Fig 1.
High-content imaging antiviral assay based on DENV-2/16681-GFP in A549 cells.
A Experimental set up of the used HCI fluorescence platform. Representative images of DENV-2/16681-GFP infective A549 cells. Images from CV7000 Yokagawa showing eGFP signal coming from the DENV-2/16681-GFP infection (eGFP, green), Hoechst staining (nuclei, blue), MitoTracker orange staining (mitochondria, Orange) and Cell-Mask Deep red staining (cytoplasm and nucleus, Red). Scale bar: 50 µM. B High level flow of the screening funnel from the primary screen to the identification of hits with orthoflavivirus broad-spectrum potential. C Morphological profiles based on 30 features for each compound, here shown for two reference compounds, Brequinar [49] (a selective inhibitor of the enzyme dihydroorotate dehydrogenase) and compound 24 [50] (a DENV NS4B inhibitor). D Results of the hit deprioritization using SIM2NIC. The maximum SIM2NIC score over all concentration at which a compound achieved ≥ 50% virus inhibition and retained cell count ≥ 30% (relative to infected control (0-line) was computed and plotted. Non-infected control cells had a SIM2NIC value of 0.99 (median of individual control wells). The SIM2NIC values obtained for the reference compounds Brequinar and compound 24 were 0.20 and 0.98, respectively. E T-SNE clustering of full feature analysis of hits from the confirmation screen identifies clusters of compounds with similar phenotypes. The phenotypic cluster of hits with a SIM2NIC > 0.8 closely correlates with the non-infected control (NIC; star symbol). The reference compounds used to select the 30 features needed for calculation of the SIM2NIC value are presented by large circles. Every small circle, represent a test compound.
Table 1.
Antiviral activity of JNJ-3644 against a set of different orthoflaviviruses.
Fig 2.
Orthoflavivirus antiviral activity evaluation of the identified hit series.
A Structure of compound JNJ-3644, JNJ-1953 and JNJ-4840. B Antiviral activity (% Inhibition eGFP expression) of JNJ-3644 against DENV-2/16681-eGFP in Vero cells, Huh7 hepatoma cells and human monocytic leukemia THP1 cells expressing the DC-SIGN receptor. C Antiviral activity (% Inhibition eGFP expression) of JNJ-3644, JNJ-1953 and JNJ-4840 against DENV-2/16681 in Vero cells. D Antiviral activity (plaque forming units per mL) of JNJ-1953 against DENV-2/Eden3295 (right Y-axis) and DENV-2/38865Y10 (left y-axis) in Huh7 cells. Data represent mean values ± SD from at least two independent experiments.
Table 2.
Antiviral activity of JNJ-4840 and JNJ-1953 against different orthoflaviviruses.
Table 3.
ADME-Tox profile of compounds JNJ-3644, JNJ-1953 and JNJ-4840.
Fig 3.
Time of drug addition studies of JNJ-1953.
Huh7 cells were either treated with JNJ-1953 (orange symbol/line) 2 h prior to DENV-2/EDEN3295 infection (pre-treatment; Pre-T), infected with DENV-2/EDEN3295 in the presence of the compound (co-treatment; Co-T), or post-treated with the compound after 1 h DENV-2/EDEN3295 infection at the indicated hours post-infection (post-treatment; Post-T) as shown in the schematic. At 24 hours post-infection, plaque quantification of the culture supernatants was assessed. NITD008 was included as reference compound [53] (gray symbol/line). Data is presented as line graph showing the average PFU/mL with standard deviation of the treated samples compared to the untreated infected control (gray dotted line) obtained from two independent experiments with duplicates.
Table 4.
Antiviral activity of JNJ-1953 against WT and mutant DENV-2 subgenomic constructs and DENV-2/EDEN3295.
Fig 4.
In vitro resistance selection experiments identify NS2A protein as target.
