Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Nano-Mole Scale Side-Chain Signal Assignment by 1H-Detected Protein Solid-State NMR by Ultra-Fast Magic-Angle Spinning and Stereo-Array Isotope Labeling

  • Songlin Wang ,

    Contributed equally to this work with: Songlin Wang, Sudhakar Parthasarathy

    Affiliation Department of Chemistry and University of Illinois at Chicago, Chicago, Illinois, United States of America

  • Sudhakar Parthasarathy ,

    Contributed equally to this work with: Songlin Wang, Sudhakar Parthasarathy

    Affiliation Department of Chemistry and University of Illinois at Chicago, Chicago, Illinois, United States of America

  • Yusuke Nishiyama,

    Affiliations JEOL RESONANCE Inc., Akishima, Tokyo, Japan, RIKEN CLST-JEOL collaboration center, RIKEN, Yokohama, Kanagawa, Japan

  • Yuki Endo,

    Affiliation JEOL RESONANCE Inc., Akishima, Tokyo, Japan

  • Takahiro Nemoto,

    Affiliation JEOL RESONANCE Inc., Akishima, Tokyo, Japan

  • Kazuo Yamauchi,

    Affiliations School of Science and Technology, Nazarbayev University, Astana, Kazakhstan, Nuclear Magnetic Resonance Core Lab., King Abdullah University of Science and Technology, Thuwal, Saudi Arabia

  • Tetsuo Asakura,

    Affiliation Department of Biotechnology, Tokyo University of Agriculture and Technology, Koganei, Tokyo, Japan

  • Mitsuhiro Takeda,

    Affiliation Structural Biology Research Center, Graduate School of Science, Furocho, Chikusa-ku, Nagoya University, Nagoya, Japan 464–8601

  • Tsutomu Terauchi,

    Affiliation SAIL Technologies Co., Inc., Tsurumi-ku, Yokohama, Kanagawa, Japan

  • Masatsune Kainosho,

    Affiliations Structural Biology Research Center, Graduate School of Science, Furocho, Chikusa-ku, Nagoya University, Nagoya, Japan 464–8601, Center for Priority Areas, Tokyo Metropolitan University, Tokyo, Japan

  • Yoshitaka Ishii

    yishii@uic.edu

    Affiliations Department of Chemistry and University of Illinois at Chicago, Chicago, Illinois, United States of America, Center for Structural Biology, University of Illinois at Chicago, Chicago, Illinois, United States of America

Nano-Mole Scale Side-Chain Signal Assignment by 1H-Detected Protein Solid-State NMR by Ultra-Fast Magic-Angle Spinning and Stereo-Array Isotope Labeling

  • Songlin Wang, 
  • Sudhakar Parthasarathy, 
  • Yusuke Nishiyama, 
  • Yuki Endo, 
  • Takahiro Nemoto, 
  • Kazuo Yamauchi, 
  • Tetsuo Asakura, 
  • Mitsuhiro Takeda, 
  • Tsutomu Terauchi, 
  • Masatsune Kainosho
PLOS
x

Abstract

We present a general approach in 1H-detected 13C solid-state NMR (SSNMR) for side-chain signal assignments of 10-50 nmol quantities of proteins using a combination of a high magnetic field, ultra-fast magic-angle spinning (MAS) at ~80 kHz, and stereo-array-isotope-labeled (SAIL) proteins [Kainosho M. et al., Nature 440, 52–57, 2006]. First, we demonstrate that 1H indirect detection improves the sensitivity and resolution of 13C SSNMR of SAIL proteins for side-chain assignments in the ultra-fast MAS condition. 1H-detected SSNMR was performed for micro-crystalline ubiquitin (~55 nmol or ~0.5mg) that was SAIL-labeled at seven isoleucine (Ile) residues. Sensitivity was dramatically improved by 1H-detected 2D 1H/13C SSNMR by factors of 5.4-9.7 and 2.1-5.0, respectively, over 13C-detected 2D 1H/13C SSNMR and 1D 13C CPMAS, demonstrating that 2D 1H-detected SSNMR offers not only additional resolution but also sensitivity advantage over 1D 13C detection for the first time. High 1H resolution for the SAIL-labeled side-chain residues offered reasonable resolution even in the 2D data. A 1H-detected 3D 13C/13C/1H experiment on SAIL-ubiquitin provided nearly complete 1H and 13C assignments for seven Ile residues only within ~2.5 h. The results demonstrate the feasibility of side-chain signal assignment in this approach for as little as 10 nmol of a protein sample within ~3 days. The approach is likely applicable to a variety of proteins of biological interest without any requirements of highly efficient protein expression systems.

