Recovering lost magnetization: polarization enhancement in biomolecular NMR

Experimental sensitivity remains a major drawback for the application of NMR spectroscopy to fragile and low concentrated biomolecular samples. Here we describe an efficient polarization enhancement mechanism in longitudinal-relaxation enhanced fast-pulsing triple-resonance experiments. By recovering undetectable ¹H polarization originating from longitudinal relaxation during the pulse sequence, the steady-state ¹⁵N polarization becomes enhanced by up to a factor of ~5 with respect to thermal equilibrium yielding significant sensitivity improvements compared to conventional schemes. The benefits of BEST-TROSY experiments at high magnetic field strength are illustrated for various protein applications, but they will be equally useful for other protonated macromolecular systems.     

Determination of the rotational dynamics and pH dependence of the hydrogen exchange rates of the arginine guanidino group using NMR spectroscopy

Summary An improved version of the constant-time HSQC experiment is presented that gives uniform sensitivity over the complete ¹³C bandwidth in ¹³C−¹H correlation experiments without creating artifacts in the methyl and aromatic regions of the spectra. The improvement is achieved by replacing the refocussing ¹³C 180° pulse in the evolution time by a combination of a full-power (22 kHz) hyperbolic secant 180° pulse that inverts and refocusses the entire ¹³C window, immediately followed by a selective 180° pulse on the CO region. Further improvement in signal-to-noise in the aromatic and methyl regions, although less spectacular, is obtained by replacing the other two 180° ¹³C pulses in the INEPT parts of the pulse sequence by full-power hyperbolic secant pulses. Results of simulations and experimental data are presented that demonstrate the excellent performance of the hyperbolic secant pulse for broadband inversion and show that refocussing of transverse magnetization occurs over the same bandwidth, albeit with a ¹³C signal phase that depends quadratically on offset. A further modification, in which one of the selective pulses on the CO region is omitted, is also presented. Implications for other 2D and 3D experiments performed at high fields, where uniform ¹³C inversion and refocussing is desirable, are discussed.     

BEST and SOFAST experiments for resonance assignment of histidine and tyrosine side chains in <Superscript>13</Superscript>C/<Superscript>15</Superscript>N labeled proteins

Aromatic amino-acid side chains are essential components for the structure and function of proteins. We present herein a set of NMR experiments for time-efficient resonance assignment of histidine and tyrosine side chains in uniformly ¹³C/¹⁵N-labeled proteins. The use of band-selective ¹³C pulses allows to deal with linear chains of coupled spins, thus avoiding signal loss that occurs in branched spin systems during coherence transfer. Furthermore, our pulse schemes make use of longitudinal ¹H relaxation enhancement, Ernst-angle excitation, and simultaneous detection of ¹H and ¹³C steady-state polarization to achieve significant signal enhancements.     

Line narrowing in spectra of proteins dissolved in a dilute liquid crystalline phase by band-selective adiabatic decoupling: Application to 1HN-15N residual dipolar coupling measurements

Residual heteronuclear dipolar couplings obtained from partially oriented protein samples can provide unique NMR constraints for protein structure determination. However, partial orientation of protein samples also causes severe ¹ H line broadening resulting from residual ¹ H-¹H dipolar couplings. In this communication we show that band-selective ¹H homonuclear decoupling during data acquisition is an efficient way to suppress residual ¹H-¹H dipolar couplings, resulting in spectra that are still amenable to solution NMR analysis, even with high degrees of alignment. As an example, we present a novel experiment with improved sensitivity for the measurement of one-bond ¹ H^N-¹⁵N residual dipolar couplings in a protein sample dissolved in magnetically aligned liquid crystalline bicelles.     

Comparison of <Superscript>13</Superscript>CH<Subscript>3</Subscript>, <Superscript>13</Superscript>CH<Subscript>2</Subscript>D,and <Superscript>13</Superscript>CHD<Subscript>2</Subscript> methyl labeling strategies in proteins

