Figure 1: The Bruker BioSpin Avance 900 MHz NMR System.

After nearly four decades of advancements in both technology and methodology, Nuclear Magnetic Resonance (NMR) spectroscopy now plays an important role in a wide range of research fields, including structural proteomics. This progress results from tight collaboration between manufacturers and the researchers who develop methods.
Initially, the use of NMR was limited to physics. Improvements in sensitivity and resolution made it a useful method for chemists. The introduction of Fourier transform (FT) NMR and, about a decade later, two- dimensional FT-NMR, allowed the study of larger, more complex compounds. In the 1970s, the first attempts to study biomolecules by NMR resulted in useful structural information, facilitating its use as a tool for biochemists and molecular biologists. Methods development over the last 20 years, together with the availability of higher magnetic fields, has expanded the application of NMR to its current role in structural proteomics. The study of proteins with molecular weights over 100 kDa is made a reality.
For all molecules that readily form single crystals, x-ray crystallography will always be the method of choice, and cryo Electron Microscopy provides spectacular data of very large assemblies. NMR, on the other hand, is uniquely suited to deliver structural information on molecules that are difficult to crystallize. NMR is also preferred for obtaining information about the dynamics of biological molecules.

Obtaining structural information
Structural information from NMR spectra can be mainly obtained from two observables — spin coupling and the nuclear Overhauser effect. Spin coupling acts through chemical bonds and primarily provides information about connectivity within molecules. A large variety of experiments have been developed to establish correlations between atoms that exhibit coupling. These will eventually allow assignment of every signal in a spectrum to a specific atom in the molecule. Because the magnitude of the coupling is affected by bond angles, information about relative orientation of bonds can be obtained and used for the calculation of structures. The nuclear Overhauser effect acts through space. Its magnitude is proportional to the distance between two atoms. With the assignment information obtained with these two methods, the spatial proximity of two atoms can be established and constraints for structural calculations can be obtained.

Figure 2: Bruker BioSpin Avance ICE 600 MHz — Integrated CryoProbe NMR System.

Leaders in the application of NMR
Much of the seminal work on the determination of protein structures by NMR was done in the laboratory of Dr. Kurt Wuethrich at the Swiss Federal Institute of Technology (ETH) in Zurich, Switzerland. This work was honored when Professor Wuethrich was awarded the 2002 Nobel Prize in Chemistry. In the early 1980s, one of the essential technological factors for the success of this work was the introduction of magnets with a field strength of 11.7 T corresponding to a proton resonance frequency of 500 MHz. To build such magnets, the processing of a new superconducting alloy, NbSn3, had to be mastered. This alloy is so brittle that it cannot be wound into coils. The only viable solution is to use a wire made of Niobium filaments in a Cu/Sn bronze matrix, wind it into a coil, and then to treat this material over long periods at elevated temperatures.
The development of experimental techniques using two-dimensional NMR, together with the higher fields, allowed protein structures with molecular weights as high as 10 — 15 kDa to be determined in the 1980s. The major factor preventing the study of larger molecules was increasing spectral complexity. Later in that decade, the use of proteins uniformly labeled with the NMR active nuclei, carbon-13 and nitrogen-15, enabled the introduction of triple resonance experiments, pioneered by NIH researcher Ad Bax and his coworkers on a Bruker AM600 MHz spectrometer. These experiments used the separation of the resonance signals along multiple chemical shift dimensions to resolve signals that are overlapped when only observing hydrogen. They also allow the collection of sequential information along the polypeptide chain by unambiguously correlating hydrogens, carbons and nitrogens, both with one amino acid and across the peptide bond.

Figure 3: Magnetic Shielding Diagram compares NMR Magnets without shielding, and with Bruker Biospin UltraShield and UltraShield Plus shielding.