Introduction

1H indirect detection was introduced to 13C and 15N biomolecular high-resolution solid-state NMR (SSNMR) about a decade ago. [14] Despite its potential as a powerful tool to enhance sensitivity and resolution, 1H-detected SSNMR is not widely used due to limits on 1H resolution, even under fast MAS, and the lack of a demonstrated sensitivity advantage over more commonly used 13C detection. The recent introduction of 1H dilution by high-level deuteration and partial back-exchange of amide 1H (10–20%) has greatly improved the resolution of 1H SSNMR for biomolecules,[5, 6] offering a practical protocol for 1H indirect detection in protein SSNMR. However, the method is limited by a gross loss of 1H signals from amide sites (80–90%) due to extensive deuteration. 1H-detected 13C SSNMR for a fully protonated protein at very fast MAS (~40 kHz) has been used to obtain signal assignments for a model protein. [7] Nevertheless, this method is still hampered by relatively broad 1H line widths (0.5–1 ppm), a resolution that is insufficient even for small proteins. More importantly, it has been difficult to improve the sensitivity of 2D 1H indirect detection over that of standard 1D 13C direct-detected SSNMR. Recent studies described 1H indirect detection under ultra-fast MAS (UFMAS) at spinning frequencies of 60 kHz, resulting in resolved amide 1H resonances for fully deuterated proteins with fully back-exchanged amide proton[8, 9] or for undeuterated proteins. [10] Although those studies demonstrated the feasibility of main-chain sequential assignments for micro-crystalline samples, no strategy for assigning side-chain resonances by 1H-detected SSNMR has yet been developed, despite the fundamental importance of side chain structures and dynamics for protein functions. Equally importantly, no previous studies established advantage of 1H indirect detection method over traditional 13C direct detection for concurrent improvement in sensitivity and resolution by a quantitative analysis. Although some previous studies demonstrated sensitivity advantage of 2D 1H-detected SSNMR over 2D 13C-detected SSNMR,[7, 11, 12] it was difficult to achieve sensitivity advantage by 1H-detected (N+1)-dimensional SSNMR over a corresponding N-dimensional 13C-detected SSNMR scheme with an additional 1H dimension for higher resolution (N = 1, 2..). To overcome these problems, in this study, we propose the use of stereo-array isotope labeling (SAIL) as a highly effective labeling scheme suitable for side-chain signal assignments by 1H-detected protein SSNMR. The SAIL scheme was originally introduced to overcome the size limitation of biomolecular solution NMR by incorporating stereo-selective deuteration to achieve isolated 1H throughout all side chains of a protein.[13] Although 1H SSNMR was attempted for a SAIL amino acid (L-valine) under fast MAS at ~30 kHz,[14] resultant 1H line widths were still in a range of 0.5–0.7 ppm; the limited 1H resolution has hampered successful use of SAIL labeling for protein SSNMR. In addition, it is not trivial to determine whether sufficient sensitivity can be achieved for a SAIL-labeled protein sample under UFMAS conditions for limited sample quantity in a smaller MAS rotor and potentially much longer 1H T1 values due to deuteration. In this work, we demonstrate that a combination of UFMAS and SAIL selective deuteration significantly improves the sensitivity of 1H-detected 2D 13C SSNMR of biomolecules over 1D and 2D 13C direct detection. It is also discussed that this combination offers extremely sensitive means of biomolecular SSNMR for side-chain assignments with resolution enhanced 1H signals having line widths of 0.1–0.2 ppm.

Results and Discussion

First, in experiments on amino-acid samples, we investigated whether the combined use of SAIL labeling with UFMAS in a high magnetic field of 17.62 T (1H frequency of 750.15 MHz) could improve the resolution of 1H SSNMR resolution. Fig 1(a, b) shows chemical structures and labeling schemes for (a) uniformly 13C- and 15N-labeled isoleucine (UL-Ile) and (b) SAIL-isoleucine (SAIL-Ile). Unlike random deuteration, in a SAIL scheme, all the protonated 13C groups are connected to a single 1H species. This feature allows preparation of strong 13C polarization for all 13C species via efficient double-quantum cross-polarization from directly bonded 1H nuclei.[15] More importantly, isolated 1H spins allow us to achieve very high resolution without the effects of strong 1H–1H dipolar couplings. Fig 1(c, d) shows the spinning-speed dependence of 1H MAS SSNMR of (c) UL-Ile and (d) SAIL-Ile, with signal assignments provided in (d). Significant improvement in resolution and sensitivity was obtained at higher spinning speeds (νR). Resolution enhancements of 2–3-fold were observed at νR = 80 kHz in (d) relative to νR = 30 kHz, which was used in previous studies of 1H SSNMR on SAIL-valine.[14] Clearly, SAIL-Ile provided much higher resolution in Fig 1dR = 80 kHz) relative to the corresponding spectrum for UL-Ile in Fig 1c. In particular, for Hβ, Hγ, and Hδ groups, the spectrum for SAIL-Ile exhibits a dramatic improvement in resolution (by a factor of 3–4) relative to the resolution for UL-Ile, with 1H widths of 0.21–0.25 ppm at 80 kHz. This narrowing can be attributed to the isolation of 1H in methylene and methyl groups by stereo-specific deuteration in the SAIL scheme.[13, 16] Much broader 1H line widths (0.7–1.0 ppm) were observed for these groups in UL-Ile (Fig 1c). These observations confirm that UFMAS itself is still not sufficient to remove broadening due to strong 1H–1H dipolar couplings within the CH2 and CH3 groups, even at νR of ~80 kHz. The combination of SAIL and UFMAS also exhibited modest narrowing (15–20%) in the line widths of O–H (0.22 ppm) and 1Hα (0.26 ppm) (Fig 1d). We also confirmed the excellent 1H resolution at νR of ~80 kHz for SAIL Thr (Fig A in S1 File).

thumbnail
Fig 1. Spinning-speed dependence of 1H MAS spectra of fully protonated and SAIL isoleucine samples.

(a, b) Chemical structures of (a) uniformly 13C- and 15N-labeled (UL) Ile and (b) SAIL-Ile. (c, d) Spinning-speed dependence of 1H MAS SSNMR spectra of (c) UL-Ile and (d) SAIL-Ile. The peak at 4.8 ppm (*) is likely due to HCl salts.[17, 18] No window functions were applied.

https://doi.org/10.1371/journal.pone.0122714.g001

The sensitivity enhancement factor (ξ) of 1H indirect detection over direct detection of a dilute X nuclei depends on the line width in the 1H dimension WH, the apodization, and the efficiency of polarization transfer (f) from X to 1H used for 1H detection, as shown in Eq (1),[1, 2] (1) where γH and γX represent the gyromagnetic ratios of the nuclei H and X, WH and WX are the line widths observed for 1H and X nuclei, and QH and QX are the quality factors for the sample coil for 1H and X detection, respectively. The factor α is 1 for a comparison of 2D 1H-detected X/1H correlation SSNMR with 2D X-detected X/1H correlation. For a comparison of 2D 1H-detected X/1H correlation with 1D 13C CPMAS, the α value becomes 2π assuming apodization with matched window functions (see SI about the details).[1, 2] Thus, there is a ~2.5-fold difference (i.e.,) between the ξ values of the 1D and 2D direct X-detection experiments. In biomolecular SSNMR, it is typical that no or minimal line broadening is applied for higher spectral resolution. When no window functions are applied in the 1H dimension, α ~ 2π2 for the comparison with 1D 13C SSNMR; thus, this suggests a 4.4-fold difference (i.e.,) between the ξ values for 1D and 2D (see SI about the details). Compared with the 1H resolution for UL-Ile at νR = 40 kHz, which was previously used in 1H-detected protein SSNMR, [7] the 1H resolution for CHD and CHD2 groups improved as much as 5–6-fold for the SAIL sample at νR = 80 kHz. We confirmed that the transfer efficiency f by 1H–13C double-quantum CP at νR = 80 kHz was comparable to the CP efficiency at νR = 20–40 kHz for SAIL-Ile. Thus, this new combination of SAIL and UFMAS offers opportunities to dramatically improve the sensitivity and resolution of 1H detection.

