- Email: [email protected]
NMR supersequences with real-time homonuclear broadband decoupling: Sequential acquisition of protein and small molecule spectra in a single experiment
Accepted Manuscript NMR supersequences with real-time homonuclear broadband decoupling: sequential acquisition of protein and small molecule spectra in a single experiment Veera Mohana Rao Kakita, Kavitha Rachineni, Mandar Bopardikar, Ramakrishna V. Hosur PII: DOI: Reference:
S1090-7807(18)30288-X https://doi.org/10.1016/j.jmr.2018.10.013 YJMRE 6389
To appear in:
Journal of Magnetic Resonance
Received Date: Revised Date: Accepted Date:
2 August 2018 8 October 2018 21 October 2018
Please cite this article as: V. Mohana Rao Kakita, K. Rachineni, M. Bopardikar, R.V. Hosur, NMR supersequences with real-time homonuclear broadband decoupling: sequential acquisition of protein and small molecule spectra in a single experiment, Journal of Magnetic Resonance (2018), doi: https://doi.org/10.1016/j.jmr.2018.10.013
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
NMR supersequences with real-time homonuclear broadband decoupling: sequential acquisition of protein and small molecule spectra in a single experiment a,
a,
b
Veera Mohana Rao Kakita, † Kavitha Rachineni, † Mandar Bopardikar, and Ramakrishna V. Hosur
a
b c
a, c
*
UM-DAE Centre for Excellence in Basic Sciences, University of Mumbai, Kalina Campus, Santacruz, Mumbai 400 098, India E-mail: [email protected] Department of Chemical Sciences, Tata Institute of Fundamental Research (TIFR), 1-Homi Bhabha Road, Colaba, Mumbai 400 005, India Department of Biosciences and Bioengineering Indian Institute of Technology Bombay Powai, Mumbai – 400076
† These authors contributed equally.
Abstract NOAH (NMR by Ordered Acquisition using 1H-detection) type of pure shift NMR pulse scheme has been designed for the efficient utilization of magnetization that presents in a spinsystem under consideration. The proposed strategy, PROSMASH-HSQC2 (PROtein-HSQC and SMAll molecule-HSQC Signals with Homodecoupling) uses the real-time BIRD pure shift NMR strategy and two HSQC spectra (13C-HSQC for small molecules and 15N-HSQC for 15N-isotopic labelled proteins) can be recorded in a single NMR experiment. Thus, this method permits precise determination of drugprotein interactions at atomic levels by monitoring the chemical shift perturbations, and will have potential applications in drug discovery programs.
Keywords HSQC; NOAH pulse scheme; pure shift NMR; real-time BIRD homodecoupling; drug-protein interactions;
1
1. Introduction The very recent advent of supersequences in NMR spectroscopy, NOAH (NMR by Ordered Acquisition using 1H-detection) methods [1, 2] has enabled significant advance in high throughput structural studies of small molecules. A single NOAH experiment is adequate to obtain the entire NMR spectral information required for solving the structure of small organic molecules. For example, in a recent NOAH experiment, a set of spectra (15N-HMQC, 13C-HMBC, 13C-HSQC, COSY, and NOESY) were acquired in a sequential manner. In fact, ~ 250 different combinations of such NOAH experiments have been proposed. [1] These experiments use an ingenious strategy, which depends on efficient utilization of magnetization at different parts of the NMR pulse sequences and all the acquisition modules utilize the same recycle delay. To begin with, the entire magnetization is along the z-axis, then at the end of the first 15N-HMQC acquisition only 0.37% of 1H magnetization is used and subsequently ~1.1% of 1H magnetization (1H attached to 13C) is used for the 13C-HSQC module of NOAH experiment. The remaining ~ 98% of 1H magnetization is used for recording both COSY and NOESY. Herein, the acquisition of COSY and NOESY spectra in NOAH supersequences is performed in accordance with the COCONOSY pulse scheme. [3] Indeed, these supersequences have superior performance over methods which have been developed earlier for similar purposes. For example, multiple receiver experiments require special hardware; [4-9] thus, they are not very common in most of the laboratories. The time-shared (TS) experiments also result in multiple spectra, [10-12] but they are different from the NOAH schemes. In order to obtain the individual spectra from TS experiments, different data sets need to be acquired in in-phase and anti-phase fashion. In fact, sequential acquisition based supersequences have earlier also been occasionally explored for solidstate applications [13] and proteins. [14] Here, we have used the idea of supersequences for acquiring two HSQC data sets in a sequential manner, for a protein-small molecule mixture. A 15N-HSQC for the 15N labelled protein and a 13C-HSQC for the small molecules at natural abundance are recorded in the same pulse sequence. The pulse sequence incorporates BIRD [15-19] homodecoupling to enhance resolution in the spectra and this enables study of interactions between proteins and a mixture of small molecules. We have named this experiment as, PROSMASH-HSQC2 (PROtein-HSQC and SMAll molecule-HSQC Signals with Homodecoupling). The efficiency of the presently proposed PROSMASHHSQC2 experiment has been demonstrated on a protein-small molecular spin-system.
2. Results and Discussion 2.1. PROSMASH-HSQC2 pulse sequence Figure 1 represents the PROSMASH-HSQC2 pulse scheme developed in the present study. Initial part of this pulse sequence uses the 15N-HSQC pulse module and the last part is the pure shift 13 C-BIRD-HSQC. The first 15N-HSQC utilizes the entire amide proton magnetization, since the protein is uniformly 15N labelled. The subsequent pulse module (13C-HSQC) does not require any amide magnetization; hence, it is a perfectly appropriate strategy to acquire both the spectra in a single experiment. The second block utilizes real-time homonuclear broadband BIRD decoupling (13C-BIRDHSQC) and that results in highly resolved chemical shift information for the small molecules.
