novocyte flow cytometer and FlowJo analysis.

Fluorescence reporter-based assays (REP)
Recombinant NiV (rNiV-ZsG) (Lo et al., 2014) and EBOV (rEBOV-ZsG) (Albariño et al., 2015) expressing ZsGreen1 fluorescent protein (ZsG) were assayed for total fluorescence intensity by using an H1 Synergy plate reader (Biotek) as previously described (Lo et al., 2021).HSAEC1-KT and TIME cells were seeded at 1-2 × 10 4 cells per well in black opaque side clear bottom 96-well plates (Corning 3603, Corning, NY) and compounds were added to the assay plates for 1 h.Assay plates were transferred to the BSL-4 suite and infected with 0.25 TCID 50 per cell of the respective virus and were read at 72 h post-infection (hpi).Fluorescence signal intensity assayed in DMSO-treated, virus-infected cells were set as 100% ZsGreen fluorescence.Data points and error bars for all reporter assays indicate the mean value and standard deviation of 3 biological replicates and are representative of at least 3 independent experiments for every compound tested.Concentrations of compound that inhibited 50% of the green fluorescence signal (EC 50 ) were calculated from dose response data fitted to the mean value of experiments performed for each concentration in the 8-point, 3-fold dilution series using a 4-parameter non-linear logistic regression curve with variable slope using Graph-Pad Prism 9 (GraphPad Software, La Jolla, CA, USA).

Cytopathic effect (CPE) assay
CPE inhibition assays were conducted as previously described (Lo et al. 2020(Lo et al. , 2021)).HSAEC1-KT and TIME cells were seeded at 1-2 × 10 4 cells per well in white opaque 96-well plates, and compounds were added to the assay plates.Assay plates were transferred to the BSL-4 suite as per biocontainment requirements, infected with 0.01-0.5 TCID 50 per cell, and were analyzed with CellTiter-Glo 2.0 (Promega, Madison, WI) at 72 hpi in a HD1 Synergy plate reader.Values were normalized to uninfected cell controls according to % viability as

Statistical analysis
All statistical analyses were performed using Prism 9.3.1 (GraphPad Software).To calculate CC 50 and EC 50 values for each compound, curves were fit using [inhibitor] vs. normalized response (variable slope).To compare EC 50 values of V2043 versus V2051-V2055, data was log transformed and a two-tailed unpaired t-test was performed.To compare EC 50 values of V2043 with V2067 and V2067, an ordinary one-way ANOVA that assumed Gaussian distribution of residuals and equal standard deviations was performed on log transformed data; this test compared the mean of each compound to that of V2043.The EC 90 values of each compound were calculated using the EC 50 and Hillslope values calculated in prism, EC 90 = EC 50 (1/9 (Hillslope) ).Data is presented as the mean ± standard deviation of at least three independent experiments.Data from Huh7.5 cells was generated from at least 3 experiments.Experiments with moDCs were generated from at least 3 different donors.A "p" value of less than 0.05 was considered significant.

Modification of R1 and R2 groups of V2043
We previously generated analogs of V2043 by modifying the R1 and/ or R2 groups (Fig. 1A and B) (Carlin et al., 2023).At R1, compounds V2043, V2051, V2067, V2068 have an octadecyl group, V2052 and V2053 have an oleyl group, and V2054 and V2055 have a hexadecyl group.The R2 benzyl group present in V2043 was modified with a 3-F-4-MeO-Bn for compounds V2051, V2053, and V2055 or a cyano group at the 4′, or 3' position for compounds V2067 and V2068, respectively.

Addition of 3-CN, 4-CN or 3-F-4-MeO substituents to the benzyl group enhances antiviral activity against dengue virus and Zika virus infection in vitro
In humans, flaviviruses initially replicate in innate immune cells, such as dendritic cells, and subsequently spread hematogenously leading to hepatocyte infection (Pόvoa et al., 2014;Schmid et al., 2014;Win et al., 2019).We tested the ability of V2043, V2051, V2052, V2053, V2054, V2055, V2067, and V2068 to inhibit replication of dengue virus (DENV) and Zika virus (ZIKV) in a human hepatocyte cell line, Huh7.5, and primary human monocyte-derived dendritic cells (moDCs).moDCs and Huh7.5 cells were treated with compound concentrations ranging from 0.0137 to 30 μM and infected with DENV and ZIKV.To maximize infection, Huh7.5 cells were infected for 48 h and moDCs were infected for 24 h (Branche et al., 2022).Infection rate was determined by immunofluorescence for infected Huh7.5 cells and flow cytometry for infected moDCs.For both immunofluorescence and flow cytometry, changes in the percentage of cells positive for intracellular viral antigen were quantified.
To determine if the addition of a CN group to the R2 benzyl of V2043 increases antiviral activity, we compared the relative activity of V2067 and V2068 to inhibit viral replication in DENV and ZIKV infected Huh7.5 cells and moDCs.We found that V2067 and V2068 induce a dose-dependent inhibition of DENV and ZIKV replication in both cell types (Fig. 2G and H, Fig. 3G and H).Compared to V2043, V2067 and V2068 have significantly lower EC 50 values in DENV and ZIKV -infected Huh7.5 cells and moDCs (Fig. 2G and H, Fig. 3G and H, Tables 1 and 2).Additionally, in both DENV and ZIKV -infected moDCs and Huh7.5 cells, the SI of V2067 and V2068 exceeds that of V2043.Altogether, this data suggests that both 3-F-4-MeO and -CN R2 substitutions increase antiviral activity against DENV and ZIKV in Huh7.5 cells and moDCs.

