Highly efficient dual-channel cytometric-detection of micron-sized particles in microfluidic device

Abstract

To demonstrate the ability to efficiently count and identify suspended micron-sized particles by simultaneously detecting their fluorescence emission and light scattering in microfabricated channel, a compact configuration that used a polydimethylsiloxane (PDMS) microfabricated channel as interrogation component, hydrodynamic focusing for particle control, and a simple free-space optical setup for signal detection, was accordingly developed. Subsequently, a quantitative count of 1.013 μm diameter fluorescently labeled beads in suspension was implemented in a microfluidic device employing both fluorescence emission and light scattering at average particle throughput ranging from 83 to 416 particles/s. As a result, the detection efficiencies above 88% for both signals and correlation percentages above 97% between them were routinely achieved. In addition, it was shown that effective differentiation of 1.013 μm fluorescently labeled beads from various unlabeled beads in mixed populations of high mixing ratios had been successfully realized in this microfluidic-device-based instrumentation. Finally, the demonstrated system was used to detect fluorescein isothiocyanate (FITC) labeled nonpathogenic bacteria of Escherichia coli (E. coli) DH5α. The results showed the detection efficiencies above 89.7% for fluorescence emission and 94.5% for light scattering signals, and a correlation of 94.9% between the two signals at an average throughput of 350 cells/s have been obtained. As a comparison, the detection accuracies of the dual-channel cytometric detection of the FITC-labeled E. coli DH5α cells in the microfluidic device are approximately 84.3% and 88.8% for fluorescence emission and light scattering respectively when compared against a manual cell count using a haemocytometer as a standard.

Introduction

Rapid and reliable detections of concentration and size distribution of biological particles in suspensions are essential for many biochemical applications, such as in food safety inspection and quality control [1], [2], drug discovery [3], environmental monitoring [4], and biomedicines [5]. Among all currently available methods, optical flow cytometry is one of the most commonly used and has been widely employed to analyze many important characteristics of biological particles by providing rapid, quantitative, and multi-parameter measurements [6]. Conventional flow cytometers are designed to handle micron-sized particles in suspension at throughputs of tens of thousands particles per second with 6–8-parameter analyses [7], [8]; however due to large consumptions of expensive labeling agents in each testing run, high operating costs and limited levels of portability and integrability, these full-sized bench-top instruments are generally dedicated to some very limited applications [9].

As a miniaturized version of conventional flow cytometers, the microchip-based flow cytometers have shown some clear advantages in many areas [10]. By replacing glass tubes with microfabricated channel as the flow channel, this miniaturized flow cytometer can not only process small volumes of biological samples but also integrate many functions, and its portability has also been improved [11], [12]. Over the last few years, microchip-based flow cytometers has rapidly become one of the intensively studied subjects in the field of micro total analysis systems (μTAS) [13], [14], [15], and significant progress have been made in the manipulation, detection, and analysis of biological cells or particles in related microfluidic devices [16], [17]. However, the measurable particles throughput and detection efficiency of these systems are still well below that of modern flow cytometers, particularly for some applications [18], in which fluorescent and light scattering signals need to be detected simultaneously to indicate different characteristics of cells at high throughput with good detection efficiency. To evaluate these capabilities in microfluidic devices, Ramsey and his colleagues [19], [20] firstly developed microchip-based flow cytometers using electrokinetic focusing to detect latex beads and bacterium Escherichia coli (E. coli) at rates of 85 Hz using both fluorescence emission and light scattering and reported a detection efficiency of 94%. Similarly Holmes et al. [21] reported a microchip-based flow cytometers in combination with dielectrophoretic focusing to detect latex beads based on their fluorescence emission at a throughput rate of 250 particles/s. Recently, Simonnet and Groisman [22] constructed a microfluidic device with a complicated hydrodynamic focusing in both in-plane and out-of-plane directions to further improve detection efficiency of fluorescent latex beads in suspension to achieve a detection efficiency of 97% with a sample linear flow rate of up to 8 m/s. In addition to testing micro-scale beads in suspensions as model particles, Sakamoto et al. [23] detected both cultured bacterial cells and bacteria in natural environment with good results to verify the detection accuracy of micro flow cytometer for the real microbiological application. While some others attempts, such as parallelizing [24] and multiplexing microchannels [25], were utilized in electrophoretic analyses and micro coulter counter to improve the throughput of microfluidic device, efficient detection and quantification based on these technologies had not been realized yet.

In this work, we present the development of highly efficient count and identification for artificial latex beads and bacterial cells in a microfluidic device using dual-channel cytometric detection incorporating one dimensional hydrodynamic focusing to spatially confine particles. Using mixtures of fluorescently labeled and unlabeled beads, quantitative count and identification were demonstrated in the microfabricated channel at different particle throughputs up to 416 particles/s along with the high detection efficiencies and good correlation between fluorescence emission and light scattering signals. In addition, efficiently identifying the “targeted” fluorescently labeled particles in suspension mixed with various “interfering” unlabeled particles with dual-channel detection was reported here as well. Finally, the system was applied to test FITC-labeled nonpathogenic E. coli DH5α to verify its detection efficiency and accuracy.