June 15, 2005 (Vol. 25, No. 12)
NMR Spectroscopy Proving Value as Supplement to X-Ray Chrystallography
Not so long ago, structural proteomics research was divided into two distinct campsx-ray crystallography and NMR spectroscopy. X-ray crystallography remains the gold standard for structural determination, but NMR capabilities are proving their value for proteins that can’t be crystallized and in screening work.
For x-ray crystallography, there are two major optionsget the sample to a synchrotron beam line or conduct the work in one’s own lab. For individual lab equipment, the trends are toward robustness and automation.
Stand-alone lab systems, like the MicroStar-H rotating anode generator from Bruker AXS (www.bruker-axs.com), let researchers perform much or all of the crystallography work that previously required traveling to a synchrotron beam line facility in their own labs, saving time and money.
“The MicroStar-H delivers x-ray intensities comparable to those of a second-generation synchrotron generator,” according to Roger Durst, Ph.D., CTO, which is an order-of-magnitude increase when compared to previous models.
That increase in sensitivity lets users solve structures in their own labs and actually provides greater stability, albeit less brightness, than the strong synchrotron beam, which fluctuates. Another advantage, Dr. Durst says, is that the MicroStar-H can solve structures using native protein.
Selenomethionine is not required, “which is a big advantage,” he says, so researchers can determine the location of sulfur atoms in the structure directly by using single wavelength anomalous diffraction (SAD).
Bruker also launched a new, high-sensitivity detector called Platinum, which regulates the x-ray spectra diffraction path. Readout time for a complete set of data is one second100 times faster than an image plate detector. It is more sensitive than an image plate detector by “about a factor of eight,” Dr. Durst says.
In terms of efficiency, “charged-coupled devices like the Platinum have a significant advantage” in that they better detect the data at the outer edge of the image, by an order of magnitude.
In April, Bruker AXS opened its X-Ray Life Sciences Laboratory for protein crystallization and macromolecular structure determination in Delft, The Netherlands. “If you can’t get a selenomat version of your protein, use ours to run a native SAD structure,” Dr. Durst offers.
Rigaku/MSC (www.rigakumsc.com), a supplier of instrumentation for macromolecular crystallography, is advancing integration and automation with several new platforms, leveraging the technologies of recently acquired automation tools manufacturer RoboDesign and macromolecular software developer Molecular Images.
According to Joe Ferrara, Ph.D., vp, product marketing and CSO, “We will use the technology completely to provide a full vertical integration for tools to solve protein structures,” thus expanding Rigaku/MSC’s capabilities in the areas of instrumentation “from crystallization of proteins all the way to refined structures.”
This summer, Rigaku/MSC will install the first AGENT (Actor Gantry Enabling Numerous Targets), which currently automates two, and within two years up to eight, ACTOR (Automated Crystal Transport, Orientation and Retrieval) detector systems simultaneously with an ACTOR robot.
By reducing the need for crystallographers to interact with the equipment, throughput is increased by “a factor of four to six in a facility operating 24/7,” Dr. Ferrara says.
Rigaku/MSC has introduced the Proteros Free Mounting System, which controls protein hydration in the crystal lattice, and thereby “optimizes a sample to a degree not normally possible in terms of resolution, mosaicity, and anisotropy, which affect data quality,” Dr. Ferrara says.
With this system, he explains, the crystals have little “mother liquor,” which reduces x-ray background noise. Consequently, researchers can flash cool crystals that otherwise couldn’t be flash cooled.
“This is exciting,” he says. “It offers a different way to optimize crystals and is orthogonal to the way people do things. This is a rescue techniquenot a high throughput technique,” Dr. Ferrara emphasizes, that helps researchers find the optimal conditions for data collection. Consequently, the Proteros Free Mounting System can help shave years off the time it takes to solve a protein structure, significantly improving overall throughput.
BSI Proteomics (www.bsiproteomics. com) has applied for a patent for its Membrane Protein Stabilization System, which crystallizes membrane proteins and works in concert with BSI’s dynamic crystallization automation system, dubbed ARD.
Dynamic crystallization technology allows the crystallization process to be started, altered, and reversed at will. ARD measures and controls conditions that are important for growth, such as pH, conductivity, and temperature.
The technology “can be used to crystallize molecules ranging in size from small organic compounds of about 1030 atoms up to large macromolecular complexes comprising tens of thousands of atoms,” according to the patent (U.S. patent 6,596,081) and can also be used to crystallize medium-sized molecules, including oligonucleotides and oligopeptides, and large macromolecular complexes, such as viruses.
Membrane-bound proteins can be crystallized by introducing a detergent into the crystallization via a micro-syringe to separate the protein from the membrane.
Bio-Xtal (www.bioxtal.com) is just beginning the crystallization phase of its Membrane Protein Network (MePnet) program. Now in its fourth year, the program is developing tools for the expression, purification, and crystallization of G-protein-coupled receptors (GPCRs).
The crystallography platform is being developed with three academic teams, but “the details are confidential at this stage,” according to Kenneth Lundstrom, Ph.D., CSO.