Higher field strengths, higher sensitivity
The next major breakthrough in NMR technology came in the early 1990s when the next generation of NMR magnets was introduced — first the 750 MHz and later the 800 MHz systems. Higher magnetic fields increase the dispersion of the signals and lead to higher sensitivity. These two factors again allow the study of larger proteins. With these systems, NMR can easily study proteins up to 20 — 25 kDa. One of the challenges of ultra high field magnets is the need for an extremely stable system that displays no fluctuations in the magnetic field strength. Bruker BioSpin Corporation achieved this by implementing a unique system in which a small cooling device, based on the Joule-Thomson effect, cools the entire coil to a temperature near 2 Kelvin. With this design, the helium vessel itself stays at regular atmospheric pressure and can be operated like traditional lower field magnets. In 2001, this technology allowed the first customer installation of a Bruker BioSpin Avance 900 MHz spectrometer, at The Scripps Research Institute in La Jolla, California. Research is underway in several laboratories and companies to reach the next level of magnetic fields.

Advances in shielding technology
Active magnet shielding technology, which reduces the stray fields around the superconducting magnet systems, has also played an important role in advancing the application of NMR. Reducing the stray field allows easier system siting and permits the use of equipment such as HPLCs, mass spectrometers and liquid handling instruments that would otherwise be affected by the magnet. Actively shielded magnets are currently available up to 800 MHz. The shielded high-field magnets exhibit a smaller stray field than a traditional 500 MHz magnet.

Cryogenics for a quantum leap in sensitivity
After years of incremental increases in sensitivity, the introduction of cryogenically cooled probes represented a quantum leap for NMR spectroscopy. These probes, with cold Rf coils and preamplfiers, increased sensitivity by a factor of four or more. The resulting increase in throughput and lower detection limits allows entirely new approaches to NMR. In the case of structural proteomics this means that data on a sample can be collected in less than one-tenth of the time or that a concentration one order of magnitude lower can be used. This addresses two of the major shortcomings of NMR spectroscopy that had persisted up to this point. In the past, it could take as long as several weeks to collect the data required for assignment and structural calculation of a protein. With a CryoProbe (Bruker BioSpin), this can be reduced in the best case to less than 24 hours. Using automated analysis routines, Dr. Gaetano Montelione of Rutgers University has shown that an actual protein structure can be obtained in less than one day. Working with lower concentration eliminates time-consuming optimization of solution conditions and often prevents the protein from forming aggregates. The benefits of the CryoProbes were recognized early on by Dr. Steve Fesik from Abbott Laboratories. The higher sensitivity allowed him to increase the throughput for his SAR by NMR method by a factor of ten.

Figure 4: What NMR systems used to look like, HX-90 NMR.

Studying larger proteins
At this point NMR was still limited to small- to medium-sized proteins for structural studies. Molecules larger than 30 — 40 kDa display very unfavorable relaxation behavior, leading to signal losses that make many of the experiments fail. The introduction of Transverse Relaxation Optimized Spectroscopy (TROSY) in 1999, by Dr. K. Wuethrich and his coworker, K. Pervushin, broke this barrier and made the study of much larger proteins possible. Proteins weighing in excess of 100 kDa have now been studied. These advances have also been facilitated by new methods in sample preparation. Replacement of most hydrogen atoms in a protein by deuterium, directed partial isotopic labeling, and segmental labeling have been used to simplify the spectra of very large molecules, further aiding the analysis of the data from very large proteins.

Conclusion
In the early days of structural proteomics, or structure determination by NMR as it was referred to then, obtaining a single structure could take over a year. Today, the same structure can be determined in a matter of days. This has resulted in NMR changing from a research method used by a few to a vital tool for a large variety of scientists.
NMR technological developments continue. The push for higher fields continues and will eventually result in the 1 GHz spectrometer. This will not only give the scientist a more powerful instrument, but also will, as has been shown in the past, make lower field instruments more effective because many of the developments required for the high field instruments can be translated down to the other spectrometers. Research continues in the improved performance of both cryogenically cooled and conventional NMR probes. Scaling down measurement volumes will increase sensitivity and reduce sample requirements. In addition to these developments, mostly driven by instrument manufacturers, NMR users will develop new applications, methods and experiments.

About the author
Clemens Anklin, Ph.D., is Vice President, Applications/Training with Bruker BioSpin Corporation.
More information regarding the use of NMR in structural proteomics is available from: Bruker BioSpin Corp.

Use InfoLINK 4A1502 or Call 800-287-0633