Blocking NS3–NS4B interaction inhibits dengue virus in non-human primates – Nature

Compound

The synthesis of JNJ-1802 (Fig. 1a) is published in patent application WO 2016/180696. The synthesis and chemical characterization of JNJ-1802 are also provided in the Supplementary Methods. For in vitro experiments, the compound was dissolved in 100% dimethyl sulfoxide (DMSO) as a 10  mg ml⁻¹ or a 5 mM stock. For oral administration in AG129 mice, JNJ-1802 was dissolved in 80% polyethylene glycol 400 (PEG400)/20% H₂O. For oral administration in rhesus macaques (M. mulatta), JNJ-1802 was dissolved in 100% PEG400. The formulations used during the study in NHPs were prepared a maximum 3 days before dosing and stored at room temperature, protected from daylight.

Cells

Vero cells (African green monkey kidney cells; European Collection of Authenticated Cell Cultures (ECACC), CL 84113001/American Type Culture Collection (ATCC) CCL-81) were maintained in Eagle’s minimum essential medium (MEM) supplemented with 10% fetal bovine serum (FBS) (Sigma-Aldrich), 2 mM l-glutamine and 0.02 mg ml⁻¹ gentamicin (Thermo Fisher Scientific). Vero E6 cells (ATCC, CRL-1586) were cultured in MEM supplemented with 10% heat-inactivated FBS, 2 mM l-glutamine and 100 U ml⁻¹ penicillin–streptomycin (Sigma-Aldrich). Huh-7 hepatoma-derived cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM), supplemented with 10% FBS, 2 mM l-glutamine and 0.02 mg ml⁻¹ gentamicin. In antiviral assays using Vero and Huh-7 cells, the culture medium contained 2% FBS and 10% FBS, respectively. In antiviral experiments with Vero E6 cells, 2% heat-inactivated FBS was used. Huh-7 replicon cells were cultured in the same medium as mentioned above, supplemented with 75 µg ml⁻¹ hygromycin B (Roche).

Huh-7 cells stably expressing T7 RNA polymerase and DENV NS2B-NS3 (Huh-7–T7/NS2B–NS3 cells) were generated by lentiviral transduction, as previously described⁵². Cells were cultured at 37 °C and 5% CO₂ in DMEM, supplemented with 10% FBS, 2 mM l-glutamine, 100 U ml⁻¹ penicillin, 100 µg ml⁻¹ streptomycin, 5 µg ml⁻¹ zeocin, 1 µg ml⁻¹ puromycin and 1% non-essential amino acids.

THP-1 cells (TIB-202, ATCC) expressing dendritic-cell-specific intercellular adhesion molecule-3-grabbing non-integrin (DC-SIGN) were propagated in Roswell Park Memorial Institute (RPMI) (Lonza) supplemented with 10% heat-inactivated FBS (F7524, Sigma-Aldrich) and 0.04% gentamicin (Gibco-Life Technologies). All cell lines (Vero, Huh-7 and THP-1/DC-SIGN) were regularly tested for mycoplasma contamination.

C6/36 mosquito cells (from Aedes albopictus; ATCC, CCL-1660) were cultivated in the absence of 5% CO₂ at 28 °C in Leibovitz’s L-15 medium (Thermo Fisher Scientific), supplemented with 10% FBS, 1% non-essential amino acids (Thermo Fisher Scientific), 1% HEPES buffer (Thermo Fisher Scientific) and 1% penicillin (100 U ml⁻¹) and streptomycin (100 μg ml⁻¹) solution (Thermo Fisher Scientific).

Antibodies

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or β-actin were used as loading controls for cell lysates (input), and were visualized using mouse monoclonal anti-GAPDH, G-9 (sc-365062, I2320, Santa Cruz Biotechnology, 1:1,000) or mouse monoclonal anti-β-actin, AC-15 (A5441, 079M4799V, Sigma-Aldrich, 1:5,000) antibodies. For immunoprecipitation experiments, mouse monoclonal anti-HA agarose beads (HA-7, A2095, 119M4756V, Sigma-Aldrich), were used and NS4B–HA was detected using a purified mouse anti-HA.11 epitope tag (16B12, 901502, B276381, BioLegend, 1:1,000). NS4B- and NS3-specific bands were visualized using custom-generated rabbit polyclonal antibodies directed against NS3 (1:2,000) or NS4B (1:1,000)⁵⁰. Anti-flavivirus group monoclonal antigen antibodies (D1-4G2-4-15, MAB10216, 2441960, Millipore/Merck) were used in the survival study in AG129 mice (injected with 100 µl, 1:50) before DENV infection to induce ADE and therefore accelerate death and in the 50% tissue culture infective dose (TCID₅₀) assay, in which virus production was determined using custom-developed colorimetric detection of viral envelope protein using the anti-flavivirus group antigen monoclonal antibody as primary antibody. The 4G2 mouse monoclonal anti-DENV capture antibody was also used to coat 96-well plastic flat-bottom plates for the IgM and IgG ELISA. Secondary antibodies used in this assay were peroxidase-labelled goat anti-monkey lgM (KPL, 074-11-031) or lgG (Sigma-Aldrich, SAB3700766).

