Protective versus pathologic pre-exposure cytokine profiles in dengue virus infection

Abstract

Background: Hyperendemic circulation of all four types of dengue virus (DENV-1-4) has expanded globally, fueling concern for increased incidence of severe dengue. While the majority of DENV infections are subclinical, epidemiologic studies suggest that type-cross-reactive immunity can influence disease outcome in subsequent infections. The mechanisms controlling these differential clinical outcomes remain poorly defined.

Methodology/principal findings: Blood samples were collected from a cohort of school-aged Thai children who subsequently experienced a subclinical DENV infection or developed dengue illness. PBMC collected prior to infection were stimulated in vitro with DENV and the secretion of 30 cytokines was measured using a multiplexed, bead-based array. Significant differences were found in cytokine production based on both the type of DENV used for stimulation and the occurrence of clinical illness. Secretion of IL-15 and MCP-1 was significantly higher by PBMC of subjects who later developed symptomatic DENV infection. In addition, IL-6 was produced by PBMC from all subjects who subsequently developed symptomatic infection, versus 59% of subjects who had subclinical infection. Secretion of IL-12, IL-2R, MIP-1α, RANTES, GM-CSF, and TNFα was significantly lower by PBMC from subjects with symptomatic infection.

Conclusions/significance: These data demonstrate significant differences in pre-existing immune responses to DENV associated with the clinical outcome of subsequent infection. The finding of higher levels of some cytokines in subjects with symptomatic infection and higher levels of other cytokines in subjects with subclinical infection supports the existence of both protective and pathologic immune profiles. Clinical-immunological correlations identified in the context of natural DENV infection may be useful for evaluating immune responses to dengue vaccines.

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1. Relative expression of cytokines in cell culture supernatants from stimulated pre-illness PBMC.

PBMC from subjects who went on to experience subclinical (n = 29) or symptomatic (n = 22) DENV infections were stimulated in vitro with anti-CD3 antibody (positive control), live DENV-1, live DENV-2, live DENV-3, live DENV-4, or uninfected Vero cell supernatant (negative control). After 6–7 days, culture supernatants were assessed by a multiplexed, bead-based array for quantification of 30 cytokines/chemokines/growth factors. Shown are three- and five-fold changes up (light and dark blue, respectively) or down (grey and pink, respectively), relative to the negative control, of each listed analyte. Each row represents responses from a single individual.

Fig 2. Increased cytokine production in subjects who subsequently had subclinical versus symptomatic infections.

(A) Levels (pg/mL) of GM-CSF, TNFα, IL-12, IL-2R, MIP-1α, and RANTES were elevated in stimulated PBMC of subjects who subsequently developed subclinical (filled circles), as compared to symptomatic (open squares), DENV infections. (B) MIP-1β and VEGF responses also showed slightly elevated production by subclinical cases, although they did not reach statistical significance. (C) IFNγ expression was largely found only in subclinical cases. Each symbol represents responses from a single sample following stimulation with uninfected supernatant (unstimulated; negative control, NC), a positive control stimulus (anti-CD3; PC), or live DENV-1 (DV-1), DENV-2 (DV-2), DENV-3 (DV-3), or DENV-4 (DV-4). Median responses are represented by the grey bars. The dotted line indicates the lower level of detection for each analyte. The Wilcoxon rank-sum test was used to compare cytokine production between the subclinical and symptomatic groups (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001).

Fig 3. Higher frequencies of DENV-specific T cells in PBMC prior to subclinical versus symptomatic infections.

Intracellular IFNγ, TNFα, and IL-2 production by (A) CD4+ and (B) CD8+ T cells was measured by flow cytometry in response to DENV-1-4 inactivated antigen stimulation of PBMC from Thai schoolchildren. Comparison groups are PBMC from children who later developed subclinical (open symbols) or symptomatic (closed symbols) infections. The Wilcoxon rank-sum test was used to compare cytokine production between the two groups (*p<0.05, **p<0.01). Data are presented relative to the negative control (responses to negative control stimulation were subtracted out).

Fig 4. Higher IL-6, IL-15, and MCP-1 production in pre-exposure PBMC from symptomatic, compared to subclinical, cases.

IL-6, IL-15, and MCP-1 levels (pg/mL) were higher in culture supernatants of stimulated PBMC from subjects who subsequently developed symptomatic (open squares), as compared to subclinical (filled circles), DENV infections. Each symbol represents responses from a single sample following stimulation with uninfected supernatant (unstimulated; negative control, NC), a positive control stimulus (anti-CD3; PC), or live DENV-1 (DV-1), DENV-2 (DV-2), DENV-3 (DV-3), or DENV-4 (DV-4). Median responses are represented by the grey bars. The dotted line indicates the upper or lower level of detection, as applicable, for each analyte. The Wilcoxon rank-sum test was used to compare cytokine production between the subclinical and symptomatic groups (*p