Next, we explored the possibility of improving the sensitivity and resolution of 13C SSNMR for SAIL proteins. As a suitable benchmark, we selected a micro-crystalline sample of ubiquitin (Ubq) that was selectively labeled with SAIL-Ile to compare the resolution with that of the amino acid data. Because there are as many as seven Ile residues in ubiquitin, the system is also suited for investigating improvements in the resolution of 1H detection. Another major challenge is the limited amount of sample used in these experiments. Because the engineering needs for UFMAS at 80 kHz limit the sample volume to only ~1 μL, it was difficult to achieve sufficient sensitivity in multi-dimensional 13C protein SSNMR, even in a high magnetic field. Fig 2a and 2b shows (a) 1H-detected and (b) 13C-detected 2D 13C/1H correlation spectra of the SAIL-Ubq sample. In the 1H-detected 2D spectrum, the sensitivity was dramatically improved (by a factor of 5.4–9.7) (Fig 2a) relative to the 13C-detected 2D data (b), as shown from the comparison of the slices corresponding to the peaks indicated by arrows (d–g). The factors were confirmed with the data collected from additional scans for (b) (see Table A in S1 File). As a result of this significant improvement, the 2D spectrum in Fig 2a was obtained after only 5 min, using sub-milligram quantities of protein sample (~0.5 mg or ~55 nmol, excluding H2O). By contrast, the corresponding 13C-detected 2D spectrum in Fig 2b had a much lower signal-to-noise ratio. Because of the excellent 1H resolution, most of the resonances are well separated in the 1H-detected 2D spectrum in Fig 2a in contrast to the significant signal overlap observed in the 1D 13C SSNMR in (c). It should be noted that the backbone 13Cα signals and some of the 13Cβ signals were weak because the protein was expressed in a D2O medium and, consequently, 1Hα and some 1Hβ were replaced by 2H; this issue can be overcome by expressing the protein in a cell-free system. It is also noteworthy that only 7.5 mg of SAIL-Ile was needed for preparing the sample at high labeling efficiency (~90%) from an E. coli cell culture of 0.5 L. The 1H line widths were 0.14–0.25 ppm and 0.10–0.22 ppm, respectively, with and without window functions. Clearly, our approach has opened an avenue for micro-SSNMR analysis of a protein sample in a minimal experimental time.

thumbnail
Fig 2. A comparison of 1H-detected and 13C-detected 2D and 1D SSNMR spectra of SAIL ubiquitin.

(a) 1H-detected and (b) 13C-detected 2D 13C/1H spectra and (c) a 1D 13C CP-MAS spectrum of SAIL-Ubq (~0.5 mg) at MAS 80 kHz. (d–g) 1D slices from 1H shifts of (d, f) 0.43 ppm and (e, g) 0.69 ppm from (d, f) 1H-detected and (e, g) 13C-detected experiments. All spectra were processed with 45°- and 60°-shifted sine-bell window functions in the 1H and 13C dimensions, respectively. The 1H and 13C line widths were 0.10–0.22 ppm and 0.66–0.94 ppm, respectively, in the absence of any window functions. The insets in (c, e, and g) show the magnified noise regions. Each spectrum in (a–c) was collected within ~5 min. The pulse sequences used for (a) and (b) are shown in Fig B and Fig C in S1 File.

https://doi.org/10.1371/journal.pone.0122714.g002

The sensitivity enhancement factors by 1H-detected data in (d, e) relative to 1D 13C CPMAS in (c) for resolved 13CHD signals at 24.5 ppm and 13CHD2 signals at 7.9 ppm (orange arrows) were 2.1 and 5.0, respectively. These factors are slightly greater than the theoretical values 1.5–2.7, which were obtained by multiplying by the experimental ξ values for the 13C-detected 2D experiment, where the factor came from the sensitivity difference between 2D and 1D experiments without window functions (see SI). The modest gain by a factor of (or ~1.2) was expected from “time saving” due to a linear prediction, (LP) which was employed to extend the indirect time-domain signals by 1.5 fold although an apparent S/N ratio with LP may be influenced by other issues such as additional “noise” for prediction artifacts. We also experimentally confirmed similar sensitivity improvement factors ξ over 1D CPMAS (ξ = 1.3–2.0) and 13C-detected 2D correlation (ξ = 3.9–10.1) for SAIL-Ile (see Fig E and Table B in S1 File). To the best of our knowledge, this is the first demonstration that 1H-detected 2D 1H/13C correlation SSNMR for a protein sample is significantly more sensitive than 1D 13C direct detection.