Viruses

A clonal stock of the strain DENV-2/16681 was produced by transfection of in vitro transcribed RNA of plasmid pFK-DVs into Huh-7 cells³. DENV-2/16681/eGFP, carrying enhanced green fluorescent protein (eGFP) at the amino terminus of the capsid protein, was produced through the transfection of in vitro transcribed RNA of plasmid pFK-DV-G2A into Huh-7 cells^3,53.

A total of 20 DENV strains, representing the available genotypes within the four DENV serotypes, were used in the in vitro genotype panel, from which 19 viruses were described previously in ref. ³ and one additional virus, DENV-1/45AZ5 (ref. ⁵⁴), was added.

For antiviral assays using Vero cells, in vitro resistance-selection experiments and in vivo efficacy studies in mice, the DENV-2 Rega Laboratory strain (referred to as DENV-2 RL) was used (GenBank: MW741553; provided by V. Deubel). For in vivo studies in mice, high-titre stocks were generated by propagating the virus in C6/36 mosquito cells and subsequently concentrating, either by ultracentrifugation or tangential flow filtration using tangential flow filtration capsules (Minimate TFF; Pall Life Sciences), according to the manufacturer’s protocol. Infectious virus titres (PFU per ml) were determined by performing plaque assays on baby hamster kidney cells (BHK-21, ATCC (CCL-10)).

Four non-dengue flaviviruses were used: ZIKV (H/PF/2013, French Polynesia, GenBank: KJ776791), JEV (CNS769-Laos 2009, Laos, GenBank: KC196115), WNV (R94224, CDC human brain 29 September 2008, Wisconsin, GenBank: MF004388) and yellow fever virus (YFV; 88-99, Bolivia, GenBank: MF004382). All other viruses, not belonging to the Flavivirus genus, are detailed in the Supplementary Methods and Supplementary Table 2.

To test the efficacy against the DENV serotypes in the AG129 mouse model (129/Sv mice deficient in both IFNα/β and IFNγ receptors) with mortality read-out, DENV-1/West Pac (genotype IV), DENV-3/C0360/94 (genotype II) or DENV-4/703/4 (genotype II) was used. The virus strains and culture procedure to generate the virus stocks have been described previously⁵⁵. For proof-of-concept studies in rhesus macaques, the virus strains DENV-1/45AZ5 (ref. ⁵⁴) and DENV-2/16681 were used. DENV-2/16681 stocks were generated by propagating the virus in C6/36 mosquito cells. Infectious virus titres were determined on Vero cells (ATCC, CCL-81) using standard procedures. DENV-1/45AZ5 was grown in fetal rhesus lung cells and the virus was not further concentrated⁵⁴.

Antiviral assays

The antiviral activity of JNJ-1802 against DENV-2/16681/eGFP was determined in a phenotypic antiviral assay with eGFP readout as a measure for viral replication. In parallel, the toxicity was measured using an ATPLite cell viability luminescence assay (PerkinElmer). The assays were performed in three different cell types (Vero, Huh-7 and THP-1/DC-SIGN) as described in ref. ³ to exclude cell-specific activity of the compound.

