Identification of stage-related and severity-related biomarkers and exploration of immune landscape for Dengue by comprehensive analyses

Abstract

Background: At present, there are still no specific therapeutic drugs and appropriate vaccines for Dengue. Therefore, it is important to explore distinct clinical diagnostic indicators.

Methods: In this study, we combined differentially expressed genes (DEGs) analysis, weighted co-expression network analysis (WGCNA) and Receiver Operator Characteristic Curve (ROC) to screen a stable and robust biomarker with diagnosis value for Dengue patients. CIBERSORT was used to evaluate immune landscape of Dengue patients. Gene Ontology (GO) enrichment, Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis and Gene set enrichment analysis (GSEA) were applied to explore potential functions of hub genes.

Results: CD38 and Plasma cells have excellent Area Under the Curve (AUC) in distinguishing clinical stages for Dengue patients, and activated memory CD4+ T cells and Monocytes have good AUC for this function. ZNF595 has acceptable AUC in discriminating dengue hemorrhagic fever (DHF) from dengue fever (DF) in whole acute stages. Analyzing any serotype, we can obtain consistent results. Negative inhibition of viral replication based on GO, KEGG and GSEA analysis results, up-regulated autophagy genes and the impairing immune system are potential reasons resulting in DHF.

Conclusions: CD38, Plasma cells, activated memory CD4+ T cells and Monocytes can be used to distinguish clinical stages for dengue patients, and ZNF595 can be used to discriminate DHF from DF, regardless of serotypes.

Keywords: Autophagy; CD38; DH; DHF; ZNF595.

Fig. 1

Flow chart diagram. RNA expression levels in peripheral blood mononuclear cell (PBMC) of dengue fever patients will change during Dengue virus (DENV) infection. We downloaded these RNA-sequence data and clinical information from National Center of Biotechnology Information (NCBI) dataset (https://www.ncbi.nlm.nih.gov/geo/). Combining R package “Limma” with “WGCNA”, stable differentially expressed genes (DEGs) among three phases and between DHF and DF for Dengue patients were screened and we also explored potential functions for these DEGs using Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Set Enrichment Analysis (GSEA) analysis. Based on gene expression data, fractions of immune cells in whole blood samples were estimated by the “CIBERSORT” website and correlations between immune cells and huh genes were also qualified. Finally, we used Area Under the Curve (AUC) to evaluate diagnosis value of huh genes and important immune cells in distinguishing clinical stages and severity

Fig. 2

A–E Volcano maps and F–J heatmaps screen out differentially expressed genes (DEGs). K–O Gene Ontology (GO) analysis shows that differentially expressed genes (DEGs) are mainly involved in biological processes and P–T Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis displays potential enrichment pathways of DEGs. (C, Convalescent stage; EA, Early Acute stage; LA, Late Acute stage)

Fig. 3

Weighted co-expression network analysis (WGCNA). A–C The scale-free fit index (left) and average connectivity (right) of different soft threshold powers. D–F Clustering dendrograms of genes. G–I Heatmaps of the module-trait display correlations among different stages. (C, Convalescent stage; EA, Early Acute stage; LA, Late Acute stage)

Fig. 4

Venn diagrams screen out shared genes: A 143 shared genes between the turquoise module and differentially expressed genes (DEGs) (between the C stage and the EA stage); B 99 shared genes between the blue module and DEGs (between the C stage and the LA stage); C 187 shared genes between the black and green module and DEGs (between the EA stage and the LA stage); D 2 shared genes among three stages. E and F showing the potential function enrichment pathways for CD38 and CDKN1C by gene set enrichment analysis (GSEA). G The boxplot showing changes in expression levels of CD38 and CDKN1C over time (***P < 0.001). H A negative correlation between CD38 and CDKN1C. (C, Convalescent stage; EA, Early Acute stage; LA, Late Acute stage)

Fig. 5

A and B Significantly different expression levels of CD38 and CDKN1C in three stages. C CD38 and CDKN1C can be used to distinguish Dengue samples from normal samples. Analyzing staging diagnosis value of CD38 and CDKN1C for Dengue by Area Under the Curve (AUC): D–F analyzed Firstly in the training group, G–I verified subsequently in the verifying group, J–L tested lastly in the testing group. (C, Convalescent stage; EA, Early Acute stage; LA, Late Acute stage)

Fig. 6

A Shared differentially expressed genes (DEGs) between Dengue Fever (DF) and Dengue Hemorrhagic Fever (DHF) in the Acute stage. Analyzing diagnosis value of LOC101928288, TCN1, DEFA4, FRG1B, LOC286087 and ZNF595 for Dengue severity by Area Under the Curve (AUC): B–C in the training dataset, D in the verifying dataset, E–F in testing datasets. (C, Convalescent stage; EA, Early Acute stage; LA, Late Acute stage)

Fig. 7

A–E Violin diagrams and F–H heatmaps display immune differences of immune cells in different comparing groups. I–J Correlation between genes (CD38 and ZNF595) and 22 types of immune cells (red rectangular: statistically significant (P < 0.05)). (C, Convalescent stage; EA, Early Acute stage; LA, Late Acute stage)

Fig. 8

Analyzing staging diagnosis value of immune cells for Dengue by Area Under the Curve (AUC). A–C in the training group and D–F in the test group. G Validating diagnosis value of immune cells between Dengue samples and normal samples. (C, Convalescent stage; EA, Early Acute stage; LA, Late Acute stage)