- Email: [email protected]
1H isotropic shift correlation experiment mediated through 1H–1H RFDR mixing on a natural abundant sample under ultrafast MAS
Journal of Magnetic Resonance 258 (2015) 96–101
Contents lists available at ScienceDirect
Journal of Magnetic Resonance journal homepage: www.elsevier.com/locate/jmr
Proton-detected 3D 14N/14N/1H isotropic shift correlation experiment mediated through 1H–1H RFDR mixing on a natural abundant sample under ultrafast MAS Manoj Kumar Pandey a, Yusuke Nishiyama a,b,⇑ a b
RIKEN CLST-JEOL Collaboration Center, RIKEN, Yokohama, Kanagawa 230-0045, Japan JEOL RESONANCE Inc., Musashino, Akishima, Tokyo 196-8558, Japan
a r t i c l e
i n f o
Article history: Received 14 May 2015 Revised 22 June 2015 Available online 17 July 2015 Keywords: Solid-state NMR Homonuclear correlation 14 N–14N correlation 1 H–14N HMQC Ultrafast MAS
a b s t r a c t In this contribution, we have demonstrated a proton detection-based approach on a natural abundant powdered L-Histidine HCl–H2O sample at ultrafast magic angle spinning (MAS) to accomplish 14N/14N correlation from a 3D 14N/14N/1H isotropic shift correlation experiment mediated through 1H finite-pulse radio frequency-driven recoupling (fp-RFDR). Herein the heteronuclear magnetization transfer between 14N and 1H has been achieved by HMQC experiment, whereas 14N/14N correlation is attained through enhanced 1H–1H spin diffusion process due to 1H–1H dipolar recoupling during the RFDR mixing. While the use of ultrafast MAS (90 kHz) provides sensitivity enhancement through increased 1H transverse relaxation time (T2), the use of micro-coil probe which can withstand strong 14 N radio frequency (RF) fields further improves the sensitivity per unit sample volume. Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction Recent advances in sophisticated nuclear magnetic resonance (NMR) instrumentation have resulted in the development of sensitivity and resolution enhancement methods in solid state, making it simpler to get finer details of structure and dynamics of complex chemical and biological systems [1–6]. A possibility of attaining ultrafast (120 kHz) magic angle spinning (MAS) of solid samples by using a very small amount (0.3 mg) of sample is an excellent example to show how far we have come as far as micro-coil NMR probe design is concerned [7–15]. This has made solid-state NMR experiment with detection of protons now a routinely employed method resulting in a remarkable improvement in resolution and sensitivity which otherwise was not possible due to 1H line broadening in the presence of strong 1H–1H homonuclear dipolar couplings [16–27]. Subsequently, indirect observation of highly abundant (99.6%) but enormously challenging 14N spins (I = 1) with strong quadrupolar interactions (in the range of a few hundreds kHz to several MHz) through 1H-detection under ultrafast MAS spinning has opened up many new possibilities toward the development of methods relying on 14N nuclei. Since 14N nuclei are associated with a spin quantum number I = 1 and is devoid of ⇑ Corresponding author at: JEOL RESONANCE Inc., Musashino, Akishima, Tokyo 196-8558, Japan. E-mail address: [email protected] (Y. Nishiyama). http://dx.doi.org/10.1016/j.jmr.2015.06.012 1090-7807/Ó 2015 Elsevier Inc. All rights reserved.
any central transition line, all allowed single quantum transitions are largely affected by the first-order broadening which makes it difficult to observe 14N resonances from powdered samples. Fortunately, MAS completely gets rid of the first-order quadrupolar broadening resulting in 14N lineshape with numerous spinning sidebands and broadened only by the higher-order quadrupolar interaction terms [28]. Nevertheless, MAS experiments require accurate setting of magic angle to avoid any reintroduction of the 14 N first-order quadrupolar couplings. Numerous sidebands generated by MAS can be folded back on to the center peak if 14N signal is indirectly detected through spin-1/2 nuclei such as 1H and 13C (also termed as spy nuclei) by implementing 2D hetero nuclear multiple quantum coherence (HMQC) experiment with rotor-synchronized acquisition [29–33]. This leads to 14N NMR signal without any first-order quadrupolar broadening in the indirect dimension. Although there has been an enormous progress in the recent past for the development of methods based on 14N–1H HMQC experiment, there is still no report to the best of our knowledge that describes a method to get 14N/14N homonuclear isotropic shift correlation despite of high natural abundance of 14N and finds major applications in the material, pharmaceutical and biophysical sciences. The major challenge to overcome in order to achieve such correlation is the presence of very weak 14N–14N dipolar couplings. Likewise, 15N/15N homonuclear correlation experiments are also challenging again due to negligible 15N–15N dipolar interactions (still larger than 14N–14N dipolar interactions) even in fully
M.K. Pandey, Y. Nishiyama / Journal of Magnetic Resonance 258 (2015) 96–101 15
N labeled samples. The experimental methods based on direct as well as indirect detection of 15N to get 15N/15N chemical shift correlation are still emerging and only a handful of reports exist in the literature [8,34–39]. Recently, 15N/15N chemical shift correlation was obtained by relayed magnetization transfer through 1 H/1H spin diffusion utilizing 1H radio-frequency recoupling (RFDR) under ultrafast MAS [8]. A similar approach can also be used to overcome the difficulty associated with the measurement of weak 14N/14N isotropic shift correlations. Unlike 15N, the experimental methods based on 14N do not require any isotopic labeling of the samples as a result 14N/14N isotropic shift correlation experiments can be performed on natural abundant samples which is of a huge advantage. To this end, we present a 3D 14N/14N/1H isotropic shift correlation experiment mediated through RFDR mixing to correlate isotropic shifts of 14N, 14N and 1H from L-Histidine HCl– H2O sample at ultrafast MAS rate of 90 kHz wherein spin diffusion process is boosted through 1H–1H homonuclear dipolar recoupling by the application of a series of rotor-synchronized 180° RFDR pulses during the mixing time.
97
period through an enhanced spin diffusion process resulting from recoupled 1H–1H dipolar interactions. After the exchange process is accomplished, 1H transverse magnetization is again transferred via J-coupling and RDS to 14N nuclei during the second excitation period (texc) and then transferred back to 1H during the second reconversion period (trec) applied in the second HMQC block and finally, data is acquired during t3 period. 14N nuclei evolve under their respective isotropic shifts during t1 and t2 periods. Through proper phase cycling of 14N pulses single-quantum (SQ) coherence transfer pathways are selected during both 14N evolution periods. The expected loss in the 14N/14N cross-correlation efficiency resulting from four magnetization transfers that take place during the excitation and reconversion periods applied in two J-HMQC blocks is minimized by the use of ultrafast MAS (increases 1H T2 relaxation time by suppressing 1H–1H dipolar couplings) and the application of finite-pulse RFDR (fp-RFDR) during the mixing period to enhance the spin diffusion process which otherwise is suppressed by MAS.
3. Experimental 2. Pulse sequence The newly designed 3D 14N/14N/1H pulse sequence used in the present study is shown in Fig. 1C. The 1H transverse magnetization is first transferred via J-coupling and residual dipolar splitting (RDS) to 14N nuclei during the excitation period (texc) and again transferred back to 1H during the reconversion period (trec) applied in the first HMQC block. Longitudinal magnetization exchange between protons is allowed to take place during the RFDR mixing
All NMR experiments were carried out on a solid-state NMR spectrometer (JEOL ECZ600R) operating at a 1H Larmor frequency of 600 MHz using a 0.75 mm double-resonance ultrafast MAS probe (JEOL RESONANCE Inc.). 0.3 mg of L-Histidine HCl–H2O was packed into a 0.75 mm zirconia rotor and all measurements were carried out at room temperature (25 °C) under 90 kHz MAS rate. The pulse sequences implemented to get 2D 1H/1H, 2D 14N/1H and 3D 14N/14N/1H isotropic shift correlations are shown in Fig. 1A, B and C, respectively. To stabilize the spin system, 8
Fig. 1. (A) Proton-detected two-dimensional (2D) RFDR pulse sequence to correlate 1H/1H isotropic chemical shifts. A train of rotor-synchronized 180° RFDR pulses with XY414 phase cycling scheme is applied during the mixing time (tmix) to enhance the spin diffusion process accomplished by the recoupling of 1H–1H homonuclear dipolar interactions. The background signals from the NMR probe are suppressed by using a rotor-synchronized spin-echo sequence (sr–180°–sr) before the data acquisition. (B) Proton-detected 2D J-HMQC pulse sequence for 1H/14N isotropic shift correlation. The 14N evolution period (t1) under isotropic shifts is rotor-synchronized with respect to the sample spinning. (C) Proton-detected 3D 14N/14N/1H pulse sequence that correlates isotropic shifts of 14N, 14N and 1H. In the 3D experiment isotropic shifts of 14N nuclei are expressed during t1 and t2 periods while z-magnetization exchange between protons is achieved through recoupling of 1H–1H dipolar couplings by implementing fp-RFDR pulses during tmix. Similar to 2D 14N/1H J-HMQC experiment the 14N evolution periods (t1 and t2) are rotor-synchronized in 3D chemical shift correlation experiment. Single quantum coherence pathways were selected during 14N evolution both in 2D and 3D experiments obtained through proper phase cycling of 14N pulses. The phase cycling schemes are as follows: (a) /1 = {18(0), 18(180)}, /2 = {0, 180}, /3 = {2(0), 2(120), 2(240)}, /4 = {6(0), 6(120), 6(240)}, and /acq = {0, 180, 240, 60, 120, 300, 120, 300, 0, 180, 240, 60, 240, 60, 120, 300, 0, 180, 180, 0, 60, 240, 300, 120, 300, 120, 180, 0, 60, 240, 60, 240, 300, 120, 180, 0}, (b) /1 = {2(0), 2(180), 2(90), 2(270)}, /2 = {0}, /3 = {0, 180}, /4 = {0}, /aqc = {0, 180, 180, 0, 270, 90, 90, 270}, and (c) /1 = {0}, /2 = {4(0), 4(120), 4(240)}, /3 = {4(0), 4(240), 4(120)}, /4 = {0}, /5 = {12(0), 12(120), 12(240)}, /6 = {0, 180}, /7 = {0}, /8 = {2(0), 2(180)}, /9 = {0}, /acq = {3(0, 180, 180, 0), 3(240, 60, 60, 240), 3(120, 300, 300, 120)}.
