Journal of Magnetic Resonance

Transcription

1 Journal of Magnetic Resonance 230 (2013) Contents lists available at SciVerse ScienceDirect Journal of Magnetic Resonance journal homeage: Communication Efficient CO CA transfer in highly deuterated roteins by band-selective homonuclear cross-olarization Veniamin Chevelkov, Karin Giller, Stefan Becker, Adam Lange Max Planck Institute for Biohysical Chemistry, Göttingen, Germany article info abstract Article history: Received 10 December 2012 Revised 8 February 2013 Available online 15 March 2013 Keywords: Cross olarization Magnetization transfer Recouling Protein resonance assignment Deuteration Robust and efficient band-selective magnetization transfer between CO and CA sins can be achieved in highly deuterated solid roteins by diolar-based homonuclear cross olarization. The aroach is designed for moderate magic-angle sinning rates and high external magnetic fields where the isotroic chemical shift difference of CO and CA considerably exceeds the sinning rate. The most efficient recouling is achieved when the sum of effective radio-frequency fields on CO and CA resonances equals two times the sinning rate. This method can be directly imlemented in roton-detected versions of interresidual correlation exeriments as needed for resonance assignment in rotein solid-state NMR sectroscoy. Ó 2013 Elsevier Inc. All rights reserved. 1. Introduction In the last decade solid-state NMR (ssnmr) sectroscoy of rotating solids has rogressed into a unique and owerful tool to study structure [1 3] and dynamics [4,5] of biological macromolecules at atomic resolution. To overcome the inherently low sensitivity and resolution, which limits the alication of magic-angle sinning (MAS) ssnmr, a number of develoments have been made. One romising aroach in biomolecular ssnmr is to emloy roton detection to make use of the high gyromagnetic ratio of rotons [6 9]. These exeriments became feasible by emloying deuterated samles and/or high samle rotation rates. The aroach resulted in a number of successful studies of dynamics [4,5] and structure [9 11] of biomolecules. An essential ste in rotein NMR studies is the resonance assignment, which requires new or modified exerimental aroaches in case of roton detected NMR comared to conventionally used carbon detected exeriments. There are numerous strategies to assign deuterated roteins in roton-detected exeriments using both scalar and diolar coulings [9,12 14]. Recent studies show that roton-detected exeriments allow for the assignment of rather large systems, namely the 16 kda rotein SOD [14] and the 21 kda rotein DsbA [12]. One crucial ste in many assignment schemes is the magnetization transfer between CO and CA nuclei, for examle in roton-detected exeriments of the tye (x)co(ca)nh. Pintacuda and coworkers made use of the relatively strong CO CA scalar couling [14] of 55 Hz in INEPT-tye exeriments, while Corresonding author. address: adla@nmr.mibc.mg.de (A. Lange). Zhou et al. alied DREAM [12], that recoules homonuclear diolar interactions, for olarization transfer. The alication of INEPT transfer bears a number of exerimental drawbacks. Due to scalar relaxation of the second kind [15,16], the transversal magnetization life time of CA bound to 2 H is relatively short in the absence of 13 C 2 H scalar decouling. Such scalar decouling has recently been achieved by Meier and coworkers [17] but in most commercial ssnmr robes the necesseray fourth channel for deuterium is missing. Furthemore, the INEPT sequence comrises 10 selective ulses and due to robe inhomogeneity this can result in oor erformance. Even moderate magnetization losses on the order of 10% for a single ulse result in overall losses of 65%. At sinning rates that can be reached with 3.2 mm robes (25 khz), olarization losses originate also from imerfectly averaged CSAs and from the interference of sinning rate and the