Single-Nucleus and Spatial Transcriptomics Revealing Host Response Differences Triggered by Mutated Virus in Severe Dengue

Abstract

Dengue virus (DENV) infection causes various disease manifestations ranging from an asymptomatic state to severe, life-threatening dengue. Despite intensive research, the molecular mechanisms underlying the abnormal host responses and severe disease symptoms caused by evolved DENV strains is not fully understood. First, the spatial structure of mutant DENV was compared via in silico molecular modeling analysis. Second, employing single-nucleus and spatial RNA sequencing, we analyzed and verified transcriptome samples in uninfected, mild (NGC group), and severe (N10 group) liver tissues from murine models. In this study, we obtained a cumulatively mutated DENV-2 N10 with enhanced capability of replication and pathogenicity post 10 serial passages in Ifnra^-/- mice. This variant caused severe damage in the liver, as compared with other organs. Furthermore, mutated DENV infection elicited stronger responses in hepatocytes. The critical host factor Nrg4 was identified. It dominated mainly via the activation of the NRG/ErbB pathway in mice with severe symptoms. We report on evolved N10 viruses with changes observed in different organisms and tissue. This evolutionary variant results in high replicability, severe pathogenicity, and strong responses in murine. Moreover, the host responses may play a role by activating the NRG/ErbB signaling pathway. Our findings provide a realistic framework for defining disturbed host responses at the animal model level that might be one of the main causes of severe dengue and the potential application value.

Keywords: Nrg4; cumulative mutation; dengue virus; severe dengue; single-nucleus RNA sequencing; spatial transcriptomics.

Figure 1

The enhanced pathogenicity and viral replication in Ifnar1^−/− mice infected with successively passed DENV strains. (A) Graphical representation of the method of generating DENV-2 N10 virus. Ifnar^−/− mice were infected intravenously with 10⁶ PFU/mL of prototypic DENV-2 NGC (n = 3). The infected mice were euthanized at 3 dpi, and serum was separated and inoculated intravenously into a new cohort of mice. This process was repeated 10 times in total. (B–D) Percentage of weight change (B), clinical score (C), and survival rate of infected mice (D). (E) Viral loads in serum at 3 dpi from infected mice were detected via qRT–PCR. (F,G) Cells were infected with NGC or N10 at a multiplicity of infection (MOI) of 0.1. The copies of viral genomic RNA of DENV2-NGC and N10 in C6/36 (F) and Vero cells (G) measured with qRT-PCR from 0 to 8 dpi. * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001.

Figure 2

Dengue virus with cumulative mutations causes pathological changes. Ifnar^−/− mice were infected intravenously with 10⁶ PFU/mL of DENV-2 NGC and N10 (n = 5). (A–E) Histopathology of the liver (A), spleen (B), brain (C), kidney (D), and intestine (E) from the uninfected, NGC-, and N10-infected mice at 8 dpi. The white arrows in the images showing H&E staining all represent the pathological changes mentioned in the results section. Histological images are representative of no fewer than five mice (scale bars, 50 µm). (F) Viral load results for liver, spleen, brain, kidney, and intestine. (G,H) Localization of the DENV NS3 in the liver (G) and quantitative results (H). (I–M) Blood biochemical analysis of uninfected, NGC-, and N10-infected mice at 6 dpi, including Alt (I), Ast (J), total protein (TP) (K), albumin (ALB) (L), and globulin (Glob) (M). Circulating cytokine analysis of uninfected, NGC-, and N10-infected mice at 8 dpi, including TNF-α, IL-17, IL-23, and IL-2 (N). * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001.

Figure 3

DENV strain N10 induces aberrant host responses in the liver. (A) Flow diagram of bioinformatics analysis. (B) The distribution of cell clusters in the livers from uninfected, NGC-, and N10- infected mice was shown in UMAP chart. (C) The distribution of typical cellular marker expression for different clusters was shown in violin plot. (D) Identified DEGs, (E,F) GO, and (G,H) KEGG upregulation and downregulation analysis in hepatocyte populations between the NGC and N10 groups. Dot sizes represent the proportion of all genes from GO analysis that were considered as markedly differentially expressed (gene ratio). Negative log10 FDR-adjusted p values related to each pathway are plotted.

Figure 4

Pseudotime analysis of hepatocyte genes in response to NGC or N10 dengue virus. (A) UMAP of hepatocytes by differentiation state inferred using CytoTRACE. (B) Pseudotime trajectory projected onto the UMAP of hepatocytes. (C,D) Pseudotime analysis of the uninfected, NGC-infected, and N10-infected hepatocytes. The trajectory spans uninfected (arrow) to NGC-infected to N10-infected hepatocytes. (E,F) Heatmaps showing different gene expression patterns during the differentiation of node 1 and node 2 over pseudotime. (G,H) The gene expression distribution of Nrg4 over pseudotime.

Figure 5

Spatial distribution analysis of hepatocyte genes in response to NGC or N10 dengue virus. (A) Spatial location of clusters in the uninfected group, NGC group, and N10 group. (B–I) Spatial expression distribution and expression quantity of candidate genes in the uninfected, NGC, and N10 groups. Candidate genes include Nrg4 (B,C), Sult1e1 (D,E), Ddit4 (F,G), Cxcl13 (H,I).

Figure 6

Cell–cell communication in dengue infection. (A) Involvement in cellular interactions of the NRG signaling pathway. (B) The outgoing (bottom) and incoming (top) communication patterns of different cells. (C) The heatmap shows the relative importance of different cell types (senders, receivers, mediators, and influencers) based on the computed network centrality measures of TGFb, BMP, EGF, and NRG signaling, respectively. (D) Relative contribution of ligand–receptor pairs.

Figure 7

N10 dengue virus infection triggers change in Nrg4. (A) An overview of molecular validation (B–E). The relative mRNA expression levels of Cxcl13 (B), Ddit4 (C), Sult1e1 (D), and Nrg4 (E) in the liver tissues (n = 10). (F–I) The relative protein expression levels of SULT1E1 (F,I) and NRG4 (G,H) in the liver tissues (n  =  5). * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001.