98
M.K. Pandey, Y. Nishiyama / Journal of Magnetic Resonance 258 (2015) 96–101
dummy scans were applied prior to the start of all the measurements. The proton 90° pulse length was set at 0.6 ls. RFDR pulse sequence with XY414 phase cycling scheme (Fig. 1A) was implemented to find the optimal magnetization exchange efficiency between protons attached to correlating nitrogens through 1 H–1H homonuclear dipolar recoupling [40,41]. The proton 180° pulse length was set at 1.22 ls during the RFDR mixing time of 2.5 ms and a total of 224 p pulses. In 2D 14N/1H and 3D 14 N/14N/1H experiments 14N pulse length, which is not necessarily equal to a 90° pulse length, and both excitation and reconversion periods were optimized carefully to maximize the 14N–1H magnetization transfer efficiency. The 14N pulse length was set at 7.0 ls while the excitation and reconversion durations were set at sexc ¼ srec ¼ 2 ms. For the 3D measurement, 8 increments were set in both t1 and t2 dimensions and 240 scans were collected for each t1 and t2 increments in the presence of RFDR mixing, while 24 scans every 16 t1 and 16 t2 increments were set for the data collected in the absence of RFDR mixing. States-TPPI method was applied in both t1 and t2 dimensions to achieve pure absorption peaks. A recycle delay of 8 s was used in all the above experiments. The 3D data with and without RFDR, were acquired in 5.7 and 2.3 days, respectively. 14N isotropic shifts were referenced relative to CH3NO2. The drive pressure was actively controlled by pneumatic regulators to stabilize the sample spinning, and the fluctuation in the sample spinning monitored from the MAS controller unit was ±20 Hz. Delta NMR software (JEOL RESONANCE Inc.) was used to process the NMR data.
4. Results and discussion To achieve 14N/14N isotropic shift correlation in L-Histidine HCl– H2O from a 3D 14N/14N/1H correlation experiment under ultrafast MAS, first we found exact 14N peak positions from a 2D 14N/1H J-HMQC experiment to set the right values of 14N offset and spectral width required for the 3D measurement. This was followed by recording a 3D 14N/14N/1H isotropic shift correlation experiment. In the end, we evaluated the resulting signal intensity from the 3D experiment by measuring the filtering efficiencies of different blocks (HMQC, RFDR, HMQC-HMQC, HMQC-RFDR-HMQC) of the 3D pulse sequence (Fig. 1C). The magnetization transfer efficiency between the two cross correlating nitrogens through 1H–1H dipolar recoupling was maximized by carrying out a series of 2D 1H/1H chemical shift correlation experiments at 90 kHz MAS recorded with different mixing times. The pulse sequence implemented for this purpose is shown in Fig. 1A. The spin diffusion process is accelerated through 1H–1H dipolar recoupling during the mixing time. By measuring cross-peak intensities of protons (refer to Fig. S1 in the Supporting Information) directly bonded to nitrogens in the imidazole ring of L-Histidine HCl–H2O, the optimal magnetization transfer was achieved at RFDR mixing time of 2.5 ms and beyond which the magnetization exchange between these protons remained almost unchanged. It is worthwhile to mention here that the optimization procedure for RFDR mixing time described above can only be adopted if 1H signals associated with two cross correlating nitrogens are well separated. Similar procedure cannot be implemented in the case of overlapped 1H resonances of interest as change in the magnetization efficiency between protons during RFDR mixing cannot be evaluated. Before we could set a 3D 14N/14N/1H isotropic shift correlation experiment so as to have maximum resolution and sensitivity of 14 N/14N cross-peaks, it was important to carefully adjust the 14N offset and the spectral width of the indirect frequency dimensions. The main objective of carrying out these exercises was to get optimal spectral width of the indirect frequency dimensions to be set
with a minimum number of t1 and t2 points in order to sufficiently resolve 14N/14N cross peaks in a 3D 14N/14N/1H isotropic shift correlation experiment. Subsequently, we carried out a 2D 14 N/1H isotropic shift correlation experiment at 90 kHz MAS by implementing proton detection-based J-HMQC pulse sequence shown in Fig. 1B to get the exact values of 14N isotropic shifts and hence the right 14N offset. This was followed by setting up a series of 2D 14N/1H isotropic shift correlation experiments by varying spectral width of the indirect frequency dimension and t1 points while keeping the 14N offset exactly at the center of the two imidazole 14N resonances (114.4 ppm) of L-Histidine HCl– H2O. From these experiments a 14N spectral width of 5 kHz and 8 t1 points were found to be sufficient enough to get well-resolved 14 N/1H correlations (data not shown). Resonance assigned 1D 1H spectrum and 2D 14N/1H HMQC spectrum of L-Histidine HCl–H2O recorded at 90 kHz MAS are shown in Fig. 2. It is to be noted that a short transverse relaxation time (T2) of NH+3 and its longer 14 N–14N distances from nitrogens in the imidazole ring restricted us to optimize experimental conditions so as to achieve 14N/14N isotropic shift correlation only between nitrogens of the imidazole ring. In addition, the 14N isotropic shifts are quite different from 15 N chemical shifts because 14N isotropic shift is not only described by its isotropic chemical shift (riso) but a combination of riso and isotropic second-order quadrupolar shift. Once all the parameters were precisely optimized a 3D 14 N/14N/1H isotropic shift correlation experiment by implementing the J-based HMQC pulse sequence mediated through RFDR (see Fig. 1C) was performed to accomplish 14N/14N cross-correlation peaks. Proton-detected 2D 14N/14N isotropic shift correlation spectra, which are projections of 3D 14N/14N/1H spectra on to 14N/14N planes in the absence and presence of RFDR mixing, are shown in Fig. 3A and B, respectively. Additionally, 2D 14N/14N planes extracted at isotropic 1H chemical shifts of 12.6 and 17.1 ppm from a 3D 14N/14N/1H spectrum recorded with RFDR mixing are shown in Fig. 3C and D, respectively. As per our expectation we do not observe any 14N/14N cross-correlation in the absence of any 1 H–1H RFDR mixing. While in the presence of 1H–1H RFDR mixing time of 2.5 ms, which provides a medium for magnetization transfer between 14N nuclei to be accomplished through 1H–1H dipolar recoupling, we clearly observe 14N/14N cross-peaks. Also seen from the 2D planes are the less sensitive peaks (circled resonances in Fig. 3C and D) around but not exactly at the 14N/14N cross-peak positions. This observation seems rather intriguing in view of the fact that 2D slices at isotropic 1H chemical shifts should result in 14 N/14N cross-peaks only along the vertical axes. The appearance of additional resonances along the horizontal axis should be attributed to the tailing of the cross-peaks resulting from the presence of the second-order quadrupolar broadening. It is important to mention here that in the absence of any RFDR mixing there could also have been a possibility of magnetization exchange between 1H nuclei but only through the application of a very long mixing time (several hundred milliseconds) required to enhance the spin diffusion process. Finally, we measured filtering efficiencies by recording proton-detected 1D spectra using RFDR, HMQC, HMQC-HMQC, and HMQC-RFDR-HMQC blocks implemented in the 3D pulse sequence and compared the signal intensity with respect to that observed from a spin-echo experiment (refer to Fig. S2 in the Supporting Information for 1D spectra). Table 1 lists the percentage filtering efficiencies of proton resonances (17.1 and 12.6 ppm) which are directly bonded to nitrogens from all the different blocks described above in comparison to a spin-echo experiment. As seen from Table 1 the loss in the magnetization efficiency resulting from RFDR with 2.5 ms mixing time is in the range of 6–10% which clearly suggests that the total 1H magnetization is well retained during the RFDR mixing. This was followed by
M.K. Pandey, Y. Nishiyama / Journal of Magnetic Resonance 258 (2015) 96–101
Fig. 2. Molecular structure and assigned 1D 1H NMR spectrum (A), and 2D 600 MHz NMR spectrometer.
