High-Throughput Screening of Response Kinetics of Genetically-Encoded Metal Sensors with Microfluidics Technology

Abstract

Metal sensors, chromophores which generate responses upon metal-ion binding, are indispensible tools to study the metal homeostasis of the cell. Compared to the traditional dye-based metal sensors, the fluorescent protein (FP)-based FRET sensors offer many distinctive advantages. They are fully genetically encoded and targeted to specific cell location, capable of controlled expression, and provide intensity-independent ratiometric measurement. Yet at present the performances of the FP-based sensors, such as the dynamic range, metal selectivity, and metal affinity, need to be markedly improved to meet the requirements and challenges of their cellular application. By combining a sensor library design strategy, the aim of our work is to screen and sort the FP-based sensors for improved photophysical and photochemical properties using a high-throughput microfluidics method. We have developed a microfluidic platform that combines high-throughput screening with versatile optical FRET detection, fast mixing, and laser trapping techniques on individual mammalian cells. We demonstrated that the FRET responses of the sensors can readily be measured and differentiated on our cytometer with a rate of >20 cells/second. At this throughput a typical small targeted sensor library (∼105 clones) can be sorted within a time frame of several hours. Using this technique we also investigated the response kinetics of various Ca2+ and Zn2+ sensors expressed at different cell locations upon exposure to metal ions in the microfluidic environment. Dissection of the relationship between sensor kinetics, construction, and location will shed light on the mechanism and dynamics of sensor function and ion transportation in the cell. The high-throughput characteristics of the microfluidics approach, integrated with fast optical detection with ms time-resolution over a broad time window (spanning orders of magnitudes from ms to seconds), can be readily adapted to other high-fidelity kinetics-based screening or sorting applications.