Research / Advanced MR methodologies

Heteronuclear detection for difficult proteins

Some of the most biologically interesting proteins are the ones proton-detected NMR handles worst. The alternative has been known for years; what it needed was better control.

The problem

Almost all biomolecular NMR detects protons, for the good reason that ¹H has the highest gyromagnetic ratio of the common nuclei and therefore the best intrinsic sensitivity. This works well for folded, well-behaved proteins.

It works badly for two important classes. Intrinsically disordered proteins lack a fixed structure, so their amide protons all experience nearly the same chemical environment and their signals pile up in a narrow region — the spectrum collapses into an unresolvable heap. Worse, disordered chains exchange amide protons rapidly with solvent, which broadens the very signals you were trying to detect. Paramagnetic proteins have an unpaired electron whose enormous magnetic moment bleaches out the nuclei nearest to it — exactly the nuclei at the metal site you most want to observe.

Detecting ¹³C instead sidesteps both problems. Carbon chemical shifts stay dispersed even in a disordered chain, and ¹³C is far less affected by paramagnetic broadening and does not exchange with solvent. The catch is sensitivity: you have given up the proton's advantage, so every subsequent inefficiency in the experiment hurts more than it would otherwise.

The approach

A ¹³C-detected experiment still moves magnetization through ¹⁵N, and the ¹⁵N pulses are a weak link. Nitrogen has a low gyromagnetic ratio, so its pulses are long; at 1.2 GHz the ¹⁵N bandwidth that must be covered is large; and long, imperfect pulses give relaxation time to destroy signal that was already scarce.

With Roberto Pierattelli and Isabella Felli at CERM, University of Florence — who have driven ¹³C-detected biomolecular NMR for years — we designed ¹⁵N optimal control pulses tailored to this role: uniform over the full nitrogen bandwidth, tolerant of B₁ inhomogeneity, and short enough that relaxation losses stay manageable.

Schematic, not measured data. Why carbon detection is worth the sensitivity penalty for disordered proteins. The optimal control contribution is not visible in a cartoon like this — it is in making the ¹⁵N transfer steps efficient enough that the right-hand spectrum is achievable in a realistic amount of instrument time.

A parallel collaboration with Haribabu Arthanari at Harvard Medical School applies the same reasoning to ¹⁵N-detected experiments, where nitrogen is the observed nucleus rather than a relay — an even more demanding sensitivity regime.

Relevant paper: Schiavina, M., Joseph, D., Griesinger, C., Felli, I. C., Pierattelli, R., “¹⁵N optimal control pulses: an efficient approach to boost ¹³C-detected NMR experiments,” J. Magn. Reson. 381, 107972 (2025).