The antiviral activity of JNJ-1802 against the DENV-2 RL strain was determined in Vero and C6/36 cells by measuring the levels of viral RNA using RT–qPCR as previously described³. A potential toxic effect on host cells was tested in parallel, while omitting virus infection, using the MTS/PMS method (Promega) for Vero cells, or by high-content imaging for C6/36 cells, as described previously³. Antiviral activity of JNJ-1802 against the DENV genotype panel (Supplementary Table 1) was assessed by measuring viral RNA using RT–qPCR as described previously^3,23. RT–qPCR data were analysed using the QuantStudio 12K Flex software (v.1.2.3) or SDS v.1.2 Applied Biosystems software. Inhibition values for antiviral molecules were plotted using KaleidaGraph plotting software (v.4.03, Synergy Software). The antiviral activity of JNJ-1802 against DENV-1/45AZ5 was tested in Vero cells by measuring viral RNA using RT–qPCR as described previously⁵⁶. Antiviral activity/toxicity assays of JNJ-1802 for all other viruses, not belonging to the flaviviruses, were performed using either reporter systems, cell viability, RT–qPCR or plaque reduction assays as readout, as further specified in the Supplementary Information and Supplementary Table 2.

DENV-2 in vitro resistance selection experiments were performed in Vero cells with gradually increasing concentrations of JNJ-1802 (ref. ³). Whole-genome sequencing was performed in-house on DENV variants collected at every fifth passage and at the end of the experiments (that is, passage 42 for experiment A and passage 50 for experiment B) as described previously³.

Transient mutant replicon assays

A panel of WT and mutant subgenomic DENV reporter replicons (sgDVs-R2A) was used to determine the compound resistance imposed by each of these mutations³.

Immunoprecipitation experiments

HA-immunoprecipitation experiments were performed as previously described³ with minor modifications. In brief, for dose–response assays, cells were transfected for 4 h, followed by addition of fresh DMEM supplemented with JNJ-1802 or an equivalent amount of DMSO and cells were collected 18 h after treatment. For the time-of-addition experiments, cells were drug treated either 4 h or 24 h after transfection and, in either case, collected 8 h after treatment. Immunoprecipitation was conducted using mouse anti-HA monoclonal agarose beads (Sigma-Aldrich). Cell lysates and captured protein complexes were analysed by western blotting using a home-made polyclonal antibody for the detection of NS3 and both a home-made polyclonal anti-NS4B-specific antibody and a monoclonal anti-HA-specific antibody for detecting NS4B-containing protein species (or HA-tagged modifications thereof). Intensities of NS3- and NS4B–HA-specific bands (the latter determined by the anti-HA-specific antibody) were quantified using ImageJ (v.2.1.0/1.53j; Wayne Rasband and contributors, National Institutes of Health, USA). To ensure equal weighting of individual experiments, signal intensities of NS3 and NS4B–HA species (NS4A-2K-NS4B–HA, 2K-NS4B–HA and NS4B–HA) were first divided by the corresponding sum across all samples (both in lysates and after pull-down). Ratios of NS3 to HA-tagged NS4B-containing species were then formed and normalized to the mean ratio of the DMSO-treated control. For dose–response assays, statistical analysis was performed using repeated-measures one-way ANOVA with subsequent Dunnett’s multiple-comparisons test. In the time-of-addition experiments, statistical analysis was performed using paired two-sided t-tests.

DENV-2 infection studies in mice

Breeding couples of AG129 mice were purchased from Marshall BioResources and bred in-house. The specific pathogen-free status of the mice was regularly checked at the KU Leuven animal facility. Mice were housed in individually ventilated cages (maximum of five mice per cage, type GM500 Sealsafe Plus, Tecniplast) at 21 °C, 55% humidity under a 12 h–12 h light–dark cycle. The mice were provided with food and water ad libitum as well as with cardboard play tunnels and cotton as extra bedding material. Allocation to experimental groups was performed randomly. Housing conditions and experimental procedures were approved by the ethics committee of KU Leuven (licence P169/2011 and P047/2017) following institutional guidelines approved by the Federation of European Laboratory Animal Science Associations.

The efficacy of JNJ-1802 on viraemia, viral kinetics and virus-induced disease (survival) was evaluated in DENV-2 infection models in AG129 mice (Supplementary Table 3) as described previously³.