99
14
N/1H HMQC spectrum (B) of L-Histidine HCl–H2O at a MAS rate of 90 kHz collected from a
Fig. 3. Representative regions of proton-detected 2D 14N/14N chemical shift correlation spectra of L-Histidine HCl–H2O in the absence (A) and presence (B) of 1H/1H RFDR mixing projected from 3D 14N/14N/1H isotropic shift correlation spectra on 14N/14N planes recorded at 90 kHz ultrafast MAS using the pulse sequence shown in Fig. 1C. 2D 14 N/14N planes extracted from a 3D 14N/14N/1H isotropic shift correlation spectrum with 1H/1H RFDR mixing at isotropic 1H chemical shifts of 12.6 ppm (C) and 17.1 ppm (D). 16 t1 and t2 points every 24 scans, while 8 t1 and t2 points every 240 scans were used for 3D 14N/14N/1H isotropic shift correlation data collected without RFDR mixing and with RFDR mixing (2.5 ms), respectively, with 5 kHz spectral width and 8 s recycle delay.
Table 1 Observed magnetization transfer (filtering) efficiencies from 1H 1D measurements utilizing RFDR, HMQC, HMQC-HMQC, and HMQC-RFDR-HMQC pulse sequences as compared to the sensitivity measured from a 1H spin-echo experiment. 1
H peak (ppm)
RFDR (%)
HMQC (%)
HMQC-HMQC (%)
HMQC-RFDR-HMQC (%)
17.1 (NH1) 12.6 (NH2)
93.79 90.00
7.55 9.75
0.73 0.99
0.22 0.26
measuring the filtering efficiency resulting from a proton-detected 1D 14N/1H HMQC experiment. In general, the magnetization transfer efficiency after one HMQC filtering should be in the range
of 1–10% of the intensity observed from a spin-echo experiment. From our experiment this value is in the range of 7.5–10% for both 1 H resonances (Table 1), which is almost close to the expected value. Likewise, if we assume similar transfer efficiency also from the second HMQC sequence then the filtering efficiency after the application of two HMQC blocks without 1H–1H RFDR mixing (tmix = 0) should be equal to 1% (10% of the 10%) of the spin-echo intensity, which is also in a close agreement with our experimental observation (Table 1). It is worth to point here that there can be a possibility of better filtering efficiency from the second HMQC as the first HMQC selects only those nuclei that have a favorable orientation for the HMQC filtering. Nonetheless, this phenomenon is
100
M.K. Pandey, Y. Nishiyama / Journal of Magnetic Resonance 258 (2015) 96–101
not observed from our measurements as we did not notice any gain in the magnetization from the second HMQC. Although the total 1H magnetization is well preserved during the RFDR as mentioned earlier, there is a substantial loss in the magnetization of the imidazole ring NH protons due to exchange with the remaining protons that are not directly bonded to nitrogens. This phenomenon significantly reduces the 1H magnetization of the NH protons after the HMQC-RFDR-HMQC filtering which is also observed from our 1D HMQC-RFDR-HMQC measurement with 2.5 ms mixing for both the resonances (Table 1). The loss in the magnetization of protons bonded to nitrogens is ca 70% of the total signal intensity observed from 1D HMQC-HMQC experiment in the absence of RFDR mixing. The above observation is in accordance with the measured signal intensities (both auto and cross-correlation peaks) of the NH protons from the 2D 1H/1H RFDR correlation spectra collected with (2.5 ms) and without mixing (refer to Fig. S1 in the Supporting Information). More importantly, the magnetization efficiency left after HMQC-RFDR-HMQC block is sufficient enough to measure a 3D 14N/14N/1H correlation experiment on a natural abundant sample. In fact, a 3D 15N/15N/1H correlation experiment with 1H–1H RFDR mixing demonstrated in our previous work on a fully 15N labeled N-acetyl-L-valyl-L-leucine (NAVL) sample, the final magnetization that survived was 1.6% of the original signal intensity which is higher as compared to the magnetization that survives for detection in the present study. The main reason for the observed difference is the survival of 31% and 10% of the total 1H magnetization after the first and second cross polarization-based 1 H–15N–1H magnetization transfer, respectively, unlike the present case wherein ca 10% and 1% of the total 1H magnetization (Table 1) survive after first and second HMQC filtering, respectively. We would like to mention here that the required time for the collection of a 3D 14N/14N/1H chemical shift correlation data to achieve good sensitivity of the 14N/14N cross-peaks is relatively longer in the present study. Apart from the inherent difficulties in the measurement due to the presence of a weak 14N–14N homonuclear dipolar interaction and the first-order quadrupolar couplings, the use of a tiny amount of sample at natural abundance can lead to loss in the sensitivity of the cross-peaks. Additionally, slight fluctuations in the sample spinning and rotor phase could also add to the loss in the sensitivity as fundamental transition frequency of 14N is extremely sensitive to exact magic angle setting and stable spinning. Better sensitivity in the 14N dimension could be achieved by implementing D-HMQC experiment wherein magnetization transfer occurs via recoupled 14N–1H heteronuclear dipolar interactions instead of J-coupling and RDS [33]. As a cautionary note we would like to mention here that D-HMQC experiment is more sensitive toward the fluctuations in the sample spinning, and in such cases J-HMQC with 1H–1H decoupling can be implemented as an alternative wherein sensitivity is enhanced through increased T2 relaxation time [42]. Sample spinning faster than 90 kHz with the application of 1H–1H decoupling can also result in the gain in the sensitivity. Furthermore, selective 1H–1H recoupling methods instead of RFDR can also be implemented for the sensitivity enhancement as loss in the magnetization due to its distribution to protons that are not directly bonded to 14N can be avoided. It is worth pointing that a limited resolution both in the 1H and 14N dimensions can hamper the application of the proposed method to large molecules like proteins due to possible signal overlaps. However, we believe that an addition of high resolution dimension like 13C can overcome this difficulty. 5. Conclusions In summary, we have presented a new proton detection-based 3D pulse sequence to get 14N/14N/1H isotropic shift correlation in natural abundant powdered L-Histidine HCl–H2O (0.3 mg) under
ultrafast MAS (90 kHz) condition by utilizing the magnetization transfer between 14N and 1H through J-coupling and RDS. We have applied 1H–1H RFDR sequence during mixing time to enhance the spin diffusion process resulting from 1H–1H dipolar recoupling so as to accomplish 14N/14N correlation. The new approach for observing 14N/14N homonuclear isotropic shift correlation demonstrates a strong ability of proton detection-based measurements with the application of RFDR-based 1H–1H dipolar recoupling under ultrafast MAS condition that can be used for structural studies of challenging chemical and biological samples at natural abundance. Although the method presented in this study to accomplish 14 N/14N correlation has a limitation with respect to the measurement time, we believe that better sensitivity in 14N dimensions can be achieved by implementing already existing and/or new state-of-the-art solid-state NMR techniques. Acknowledgment This work was supported by the NMR Platform from the Ministry of Education, Culture, Sports, Science and Technology, Japan. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jmr.2015.06.012. References [1] S. Parthasarathy, Y. Nishiyama, Y. Ishii, Sensitivity and resolution enhanced solid-state NMR for paramagnetic systems and biomolecules under very fast magic angle spinning, Acc. Chem. Res. 46 (2013) 2127–2135. [2] J.P. Demers, V. Chevelkov, A. Lange, Progress in correlation spectroscopy at ultra-fast magic-angle spinning: basic building blocks and complex experiments for the study of protein structure and dynamics, Solid State Nucl. Magn. Reson. 40 (2011) 101–113. [3] P. Paluch, T. Pawlak, J.P. Amoureux, M.J. Potrzebowski, Simple and accurate determination of X–H distances under ultra-fast MAS NMR, J. Magn. Reson. 233 (2013) 56–63. [4] S. Wang, R.A. Munro, L. Shi, I. Kawamura, T. Okitsu, A. Wada, S.Y. Kim, K.H. Jung, L.S. Brown, V. Ladizhansky, Solid-state NMR spectroscopy structure determination of a lipid-embedded heptahelical membrane protein, Nat. Methods 10 (2013) 1007–1012. [5] M.E. Ward, L. Shi, E. Lake, S. Krishnamurthy, H. Hutchins, L.S. Brown, V. Ladizhansky, Proton-detected solid-state NMR reveals intramembrane polar networks in a seven-helical transmembrane protein proteorhodopsin, J. Am. Chem. Soc. 133 (2011) 17434–17443. [6] I. Bertini, C. Luchinat, G. Parigi, E. Ravera, B. Reif, P. Turano, Solid-state NMR of proteins sedimented by ultracentrifugation, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 10396–10399. [7] Y.Q. Ye, M. Malon, C. Martineau, F. Taulelle, Y. Nishiyama, Rapid measurement of multidimensional 1H solid-state NMR spectra at ultra-fast MAS frequencies, J. Magn. Reson. 239 (2014) 75–80. [8] Y. Nishiyama, M. Malon, Y. Ishii, A. Ramamoorthy, 3D (1)(5)N/(1)(5)N/(1)H chemical shift correlation experiment utilizing an RFDR-based 1H/1H mixing period at 100 KHz MAS, J. Magn. Reson. 244 (2014) 1–5. [9] V. Agarwal, S. Penzel, K. Szekely, R. Cadalbert, E. Testori, A. Oss, J. Past, A. Samoson, M. Ernst, A. Bockmann, B.H. Meier, De novo 3D structure determination from sub-milligram protein samples by solid-state 100 KHz MAS NMR spectroscopy, Angew. Chem. 53 (2014) 12253–12256. [10] J.M. Lamley, D. Iuga, C. Oster, H.J. Sass, M. Rogowski, A. Oss, J. Past, A. Reinhold, S. Grzesiek, A. Samoson, J.R. Lewandowski, Solid-state NMR of a protein in a precipitated complex with a fulllength antibody, J. Am. Chem. Soc. 136 (2014) 16800–16806. [11] T. Kobayashi, K. Mao, P. Paluch, A. Nowak-Krol, J. Sniechowska, Y. Nishiyama, D.T. Gryko, M.J. Potrzebowski, M. Pruski, Study of intermolecular interactions in the corrole matrix by solid-state NMR under 100 KHz MAS and theoretical calculations, Angew. Chem. 52 (2013) 14108–14111. [12] Y. Nishiyama, T. Kobayashi, M. Malon, D. Singappuli-Arachchige, I.I. Slowing, M. Pruski, Studies of minute quantities of natural abundance molecules using 2D heteronuclear correlation spectroscopy under 100 KHz MAS, Solid State Nucl. Magn. Reson. 66–67 (2015) 56–61. [13] A. Bockmann, M. Ernst, B.H. Meier, Spinning proteins, the faster, the better?, J Magn. Reson. 253 (2015) 71–79. [14] V. Agarwal, T. Tuherm, A. Reinhold, J. Past, A. Samoson, M. Ernst, B.H. Meier, Amplitude-modulated low-power decoupling sequences for fast magic-angle spinning NMR, Chem. Phys. Lett. 583 (2013) 1–7.
