2D DARR spectra of CT fiber. The secondary structure components of the amino acid residues of CT in fiber form based on the chemical shifts are indicated by the colors: β-sheet in red, α-helix in blue and random coil in black.


METHODS
The isotopically labelled fibers were packed into 3.2 mm zirconia MAS rotors using a spiNpack  ultracentrifugal packing device (Giotto Biotech, Italy), which was then placed in a Beckman Coulter Optima L-100 XP with swinging-bucket SW 28 Ti ultracentrifuge rotor and the sample was first centrifuged at 141,000 x g over 1 h. About 1.5 mL of the buffer was used to rinse the original sample tube and centrifuged again before the rotor was sealed for MAS NMR measurements. To restrict the samples to the central third of the rotor, a thick-bottomed 3.2 mm rotor with a 12 uL Teflon spacer above the sample was used. Uniformly 13C-15N-labelled CT domain protein fiber (sample III), and a mixture of 13C-labeled CT and 15N-labeled CT protein fiber 1:1 (sample IV) were prepared. Experiments were performed on a 850 MHz Avance III HD spectrometer (Bruker Biospin) equipped with a 3.2 mm triple resonance MAS probe. All experiments involving labeled fibers were performed at a MAS spinning rate of 17.5 kHz with a sample temperature of 278 K unless stated otherwise. To gain access to intermolecular and intramolecular structural information of the fibers, a series of 13C-13C 2D DARR (Dipolar Assisted Rotational Resonance)79 NMR data were acquired with 25, 50, 100, 200, and 400 ms mixing times. The following spectral parameters were used to setup the experiments: 1H 90 deg pulse length of 3.0 us, 13C 90 deg pulse length of 3.5 us, 1H-13C CP was implemented using a linearly ramped RF amplitude centered around 65 kHz on the 13C channel while a constant RF field of 84 kHz on the 1H channel for 1.5 ms (-1 spinning side band), a recycle delay (RD) of 2 s and an applied Spinal-64 decoupling on 1H at 84 kHz. DARR spectra were recorded by acquiring 192 points in the t1 dimension with a dwell time of 11.0 us and 40 scans per free induction decay (FID).
For 15N assignments, both one-bond NCa and NCO, and multibond NCX correlation data to side chain carbons were acquired. The double-CP (DCP)80 experiment correlate 13C and 15N chemical shifts. During the initial CP transfer (1H to 15N) in NCO and NCA double CP experiments, a constant 1H RF amplitude of 59 kHz is applied along with a linear ramp from 41 and 46 kHz for 15N RF field over 0.4 ms (contact time). In the second CP step, the magnetization is transferred from 15N to 13C selectively by setting the carrier frequency to be on-resonance with either CO or Cα. For the NC transfer, from 15N to 13C, a constant RF amplitude of 44 kHz and 26.5 kHz are applied on 15N and 13C (-1 sideband), respectively, along with 1H CW decoupling at 70 kHz for 2.5 ms. A Spinal-64 decoupling on 1H at 83 kHz was applied during 13C acquisition of 24 points in the t1 dimension with a dwell time of 11.0 us and 800 scans per FID. To establish long distance intermolecular contacts between 13C and 15N, proton assisted insensitive nuclei (PAIN) CP experiments were performed on sample IV. PAIN-CP spectra were measured with a 4 ms mixing time, using RF fields of 45 and 44 kHz for 13C and 15N (2.5x spin rate 17.5 kHz), respectively, and 1H RF field at about 55 kHz.

13C NMR chemical shifts were calibrated externally to tetramethylsilane, TMS (iso=0 ppm) by adjusting the field to align the high frequency peak of adamantane at 38.48 ppm. This setting establishes the 15N chemical shift of powdered NH4Cl at 39.3 ppm relative to liq. NH3. The reported chemical shifts were corrected for the difference between 2,2-dimethyl-2-silapentane-5-sulphonic acid (DSS) and TMS by adding 2.67 ppm.  All data was processed using the TopSpin 3.5 software package. The 2D spectral analysis and peak assignments were performed with CCPNMR 2.5.2/3.2.082. All reported shifts were validated in the full set of recorded spectra. Assignments with apparent inconsistencies were excluded from the final set of reported assignments. Interpretations of chemical shifts in terms of secondary structure were based on the “Wishart” protein secondary structure parameters. 



FILE FORMAT
solid state NMR data (2D DARR spectra with mixing times between 50 ms and 400 ms, Bruker Topspin format).