In brief, in day 3 viraemia studies, DENV female mice (7–11 weeks old) were challenged intraperitoneally (i.p.) with 10⁶ PFU DENV-2 RL strain. Mice were treated b.i.d. by oral gavage for 3 consecutive days with either vehicle (PEG400:water (80:20); n = 24) or various doses of JNJ-1802 (60, 20, 6, 2, 0.6 or 0.2 mg per kg per day; n = 8 (60 mg per kg per day dosing group) and n = 16 (all other dosing groups)), with the first administration 1 h before DENV challenge. In a separate viraemia study, in which mice were orally treated for 3 days with 30, 3 or 0.3 mg per kg per day JNJ-1802 (n = 8 for each group), the treatment was administered once daily. On day 3 after infection, the mice were euthanized, and blood was collected and stored at −80 °C until further use. The effect of the compound on viral RNA levels in the blood on various days after infection was monitored in an in vivo kinetics study. AG129 female mice (aged 7–11 weeks, n = 16 per group) were inoculated i.p. with 10² PFU DENV-2 RL strain. Mice were treated b.i.d. through oral gavage with vehicle or JNJ-1802 using three different doses: 20, 2 or 0.2 mg per kg per day. In the kinetics studies, treatment was initiated 1 h before DENV infection and continued for 6 consecutive days. Each group was subdivided into two smaller groups (A and B; n = 8 each), from which blood was collected on alternating days: on day 1, 3 and 5 (A groups) or on day 2, 4 and 6 (B groups). On day 8 and day 11 after infection, the mice from the A and B groups, respectively, were euthanized, and blood was collected and stored at −80 °C until further use. The survival study was performed as described previously³ with b.i.d. dosing of AG129 mice (aged 7–11 weeks, female, n = 10 per group) at 20, 6, 2 or 0.6 mg per kg per day (start of treatment 1 h before DENV challenge).

In delayed-treatment studies (therapeutic setting), AG129 female mice (aged 7–11 weeks, n = 10 per group) were inoculated i.p. with 10² PFU DENV-2 RL strain. Treatment with JNJ-1802 (60 mg per kg per day, b.i.d.) was initiated on day 4 or 5 after infection and continued for 6 days. Mice that were treated with vehicle or JNJ-1802 on the day of infection (that is, 1 h before infection) were included as controls. Each group was subdivided into two smaller groups (A and B; n = 5 per group), from which blood was collected on alternating days: on day 3, 5, 7, and 9 (A groups) or on day 4, 6, 8 and 10 (B groups). On day 12 and day 14 after infection, mice from the A and B groups, respectively, were euthanized, and blood was collected and stored at −80 °C until further use.

Viral RNA isolation from plasma was performed using the NucleoSpin RNA virus kit (Macherey-Nagel) according to the manufacturer’s instructions. Details of the primers, the probe and the RT–qPCR were described previously^57,58. The LLOQ of this RT–qPCR assay was defined as 3.8 log₁₀ copies per ml, which corresponds to a median cycle threshold value of 36.3.

DENV-1, DENV-3 and DENV-4 infection in mice

Studies to evaluate the in vivo efficacy of JNJ-1802 against DENV-1, DENV-3 and DENV-4 were undertaken at the University of Texas Medical Branch (UTMB) in laboratories managed by the Animal Resources Center. The UTMB is an Association for the Assessment and Accreditation of Laboratory and Care (AAALAC) International accredited facility. All of the procedures were reviewed and approved by the UTMB Institutional Animal Care and Use Committee. The studies were carried out in strict compliance with the recommendations of the Guide for the Care and Use of Laboratory animals published by the National Research Council. Male and female AG129 mice (n = 10 per group; aged 6–8 weeks) were housed at a maximum of 5 mice per cage in individually ventilated cages with food and water provided ad libitum. Allocation to experimental groups was performed randomly. For the in vivo efficacy of JNJ-1802 against DENV-1, DENV-3 and DENV-4, AG129 mice (aged 6–8 weeks) were treated b.i.d. by oral gavage with JNJ-1802 (60 or 6 mg per kg per day in 0.1 ml per dose, (n = 10 each)) or vehicle (80% PEG400 + 20% H₂O in 0.1 ml) (n = 10) for 5 consecutive days. Approximately 1 h after the first treatment, all of the animals were challenged by i.p. injection with 10⁷ PFU DENV-1/West Pac, DENV-3/C0360/94 or DENV-4/703/4. Pretreatment with anti-flavivirus antibody 4G2 to increase disease severity by induction of ADE was not required as infection with DENV-1/West Pac, DENV-3/C0360/94 or DENV-4/703/4 in our models leads to severe disease (requiring euthanasia after >20% loss of body weight). Each treatment group was subdivided into two smaller groups (groups A and B; n = 5 each). Blood was collected on day 3 after infection from group A to measure viral RNA; samples were collected by retro-orbital bleed on day 3 after infection around 30 min before the morning treatment. In parallel, animals (groups A and B) were followed for lethality throughout the experiment. Viral RNA isolation from plasma and RT–qPCR analysis was performed using the methods described previously⁵⁵.

