NMR technology has become more accessible, but researchers want to expand its role in discovery by increasing its sensitivity and throughput.
Nuclear magnetic resonance (NMR), a spectroscopic technique used to deduce facts about the structure of targeted molecules, has played a key role in drug discovery over the years. “Any chemist would tell you that the kind of information that you get from NMR is probably the single most valuable piece of analytical data in terms of structural analysis on a small molecule. That application, in terms of medicinal chemistry support, has been around the industry a lot longer than I have,” says Michael Reily, PhD, who has been working with NMR for 17 years.
Reily, research fellow, Pfizer Global Research & Development, Ann Arbor, Mich., says he has done NMR research ranging from small-molecule analysis, to protein structure, to his present work, metabonomics. Over the years, there have been a number of advances that made NMR more powerful and sensitive while making the systems more compact, including the development of stronger magnets and shielded magnets, and the emergence of probes cooled to cryogenic temperatures.
Due to those advances, NMR technology has become more accessible, Reily says. “When I started here at Pfizer, you had to submit your sample, the technician would run it, look at the spectrum, sign the data, and turn it back to the chemist. Nowadays, with automation and relatively available high-field equipment that can fit into smaller spaces, chemists can walk up directly with their sample, drop it in, and have a high-resolution, high-field proton spectrum sent to their desktop.”
Born in the ’50s
NMR has been around since the 1950s in the chemical industry. In the 1970s, researchers started to become more interested in liquid NMR with proton detection, and the technology improved so much that it became more applicable to a laboratory setting, says Timothy Peck, PhD, division president of MRM Corp., Savoy, Ill. The company, which is a division of Protasis Corp., Marlboro, Mass., produces NMR probes and sample loaders.
During the late 1980s and early 1990s, scientists began to realize that the information provided by NMR, such as atom-to-atom connectivity, was going to be very useful in studying protein-structure function and dynamics, says Iain Green, PhD, product development marketing manager at Varian Inc., Palo Alto, Calif. Varian and Bruker BioSpin Corp., Billerica, Mass., are world leaders in NMR systems. “For many years, up to the year 2000 or so, the development of higher and higher frequency NMR systems, which gave better sensitivity and better peak dispersion, was one of the fundamental driving forces for NMR for biomolecular studies,” says Green.
NMR became very important, in many ways more important than X-ray or mass spectrometry, in investigating proteins as drug targets. That’s because it can be used to understand not only the structure of proteins, which determines where the binding sites are positioned, but also protein function and dynamics, says Green. NMR produces data on how proteins fold or move in cells and also how they fold around potential drugs or drug metabolites or other foreign species in the body. Now that much more is known about protein families, tools such as mass spectrometry and X-ray spectrometry are playing more dominant roles because of their speed and sensitivity, while NMR is moving back to its more traditional role in small-molecule analysis.
As the role of NMR in discovery has evolved, so too has the technology. “The availability of higher-field magnets has been key, but I think equally important has been the advent of shielded magnet systems, so you can have a much smaller magnetic fringe-field footprint and can actually site a 600-MHz NMR in a much smaller space than you used to be able to,” says Reily. While past systems required 1,600 square feet, some newer ones require only about 400 square feet. “Back in the 1980s, it was a pretty big deal if you had a magnet that had a field value of 7 tesla,” says Peck, “and now it’s very common to have magnets with over 20 tesla. The reason that’s helpful is that the signal quality basically goes up as field squared, so there are big advances there.”
Clemens Anklin, PhD, vice president of the NMR applications group, Bruker BioSpin, says Bruker is continually working on developing magnets with stronger fields. A few years ago, the company introduced a 900-MHz system, and now they and others are working to develop a 1,000-MHz (or 1 GHz) system. Although Bruker and Varian offer systems in the 800- and 900-MHz range and offer sensitivity needed for biomolecular and metabolite research groups, not everyone needs that power, says Green.
The bulk of NMR analysis is used to complement mass spectrometry, liquid chromatography/mass spectrometry, and gas chromatography/mass spectrometry—cases in which high-field magnets are not necessarily needed and 400- and 500-MHz systems will suffice. When Varian introduced its new NMR system early last year, the company focused on features that would help the routine chemist perform more definitive experiments using push-button technologies, says Green. “We’ve focused on things like improving the baseline and improving techniques that we call selective excitation. Selective excitation techniques focus on speeding up more highly deterministic experiments by using what we call shaped pulses.”
