Flow-through sampling for electrophoresis-based microchips and their applications for protein analysis

This work presents a model behind the operation of a flow- through sampling chip and its application for immunoseparation, as well as its integration with a wash/elution bed for protein purification, concentration, and detection. This device used hydrodynamic pressure to drive the sample flow, and a gating voltage was applied to the electrophoretic channel on the microchip to control the sample loading for the separation and to inhibit sample leakage. The deduced model indicates that the critical gating voltage (V^c) that is defined as the minimum gating voltage applied to the microchip for sampling is a function of the pump flow rate, the configuration of the microchannel on the chip, and the electroosmosis of the buffer solution. It was found that the theoretical V^c values calculated from the measured electroosmotic mobilities and flow split ratios were comparable to those experimentally obtained from two microchips with different sampling channel sizes. This had an error percentage ranging from 1 to 20%. Because the hydrodynamic flow is insensitive to electrophoretic mobility, this electrophoresis-based microchip device was free of injection bias due to different ionic strength and electrophoretic mobility in the sample. Additionally, the usefulness of this device was demonstrated for the study of affinity interactions. Mixtures of Cy5-labeled bovine serum albumin (Cy5-BSA) and anti-BSA in various proportions were introduced into the microchip via a syringe pump, and the immunocomplex was electrophoretically separated from the free Cy5-BSA on the microchip. Based on the relative intensity of the free and complex BSA, the binding constant of BSA and anti-BSA was estimated as 3.3 x 10⁷ M^-1. Furthermore, a C18 microcartridge (20 μL) was connected to the hydrodynamic inlet of the microchip. Using this device, the wash/elution step can be integrated on-line with the electrophoretic separation and detection on the microchip. Results show that the calibration curve of Cy5-BSA obtained from this integrated device has an R² value greater than 0.99 and a minimum of quantitation at ∼10 ng. This direct sampling method is another means of subfractionation, resulting in a relatively greater concentration factor than the average concentration of the whole fraction. Moreover, the electrical field-free bed ensures that the protein interaction will not be affected by the electric field during the wash/elution step. © 2002 American Chemical Society