DENV-2 study in NHPs

Pre-exposure prophylaxis to DENV-2 was assessed in 12 rhesus macaques (M. mulatta) at the Biomedical Primate Research Centre (BPRC), Rijswijk, the Netherlands. BPRC is an AAALAC International accredited facility, and the research protocol was approved by appropriate national authorities (CCD, Central Committee for Animal Experiments; licence number AVD5020020172884) and by the institutional Animal Welfare Body (AWB).

The animals were selected from the experimental stock from the self-sustainable BPRC colony. Before being placed on the protocol, all animals tested negative for antibodies to DENV serotypes 1, 2, 3 and 4, WNV and ZIKV. All animals were healthy, adult, male and female, Indian-origin rhesus monkeys with a minimum age of 5 years and a minimum body weight of 7 kg. During the experiment, the animals were housed in pairs with a socially compatible cage mate under biosafety level 3 (BSL3) conditions. The animals were offered a daily diet consisting of monkey food pellets, fruit and bread. Enrichment (toys, extra food) was offered daily. Drinking water was available ad libitum through an automatic watering system. The animals were randomly allocated to treatment groups according to a randomized block design on the basis of the sex and body weight of the animals. To assess the efficacy of JNJ-1802 in a prophylactic setting, rhesus macaques received daily vehicle (100% PEG400, n = 3), or compound JNJ-1802 at a dose of 0.01 mg per kg per day (n = 3), 0.18 mg per kg per day (n = 3) or 3 mg per kg per day (n = 3) dissolved in 100% PEG400 through oral gavage, starting from 1 day before experimental infection until 10 days after infection. On day 0, the animals were exposed through intradermal inoculation in the upper back with 100 TCID₅₀ of strain DENV-2/16681. In the DENV-1 NHP study that was conducted first, infection with DENV-1 was performed 3 days after oral administration of JNJ-1802. As a rapid increase in drug levels was obtained 12 h after dosing and high concentration persisted up to the end of the dosing period, animals in the DENV-2 NHP study were infected already one day after the start of oral administration with JNJ-1802. The animals were monitored daily during the study period for general behaviour until day 28. Blood samples were taken at regular timepoints for both virological assessments and determination of plasma compound concentrations. Virological assessments included the quantification of DENV-2 RNA using RT–qPCR, the detection of DENV NS1 antigen by ELISA, the quantification of infectious DENV-2 by TCID₅₀ assay in Vero cells and the detection of anti-DENV IgM/IgG by ELISA. Details of the assessment of haematological parameters are provided in the Supplementary Information.

Quantification of DENV RNA using RT–qPCR

Plasma samples from ethylenediaminetetraacetic acid (EDTA)-treated blood of macaques were checked for the presence of DENV-2 RNA using RT–qPCR as described in ref. ⁵⁹. Viral RNA was isolated from 140 µl plasma using the QIAamp Viral RNA Mini Kit (Qiagen Benelux) according to the manufacturer’s instructions. Next, 10 µl taken from the 60 µl eluate was used for cDNA synthesis and PCR amplification. This was performed using the Brilliant II QRT-PCR Core Reagent Kit, 1-Step kit (Agilent), according to the manufacturer’s instructions. The amount of viral RNA in the plasma was evaluated on a Bio-Rad CFX Connectreal-time system. A multiplication factor of 42.86 was used to calculate the amount of viral RNA per ml of plasma. The amount of viral RNA per ml plasma was determined in three independent assays. The LLOQ was 30 genomic copies per reaction.

DENV IgM and IgG ELISA

Anti-DENV IgM and IgG were assessed in the serum of DENV-2-infected macaques using the Dengue Virus IgM Capture DxSelect ELISA test (EL1500M; Focus Diagnostics) and the Dengue Virus IgG DxSelect ELISA test (EL1500G; Focus Diagnostics) according to the manufacturer’s instructions in two independent experiments⁶⁰. Both are qualitative assays for the detection of human IgM and IgG antibodies to DENV expressed as an index value relative to the cut-off calibrator as described in the instruction manual of the assay. To calculate index values, specimen optical density (OD) values (corrected for blank readings) were divided by the mean of the corrected cut-off calibrator absorbance values. As no specific assays were available to detect macaque antibodies in response to DENV infection, we made use of the broad cross-reactivity between the macaque antibodies and the detecting conjugates to evaluate the anti-DENV humoral immune responses in macaques.