The results described above suggest that most standard 13C-detected 2D and 3D SSNMR involving side-chain signals can be replaced by 1H-detected 3D and 4D SSNMR, respectively, with significantly enhanced resolution and sensitivity. To test this possibility, we performed 1H-detected 3D 13C/13C/1H correlation SSNMR on the SAIL-Ubq sample. Fig 3 shows (a,b) a 2D 13C/13C projection of the 3D data and (c–e) strip plots of 13C/13C 2D slices corresponding to 1H chemical shifts of (c) 1.57 ppm, (d) 1.73 ppm, and (e) 1.41 ppm. The 3D spectrum in Fig 3 was obtained in 2.5 h, despite the deuteration of Cα and partial deuteration of Cβ. We did not attempt a 3D experiment by 13C detection, as it would have taken up to 10 days. All the 13C resonances for the seven Ile residues are observed in Fig 3a, including those for nearly fully deuterated 13Cα (dotted circle). These 13Cα resonances were detected in the t1 period by polarization transfer from remote 1H, correlation to 13Cβ in t2, and final detection at 1Hβ in the t3 period. In the 2D 13C/13C projection, the signal overlap could not be completely eliminated. For example, the three signals for Ile-3, Ile-13, and Ile-44 are nearly overlapping at (f1, f2) ~ (59 ppm, 42 ppm) in the projection shown in Fig 3b. However, in 2D slices at three different 1H shifts (Fig 3c3e), these overlapped peaks are clearly separated by well dispersed 1Hβ shifts. All the side-chain 1H and 13C resonances of Ile, except for 1Hγ1 of Ile-13, were assigned with excellent resolution (Table C in S1 File), as shown for an example of Ile-61 in Fig 3f. Although the 13Cα and 13Cβ signals were assigned here to specific residues based on previous 13C-detected SSNMR studies,[19, 20] it is possible to connect side-chain resonances to main-chain and 13Cβ resonances, which can be assigned from 1H-detected SSNMR for a uniformly 13C- and 15N-labeled protein as previously discussed. [10] It should be noted that the main-chain assignment strategy by 1H-detection [10] is likely to be applicable to uniformly SAIL-labeled proteins. In that case, spectral assignments for the main chain as well as side chains could be obtained by the 1H-detection method on the uniformly SAIL-labeled protein. Alternatively, site-directed mutagenesis can be used for assignments of some specific residues. Thus, the 1H-detected high-field SSNMR approach using SAIL-labeled protein and UFMAS is highly effective for side-chain assignments. The data suggest that side-chain assignments by 3D SSNMR analysis can be obtained from 10 nmol (or ~90 μg for Ubq) within only ~3 days in our approach.

thumbnail
Fig 3. Resolution and side-chain assignments from 3D 13C/13C/1H SSNMR of SAIL-Ubq.

(a, b) 2D 13C/13C 2D projection spectra from a 1H-detected 3D 13C/13C/1H SSNMR of SAIL-Ubq at MAS 80 kHz. All the peaks including minor ones in (a) are attributed to intra-residue cross peaks within the Ile residues. (c–e) Representative 2D 13C/13C slices corresponding to 1H chemical shifts of (c) 1.57 ppm, (d) 1.73 ppm, and (e) 1.41 ppm. The data show clear separation of signals for (c) Ile-3, (d) Ile-13, and (e) Ile-44 by 1H shifts. The spectrum was processed with 45°- and 60°-shifted sine-bell window functions in the 1H and 13C dimensions, respectively. (f) 13C/1H assignments for Ile-61 from the 3D data. The pulse sequence is listed in Fig D in S1 File.

https://doi.org/10.1371/journal.pone.0122714.g003

Conclusion

In this work, we discussed a general approach to achieve side-chain spectral assignments of SAIL-labeled proteins by 1H-detected protein SSNMR approaches. This work raises new prospects for high-field protein SSNMR in two areas. First, we presented the advantage of 1H-detected high-field SSNMR by demonstrating feasibility of side-chain assignments from 10–50 nmol of SAIL-labeled proteins under UFMAS at 80 kHz. To date no approach has been able to achieve efficient side-chain assignments through 1H-detected biomolecular SSNMR. Our data clearly demonstrate that, using this approach, heavily overlapped 13C side-chain signals can be resolved by 1H shifts for all seven Ile residues in SAIL-labeled ubiquitin with minimal sample requirements. Unlike methyl-selective labeling,[21] 1H-detected SSNMR for SAIL-labeled proteins offers 13C–13C connectivities and high-resolution 1H SSNMR for stereo-selectively labeled CHD groups for side-chain assignments. The method can be applied to structural analyses of a variety of proteins that are either selectively labeled with a set of different SAIL amino acids or uniformly labeled with SAIL amino acids. Although preparation of a SAIL-labeled protein in the large quantities required for conventional SSNMR experiments (0.5–1 μmol) is cost-prohibitive, the minimal sample requirement (10–50 nmol) of our approach makes such an approach very practical. The demonstrated reduction of sample requirements will make it feasible to implement even more advanced isotope labeling schemes or multiple sets of differently SAIL-labeled samples for future biomolecular SSNMR.

Secondly, we experimentally demonstrated that 1H indirect detection in 2D–3D SSNMR experiments on the SAIL-labeled protein notably improved sensitivity as well as additional 1H spectral resolution, over traditional 13C-direct detection in the corresponding 1D–2D experiments. This suggests that in our approach combining uses of UFMAS at 80 kHz and SAIL proteins, many of traditional 2D-3D 13C-detected experiments can be replaced by 1H-detected 3D-4D experiments. As discussed above, in most of previous studies the sensitivity advantage of 1H indirect detected SSNMR was described for less useful 15N SSNMR, rather than for 13C SSNMR, which is 4-fold more sensitive. In this work, we quantitatively demonstrated the sensitivity and resolution advantage of this method over traditional 13C-detected SSNMR in a high field (17.6 T), which is now widely available to the scientific community. We reported up to 10-fold sensitivity improvements by 1H detection and UFMAS at 80 kHz when a 1H-detected 2D–3D experiment is compared with an equivalent 13C-detected 2D–3D experiment. It will be feasible to reach the detection limit of a few nmol or sub-nmol of SAIL-labeled proteins by further sensitivity enhancement using paramagnetic doping[19, 22], modified polarization transfer schemes,[23], non-uniform sampling[24], and even faster MAS.[15, 25, 26] This approach is likely to applicable to a variety of micro/nano-crystalline proteins in order to drastically speed up side-chain spectral and structural analysis. Although it is outside the scope of this study, a similar approach using 1H- or 19F-detection should be possible for amyloid aggregates with reasonable structural homogeneity[2733] and bioinorganic samples.[34, 35]

Materials and Methods

The SAIL-Ile and UL-Ile used for these experiments were recrystallized in 20% DCl in D2O[17] before packing into a MAS rotor. SAIL-Ubq samples were expressed in E. coli BL21 cells harboring a plasmid containing the chlorella Ubq gene, cultured in D2O/M9 medium supplemented with SAIL-Ile, as described in the SI. The sample was crystallized by dissolving 2 mg of the lyophilized powder in 160 μL of citrate buffer in D2O and then precipitating with 240 μL d12-MPD (2-methyl-2,4-pentanediol).[20] Further details about the sample preparation are discussed in the SI.

