The basic heteronuclear single-quantum correlation exper-somewhat similar grounds, but due to differences in the iment (HSQC) has been proposed in diverse variations: with experimental approach, we arrive at different conclusions. coherence selection by pulsed field gradients (1), with sensi-This work concentrates on CHa signals, and, with the experitivity enhancement through the use of the technique of presments utilized, signals from CH 2 and CH 3 should not show ervation of equivalent pathways (2), or the combination of any change (7). This study was motivated by the wide both techniques (3). We have been investigating these varispread of natural-abundance 13 C protein studies: dynamic ous approaches in terms of overall sensitivity improvement. investigations by 13 C relaxation measurement or triple-reso-We report here the sensitivity figures obtained for 13 C natunance experiments. Indeed, these very lengthy experiments ral-abundance spectroscopy and show that sensitivity gains can benefit from sensitivity enhancement. better than 50% can be obtained under these conditions.
We worked on an 8 mM protein sample of Pike parval-Differences with previous studies are shown and discussed. bumin in solution in H 2 O, with 10% D 2 O. Parvalbumin is a
The recently introduced PEP technique (preservation of 109-residues calcium-binding protein, with two calcium sites equivalent pathways) (4, 5), when applied to the HSQC structured as two EF hands; the total mass is 11.5 kDa. The experiment (2, 6), permits one to optimize the sensitivity complete 1 H assignment and the 3D structure in solution of the experiment. This method permits a theoretical enhave been determined by Padilla et al. (9). This study has hancement of 2 of the detected signal by preserving magnebeen performed on the natural-abundance 1 H-13 C spectrum, tization from coherence levels which are otherwise diswhich has been recently assigned (10). carded. More recently (3, 7), it was proposed to adapt this All the experiments presented have been acquired in an technique to experiments which make use of pulsed field overall time of 9.2 h, on a Bruker AMX600 spectrometer, gradients (PFG) for coherence selection. With this addiequipped with a triple-channel Bruker probe with z gradient. tional enhancement, PFG filtering can be used in HSQC-Data sets are 2K t 2 points by 512 t 1 values. Spectral widths type experiments in which the sensitivity gain of the PEP were respectively 14.12 ppm for 1 H and 66.26 ppm for 13 C. technique is fully conserved. Such experiments should pres-A relaxation delay time of 0.8 s was used between each ent a 2 signal-to-noise enhancement when compared to scan. A delay of 3.4 ms was used for the refocusing of the regular HSQC experiments, and a gain of 2 when compared J coupling. All experiments were acquired with a four-step to regular PFG-filtered HSQC experiments.
phase cycle, permitting the cancellation of the 12 C-bound In the work presented here, a comparison of the relative protons. sensitivity of the different HSQC experiments has been done
The regular HSQC (11) was acquired with two FIDs per on a protein sample, in natural-abundance 13 C spectroscopy. t 1 value and 64 scans per FID. The sensitivity-enhanced Lately, in a remarkable work, Kontaxis et al. (8) have stud-HSQC (termed PEP-HSQC) was acquired with four FIDs ied experimentally the sensitivity gains obtained when using per t 1 value and 32 scans per FID. Both experiments were such sequences. They mostly confirmed the theoretical figacquired in States-TPPI mode (12). Presaturation of the ures with a set of experimental measurements on a 15 Nwater line was applied during the relaxation delay, with a labeled protein sample; nevertheless they observed deparsaturation power of 31.5 Hz. The PFG-filtered experiment tures from the theoretical values. Our work is based on (termed PFG-HSQC) was acquired with the N and P experiments interleaved, resulting in two FIDs per t 1 value and 64 scans per FID. Field gradients of 1 ms duration were applied,