A Allele frequency distributions of the polyprotein of two passaged (p18 and p28) DENV-2/16681 viruses which were exposed to increasing drug concentrations of JNJ-1953, highlighting mutations within NS2A protein (pink region). B Sequence alignment of a part of the NS2A proteins among DENV 4 serotypes, ZIKV, WNV, JEV, and YFV. The consensus and conservation percentage are indicated below each amino acid. C Effect of the single, double and triple NS2A resistance mutations (F18L, E21G, A32V, E21G/A32V and F18L/E21G/A32V) on the replication fitness of the subgenomic DENV-2/16681 reporter replicon. Data shown represents two independent experiments with two biological replicates each (total n = 4). Values were Log2-transformed and comparisons between WT and each mutant were performed by one-way ANOVA followed by Dunnett’s multiple comparisons test. D Level of compound resistance to JNJ-1953 imposed by NS2A resistance mutations (NS2AF18L, NS2AE21G, NS2AA32V, NS2AE21G/A32V and NS2AF18L/E21G/A32V). E Intracellular viral RNA replication of WT and the NS2A mutant (NS2AE21G (green), NS2AA32V (purple) and NS2AE21G/A32V (red)) DENV-2/EDEN3295 viruses determined by RT-qPCR. F Infectious-virus titers of WT and mutant DENV-2/EDEN3295 viruses (NS2AE21G (green), NS2AA32V (purple) and NS2AE21G/A32V (red)) in the culture supernatants of electroporated cells as determined by standard BHK-21 plaque assay. Data presented in E and F are from two independent experiments with technical readings. The difference in the kinetics of the (E) intracellular viral RNA synthesis and (F) virus titers between WT and the various mutants are compared by two-way ANOVA with Geisser-Greenhouse correction and statistical significance are indicated (* - p < 0.05; ** - p < 0.01; *** - p < 0.001; **** - p < 0.0001).
Fig 5.
JNJ-1953 reduces viral RNA synthesis and impairs RNA packaging.
A Schematic. B,C Delayed time of drug addition study of JNJ-1953 on DENV-2/EDEN3295 replication. Huh7 cells were infected with DENV-2/EDEN3295 at MOI 1 for 1 h and treated with 5 µM of the respective compounds at 6 h post-infection (hpi). Samples were harvested at the indicated timepoints and subjected to (B) intracellular viral RNA quantification by RT-qPCR and (C) infectious virus production determined by plaque assay. Data presented in B and C are representative results from two independent experiments with technical readings. The difference in the kinetics of the (B) intracellular viral RNA synthesis and (C) virus titers between virus control (VC) and JNJ-1953 treatment are compared by two-way ANOVA with Geisser-Greenhouse correction. D,E Effect of JNJ-1953 on virion assembly. Huh7 cells were infected with DENV-2/Eden3295 at MOI 1 followed by treatment with 5 µM of indicated inhibitors. Cell lysates were harvested 24 hours post infection and subjected to treatment with RNaseA/T1 to remove any unpackaged RNA. (D) Intracellular viral RNA quantification of RNaseA/T1 treated and non-treated samples by qRT-PCR. (E) Tabulated ratio of the non-RNase treated to RNase treated which provide an indication of the proportion of viral RNA packaged into virions upon treatment with the various inhibitors. Data in D and E are presented as bar graphs showing mean with standard deviation from 2 independent experiments (n = 4). The difference in intracellular viral RNA between DMSO-treated virus control (VC) and JNJ-1953 treatment matched by RNase treatment is compared using two-way ANOVA with Geisser-Greenhouse correction. The difference in the non-RNase to RNase-treated ratio between JNJ-1953 and VC was compared by unpaired Student’s t-test with df = 6. Statistical significances are indicated (** - p < 0.01; *** - p < 0.001; **** - p < 0.0001).
Fig 6.
JNJ-1953 destabilizes prM binding to NS2A protein.
A Schematic showing the construct design of DENV-2 NS2A and prM. The constructs design followed Xie et al. [2] B HEK293T cells were transfected with 5 µg each of Myc-prM and Flag-WT or NS2AE21G/A32V, followed by treatment with the active JNJ-1953 compound or with its inactive enantiomer, JNJ-2005, at 6 hours post-transfection. Cell lysates were harvested at 44 hours post-transfection and subjected to co-immunoprecipitation using Myc or IgG control antibody. Western blots showing the detection of Flag-NS2A WT and Flag-NS2AE21G/A32V immunoprecipitated with Myc-prM in the absence and presence of JNJ-1953 or JNJ-2005. GAPDH was used as a loading control. C Densitometric analysis of the band intensities of WT or NS2AE21G/A32V normalized to prM for the co-immunoprecipitated samples upon treatment with JNJ-1953 or JNJ-2005 compared to untreated. Data is presented as bar graphs showing mean with standard deviation from 4 independent experiments (n = 4) and differences between WT and NS2AE21G/A32V is compared by unpaired Student’s t-test with df = 6 (* - p < 0.05).
Table 5.
In vivo pharmacokinetic properties of oral administration of JNJ-1953.