2
Figure 1: Schematic representation of PROSMASH-HSQC2 pulse sequence developed in the present study. Initial part of this pulse scheme utilizes the 15N-HSQC (for uniformly 15N labelled proteins) and the latter part uses 13C-HSQC with real-time homonuclear broadband BIRD decoupling blocks. The narrow and wide rectangles, respectively, represent the 90o and 180o hard pulses. The hard 180o pulses on 13C nuclei may be replaced by adiabatic pulses, if the sample demands large excitation bandwidths. For the INEPT, 13
for 15N-HSQC) and
(
C-HSQC), and for the real-time homonuclear broadband BIRD decoupling,
(
for evolution
periods were used. Acquisition parameters: spectral widths, 9600 Hz (1H, direct dimension)/ 8000 Hz (13C, indirect dimension); number of points, 3072 (1H, direct dimension)/ 256 (13C, indirect dimension); acquisition times, 0.159s (1H, indirect dimension)/ 0.032s (13C, indirect dimension); ; phase cycling,
,
,
,
,
, , , ; gradients G1=80%, G2=8.1%, G3=20.1% (in 1ms duration), G4= 5% (in 0.3ms duration and at a maximum of 53.5 G/cm); relaxation delay, d1=1 s; number of scans, 8; number of BIRD decoupling interruptions is equal to ‘n’ and during 1H acquisition, garp heteronuclear broadband decoupling was applied. Quadrature detection has been achieved by echo-anti echo manner. All the data sets were acquired on a 800 MHz Bruker NMR spectrometer equipped with a room temperature probe. Total experimental time required for the acquisition of each PROSMASH-HSQC2 experiment was ~54 min. For data separation, Bruker inbuilt AU macro, ‘split’ was used and then 15N axis of 15N-HSQC was recalibrated with the ‘fixF1n’ AU program that has been downloaded from the Bruker User library. 3
2.2. PROSMASH-HSQC2 for monitoring the interactions of α-synuclein (α-syn) and a mixture of monosaccharides In order to demonstrate the efficacy of the presently proposed PROSMASH-HSQC2 method, α-syn and a mixture of monosaccharides (glucose, mannose, and xylose) have been used. In aqueous solvent medium, these three monosaccharides convert into a mixture of total six anomer constituents (α and β, Fig.2a). α-syn is a 140 amino acid residue length neuronal protein, which is highly expressed in the brain. This native protein is intrinsically disordered. It has been identified as a major constituent of neuronal inclusion/ lewy bodies in aggregated form of insoluble fibril.[20] Its process of aggregation into fibrilar form is complex, and depends on many cellular conditions, viz., cellular composition,[21] pH,[22] temperature[23] etc. As is well known, sugars act as stabilizing agents, which act by changing the solvent surface tension around the proteins.[24] In general, cellular composition may have several sugars with different conformational and configurational isomers, but in the present studies, we have arbitrarily chosen the said monosaccharides to monitor their interactions with α-syn. Initially, two independent PROSMASH-HSQC2 (while switching off the BIRD homodecoupling, n=1) spectra have been recorded for the individual samples of α-syn (only 15N labelled, 300 M, at pH 7.4) and the monosaccharide mixture (10 mg of each, at pH 7.4, in 0.5 ml of 1 (D2O):9 (H2O)), wherein each anomer concentration is equal to 55 mM (i.e., per residue of α-syn ~ 390 M), which is much less when compared with the studies that were earlier carried out to mimic the crowding environment (α-syn in 2M glucose).[25] They have resulted in 15N-HSQC (for α-syn, Fig.2b, blue colour contours) and 13C-HSQC (for monosaccharides, Fig. 2c) spectra. From Fig. 2b, it is clear from peak count that several peaks of α-syn are missing in the resultant 15N-HSQC spectrum (blue contours); this may be a consequence of exchange processes occurring in the protein. In the 13CHSQC (Fig. 2c) chemical shift resolution for most of the peaks is found to be poor, and this is because all the constituents have similar molecular structures. Then, for the same pure monosaccharide mixture, PROSMASH-HSQC2 experiments have been repeated with n=6, which has resulted in a BIRD homodecoupled 13C-HSQC spectrum and that has allowed us to identify all the chemical sites belonging to all the constituents (Fig. 2d). In the first sample, i.e., α-syn, 13C-HSQC part of the PROSMASH-HSQC2 results in very weak peaks (obscure in the noise level) at the natural abundance of 13C spins. On the other hand, for the second sample, i.e., 15N-HSQC part of the PROSMAS H-HSQC2 does not show any signals, as there are no amide resonances in the considered monosaccharides. Further, 1D-traces obtained from the 1H-1H coupled (Fig. 2f) and BIRD-homodecoupled 13C-HSQC (Fig. 2g) part of the PROSMASH-HSQC2 are also compared. This reveals that there is a significant enhancement in the chemical shift resolution, and signal sensitivity improvements are also evident. Clearly, this kind of homodecoupled spectra provide unambiguous measurement of chemical shift perturbations as demonstrated earlier by Zangger and co-workers.[26] Subsequently, another sample has been prepared by just adding the said monosaccharides (10 mg of each) to α-syn protein (300 M, at pH 7.4) and PROSMASH-HSQC2 experiment has been repeated in the BIRD homodecoupling mode. This time acquisition of only one PROSMASH-HSQC2 experiment is adequate to obtain both the 15N-HSQC (first part of the pulse scheme, for α-syn, Fig.2b, red colour contours), and 13C-BIRD-HSQC (second part of the pulse sequence, for the monosaccharide mixture, Fig.2e). The resultant 15N-HSQC spectrum (Fig.2b, red colour contours) of α-syn has shown many additional peaks when compared with the previously recorded 15N-HSQC for 4
the pure α-syn sample (Fig. 2b, blue colour contours). This improvement in the 15N-HSQC spectrum may be attributed to stabilization of conformation of α-syn in the presence of monosaccharides. At the same time, very clear information of chemical sites has been monitored from the 13C-BIRD-HSQC (Fig. 2e and Fig. 2h) of monosaccharides and this spectrum is exactly similar to the 13C-BIRD-HSQC that has earlier been recorded for the pure monosaccharide mixture (Fig. 2d and Fig. 2g). This could be due to high concentrations of sugars relative to the protein concentration and possibly also weak interactions between protein and small molecules; hence, the chemical shift resolution is still sustained for all the small molecules in the presence of α-syn protein. Here, it is worthwhile to discuss the sensitivity of 13C-HSQC resonances of α-syn (20 times scaled up in Fig. 2i) relative to that of the monosaccharide peaks (Fig. 2e). As a proof of principle, PROSMASH-HSQC2 experiments have been demonstrated at rather high concentrations of small molecules. Nevertheless, even at low concentrations of small molecules, this method would work reasonably well, as the 13C-HSQC of PROSMASH-HSQC2 exploits the benefits of relaxation differences between proteins and small molecules. In some cases, there is a possibility of overlap between protein (at the level of detectable signals) and small molecule chemical shifts in the 13C-HSQC of PROSMASH-HSQC2, but these could be unravelled by suitable controls with 13C-BIRD-HSQC in the PROSMASH-HSQC2 manner for pure small molecules (reference spectrum), which allows identification of small molecule chemical shifts, even when they are overlapped with protein signals. But, it must be realized that the present method does not work well with doubly labelled (13C and 15 N) proteins in the presence of unlabelled small molecules, as the 13C-HSQC of PROSMASH-HSQC2 shows both the protein and small molecule signals, and the BIRD decoupling does not work for the 13 C labelled proteins. Chemical shift perturbation values for both the protein (15N-HSQC) and monosaccharide mixture (13C-HSQC) have been measured,[27] while taking the chemical shift differences between pure compounds and mixtures. Here the CSP’s [
] for the observed peaks of α-syn
protein are found to be negligible (very small, Fig. 2j) and similarly most of the spins of monosaccharide mixture [
] have also not shown any considerable perturbations
(Fig. 2k). However, the peak intensities of several peaks in the protein spectra are significantly affected. Most of these peaks belong to Threonines or their neighbours, which suggests a possible interaction between OH groups of small molecules with the OH groups of the threonine residues of α-syn protein.