alied 13 C RF field with an amlitude of 6 8 khz used for the band-selective ulses. Alternatively, DREAM [18] can be emloyed in a straightforward manner and it is very efficient, but requires magic-angle sinning rates that are higher then the chemical shift difference between CO and CA (e.g. 120 m = 24 khz at 800 MHz roton Larmor frequency) [19]. This condition is not always achievable at moderate sinning rates and high magnetic fields. Here, we resent an alternative robust exerimental method for diolar-based CO CA magnetization transfer that works well at moderate MAS rates (ca. 20 khz) that are smaller than the isotroic chemical shift difference between the CO and CA bands. The method relies on efficient band-selective homonuclear cross-olarization (BSH-CP). Both zero- and double-quantum homonuclear recouling can be achieved by continuous RF irradiation, if the Hartmann Hahn condition is satisfied, i.e. if the sum or difference /$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. htt://dx.doi.org/ /j.jmr

2 206 Communication / Journal of Magnetic Resonance 230 (2013) of the effective fields acting on CO and CA is equal to one or two times the sinninig rate. Under the exerimental conditions of the resent study, we obtained efficient CO CA magnetization transfer when (i) the sum was equal to two times the sinning rate or (ii) when the difference was equal to the sinning rate, resectively. Exerimentally, we achieved magnetization transfer rates of u to 50%, allowing us to record 2D CO CA correlation sectra of highly deuterated PrgI needles with high sensitivity [20]. The described method is concetually similar to well-known recouling techniques under continuous RF irradiation such as HORROR [21], DREAM [18], and rotational resonance in the tilted frame [22 24], but uses different recouling conditions, roviding a new tool for efficient band-selective homonuclear magnetization transfer at moderate sinning rates (ca khz) and high magnetic fields (ca MHz 1 H Larmor frequency). 2. Materials and methods Exression, urification and olymerization of erdeuterated 15 N- and 13 C-labelled wild-tye PrgI rotein was erformed as described before [3,20]. A low degree of rotonation of the samle was achieved according to the aroach described earlier by Reif and coworkers [6]. Proton back-exchange took lace in a buffer containing a mixture of 10% 1 H 2 O and 90% 2 H 2 O. Aroximately 12 mg of material were acked into a 3.2 mm rotor. All exeriments were conducted on a 20 T wide-bore sectrometer (Bruker Biosin, Germany) equied with an Avance III console and a ( 1 H, 13 C, 15 N) trile-resonance 3.2 mm robe. The MAS rate was adjusted to 20 khz and the effective samle temerature was set to 12 ± 2 C as measured by the temerature-deendent water roton resonance relative to an internal DSS reference. Hard roton and carbon ulses emloyed RF field strengths of 70 khz and 50 khz, resectively. WALTZ-16 [25] decouling with an RF field strength of 2.5 khz was alied on the 15 N channel during 13 C evolution and detection eriods to remove heteronuclear scalar coulings. A contact time of 3.2 ms and 2.35 ms was used for the H-CO and H-CA heteronuclear CP stes, resectively. A constant roton field strength of 49 khz was used, whereas the 13 C RF field strength was ramed down to 80% from an initial value of 32 khz. 3. Theoretical background In this section we briefly outline the theory of the sin dynamics of two diolar-couled homonuclear sins I and S in a owdered solid under MAS and continuous RF irradiation. A detailed descrition has been resented elsewhere [22 24]. For the sake of simlicity we neglect the chemical shift anisotroy (CSA). The total sin Hamiltonian (SH) in the rotating frame is given by the sum of the time-deendent diolar art H D (t) as well as the Zeeman and RF field art H Z : HðtÞ ¼H Z þ H D ðtþ with the Zeeman and RF field interactions: H Z ¼ Dx I I Z Dx S S Z x 1 