NS1 ELISA

Detection of NS1 antigen in serum of DENV-2 infected macaques was performed using the Alere Panbio Dengue Early ELISA (01PE40; Alere) and performed according to the manufacturers’ instructions. DENV NS1 was detected in two independent assays.

Infectious virus quantification

Quantitative virus isolation was performed using the TCID₅₀ assay. To determine the TCID₅₀ per ml, plasma samples from DENV-2-infected macaques were titrated on Vero cells. In brief, the samples were 1:1 serially diluted in culture medium and added in triplicate to a monolayer of Vero cells in microtitre plates. Undiluted virus and medium alone were used as positive and negative controls, respectively. After incubation for 6 days under 5% CO₂ and at 37 °C, the medium was discarded and the cells were fixed with cold acetone. Next, virus production was determined using a custom-developed colorimetric detection of viral envelope protein using an anti-flavivirus group antigen monoclonal antibody (D1-4G2-4-15) as the primary antibody. The plates were measured at an OD of 490 nm on the Bio-Rad iMark microplate reader. A well was scored positive when OD_{sample well} > OD_{negative control} + 0.75 × (OD_{positive control} − OD_{negative control}). The 50% endpoint was calculated using the Spearman & Kärber algorithm^61,62.

The colorimetric screening assay was validated by a direct comparison with a similar test but based on microscopic scoring of infected wells.

DENV-1 study in NHPs

Antiviral activity of JNJ-1802 was also assessed in a prophylactic setting in rhesus macaques infected with the DENV-1/45AZ5 virus strain. The study was approved by the Walter Reed Army Institute of Research Institutional Animal Care and Use Committee (WRAIR IACUC). Research was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations pertaining to animals, and the work was performed in accordance with the principles stated in the Guide for the Care and Use of Laboratory Animals, the National Research Council. WRAIR is fully accredited by AAALAC International. The animals were US-colony bred and procured from Covance Research Products. Before being placed on the protocol, all of the animals tested negative for antibodies to DENV serotypes 1–4, WNV and ZIKV by neutralization assays performed at WRAIR. The animals also tested negative for simian retroviruses, simian immunodeficiency virus and simian T cell leukaemia virus with testing performed by the vendor before shipment. The animals were single-housed under a 12 h–12 h light–dark cycle and with 10–15 room air changes per hour. The animals were fed Old World Primate Chow 5038 (Quality Lab Products) b.i.d., fresh fruit at least three times a week and water ad libitum. Environmental enrichment was provided in the form of cage complexities, food treats, opportunities to forage, a rotation of several toys and puzzles, periodic access to an activity cage that permits climbing, jumping and swinging, and alternating days of television and music. Animals were anaesthetized by intramuscular injection with a combination of ketamine (5 mg per kg) and dexmedetomidine (0.01 mg per kg) before all procedures, and anaesthesia was reversed by intramuscular injection of atipamezole (0.01 mg per kg). The studies performed under this protocol were non-terminal. Twelve healthy, adult, Indian-strain rhesus macaques, males and females, at least 5 kg in weight and aged 4–10 years were selected for the study. The animals were allocated into two experimental groups (n = 6 per group) in no particular order, while attempting to balance sex, ages and weights. Group 1 received 6 mg per kg per day JNJ-1802 and group 2 received vehicle (100% PEG400). Oral administration of compound JNJ-1802 started 3 days before experimental infection (day −3) and continued with daily administrations until day 10 after infection. On study day 0, which is 3 days after the start of treatment, the challenge virus (0.5 ml of DENV-1/45AZ5; 1.2 × 10⁵ PFU per ml) was administered subcutaneously into the upper arm. Whole-blood samples were collected on days −28, −3, −2, 0–11, 14 and 29 (Fig. 4a). In addition to veterinary and husbandry care, the animals were observed visually at least once daily throughout the study by research personnel.