5
Figure 2: Molecular structures of various monosaccharides used in the present study (a). Superimposition of 15N-HSQC spectra recorded for α-syn protein in the PROSMASH-HSQC2 NMR pulse scheme (b). Here, blue and red colour contours represent the 15N-HSQC spectra obtained in absence and in presence of the monosaccharide mixture, respectively. The PROSMASH-HSQC2 resultant 13CHSQC spectra of monosaccharide mixture at different conditions are shown in (c-e). Further, the corresponding 1D-internal traces have also been depicted in (f-h). The residual 13C-BIRD-HSQC signals of protein at natural abundant level are shown (i), while raising the vertical scale of this spectrum by 20 folds with respect to the monosaccharide peaks (e). The calculated chemical shift perturbations of α-syn protein (from 15N-HSQC, as the residue number of α-syn) and monosaccharide mixture (from 13 C-BIRD-HSQC, as the spin number of monosaccharides; M: mannose, G: glucose, X: xylose) are depicted in (j) and (k), respectively.
The same α-syn and monosaccharide mixture sample has been incubated initially at room temperature for a week and then at 4oC for 23 days. Subsequently, PROSMASH-HSQC2 experiment has been repeated under the identical experimental conditions (which we have initially used earlier for the same sample). The resultant 15N-HSQC (Fig. 3a) and 13C-BIRD-HSQC (Fig. 3b) spectra of α-syn, and monosaccharide mixture, respectively, have been used for calculating the CSP’s. From the Fig. 3c, between immediately prepared and 30 days incubated samples, a nice correlation (~ 0.9) in the CSP’s is noticed with a slope of ~1.0 (Fig. 3c). Similarly, the measured CSP’s for most of the spins of monosaccharide mixture are also found to be very small (Fig. 3d), as in the immediately prepared sample. Indeed, in order to obtain the very important backbone chemical shift assignments of α-syn like intrinsically disordered proteins (IDP’s), to avoid the conformational flexibility (i.e., missing peaks), it would be a nice idea to consider the monosaccharides as stabilizing agents (as osmolytes with minimal interactions) and such samples can be used for acquiring the HNN (one experiment is adequate to obtain the backbone chemical shift assignments of IDP’s) [28] type of experiments. 6
Figure 3: The PROSMASH-HSQC2 resultant 15N-HSQC and 13C-BIRD-HSQC spectra of α-syn and the monosaccharide mixture (after 30 days incubation) are shown in (a) and (b), respectively. The calculated CSP’s of α-syn (from 15N-HSQC) protein at this condition have shown nice correlation with the CSP’s measured for the immediately prepared sample (c). Finally, 13C-BIRD-HSQC derived CSP’s of monosaccharide mixture are shown in (d).
From these findings, it is clear that the monosaccharides have a significant influence on the behaviour of α-syn in solution and this could be via change in viscosity of solution or/and interactions between OH groups belonging to threonine residues of α-syn and OH groups of monosaccharide moieties. As demonstrated in the present example, the PROSMASH-HSQC2 method is very useful for precise determination of CSP’s for both proteins and small molecules. As BIRD homodecoupling is an inbuilt part of the small molecule 13C-HSQC, chemical shift details can be obtained with improved sensitivities and high accuracies.
3. Conclusions The present work demonstrates the development and applications of a new NOAH experiment, namely, homodecoupled PROSMASH-HSQC2 experiment. The initial part of this pulse scheme acquires 15N-HSQC (for 15N labelled proteins) and the second part records 13C-HSQC (at natural abundance for small molecules) in the real-time BIRD homonuclear broadband decoupling fashion. An interesting example of protein and small molecular spin-system, i.e., α-syn and a mixture of monosaccharide mixture (glucose, mannose, and xylose) have been considered to demonstrate the efficacy of the PROSMASH-HSQC2 experiment. This method helps in the precise determination of chemical shift perturbations of both the small molecules and the proteins, from a single NMR experiment. Therefore, the present method saves the spectrometer time by two fold. On demand, these experimental times can be further reduced by invoking the non-uniform sampling schemes [29-31]. The utility of the presently proposed method lies in unambiguous determination of proteinsmall molecule (mixture) interactions and that will have an immense impact on drug discovery programs. Further, in some cases, simultaneous acquisition of both protein and small molecule signals may be essential; in such situations, the proposed method would be greatly helpful. Like in other BIRD homodecoupled pure shift NMR methods, the present PROSMASH-HSQC2 also does not work well for 13C isotopically labelled spin-systems and cannot decouple the diastereomeric protons.