In the first phase of the project, “We studied 100 GPCRs in three expressions systems,” which resulted in “high expression levels of about 510 mg/L for approximately 60 GPCRs,” Dr. Lundstrom says. Typically, expression levels for membrane proteins have been much lower, he says. “This brings them up to levels applicable to structural biology.”
The goal, he continues, is to “produce recombinant GPCRs of better quantity and quality and to develop crystallization methods in a high throughput format.” Quality and quantity are both addressed by using rapid recombinant protein expression in mammalian cells.
“Using the replication-deficient Semliki Forest virus vectors [originally isolated from a virus found in the Semlike forest in Uganda], we can express target proteins in a broad range of host cells. Expression evaluation can be performed within a few weeks, and the scale-up for large-scale production is established.”
Bio-Xtal also is in the early stages of developing selenomethionine labeling for mammalian cells, as a way of attaching an isotope label to the protein of interest. “Methionine is a natural amino acid that cells need for their growth, so it fools them into taking up the radioisotope,” Dr. Lundstrom explains.
Mammalian cells are sensitive to exposure to toxic compounds, such as selenomethionine, which has made the labeling process difficult and inefficient. The Semlike Forest virus essentially steals the cell’s synthesis machinery and begins producing recombinant protein very rapidly, thus improving the chance of incorporating the selenomethionine.
Beam Line Services
When researcher need a synchrotron beam line, they have the option of either contacting synchrotron facilities (such as Argonne National Laboratory in the U.S., Daresbury Laboratory in the U.K., or the European Synchrotron Radiation Facility in France and SLS at the Paul Sherer Institute in Switzerland) or working through companies like Shamrock Structures (www. shamrockstructures.com), which provide access at some of these facilities and also additional services.
Shamrock Structures, formed near Argonne National Laboratory in 2003, has “built a growing client base,” according to CEO Steve Schiltz, by providing “the flexibility to choose from a range of capabilities” that includes complete structure solutions from proteomic or genomic sources, or point-to-point steps along the way.
Shamrock’s key competencies are focused on its dedicated access to the third-generation synchrotron beam line at Argonne National Laboratory. Researchers, Schiltz says, may either mail their samples to Shamrock through the company’s Crystal Express system or purchase beam time and come themselves. The Chicago area is fast gaining proteomics capabilities, he says, citing the new 900-MHz, actively shielded NMR equipment recently installed at the University of Illinois.
Membrane proteins, proteins with large disordered sections, and protein complexes, however, do not readily form crystals. “NMR spectroscopy always will be the primary method of structure determination when the protein does not form crystals and is of the right size for NMR, according to Clemens Anklin, Ph.D., vp, applications/training, Bruker BioSpin (www.bruker-biospin. com). Additionally, “NMR has proven valuable in providing information on protein dynamics and protein folding,” he says.
NMR capabilities have increased significantly, and Bruker BioSpin is currently concentrating on increasing both magnet field strength and detection sensitivity. A 1-GHz magnet is on the drawing board. “There is a lot of materials research involved,” Dr. Anklin says, so he won’t speculate on when it will be ready for commercialization. The benefit, however, may be the ability to study proteins larger than the 25 to 40 kDa protein sizes possible with the currently available magnets (up to 900 MHz).
Bruker BioSpin recently installed the first actively shielded 900 MHz NMR magnet, the 900 US2 at the University of Illinois at Chicago. “Active shielding reduces the stray magnet field, so you can set up more magnets in a facility,” he says, and such other equipment as HPLC, mass spectrometers, and liquids handlers can be placed closer to the NMR.
The active shielding also enhances personnel safety, he says, noting that magnetic fields from unshielded magnets may extend two floors above and below the equipment, making those spaces off limits to persons with implanted medical devices such as pacemakers.
Bruker BioSpin also introduced the 850 US2 MHz wide bore actively shielded magnet NRM, which has a large, 89-mm opening for solid samples. The magnet is based on the already-released 900 US2 MHz standard bore magnet
The most significant recent increase in sensitivity can be traced to the development of cryo-probes. After years of delivering cryo-probes used mainly for protein research, Bruker BioSpin is expanding its range of products to include models designed specifically for nucleic acids, chemical biology, systems biology, metabolomics, and toxicology studies. It introduced the first microimaging cryo-probe this year.
Korea-based CrystalGenomics is exploring structural chemoproteomics. “It is similar to proteomics, but has a different concept,” says Seonggu Ro, Ph.D., vp, technology. “Structural chemoproteomics focuses not on the protein structure but on the protein/chemical interaction.”
In this field, target proteins are sorted into their gene families using sequence homology. Structural analysis yields the folding patterns of the protein family and the structural information of the active sites or the interaction sites. Virtual screening then identifies master scaffolds that are chemical moieties to recognize the common structure of the active sites.
“We have already identified many master scaffolds for several protein families, including kinases, phosphodiesterases, and peroxisome proliferator activated receptors,” Dr. Ro says. To discover drug candidates with high activity and selectivity to a specific target protein, CrystalGenomics has optimized these scaffolds on the basis of the structural chemoproteomic studies of the corresponding protein family.
CrystalGenomics also provides services in the area of x-ray crystallography for protein/ligand complexes and lead generation.