Promising to alleviate some of the shortcomings of 2D gels, Lynx Therapeutics is developing a two-dimensional microchannel electrophoresis system that may reverse — quite literally — a few conventions in the process.
The system, called Protein ProFiler, is still in development, but Lynx has already been drumming up interest at meetings, including this year’s HUPO conference in San Diego. It aims to improve reproducibility, protein recovery, resolution, sensitivity, and signal dynamic range for differential protein expression studies. “We are very excited by the prospect of this system [eventually] replacing all 2D electrophoresis,” said John Wiktorowicz, director of proteomics at Lynx and a former group leader for bioseparation at Applied Biosystems. But as of now, no data demonstrating the system’s capabilities have been published in a scientific journal, and Lynx is currently seeking collaborators to evaluate the technology.
“We did have the luxury of having a clean sheet of paper” to start, said Wiktorowicz, one of the inventors of the new electrophoresis apparatus, which is protected by two patents. The device itself consists of two 6 x 6 inch sealed glass plates with etched inner surfaces, thus forming a system of intersecting internal channels of 100 microns in diameter. A single channel, filled with a liquid gel of entangled polymers, defines the first electrophoresis dimension, and in contrast to conventional 2D gels it separates proteins largely by size, not charge.
Samples are loaded from an injection complex at the edge of this channel; perpendicular to this and attached to it like teeth to a comb is an array of buffer-filled channels for the second dimension of separation: isoelectric focusing. Molecules covalently bound to the walls of these channels produce stable pH gradients.
Protein samples are fluorescently labeled prior to the run, an approach similar to Amersham Bioscience’s two-dimensional gel electrophoresis labeling technique (2D DIGE). Three differently colored labels allow Lynx to run two protein samples simultaneously for differential analysis, as well as a panel of internal peptide standards.
The labeled proteins can be traced during the run and spotted at the end with a CCD camera. After they have reached their final positions in the pH gradient, bands selected for recovery are coupled to beads in the buffer that contain a photoreactive group. When the channel content is pushed out, the bead-coupled proteins are bound to a filter and can be identified by mass spectrometry, or subjected to other procedures like Western blotting.
According to Wiktorowicz, the system has several advantages over conventional 2D gel electrophoresis: separating the proteins in liquid phase instead of a crosslinked gel simplifies their recovery, and using intersecting channels means no movement of equipment is necessary to switch from one dimension to the other.
Heat dissipation in the channels is very efficient, allowing for higher voltages, better resolution and much shorter run times. Using isoelectric focusing as the second separation step instead of the first means that diffusion does not limit the final resolution and proteins will not run off the plate.
Furthermore, the gradient-forming chemicals are currently stable enough to perform about 10 runs, thus increasing reproducibility, Wiktorowicz said. Finally, the linear dynamic range of the fluorescent label can reach up to five orders of magnitude, more than other staining methods, and its sensitivity is 70 times higher — for half-second exposures — than that of silver staining. Since a researcher can monitor the proteins during the run, the switch to the second dimension can be made according to the position of the peptide standards.
However, Wiktorowicz admitted that some of the limitations of 2D gel electrophoresis remain, such as sensitivity to ion concentration and the tendency for some proteins to precipitate at their isoelectric point. But, he said, “everything is labeled, so you can watch what happens.”
Peter James, a professor at Lund University in Sweden who has heard Wiktorowicz give presenations on the technology, noted further that the system “suffers all the problems of standard 2D,” including problems with high and low pI proteins, large and small proteins, and membrane proteins.
So far Lynx has tested its apparatus mostly with well-characterized 1 kDa peptide standards. To analyze more complex protein samples, Wiktorowicz said, pre-fractionation is necessary to remove highly abundant proteins.
At the moment, Lynx is looking for biological examples to demonstrate the instrument’s capabilities, both in-house and through collaborations. Wiktorowicz declined to give a time-line for commercialization but noted that someone other than Lynx might be selling the instrument. “Our goal is of course to get it out there as soon as we possibly can because we recognize the need for it,” he said.