Courtesy of Fluidigm

Fluidigm's TOPAZ 1.96 screening chips employ microscale channels and valves for diffusive mixing of protein and crystallization reagents. Future chip designs will steadily increase parallel throughput.

Protein structure determination using X-ray crystallography typically suffers from two major bottlenecks: producing sufficient quantities of material, and finding appropriate crystallization conditions. The TOPAZ™ Crystallizer, released last year by microfluidics start-up Fluidigm http://www.fluidigm.com, addresses both of these concerns. Now the South San Francisco-based company has enhanced and expanded its TOPAZ line, providing reagents and newly designed chips as well as hardware to set up and image the crystallization reactions. It's all part of the company's strategy of providing "everything from screen to beam," says product manager Kristin Spataro.

Fluidigm's first-generation Crystallizer was a manual, free-interface diffusion (FID)-based system that allowed researchers to screen 48 crystallization conditions on a single chip, at three different concentration ratios, for a total...


The TOPAZ line now centers on the new 1.96 screening chip. Mounted in a microtiter plate-sized carrier, the chip can screen 96 crystallization conditions for a single protein (the "1" in "1.96"), using just 2 μl of protein. Spataro says the carrier format gives the company the flexibility to design chips to run as many as 768 reactions while testing 96 conditions on eight proteins, or 8.96. Such a high-density chip is not currently available, but a new 4.96 chip able to run 384 reactions will be available this quarter, says Spataro.

Fluidigm's chips are bilayered, gas-permeable consumables built of pliable, spongy polydimethylsiloxane. Protein and precipitants flow in the bottom layer, while the top layer contains control valves. Because they are built on a microtiter format, users can load these chips using an automated liquid-handling robot or multichannel pipettor. To actually start the reactions on the chips, though, they must use the new FID Crystallizer.

This instrument primes and sets up the chips, says Spataro, opening and closing the appropriate valves in sequence, first to load the reaction chambers, then to isolate them from one another, and finally to allow the protein and precipitant to mix. Each FID Crystallizer can process up to four plates simultaneously, and a single computer can control up to four instruments, allowing users to setup as many as 16 chips at one time.


Researchers can monitor each reaction's progress with Fluidigm's new AutoInspeX™ workstation. This instrument automatically images and scores each reaction for crystal growth. The chips feature alignment marks that make it easier for the system to focus on the growth chambers, and also to acquire and compare images over time. Typically users collect images on days 1, 2, 4, and 7, allowing them to watch the nucleation and growth of crystals, something that is "pretty difficult to do with conventional techniques," Spataro says.

She explains that imaging FID reactions on-chip is simpler than imaging drop-based crystallization methods such as vapor diffusion and microbatch. "There's an optical clarity that you get within the silicone that you can't get with plate-based technologies," Spataro says. "Our chips provide a unique, two-dimensional environment for imaging crystals that is not subject to the same lensing effects as 3-D drops with vapor diffusion and microbatch."

Rounding out Fluidigm's screening tools, the company has designed four sets of reagents, 96 solutions each, optimized for use with its system. According to Spataro, two of the sets are sparse matrices, one is a PEG/ion screen, and one is a high-salt screen.


Once researchers identify suitable crystallization conditions, they can migrate to Fluidigm's new "growth chip" to produce crystals of sufficient size for X-ray work. Featuring 27 chambers, the growth chips, mounted in the carrier housing, are thinner than the screening chips and can be mounted on an X-ray source or in a synchrotron directly.

The growth chips can help scientists determine if a crystal is protein or salt, and if it will diffract well, says Spataro. At that point, though, she recommends that a researcher open the chip, extract the crystal, and then mount it.

Brent Segelke, biomedical scientist at Lawrence Livermore National Laboratory, uses Fluidigm's original Crystallizer system in his work in the Tuberculosis Structural Genomics Consortium, but says he has not upgraded to the FID Crystallizer for budgetary reasons (the company declined to disclose pricing information). Segelke says Fluidigm's approach could have a major impact in structural genomics research because of the small amounts of protein it requires. "Crystallography is a huge burden on protein production," he says, citing a recent example in which his lab set up 384 vapor-diffusion trials. Each reaction required 0.5 μl of protein at 10 mg/ml, or about 1.9 mg protein total. Doing the same trial using FID, however, would have required only 180 micrograms, a 10-fold reduction.

And that thriftiness, Segelke says, could pave the way for massively parallel protein production and crystallization. "The amount you need at the end sets the bar for the amount you need to start with."

- Jeffrey M. Perkel

Interested in reading more?

Magaizne Cover

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?