Comparison of Static and Microfluidic Protease Assays Using Modified Bioluminescence Resonance Energy Transfer Chemistry

BACKGROUND Fluorescence and bioluminescence resonance energy transfer (F/BRET) are two forms of Förster resonance energy transfer, which can be used for optical transduction of biosensors. BRET has several advantages over fluorescence-based technologies because it does not require an external light source. There would be benefits in combining BRET transduction with microfluidics but the low luminance of BRET has made this challenging until now. METHODOLOGY We used a thrombin bioprobe based on a form of BRET (BRET(H)), which uses the BRET(1) substrate, native coelenterazine, with the typical BRET(2) donor and acceptor proteins linked by a thrombin target peptide. The microfluidic assay was carried out in a Y-shaped microfluidic network. The dependence of the BRET(H) ratio on the measurement location, flow rate and bioprobe concentration was quantified. Results were compared with the same bioprobe in a static microwell plate assay. PRINCIPAL FINDINGS The BRET(H) thrombin bioprobe has a lower limit of detection (LOD) than previously reported for the equivalent BRET(1)-based version but it is substantially brighter than the BRET(2) version. The normalised BRET(H) ratio of the bioprobe changed 32% following complete cleavage by thrombin and 31% in the microfluidic format. The LOD for thrombin in the microfluidic format was 27 pM, compared with an LOD of 310 pM, using the same bioprobe in a static microwell assay, and two orders of magnitude lower than reported for other microfluidic chip-based protease assays. CONCLUSIONS These data demonstrate that BRET based microfluidic assays are feasible and that BRET(H) provides a useful test bed for optimising BRET-based microfluidics. This approach may be convenient for a wide range of applications requiring sensitive detection and/or quantification of chemical or biological analytes.