CpG oligodeoxynucleotides and pan-serotype inhibitors control neurotropic dengue infection in novel immune competent neonatal mouse model

Abstract

Dengue virus (DENV) is a growing global public health threat. The lack of symptomatic immune-competent animal models for dengue has hindered the screening and development of effective therapeutics that can be used to control dengue virus replication and thereby control the progression to severe dengue disease. To address this, we established an infection model in neonatal C57BL/6 mice and showed that a systemic Dengue challenge leads to ataxia, seizures, paralysis, and death within 15 days. The virus was found predominantly in the eye and brain where DENV infects neurons but not astrocytes and causes extensive infiltration of macrophages and microglia activation. The response to infection included upregulation of multiple genes linked to interferons (Ifna, Ifnb, Ifng, Irf7, Irf8, Mx1, Stat1 and Bst2), inflammation (Il6,Tnfa), complement (Cfb,C1ra,C2, C3), cytolysis (Gzma, Gzmb, Prf1) consistent with antiviral responses and inflammation together with neuroprotective regulatory signals (Il27, Il10, and stat2). The increased proinflammatory signature was associated with downregulation neurodevelopmental genes (Calb2, Pvalb, Olig1 and Olig2). We tested the utility of this mouse model by assessing the protection conferred by direct antivirals JNJ-A07 and ST-148 and host-directed antiviral immunomodulatory CpG oligodeoxynucleotide (ODN), alone or in combination against lethal dengue viral infection. The data showed that immunomodulatory CpG ODN and antiviral JNJ-A07 improved the survival of neonatal mice, and protection from lethal neurotropic infection was optimal when treatments were combined. This study suggests that a combination of an effective dengue antiviral along with a host-directed therapeutic may be a useful strategy to protect against dengue virus infections.

Keywords: Dengue; immunocompetent; immunomodulatory; mouse model; neurotropic; virus.

Figure 1.

Susceptibility of neonatal immunocompetent mice to DENV2. (A) Experimental scheme created in BioRender.com/o93c834. B6 wildtype mice were subcutaneously infected with DENV2 24 h after birth. (B) White blood cell (WBC) and platelet (PLT) counts were determined in control and DENV2-infected mice at 8 dpi. (C) Mice were monitored for signs of disease, mortality over time and scored from normal (0, −/+) to severe (3, +++). (D) Mice were monitored for weight loss for 3500 TCID₅₀ DENV2 infected (n = 10), 150 TCID₅₀ DENV2 infected (n = 3), and uninfected (n = 3), and mortality (E) of WT neonatal B6 mice infected with 3500 TCID₅₀ DENV2 (n = 13), 150 TCID₅₀ DENV2 (n = 3), and uninfected (n = 6). Graphs show mean ± SD; *denotes the weight difference between DENV2 and uninfected mice at 9DPI (p < 0.05) as measured by 2-tailed unpaired Student’s t test.

Figure 2.

Viral load of DENV2 in B6 WT neonatal mice infected with DENV2. (A) DENV2 RNA was quantified in the CNS and peripheral organs of infected B6 WT mice at 1, 2, 3, 6 and 8 dpi (N = 3–6 per group) using quantitative real-time PCR. Values are presented as the number of viral RNA copies/mg of RNA. (B) Viral loads in the brain and eyes of DENV2-infected mice were evaluated by TCID50 at 3, 6 and 8 dpi. Virus-infected cells in the brain of B6 WT mice infected with DENV2 and uninfected control. (C) IF-IHC staining of brain of B6 WT uninfected (bottom) and infected (top) mice with DENV2 (8 dpi). DENV2 virus labelled with 4G2 ab (orange). Scale bar: 1 mm. The scale bar is the same for all figures. (D) Confocal imaging of neurons (Fox2, green), astrocytes (GFAP, green), and DENV2 virus (4G2 ab, red) staining in the cortex of infected or uninfected mice. Arrows indicate infected neurons. Scale bar: 50 μm. The scale bar is the same for all figures.

Figure 3.