A comparison of three labeling strategies for studies involving side chain methyl groups in high molecular weight proteins, using ¹³CH₃,¹³CH₂D, and ¹³CHD₂ methyl isotopomers, is presented. For each labeling scheme, ¹H–¹³C pulse sequences that give optimal resolution and sensitivity are identified. Three highly deuterated samples of a 723 residue enzyme, malate synthase G, with ¹³CH₃,¹³CH₂D, and ¹³CHD₂ labeling in Ile δ1 positions, are used to test the pulse sequences experimentally, and a rationalization of each sequence’s performance based on a product operator formalism that focuses on individual transitions is presented. The HMQC pulse sequence has previously been identified as a transverse relaxation optimized experiment for ¹³CH₃-labeled methyl groups attached to macromolecules, and a zero-quantum correlation pulse scheme (¹³CH₃ HZQC) has been developed to further improve resolution in the indirectly detected dimension. We present a modified version of the ¹³CH₃ HZQC sequence that provides improved sensitivity by using the steady-state magnetization of both ¹³C and ¹H spins. The HSQC and HMQC spectra of ¹³CH₂D-labeled methyl groups in malate synthase G are very poorly resolved, but we present a new pulse sequence, ¹³CH₂D TROSY, that exploits cross-correlation effects to record ¹H–¹³C correlation maps with dramatically reduced linewidths in both dimensions. Well-resolved spectra of ¹³CHD₂-labeled methyl groups can be recorded with HSQC or HMQC; a new ¹³CHD₂ HZQC sequence is described that provides improved resolution with no loss in sensitivity in the applications considered here. When spectra recorded on samples prepared with the three isotopomers are compared, it is clear that the ¹³CH₃ labeling strategy is the most beneficial from the perspective of sensitivity (gains ≥2.4 relative to either ¹³CH₂D or ¹³CHD₂ labeling), although excellent resolution can be obtained with any of the isotopomers using the pulse sequences presented here.     

Improved detection of long-range residual dipolar couplings in weakly aligned samples by Lee–Goldburg decoupling of homonuclear dipolar truncation

Homonuclear ¹H residual dipolar couplings (RDCs) truncate the evolution of transverse ¹H magnetization of weakly aligned molecules in high-resolution NMR experiments. This leads to losses in sensitivity or resolution in experiments that require extended ¹H evolution times. Lee–Goldburg decoupling schemes have been shown to remove the effects of homonuclear dipolar couplings, while preserving chemical shift evolution in a number of solid-state NMR applications. Here, it is shown that the Lee-Goldburg sequence can be effectively incorporated into INEPT- or HMQC-type transfer schemes in liquid state weak alignment experiments in order to increase the efficiency of the magnetization transfer. The method is applied to the sensitive detection of ¹H^N–¹³C long-range RDCs in a three-dimensional HCN experiment. As compared to a conventional HCN experiment, an average sensitivity increase by a factor of 2.4 is obtained for a sample of weakly aligned protein G. This makes it possible to detect 170 long-range ¹H^N–¹³C RDCs for distances up to 4.9 Å     

Broadband homonuclear chemical shift correlation at high MAS frequencies: a study of tanh/tan adiabatic RF pulse schemes without $${^{{\bf 1}}\hbox{{\bf H}}}$$ decoupling during mixing

At high magic angle spinning (MAS) frequencies the potential of tanh/tan adiabatic RF pulse schemes for ¹³C chemical shift correlation without ¹H decoupling during mixing has been evaluated. It is shown via numerical simulations that a continuous train of adiabatic ¹³C inversion pulses applied at high RF field strengths leads to efficient broadband heteronuclear decoupling. It is demonstrated that this can be exploited effectively for generating through-bond and through-space, including double-quantum, correlation spectra of biological systems at high magnetic fields and spinning speeds with no ¹H decoupling applied during the mixing period. Experiments carried out on a polycrystalline sample of histidine clearly suggest that an improved signal to noise ratio can be realised by eliminating ¹H decoupling during mixing.     

A spin-state-selective experiment for measuring heteronuclear one-bond and homonuclear two-bond couplings from an HSQC-type spectrum

Recently, a set of selective 1D experiments with spin-state-selective excitation for CH spin systems was introduced by Parella and Belloc (J. Magn. Reson., 148, 78–87 (2001)). We have expanded and generalized this concept further, and demonstrated that a very simple experiment utilizing spin-state-selective filtering can be used for simultaneous measurement of heteronuclear ¹ J _NH (or ¹ J _CH) and geminal ² J _HH couplings from two-dimensional ¹⁵N-¹H (or ¹³C-¹H) correlation spectrum. The experiment has very high sensitivity owing to the preservation of equivalent coherence transfer pathways analogous to the sensitivity and gradient enhanced HSQC experiment. However, overall length of the pulse sequence is 1/(2J) shorter than the gradient selected SE-HSQC experiment. Furthermore, the spin-state-selection can be utilized between NH and NH₂ (or CH and CH₂) moieties by changing the phase of only one pulse. The pulse scheme will be useful for the measurement of scalar and residual dipolar couplings in wide variety of samples, due to its high sensitivity and artifact suppression efficiency. The method is tested on NH₂ and CH₂ moieties in ¹⁵N- and ¹⁵N/¹³C–labeled ubiquitin samples.     

