All experiments were carried out using a Bruker ELEXSYS E580 spectrometer operating at X-band with a dielectric ring resonator 4118X-MD4 and a Bruker 400U microwave source unit. All measurements were made at 50 K with the sample in a frozen glassy state. The resonator was over-coupled giving a Q factor of approximately 100. The video bandwidth

was set to 20 MHz. Experiments to determine the phase memory time (Tm) were performed by measuring the intensity of a Hahn echo as it decayed with increasing inter-pulse delay. The pulse sequence used was π/2–t1–π, where the π pulse was 32 ns and the initial time delay PD-0332991 manufacturer t1 was 400 ns, in addition two-step phase cycling was employed to eliminate receiver offsets. Timings and delays were used appropriately for each sample. The experiment repetition time was 4 ms and 50 shots were taken at each time point. Echo decay curves in a deuterated medium are dominated, initially, by ESEEM oscillations and so Tm was estimated by fitting

of Eq. (1) to the tail end of the data that is largely free of ESEEM. Histone octamers used to make relaxation measurements were full length and as such contained unstructured see more tails that are not defined by X-ray crystallography. In order to calculate the sum(1/r3) values depicted in Fig. 3, the unstructured tails were built onto the crystal structure (PDB code 1TZY) using a simulated annealing protocol within Xplor-NIH [14]. High temperature dynamics with only the unstructured tail regions allowed to move freely, gave a large ensemble of structures. The position of the spin-label was determined my molecular dynamics as previously described [15]. Distances between the nitrogen atom (averaged position) of the spin label to the positions of any remaining (after

deuteration) proton positions were measured and their 1/r3 values were averaged and summed SPTBN5 to provide a single term describing the proton distribution around the unpaired electron and the protons of the protein. Echo decay curves were measured and Fig. 2 shows the complete echo decay curves for all protein constructs discussed here. Echo decay curves were fitted using a stretched exponential (Eq. (1)) [3]. At X-band the beginning of the decay curves are obscured by deuterium ESEEM signals and so fitting and extraction of Tm values was done using cropped decay curves ( Fig. S3). Line fitting was also hindered by the presence of a low frequency oscillation, derived from dipolar coupling, that was especially prominent in the fully deuterated sample [16]. The estimated Tm for the non-deuterated, octameric complex (in deuterated solvent) is 6.9 μs, which is at the high end of reported Tm values for a spin label situated on the surface of a protein dissolved in deuterated buffer [1]. Deuteration of H3 leads to an approximate doubling of the Tm to 13.6 μs.