Acute Zika Virus Infection in an Endemic Area Shows Modest Proinflammatory Systemic Immunoactivation and Cytokine-Symptom Associations

Abstract

An early immune response to Zika virus (ZIKV) infection may determine its clinical manifestation and outcome, including neurological effects. However, low-grade and transient viremia limits the prompt diagnosis of acute ZIKV infection. We have investigated the plasma cytokine, chemokine, and growth factor profiles of 36 individuals from an endemic area displaying different symptoms such as exanthema, headache, myalgia, arthralgia, fever, hyperemia, swelling, itching, and nausea during early-phase infection. These profiles were then associated with symptoms, revealing important aspects of the immunopathophysiology of ZIKV infection. The levels of some cytokines/chemokines were significantly higher in acute ZIKV-infected individuals compared to healthy donors, including interferon (IFN) gamma-induced protein 10 (IP-10), regulated on activation, normal T cell expressed and secreted (RANTES), IFN-γ, interleukin (IL)-9, IL-7, IL-5, and IL-1ra, including some with predominantly immunoregulatory activity. Of note, we found that higher levels of IP-10 and IL-5 in ZIKV-infected individuals were strongly associated with exanthema and headache, respectively. Also, higher levels of IL-1ra were associated with subjects with arthralgia, whereas those with fever showed lower levels of granulocyte-colony stimulating factor (G-CSF). No correlation was observed between the number of symptoms and ZIKV viral load. Interestingly, only IP-10 showed significantly decreased levels in the recovery phase. In conclusion, our results indicate that acute ZIKV infection in a larger cohort resident to an endemic area displays a modest systemic immune activation profile, involving both proinflammatory and immunoregulatory cytokines and chemokines that could participate of virus control. In addition, we showed that differential cytokine/chemokine levels are related to specific clinical symptoms, suggesting their participation in underlying mechanisms.

Keywords: acute Zika virus infection; chemokines; cytokines; immune response; immunoactivation; low viremia; symptoms.

Figure 1

Cytokine/chemokine profiling in acute Zika virus (ZIKV) infection. (A) Differences in plasma cytokine/chemokine levels between acute ZIKV-infected and healthy subjects. Names in red represent cytokines with significantly higher levels upon infection (p < 0.05 and greater than twofold induction). The size of the circles is proportional to the significance. p-Values are shown in blue. (B) Individual levels of the cytokines upregulated in acute ZIKV infection. Each column represents a study subject. The levels of each cytokine/chemokine (rows) were z-normalized.

Figure 2

Plasma concentration and frequency of differentially expressed cytokines in acute Zika virus (ZIKV)-infected individuals compared to healthy controls. (A) Cytokines measured in the plasma of acute ZIKV-infected individuals and healthy controls and the concentrations of differentially expressed cytokines are shown. The dots on the graphs represent individuals. p-Values are indicated at the top of each graph. The non-parametric Mann–Whitney test was used and considered significant when p < 0.05. The cytokine frequencies of ZIKV-infected individuals and healthy controls showed significant differences considering the limit of detection of the test (B) and based on the standard curve range (C). Chi-square tests were used (B,C) and p < 0.05 was considered significant.

Figure 2

Plasma concentration and frequency of differentially expressed cytokines in acute Zika virus (ZIKV)-infected individuals compared to healthy controls. (A) Cytokines measured in the plasma of acute ZIKV-infected individuals and healthy controls and the concentrations of differentially expressed cytokines are shown. The dots on the graphs represent individuals. p-Values are indicated at the top of each graph. The non-parametric Mann–Whitney test was used and considered significant when p < 0.05. The cytokine frequencies of ZIKV-infected individuals and healthy controls showed significant differences considering the limit of detection of the test (B) and based on the standard curve range (C). Chi-square tests were used (B,C) and p < 0.05 was considered significant.

Figure 3

Cytokine/chemokine level correlations in Zika virus-infected subjects. Connections (edges) between two given cytokines represent direct significant correlations (Spearman p-value <0.05 and R > 0.5). The width of each edge is directly proportional to the R value. The node colors represent the fold-changes in cytokine/chemokine levels between infected and healthy subjects.

Figure 4

Cytokine and chemokine level changes associated with clinical symptoms. (A) Clinical symptoms in Zika virus (ZIKV)-infected subjects. The heat map represents the presence (red) or absence (gray) of symptoms (columns) in ZIKV-infected subjects (rows). RNAemia levels for each subject are shown in the bar graph. (B) Different cytokine levels were associated with different symptoms. The green pentagons represent clinical symptoms and the nodes represent the cytokines and the edges represent putative associations among the pentagons. Red arrows indicate that the level of a given cytokine is higher in ZIKV-infected subjects with the symptoms compared to subjects without them. Blue arrows indicate the cytokine level is lower in patients with the associated symptoms compared to those without them. The box plots show the levels of granulocyte-colony stimulating factor and IL-5 in subjects with fever and headache, respectively. Dots on the graphs represent individuals. p-Values are indicated at the top of each box plot.

Figure 5

Cytokine levels can predict the clinical symptoms of Zika virus-infected subjects. The green pentagons represent clinical symptoms, the nodes represent cytokines and the edges represent the mean decrease of the Gini index (MDGI). Arrow thickness is proportional to the MDGI index.

Figure 6

Comparison of plasma cytokine levels during the acute and recovery phases. (A) Cytokine measurements were performed in the plasma of six individuals in the recovery phase (2–3 weeks after the onset of symptoms) and compared to those performed during the acute phase. The dots on the graphs represent individuals. The Wilcoxon matched-pair test was used. p-Values are indicated at the top of the graph and were considered significant when p = 0.0312. (B) The levels of cytokines with fold-changes >2 over healthy control levels in the acute or recovery phases are represented. The node color represents the fold-change in cytokine levels between infected and healthy subjects.