7
4. Acknowledgements The authors thank National NMR Facility, TIFR, Mumbai. VMRK (PDF/2016/000365), and KR (PDF/2016/002063) are grateful to DST-SERB (Department of Science and Technology-Science and Engineering Research Board), Government of India, New Delhi, for the award of National Postdoctoral Fellowships. RVH is a DST sponsored JC Bose National Fellow. 5. References [1]. Ē. Kupče, and T. D. W. Claridge, NOAH: NMR Supersequences for Small Molecule Analysis and
Structure Elucidation, Angew. Chem. Int. Ed. Engl. 56 (2017) 11779–11783. [2]. Ē. Kupče, and T. D. W. Claridge, Molecular structure from a single NMR supersequence, Chem. Commun. 54 (2018) 7139–7142. [3]. C. Haasnoot, F. J. M. van de Ven, and C. W. Hilbers, COCONOSY. Combination of 2D correlated and 2D nuclear overhauser enhancement spectroscopy in a single experiment, J. Magn. Reson. 56 (1984) 343–349. [4]. Ē. Kupče, R. Freeman, and B. K. John, J. Am. Chem. Soc. 128 (2006) 9606–9607. [5]. Ē. Kupče, in Modern NMR Methodology, Vol. 335 (Eds.: H. Heise, S. Matthews), Springer, Berlin, 2013, pp 71-96. [6]. H. Kovacs, and Ē. Kupče, Magn. Reson. Chem. 54 (2016) 544–560. [7]. Ē. Kupče, and R. Freeman, J. Am. Chem. Soc. 130 (2008) 10788–10792. [8]. K. J. Donovan, Ē. Kupče, and L. Frydman, Multiple parallel 2D NMR acquisitions in a single scan, Angew. Chem. Int. Ed. 52 (2013) 4152–4155. [9]. J. G. Reddy, and R. V. Hosur, Parallel acquisition of 3D-HA(CA)NH and 3D-HACACO spectra, J. Biomol. NMR. 56 (2013) 77–84. 13 15 [10]. B. T. Farmer II, Simultaneous [ C, N]-HMQC, a pseudo-triple-resonance experiment, J. Magn. Reson. 93 (1991) 635–641. [11]. L. E. Kay, M. Wittekind, M. A. McCoy, M. S. Friedrichs, and L. Mueller, 4D NMR triple-resonance experiments for assignment of protein backbone nuclei using shared constant-time evolution periods, J. Magn. Reson. 98 (1992) 443–450. [12]. P. Nolis, M. Pérez-Trujillo, and T. Parella, Multiple FID Acquisition of Complementary HMBC Data, Angew. Chem. Int. Ed. 46 (2007) 7495–7497. [13]. T. Gopinath, and G. Veglia, Multiple acquisition of magic angle spinning solid-state NMR experiments using one receiver: Application to microcrystalline and membrane protein preparations, J. Magn. Reson. 253 (2015) 143–153. [14]. C. Wiedemann, P. Bellstedt, A. Kirschstein, S. Häfner, C. Herbst, M. Görlach, and R. Ramachandran, Sequential protein NMR assignments in the liquid state via sequential data acquisition, J. Magn. Reson. 239 (2014) 23–28. [15]. J. R. Garbow, D. P. Weitekamp, and A. Pines, Bilinear rotation decoupling of homonuclear scalar interactions, Chem. Phys. Lett. 93 (1982) 504–509. [16]. A. Lupulescu, G.L. Olsen, and L. Frydman, Toward single-shot pure-shift solution 1H NMR by trains of BIRD-based homonuclear decoupling, J. Magn. Reson. 218 (2012) 141–146. [17]. J. A. Aguilar, M. Nilsson, and G. A. Morris, Simple Proton Spectra from Complex Spin Systems: Pure Shift NMR Spectroscopy Using BIRD, Angew. Chem. Int. Ed. 50 (2011) 9716–9717. [18]. L. Paudel, R. W. Adams, P. Király, J. A. Aguilar, M. Foroozandeh, M. J. Cliff, M. Nilsson, P. Sándor, J. P. Waltho, and G. A. Morris, Simultaneously Enhancing Spectral Resolution and Sensitivity in Heteronuclear Correlation NMR Spectroscopy, Angew. Chem. Int. Ed. 52 (2013) 11616–11619. [19]. K. J. Donovan, and L. Frydman. HyperBIRD: a sensitivity-enhanced approach to collecting homonuclear-decoupled proton NMR spectra, Angew. Chem. Int. Ed. 54 (2015) 594–598.
8
[20]. M. G. Spillantini, R. A. Crowther, R. Jakes, M. Hasegawa, and M. Goedert, α-Synuclein in filamentous
inclusions of Lewy bodies from Parkinson’s disease and dementia with Lewy bodies, Proc. Natl. Acad. Sci. 95 (1998) 6469–6473. [21]. E. Ihse, H. Yamakado, X. M. van Wijk, R. Lawrence, J. D. Esko, and E. Masliah, Cellular internalization of alpha-synuclein aggregates by cell surface heparan sulfate depends on aggregate conformation and cell type, Sci. Rep. 7 (2017) 9008. [22]. A. K. Buell, C. Galvagnion, R. Gaspar, E. Sparr, M. Vendruscolo, T. P. J. Knowles, S. Linse, and C. M. Dobson, Solution conditions determine the relative importance of nucleation and growth processes in α-synuclein aggregation, Proc. Natl. Acad. Sci. 111 (2014) 7671–7676. [23]. W. Ariesandi, C. Chang, T. Chen, and Y. Chen , Temperature-Dependent Structural Changes of Parkinson's Alpha-Synuclein Reveal the Role of Pre-Existing Oligomers in Alpha-Synuclein Fibrillization, PLOS ONE. 8 (2013) e53487. [24]. T. Arakawa, and S. N. Timasheff, Stabilization of protein structure by sugars, Biochemistry, 25 (1982) 6536–6544. [25]. K-P. Wu, and J. Baum, Backbone Assignment and Dynamics of Human Α-Synuclein in Viscous 2 M Glucose Solution, Biomol. NMR. Assign. 5 (2011) 43–46. [26]. N. H. Meyer, and K. Zangger, Enhancing the resolution of multi-dimensional heteronuclear NMR spectra of intrinsically disordered proteins by homonuclear broadband decoupling, Chem. Commun. 50 (2014) 1488–1490. [27]. M. P. Williamson, Using chemical shift perturbation to characterise ligand binding, Prog. Nucl. Magne. Reson. Spec. 73 (2013) 1–16. [28]. V. M. R. Kakita, M. Bopardikar, V. K. Shukla, K. Rachineni, P. Ranjan, J. S. Singh, and R. V. Hosur, An efficient combination of BEST and NUS methods in multidimensional NMR spectroscopy for high throughput analysis of proteins, RSC Adv. 8 (2018) 17616–17621. [29]. A. Hansen, and R. Bruschweiler, Absolute Minimal Sampling in High-Dimensional NMR Spectroscopy, Angew. Chem. Int. Ed. 55 (2016) 14169–14172. [30]. M. Mobli, and J. C. Hoch, Nonuniform sampling and non-Fourier signal processing methods in multidimensional NMR, Prog. Nucl. Magn. Reson. Spec. 83 (2014) 21–41. [31]. Q. Wu, B. E. Coggins, and P. Zhou, Unbiased measurements of reconstruction fidelity of sparsely sampled magnetic resonance spectra, Nature Commun. 7 (2016) 12281.
