PROGRAMMABLE DIELECTROPHORETIC μTAS SAMPLE HANDLING

Abstract

We present the concept of a general-purpose sample analysis platform (GSAP) based on dielectrophoretic methods. The platform architecture comprises integrated functional blocks that can be programmed to perform a diverse range of analysis steps, including the on-device preparation of real world samples.

1. Introduction

Micro total analysis systems for molecular screening will revolutionize the diagnosis and prognosis of disease, as well as industrial, agricultural and environmental applications, and will enable automated, continuous monitoring at points-of-care and in the field. Although great progress has been made in realizing chip-scale analysis, the inability to effectively prepare real world samples on-device is a major shortcoming in many μTAS designs. Here we present the concept of a general-purpose molecular analysis platform based on AC electrokinetic methods. This can be programmed to perform diverse sample processing and analysis tasks, and can accommodate unprocessed, real-world samples. Devices capable of general purpose molecular analyses are the μTAS equivalents of microprocessors — integrated chips that can be adapted to a wide range of applications through appropriate interfaces and software. Microprocessors employ multiple functional blocks used only as required. A similar architecture for μTAS would reduce application development times and costs by being generally adaptable to disparate applications. The functional blocks of a general-purpose sample analysis platform (GSAP) must be able to perform any functions (see Fig. 1) that may be required during analysis starting from a raw sample (e.g., a complex mixture of cells, debris, and interfering ions and molecules).

Figure 1.

Functional steps needed for a general purpose sample preparation and analysis system. In an integrated μTAS, each step can be considered to be a functional module.

2. Experimental and Results

Fully three-quarters of the functional blocks in Fig. 1 are forms of sample manipulation. Because they provide electrically programmable discrimination for handling matter, we applied dielectrophoretic filtering [1] and magnetophoretic-dielectrophoretic-field-flow fractionation [2] for this purpose. These methods were integrated with an AC impedance sensor [3] and external optical sensing. By programming the integrated modules and bypassing some of them as appropriate, the functional block architecture satisfied several analysis applications requiring widely different operational parameters. Real time responses of viable HL-60 cells to exposure with toxicants (Fig. 2) were measured by dielectrophoretically-trapping the cells and then analyzing their DEP-FFF elution times with the AC impedance sensor module. The discrimination and separation of prestained transformed from non-transformed rat kidney cells (Fig. 3) and of live from dead Tilapia spermatozoa were both accomplished by flowing samples past the inactivated DEP trapping and impedance sensor modules while utilizing the DEPPFF and fluorescence modules (sperm were stained with Molecular Probes live/dead sperm kit). Finally, malarially-parasitized cells were detected in human blood by prefiltering samples using DEP-trapping, subsequent fractionation by high discrimination DEP-FFF, and then detecting the fluorescence of the parasitized cells stained with Molecular Probes Sybro 14 (Fig. 4). The frequencies of the DEP trapping and DEP-FFF modules were adjusted to appropriate values for the widely different cell types in the different tasks.

Figure 2.

DEP-FFF elution profile of HL-60 cells following exposure to 15 mM paraquat on the μTAS. The peak marked as apoptotic is absent in cells unexposed to the toxin. Cells were counted by an integrated AC impedance sensor.

Figure 3.

Separation of transformed from non-transformed 6M2 rat kidney cells. Cells were separated by DEP-FFF and detected using the fluorescence detection scheme.

Figure 4.

Detection of Plasmodium falciparum-infected human red blood cells by DEP followed by fluorescence detection of parasite DNA stained by Sybro 14.

3. Conclusions

Although consisting of a limited number of programmable sample preparation and detection modules, our device illustrates the feasibility of utilizing subsets of programmable subunits according the modular architecture of Fig. 1 to achieve widely different applications. We believe this approach can be readily extended to include molecular detection modules and that the range of applications can include bacterial and other pathogen types. A suitable programmable molecular detection module would be the droplet processor we reported earlier [4] for embedded chemical synthesis and analysis applications, for example.