We report the demonstration of miniaturized capillary isoelectric focusing (CIEF) in plastic microfluidic devices. Conventional CIEF technique was adapted to the microfluidic devices to separate proteins and to detect protein-protein interactions. Both acidic and basic proteins with isoelectric points (pI) ranging from 5.4 to 11.0 were rapidly focused, mobilized, and detected in a 1.2 cm long channel (50 microm deep x 120 microm wide) with a total analysis time of 150 s. In a device with a focusing distance of 4.7 cm, the separation efficiency for a basic protein, lysozyme, was achieved as high as 1.5 x 10(5) plates, corresponding to 3.2 million plates per meter. We also experimentally confirmed that IEF resolution is essentially independent of focusing length when the applied voltage is kept the same and within a range that it does not cause Joule heating. Further, we demonstrated the use of miniaturized CIEF to study the interactions between two pairs of proteins, immunoglobulin G (IgG) with protein G and anti-six histidine (anti-6xHis) with 6xHis-tagged green fluorescent protein (GFP). Using this approach, protein-protein interactions can be detected for as little as 50 fmol of protein. We believe miniaturized CIEF is useful for studying protein-protein interactions when there is a difference in pI between a protein-protein complex and its constitutent proteins.

On-board generation of a set of calibration standards was demonstrated within a microfluidic device designed to perform immunoassay. Electrokinetic flow was used to proportionally mix the antibody (Ab) to bovine serum albumin (BSA) and a diluting buffer, to provide varying Ab concentrations for downstream mixing with fluorescently labeled BSA (BSA*). Mixing ratios were determined from electrical impedance modeling of the fluidic network using P-SPICE software, and peak heights for the labeled species were analyzed relative to the concentration calculated from the model. For dilution and separation of fluorescently labeled amino acids, a linear calibration curve was obtained for mixing ratios of 0.118 to 7.46. A linear calibration curve was obtained for the immunoassay calibration using dilution ratios between 0.197 and 5.077. Deviations were observed at larger extremes, possibly due to leakage effects at intersections. Peak height reproducibility was +/- 3% for the immunoassay, using diluted monoclonal Ab in mouse ascites fluid as the analyte. Recovery for on-chip calibration was 92 +/- 6% versus calibrants prepared off-chip, indicating a small bias.

No abstract