M.K. Pandey, Y. Nishiyama / Journal of Magnetic Resonance 258 (2015) 96–101 [15] K. Mroue, Y. Nishiyama, M.K. Pandey, B. Gong, E. McNerny, D. Kohn, M. Morris, A. Ramamoorthy, Proton-detected solid-state NMR spectroscopy of bone with ultrafast magic angle spinning, Sci. Rep. 5 (2015) 11991. [16] B. Reif, R.G. Griffin, 1H detected 1H,15N correlation spectroscopy in rotating solids, J. Magn. Reson. 160 (2003) 78–83. [17] D.H. Zhou, G. Shah, M. Cormos, C. Mullen, D. Sandoz, C.M. Rienstra, Protondetected solid-state NMR spectroscopy of fully protonated proteins at 40 KHz magic-angle spinning, J. Am. Chem. Soc. 129 (2007) 11791–11801. [18] Y. Ishii, R. Tycko, Sensitivity enhancement in solid state 15N NMR by indirect detection with high-speed magic angle spinning, J. Magn. Reson. 142 (2000) 199–204. [19] K. Mao, J.W. Wiench, V.S.Y. Lin, M. Pruski, Indirectly detected through-bond chemical shift correlation NMR spectroscopy in solids under fast MAS: studies of organic-inorganic hybrid materials, J. Magn. Reson. 196 (2009) 92–95. [20] S.M. Althaus, K.M. Mao, J.A. Stringer, T. Kobayashi, M. Pruski, Indirectly detected heteronuclear correlation solid-state NMR spectroscopy of naturally abundant 15N nuclei, Solid State Nucl. Magn. Reson. 57–58 (2014) 17–21. [21] K.M. Mao, M. Pruski, Directly and indirectly detected through-bond heteronuclear correlation solid-state NMR spectroscopy under fast MAS, J. Magn. Reson. 201 (2009) 165–174. [22] V. Agarwal, A. Diehl, N. Skrynnikov, B. Reif, High resolution 1H detected 1H,13C correlation spectra in MAS solid-state NMR using deuterated proteins with selective 1H,2H isotopic labeling of methyl groups, J. Am. Chem. Soc. 128 (2006) 12620–12621. [23] S. Asami, M. Rakwalska-Bange, T. Carlomagno, B. Reif, Protein-RNA interfaces probed by 1H-detected MAS solid-state NMR spectroscopy, Angew. Chem. 52 (2013) 2345–2349. [24] S. Asami, B. Reif, Proton-detected solid-state NMR spectroscopy at aliphatic sites: application to crystalline systems, Acc. Chem. Res. 46 (2013) 2089–2097. [25] R. Linser, V. Chevelkov, A. Diehl, B. Reif, Sensitivity enhancement using paramagnetic relaxation in MAS solid-state NMR of perdeuterated proteins, J. Magn. Reson. 189 (2007) 209–216. [26] R. Linser, M. Dasari, M. Hiller, V. Higman, U. Fink, J.M. Lopez del Amo, S. Markovic, L. Handel, B. Kessler, P. Schmieder, D. Oesterhelt, H. Oschkinat, B. Reif, Proton-detected solid-state NMR spectroscopy of fibrillar and membrane proteins, Angew. Chem. 50 (2011) 4508–4512. [27] Y. Nishiyama, M. Malon, M.J. Potrzebowski, P. Paluch, J.P. Amoureux, Accurate NMR determination of C–H or N–H distances for unlabeled molecules, Solid State Nucl. Magn. Reson. in press. [28] E. Konstantin, B.M. Fung, High-resolution 14N NMR in polycrystalline solids, J. Chem. Phys. 110 (1999) 7977–7982.
101
[29] S. Cavadini, Indirect detection of nitrogen-14 in solid-state NMR spectroscopy, Prog. Nucl. Magn. Reson. Spectrosc. 56 (2010) 46–77. [30] S. Cavadini, S. Antonijevic, A. Lupulescu, G. Bodenhausen, Indirect detection of nitrogen-14 in solids via protons by nuclear magnetic resonance spectroscopy, J. Magn. Reson. 182 (2006) 168–172. [31] S. Cavadini, A. Lupulescu, S. Antonijevic, G. Bodenhausen, Nitrogen-14 NMR spectroscopy using residual dipolar splittings in solids, J. Am. Chem. Soc. 128 (2006) 7706–7707. [32] Z.H. Gan, Measuring amide nitrogen quadrupolar coupling by high-resolution 14 N/13C NMR correlation under magic-angle spinning, J. Am. Chem. Soc. 128 (2006) 6040–6041. [33] Y. Nishiyama, Y. Endo, T. Nemoto, H. Utsumi, K. Yamauchi, K. Hioka, T. Asakura, Very fast magic angle spinning 1H–14N 2D solid-state NMR: sub-micro-liter sample data collection in a few minutes, J. Magn. Reson. 208 (2011) 44–48. [34] J. Xu, J. Struppe, A. Ramamoorthy, Two-dimensional homonuclear chemical shift correlation established by the cross-relaxation driven spin diffusion in solids, J. Chem. Phys. 128 (2008). [35] A. Ramamoorthy, L.M. Gierasch, S.J. Opella, 4-dimensional solid-state NMR experiment that correlates the chemical-shift and dipolar-coupling frequencies of 2 heteronuclei with the exchange of dilute-spin magnetization, J. Magn. Reson. B 109 (1995) 112–116. [36] Y.F. Wei, A. Ramamoorthy, 2D 15N–15N isotropic chemical shift correlation established by 1H–1H dipolar coherence transfer in biological solids, Chem. Phys. Lett. 342 (2001) 312–316. [37] J.R. Lewandowski, G. De Paepe, M.T. Eddy, R.G. Griffin, N-15-N-15 proton assisted recoupling in magic angle spinning NMR, J. Am. Chem. Soc. 131 (2009) 5769–5776. [38] M. Giffard, S. Hediger, J.R. Lewandowski, M. Bardet, J.P. Simorre, R.G. Griffin, G. De Paepe, Compensated second-order recoupling: application to third spin assisted recoupling, Phys. Chem. Chem. Phys. 14 (2012) 7246–7255. [39] A.K. Khitrin, J.D. Xu, A. Ramamoorthy, Cross-correlations between low-gamma nuclei in solids via a common dipolar bath, J. Magn. Reson. 212 (2011) 95–101. [40] Y. Nishiyama, R. Zhang, A. Ramamoorthy, Finite-pulse radio frequency driven recoupling with phase cycling for 2D 1H/1H correlation at ultrafast MAS frequencies, J. Magn. Reson. 243 (2014) 25–32. [41] R. Zhang, Y. Nishiyama, P. Sun, A. Ramamoorthy, Phase cycling schemes for finite-pulse-RFDR MAS solid state NMR, J. Magn. Reson. (2015), http:// dx.doi.org/10.1016/j.jmr.2014.12.010. [42] Y. Nishiyama, X. Lu, J. Trebosc, O. Lafon, Z. Gan, P.K. Madhu, J.P. Amoureux, Practical choice of 1H–1H decoupling schemes in through-bond 1H–{X} HMQC experiments at ultra-fast MAS, J. Magn. Reson. 214 (2012) 151–158.
