|Home | About | Journals | Submit | Contact Us | Français|
Pulsed Electron Double Resonance (PELDOR)/Double Electron-Electron Resonance (DEER) spectroscopy is a very powerful structural biology tool in which the dipolar coupling between two unpaired electron spins (site-directed nitroxide spin labels) is measured. These measurements are typically conducted at X-band (9.4 GHz) microwave excitation using the 4-pulse DEER sequence and can often require up to 12+ hours of signal averaging for biological samples (depending upon spin label concentration). In this rapid report, we present for the first time, a substantial increase in DEER sensitivity obtained by collecting DEER spectra at Q-band (34 GHz), when compared to X-band. The huge boost in sensitivity (factor of 13) demonstrated at Q-band represents 169-fold decrease in data-collection time, reveals greatly improved frequency spectrum, higher quality distance data, and significantly increases sample throughput. Thus, the availability of Q-band DEER spectroscopy should have a major impact on structural biology studies using site-directed spin labeling EPR techniques.
Pulsed Electron Double Resonance (PELDOR)/Double Electron-Electron Resonance (DEER) spectroscopy is a rapidly emerging, powerful structural biology technique in which the dipolar coupling between two unpaired electron spins (usually site-directed nitroxide spin labels) is measured(1-3). The strength of the dipolar coupling can then be used to determine the distance between the two spins in the range of 2-8 nm(4-9). This allows researchers to gain valuable structural information from samples in which other techniques like solution NMR or X-ray crystallography prove difficult or impossible(10-12). These measurements are typically conducted with X-band (9.4 GHz) microwave excitation using the 4-pulse DEER sequence and can often require up to 12+ hours of signal averaging (depending upon concentration) for typical biological samples. In this communication, we report for the first time a substantial increase in sensitivity, that is obtained by collecting DEER spectra using a pulse Q-band (34 GHz) EPR spectrometer. The huge boost in sensitivity at Q-band reveals higher quality data and significantly reduces data acquisition time.
For this study, an α-helical coiled coil peptide with a Leucine Zipper (LZ) motif (residues 245-281) of the yeast transcriptional activator GCN4(13, 14) (PDB entry 1YSA) was used as a model. The peptide was synthesized on a solid-state peptide synthesizer with a single TOAC nitroxide spin label at position 248 with Gln→TOAC substitution for the distance measurements between the two monomers. The TOAC(15-19) spin label was chosen over the traditional MTSL (1, 14) to eliminate disproportionation of the two label disulfide bonds to form a linked peptide dimer, and to avoid the motional flexibility of the MTSL nitroxide group for more accurate inter-coil backbone distance measurements. In previously published results, the distance between two spin labels as determined via DEER studies matches the predicted distance from the X-ray crystal structure (13, 20).
Samples for DEER measurements were prepared at a concentration of 200 μM peptide in phosphate buffer consisting of 40 mM potassium phosphate, 50 mM NaCl and 30% sucrose which was added as a cryoprotectant. at pH 7.0. The data were collected at the Ohio Advanced EPR Laboratory using the newly installed Bruker ELEXSYS E580 spectrometer equipped with a SuperQ-FT pulse Q-band system and EN5107D2 resonator. The electron spin echo-detected EPR spectrum of the doubly labeled GCN4 peptide is shown in Figure 1. The high signal-to-noise was achieved in a single scan with 200 averaged echoes per data point (100 each with a 2-step phase cycling).
Figure 2 shows the DEER data sets collected with 10, 100, and 1000 scans at both X- and Q-bands scaled for noise level comparisons. It is immediately obvious that the data collected at Q-band shows a much higher signal-to-noise ratio. The signal enhancement at Q-band was measured to be approximately 13 fold. The S/N boost is quite remarkable considering that the greatly reduced sample volumes used in the Q-band experiments (5 μL vs 20 μL for X-band). We have observed comparable S/N enhancements on MTSL proteins at Q-band experiments (data not shown). The enhanced sensitivity observed at Q-band stems from a increase in the Boltzmann population difference between the spin states at the higher Zeeman energy, which contributes a factor of about four, and increased sensitivity of the instrument/resonator at the higher frequency (21, 22), (Peter Höfer, personal communication).
Figure 3 shows Pake patterns in the frequency domain and the corresponding distance distributions obtained at X-band (Figs. 3A and 3B) and Q-band (Figs. 3C and 3D). The increased sensitivity in the time domain translates directly into an improvement in the frequency spectrum and a more accurate distance distribution plot. The spectra were analyzed using Matlab and the DeerAnalysis2008 package provided by the Jeschke laboratory(23, 24). For the Q-band data, the complete Pake pattern is very well resolved; in addition, small inflections corresponding to different subpopulations of the distance distribution are clearly resolved above the noise level. The main peak in the Q-band distance distribution is better resolved (full-width at half-height, 1.85 nm), yielding a distance of 2.32 nm. In contrast, a relatively broader peak (full-width at half-height, 2.55 nm), corresponding to a distance of 2.27 nm is observed at X-band.
The improved sensitivity observed at Q-band will enable low frequency components to be more easily detected out of the noise level. Thus, DEER data can be collected for longer pulse delay times, enabling measurement of weaker dipolar couplings (longer distances), resulting in better L-curves for more accurate distance measurements.
Over the last few years, one of the biggest technological breakthroughs in solution NMR spectroscopy and structural biology has been the development of the cryoprobe technology, which increases the S/N in NMR spectra by 3 to 4 depending on the amount of the salt content (25). This has significantly increased the productivity in solution NMR labs by cutting down signal acquisition time and increased sample throughput by 9 to 16 fold. In a much more dramatic fashion, the 13× boost in sensitivity for the Q-band DEER measurements shown in this work theoretically represents 169 fold decrease in data acquisition time. It is apparent from Figure 2 that the data collected with 1000 scans at X-band are comparable to the data collected with only 10 scans at Q-band. Thus, research labs conducting DEER experiments at Q-band will be able to analyze multiple biological samples in a single day, which will dramatically improve productivity. The remarkable boost in sensitivity at Q-band will also enable researchers to conduct experiments at much smaller, more biologically relevant sample concentrations.
In conclusion, we have clearly demonstrated a significant boost in the sensitivity of DEER spectroscopy at Q-band when compared to X-band. Considering the overall improvement in the S/N ratio, sample volumes requirement (≈ 4-5 μL at Q-band vs ≈ 20 μL at X-band), the signal averaging time, and ease of data analysis, conducting DEER experiments at Q-band is much more advantageous for measuring SL-SL distances in proteins/peptides than the X-band counterpart. The availability of Q-band DEER can thus be expected to have a major impact on structural biology studies based on site-directed spin labeling.
This work is supported by NIGMS/NIH (GM60259-01), Army CDMRP grant W81XWH-06-1-0551 and the NSF (CHE-0645709 and MRI-0722403). We would also like to thank Ralph Weber and Peter Höfer from Bruker Biospin for critical discussion.