All SSNMR experiments were performed on a Bruker Avance III 750MHz spectrometer at the UIC Center for Structural Biology using a JEOL 1 mm 1H/13C/15N/2H quad-resonance MAS probe. All the 13C–1H polarization transfers in this work were performed using adiabatic double-quantum cross-polarization (DQ-CP) schemes[19, 36], with the sum of the rf field strengths for I and S hetero-nuclear spins matched to ωR (i.e., ω1I + ω1S = ωR) so that sample heating was minimized at high repetition rates by the low-power CP scheme.[10, 15] For 1H decoupling, a low-power decoupling scheme[37] with SPINAL-64[38] at 10 kHz was applied. For 13C, 15N, and 2H decoupling, a WALTZ-16 scheme was used at RF field strengths of 10, 2, and 5 kHz, respectively. All the multi-dimensional NMR data were processed using the nmrPipe software.[39] Unless stated otherwise, all indirect time-domain signals in the 2D and 3D data were extended to 1.5-fold and 2-fold by linear prediction, respectively. The multi-dimensional SSNMR data were apodized with 45°- and 60°-shifted sine-bell window functions in the 1H and 13C dimensions, respectively, to balance sensitivity and resolution. For the SAIL-Ile and Ubq samples, 13C decoupling was applied during the 1H detection/evolution periods, whereas the 1H and 2H decoupling sequences were applied during the 13C detection/evolution periods. For the SAIL samples, the sample temperature under UFMAS at 80 kHz was kept at ~37°C using a FTS cooler unit with N2 gas at ˗12°C.

The 1H NMR spectra in Fig 1 were collected with π/2-pulse direct excitation with WALTZ-16 13C decoupling [3] at an RF field strength of 10 kHz with a recycle delay of 10 s. The details of the pulse sequences used for Figs 2 and 3 are listed in the SI (Figs B–D in S1 File). For the data in Fig 2, 13C polarization was prepared by DQ-CP transfer with a constant 13C RF field strength of ~2νR/5 and a downward-ramped 1H RF field with an average strength of ~3νR/5, using a contact time of 1.5 ms. Similar conditions, but with an upward 1H ramp, were used to transfer 13C polarization back to 1H for the 1H detection experiments in Fig 2a. The maximum t1 and t2 periods for Fig 2(a) and 2(b) were 5 ms and 10 ms, respectively. The recycle delay was set to 0.54 s for SAIL-Ubq, which had relatively short 1H T1 values (~0.25 s). For data processing, the indirect time-domain data along the t1 period were extended to 10 ms by linear prediction.

The 3D data in Fig 3 were obtained with a recycle delay of 0.52 s using the DQ—CP scheme (Fig C in S1 File) used for Fig 2a. After the t1 period for 13C evolution, the fpRFDR sequence[40] was employed for 13C–13C mixing. Subsequently, the 13C signal was recorded in the t2 period. Then, the 1H signal was acquired after DQ—CP transfer of the 13C polarization to 1H spins with a contact time of 0.5 ms. The signals were collected with maximum t1 and t2 periods of 2.4 ms and an acquisition period (t3) of 10.2 ms. The relatively short t1 and t2 periods were employed to optimize the sensitivity for the detection of weak deuterated 13Cα signals. The t1 and t2 data were extended to 3.6 ms by linear prediction and was processed as discussed above for Fig 2.

Supporting Information

S1 File. Fig. A, Spinning speed dependence of 1H MAS SSNMR of SAIL-threonine (SAIL-Thr) with its chemical structure and labeling scheme.

