Micromachines (Nov 2021)

A 3D Microfluidic ELISA for the Detection of Severe Dengue: Sensitivity Improvement and Vroman Effect Amelioration by EDC–NHS Surface Modification

  • Hinata Maeno,
  • Pooi-Fong Wong,
  • Sazaly AbuBakar,
  • Ming Yang,
  • Sing-Sin Sam,
  • Juraina Jamil-Abd,
  • Anusha Shunmugarajoo,
  • Mahiran Mustafa,
  • Rosaida Md Said,
  • Eashwary Mageswaren,
  • Azureen Azmel,
  • Anilawati Mat Jelani

DOI
https://doi.org/10.3390/mi12121503
Journal volume & issue
Vol. 12, no. 12
p. 1503

Abstract

Read online

Serum is commonly used as a specimen in immunoassays but the presence of heterophilic antibodies can potentially interfere with the test results. Previously, we have developed a microfluidic device called: 3D Stack for enzyme-linked immunosorbent assay (ELISA). However, its evaluation was limited to detection from a single protein solution. Here, we investigated the sensitivity of the 3D Stack in detecting a severe dengue biomarker—soluble CD163 (sCD163)—within the serum matrix. To determine potential interactions with serum matrix, a spike-and-recovery assay was performed, using 3D Stacks with and without surface modification by an EDC–NHS (N-ethyl-N′-(3-(dimethylamino)propyl)carbodiimide/N-hydroxysuccinimide) coupling. Without surface modification, a reduced analyte recovery in proportion to serum concentration was observed because of the Vroman effect, which resulted in competitive displacement of coated capture antibodies by serum proteins with stronger binding affinities. However, EDC–NHS coupling prevented antibody desorption and improved the sensitivity. Subsequent comparison of sCD163 detection using a 3D Stack with EDC–NHS coupling and conventional ELISA in dengue patients’ sera revealed a high correlation (R = 0.9298, p < 0.0001) between the two detection platforms. Bland–Altman analysis further revealed insignificant systematic error between the mean differences of the two methods. These data suggest the potentials of the 3D Stack for further development as a detection platform.

Keywords