Feasibility Study of an Optical Caustic Plasmonic Light Scattering Sensor for Human Serum Anti-Dengue Protein E Antibody Detection

Antibody detection and accurate diagnosis of tropical diseases is essential to help prevent the spread of disease. However, most detection methods lack cost-effectiveness and field-portability, which are essential features for achieving diagnosis in a timely manner in developing countries. To address this problem, transparent 3D printed sample chambers with a total volume of 700 microliters and an oblate spheroid shape were fabricated to measure green light scattering of gold nanoparticles using an optical caustic focus to detect antibodies. Scattering signals from 90 degree scattering of 20, 40, 50, 60, 80, 100, and 200 nm gold nanoparticles using a green laser and standard quartz cuvette were compared to the scattering signals from a green LED light source with an oblate spheroid sample chamber and to Mie theory by fitting the data to a logistic curve. The change in signal from 60 nm to 120 nm decreased in the order of Mie Theory > Optical Caustic scattering > standard laser 90 degree scattering. These results suggested that conjugating 60 nm gold nanoparticles with Dengue Protein E and using an optical caustic system to detect plasmonic light scattering would result in a sensitive test for detecting human antibodies against Dengue Protein E in serum. To explore this possibility, we studied the light scattering response of protein E conjugated gold nanoparticles exposed to different concentrations of anti-protein E antibody, and posteriorly via a feasibility study consisting of 10 human serum samples using a modified dot blot protocol and a handheld optical caustic-based sensor device. The overall agreement between the benchtop light scattering and dot blot results and the handheld optical caustic sensor suggest that the new sensor concept shows promise to detect gold nanoparticle aggregation caused by the presence of the antibody using a homogeneous assay. Further testing and protocol optimization is needed in order to draw conclusions on the positive predictive and negative predictive values for this new testing system.