Speeding Up Protein NMR

Recording all required NMR experiments for a protein structure calculation, dynamic analysis or even backbone assignment can be very time consuming. By using a combination of different techniques for data acquisition and processing, the time required to record almost all experimental data needed to characterize a protein can be shortened significantly.

Looking at a regular NMR experiment, most of the experimental time is spent waiting for the magnetization to recover between scans. This is determined by the spin-lattice relaxation time T1, which is usually of the order of one second for a backbone amide proton in a small- to medium-sized protein. This means that it would take about 5 seconds for 99 % of the magnetization to recover, whereas a usual recovery delay would be in the order of 1.5 seconds (~80 % recovery).

Multidimensional NMR experiments also take significantly longer as, using conventional techniques, introducing an additional dimension usually increases the time necessary to perform an experiment by an order of magnitude. This means that if a two dimensional experiment can be recorded within an hour, a 3D experiment will take a day, 4D a week, 5D a month, and so on. In terms of the signal-to-noise ratio, most of the time – especially when recording high dimensional spectra – a much shorter experiment time is sufficient.

When going to higher field strengths, it might be worth considering TROSY-based, rather than HSQC-based, experiments[5,6]. The final spectrum will not only show sharper lines, but the experiments also use lower duty cycles, since nitrogen is not decoupled during the acquisition. In addition, more recent work has shown that it is possible to recover steady state 15N magnetization to further increase the sensitivity.

Figure 1 speed up nmr

Figure 1: (left) SOFAST HMQC recorded in 20 seconds using a recovery delay of 10 ms and a 40 ms acquisition time, recording 512 x 64 complex points. (right) BEST TROSY recorded in 3 minutes using a recovery delay of 150 ms and 80 ms acquisition time recording 1024 x 128 complex points.

Daniel Mathieu
Bruker BioSpin GmbH, Rheinstetten, Germany