Zwitterionic polymer-modified silicon microring resonators for label-free biosensing in undiluted human plasma

Abstract

A widely acknowledged goal in personalized medicine is to radically reduce the costs of highly parallelized, small fluid volume, point-of-care and home-based diagnostics. Recently, there has been a surge of interest in using complementary metal-oxide-semiconductor (CMOS)-compatible silicon photonic circuits for biosensing, with the promise of producing chip-scale integrated devices containing thousands of orthogonal sensors, at minimal cost on a per-chip basis. A central challenge in biosensor translation is to engineer devices that are both sensitive and specific to a target analyte within unprocessed biological fluids. Despite advances in the sensitivity of silicon photonic biosensors, poor biological specificity at the sensor surface remains a significant factor limiting assay performance in complex media (i.e. whole blood, plasma, serum) due to the non-specific adsorption of proteins and other biomolecules. Here, we chemically modify the surface of silicon microring resonator biosensors for the label-free detection of an analyte in undiluted human plasma. This work highlights the first application of a non-fouling zwitterionic surface coating to enable silicon photonic-based label-free detection of a protein analyte at clinically relevant sensitivities in undiluted human plasma.

Fig. 1

(a) Schematic detailing the deposition of DpC polymer chains on the sensor’s oxide surface. (b) The relative shift in resonance wavelength over time during surface modification. Microrings were exposed to (1) a solution of DpC or BSA and then returned to (2) deposition buffer. DpC-coated microring resonators were (3) washed with buffer to remove loosely bound polymer, and returned to (4) deposition buffer to quantify the adsorbate. (c) Schematic illustrating the resistance of DpC-coated surfaces to protein fouling in complex media. (d) Microring resonators were exposed to undiluted human plasma for 15 min prior to returning to buffer to determine the amount of non-specific adsorption based on the overall shift in resonance wavelength before and after exposure to plasma. Note: illustrations are not to scale.

Fig. 2

(a) Specific binding of SA to DpC-immobilized antiSA in buffer. Note: illustrations are not to scale. (b) Upon exposure to a solution of SA in buffer, antiSA-DpC microrings demonstrate specific binding of SA as well as the appropriate dissociation response upon returning to buffer. As expected, IgGi-DpC microrings show no response to the buffer solution of SA. (c) AntiSA-DpC sensors exhibit concentration dependent binding of SA, with minimal non-specific binding to IgGi-DpC sensors.

Fig. 3

(a) Sensor response to increasing concentrations of SA in undiluted human plasma separated by brief washes with buffer (inset, top). Specific SA binding was defined as the difference in signal between antiSA-DpC and IgGi-DpC microrings (inset, bottom). (b) A Langmuir best-fit binding curve for the differential binding response between antiSA-DpC microrings and IgGi-DpC negative control sensors (n ≥ 30). The gray box represents the differential sensor response to an unspiked plasma control (mean ± SD).