ðs X þ I X Þ ð2þ where Dx X is the isotroic chemical shift offset from the carrier frequency for the sin X (I or S). The cartesian comonents x, y, z of sin oerators S and I are reresented by S X, S Y, S Z and I X, I Y, I Z, resectively. The amlitude of the alied continuous RF irradiation is x 1. The diolar interaction in the rotating frame is reresented by: H D ðtþ ¼DðtÞ S Z I Z 1 ð 4 I þs þ S þ I Þ ð3þ ð1þ where the satial art of the diolar couling D(t) is eriodically modulated by MAS with the sinning rate x r and can be exressed as: DðtÞ ¼ X ð4þ n¼1;2 x D;n exði nx r tþ 2 x D;1 ¼ 2 d sin ð 2h DÞexði u D Þ x D;2 ¼ 1 2 d sin2 ð h D Þ exð2i u D Þ where the olar angles h D and u D define the direction of the internuclear S I vector in a rotor-fixed frame and the diolar couling constant d ¼ ðl 0 =4Þðc 2 h=r 3 Þ deends on the internuclear distance r and the gyromagnetic ratio c. l 0 and h are the ermeability of free sace and Planck s constant, resectively. The effective fields x X,eff for sins I and S are determined by the alied RF field x 1 and the chemical shift offsets Dx X : x X;eff ¼ x 2 1 þ 1=2 Dx2 X ð6þ and are tilted with resect to the static external magnetic field by the angles: b X ¼ arctanðx 1 =Dx X Þ Transformation into a tilted rotating frame determined by the effective RF fields can be achieved using the unitary oerator: U 1;eff ¼ exðib I I Y Þexðib S S Y Þ The resulting Zeeman and diolar terms are then given by: H 0 Z ¼ x I;eff I 0 Z x S;eff S 0 Z H 0 D ðtþ ¼DðtÞ A S0 Z I0 Z 1 4 B S0 þ I0 þ S0 I0 þ þ 1 4 Q S0 þ I0 þ þ S0 I0 þ R S 0 Z I0 þ P S0 I0 Z ð5þ ð7þ ð8þ ð9þ ð10þ The geometrical scaling factors A, B, Q, R and P have been described elsewhere [22 24]. The relevant ones for zero- and double-quantum recouling (ZQ and DQ, resectively) are B and Q: B ¼ 1 ð 2 1 þ cos b S cos b I 2 sin b S sin b I Þ Q ¼ 1 ð 2 1 cos b S cos b I þ 2 sin b S sin b I Þ Transition into the interaction frame by the unitary oerator U 2;eff ðtþ ¼exð ix I;eff I 0 Z tþexð ix S;eff S 0 Z tþ ð11þ ð12þ reduces the SH to the diolar interaction only. Using the transformation roerty of raising and lowering oerators, exð ix S z tþs exðix S z tþ¼s exðix tþ, one can readily obtain the SH in the interaction frame: ( H 00 D ðtþ ¼U 2;eff ðtþh 0 D ðtþu 1 2;eff ¼ X ) x D;n exðinx r tþ n¼1;2 A S 00 Z I00 Z 1 4 B S00 þ I00 exð id eff tþ þ S 00 I00 þ exðid eff tþ þ 1 4 Q S00 þ I00 þ exð ir eff tþ þ S 00 I00 exðir eff tþ þ R S 00 Z I00 exðix I;eff tþ þ P S 00 I00 Z exðix S;eff tþ ð13þ

3 Communication / Journal of Magnetic Resonance 230 (2013) Fig. 1. (A) Pulse sequence for CO CA 2D correlation exeriments based on band-selective homonuclear (BSH) CP transfer. (B) Effective 13 C RF field strength for several carbon resonances with different chemical shifts as a function of the 13 C RF field strength alied at 58 m. Deviations from Hartmann Hahn conditions as defined in the text are lotted as well. (C) Scaling factor B for n = 1 ZQ recouling, and scaling factor Q= 2 for n = 2 DQ recouling. Circles indicate the scaling factors for RF field strengths that fulfill the recouling conditions. (D) Required RF field strength for homonuclear CO CA CP at the n = 2 DQ recouling condition as a function of carrier frequency offset from 58 m. (E) Scaling factor Q= 2 for CO CA transfer at the n = 2 DQ recouling condition as a function of carrier frequency offset from 58 m. (F) Deviation from the recouling condition for n = 2, DQ BSH-CP as defined in the text as a function of carrier frequency offset from 58 m. where the sum and difference of effective fields are given by R eff = x S,eff + x I,eff and D eff = x S,eff x I,eff. Recouling takes lace at various conditions when the modulations of the satial and sin arts of the SH mutually cancel each