Contents lists available at ScienceDirect
Journal of Magnetic Resonance journal homepage: www.elsevier.com/locate/jmr
Proton-detected 3D 14N/14N/1H isotropic shift correlation experiment mediated through 1H–1H RFDR mixing on a natural abundant sample under ultrafast MAS Manoj Kumar Pandey a, Yusuke Nishiyama a,b,⇑ a b
RIKEN CLST-JEOL Collaboration Center, RIKEN, Yokohama, Kanagawa 230-0045, Japan JEOL RESONANCE Inc., Musashino, Akishima, Tokyo 196-8558, Japan
a r t i c l e
i n f o
Article history: Received 14 May 2015 Revised 22 June 2015 Available online 17 July 2015 Keywords: Solid-state NMR Homonuclear correlation 14 N–14N correlation 1 H–14N HMQC Ultrafast MAS
a b s t r a c t In this contribution, we have demonstrated a proton detection-based approach on a natural abundant powdered L-Histidine HCl–H2O sample at ultrafast magic angle spinning (MAS) to accomplish 14N/14N correlation from a 3D 14N/14N/1H isotropic shift correlation experiment mediated through 1H finite-pulse radio frequency-driven recoupling (fp-RFDR). Herein the heteronuclear magnetization transfer between 14N and 1H has been achieved by HMQC experiment, whereas 14N/14N correlation is attained through enhanced 1H–1H spin diffusion process due to 1H–1H dipolar recoupling during the RFDR mixing. While the use of ultrafast MAS (90 kHz) provides sensitivity enhancement through increased 1H transverse relaxation time (T2), the use of micro-coil probe which can withstand strong 14 N radio frequency (RF) fields further improves the sensitivity per unit sample volume. Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction Recent advances in sophisticated nuclear magnetic resonance (NMR) instrumentation have resulted in the development of sensitivity and resolution enhancement methods in solid state, making it simpler to get finer details of structure and dynamics of complex chemical and biological systems [1–6]. A possibility of attaining ultrafast (120 kHz) magic angle spinning (MAS) of solid samples by using a very small amount (0.3 mg) of sample is an excellent example to show how far we have come as far as micro-coil NMR probe design is concerned [7–15]. This has made solid-state NMR experiment with detection of protons now a routinely employed method resulting in a remarkable improvement in resolution and sensitivity which otherwise was not possible due to 1H line broadening in the presence of strong 1H–1H homonuclear dipolar couplings [16–27]. Subsequently, indirect observation of highly abundant (99.6%) but enormously challenging 14N spins (I = 1) with strong quadrupolar interactions (in the range of a few hundreds kHz to several MHz) through 1H-detection under ultrafast MAS spinning has opened up many new possibilities toward the development of methods relying on 14N nuclei. Since 14N nuclei are associated with a spin quantum number I = 1 and is devoid of ⇑ Corresponding author at: JEOL RESONANCE Inc., Musashino, Akishima, Tokyo 196-8558, Japan. E-mail address: [email protected] (Y. Nishiyama). http://dx.doi.org/10.1016/j.jmr.2015.06.012 1090-7807/Ó 2015 Elsevier Inc. All rights reserved.
any central transition line, all allowed single quantum transitions are largely affected by the first-order broadening which makes it difficult to observe 14N resonances from powdered samples. Fortunately, MAS completely gets rid of the first-order quadrupolar broadening resulting in 14N lineshape with numerous spinning sidebands and broadened only by the higher-order quadrupolar interaction terms [28]. Nevertheless, MAS experiments require accurate setting of magic angle to avoid any reintroduction of the 14 N first-order quadrupolar couplings. Numerous sidebands generated by MAS can be folded back on to the center peak if 14N signal is indirectly detected through spin-1/2 nuclei such as 1H and 13C (also termed as spy nuclei) by implementing 2D hetero nuclear multiple quantum coherence (HMQC) experiment with rotor-synchronized acquisition [29–33]. This leads to 14N NMR signal without any first-order quadrupolar broadening in the indirect dimension. Although there has been an enormous progress in the recent past for the development of methods based on 14N–1H HMQC experiment, there is still no report to the best of our knowledge that describes a method to get 14N/14N homonuclear isotropic shift correlation despite of high natural abundance of 14N and finds major applications in the material, pharmaceutical and biophysical sciences. The major challenge to overcome in order to achieve such correlation is the presence of very weak 14N–14N dipolar couplings. Likewise, 15N/15N homonuclear correlation experiments are also challenging again due to negligible 15N–15N dipolar interactions (still larger than 14N–14N dipolar interactions) even in fully
M.K. Pandey, Y. Nishiyama / Journal of Magnetic Resonance 258 (2015) 96–101 15
N labeled samples. The experimental methods based on direct as well as indirect detection of 15N to get 15N/15N chemical shift correlation are still emerging and only a handful of reports exist in the literature [8,34–39]. Recently, 15N/15N chemical shift correlation was obtained by relayed magnetization transfer through 1 H/1H spin diffusion utilizing 1H radio-frequency recoupling (RFDR) under ultrafast MAS [8]. A similar approach can also be used to overcome the difficulty associated with the measurement of weak 14N/14N isotropic shift correlations. Unlike 15N, the experimental methods based on 14N do not require any isotopic labeling of the samples as a result 14N/14N isotropic shift correlation experiments can be performed on natural abundant samples which is of a huge advantage. To this end, we present a 3D 14N/14N/1H isotropic shift correlation experiment mediated through RFDR mixing to correlate isotropic shifts of 14N, 14N and 1H from L-Histidine HCl– H2O sample at ultrafast MAS rate of 90 kHz wherein spin diffusion process is boosted through 1H–1H homonuclear dipolar recoupling by the application of a series of rotor-synchronized 180° RFDR pulses during the mixing time.
97
period through an enhanced spin diffusion process resulting from recoupled 1H–1H dipolar interactions. After the exchange process is accomplished, 1H transverse magnetization is again transferred via J-coupling and RDS to 14N nuclei during the second excitation period (texc) and then transferred back to 1H during the second reconversion period (trec) applied in the second HMQC block and finally, data is acquired during t3 period. 14N nuclei evolve under their respective isotropic shifts during t1 and t2 periods. Through proper phase cycling of 14N pulses single-quantum (SQ) coherence transfer pathways are selected during both 14N evolution periods. The expected loss in the 14N/14N cross-correlation efficiency resulting from four magnetization transfers that take place during the excitation and reconversion periods applied in two J-HMQC blocks is minimized by the use of ultrafast MAS (increases 1H T2 relaxation time by suppressing 1H–1H dipolar couplings) and the application of finite-pulse RFDR (fp-RFDR) during the mixing period to enhance the spin diffusion process which otherwise is suppressed by MAS.
3. Experimental 2. Pulse sequence The newly designed 3D 14N/14N/1H pulse sequence used in the present study is shown in Fig. 1C. The 1H transverse magnetization is first transferred via J-coupling and residual dipolar splitting (RDS) to 14N nuclei during the excitation period (texc) and again transferred back to 1H during the reconversion period (trec) applied in the first HMQC block. Longitudinal magnetization exchange between protons is allowed to take place during the RFDR mixing
All NMR experiments were carried out on a solid-state NMR spectrometer (JEOL ECZ600R) operating at a 1H Larmor frequency of 600 MHz using a 0.75 mm double-resonance ultrafast MAS probe (JEOL RESONANCE Inc.). 0.3 mg of L-Histidine HCl–H2O was packed into a 0.75 mm zirconia rotor and all measurements were carried out at room temperature (25 °C) under 90 kHz MAS rate. The pulse sequences implemented to get 2D 1H/1H, 2D 14N/1H and 3D 14N/14N/1H isotropic shift correlations are shown in Fig. 1A, B and C, respectively. To stabilize the spin system, 8
Fig. 1. (A) Proton-detected two-dimensional (2D) RFDR pulse sequence to correlate 1H/1H isotropic chemical shifts. A train of rotor-synchronized 180° RFDR pulses with XY414 phase cycling scheme is applied during the mixing time (tmix) to enhance the spin diffusion process accomplished by the recoupling of 1H–1H homonuclear dipolar interactions. The background signals from the NMR probe are suppressed by using a rotor-synchronized spin-echo sequence (sr–180°–sr) before the data acquisition. (B) Proton-detected 2D J-HMQC pulse sequence for 1H/14N isotropic shift correlation. The 14N evolution period (t1) under isotropic shifts is rotor-synchronized with respect to the sample spinning. (C) Proton-detected 3D 14N/14N/1H pulse sequence that correlates isotropic shifts of 14N, 14N and 1H. In the 3D experiment isotropic shifts of 14N nuclei are expressed during t1 and t2 periods while z-magnetization exchange between protons is achieved through recoupling of 1H–1H dipolar couplings by implementing fp-RFDR pulses during tmix. Similar to 2D 14N/1H J-HMQC experiment the 14N evolution periods (t1 and t2) are rotor-synchronized in 3D chemical shift correlation experiment. Single quantum coherence pathways were selected during 14N evolution both in 2D and 3D experiments obtained through proper phase cycling of 14N pulses. The phase cycling schemes are as follows: (a) /1 = {18(0), 18(180)}, /2 = {0, 180}, /3 = {2(0), 2(120), 2(240)}, /4 = {6(0), 6(120), 6(240)}, and /acq = {0, 180, 240, 60, 120, 300, 120, 300, 0, 180, 240, 60, 240, 60, 120, 300, 0, 180, 180, 0, 60, 240, 300, 120, 300, 120, 180, 0, 60, 240, 60, 240, 300, 120, 180, 0}, (b) /1 = {2(0), 2(180), 2(90), 2(270)}, /2 = {0}, /3 = {0, 180}, /4 = {0}, /aqc = {0, 180, 180, 0, 270, 90, 90, 270}, and (c) /1 = {0}, /2 = {4(0), 4(120), 4(240)}, /3 = {4(0), 4(240), 4(120)}, /4 = {0}, /5 = {12(0), 12(120), 12(240)}, /6 = {0, 180}, /7 = {0}, /8 = {2(0), 2(180)}, /9 = {0}, /acq = {3(0, 180, 180, 0), 3(240, 60, 60, 240), 3(120, 300, 300, 120)}.