Quantification of DENV RNA using RT–qPCR

Serum samples from macaques were checked for the presence of DENV-1 RNA. Viral RNA for RT–qPCR was extracted from 100–200 µl of serum using the QIAcube instrumentation and the QIAGEN Viral RNA extraction kits. The primers and fluorogenic probes used in the RT–qPCR assays have been previously described^63,64. DENV-1 RNA was detected in a 25 µl reaction containing 5 µl of extracted RNA sample, 10 pmol of dengue-specific primers and 5 pmol dengue probe in the reaction mixture from Superscript Ill Platinum One-Step Quantitative RT-PCR Kit (Invitrogen, Thermo Fisher Scientific), consisting of 0.5 µl Superscript III RT/Platinum mix, 12.5 µl 2× reaction mix and 0.05 µl ROX reference dye. The one-step RT–qPCR comprises a 30 min RT step at 50 °C and 2 min of Taq polymerase activation at 94 °C, followed by 40 cycles of PCR at 94 °C for 5 s and 53 °C for 35 s performed in an ABI 7500DX Fast Real Time PCR system. Each assay included positive, negative and non-template controls. The RNA copy number was calculated from a standard curve of tenfold serial dilutions made from RNA in vitro transcripts consisting of DENV-1 with known copy numbers. The LLOQ was 100 genomic copies per reaction.

DENV IgM and IgG ELISA

The IgM/IgG ELISA assays were performed on sera from DENV-1 infected macaques. For IgM and IgG ELISA, 96-well plastic flat-bottom plates were coated with 4G2 mouse monoclonal anti-DENV capture antibody. After blocking, the plates were incubated with partially purified DENV-1 (WP-74 strain) virus antigen, washed and then incubated with serial dilutions of NHP serum (100-fold to 218,700-fold dilutions). After washing, the plates were incubated with the appropriate secondary antibody—either peroxidase-labelled goat anti-monkey IgM or IgG in 50% glycerol. After washing, the plates were developed using the TMB Microwell Peroxidase Substrate System (KPL, 50-76-00). Once the reaction was stopped, using an acidic solution, the enzymatic turnover of the substrate was determined by OD measurement at 450 nm. The samples were assayed in duplicate. The background was calculated as the average of all day 0 (pre-infection) sera for each dilution.

Pharmacokinetics studies

The pharmacokinetics profile of JNJ-1802 was evaluated in a separate experiment in fed male CD-1 mice (n = 3 per group, 27–33 g body weight, aged around 6–8 weeks; Charles River Laboratories) at Janssen Pharmaceutica NV. The housing conditions and experimental procedures were approved by the ethics committee on animal experiments of Janssen Pharmaceutica NV (license number LA1100119). Janssen Pharmaceutica NV has full AAALAC accreditation. The mice were intravenously injected into the tail vein with 2.5 mg per kg of JNJ-1802, which was formulated as an 0.5 mg ml⁻¹ solution in 70% PEG400/30% H₂O, and blood samples were collected (in EDTA-containing microcentrifuge tubes) from the saphenous vein at 0.12, 0.33, 1, 2, 4, 7 and 24 h after dosing. Moreover, JNJ-1802 was administered by oral gavage at 1, 3, 10 and 30 mg per kg, formulated as a solution in 80% PEG400/20% H₂O, and the blood samples were collected from the saphenous vein at 0.5, 1, 2, 4, 7 and 24 h after dosing. The blood samples were immediately centrifuged at 4 °C and the plasma was stored at −20 °C. Compound concentrations in the plasma samples were determined using the API 5500 LC–MS/MS system mass spectrometer (Applied Biosystems). Individual plasma concentration–time profiles were processed for a non-compartmental pharmacokinetics analysis using Phoenix WinNonlin v.6.3. (Certara).

The pharmacokinetics profile of JNJ-1802 was evaluated in female monkeys after single intravenous administration and in DENV-infected monkeys after repeated oral administration. After single intravenous administration, the monkey blood was collected at the BPRC just before dosing, and at 0.05, 0.15, 0.5, 4, 6, 8, 24, 48 and 264 h after dosing. After oral administration, the monkey blood was collected at the BPRC on the first and last day of dosing just before dosing and at 1, 2, 4, 6, 8 and 24 h after dosing that day. Moreover, blood samples were taken on days 3, 6 and 7 after infection just before dosing. For the study at WRAIR, the monkey blood was collected just before dosing and at 1, 4, 8 and 24 h after the first day of dosing and before dosing and at 2, 5, 8, 24 and 96 h after the last day of dosing. Furthermore, blood samples were taken on days 0, 2, 5 and 8 after infection just before dosing. Blood was drawn from the femoral vein (or a peripheral vein, for example, saphenous or cephalic) The volume of blood collected did not exceed 7.5% of an animal’s total blood volume over a 7 day period or 10% over a 2 week period. A volume of 1 ml blood was collected into a blood collection tube coated with EDTA as anti-coagulant. Within 1 h after collection, the blood samples were centrifuged in a precooled centrifuge for 10 min at 1,500g (4 °C). Within 2 h after blood collection, the plasma samples were stored at −80 °C until shipment. Individual plasma concentration–time profiles were processed for a non-compartmental pharmacokinetics analysis using Phoenix WinNonlin v.6.3. (Certara).