4 208 Communication / Journal of Magnetic Resonance 230 (2013) other. Here, we are interested in recouling of the fli flo and flo flo terms that corresond to ZQ and DQ transitions, resectively, and which occur at: D eff ¼jx S;eff x I;eff j¼n x r and R eff ¼ x S;eff þ x I;eff ¼ n x r ð14þ ð15þ where n =1,2. Recouling techniques that fulfill Eqs. (14) and (15) are (I) rotational resonance in the tilted frame [22 24] which is a frequencyselective recouling method and (II) HORROR [21] and DREAM [18] which require that the recouling RF field strength is set to aroximately half of the sinning rate, which directly defines the recouling band width. These techniques are alicable when the isotroic chemical shift differences between the recouled sins are considerably smaller than the sinning rate. Here, we show that homonuclear Hartman Hahn CP transfer fulfilling also Eqs. (14) and (15) at moderate MAS rates and at high magnetic fields can be efficiently emloyed for band-selective CO CA magnetization transfer in deuterated roteins even though the isotroic chemical shift difference of CO and CA considerably exceeds the sinning rate. As will be shown below, the resulting relatively weak scaling factors for ZQ and DQ recouling entail the use of long CP contact times. Nevertheless, in highly deuterated roteins, exerimental drawbacks that would usually be associated with such long contact times such as significant loss of magnetization, samle heating and eventually robe damage, can be avoided because no high ower roton decouling is required. 4. Exerimental results and discussion Fig. 1A resents the ulse rogram for 2D CO CA correlation sectroscoy based on band-selective homonuclear CP (BSH-CP) transfer. Directly excited roton magnetization is selectively transferred to CO by CP followed by a roton fli back (FB) ulse in order to reserve and reuse remaining roton magnetization [26]. Emloying a roton FB allows reducing the exerimental time by a factor of 2.2. After subsequent chemical shift evolution, the CO magnetization is flied by a hard trim ulse along the Z 0 axis of the tilted rotating frame determined by the 13 C RF field alied on CA for BSH-CP. The fli angle is (90-b CO ) and determined by Eq. (7). Polarization from CO is transferred to CA under continuous RF irradiation, chosen to satisfy the n = 2, DQ recouling condition for CO/CA sin airs. During 13 C evolution and detection eriods, WALTZ-16 [25] is alied on the 15 N channel to remove 15 N 13 C scalar coulings. No roton decouling is used during any stage of the exeriment. The calculation results resented in Fig. 1B F assume an external magnetic field corresonding to 850 MHz 1 H Larmor frequency and an MAS rate of 20 khz. Fig. 1B shows the effective 13 C RF field strength for a number of carbon resonances with different chemical shifts as a function of the 13 C RF field strength alied at 58 m as determined by Eq. (6). The red and black lines reresent the effective 13 C fields at 58 m and 68 m, resectively. Within the band of the CA resonances the effective fields are not very deendent on the resonance osition which can be seen from the fact that the lines lie very close together. The chemical shift variations in the CO band affect the effective RF field more strongly, as demonstrated by the blue and green dashed lines corresonding to resonances at 178 m and 183 m, resectively. The difference is about 1 khz and can mismatch the BSH-CP conditions for some resonances due to chemical shift disersion. Solid blue and green lines show the deviation of two resonances at 58 m and 178 m from ideal DQ, n = 2 and ZQ, n = 1 conditions at different 13 C field strengths alied at 58 m. The calculations were made according to Eqs. (14) and (15). The deviations from DQ, n =2 (DCP DQ2 ) and ZQ, n =1(DCP ZQ1 ) recouling conditions were calculated as DCP DQ2 = x 178m,eff + x 58m,eff 2x r and DCP ZQ1 = x 178m,eff x 58m,eff x r resectively. These values show qualitatively how robust the recouling is with resect to ossible RF field variations. The lot shows that ZQ and DQ recouling take lace at RF strengths of 6.5 and 11.8 khz, resectively. The efficiency of the magnetization transfer is determined by the scaling factors of the recouled interactions, rovided by Eq. (11). Green and blue lines in Fig. 1C reresent the scaling factor B for ZQ recouling, n = 1 and the factor Q= 2 for DQ recouling, n = 2, resectively, as a function of the alied RF field amlitude. In the latter case we scaled the factor Q by 2 to allow for a direct comarison of the considered scaling factors, because the amlitude of the recouled interaction at conditions with n = 2 is smaller by 2 comared to the amlitude at conditions with n = 1, which follows from Eq. (5). The low steeness of the lines indicates that the recouling conditions are robust with resect to RF field amlitude variations. Fig. 1D shows the required RF field strength for homonuclear CP between two resonances at 58 m and 178 m as a function of the RF carrier offset from 58 m to higher m values. Fig. 1E reresents the scaling factor Q= 2 for the n = 2, DQ recouling condition between two resonances at 58 m and 178 m as a function of RF carrier offset from 58 m. Fig. 1F shows the difference between double MAS rate (2x r = 40 khz) and sum of effective fields at 183 m and 68 m when the alied RF field matches the DQ, n = 2 CP condition for two resonances at 58 m and 178 m. The lot reveals the deviation from the ideal recouling condition due to ossible isotroic chemical shift variations of CA and CO. Fig. 1D and E indicate that both scaling factor and RF field strength for the DQ, n = 2 recouling condition are well accetable at any RF carrier osition, while the deviation from the recouling condition due to chemical shift disersion is minimal when the RF carrier is set to the center of the CA band, which is imosed by the larger chemical shift disersion of this carbon tye comared to CO. Fig. 2. To: CA signal after CO CA BSH-CP as a function of maximal RF field strength alied at 58 m. Bottom: CA magnetization build-u as a function of contact time during BSH-CP at the n = 2 DQ recouling condition.

5 Communication / Journal of Magnetic Resonance 230 (2013) Fig. 2 demonstrates CO CA homonuclear magnetization transfer as a function of the alied RF field strength (to) and for the n = 2 DQ recouling condition as a function of the CP contact time (bottom). One-dimensional exeriments were erformed according to the ulse scheme shown in Fig. 1A, with the 13 C field ramed down to 80% amlitude during BSH-CP. The to anel shows magnetization transfer at both recouling conditions, namely ZQ, n = 1 and DQ, n = 2 at maximum RF field strengths of ca. 7 khz and 12.8 khz, resectively, in agreement with the theoretical exectations (Fig. 1 and Eqs. (14) and (15)). The trim ulse was set to 70 during the RF field strength otimization (as shown in the to anel), as a comromise between otimal values of 76 and 65 for ZQ and DQ transfer, resectively. A contact time of 14 ms was used. The otimization reveals that the DQ transfer rovides higher transfer efficiency comared to ZQ transfer, which is in line with theoretical exectations concerning the scaling factor and the recouling bandwidth. The magnetization builds u relatively slowly and reaches a maximum after 11 ms for DQ, n = 2 recouling, as lotted in the bottom anel. Notably, such long contact times are not a severe limitation for the alication to highly deuterated roteins, due to the resulting extremely long coherence lifetime in the absence of high-ower roton decouling. Furthermore, the moderately high RF strength for DQ, n = 2 recouling makes the transfer quite robust in terms of hardware instability and RF inhomogeneity. The build-u dynamics are consistent with results from numerical simulations as resented in the subsequent section. Fig. 3 comares reference 1D 13 C sectra obtained by bandselective H-CO or H-CA CP (anels B, D, F) with sectra obtained by n = 2, DQ homonuclear CP following band-selective H-CO or H-CA CP (anels A, C, E). The 1D version of the ulse scheme in Fig. 1A was emloyed for the sectra in anels A and C. The CO sectrum in anel E was recorded by means of a straightforward adatation of the above mentioned ulse sequence, which is deicted in Fig. 1 of the Suorting information (SI). During homonuclear CP the carbon RF field was ramed down to 70% in C and E and to 80% in A. Panels B, D and F give 1D CA and CO reference signals obtained by standard band-selective heteronuclear 1 H 13 CA/ 13 CO CP. The same exerimental sectrum is lotted in B and D. The number of scans for BSH-CP exeriments in A, C and E was two times more then for the reference sectra in B, D and F. For the maximal 13 C RF field strength that was ramed down to 80% and 70% during BSH-CP otimal values of 12.8 khz and 14.4 khz were found and the otimum contact times were 11.4 ms and 13.5 ms, resectively. Fig. 3 reveals that the carbon signal is smaller by factor of in the homonuclear CP sectra comared to the reference sectra. The data also show the advantage of using a less stee ram on the 13 C RF field. Integral values for the regions m or m are rovided in the anels, without accounting for the number of scans. The values in arentheses refer to the estimated efficiency of homonuclear CP. This is given by the ratio of the integral value of the current sectrum and the integral value of the reference sectrum of the carbon Fig. 3. (A, C and E) CA and CO signal after BSH-CP in comarison with (B, D and F) reference heteronuclear CP sectra. The number of scans for homonuclear CP exeriments (A, C and E) was two times more than for the reference exeriments (B, D and F). The numbers in the anel refer to the integrated signal in the regions m (CA) or m (CO); the different number of scans was not accounted for. The estimated efficiency of BSH-CP in each exeriment is given in arentheses. Fig. 4. A 2D CO CA correlation sectrum of U-[ 2 H, 13 C, 15 N]-labeled tye three secretion system (T3SS) needles (at 10% back-exchange of amide rotons). The CO CA transfer relied on BSH-CP at the n = 2 DQ recouling condition. The sectrum was recorded on an 850 MHz sectrometer at 20 khz MAS.

6 210 Communication / Journal of Magnetic Resonance 230 (2013) region, that was olarized rior homonuclear CP transfer. To calculate the efficiency we normalized the integral values by the number of scans used. In detail, the values from the following anels were comared: A and F, C and F, E and B. The CA region after heteronuclear CP contains 11 CB resonances of Ser and Thr, so we estimated that 88% of the intensity in the sectrum corresonds to CA resonances by assuming equal olarization of CA and CB and no olarization of the flexible residues M1 and A2 [20]. We neglected a ossible contribution of the side chain signals of As, Asn, Glu and Gln to the CO region after H-CO CP. In any case, including this contribution into the analysis would only increase the estimated efficiency of CO CA transfer. The resented results show a high efficiency of u to 50% for BSH-CP transfer which suggests its use as a building block in N i CO i 1 CA i 1 exeriments that are of aramount imortance in rotein resonance assignment. Fig. 4 shows a CO CA correlation sectrum acquired within 35 h according to the ulse scheme resented in Fig. 1A. For the BSH-CP transfer, 13 C RF ower was alied during 11.4 ms at 58 m and ramed down to 80% from an initial RF strength of 12.8 khz. Maximum acquisition times were 19.8 ms and ms for the direct and the indirect dimensions, resectively. Prior to Fourier-transformation the data were aodized in both dimensions with a 45 -shifted squared sine bell window function. The sectrum demonstrates excellent signal to noise and resolution. Also the fact that it exhibits cross eaks from the edges of the CA and CO bands, demonstrates the broad recouling bandwidth. 