9
Highlights
NOAH type of pure shift NMR pulse scheme, PROSMASH-HSQC2 has been designed 15 N-HSQC and 13C-BIRD-HSQC spectra can be recorded sequentially Permits precise determination of drug-protein interactions through CSP’s
10
S1090-7807(18)30288-X https://doi.org/10.1016/j.jmr.2018.10.013 YJMRE 6389
To appear in:
Journal of Magnetic Resonance
Received Date: Revised Date: Accepted Date:
2 August 2018 8 October 2018 21 October 2018
Please cite this article as: V. Mohana Rao Kakita, K. Rachineni, M. Bopardikar, R.V. Hosur, NMR supersequences with real-time homonuclear broadband decoupling: sequential acquisition of protein and small molecule spectra in a single experiment, Journal of Magnetic Resonance (2018), doi: https://doi.org/10.1016/j.jmr.2018.10.013
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
NMR supersequences with real-time homonuclear broadband decoupling: sequential acquisition of protein and small molecule spectra in a single experiment a,
a,
b
Veera Mohana Rao Kakita, † Kavitha Rachineni, † Mandar Bopardikar, and Ramakrishna V. Hosur
a
b c
a, c
*
UM-DAE Centre for Excellence in Basic Sciences, University of Mumbai, Kalina Campus, Santacruz, Mumbai 400 098, India E-mail: [email protected] Department of Chemical Sciences, Tata Institute of Fundamental Research (TIFR), 1-Homi Bhabha Road, Colaba, Mumbai 400 005, India Department of Biosciences and Bioengineering Indian Institute of Technology Bombay Powai, Mumbai – 400076
† These authors contributed equally.
Abstract NOAH (NMR by Ordered Acquisition using 1H-detection) type of pure shift NMR pulse scheme has been designed for the efficient utilization of magnetization that presents in a spinsystem under consideration. The proposed strategy, PROSMASH-HSQC2 (PROtein-HSQC and SMAll molecule-HSQC Signals with Homodecoupling) uses the real-time BIRD pure shift NMR strategy and two HSQC spectra (13C-HSQC for small molecules and 15N-HSQC for 15N-isotopic labelled proteins) can be recorded in a single NMR experiment. Thus, this method permits precise determination of drugprotein interactions at atomic levels by monitoring the chemical shift perturbations, and will have potential applications in drug discovery programs.
Keywords HSQC; NOAH pulse scheme; pure shift NMR; real-time BIRD homodecoupling; drug-protein interactions;
1
1. Introduction The very recent advent of supersequences in NMR spectroscopy, NOAH (NMR by Ordered Acquisition using 1H-detection) methods [1, 2] has enabled significant advance in high throughput structural studies of small molecules. A single NOAH experiment is adequate to obtain the entire NMR spectral information required for solving the structure of small organic molecules. For example, in a recent NOAH experiment, a set of spectra (15N-HMQC, 13C-HMBC, 13C-HSQC, COSY, and NOESY) were acquired in a sequential manner. In fact, ~ 250 different combinations of such NOAH experiments have been proposed. [1] These experiments use an ingenious strategy, which depends on efficient utilization of magnetization at different parts of the NMR pulse sequences and all the acquisition modules utilize the same recycle delay. To begin with, the entire magnetization is along the z-axis, then at the end of the first 15N-HMQC acquisition only 0.37% of 1H magnetization is used and subsequently ~1.1% of 1H magnetization (1H attached to 13C) is used for the 13C-HSQC module of NOAH experiment. The remaining ~ 98% of 1H magnetization is used for recording both COSY and NOESY. Herein, the acquisition of COSY and NOESY spectra in NOAH supersequences is performed in accordance with the COCONOSY pulse scheme. [3] Indeed, these supersequences have superior performance over methods which have been developed earlier for similar purposes. For example, multiple receiver experiments require special hardware; [4-9] thus, they are not very common in most of the laboratories. The time-shared (TS) experiments also result in multiple spectra, [10-12] but they are different from the NOAH schemes. In order to obtain the individual spectra from TS experiments, different data sets need to be acquired in in-phase and anti-phase fashion. In fact, sequential acquisition based supersequences have earlier also been occasionally explored for solidstate applications [13] and proteins. [14] Here, we have used the idea of supersequences for acquiring two HSQC data sets in a sequential manner, for a protein-small molecule mixture. A 15N-HSQC for the 15N labelled protein and a 13C-HSQC for the small molecules at natural abundance are recorded in the same pulse sequence. The pulse sequence incorporates BIRD [15-19] homodecoupling to enhance resolution in the spectra and this enables study of interactions between proteins and a mixture of small molecules. We have named this experiment as, PROSMASH-HSQC2 (PROtein-HSQC and SMAll molecule-HSQC Signals with Homodecoupling). The efficiency of the presently proposed PROSMASHHSQC2 experiment has been demonstrated on a protein-small molecular spin-system.
2. Results and Discussion 2.1. PROSMASH-HSQC2 pulse sequence Figure 1 represents the PROSMASH-HSQC2 pulse scheme developed in the present study. Initial part of this pulse sequence uses the 15N-HSQC pulse module and the last part is the pure shift 13 C-BIRD-HSQC. The first 15N-HSQC utilizes the entire amide proton magnetization, since the protein is uniformly 15N labelled. The subsequent pulse module (13C-HSQC) does not require any amide magnetization; hence, it is a perfectly appropriate strategy to acquire both the spectra in a single experiment. The second block utilizes real-time homonuclear broadband BIRD decoupling (13C-BIRDHSQC) and that results in highly resolved chemical shift information for the small molecules.