Gene expression in the CNS in response to DENV2 infection. (A) Nanostring mouse immunology panel. Heatmap shows gene expression (normalized log-scaled counts) in the brain DENV2 infected (purple) and uninfected age-matched control (green) at 6 dpi (white) and 8 dpi (black). Each column represents an individual mouse (n = 2–4/group) and the green and purple lines on the right depict the average normalized gene expression. Both Genes and Mice are clustered based on distances between the samples and clusters, and the similarity is shown by the respective dendrograms. (B) Heatmaps depict the geometric mean change in gene expression in brain tissue of DENV2-infected mice relative to age-matched uninfected mice.

Figure 4.

Immune cells infiltrating the CNS in response to DENV2 infection. (A) IF-IHC staining of the brain of B6 WT uninfected (top) and infected (bottom) with DENV2 (8 dpi). Activated microglia (Iba-1, green) and infiltrating immune cells (CD45+, white) are present in the brains of infected mice. Scale bar: 100μm. (B) Flow cytometry was performed on cells isolated from the CNS (n = 8) of DENV2-infected B6 WT at 8 dpi. Live cells were gated based on CD45 expression. (C) Percent of CD45^high cells in total live cells. (D) Phenotype of CD45^high infiltrating cells in CNS. (E) MHCII expression in CD45^low microglial cells from the CNS of mice infected with DENV2 (red) compared to uninfected age-matched control (blue).

Figure 5.

Anti-viral JNJ-A07 improves the survival rate of DENV-infected mice but antiviral ST-148 does not. (A) ST-148 reduces viral load in Vero cells challenged with DENV2. VERO E6 cells were pretreated with ST-148 2 h before challenge with DENV2 (MOI 0.1). (B) VERO E6 cells were treated with JNJ-A07 2 h after challenge with DENV2 (MOI 1). (C) Schematic of experimental design created in BioRender.com/o93c834. P0 WT B6 mice were treated with ST-148 (50 mg/kg) 24 h before challenge with DENV2 and received four additional doses of ST-148 every 24 h thereafter. Survival curve of mice treated with ST-148 and challenged with DENV2. (D) Schematic of experimental design. P0 WT B6 mice were treated with JNJ-A07 i.p. 24 h after infection with a lethal dose of DENV2 and received four additional doses of JNJ-A07 every 24 h thereafter. Survival curve of mice treated with JNJ-A07at 15 mg/kg (dashed blue line) and 3 mg/kg (solid blue line) 24 h after challenge with DENV2.

Figure 6.

CpG ODN improves the survival rate of DENV- infected mice. (A) Schematic of experimental design created in BioRender.com/o93c834. Top: P0 WT B6 mice were treated with CPG ODN (50ug) subcutaneously and after 24 h infected with a lethal dose of DENV2. Bottom: P1 WT W6 mice were infected with a lethal dose of DENV2 and after 24 h mice were treated with CpG ODN (50ug) subcutaneously. (B) Survival curve of mice treated with control ODN 1556 (50 mg) (n = 11), CpG ODN 1555 (50 mg) locally (green solid line; n = 16) and systemically 24 h before challenge with DENV2 (green dashed line; n = 6). Mice were treated locally with CpG ODN 1555 (50 mg) (green dotted line; n = 8) 24 h after challenge with DENV2. (C) DENV2 RNA levels in the CNS and peripheral organs of infected B6 WT mice that received CpG ODN treatment locally (green triangle) and systematically (solid green circle) 24 h before challenge with DENV2 and survived. (D–G) Survival curve of B6-RAG KO (D), B6-129 IL15RA KO (E), B6-GP49KO (F), and B6-IFNbKO (G) mice treated with CpG ODN (50 µg) locally 24 h before challenge with DENV2.

Figure 7.

Combination therapy of JNJ-A07 and CpG ODN reduces DENV viral replication on VERO E6 cells and improves the survival rate of DENV-infected mice. (A) Conditioned media from PBMCs stimulated with CpG ODN K3 (diamond) and D35 (square) and JNJ-A07 (blue) at 0.01 μM were added to VERO E6 cells infected with DENV2 at an MOI of 0.1. (B) Schematic of experimental design created in BioRender.com/o93c834. P0 B6 WT mice were treated with CpG ODN (50 µg) systematically 24 h before challenge with DENV2 and received JNJ-A07 at 3 mg/kg on day P1 of life and received four additional doses of JNJ-A07. (C) Survival curve of mice treated with monotherapies of CpG ODN, JNJ-A07 and a combination of both.