5. Numerical simulations In order to investigate the dynamics of magnetization transfer during BSH-CP at variations of 13 C isotroic chemical shift and RF field strength we erformed SIMPSON [27] simulations for a sin system comrising a directly bound CO and CA air with an internuclear distance of 1.53 Å. This corresonds to a diolar couling of 2121 Hz. For comarison with the exerimental data, 20 khz MAS and an external magnetic field corresonding to 850 MHz roton Larmor frequency were used. Other arameters of the system and calculations are given in the SI (sin system section of SIMPSON inut file). Fig. 5 shows the CA magnetization build-u as a function of contact time. The initial magnetization of CO is calibrated to 1. The simulation describes the ulse sequence of Fig. 1A with the 13 C RF field ramed down to 80%. The curves are color coded according to maximal RF amlitudes, as given in the figures. The to anel shows simulation results for CA and CO sins shifted by 1 m and 120 m from the carrier frequency at 0 m, while the bottom anel shows results for resonances at 16 m and 125 m. More simulation results can be found in the Suorting information, e.g. for other sets of chemical shifts, as well as for a 13 C RF field ramed down to 70% during BSH-CP. In summary, the simulations show that BSH-CP is robust with resect to robe inhomogeneity and chemical shift disersion. The comarison to the exerimentally obtained data (see Fig. 2) reveals similar magnetization transfer dynamics, in articular regarding the build-u time. 6. Conclusions We could show that band-selective homonuclear CP (BSH-CP) can be alied for efficient olarization transfer between CO and CA regions in highly deuterated roteins at moderate MAS rates and high magnetic fields. The most efficient recouling is achieved when the sum of effective RF fields on CA and CO resonances is equal to two times the sinning rate. This method can be directly imlemented in roton-detected variants of inter-residual correlation exeriments without the requirement of ultra-fast MAS. Acknowledgments This work was suorted by the Max Planck Society, the DFG (Emmy Noether Fellowshi to A.L.), and the Euroean Union Seventh Framework Program under Grant Agreement (Bio- NMR). Aendix A. Sulementary material Sulementary data associated with this article can be found, in the online version, at htt://dx.doi.org/ /j.jmr Fig. 5. CA magnetization dynamics during BSH-CP at different RF field strengths, as indicated in the figure. The to anel reresents simulation results for CA and CO sins shifted by 1 m and 120 m from the RF carrier frequency at 0 m, the bottom anel shows results for CA and CO resonances at 16 m and 125 m, resectively. References [1] A. McDermott, Structure and dynamics of membrane roteins by magic angle sinning solid-state NMR, Annu. Rev. Biohys. 38 (2009)

7 Communication / Journal of Magnetic Resonance 230 (2013) [2] C. Wasmer, A. Lange, H. Van Melckebeke, A.B. Siemer, R. Riek, B.H. Meier, Amyloid fibrils of the HET-s( ) rion form a beta solenoid with a triangular hydrohobic core, Science 319 (2008) [3] A. Loquet, N.G. Sgourakis, R. Guta, K. Giller, D. Riedel, C. Goosmann, C. Griesinger, M. Kolbe, D. Baker, S. Becker, A. Lange, Atomic model of the tye III secretion system needle, Nature 486 (2012) [4] V. Chevelkov, U. Fink, B. Reif, Quantitative analysis of backbone motion in roteins using MAS solid-state NMR sectroscoy, J. Biomol. NMR 45 (2009) [5] P. Schanda, B.H. Meier, M. Ernst, Quantitative analysis of rotein backbone dynamics in microcrystalline ubiquitin by solid-state NMR sectroscoy, J. Am. Chem. Soc. 132 (2010) [6] V. Chevelkov, K. Rehbein, A. Diehl, B. Reif, Ultrahigh resolution in roton solidstate NMR sectroscoy at high levels of deuteration, Angew. Chem., Int. Ed. 45 (2006) [7] V. Chevelkov, B.J.V. Rossum, F. Castellani, K. Rehbein, A. Diehl, M. Hohwy, S. Steuernagel, F. Engelke, H. Oschkinat, B. Reif, H-1 