2
Figure 1: Schematic representation of PROSMASH-HSQC2 pulse sequence developed in the present study. Initial part of this pulse scheme utilizes the 15N-HSQC (for uniformly 15N labelled proteins) and the latter part uses 13C-HSQC with real-time homonuclear broadband BIRD decoupling blocks. The narrow and wide rectangles, respectively, represent the 90o and 180o hard pulses. The hard 180o pulses on 13C nuclei may be replaced by adiabatic pulses, if the sample demands large excitation bandwidths. For the INEPT, 13
for 15N-HSQC) and
(
C-HSQC), and for the real-time homonuclear broadband BIRD decoupling,
(
for evolution
periods were used. Acquisition parameters: spectral widths, 9600 Hz (1H, direct dimension)/ 8000 Hz (13C, indirect dimension); number of points, 3072 (1H, direct dimension)/ 256 (13C, indirect dimension); acquisition times, 0.159s (1H, indirect dimension)/ 0.032s (13C, indirect dimension); ; phase cycling,
,
,
,
,
, , , ; gradients G1=80%, G2=8.1%, G3=20.1% (in 1ms duration), G4= 5% (in 0.3ms duration and at a maximum of 53.5 G/cm); relaxation delay, d1=1 s; number of scans, 8; number of BIRD decoupling interruptions is equal to ‘n’ and during 1H acquisition, garp heteronuclear broadband decoupling was applied. Quadrature detection has been achieved by echo-anti echo manner. All the data sets were acquired on a 800 MHz Bruker NMR spectrometer equipped with a room temperature probe. Total experimental time required for the acquisition of each PROSMASH-HSQC2 experiment was ~54 min. For data separation, Bruker inbuilt AU macro, ‘split’ was used and then 15N axis of 15N-HSQC was recalibrated with the ‘fixF1n’ AU program that has been downloaded from the Bruker User library. 3
2.2. PROSMASH-HSQC2 for monitoring the interactions of α-synuclein (α-syn) and a mixture of monosaccharides In order to demonstrate the efficacy of the presently proposed PROSMASH-HSQC2 method, α-syn and a mixture of monosaccharides (glucose, mannose, and xylose) have been used. In aqueous solvent medium, these three monosaccharides convert into a mixture of total six anomer constituents (α and β, Fig.2a). α-syn is a 140 amino acid residue length neuronal protein, which is highly expressed in the brain. This native protein is intrinsically disordered. It has been identified as a major constituent of neuronal inclusion/ lewy bodies in aggregated form of insoluble fibril.[20] Its process of aggregation into fibrilar form is complex, and depends on many cellular conditions, viz., cellular composition,[21] pH,[22] temperature[23] etc. As is well known, sugars act as stabilizing agents, which act by changing the solvent surface tension around the proteins.[24] In general, cellular composition may have several sugars with different conformational and configurational isomers, but in the present studies, we have arbitrarily chosen the said monosaccharides to monitor their interactions with α-syn. Initially, two independent PROSMASH-HSQC2 (while switching off the BIRD homodecoupling, n=1) spectra have been recorded for the individual samples of α-syn (only 15N labelled, 300 M, at pH 7.4) and the monosaccharide mixture (10 mg of each, at pH 7.4, in 0.5 ml of 1 (D2O):9 (H2O)), wherein each anomer concentration is equal to 55 mM (i.e., per residue of α-syn ~ 390 M), which is much less when compared with the studies that were earlier carried out to mimic the crowding environment (α-syn in 2M glucose).[25] They have resulted in 15N-HSQC (for α-syn, Fig.2b, blue colour contours) and 13C-HSQC (for monosaccharides, Fig. 2c) spectra. From Fig. 2b, it is clear from peak count that several peaks of α-syn are missing in the resultant 15N-HSQC spectrum (blue contours); this may be a consequence of exchange processes occurring in the protein. In the 13CHSQC (Fig. 2c) chemical shift resolution for most of the peaks is found to be poor, and this is because all the constituents have similar molecular structures. Then, for the same pure monosaccharide mixture, PROSMASH-HSQC2 experiments have been repeated with n=6, which has resulted in a BIRD homodecoupled 13C-HSQC spectrum and that has allowed us to identify all the chemical sites belonging to all the constituents (Fig. 2d). In the first sample, i.e., α-syn, 13C-HSQC part of the PROSMASH-HSQC2 results in very weak peaks (obscure in the noise level) at the natural abundance of 13C spins. On the other hand, for the second sample, i.e., 15N-HSQC part of the PROSMAS H-HSQC2 does not show any signals, as there are no amide resonances in the considered monosaccharides. Further, 1D-traces obtained from the 1H-1H coupled (Fig. 2f) and BIRD-homodecoupled 13C-HSQC (Fig. 2g) part of the PROSMASH-HSQC2 are also compared. This reveals that there is a significant enhancement in the chemical shift resolution, and signal sensitivity improvements are also evident. Clearly, this kind of homodecoupled spectra provide unambiguous measurement of chemical shift perturbations as demonstrated earlier by Zangger and co-workers.[26] Subsequently, another sample has been prepared by just adding the said monosaccharides (10 mg of each) to α-syn protein (300 M, at pH 7.4) and PROSMASH-HSQC2 experiment has been repeated in the BIRD homodecoupling mode. This time acquisition of only one PROSMASH-HSQC2 experiment is adequate to obtain both the 15N-HSQC (first part of the pulse scheme, for α-syn, Fig.2b, red colour contours), and 13C-BIRD-HSQC (second part of the pulse sequence, for the monosaccharide mixture, Fig.2e). The resultant 15N-HSQC spectrum (Fig.2b, red colour contours) of α-syn has shown many additional peaks when compared with the previously recorded 15N-HSQC for 4
the pure α-syn sample (Fig. 2b, blue colour contours). This improvement in the 15N-HSQC spectrum may be attributed to stabilization of conformation of α-syn in the presence of monosaccharides. At the same time, very clear information of chemical sites has been monitored from the 13C-BIRD-HSQC (Fig. 2e and Fig. 2h) of monosaccharides and this spectrum is exactly similar to the 13C-BIRD-HSQC that has earlier been recorded for the pure monosaccharide mixture (Fig. 2d and Fig. 2g). This could be due to high concentrations of sugars relative to the protein concentration and possibly also weak interactions between protein and small molecules; hence, the chemical shift resolution is still sustained for all the small molecules in the presence of α-syn protein. Here, it is worthwhile to discuss the sensitivity of 13C-HSQC resonances of α-syn (20 times scaled up in Fig. 2i) relative to that of the monosaccharide peaks (Fig. 2e). As a proof of principle, PROSMASH-HSQC2 experiments have been demonstrated at rather high concentrations of small molecules. Nevertheless, even at low concentrations of small molecules, this method would work reasonably well, as the 13C-HSQC of PROSMASH-HSQC2 exploits the benefits of relaxation differences between proteins and small molecules. In some cases, there is a possibility of overlap between protein (at the level of detectable signals) and small molecule chemical shifts in the 13C-HSQC of PROSMASH-HSQC2, but these could be unravelled by suitable controls with 13C-BIRD-HSQC in the PROSMASH-HSQC2 manner for pure small molecules (reference spectrum), which allows identification of small molecule chemical shifts, even when they are overlapped with protein signals. But, it must be realized that the present method does not work well with doubly labelled (13C and 15 N) proteins in the presence of unlabelled small molecules, as the 13C-HSQC of PROSMASH-HSQC2 shows both the protein and small molecule signals, and the BIRD decoupling does not work for the 13 C labelled proteins. Chemical shift perturbation values for both the protein (15N-HSQC) and monosaccharide mixture (13C-HSQC) have been measured,[27] while taking the chemical shift differences between pure compounds and mixtures. Here the CSP’s [
] for the observed peaks of α-syn
protein are found to be negligible (very small, Fig. 2j) and similarly most of the spins of monosaccharide mixture [
] have also not shown any considerable perturbations
(Fig. 2k). However, the peak intensities of several peaks in the protein spectra are significantly affected. Most of these peaks belong to Threonines or their neighbours, which suggests a possible interaction between OH groups of small molecules with the OH groups of the threonine residues of α-syn protein.
