Expert Advice: Performing and analyzing nuclear magnetic resonance experiments

Reviewed for scientific accuracy by Gary Shaw, PhD

Since its first development in the 1950s, nuclear magnetic resonance (NMR) spectroscopy has been a valuable tool for elucidating the structural and dynamic characteristics of molecules, such as small organic compounds and large biological macromolecules. In an NMR experiment, molecules are exposed to a strong magnetic field, causing certain atomic nuclei within the molecules to behave like small magnets. These atomic nuclei absorb and emit energy at specific radiofrequencies. The NMR spectrometer captures these signals and converts them into NMR spectra, which are graphs depicting signal intensity across different frequencies. The positions and patterns of peaks in these spectra provide information about the nuclei’s local chemical environment, revealing the molecule's structural arrangement.

Driven by an interest in medicine and biology, Gary Shaw, a biochemist at Western University, has used NMR since the 1980s. He started by investigating the chemical structures of small organic molecules and transitioned to elucidating the three-dimensional (3D) structures of large macromolecules. The 3D structures of proteins define their molecular activities and roles in human health and disease. 

“It’s very much like a puzzle,” Shaw said. “You acquire some data, and there's only one way to solve the data.” With a shared passion for solving puzzles, Shaw’s research team utilizes NMR spectroscopy to investigate the structural properties and conformational changes of calcium binding proteins in muscle contraction. Over the years, rapid advances in NMR have enabled them to closely examine the underlying mechanisms that govern critical biological processes, including the ubiquitin-mediated degradation pathway in Parkinson's disease and calcium signaling in membrane repair. 

What information does NMR provide about proteins?

Using NMR to study proteins allows us to do a wide range of experiments, such as determining proteins’ 3D structures. We also use data to investigate the interactions between proteins and small molecules like drugs, DNA, and other proteins. Additionally, NMR spectra offer insights into protein dynamics, such as how fast the protein moves in solution and what conformational changes it undergoes. The NMR method is non-destructive, especially when studying stable proteins. This means that after we collect data on our favorite protein and take the protein out of the NMR spectrometer, the protein is still in solution and available for other experiments. 

What are the steps involved in an NMR experiment?

The first step is cloning the protein gene into a plasmid and expressing the protein from bacteria, typically Escherichia coli. During this step, we introduce isotopes by adding ¹⁵N-labeled ammonium chloride or ¹³C labeled-glucose into the media. This allows the bacteria to synthesize proteins carrying these desired isotopes. Then we purify the isotopically labeled protein and prepare it in a buffer with a neutral pH.

We then place the protein in a small five-millimeter NMR tube with a sample volume of 600 microliters and insert the tube into the superconducting magnet of the NMR spectrometer. To assess the data quality, we first acquire a simple proton spectrum. The proton spectrum displays signals, each corresponding to specific protons in the sample. Achieving a high quality proton spectrum may involve iterative adjustments to the sample, such as modifying pH or adjusting salt and protein concentrations, to ensure well resolved peaks. Once we’ve got an optimal proton spectrum, the sample is ready to perform a variety of NMR experiments depending on what information we want to find out. 

What are the key considerations when preparing protein samples for NMR analysis?

Although we usually start with a neutral pH that resembles biological conditions, adjustments in pH may be necessary to enhance protein solubility and behaviour, depending on the protein's acidity or basicity. This could involve lowering the pH to around five or six or raising it to eight. Protein concentration is also a consideration since NMR is a relatively insensitive technique. We often analyze protein samples at higher concentrations than those used in other techniques like absorbance spectroscopy. Sometimes, we see signs of aggregation in the spectrum. To address this, we decrease the concentration or change other sample conditions to achieve a homogeneous sample and obtain a more desirable result.

How can researchers analyze and improve data quality during NMR analysis?

Once we acquire a proton spectrum, we assess the spectrum's quality by examining the widths and intensities of the signals. To gauge the widths, we typically put one or two standard agents in the protein samples. Since water is the most concentrated component in protein samples, we apply specific radiofrequency pulses during data acquisition to suppress its signal while preserving the narrowest possible protein signals. 

To optimize the spectra, we spend considerable time shimming the magnet. Shimming is a process of making small changes to the coils surrounding the sample by adjusting the current passing through each coil. Shimming ensures that the magnetic field felt by the sample, which is a cylindrical structure, remains uniform. These small adjustments can significantly influence the spectrum's quality. It's common for researchers to dedicate a few hours to shimming for long experiments. This initial investment to achieve the best quality spectrum will have a big impact on the overall data quality at the end of the experiment.

How can researchers interpret NMR spectra to derive structural information about proteins?

If the goal is to determine the 3D structure of a protein using NMR data, the task is assigning specific signals in the spectrum to the corresponding atoms in the protein, including protons, ¹³C, and ¹⁵N. This could be challenging, especially for larger proteins. For instance, a 100-amino-acid protein may have over 100 nitrogens, around 500 carbons, and potentially over 1000 protons. To achieve this, researchers must identify the signals for each nitrogen, carbon, and proton atom across a large number of collected NMR spectra. This process involves extensive analysis and data interpretation on a computer, like solving a complex puzzle that demands time and experience. Depending on the nature of the data, the spectrum quality, and protein size, the process could take a couple of days to months.

Over the years, the assignment of NMR spectra has benefited from an increasing number of fully assigned proteins. This means that researchers have identified all proton, nitrogen, and carbon resonances in their proteins, and deposited this data into the Biological Magnetic Resonance Data Bank. This database can help researchers more quickly identify the signals in their spectrum. There are also effective software and automation tools available for interpreting NMR data. I believe we are getting much closer to fully automated NMR spectra analysis.

Are there any emerging advances in protein NMR techniques that researchers should be aware of?

The electronics for many NMR systems are getting better, which helps improve instrument sensitivity. Simultaneously, there is a trend toward creating and implementing higher magnetic fields. The higher the magnetic field, the better the spectral dispersion. Complementing these developments, recently developed helium- and nitrogen-cooled probes effectively suppress noise in the spectra, providing researchers with better sensitivity. Scientists around the world are always designing new NMR experiments. Some of these novel experiments offer distinct advantages in data acquisition, contributing to an ongoing improvement of spectrum quality.

What advice would you give to researchers who are new to the NMR technique?

It seems many people get scared when hearing about techniques like NMR, often seeing it as a complex and difficult method. While this is true, the theory, principles, and experiments associated with NMR are incredibly well developed. Researchers have reached a point where they can design experiments precisely to see the exact nuclei they want to see in an NMR spectrum. This makes NMR unique compared to many other techniques. It’s very much an exact science. You acquire the data, and the answer is sitting right in front of you. There's only one way to figure it out. Once you’ve got the correct answer, it’s very rewarding. 

This interview has been condensed and edited for clarity.