The 1H NMR spectra were obtained using WALTZ-16 13C decoupling with an RF field strength of 10 kHz. All the spectra were obtained with 2 scans with a pulse delay of 15 s, and the data were processed without any window function. The 1H line widths for Hα, Hβ, Hγ OH/NH are 0.22 ppm, 0.22 ppm, 0.24 ppm, 0.20 ppm, respectively. Fig. B, A pulse sequence for 1H-detected 2D 1H/13C chemical-shift correlation spectroscopy used for Fig 2a. In this sequence, 13C spin polarization was prepared with adiabatic double-quantum cross polarization (DQ-CP) using an amplitude-modulated shaped pulse with a downward tangential ramp for the 1H channel and a rectangular pulse for the 13C channel. The 1H RF field strength was swept from 66.0 kHz to 26.4 kHz with the average rf field set at 46.2 kHz (~3νR/5) while the 13C RF field amplitude was kept constant at 32.0 kHz (~2νR/5). The contact time of the first CP was 1.5ms. During the t1 period, SPINAL-64 1H decoupling[38] and WALTZ-16 2H decoupling were applied with RF field strengths of 10 kHz and 5 kHz, respectively. The t1 period was incremented up to 5.1 ms with an increment of 37 μs. After the t1 period, a pair of π/2-pulses were applied as a Z-filter in order to select the real or imaginary component of the 13C polarization, which was transferred back to 1H spins with the second adiabatic CP using a reversed upward tangential ramp for the 1H channel and the same rectangular pulse for the 13C channel. The contact time of the second CP was 1.5 ms. During the acquisition (t2) period of 10.2 ms, 1H signals were acquired with dwell times of 5 μs under 13C decoupling using WALTZ-16 sequence[41] with an RF field strength of 10 kHz. The phase cycles for the pulse sequence were as follows: ϕ1 = y; ϕ2 = x; ϕ3 = x, x, -x, -x; ϕ4 = y, y, y, y, -y, -y, -y, -y; ϕ5 = x; ϕ6 = y, -y; ϕ7 = y; ϕ8 = x, -x, -x, x, -x, x, x, -x. The phase ϕ3 and the receiver phase were incremented along the t1 points using the States-TPPI data collection mode. Fig. C, A pulse sequence used for 13C-detected 2D 1H/13C chemical-shift correlation spectroscopy in Fig 2b. A pulse sequence used for 13C-detected 2D 1H/13C chemical-shift correlation spectroscopy in Fig 2b. After excitation by a π/2-pulse, 1H spin polarization evolved under 1H chemical-shift interactions during the t1 period under WALTZ-16 13C decoupling with an RF field strength of 10 kHz. The t1 period was incremented up to 5.1 ms with a t1 increment of 0.15 ms. The 1H polarization was transferred to the 13C spins by adiabatic tangential double-quantum cross polarization (DQ-CP), which was identical to the first CP scheme in Fig B in S1 File. The contact time for CP was 1.5ms. During the acquisition (t2) period of 10.2 ms, SPINAL-64 1H decoupling and WALTZ-16 2H decoupling were applied with RF strengths of 10 kHz and 5 kHz, respectively. The t2 dwell time was 5 μs. The phase cycles for the pulse sequence were as follows: ϕ1 = y, -y; ϕ2 = x, x, -x, -x; ϕ3 = y; ϕ4 = x, -x, -x, x. The phase ϕ1 and the receiver phase were incremented along the t1 points using the States-TPPI data collection mode. Fig. D, A pulse sequence used for 1H-detected 3D 13C/13C/1H correlation spectroscopy in Fig 3. 13C spin polarization was prepared by adiabatic double-quantum cross polarization (DQ-CP) using the same parameters as discussed in Fig B in S1 File. During the t1 period, SPINAL-64 1H decoupling and WALTZ-16 2H decoupling were applied with RF field strengths of 10 kHz and 5 kHz, respectively. After the t1 period, a transverse component of the 13C polarization was stored along the z-axis and the unnecessary component in the transverse plane is dephased during a z-filter period τ of 2 ms. Then, 13C polarization transfer was achieved by 13C-13C dipolar couplings using the fpRFDR sequence without 1H rf irradiation. A π-pulse train with the XY-16 phase cycle was rotor-synchronously applied to the 13C channel so that a π-pulse was applied at the center of every rotor cycle. The π-pulse width in the fpRFDR mixing was 6.6 μs, and n = 96. After a z-filter and excitation by a π/2-pulse, 13C signals were recorded during the t2 period under SPINAL-64 1H decoupling and WALTZ-16 2H decoupling, as mentioned above for the t1 period. Then, a transverse component of the 13C polarization was transferred back to 1H spins by an adiabatic DQ-CP scheme before the acquisition of 1H signals in the t3 period. The 1H RF field strength was swept from 26.4 kHz to 66.0 kHz with the average rf field at 46.2 kHz (~3νR/5) while the 13C RF field amplitude was set kept constant at 32.0 kHz (~2νR/5). The contact time of the second CP period was 0.5 ms. The t1 and t2 periods were both incremented up to 2.4 ms with an increment of 75 μs. The t3 acquisition time was 10.2 ms with 5 μs dwell time. The phase cycles for the pulse sequence were as follows: ϕ1 = y; ϕ2 = x; ϕ3 = x, x, -x, -x; ϕ4 = y, y, y, y, -y, -y, -y, -y; ϕ5 = y; ϕ6 = x, -x; ϕ7 = x; ϕ8 = x, -x, -x, x, -x, x, x, -x. The phases ϕ3 and ϕ5 and the receiver phase were incremented along the t1 and t2 points using the States-TPPI data collection mode. Fig. E, a) 1H-detected 13C/1H 2D correlation and b) 13C-detected 13C/1H 2D correlation, and c) 1D 13C CP-MAS spectra of SAIL-Ile respectively. 1D slices at various 1H chemical shifts (indicated in the fig.) from the 1H and 13C detected 2D 13C/1H correlation spectra are compared. The 1D slices and 1D spectrum in (c) are scaled so that all the 1D spectra show a common noise level for sensitivity comparisons. The experimental time was 5 min each. The pulse sequences used for (a) 13C-detected and (b) 1H-detected 2D 1H/13C chemical-shift correlation experiments are shown in Fig C and Fig B in S1 File, respectively. The CP and decoupling conditions for these experiments were similar to those for the data for the SAIL Ile labeled ubiquitin sample in Fig 2. The 13C detection/evolution periods was 10 ms, while 1H detection/evolution periods was 6.5 ms for a) and b). These periods were matched to the inverse of the average line widths of 13C and 1H. Although 1H T1 value for this sample was ~3 s, the recycle delay was set to 0.3 s as sufficient signal-to-noise ratios can be obtained for all of (a-c). All the spectra in Fig. E were processed with 45̊- and 60̊-shifted sinebell functions on the 1H and 13C dimensions respectively without linear prediction. Fig. F, A comparison of 1D 13C MAS spectra of SAIL-Ile by a) π/2-pulse direct excitation and b) cross-polarization (CP) from 1H spins. The pulse sequence for cross polarization experiments used in (b) was the same as showed in Fig C in S1 File except that the t1 value was set to 0.1 μs and t2 was used as an acquisition period. The cross polarization transfer was optimized for protonated carbons for 1H-detected experiments. The 1H RF field strength was swept from 70.0 kHz to 28.0 kHz with the average rf field set at 49.0 kHz (~5νR/8) while the 13C RF field strength was kept constant at 30.0 kHz (~3νR/8). The contact time of CP was 1.5 ms. The 13C detection periods was 3.1 ms for both (a) and (b). Recycle delays were set to 6000 s and 20 s for (a) and (b), respectively. The long delays were employed to ensure that the signals were fully recovered. No window functions were applied to the spectra. The CP-transfer efficiency for Cα, Cβ, Cγ1, Cγ2 and Cδ were 55%, 60%, 68%, 45% and 40%, respectively. The values were obtained by dividing the ratio of the integral peak intensity in (b) to that of the corresponding peak in (a) by γHC, where γH and γC are the gyromagnetic ratios of 1H and 13C, respectively. Table A, The comparison of S/N for 1H-detected 2D 13C/1H correlation, 13C-detected 13C/1H correlation, and 13C 1D CPMAS experiments of the SAIL-Ile labeled ubiquitin sample. Table B, The comparison of signal-to-noise ratios (S/N) for 1H-detected 2D 13C/1H correlation, 13C-detected 13C/1H correlation, and 13C 1D CPMAS experiments of SAIL Ile sample. Table C, Preliminary 1H and 13C signal assignments of SAIL-Ile labeled ubiquitin.

https://doi.org/10.1371/journal.pone.0122714.s001

(PDF)

Author Contributions

Conceived and designed the experiments: SW SP MK YI. Performed the experiments: SW SP YN YE TN KY MT YI. Analyzed the data: SW SP YI. Contributed reagents/materials/analysis tools: SW SP YN YE TN KY TA MT TT MK YI. Wrote the paper: SW SP YN YE TN KY TA MT TT MK YI.