An enhanced sensitivity methyl <Superscript>1</Superscript>H triple-quantum pulse scheme for measuring diffusion constants of macromolecules

We present a pulse scheme that exploits methyl ¹H triple-quantum (TQ) coherences for the measurement of diffusion rates of slowly diffusing molecules in solution. It is based on the well-known stimulated echo experiment, with encoding and decoding of TQ coherences. The size of quantifiable diffusion coefficients is thus lowered by an order of magnitude with respect to single-quantum (SQ) approaches. Notably, the sensitivity of the scheme is high, approximately ¾ that of the corresponding single quantum experiment, neglecting relaxation losses, and on the order of a factor of 4 more sensitive than a previously published sequence for AX₃ spin systems (Zheng et al. in JMR 198:271–274, 2009) for molecules that are only ¹³C labeled at the methyl carbon position. Diffusion coefficients measured from TQ- and SQ-based experiments recorded on a range of protein samples are in excellent agreement. We present an application of this technique to the study of phase-separated proteins where protein concentrations in the condensed phase can exceed 400 mg/mL, diffusion coefficients can be as low as ~10^?9 cm²s^?1 and traditional SQ experiments fail.     

HN(CA)N and HN(COCA)N experiments for assignment of large disordered proteins

Two new 3D HN-based experiments are proposed for backbone assignment of large disordered proteins. The spectra obtained with the new pulse schemes are free of redundant diagonal peaks (H_iN_i–N_i) and provide sequential correlations (H_iN_i–N_i+1 and H_iN_i–N_i?1) not only between adjacent non-proline residues but also between non-proline and proline residues. The experiments have been demonstrated on an intrinsically disordered protein with 306 amino acids including 64 proline residues. Using the two experiments, we obtained nearly complete assignments of backbone amides and proline ¹⁵N spins except for 4 proline and 4 non-proline residues.     

Acceleration of protein backbone NMR assignment by combinatorial labeling: Application to a small molecule binding study

Selective labeling with stable isotopes has long been recognized as a valuable tool in protein NMR to alleviate signal overlap and sensitivity limitations. In this study, combinatorial ¹⁵N‐, ¹³C^α‐, and ¹³C'‐selective labeling has been used during the backbone assignment of human cyclophilin D to explore binding of an inhibitor molecule. Using a cell‐free expression system, a scheme that involves ¹⁵N, 1‐¹³C, 2‐¹³C, fully ¹⁵N/¹³C, and unlabeled amino acids was optimized to gain a maximum of assignment information from three samples. This scheme was combined with time‐shared triple‐resonance NMR experiments, which allows a fast and efficient backbone assignment by giving the unambiguous assignment of unique amino acid pairs in the protein, the identity of ambiguous pairs and information about all 19 non‐proline amino acid types. It is therefore well suited for binding studies where de novo assignments of amide ¹H and ¹⁵N resonances need to be obtained, even in cases where sensitivity is the limiting factor.     

Direct amide <Superscript>15</Superscript>N to <Superscript>13</Superscript>C transfers for solid-state assignment experiments in deuterated proteins

The assignment of protein backbone and side-chain NMR chemical shifts is the first step towards the characterization of protein structure. The recent introduction of proton detection in combination with fast MAS has opened up novel opportunities for assignment experiments. However, typical 3D sequential-assignment experiments using proton detection under fast MAS lead to signal intensities much smaller than the theoretically expected ones due to the low transfer efficiency of some of the steps. Here, we present a selective 3D experiment for deuterated and (amide) proton back-exchanged proteins where polarization is directly transferred from backbone nitrogen to selected backbone or sidechain carbons. The proposed pulse sequence uses only ¹H–¹⁵N cross-polarization (CP) transfers, which are, for deuterated proteins, about 30% more efficient than ¹H–¹³C CP transfers, and employs a dipolar version of the INEPT experiment for N–C transfer. By avoiding H^N–C (H^N stands for amide protons) and C–C CP transfers, we could achieve higher selectivity and increased signal intensities compared to other pulse sequences containing long-range CP transfers. The REDOR transfer is designed with an additional selective π pulse, which enables the selective transfer of the polarization to the desired ¹³C spins.     

