Background
Dengue is a mosquito-borne viral disease caused by four serotypes of dengue virus and is endemic in tropical and subtropical areas including Southeast Asia [
1]. A report from World Health Organization indicated the incidence of dengue fever has risen 30-fold in the past 50 years [
2]. A recent study revealed that 390 million people were estimated to be infected with dengue per year [
3]. According to the Taiwan National Infectious Disease Statistics System, more than 20,000 cases of dengue fever were diagnosed in southern Taiwan, including Kaohsiung city and Tainan city, since 2014 [
4]. It is therefore a serious threat to public health in Taiwan and tropical countries.
Infection with any dengue serotype causes a wide spectrum of symptoms, ranging from a mild flu-like syndrome (dengue fever) to a severe syndrome (dengue hemorrhagic fever and dengue shock syndrome) [
2]. Onset of fever is observed in dengue patients during the acute febrile phase of dengue. The acute phase of illness lasts for 3–7 days, and then the convalescent phase lasts for several days to several weeks [
5,
6]. Because dengue virus infects various cell types, including peripheral leukocytes and endothelial cells [
7], the levels of multiple cytokines and chemokines such as macrophage migration inhibitory factor (MIF), tumor necrosis factor alpha (TNF-α), interleukin (IL)-1β, and interferon-γ (IFN-γ), increase at different time points during the disease course [
8‐
10]. The serum level of neutrophils-secreted elastase in patients with dengue shock syndrome is higher than in dengue patients without shock [
11,
12]. In addition, dengue virus-infected patients have elevated levels of soluble intercellular cell adhesion molecule 1 (sICAM-1) and soluble vascular adhesion molecule-1 (sVCAM-1), which may regulate the activation and damage of endothelial cells [
13]. Fas/Fas ligand (FasL) pathways participate in dengue virus-induced apoptosis of endothelial cells [
14]. On the other hand, low platelet count and low leukocyte count are often observed in dengue patients [
15,
16]. Platelets-secreted molecules may play a role in both dengue pathogenesis and inflammation regulation because dengue virus also alters platelets-leukocytes and platelets-endothelial cells interactions [
17]. Activated platelets affect the production of IL-10 and TNF-α from mononuclear cells [
18]. Because some of the aforementioned molecules were common markers in both sepsis and dengue [
19], we were interested in whether other sepsis-associated molecules were also associated with dengue infection.
This study aimed to determine novel dengue-associated molecules in febrile patients. The Luminex assays were used for analyzing the sepsis-associated molecules in serum of dengue patients and healthy controls. In order to determine the potential association between these molecules and cell types (such as leukocytes and endothelial cells), the expression pattern of serum molecules in online available dataset was analyzed, either. Because low white blood cell count and low platelet count are clinical features of dengue, we also investigated whether the concentration of these molecules were correlated with white blood cell count and platelet count.
Methods
Sample collection
Adult patients (age >18 years old) who presented to the emergency department of Kaohsiung Medical University Hospital with fever (tympanic temperature >38.3 °C) from Sep 2014 to Dec 2014 were eligible for the study. After obtaining informed consent, 10 ml of blood was drawn and then the serum separated and stored in aliquots in −80 °C. Because dengue is a notifiable disease in Taiwan, serum should be collected from all patients with suspected dengue infection and sent to the Taiwan Centers for Disease Control (CDC), where final confirmation of the diagnosis was made (The commercial DENV Ag NS1 Strip [Bio-Rad] was used). Patients with dengue confirmed by CDC were selected and their sera were used for the following Luminex assay. The sera from healthy controls were also collected after obtaining informed consent.
Quantification of sepsis-associated molecules
The concentration of serum molecules was determined by using Luminex technology. MILLIPLEX MAP Human Sepsis Magnetic Bead Panel 1 (macrophage migration inhibitory factor [MIF], soluble intercellular adhesion molecule 1 [sICAM-1], soluble Fas [sFas], soluble Fas ligand [sFasL], soluble vascular cell adhesion molecule 1 [sVCAM-1], and total plasminogen activator inhibitor-I [tPAI-I]) (Millipore, Billerica, MA, USA), MILLIPLEX MAP Human Sepsis Magnetic Bead Panel 3 (lactoferrin, elastase 2, neutrophil gelatinase–associated lipocalin [NGAL], resistin, and thrombospondin-1) (Millipore, Billerica, MA, USA) and Magnetic Luminex Performance Assay (Human High Sensitivity Cytokine Base Kit A, interferon-γ [IFN-γ]) (R&D Systems, Minneapolis, MN, USA) were used for determining the serum levels of these molecules according to the manufacturer’s instructions. Data was acquired on Luminex xMAP technology (Millipore, St Charles, MO, US). For concentration calculation, the calibration curve for each serum molecule was analyzed with a five parameter logistic curve fit curve through the Milliplex Analyst Software (Viagene Tech, Carlisle, MA, USA).
Collection of independent microarrays dataset regarding dengue virus infection
The publicly available microarray dataset was obtained from NCBI Gene Expression Omnibus datasets (GEO dataset,
http://www.ncbi.nlm.nih.gov/gds/) [
20]. Briefly, we searched “dengue” at GEO datasets and then the results were filtered by the criteria “human” and “expression profiling by array.” After excluding the datasets from cell lines and peripheral blood mononuclear cells (PBMC) and datasets without healthy control, the dataset with GEO Series accession number of GSE51808 (
http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=gse51808) was chosen. The dataset contains healthy controls (9 healthy controls) and dengue patients (18 dengue fever patients and 10 dengue hemorrhagic fever patients) [
21]. Expression values were adapted from GEO2R.