Another innovation that has improved the sensitivity of NMR is cryogenically cooled probes. The first of these was the CryoProbe from Bruker, which features 500-MHz probes in two configurations for 5 mm samples. The first installation of these units were completed in 1999, followed in 2000 by the introduction of a 600-MHz CryoProbe. Why is a colder probe better? A probe is a tube in which the sample is placed before it is slid into the magnet, says Green. The sample is then sitting in the magnet’s field and the probe contains electronics which pulses the radiofrequency field. This makes the nuclei absorb and dissipate energy; this dissipation is measured as NMR signals. A probe essentially provides a characteristic called sensitivity, which is the measure of the signal, divided by the measure of the noise inherent in the system.
“As you go higher in frequency from a 500-MHz magnet to an 800-MHz magnet, you will gain signal and your noise will remain approximately the same, so overall you gain sensitivity,” Green says. But that sensitivity can be affected when probes are room temperature because of the noise that the electronics emit.
In CryoProbes, the probe or sensor is cooled to cryogenic temperatures with liquid nitrogen, as is the preamplifier part of the electronics, Anklin says. “These two components bring an improvement in sensitivity and also a massive reduction in noise, so overall your signal-to-noise ratio, the sensitivity of the instruments, gets better. We’re typically looking at a factor of four or better in improvement.”
Since their introduction, Cryo-Probes have become popular with researchers working in biomolecular NMR and those performing studies of low-molecular-weight compounds where sample quantities may be in the low-microgram range. Green says that improved sensitivity is particularly important to those studying proteins or metabolites. Also, the trend toward miniaturization in chemical synthesis and screening increased the demand for the analytical sensitivity provided by tools such as the CryoProbe. Since Bruker introduced CryoProbes, other companies such as Varian have begun to offer cryogenically cooled probes.
Automation is the latest trend in the evolution of NMR. Protasis has been touting its One-Minute NMR automation platform, something that Peck says realizes the scaling advantages in performance, efficiency, and ease of use that accompany size reduction. “NMR has traditionally been more in a ‘V10-Hummer’ mentality, where the instrument has been designed to do anything and everything imaginable and have lots of horsepower in reserve at the price of requiring considerably more sample and reagent than other analogous techniques such as mass spectrometry, and not being particularly well automated.”
Bernhard Geierstanger, PhD, NMR group leader at the Genomics Institute of the Novartis Research Foundation, San Diego, published a paper in Analytical Chemistry on NMR automation. His team focused on a micro-flow CapNMR probe from Protasis, a capillary probe featuring small radiofrequency microcoils that wrap directly around the NMR flow cell. They put the probe on a 400-MHz NMR spectrometer and connected it to an automated liquid handler.
They then used this setup to perform two tasks: open-access analysis of single samples from LC/MS vials and high-throughput analysis and quality control of registered medicinal chemistry compounds and compound libraries from 384-well plates. When considering open-access analysis of single samples, they found it was possible to obtain LC/MS and NMR data from the same vial using 20 µL of 3 mM solutions of compounds. They also found that only 10 µL of these samples is required for NMR analysis and that they could accomplish a throughput of about 130 samples daily for quality control of medicinal chemistry and library compounds.
Finally, they concluded that the main advantage of this approach was that 10 µL of samples can be prepared in and measured automatically from 384-well plates that are filled robotically. “It gives us good results in a time frame that we can feel comfortable with, and with the manpower issue, it would be a lot more work to make hundreds of samples or plates full of samples by hand in tubes and recover material.”
While such research and other advances in technologies and tools have helped the advance of NMR, sensitivity is still an issue, Geierstanger says. “The fact that you can’t work with micromolar concentrations of materials, I think that’s a real issue.”
Reily agrees, saying sensitivity will always be an issue for NMR spectroscopy: “We’ve had an almost quantum leap with the introduction of the CryoProbes, but when you compare NMR with something like mass spectrometry, it is much less sensitive. In terms of getting as much information as you can out of a sample, multiple technologies will always be necessary.” Reily also believes NMR will become a stable alternative-screening paradigm. For example, in certain cases where targets aren’t amenable to traditional high-throughput readouts, NMR spectroscopy can provide an immediate answer as to whether a small molecule is binding to a protein or not. “You’ll never run a million compounds with it, but in cases where things aren’t amenable to traditional HTS, NMR can really provide a completely different window in terms of ligand binding.”
Filed Under: Drug Discovery