detection in MAS solid-state NMR sectroscoy of biomacromolecules emloying ulsed field gradients for residual solvent suression, J. Am. Chem. Soc. 125 (2003) [8] E.K. Paulson, C.R. Morcombe, V. Gaonenko, B. Dancheck, R.A. Byrd, K.W. Zilm, Sensitive high resolution inverse detection NMR sectroscoy of roteins in the solid state, J. Am. Chem. Soc. 125 (2003) [9] D.H. Zhou, J.J. Shea, A.J. Nieuwkoo, W.T. Franks, B.J. Wylie, C. Mullen, D. Sandoz, C.M. Rienstra, Solid-rate rotein-structure determination with rotondetected trile-resonance 3D magic-angle-sinning NMR sectroscoy, Angew. Chem., Int. Ed. 46 (2007) [10] M. Huber, S. Hiller, P. Schanda, M. Ernst, A. Bockmann, R. Verel, B.H. Meier, A roton-detected 4D solid-state NMR exeriment for rotein structure determination, ChemPhysChem 12 (2011) [11] R. Linser, B. Bardiaux, V. Higman, U. Fink, B. Reif, Structure calculation from unambiguous long-range amide and methyl (1)H (1)H distance restraints for a microcrystalline rotein with MAS solid-state NMR sectroscoy, J. Am. Chem. Soc. 133 (2011) [12] D.H. Zhou, A.J. Nieuwkoo, D.A. Berthold, G. Comellas, L.J. Serling, M. Tang, G. Shah, E.J. Brea, L.R. Lemkau, C.M. Riensta, Solid-state NMR analysis of membrane roteins and rotein aggregates by roton detected sectroscoy, J. Biomol. NMR 54 (2012) [13] R. Linser, U. Fink, B. Reif, Proton-detected scalar couling based assignment strategies in MAS solid-state NMR sectroscoy alied to erdeuterated roteins, J. Magn. Reson. 193 (2008) [14] M.J. Knight, A.L. Webber, A.J. Pell, P. Guerry, E. Barbet-Massin, I. Bertini, I.C. Felli, L. Gonnelli, R. Pierattelli, L. Emsley, A. Lesage, T. Herrmann, G. Pintacuda, Fast resonance assignment and fold determination of human sueroxide dismutase by high-resolution roton-detected solid-state MAS NMR sectroscoy, Angew. Chem., Int. Ed. 50 (2011) [15] A. Abragam, Princiles of Nuclear Magnetism, At the Clarendon Press, [16] R.E. London, A theoretical evaluation of the significance of scalar relaxation in couled sin-1/2 systems in macromolecules, J. Magn. Reson. 86 (1990) [17] M. Huber, O. With, P. Schanda, R. Verel, M. Ernst, B.H. Meier, A sulementary coil for H-2 decouling with commercial HCN MAS robes, J. Magn. Reson. 214 (2012) [18] R. Verel, M. Ernst, B.H. Meier, Adiabatic diolar recouling in solid-state NMR: the DREAM scheme, J. Magn. Reson. 150 (2001) [19] J.-P. Demers, V. Chevelkov, A. Lange, Progress in correlation sectroscoy at ultra-fast magic-angle sinning: basic building blocks and comlex exeriments for the study of rotein structure and dynamics, Solid State Nucl. Magn. Reson. 40 (2011) [20] A. Loquet, G. Lv, K. Giller, S. Becker, A. Lange, C-13 sin dilution for simlified and comlete solid-state NMR resonance assignment of insoluble biological assemblies, J. Am. Chem. Soc. 133 (2011) [21] N.C. Nielsen, H. Bildsoe, H.J. Jakobsen, M.H. Levitt, Double-quantum homonuclear rotary resonance: efficient diolar recovery in magic-angle sinning nuclear magnetic resonance, J. Chem. Phys. 101 (1994) [22] V. Ladizhansky, R.G. Griffin, Band-selective carbonyl to alihatic side chain C- 13 C-13 distance measurements in U-C-13, N-15-labeled solid etides by magic angle sinning NMR, J. Am. Chem. Soc. 126 (2004) [23] K. Takegoshi, K. Nomura, T. Terao, Rotational resonance in the tilted rotatingframe, Chem. Phys. Lett. 232 (1995) [24] K. Takegoshi, K. Nomura, T. Terao, Selective homonuclear olarization transfer in the tilted rotating frame under magic angle sinning in solids, J. Magn. Reson. 127 (1997) [25] A.J. Shaka, J. Keeler, T. Frenkiel, R. Freeman, An imroved sequence for broadband decouling WALTZ-16, J. Magn. Reson. 52 (1983) [26] J. Tegenfeldt, U. Haeberlen, Cross olarization in solids with fli-back of I-sin magnetization, J. Magn. Reson. 36 (1979) [27] M. Bak, J.T. Rasmussen, N.C. Nielsen, SIMPSON: a general simulation rogram for solid-state NMR sectroscoy, J. Magn. Reson. 147 (2000)