Statistical analysis
Differences between two independent groups were analyzed with student’s T test (Two tailed). Comparisons between three groups were done with one-way ANOVA test, followed by Bonferroni’s multiple comparison test. The results were expressed as the median and the inter quartile range. Correlation was estimated with Spearman’s correlation. Graphs and statistical analysis were carried out using GraphPad Prism version 5.03 (GraphPad Software, San Diego, CA). P-value < 0.05 was considered significant.
Discussion
In 2014, there was a dengue fever outbreak in Taiwan and 96 % of patients were from Kaohsiung city [
22]. A recent report indicates dengue virus 1 is dominant serotype [
23]. Because patients were enrolled from Sep 2014 to Dec 2014 in Kaohsiung city, we supposed serotype I was predominant serotype in our study. Furthermore, a report indicated around 70 % of non-dengue virus 2 infection (dengue virus 1 and dengue virus 3) is primary infection in Taiwan [
24]. It might imply that enrolled patients with primary infection were dominant. Our results showed the serum concentrations of MIF, sVCAM-1, sFasL, resistin, and IFN-γ were significantly higher in dengue patients than that in healthy controls. Previous studies have indicated that elevated serum levels of MIF, sVCAM-1, sFas, sFasL, and IFN-γ are associated with dengue [
9,
10,
13,
14]. Similar expression pattern was observed in the present study. To the best of our knowledge, this is the first report describing elevated serum levels of resistin in dengue patients.
Previous studies have shown that elastase 2, lactoferrin, and NGAL are mainly secreted from neutrophils [
25]. Elastase 2 and lactoferrin locate in azurophilic granules and specific granules, respectively [
26]. Increasing serum levels of elastase 2 and lactoferrin are correlated with neutrophil degranulation and IL-8 is one of the important regulators for neutrophil degranulation in dengue-infected children [
11,
27]. NGAL is a kind of protein inhibiting bacterial growth [
28]. The role of NGAL is unclear in dengue infection. The plasma and urine NGAL level in rotavirus-induced dehydration is higher than that in healthy controls [
29]. However, the function of NGAL is not well-known in dengue infection. In addition, NGAL is a biomarker for human acute kidney injury [
30]. Although dengue hemorrhagic fever is reported to be a risk factor of acute kidney factor in children [
31], the NGAL level does not reveal significant difference between dengue patients and healthy controls in adults. In this study, the dengue patients were enrolled after fever onset. We assumed that most of enrolled patients might be in the acute febrile phase and critical phase of dengue. Relatively low level of all three molecules was observed in dengue patients with leukopenia and thrombocytopenia (Tables
3 and
4). All molecules were significantly correlated with leukocyte and platelet count (Table
5). The highest virus titer is in the febrile phase of dengue (1–3 days post fever onset) and the number of leukocyte and platelet decreases is in critical phase (4–6 days post fever onset) [
32]. It might suggest elastase 2, lactoferrin and NGAL play a role in anti-dengue immune responses in febrile phase.
Potts and colleagues report that dengue patients had lower platelet, white blood cell and neutrophil counts [
33]. Although we did not count neutrophils in this study, we supposed serum levels of elastase 2, lactoferrin, and NGAL correlated with neutrophil counts in adult dengue patients. A recent study indicated that the concentration of the endothelial cell-related molecule VCAM-1 was negatively correlated with neutrophil count [
34]. In Tables
3 and
4, the level of sVCAM-1, IFN-γ and three neutrophils-related molecules showed opposite expression pattern in the leukopenic group and thrombocytopenic group. Thrombospondin-1 which is an inflammatory molecule in activated platelet shows similar expression pattern with elastase 2, lactoferrin, and NGAL in this study [
35]. It needs to further investigate whether these molecules involve in the regulation of the interaction between T cells, neutrophils, platelet, and endothelial cells during dengue infection in the future.
Resistin is a kind of adipokine and is involved in various inflammatory processes [
36,
37]. Peripheral blood mononuclear cells (PBMCs), macrophages, and bone marrow cells are major sources of resistin in human [
38]. Circulating levels of IL-6, IL-10 and IFN-γ associated with the level of resistin in an obese mouse model [
39]. Pro-inflammatory cytokines, such as IL-6 and TNF-α induce mRNA expression of resistin in human PBMCs. Resistin is significantly correlated with IL-6 and ICAM-1 in patients with obstructive sleep apnea syndrome [
40]. The present study showed that serum resistin level significantly increased in dengue patients. Our results indicated that resistin did not correlate with leukocyte and platelet count although human PBMC and macrophage are reported to secret resistin [
38]. Persistent human papillomavirus infection increases the resistin level in plasma of older women [
41]. However, the function of resistin in dengue virus and human papillomavirus infection is still unknown. In addition, we observed high expression level of resistin, MIF and sFasL in dengue patients and no significant correlation between these molecules and leukocyte and platelet count. It suggests that serum resistin and sFasL might be potential biomarkers for dengue infection since MIF is correlated with disease severity and clinical outcome in dengue [
10].