98
M.K. Pandey, Y. Nishiyama / Journal of Magnetic Resonance 258 (2015) 96–101
dummy scans were applied prior to the start of all the measurements. The proton 90° pulse length was set at 0.6 ls. RFDR pulse sequence with XY414 phase cycling scheme (Fig. 1A) was implemented to find the optimal magnetization exchange efficiency between protons attached to correlating nitrogens through 1 H–1H homonuclear dipolar recoupling [40,41]. The proton 180° pulse length was set at 1.22 ls during the RFDR mixing time of 2.5 ms and a total of 224 p pulses. In 2D 14N/1H and 3D 14 N/14N/1H experiments 14N pulse length, which is not necessarily equal to a 90° pulse length, and both excitation and reconversion periods were optimized carefully to maximize the 14N–1H magnetization transfer efficiency. The 14N pulse length was set at 7.0 ls while the excitation and reconversion durations were set at sexc ¼ srec ¼ 2 ms. For the 3D measurement, 8 increments were set in both t1 and t2 dimensions and 240 scans were collected for each t1 and t2 increments in the presence of RFDR mixing, while 24 scans every 16 t1 and 16 t2 increments were set for the data collected in the absence of RFDR mixing. States-TPPI method was applied in both t1 and t2 dimensions to achieve pure absorption peaks. A recycle delay of 8 s was used in all the above experiments. The 3D data with and without RFDR, were acquired in 5.7 and 2.3 days, respectively. 14N isotropic shifts were referenced relative to CH3NO2. The drive pressure was actively controlled by pneumatic regulators to stabilize the sample spinning, and the fluctuation in the sample spinning monitored from the MAS controller unit was ±20 Hz. Delta NMR software (JEOL RESONANCE Inc.) was used to process the NMR data.
4. Results and discussion To achieve 14N/14N isotropic shift correlation in L-Histidine HCl– H2O from a 3D 14N/14N/1H correlation experiment under ultrafast MAS, first we found exact 14N peak positions from a 2D 14N/1H J-HMQC experiment to set the right values of 14N offset and spectral width required for the 3D measurement. This was followed by recording a 3D 14N/14N/1H isotropic shift correlation experiment. In the end, we evaluated the resulting signal intensity from the 3D experiment by measuring the filtering efficiencies of different blocks (HMQC, RFDR, HMQC-HMQC, HMQC-RFDR-HMQC) of the 3D pulse sequence (Fig. 1C). The magnetization transfer efficiency between the two cross correlating nitrogens through 1H–1H dipolar recoupling was maximized by carrying out a series of 2D 1H/1H chemical shift correlation experiments at 90 kHz MAS recorded with different mixing times. The pulse sequence implemented for this purpose is shown in Fig. 1A. The spin diffusion process is accelerated through 1H–1H dipolar recoupling during the mixing time. By measuring cross-peak intensities of protons (refer to Fig. S1 in the Supporting Information) directly bonded to nitrogens in the imidazole ring of L-Histidine HCl–H2O, the optimal magnetization transfer was achieved at RFDR mixing time of 2.5 ms and beyond which the magnetization exchange between these protons remained almost unchanged. It is worthwhile to mention here that the optimization procedure for RFDR mixing time described above can only be adopted if 1H signals associated with two cross correlating nitrogens are well separated. Similar procedure cannot be implemented in the case of overlapped 1H resonances of interest as change in the magnetization efficiency between protons during RFDR mixing cannot be evaluated. Before we could set a 3D 14N/14N/1H isotropic shift correlation experiment so as to have maximum resolution and sensitivity of 14 N/14N cross-peaks, it was important to carefully adjust the 14N offset and the spectral width of the indirect frequency dimensions. The main objective of carrying out these exercises was to get optimal spectral width of the indirect frequency dimensions to be set
with a minimum number of t1 and t2 points in order to sufficiently resolve 14N/14N cross peaks in a 3D 14N/14N/1H isotropic shift correlation experiment. Subsequently, we carried out a 2D 14 N/1H isotropic shift correlation experiment at 90 kHz MAS by implementing proton detection-based J-HMQC pulse sequence shown in Fig. 1B to get the exact values of 14N isotropic shifts and hence the right 14N offset. This was followed by setting up a series of 2D 14N/1H isotropic shift correlation experiments by varying spectral width of the indirect frequency dimension and t1 points while keeping the 14N offset exactly at the center of the two imidazole 14N resonances (114.4 ppm) of L-Histidine HCl– H2O. From these experiments a 14N spectral width of 5 kHz and 8 t1 points were found to be sufficient enough to get well-resolved 14 N/1H correlations (data not shown). Resonance assigned 1D 1H spectrum and 2D 14N/1H HMQC spectrum of L-Histidine HCl–H2O recorded at 90 kHz MAS are shown in Fig. 2. It is to be noted that a short transverse relaxation time (T2) of NH+3 and its longer 14 N–14N distances from nitrogens in the imidazole ring restricted us to optimize experimental conditions so as to achieve 14N/14N isotropic shift correlation only between nitrogens of the imidazole ring. In addition, the 14N isotropic shifts are quite different from 15 N chemical shifts because 14N isotropic shift is not only described by its isotropic chemical shift (riso) but a combination of riso and isotropic second-order quadrupolar shift. Once all the parameters were precisely optimized a 3D 14 N/14N/1H isotropic shift correlation experiment by implementing the J-based HMQC pulse sequence mediated through RFDR (see Fig. 1C) was performed to accomplish 14N/14N cross-correlation peaks. Proton-detected 2D 14N/14N isotropic shift correlation spectra, which are projections of 3D 14N/14N/1H spectra on to 14N/14N planes in the absence and presence of RFDR mixing, are shown in Fig. 3A and B, respectively. Additionally, 2D 14N/14N planes extracted at isotropic 1H chemical shifts of 12.6 and 17.1 ppm from a 3D 14N/14N/1H spectrum recorded with RFDR mixing are shown in Fig. 3C and D, respectively. As per our expectation we do not observe any 14N/14N cross-correlation in the absence of any 1 H–1H RFDR mixing. While in the presence of 1H–1H RFDR mixing time of 2.5 ms, which provides a medium for magnetization transfer between 14N nuclei to be accomplished through 1H–1H dipolar recoupling, we clearly observe 14N/14N cross-peaks. Also seen from the 2D planes are the less sensitive peaks (circled resonances in Fig. 3C and D) around but not exactly at the 14N/14N cross-peak positions. This observation seems rather intriguing in view of the fact that 2D slices at isotropic 1H chemical shifts should result in 14 N/14N cross-peaks only along the vertical axes. The appearance of additional resonances along the horizontal axis should be attributed to the tailing of the cross-peaks resulting from the presence of the second-order quadrupolar broadening. It is important to mention here that in the absence of any RFDR mixing there could also have been a possibility of magnetization exchange between 1H nuclei but only through the application of a very long mixing time (several hundred milliseconds) required to enhance the spin diffusion process. Finally, we measured filtering efficiencies by recording proton-detected 1D spectra using RFDR, HMQC, HMQC-HMQC, and HMQC-RFDR-HMQC blocks implemented in the 3D pulse sequence and compared the signal intensity with respect to that observed from a spin-echo experiment (refer to Fig. S2 in the Supporting Information for 1D spectra). Table 1 lists the percentage filtering efficiencies of proton resonances (17.1 and 12.6 ppm) which are directly bonded to nitrogens from all the different blocks described above in comparison to a spin-echo experiment. As seen from Table 1 the loss in the magnetization efficiency resulting from RFDR with 2.5 ms mixing time is in the range of 6–10% which clearly suggests that the total 1H magnetization is well retained during the RFDR mixing. This was followed by
M.K. Pandey, Y. Nishiyama / Journal of Magnetic Resonance 258 (2015) 96–101
Fig. 2. Molecular structure and assigned 1D 1H NMR spectrum (A), and 2D 600 MHz NMR spectrometer.