Estimates of methyl 13C and 1H CSA values (Δσ) in proteins from cross-correlated spin relaxation

Simple pulse schemes are presented for the measurement of methyl ¹³C and ¹H CSA values from ¹H–¹³C dipole/¹³C CSA and ¹H–¹³C dipole/¹H CSA cross-correlated relaxation. The methodology is applied to protein L and malate synthase G. Average ¹³C CSA values are considerably smaller for Ile than Leu/Val (17 vs 25 ppm) and are in good agreement with previous solid state NMR studies of powders of amino acids and dipeptides and in reasonable agreement with quantum-chemical DFT calculations of methyl carbon CSA values in peptide fragments. Small averaged ¹H CSA values on the order of 1 ppm are measured, consistent with a solid state NMR determination of the methyl group ¹H CSA in dimethylmalonic acid.     

Development of NMR spectroscopic methods for dynamic detection of acetylcholine synthesis by choline acetyltransferase in hippocampal tissue

Choline acetyltransferase (ChAT) is the key enzyme for acetylcholine (ACh) synthesis and constitutes a reliable marker for the integrity of cholinergic neurons. Cortical ChAT activity is decreased in the brain of patients suffering from Alzheimer's and Parkinson's diseases. The standard method used to measure the activity of ChAT enzyme relies on a very sensitive radiometric assay, but can only be performed on post‐mortem tissue samples. Here, we demonstrate the possibility to monitor ACh synthesis in rat brain homogenates in real time using NMR spectroscopy. First, the experimental conditions of the radiometric assay were carefully adjusted to produce maximum ACh levels. This was important for translating the assay to NMR, which has a low intrinsic sensitivity. We then used ¹⁵N‐choline and a pulse sequence designed to filter proton polarization by nitrogen coupling before ¹H‐NMR detection. ACh signal was resolved from choline signal and therefore it was possible to monitor ChAT‐mediated ACh synthesis selectively over time. We propose that the present approach using a labeled precursor to monitor the enzymatic synthesis of ACh in rat brain homogenates through real‐time NMR represents a useful tool to detect neurotransmitter synthesis. This method may be adapted to assess the state of the cholinergic system in the brain in vivo in a non‐invasive manner using NMR spectroscopic techniques.     

TROSY-based HNCO pulse sequences for the measurement of 1HN-15N, 15N-13CO, 1HN-13CO, 13CO-13Cα and 1HN-13Cα dipolar couplings in 15N, 13C, 2H-labeled proteins

HNCO-based 3D pulse schemes are presented for measuring ¹HN-¹⁵N,¹⁵N-¹³CO, ¹HN-¹³CO,¹³CO-¹³C and ¹HN-¹³C dipolar couplings in ¹⁵N,¹³C,²-labeled proteins. The experiments are based on recently developed TROSY methodology for improving spectral resolution and sensitivity. Data sets recorded on a complex of Val, Leu, Ile (1 only) methyl protonated ¹⁵N,¹³C,²H-labeled maltose binding protein and -cyclodextrin as well as ¹⁵N,¹³C,²H-labeled human carbonic anhydrase II demonstrate that precise dipolar couplings can be obtained on proteins in the 30–40 kDa molecular weight range. These couplings will serve as powerful restraints for obtaining global folds of highly deuterated proteins.     

A new class of CEST experiment based on selecting different magnetization components at the start and end of the CEST relaxation element: an application to <Superscript>1</Superscript>H CEST

Chemical exchange saturation transfer (CEST) experiments are becoming increasingly popular for investigating biomolecular exchange dynamics with rates on the order of approximately 50–500 s^?1 and a rich toolkit of different methods has emerged over the past few years. Typically, experiments are based on the evolution of longitudinal magnetization, or in some cases two-spin order, during a fixed CEST relaxation delay, with the same class of magnetization prepared at the start and selected at end of the CEST period. Here we present a pair of TROSY-based pulse schemes for recording amide and methyl ¹H CEST profiles where longitudinal magnetization at the start evolves to produce two-spin order that is then selected at the completion of the CEST element. This selection process subtracts out contributions from ¹H–¹H cross-relaxation on the fly that would otherwise complicate analysis of the data. It also obviates the need to record spin-state selective CEST profiles as an alternative to eliminating NOE effects, leading to significant improvements in sensitivity. The utility of the approach is demonstrated on a sample of a cavity mutant of T4 lysozyme that undergoes chemical exchange between conformations where the cavity is free and occupied.     