5
Figure 2: Molecular structures of various monosaccharides used in the present study (a). Superimposition of 15N-HSQC spectra recorded for α-syn protein in the PROSMASH-HSQC2 NMR pulse scheme (b). Here, blue and red colour contours represent the 15N-HSQC spectra obtained in absence and in presence of the monosaccharide mixture, respectively. The PROSMASH-HSQC2 resultant 13CHSQC spectra of monosaccharide mixture at different conditions are shown in (c-e). Further, the corresponding 1D-internal traces have also been depicted in (f-h). The residual 13C-BIRD-HSQC signals of protein at natural abundant level are shown (i), while raising the vertical scale of this spectrum by 20 folds with respect to the monosaccharide peaks (e). The calculated chemical shift perturbations of α-syn protein (from 15N-HSQC, as the residue number of α-syn) and monosaccharide mixture (from 13 C-BIRD-HSQC, as the spin number of monosaccharides; M: mannose, G: glucose, X: xylose) are depicted in (j) and (k), respectively.
The same α-syn and monosaccharide mixture sample has been incubated initially at room temperature for a week and then at 4oC for 23 days. Subsequently, PROSMASH-HSQC2 experiment has been repeated under the identical experimental conditions (which we have initially used earlier for the same sample). The resultant 15N-HSQC (Fig. 3a) and 13C-BIRD-HSQC (Fig. 3b) spectra of α-syn, and monosaccharide mixture, respectively, have been used for calculating the CSP’s. From the Fig. 3c, between immediately prepared and 30 days incubated samples, a nice correlation (~ 0.9) in the CSP’s is noticed with a slope of ~1.0 (Fig. 3c). Similarly, the measured CSP’s for most of the spins of monosaccharide mixture are also found to be very small (Fig. 3d), as in the immediately prepared sample. Indeed, in order to obtain the very important backbone chemical shift assignments of α-syn like intrinsically disordered proteins (IDP’s), to avoid the conformational flexibility (i.e., missing peaks), it would be a nice idea to consider the monosaccharides as stabilizing agents (as osmolytes with minimal interactions) and such samples can be used for acquiring the HNN (one experiment is adequate to obtain the backbone chemical shift assignments of IDP’s) [28] type of experiments. 6
Figure 3: The PROSMASH-HSQC2 resultant 15N-HSQC and 13C-BIRD-HSQC spectra of α-syn and the monosaccharide mixture (after 30 days incubation) are shown in (a) and (b), respectively. The calculated CSP’s of α-syn (from 15N-HSQC) protein at this condition have shown nice correlation with the CSP’s measured for the immediately prepared sample (c). Finally, 13C-BIRD-HSQC derived CSP’s of monosaccharide mixture are shown in (d).
From these findings, it is clear that the monosaccharides have a significant influence on the behaviour of α-syn in solution and this could be via change in viscosity of solution or/and interactions between OH groups belonging to threonine residues of α-syn and OH groups of monosaccharide moieties. As demonstrated in the present example, the PROSMASH-HSQC2 method is very useful for precise determination of CSP’s for both proteins and small molecules. As BIRD homodecoupling is an inbuilt part of the small molecule 13C-HSQC, chemical shift details can be obtained with improved sensitivities and high accuracies.
3. Conclusions The present work demonstrates the development and applications of a new NOAH experiment, namely, homodecoupled PROSMASH-HSQC2 experiment. The initial part of this pulse scheme acquires 15N-HSQC (for 15N labelled proteins) and the second part records 13C-HSQC (at natural abundance for small molecules) in the real-time BIRD homonuclear broadband decoupling fashion. An interesting example of protein and small molecular spin-system, i.e., α-syn and a mixture of monosaccharide mixture (glucose, mannose, and xylose) have been considered to demonstrate the efficacy of the PROSMASH-HSQC2 experiment. This method helps in the precise determination of chemical shift perturbations of both the small molecules and the proteins, from a single NMR experiment. Therefore, the present method saves the spectrometer time by two fold. On demand, these experimental times can be further reduced by invoking the non-uniform sampling schemes [29-31]. The utility of the presently proposed method lies in unambiguous determination of proteinsmall molecule (mixture) interactions and that will have an immense impact on drug discovery programs. Further, in some cases, simultaneous acquisition of both protein and small molecule signals may be essential; in such situations, the proposed method would be greatly helpful. Like in other BIRD homodecoupled pure shift NMR methods, the present PROSMASH-HSQC2 also does not work well for 13C isotopically labelled spin-systems and cannot decouple the diastereomeric protons.