99
14
N/1H HMQC spectrum (B) of L-Histidine HCl–H2O at a MAS rate of 90 kHz collected from a
Fig. 3. Representative regions of proton-detected 2D 14N/14N chemical shift correlation spectra of L-Histidine HCl–H2O in the absence (A) and presence (B) of 1H/1H RFDR mixing projected from 3D 14N/14N/1H isotropic shift correlation spectra on 14N/14N planes recorded at 90 kHz ultrafast MAS using the pulse sequence shown in Fig. 1C. 2D 14 N/14N planes extracted from a 3D 14N/14N/1H isotropic shift correlation spectrum with 1H/1H RFDR mixing at isotropic 1H chemical shifts of 12.6 ppm (C) and 17.1 ppm (D). 16 t1 and t2 points every 24 scans, while 8 t1 and t2 points every 240 scans were used for 3D 14N/14N/1H isotropic shift correlation data collected without RFDR mixing and with RFDR mixing (2.5 ms), respectively, with 5 kHz spectral width and 8 s recycle delay.
Table 1 Observed magnetization transfer (filtering) efficiencies from 1H 1D measurements utilizing RFDR, HMQC, HMQC-HMQC, and HMQC-RFDR-HMQC pulse sequences as compared to the sensitivity measured from a 1H spin-echo experiment. 1
H peak (ppm)
RFDR (%)
HMQC (%)
HMQC-HMQC (%)
HMQC-RFDR-HMQC (%)
17.1 (NH1) 12.6 (NH2)
93.79 90.00
7.55 9.75
0.73 0.99
0.22 0.26
measuring the filtering efficiency resulting from a proton-detected 1D 14N/1H HMQC experiment. In general, the magnetization transfer efficiency after one HMQC filtering should be in the range
of 1–10% of the intensity observed from a spin-echo experiment. From our experiment this value is in the range of 7.5–10% for both 1 H resonances (Table 1), which is almost close to the expected value. Likewise, if we assume similar transfer efficiency also from the second HMQC sequence then the filtering efficiency after the application of two HMQC blocks without 1H–1H RFDR mixing (tmix = 0) should be equal to 1% (10% of the 10%) of the spin-echo intensity, which is also in a close agreement with our experimental observation (Table 1). It is worth to point here that there can be a possibility of better filtering efficiency from the second HMQC as the first HMQC selects only those nuclei that have a favorable orientation for the HMQC filtering. Nonetheless, this phenomenon is
100
M.K. Pandey, Y. Nishiyama / Journal of Magnetic Resonance 258 (2015) 96–101
not observed from our measurements as we did not notice any gain in the magnetization from the second HMQC. Although the total 1H magnetization is well preserved during the RFDR as mentioned earlier, there is a substantial loss in the magnetization of the imidazole ring NH protons due to exchange with the remaining protons that are not directly bonded to nitrogens. This phenomenon significantly reduces the 1H magnetization of the NH protons after the HMQC-RFDR-HMQC filtering which is also observed from our 1D HMQC-RFDR-HMQC measurement with 2.5 ms mixing for both the resonances (Table 1). The loss in the magnetization of protons bonded to nitrogens is ca 70% of the total signal intensity observed from 1D HMQC-HMQC experiment in the absence of RFDR mixing. The above observation is in accordance with the measured signal intensities (both auto and cross-correlation peaks) of the NH protons from the 2D 1H/1H RFDR correlation spectra collected with (2.5 ms) and without mixing (refer to Fig. S1 in the Supporting Information). More importantly, the magnetization efficiency left after HMQC-RFDR-HMQC block is sufficient enough to measure a 3D 14N/14N/1H correlation experiment on a natural abundant sample. In fact, a 3D 15N/15N/1H correlation experiment with 1H–1H RFDR mixing demonstrated in our previous work on a fully 15N labeled N-acetyl-L-valyl-L-leucine (NAVL) sample, the final magnetization that survived was 1.6% of the original signal intensity which is higher as compared to the magnetization that survives for detection in the present study. The main reason for the observed difference is the survival of 31% and 10% of the total 1H magnetization after the first and second cross polarization-based 1 H–15N–1H magnetization transfer, respectively, unlike the present case wherein ca 10% and 1% of the total 1H magnetization (Table 1) survive after first and second HMQC filtering, respectively. We would like to mention here that the required time for the collection of a 3D 14N/14N/1H chemical shift correlation data to achieve good sensitivity of the 14N/14N cross-peaks is relatively longer in the present study. Apart from the inherent difficulties in the measurement due to the presence of a weak 14N–14N homonuclear dipolar interaction and the first-order quadrupolar couplings, the use of a tiny amount of sample at natural abundance can lead to loss in the sensitivity of the cross-peaks. Additionally, slight fluctuations in the sample spinning and rotor phase could also add to the loss in the sensitivity as fundamental transition frequency of 14N is extremely sensitive to exact magic angle setting and stable spinning. Better sensitivity in the 14N dimension could be achieved by implementing D-HMQC experiment wherein magnetization transfer occurs via recoupled 14N–1H heteronuclear dipolar interactions instead of J-coupling and RDS [33]. As a cautionary note we would like to mention here that D-HMQC experiment is more sensitive toward the fluctuations in the sample spinning, and in such cases J-HMQC with 1H–1H decoupling can be implemented as an alternative wherein sensitivity is enhanced through increased T2 relaxation time [42]. Sample spinning faster than 90 kHz with the application of 1H–1H decoupling can also result in the gain in the sensitivity. Furthermore, selective 1H–1H recoupling methods instead of RFDR can also be implemented for the sensitivity enhancement as loss in the magnetization due to its distribution to protons that are not directly bonded to 14N can be avoided. It is worth pointing that a limited resolution both in the 1H and 14N dimensions can hamper the application of the proposed method to large molecules like proteins due to possible signal overlaps. However, we believe that an addition of high resolution dimension like 13C can overcome this difficulty. 5. Conclusions In summary, we have presented a new proton detection-based 3D pulse sequence to get 14N/14N/1H isotropic shift correlation in natural abundant powdered L-Histidine HCl–H2O (0.3 mg) under
ultrafast MAS (90 kHz) condition by utilizing the magnetization transfer between 14N and 1H through J-coupling and RDS. We have applied 1H–1H RFDR sequence during mixing time to enhance the spin diffusion process resulting from 1H–1H dipolar recoupling so as to accomplish 14N/14N correlation. The new approach for observing 14N/14N homonuclear isotropic shift correlation demonstrates a strong ability of proton detection-based measurements with the application of RFDR-based 1H–1H dipolar recoupling under ultrafast MAS condition that can be used for structural studies of challenging chemical and biological samples at natural abundance. Although the method presented in this study to accomplish 14 N/14N correlation has a limitation with respect to the measurement time, we believe that better sensitivity in 14N dimensions can be achieved by implementing already existing and/or new state-of-the-art solid-state NMR techniques. Acknowledgment This work was supported by the NMR Platform from the Ministry of Education, Culture, Sports, Science and Technology, Japan. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jmr.2015.06.012. References [1] S. Parthasarathy, Y. Nishiyama, Y. Ishii, Sensitivity and resolution enhanced solid-state NMR for paramagnetic systems and biomolecules under very fast magic angle spinning, Acc. Chem. Res. 46 (2013) 2127–2135. [2] J.P. Demers, V. Chevelkov, A. Lange, Progress in correlation spectroscopy at ultra-fast magic-angle spinning: basic building blocks and complex experiments for the study of protein structure and dynamics, Solid State Nucl. Magn. Reson. 40 (2011) 101–113. [3] P. Paluch, T. Pawlak, J.P. Amoureux, M.J. Potrzebowski, Simple and accurate determination of X–H distances under ultra-fast MAS NMR, J. Magn. Reson. 233 (2013) 56–63. [4] S. Wang, R.A. Munro, L. Shi, I. Kawamura, T. Okitsu, A. Wada, S.Y. Kim, K.H. Jung, L.S. Brown, V. Ladizhansky, Solid-state NMR spectroscopy structure determination of a lipid-embedded heptahelical membrane protein, Nat. Methods 10 (2013) 1007–1012. [5] M.E. Ward, L. Shi, E. Lake, S. Krishnamurthy, H. Hutchins, L.S. Brown, V. Ladizhansky, Proton-detected solid-state NMR reveals intramembrane polar networks in a seven-helical transmembrane protein proteorhodopsin, J. Am. Chem. Soc. 133 (2011) 17434–17443. [6] I. Bertini, C. Luchinat, G. Parigi, E. Ravera, B. Reif, P. Turano, Solid-state NMR of proteins sedimented by ultracentrifugation, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 10396–10399. [7] Y.Q. Ye, M. Malon, C. Martineau, F. Taulelle, Y. Nishiyama, Rapid measurement of multidimensional 1H solid-state NMR spectra at ultra-fast MAS frequencies, J. Magn. Reson. 239 (2014) 75–80. [8] Y. Nishiyama, M. Malon, Y. Ishii, A. Ramamoorthy, 3D (1)(5)N/(1)(5)N/(1)H chemical shift correlation experiment utilizing an RFDR-based 1H/1H mixing period at 100 KHz MAS, J. Magn. Reson. 244 (2014) 1–5. [9] V. Agarwal, S. Penzel, K. Szekely, R. Cadalbert, E. Testori, A. Oss, J. Past, A. Samoson, M. Ernst, A. Bockmann, B.H. Meier, De novo 3D structure determination from sub-milligram protein samples by solid-state 100 KHz MAS NMR spectroscopy, Angew. Chem. 53 (2014) 12253–12256. [10] J.M. Lamley, D. Iuga, C. Oster, H.J. Sass, M. Rogowski, A. Oss, J. Past, A. Reinhold, S. Grzesiek, A. Samoson, J.R. Lewandowski, Solid-state NMR of a protein in a precipitated complex with a fulllength antibody, J. Am. Chem. Soc. 136 (2014) 16800–16806. [11] T. Kobayashi, K. Mao, P. Paluch, A. Nowak-Krol, J. Sniechowska, Y. Nishiyama, D.T. Gryko, M.J. Potrzebowski, M. Pruski, Study of intermolecular interactions in the corrole matrix by solid-state NMR under 100 KHz MAS and theoretical calculations, Angew. Chem. 52 (2013) 14108–14111. [12] Y. Nishiyama, T. Kobayashi, M. Malon, D. Singappuli-Arachchige, I.I. Slowing, M. Pruski, Studies of minute quantities of natural abundance molecules using 2D heteronuclear correlation spectroscopy under 100 KHz MAS, Solid State Nucl. Magn. Reson. 66–67 (2015) 56–61. [13] A. Bockmann, M. Ernst, B.H. Meier, Spinning proteins, the faster, the better?, J Magn. Reson. 253 (2015) 71–79. [14] V. Agarwal, T. Tuherm, A. Reinhold, J. Past, A. Samoson, M. Ernst, B.H. Meier, Amplitude-modulated low-power decoupling sequences for fast magic-angle spinning NMR, Chem. Phys. Lett. 583 (2013) 1–7.