Fast hydrogen exchange affects 15N relaxation measurements in intrinsically disordered proteins

Unprotected amide protons can undergo fast hydrogen exchange (HX) with protons from the solvent. Generally, NMR experiments using the out-and-back coherence transfer with amide proton detection are affected by fast HX and result in reduced signal intensity. When one of these experiments, ¹H–¹⁵N HSQC, is used to measure the ¹⁵N transverse relaxation rate (R₂), the measured R₂ rate is convoluted with the HX rate (k_HX) and has higher apparent R₂ values. Since the ¹⁵N R₂ measurement is important for analyzing protein backbone dynamics, the HX effect on the R₂ measurement is investigated and described here by multi-exponential signal decay. We demonstrate these effects by performing ¹⁵N R ₂ ^CPMG experiments on α-synuclein, an intrinsically disordered protein, in which the amide protons are exposed to solvent. We show that the HX effect on R ₂ ^CPMG can be extracted by the derived equation. In conclusion, the HX effect may be pulse sequence specific and results from various sources including the J coupling evolution, the change of steady state water proton magnetization, and the D₂O content in the sample. To avoid the HX effect on the analysis of relaxation data of unprotected amides, it is suggested that NMR experimental conditions insensitive to the HX should be considered or that intrinsic R ₂ ^CPMG values be obtained by methods described herein.     

Off-resonance rotating-frame amide proton spin relaxation experiments measuring microsecond chemical exchange in proteins

NMR spin relaxation in the rotating frame (R_1ρ) is a unique method for atomic-resolution characterization of conformational (chemical) exchange processes occurring on the microsecond time scale. Here, we use amide ¹H off-resonance R_1ρ relaxation experiments to determine exchange parameters for processes that are significantly faster than those that can be probed using ¹⁵N or ¹³C relaxation. The new pulse sequence is validated using the E140Q mutant of the C-terminal domain of calmodulin, which exhibits significant conformational exchange contributions to the transverse relaxation rates. The ¹H off-resonance R_1ρ data sample the entire relaxation dispersion profiles for the large majority of residues in this protein, which exchanges between conformations with a time constant of approximately 20 μs. This is in contrast to the case for ¹⁵N, where additional laboratory-frame relaxation data are required to determine the exchange parameters reliably. Experiments were performed on uniformly ¹⁵N-enriched samples that were either highly enriched in ²H or fully protonated. In the latter case, dipolar cross-relaxation with aliphatic protons were effectively decoupled to first order using a selective inversion pulse. Deuterated and protonated samples gave the same results, within experimental errors. The use of deuterated samples increases the sensitivity towards exchange contributions to the ¹H transverse relaxation rates, since dipolar relaxation is greatly reduced. The exchange correlation times determined from the present ¹H off-resonance R_1ρ experiments are in excellent agreement with those determined previously using a combination of ¹⁵N laboratory-frame and off-resonance R_1ρ relaxation data, with average values of and 21 ± 3 μs, respectively.     

HMQC and HSQC experiments with water flip-back optimized for large proteins

An HMQC experiment is proposed, dubbed FHMQC, where water flip-back is achieved by a single water-selective pulse preceding the basic HMQC pulse sequence. The scheme is demonstrated with a ¹⁵N, ¹H-HMQC spectrum of uniformly ¹⁵N/²H-labelled S. aureus DNA gyrase B with a molecular weight of 45 kDa for the unlabelled protein. The sensitivity of the experiment is improved compared to that of an FHSQC spectrum. It is further shown that the original FHSQC experiment can be shortened by the use of bipolar gradients. Relaxation times of different ¹⁵N magnetizations and coherences were measured. The new FHMQC scheme is implemented in 3D NOESY-¹⁵N-HMQC and 3D¹⁵ N-HMQC-NOESY-¹⁵N-HMQC pulse sequences which are demonstrated with a 24 kDa fragment of uniformly ¹⁵N/¹³C/²H-labelled S. aureus DNA gyrase B.     

High-level 2H/13C/15N labeling of proteins for NMR studies

Summary The protein human carbonic anhydrase II (HCA II) has been isotopically labeled with ²H, ¹³C and ¹⁵N for high-resolution NMR assignment studies and pulse sequence development. To increase the sensitivity of several key ¹H/¹³C/¹⁵N triple-resonance correlation experiments, ²H has been incorporated into HCA II in order to decrease the rates of ¹³C and ¹H_N T₂ relaxation. NMR quantities of protein with essentially complete aliphatic ²H incorporation have been obtained by growth of E. coli in defined media containing D₂O, [1,2-¹³C₂, 99%] sodium acetate, and [¹⁵N, 99%] ammonium chloride. Complete aliphatic deuterium enrichment is optimal for ¹³C and ¹⁵N backbone NMR assignment studies, since the ¹³C and ¹H_N T₂ relaxation times and, therefore, sensitivity are maximized. In addition, complete aliphatic deuteration increases both resolution and sensitivity by eliminating the differential ²H isotopic shift observed for partially deuterated CH_nD_m moieties.