Most researchers trying their hands at protein arrays have resorted to spotters, widely used in the DNA array field, as their tool of choice for placing proteins at their assigned locations. The technology is reliable and fairly reproducible, but it suffers a few drawbacks: it requires access to a spotter, the arrays are not identical, and high-volume spotting can take a long time. To address this problem, at least three groups of researchers have developed an alternative that might be called sushi arrays. In this approach, protein samples fill the inside of channels or rods that are then sliced cross-sectionally like a rice roll, yielding hundreds of virtually identical arrays.
Robert Star, a senior investigator at the NIH and chief of its renal diagnostics and therapeutics unit, said he came up with his version of this technology — frozen protein arrays — more than a year ago while pondering how to screen a large number of protein samples without access to “high-tech, very expensive” spot arrayer equipment. Instead, he started with an ordinary block of tissue-embedding component, punched 25 steel needles into it, froze the block, and then pulled the needles out. After filling each of the remaining wells with three microliters of liquid protein solution, he refroze the block and cut it into 10 micrometer slices with a cryostat, a device routinely used in hospitals for making tissue sections.
Start said he gets up to 800 slices from a single block, with each slice containing spots of about 300 microns in diameter. Each spot contains 2.7 microliters of sample solution. To put the arrays to use, Start transfers the slices to a piece of Scotch tape, then to a glass-backed nitrocellulose slide, and keeps them frozen and desiccated until time for an experiment.
Star is publishing a proof-of-concept study on these arrays in an upcoming issue of the journal Proteomics. So far, he has been able to detect a purified protein down to a limit of 400 femtograms, as well as changes in concentration of a protein in a tissue homogenate on his arrays. “It’s like a dot blot,” Star said. “If you have a good antibody…this is a way of rapidly screening for a number of targets on the same set of samples.”
Apart from using the arrays to screen biological samples for the presence of proteins, Star hopes to perform biochemical or protein interaction assays with the technology, but has yet to test it for these applications. Manufacturing the chips while the proteins are frozen is “protein friendly,” he said, and might allow the proteins to retain their native fold. But several challenges remain: not all proteins can be expected to survive the freeze-thaw cycle in their functional form, and it is unclear if he can deposit enough protein on the membrane to do interaction studies.
For Star, however, the main advantage of the technology is that it’s cheap. “This is all very low-tech, using things that are commonly around in most labs,” he said. The technology is up for grabs for licensing from the NIH, which has filed a patent on it.
The NIH is not the only inventor of sushi-arrays: Mitsubishi Rayon of Japan has ventured to create DNA arrays using the technology, and Vacaville, Calif.-based Large Scale Biology Corporation has developed protein arrays based on the same principle.
LSBC Bundles Rods, But Waits for Partner
Using more of a California-roll type approach, LSBC has bundled polymer rods, each containing a different antibody or other assay reagent. These rods are connected with a polymer, and are sliced with a commercially available microtome. The rods serve as solid support systems; they are either filled with a porous material carrying the proteins or they come as hollow tubes coated with proteins on the inside, said Manfred Scholz, LSBC’s head of business development for the proteomics division. The company also has other undisclosed methods of loading the rods with probes, he said.
Like Star at NIH, LSBC also chose to develop sushi arrays because they cost relatively little to manufacture. Although the prototypes only consist of up to a few dozen rods of a few centimeters length, they could potentially be scaled up in number and length. “You make it by the meter and sell it by the micrometer,” Scholz said, adapting 3M’s business strategy for adhesive tape. But he cautioned that the technology is still years away from earning LSBC a profit: “The need for a low-cost array is not there yet,” he said. Once a clinical array is required, “then cost becomes the main pressure.”
LSBC, Scholz said, mainly developed the array to “detect small quantities [of proteins] in very diverse samples.” Thus far, LSBC researchers have only tested the array in a few model assays to measure protein concentrations. However, Scholz said he can imagine other uses, such as screening chemical libraries for compounds that bind to proteins on the array. The company’s patent application even claims the use of nucleic acids in the rods.
But LSBC will not pursue the technology further without a partner with access to a market that will help develop the array for a specific application, Scholz said, a decision the company only recently stated clearly. “For our own applications, we have decided we will not take it forward,” he said. “It will only go forward if we have a partner.”