References

  1. 1. Ishii Y, Tycko R. Sensitivity enhancement in solid state N-15 NMR by indirect detection with high-speed magic angle spinning. J. Magn. Reson. 2000;142(1):199–204. pmid:10617453
  2. 2. Ishii Y, Yesinowski JP, Tycko R. Sensitivity enhancement in solid-state C-13 NMR of synthetic polymers and biopolymers by H-1 NMR detection with high-speed magic angle spinning. J. Am. Chem. Soc. 2001;123(12):2921–2922. pmid:11456995
  3. 3. Paulson EK, Morcombe CR, Gaponenko V, Dancheck B, Byrd RA, Zilm KW. Sensitive high resolution inverse detection NMR spectroscopy of proteins in the solid state. J. Am. Chem. Soc. 2003;125(51):15831–15836. pmid:14677974
  4. 4. Schnell I, Saalwachter K. N-15-H-1 bond length determination in natural abundance by inverse detection in fast-MAS solid-state NMR spectroscopy. J. Am. Chem. Soc. 2002;124(37):10938–10939. pmid:12224915
  5. 5. Chevelkov V, Rehbein K, Diehl A, Reif B. Ultrahigh resolution in proton solid-state NMR spectroscopy at high levels of deuteration. Angew. Chem. Int. Edit. 2006;45(23):3878–3881. pmid:16646097
  6. 6. Reif B. Ultra-high resolution in MAS solid-state NMR of perdeuterated proteins: Implications for structure and dynamics. J. Magn. Reson. 2012;216:1–12. pmid:22280934
  7. 7. Zhou DH, Shah G, Cormos M, Mullen C, Sandoz D, Rienstra CM. Proton-detected solid-state NMR spectroscopy of fully protonated proteins at 40 kHz magic angle spinning. J. Am. Chem. Soc. 2007;129:11791–11801. pmid:17725352
  8. 8. Knight MJ, Webber AL, Pell AJ, Guerry P, Barbet-Massin E, Bertini I, et al. Fast Resonance Assignment and Fold Determination of Human Superoxide Dismutase by High-Resolution Proton-Detected Solid-State MAS NMR Spectroscopy. Angew. Chem.-Int Edit. 2011;50(49):11697–11701. pmid:21998020
  9. 9. Barbet-Massin E, Pell AJ, Retel JS, Andreas LB, Jaudzems K, Franks WT, et al. Rapid Proton-Detected NMR Assignment for Proteins with Fast Magic Angle Spinning. J. Am. Chem. Soc. 2014;136:12489–12497. pmid:25102442
  10. 10. Marchetti A, Jehle S, Felletti M, Knight MJ, Wang Y, Xu Z-Q, et al. Backbone Assignment of Fully Protonated Solid Proteins by 1H Detection and Ultrafast Magic-Angle-Spinning NMR Spectroscopy. Angew. Chem. Int. Edit. 2012;51(43):10756–10759. pmid:23023570
  11. 11. Ishii Y, Yesinowski JP, Tycko R. Sensitivity enhancement in solid-state C-13 NMR of synthetic polymers and biopolymers by H-1 NMR detection with high-speed magic angle spinning. J. Am. Chem. Soc. 2001;123(12):2921–2922. pmid:11456995
  12. 12. Guo C, Hou G, Lu X, O'Hare B, Struppe J, Polenova T. Fast magic angle spinning NMR with heteronucleus detection for resonance assignments and structural characterization of fully protonated proteins. J. Biomol. NMR. 2014;60(4):219–229. pmid:25381566
  13. 13. Kainosho M, Torizawa T, Iwashita Y, Terauchi T, Ono AM, Guntert P. Optimal isotope labelling for NMR protein structure determinations. Nature. 2006;440(7080):52–57. pmid:16511487
  14. 14. Takahashi H, Kainosho M, Akutsu H, Fujiwara T. H-1-detected H-1-H-1 correlation spectroscopy of a stereo-array isotope labeled amino acid under fast magic-angle spinning. J. Magn. Reson. 2010;203(2):253–256. pmid:20129804
  15. 15. Parathasarathy S, Nishiyama Y, Ishii Y. Sensitivity and resolution enhanced solid-state NMR for paramagnetic systems and biomoleculesunder very fast magic agle spinning. Acc. Chem. Res. 2013;46:2127–2135. pmid:23889329
  16. 16. Kainosho M, Guentert P. SAIL—stereo-array isotope labeling. Q. Rev. Biophys. 2009;42(4):247–300. pmid:20370954
  17. 17. Torii K, Iitaka Y. Crystal structure of L-isoleucine. Acta Crystallographica Section B-Structural Crystallography and Crystal Chemistry. 1971;B 27(NOV15):2237-&.
  18. 18. Varughese KI, Srinivasan R. Studies in molecular-structure, symmetry and conformation. 14. Crystal and molecular-structure of L-isoleucine hydrochloride monohydrate form-2. Pramana. 1976;6(4):189–195.
  19. 19. Wickramasinghe NP, Parthasarathy S, Jones CR, Bhardwaj C, Long F, Kotecha M, et al. Nanomole-scale protein solid-state NMR by breaking intrinsic H-1 T-1 boundaries. Nature Methods. 2009;6(3):215–218. pmid:19198596
  20. 20. Igumenova TI, McDermott AE, Zilm KW, Martin RW, Paulson EK, Wand AJ. Assignments of carbon NMR resonances for microcrystalline ubiquitin. J. Am. Chem. Soc. 2004;126(21):6720–6727. pmid:15161300
  21. 21. Huber M, Boeckmann A, Hiller S, Meier BH. 4D solid-state NMR for protein structure determination. Physical Chemistry Chemical Physics. 2012;14(15):5239–5246. pmid:22402636