V2043 and modified analogs exhibit broad spectrum antiviral efficacy
To establish if 3-F-4-MeO and -CN R2 modified compounds exhibit antiviral efficacy against additional RNA viruses of medical importance, we tested the ability of these compounds to inhibit viruses in the Filoviridae, Paramyxoviridae, Pneumoviridae, and Coronaviridae families in vitro.
difference between the EC 50 values of V2043 and V2067-V2068, the SI of V2067 and V2068 exceed that of V2043.Overall, this suggests that 3-F-4-MeO modified V2053 has significantly higher antiviral efficacy than V2043 against RSV in HEp-2 cells.
When comparing the efficacy of these compounds across all viruses and cell types tested, we observe that 3-F-4-MeO and CN R2 modified

Discussion
Our lipid RVn prodrugs, including V2043, have broad spectrum activity against many RNA viruses of clinical concern in vitro and SARS-CoV-2 in vivo.Further they overcome many limitations of RDV, including requirement for IV administration and plasma instability (Schooley et al., 2021).Here, we demonstrate that modifying the R2 benzyl of V2043 to contain 3-F-4-MeO, 3-CN, or 4-CN substitutions significantly improves the antiviral activity against DENV and ZIKV in Huh7.5 cells and moDCs without significantly modifying cytotoxicity.Compounds with sn-2 benzyl 3-F-4-MeO, 3-CN, and 4-CN modifications were also efficacious at inhibiting the replication of RNA viruses in the Filoviridae, Pneumoviridae, Paramyxoviridae, and Coronaviridae families.Collectively, 3-F-4-MeO and -CN R2 modified compounds generally improve broad-spectrum antiviral activity relative to V2043.
Several oral RDV prodrugs GS-5245 (ATV006, Obeldesivir), GS-621763, and VV116 are in various stages of clinical development for the treatment of SARS-CoV-2 (Cao et al., 2022;Cao et al., 2023;Cox et al., 2021;Martinez et al., 2023 [preprint], Schäfer et al., 2022).GS-5245 and GS-621763 are rapidly metabolized to RVn (GS-441524) prior to cellular uptake in tissues, including the lung.After entry into cells, GS-5245, GS-621763, and VV116 require an initial phosphorylation step, which for nucleosides is rate-limiting (slow) and believed to account for the reduced antiviral activity of RVn compared to RDV (Eastman et al., 2020).In contrast, our oral lipid prodrugs are absorbed and circulate in plasma intact (Schooley et al., 2021).For example, Syrian hamsters administered oral V2043 maintained plasma  Mean values with ±standard deviation values were derived from a minimum of at least 2 independent experiments performed in biological triplicates.EC 50 , EC 90 , and CC 50 values were calculated using Graphpad Prism 9 software.ODBG-P-RVn levels significantly above the EC 90 of SARS-CoV-2 12 h post-administration with minimal conversion to RVn (Schooley et al., 2021).Additionally, V2043 and its analogs are metabolized intracellularly to RVn-monophosphate, thereby bypassing the rate-limiting initial phosphorylation step.Given that V2043 and its related analogs containing sn-2 benzyl 3-F-4-MeO, 3-CN, or 4-CN modifications are significantly more potent in vitro than RVn, we expect that these compounds will be effective oral antivirals for the treatment of many RNA viruses.Prodrug approaches which rely on extracellular RVn may be less effective against viruses which have EC 50 values 10 to 100 times greater than V2053, 2067 and 2068 (Fig. 6B) including Nipah, Ebola, Hendra, Zika and Dengue.
Our previous work has focused on the antiviral efficacy of these analogs to SARS-CoV-2 and related coronaviruses.This is the first study to investigate the broad-spectrum antiviral activity of 3-F-4-MeO, 3-CN, and 4-CN -modified V2043 analogs.Importantly, we find that these analogs have antiviral activity against viruses with pandemic potential that currently have no antiviral therapies including dengue virus, Zika virus, Ebola virus, Hendra virus and Nipah virus.A limitation of our study is that all testing was performed in vitro.However, we previously established that V2043 is effective in limiting SARS-CoV-2 infection in mice, and future studies by our labs will test the in vivo pharmacokinetics and antiviral efficacy of these compounds (Carlin et al., 2023).This work, in combination with our previous studies, provide promising evidence that V2043 and 3-F-4-MeO, 3-CN, and 4-CN modified compounds should continue development as broad-spectrum oral antivirals for the treatment of clinically significant RNA viruses.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper

Fig. 2 .
Fig. 2. 3-F-4-MeO-Bn, 3-CN, and 4-CN R 2 substituted compounds have increased antiviral activity against DENV and ZIKV in Huh7.5, a human hepatocyte cell line.(A-F) Antiviral dose-response curves comparing compounds that differ solely by a 3-F-4-MeO-Bn modification (A and B V2043 versus V2051), (C and D V2052 versus V2053), and (E and F V2054 versus V2055) during (A, C, E) DENV and (B, D, F) ZIKV infection of Huh7.5 cells.Antiviral dose-response curves comparing compounds that differ solely by a 3-CN-Bn or 4-CN-Bn compared to V2043 in (G) DENV or (H) ZIKV infection of Huh7.5 cells.Immunofluorescence was used to quantify changes in intracellular viral antigen, detected by staining for pan-flavivirus envelope epitope, 4G2.Data shown are generated from at least 3 independent experiments performed in duplicate.Error bars represent standard deviation.Log 10 EC 50 values from each experiment were compared by unpaired t-test (A-F) or compared to V2043 by one-way ANOVA (G and H) with Dunnett's correction for multiple comparisons * P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 3
Fig. 3. 3-F-4-MeO-Bn, 3-CN, and 4-CN R 2 substituted compounds have increased antiviral activity against DENV and ZIKV in primary human monocytederived dendritic cells (moDC).(A-H) Antiviral dose-response curves for indicated compounds during (A, C, E, G) DENV and (B, D, F, H) ZIKV infection of moDCs.Flow cytometry was used to quantify changes in intracellular viral antigen, detected by staining for pan-flavivirus envelope epitope, 4G2.(I-J) Data shown are generated from at least 3 independent experiments performed in duplicate.Error bars represent standard deviation.Log 10 EC 50 values from each experiment were compared by unpaired t-test (A-F) or compared to V2043 by one-way ANOVA (G and H) with Dunnett's correction for multiple comparisons * P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 4
Fig. 4. 3-F-4-MeO-Bn, 3-CN, and 4-CN R 2 substituted compounds have increased antiviral activity over V2043 against EBOV, NiV, and HeV.Antiviral doseresponse curves for indicated compounds in either HSAEC1-KT or TIME cells infected with recombinant reporter rEBOV-ZsG (A, E) and rNiV-ZsG (B, F) viruses or with wild-type NiV-B (C) and HeV (D).Cells infected with recombinant viruses expressing ZsGreen1 protein (ZsG) were assayed for levels of fluorescence normalized to levels observed from DMSO-treated virus infected controls.HSAEC1-KT cells infected with wild-type NiV-B and HeV were assayed for cytopathic effect (CPE) based on cellular ATP levels using CellTiterGlo 2.0.CPE was normalized to cellular ATP levels observed in DMSO-treated uninfected controls.Data shown are representative at least 3 independent experiments performed in triplicate.Error bars represent standard deviation.LogEC 50 values from each experiment were compared to V2043 by one-way ANOVA with Dunnett's correction for multiple comparisons * P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 5 .
Fig. 5. Antiviral activity of RDV, V2043, and 3-F-4-MeO-Bn and CN R 2 substituted compounds against RSV and human coronavirus 229E.Antiviral doseresponse curves for indicated compounds during (A-B) RSV and (C) HCoV 299E infection of indicated cell types.Data shown are generated from at least 3 independent experiments performed in duplicate.Error bars represent standard deviation.Log 10 EC 50 values from each experiment were compared to V2043 by one-way ANOVA with Dunnett's correction for multiple comparisons * P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 6 .
Fig. 6.Relative antiviral activity of 3-F-4-MeO-Bn and CN R 2 substituted compounds compared to V2043 and RVn against RNA viruses of clinical concern in several cell types.The relative average EC 50 of lipid monophosphate prodrugs for the indicated virus and cell type compared to (A) V2043 or (B) RVn and median are shown.Average EC 50 data are derived from at least 3 independent experiments performed in duplicate.

Table 1
Antiviral activity of V2043 analogs to DENV and ZIKV in Huh7.5 cells.CC 50 /EC 50 ; Mean values with ±standard deviation values were derived from a minimum of 3 independent experiments performed in biological duplicate or triplicate.EC 50 , EC 90 , and CC 50 values were calculated using Graphpad Prism 9 software.

Table 2
Antiviral activity of V2043 analogs to DENV and ZIKV in moDCs.

Table 3
Antiviral activity of V2043 analogs to filoviruses and paramyxoviruses in HSAEC1-KT cells.Mean values with ±standard deviation values were derived from a minimum of at least 3 independent experiments performed in biological triplicates.EC 50 , EC 90 , and CC 50 values were calculated using Graphpad Prism 9 software.*RVnvalues are fromLo et al.Microbiol Spectr 2021 (PMID 34817209).

Table 4
Antiviral activity of V2043 analogs to EBOV and NiV in TIME cells.

Table 5
Antiviral activity of V2043 analogs to RSV and HCoV in human cells.