7
4. Acknowledgements The authors thank National NMR Facility, TIFR, Mumbai. VMRK (PDF/2016/000365), and KR (PDF/2016/002063) are grateful to DST-SERB (Department of Science and Technology-Science and Engineering Research Board), Government of India, New Delhi, for the award of National Postdoctoral Fellowships. RVH is a DST sponsored JC Bose National Fellow. 5. References [1]. Ē. Kupče, and T. D. W. Claridge, NOAH: NMR Supersequences for Small Molecule Analysis and
Structure Elucidation, Angew. Chem. Int. Ed. Engl. 56 (2017) 11779–11783. [2]. Ē. Kupče, and T. D. W. Claridge, Molecular structure from a single NMR supersequence, Chem. Commun. 54 (2018) 7139–7142. [3]. C. Haasnoot, F. J. M. van de Ven, and C. W. Hilbers, COCONOSY. Combination of 2D correlated and 2D nuclear overhauser enhancement spectroscopy in a single experiment, J. Magn. Reson. 56 (1984) 343–349. [4]. Ē. Kupče, R. Freeman, and B. K. John, J. Am. Chem. Soc. 128 (2006) 9606–9607. [5]. Ē. Kupče, in Modern NMR Methodology, Vol. 335 (Eds.: H. Heise, S. Matthews), Springer, Berlin, 2013, pp 71-96. [6]. H. Kovacs, and Ē. Kupče, Magn. Reson. Chem. 54 (2016) 544–560. [7]. Ē. Kupče, and R. Freeman, J. Am. Chem. Soc. 130 (2008) 10788–10792. [8]. K. J. Donovan, Ē. Kupče, and L. Frydman, Multiple parallel 2D NMR acquisitions in a single scan, Angew. Chem. Int. Ed. 52 (2013) 4152–4155. [9]. J. G. Reddy, and R. V. Hosur, Parallel acquisition of 3D-HA(CA)NH and 3D-HACACO spectra, J. Biomol. NMR. 56 (2013) 77–84. 13 15 [10]. B. T. Farmer II, Simultaneous [ C, N]-HMQC, a pseudo-triple-resonance experiment, J. Magn. Reson. 93 (1991) 635–641. [11]. L. E. Kay, M. Wittekind, M. A. McCoy, M. S. Friedrichs, and L. Mueller, 4D NMR triple-resonance experiments for assignment of protein backbone nuclei using shared constant-time evolution periods, J. Magn. Reson. 98 (1992) 443–450. [12]. P. Nolis, M. Pérez-Trujillo, and T. Parella, Multiple FID Acquisition of Complementary HMBC Data, Angew. Chem. Int. Ed. 46 (2007) 7495–7497. [13]. T. Gopinath, and G. Veglia, Multiple acquisition of magic angle spinning solid-state NMR experiments using one receiver: Application to microcrystalline and membrane protein preparations, J. Magn. Reson. 253 (2015) 143–153. [14]. C. Wiedemann, P. Bellstedt, A. Kirschstein, S. Häfner, C. Herbst, M. Görlach, and R. Ramachandran, Sequential protein NMR assignments in the liquid state via sequential data acquisition, J. Magn. Reson. 239 (2014) 23–28. [15]. J. R. Garbow, D. P. Weitekamp, and A. Pines, Bilinear rotation decoupling of homonuclear scalar interactions, Chem. Phys. Lett. 93 (1982) 504–509. [16]. A. Lupulescu, G.L. Olsen, and L. Frydman, Toward single-shot pure-shift solution 1H NMR by trains of BIRD-based homonuclear decoupling, J. Magn. Reson. 218 (2012) 141–146. [17]. J. A. Aguilar, M. Nilsson, and G. A. Morris, Simple Proton Spectra from Complex Spin Systems: Pure Shift NMR Spectroscopy Using BIRD, Angew. Chem. Int. Ed. 50 (2011) 9716–9717. [18]. L. Paudel, R. W. Adams, P. Király, J. A. Aguilar, M. Foroozandeh, M. J. Cliff, M. Nilsson, P. Sándor, J. P. Waltho, and G. A. Morris, Simultaneously Enhancing Spectral Resolution and Sensitivity in Heteronuclear Correlation NMR Spectroscopy, Angew. Chem. Int. Ed. 52 (2013) 11616–11619. [19]. K. J. Donovan, and L. Frydman. HyperBIRD: a sensitivity-enhanced approach to collecting homonuclear-decoupled proton NMR spectra, Angew. Chem. Int. Ed. 54 (2015) 594–598.
8
[20]. M. G. Spillantini, R. A. Crowther, R. Jakes, M. Hasegawa, and M. Goedert, α-Synuclein in filamentous
inclusions of Lewy bodies from Parkinson’s disease and dementia with Lewy bodies, Proc. Natl. Acad. Sci. 95 (1998) 6469–6473. [21]. E. Ihse, H. Yamakado, X. M. van Wijk, R. Lawrence, J. D. Esko, and E. Masliah, Cellular internalization of alpha-synuclein aggregates by cell surface heparan sulfate depends on aggregate conformation and cell type, Sci. Rep. 7 (2017) 9008. [22]. A. K. Buell, C. Galvagnion, R. Gaspar, E. Sparr, M. Vendruscolo, T. P. J. Knowles, S. Linse, and C. M. Dobson, Solution conditions determine the relative importance of nucleation and growth processes in α-synuclein aggregation, Proc. Natl. Acad. Sci. 111 (2014) 7671–7676. [23]. W. Ariesandi, C. Chang, T. Chen, and Y. Chen , Temperature-Dependent Structural Changes of Parkinson's Alpha-Synuclein Reveal the Role of Pre-Existing Oligomers in Alpha-Synuclein Fibrillization, PLOS ONE. 8 (2013) e53487. [24]. T. Arakawa, and S. N. Timasheff, Stabilization of protein structure by sugars, Biochemistry, 25 (1982) 6536–6544. [25]. K-P. Wu, and J. Baum, Backbone Assignment and Dynamics of Human Α-Synuclein in Viscous 2 M Glucose Solution, Biomol. NMR. Assign. 5 (2011) 43–46. [26]. N. H. Meyer, and K. Zangger, Enhancing the resolution of multi-dimensional heteronuclear NMR spectra of intrinsically disordered proteins by homonuclear broadband decoupling, Chem. Commun. 50 (2014) 1488–1490. [27]. M. P. Williamson, Using chemical shift perturbation to characterise ligand binding, Prog. Nucl. Magne. Reson. Spec. 73 (2013) 1–16. [28]. V. M. R. Kakita, M. Bopardikar, V. K. Shukla, K. Rachineni, P. Ranjan, J. S. Singh, and R. V. Hosur, An efficient combination of BEST and NUS methods in multidimensional NMR spectroscopy for high throughput analysis of proteins, RSC Adv. 8 (2018) 17616–17621. [29]. A. Hansen, and R. Bruschweiler, Absolute Minimal Sampling in High-Dimensional NMR Spectroscopy, Angew. Chem. Int. Ed. 55 (2016) 14169–14172. [30]. M. Mobli, and J. C. Hoch, Nonuniform sampling and non-Fourier signal processing methods in multidimensional NMR, Prog. Nucl. Magn. Reson. Spec. 83 (2014) 21–41. [31]. Q. Wu, B. E. Coggins, and P. Zhou, Unbiased measurements of reconstruction fidelity of sparsely sampled magnetic resonance spectra, Nature Commun. 7 (2016) 12281.
9
Highlights
NOAH type of pure shift NMR pulse scheme, PROSMASH-HSQC2 has been designed 15 N-HSQC and 13C-BIRD-HSQC spectra can be recorded sequentially Permits precise determination of drug-protein interactions through CSP’s
10