Statistical analysis for in vivo studies

Statistical power calculations considered the number of mice required to detect a significant reduction in viraemia compared with vehicle-treated controls. With groups of n = 8, a reduction of at least 0.8 log₁₀ in viral RNA can be detected according to the independent t-test (with α = 0.05, power = 80% and a s.d. value of 0.5). Moreover, statistical calculations considered the number of mice that were required to detect a significant improvement in survival compared with vehicle-treated controls. With groups of n = 11, a minimal survival rate of 60% for treated mice versus 0% in the untreated, infected control group can be demonstrated according to the Fisher’s exact test (with α = 0.05 and power = 80%). The experiments were not randomized, and investigators were not blinded to allocation during experiments and outcome assessment.

To assess the effect of JNJ-1802 treatment on viral load in DENV-2-infected mice for each treatment group compared with the vehicle-treated mice (viraemia studies), ordinary one-way ANOVA was used with Dunnett’s test to correct for multiple comparisons (experiment with q.d. treatment) and the pooled data of three independent studies were analysed using two-way ANOVA with Dunnett’s test to correct for multiple comparisons (experiment with b.i.d. treatment).

To assess the effect of JNJ-1802 treatment on viral load in DENV-1-, DENV-3- or DENV-4-infected mice for each treatment group compared with the vehicle-treated mice (viraemia studies), ordinary one-way ANOVA was applied with Dunnett’s test to correct for multiple comparisons (DENV-1), and Kruskal–Wallis tests with Dunn’s multiple-comparison test (DENV-3 and DENV-4). For the viral kinetics studies and the delayed-treatment studies, a batch approach was applied to calculate the viral load AUC using the PΚ R package⁶⁵. This package estimates a mean AUC value for settings in which the animals are measured at varying timepoints within a treatment group. Within each experiment, the mean AUC value and 95% CIs were determined for each group. The AUC was calculated using the LOD (2.6 log₁₀ copies per ml) as the lowest limit. In case the CI of a compound-treated group did not overlap with that of another group (vehicle or treatment), the groups were considered to be statistically different. In the delayed-treatment studies, the total viral load AUC for each of the compound-treated groups was calculated and compared with that of the vehicle-treated group. Fisher’s exact tests were used to determine whether the survival rate in DENV-2-infected mice on day 25 for each compound treatment group differed significantly from that of the vehicle group. P values were adjusted using the Bonferroni multiple-comparison correction method. Note that P values lower than 0.0001 were rounded to 0.0001. For survival studies of DENV-1-, DENV-3- or DENV-4-infected mice, Fisher’s exact tests were used on day 6. P values were adjusted using the Bonferroni multiple-comparison correction method.

To assess the in vivo efficacy of JNJ-1802 against DENV-2 in NHPs, no formal sample size calculation was performed. A Bayesian nonlinear inhibitory sigmoid E_max model was used to quantify the dose–response relationship between JNJ-1802 and the induced peak viral load DENV-2 in monkeys. The model estimates the mean peak viral load under placebo (dose zero) and the nonlinear decrease in function of increasing dose of JNJ-1802 under the assumptions of 100% maximum drug inhibition at infinite high dose levels. The animal caretakers who performed all animal handling and blood collections were not blinded to the DENV-2 NHP studies. However, the samples were processed by laboratory technicians who were not involved in the animal treatment/manipulation and who received the blood samples in coded tubes for further analysis.

To assess the antiviral activity of JNJ-1802 in rhesus macaques infected with the DENV-1/45AZ5 virus strain, the sample size was calculated using one-sided Fisher exact tests and an α = 0.05 targeting a power of at least 80%. This sample size was sufficient to detect a significant difference in the number of infected animals in each of the groups as low as 83%. Although the DENV-3 NHP studies were not formally blinded, the veterinary staff were not told which material (JNJ-1802 or placebo) they were delivering to a specific animal. Sample processing and conducting of assays was performed by laboratory technicians not involved in the treatment/manipulation of the animals. Blood samples were labelled with animal ID numbers and processed in a non-specific manner simultaneously. All assays were performed on samples of all animals by laboratory technicians; the samples and raw data were not labelled with their treatment group.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.