M.K. Pandey, Y. Nishiyama / Journal of Magnetic Resonance 258 (2015) 96–101 [15] K. Mroue, Y. Nishiyama, M.K. Pandey, B. Gong, E. McNerny, D. Kohn, M. Morris, A. Ramamoorthy, Proton-detected solid-state NMR spectroscopy of bone with ultrafast magic angle spinning, Sci. Rep. 5 (2015) 11991. [16] B. Reif, R.G. Griffin, 1H detected 1H,15N correlation spectroscopy in rotating solids, J. Magn. Reson. 160 (2003) 78–83. [17] D.H. Zhou, G. Shah, M. Cormos, C. Mullen, D. Sandoz, C.M. Rienstra, Protondetected solid-state NMR spectroscopy of fully protonated proteins at 40 KHz magic-angle spinning, J. Am. Chem. Soc. 129 (2007) 11791–11801. [18] Y. Ishii, R. Tycko, Sensitivity enhancement in solid state 15N NMR by indirect detection with high-speed magic angle spinning, J. Magn. Reson. 142 (2000) 199–204. [19] K. Mao, J.W. Wiench, V.S.Y. Lin, M. Pruski, Indirectly detected through-bond chemical shift correlation NMR spectroscopy in solids under fast MAS: studies of organic-inorganic hybrid materials, J. Magn. Reson. 196 (2009) 92–95. [20] S.M. Althaus, K.M. Mao, J.A. Stringer, T. Kobayashi, M. Pruski, Indirectly detected heteronuclear correlation solid-state NMR spectroscopy of naturally abundant 15N nuclei, Solid State Nucl. Magn. Reson. 57–58 (2014) 17–21. [21] K.M. Mao, M. Pruski, Directly and indirectly detected through-bond heteronuclear correlation solid-state NMR spectroscopy under fast MAS, J. Magn. Reson. 201 (2009) 165–174. [22] V. Agarwal, A. Diehl, N. Skrynnikov, B. Reif, High resolution 1H detected 1H,13C correlation spectra in MAS solid-state NMR using deuterated proteins with selective 1H,2H isotopic labeling of methyl groups, J. Am. Chem. Soc. 128 (2006) 12620–12621. [23] S. Asami, M. Rakwalska-Bange, T. Carlomagno, B. Reif, Protein-RNA interfaces probed by 1H-detected MAS solid-state NMR spectroscopy, Angew. Chem. 52 (2013) 2345–2349. [24] S. Asami, B. Reif, Proton-detected solid-state NMR spectroscopy at aliphatic sites: application to crystalline systems, Acc. Chem. Res. 46 (2013) 2089–2097. [25] R. Linser, V. Chevelkov, A. Diehl, B. Reif, Sensitivity enhancement using paramagnetic relaxation in MAS solid-state NMR of perdeuterated proteins, J. Magn. Reson. 189 (2007) 209–216. [26] R. Linser, M. Dasari, M. Hiller, V. Higman, U. Fink, J.M. Lopez del Amo, S. Markovic, L. Handel, B. Kessler, P. Schmieder, D. Oesterhelt, H. Oschkinat, B. Reif, Proton-detected solid-state NMR spectroscopy of fibrillar and membrane proteins, Angew. Chem. 50 (2011) 4508–4512. [27] Y. Nishiyama, M. Malon, M.J. Potrzebowski, P. Paluch, J.P. Amoureux, Accurate NMR determination of C–H or N–H distances for unlabeled molecules, Solid State Nucl. Magn. Reson. in press. [28] E. Konstantin, B.M. Fung, High-resolution 14N NMR in polycrystalline solids, J. Chem. Phys. 110 (1999) 7977–7982.
101
[29] S. Cavadini, Indirect detection of nitrogen-14 in solid-state NMR spectroscopy, Prog. Nucl. Magn. Reson. Spectrosc. 56 (2010) 46–77. [30] S. Cavadini, S. Antonijevic, A. Lupulescu, G. Bodenhausen, Indirect detection of nitrogen-14 in solids via protons by nuclear magnetic resonance spectroscopy, J. Magn. Reson. 182 (2006) 168–172. [31] S. Cavadini, A. Lupulescu, S. Antonijevic, G. Bodenhausen, Nitrogen-14 NMR spectroscopy using residual dipolar splittings in solids, J. Am. Chem. Soc. 128 (2006) 7706–7707. [32] Z.H. Gan, Measuring amide nitrogen quadrupolar coupling by high-resolution 14 N/13C NMR correlation under magic-angle spinning, J. Am. Chem. Soc. 128 (2006) 6040–6041. [33] Y. Nishiyama, Y. Endo, T. Nemoto, H. Utsumi, K. Yamauchi, K. Hioka, T. Asakura, Very fast magic angle spinning 1H–14N 2D solid-state NMR: sub-micro-liter sample data collection in a few minutes, J. Magn. Reson. 208 (2011) 44–48. [34] J. Xu, J. Struppe, A. Ramamoorthy, Two-dimensional homonuclear chemical shift correlation established by the cross-relaxation driven spin diffusion in solids, J. Chem. Phys. 128 (2008). [35] A. Ramamoorthy, L.M. Gierasch, S.J. Opella, 4-dimensional solid-state NMR experiment that correlates the chemical-shift and dipolar-coupling frequencies of 2 heteronuclei with the exchange of dilute-spin magnetization, J. Magn. Reson. B 109 (1995) 112–116. [36] Y.F. Wei, A. Ramamoorthy, 2D 15N–15N isotropic chemical shift correlation established by 1H–1H dipolar coherence transfer in biological solids, Chem. Phys. Lett. 342 (2001) 312–316. [37] J.R. Lewandowski, G. De Paepe, M.T. Eddy, R.G. Griffin, N-15-N-15 proton assisted recoupling in magic angle spinning NMR, J. Am. Chem. Soc. 131 (2009) 5769–5776. [38] M. Giffard, S. Hediger, J.R. Lewandowski, M. Bardet, J.P. Simorre, R.G. Griffin, G. De Paepe, Compensated second-order recoupling: application to third spin assisted recoupling, Phys. Chem. Chem. Phys. 14 (2012) 7246–7255. [39] A.K. Khitrin, J.D. Xu, A. Ramamoorthy, Cross-correlations between low-gamma nuclei in solids via a common dipolar bath, J. Magn. Reson. 212 (2011) 95–101. [40] Y. Nishiyama, R. Zhang, A. Ramamoorthy, Finite-pulse radio frequency driven recoupling with phase cycling for 2D 1H/1H correlation at ultrafast MAS frequencies, J. Magn. Reson. 243 (2014) 25–32. [41] R. Zhang, Y. Nishiyama, P. Sun, A. Ramamoorthy, Phase cycling schemes for finite-pulse-RFDR MAS solid state NMR, J. Magn. Reson. (2015), http:// dx.doi.org/10.1016/j.jmr.2014.12.010. [42] Y. Nishiyama, X. Lu, J. Trebosc, O. Lafon, Z. Gan, P.K. Madhu, J.P. Amoureux, Practical choice of 1H–1H decoupling schemes in through-bond 1H–{X} HMQC experiments at ultra-fast MAS, J. Magn. Reson. 214 (2012) 151–158.