A Proteomic Approach Identifies Candidate Early Biomarkers to Predict Severe Dengue in Children

Abstract

Background: Severe dengue with severe plasma leakage (SD-SPL) is the most frequent of dengue severe form. Plasma biomarkers for early predictive diagnosis of SD-SPL are required in the primary clinics for the prevention of dengue death.

Methodology: Among 63 confirmed dengue pediatric patients recruited, hospital based longitudinal study detected six SD-SPL and ten dengue with warning sign (DWS). To identify the specific proteins increased or decreased in the SD-SPL plasma obtained 6-48 hours before the shock compared with the DWS, the isobaric tags for relative and absolute quantification (iTRAQ) technology was performed using four patients each group. Validation was undertaken in 6 SD-SPL and 10 DWS patients.

Principal findings: Nineteen plasma proteins exhibited significantly different relative concentrations (p<0.05), with five over-expressed and fourteen under-expressed in SD-SPL compared with DWS. The individual protein was classified to either blood coagulation, vascular regulation, cellular transport-related processes or immune response. The immunoblot quantification showed angiotensinogen and antithrombin III significantly increased in SD-SPL whole plasma of early stage compared with DWS subjects. Even using this small number of samples, antithrombin III predicted SD-SPL before shock occurrence with accuracy.

Conclusion: Proteins identified here may serve as candidate predictive markers to diagnose SD-SPL for timely clinical management. Since the number of subjects are small, so further studies are needed to confirm all these biomarkers.

Fig 1. Study profile.

Of 113 patients enrolled, 63 had laboratory-confirmed dengue infection. One patient hospitalized in shock presentation was excluded. Six SD-SPL and ten DWS, with early stage plasma collected, were included in the current study.

Fig 2. Identification of angiotensinogen (AGT) and antithrombin III (AT III) proteins.

Representative precursor ion and MS/MS spectra of reporter ions from peptide ALQDQLVLVAAK and DPTFIPAPIQAK of AGT (A) and from peptide RVWELSK and ATEDEGSEQKIPEATNR of AT III (B). Quantification was derived from the signal intensities of eight iTRAQ reporter ions.

Fig 3. Western-blot analysis validating iTRAQs results for angiotensinogen and antithrombin III.

Representative protein bands in DWS (n = 10) and SD-SPL (n = 6) for AGT (A) and for AT III (C). Quantification of relative protein expression of AGT (B) and AT III (D), based on normalized densitometry to healthy control, *p < 0.05.

Fig 4. Receiver operating characteristic (ROC) curves comparing antithrombin III and angiotensinogen markers for SD-SPL prediction.

Area under the curve (AUC) for antithrombin III = 0.85, AUC for angiotensinogen = 0.83 and AUC for combination of two markers = 0.87.