There is a strong interest in the simultaneous detection of a number of different proteins in a single biological sample. Limitations on the size of the collected sample require that these measurements be done on as small a volume of fluid as possible. This interest has been one of the driving forces behind the development of microfluidic devices for biomedical applications. The move to these smaller-scale systems has a number of advantages. First, they are capable of analyzing smaller volumes of sample. Second, in applications such as capillary electrophoresis, the microfluidic system can achieve the same separation resolution in much less time than a larger-scale system. Finally, the reduced size of the analysis setup raises the possibility of developing portable analytical devices. In collaboration with scientists at NIST, DBEPS is developing a microfluidic device for immunoaffinity electrophoresis, in which multiple proteins will be simultaneously isolated and detected. Using the microfabrication facilities at NIST, we are able to make micrometer-scale glass-encapsulated microfluidic systems with any desired two-dimensional configuration. The prototype devices consist of a long glass-encapsulated channel, 50mm x 15 mm x 30cm, with a serpentine pattern. Side ports are used for pressure-driven loading of different biotinylated antibodies into each segment of the channel; these antibodies bind to streptavidin that has been covalently linked to the channel walls. After the antibodies have been immobilized, the sample under analysis flows through the entire device. Electrical control of the sample flow permits adjustment of the residence time in each segment in order to optimize binding. Initially, the captured proteins will be detected optically, using fluorescent tags. After detection, the captured proteins may be eluted from the channel for further analysis. The channel device architecture has several advantages over existing array technology: the proteins are detected by single-point capture, and much smaller sample volumes can be used. In addition, we hope to be able to reuse the channels with the bound antibodies for multiple samples. The immediate clinical need that motivates this project is an epidemiological study of the immune response to Human Papillomavirus (HPV) infection, and the relationship between this response and the development of cervical cancer. Studies of lymphocytes taken from peripheral blood samples have suggested a difference between patients whose cells respond to HPV peptides with a T lymphocyte helper cell type 1 response (Th1-type response) compared to those with a Th-2 type response. In order to better understand these differences at a molecular level, it is necessary to detect and quantify a number of selected immune regulatory molecules directly in cervical secretions, as a local probe, rather than blood samples as a systemic probe of the immune response. From these requirements, the need for a microfluidic device such as the one being developed becomes clear. The volume of cervical secretions that can be collected is quite small, typically 25-50 microliters, and the difficulty in collecting a sufficient number of fluid samples for an epidemiological study makes it imperative to extract as much information as possible from a single sample Conventional diagnostic techniques, such as ELISA, only allow for the analysis of one or two analytes from a sample of this size. The proposed device will be able to detect over a dozen proteins from a one-microliter sample volume. Although the functionality of the proposed device is being targeted for this particular application, there is every reason to expect that, once this device is realized, many other biological and clinical applications could also be addressed. The small length scale of microfluidic systems permits the use of smaller sample sizes, leading to decreased reagent costs,, as well as faster analysis times. Currently, the most common detection method used in microfluidic systems is fluorescence detection. While fluorescence offers high sensitivity, chemical labeling of analytes with a fluorophores is generally required. In addition, non-specific labeling and autofluorescence is often a problem. The use of patterned gold colloids in a microdevice for surface-enhanced Raman spectroscopy (SERS) detection has been achieved and offers an attractive potential alternative to fluorescence. This detection technique offers direct measurement of biomolecules in aqueous solution. In making multifunctional devices, there is a need to create spatial variations of the surface properties along the length of a sealed microchannel. The ability to tailor these properties could lead to a variety of applications, such as improved electroosmotic pumping from highly charged channel regions, or the targeted patterning of capture ligands. Covalently attached photoprotection groups, which are removed from the surface under exposure to ultraviolet light, have been attached to microfluidic channels to selectively tether charged groups to specific regions of the microdevice. Elastomeric microchannels have been used to temporarily create a microdevice on an atomic force microscopy (AFM) substrate. Because flow in microchannels is laminar, all mixing is diffusive, which allows us to create a well-controlled buffer gradient, either in ionic strength or pH, over a few hundred microns. This permits a rapid sampling of buffer conditions for protein deposition on AFM substrates and offers the potential of reducing substantially the time required to optimize binding to the AFM substrate. Miniature ball lenses coupled to optical fibers have been used as a high numerical aperture system for fluorescent detection from microfluidic channels. Coupled to an imaging spectrograph to provide spectral separation, the fiber optic approach is readily multiplexed to report on several channels simultaneously.