  22. 22. Nadaud PS, Helmus JJ, Sengupta I, Jaroniec CP. Rapid Acquisition of Multidimensional Solid-State NMR Spectra of Proteins Facilitated by Covalently Bound Paramagnetic Tags. J. Am. Chem. Soc. 2010;132(28):9561–9563. pmid:20583834
  23. 23. Althaus SM, Mao K, Stringer JA, Kobayashi T, Pruski M. Indirectly detected heteronuclear correlation solid-state NMR spectroscopy of naturally abundant N-15 nuclei. Solid State Nucl. Magn. Reson. 2014;57–58:17–21. pmid:25466354
  24. 24. Suiter CL, Paramasivam S, Hou G, Sun S, Rice D, Hoch JC, et al. Sensitivity gains, linearity, and spectral reproducibility in nonuniformly sampled multidimensional MAS NMR spectra of high dynamic range. J. Biomol. NMR. 2014;59(2):57–73. pmid:24752819
  25. 25. Kobayashi T, Mao K, Paluch P, Nowak-Krol A, Sniechowska J, Nishiyama Y, et al. Study of Intermolecular Interactions in the Corrole Matrix by Solid-State NMR under 100 kHz MAS and Theoretical Calculations. Angew. Chem. Int. Edit. 2013;52(52):14108–14111. pmid:24227750
  26. 26. Nishiyama Y, Malon M, Ishii Y, Ramamoorthy A. 3D N-15/N-15/H-1 chemical shift correlation experiment utilizing an RFDR-based H-1/H-1 mixing period at 100 kHz MAS. J. Magn. Reson. 2014;244:1–5. pmid:24801998
  27. 27. Rienstra CM, Tucker-Kellogg L, Jaroniec CP, Hohwy M, Reif B, McMahon MT, et al. De novo determination of peptide structure with solid-state magic-angle spinning NMR spectroscopy. Proc. Natl. Acad. Sci. U. S. A. 2002;99(16):10260–10265. pmid:12149447
  28. 28. Heise H, Hoyer W, Becker S, Andronesi OC, Riedel D, Baldus M. Molecular-level secondary structure, polymorphism, and dynamics of full-length alpha-synuclein fibrils studied by solid-state NMR. Proc. Natl. Acad. Sci. U. S. A. 2005;102(44):15871–15876. pmid:16247008
  29. 29. Wasmer C, Lange A, Van Melckebeke H, Siemer AB, Riek R, Meier BH. Amyloid fibrils of the HET-s(218–289) prion form a beta solenoid with a triangular hydrophobic core. Science. 2008;319(5869):1523–1526. pmid:18339938
  30. 30. Helmus JJ, Surewicz K, Nadaud PS, Surewicz WK, Jaroniec CP. Molecular conformation and dynamics of the Y145Stop variant of human prion protein. Proc. Natl. Acad. Sci. U. S. A. 2008;105(17):6284–6289. pmid:18436646
  31. 31. Schneider R, Schumacher MC, Mueller H, Nand D, Klaukien V, Heise H, et al. Structural Characterization of Polyglutamine Fibrils by Solid-State NMR Spectroscopy. J. Mol. Biol. 2011;412(1):121–136. pmid:21763317
  32. 32. Lu J-X, Qiang W, Yau W-M, Schwieters CD, Meredith SC, Tycko R. Molecular Structure of beta-Amyloid Fibrils in Alzheimer's Disease Brain Tissue. Cell. 2013;154(6):1257–1268. pmid:24034249
  33. 33. Hoop CL, Lin H-K, Kar K, Hou Z, Poirier MA, Wetzel R, et al. Polyglutamine amyloid core boundaries and flanking domain dynamics in huntingtin fragment fibrils determined by solid-state nuclear magnetic resonance. Biochemistry. 2014;53(42):6653–6666. pmid:25280367
  34. 34. Goobes G, Goobes R, Schueler-Furman O, Baker D, Stayton PS, Drobny GP. Folding of the C-terminal bacterial binding domain in statherin upon adsorption onto hydroxyapatite crystals. Proc. Natl. Acad. Sci. U. S. A. 2006;103(44):16083–16088. pmid:17060618
  35. 35. Vyalikh A, Simon P, Rosseeva E, Buder J, Kniep R, Scheler U. Intergrowth and Interfacial Structure of Biomimetic Fluorapatite- Gelatin Nanocomposite: A Solid-State NMR Study. J. Phys. Chem. B. 2014;118(3):724–730. pmid:24354406
  36. 36. Lange A, Scholz I, Manolikas T, Ernst M, Meier BH. Low-power cross polarization in fast magic-angle spinning NMR experiments. Chem. Phys. Lett. 2009;468(1–3):100–105.
  37. 37. Kotecha M, Wickramasinghe NP, Ishii Y. Efficient low-power heteronuclear decoupling in 13C high-resolution solid-state NMR under fast magic angle spinning. Magn. Reson. Chem. 2007;45:S221–230. pmid:18157841
  38. 38. Fung BM, Khitrin AK, Ermolaev K. An improved broadband decoupling sequence for liquid crystals and solids. J. Magn. Reson. 2000;142(1):97–101. pmid:10617439
  39. 39. Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A. Nmrpipe—a Multidimensional Spectral Processing System Based on Unix Pipes. J. Biomol. NMR. 1995;6(3):277–293. pmid:8520220
  40. 40. Ishii Y. 13C-13C dipolar recoupling under very fast magic angle spinning in solid-state NMR: Applications to distance measurements, spectral assignments, and high-throughput secondary-structure elucidation. J. Chem. Phys. 2001;114:8473–8483.
  41. 41. Shaka AJ, Keeler J, Frenkiel T, Freeman J. An improved sequence for broadband decoupling: WALTZ-16. J. Magn. Reson. 1983;52:335–338.