Optimization of Pulsed DEER Measurements for Gd-Based Labels: Choice of Operational

Supplementary materials to:

Optimization of pulsed DEER measurements for Gd-based labels: choice of operational

frequencies, pulse durations and positions and temperature.

A. Raitsimringa,*, A. V. Astashkina, J.H.Enemarka, I. Kaminkerb, D. Goldfarbb , E. D. Walterc, Y. Songd and T. J. Meaded.

a Department of Chemistry and Biochemistry, University of Arizona, Tucson, Arizona 85721-0041,USA.

bDepartment of Chemical Physics, Weizmann Institute of Science, Rehovot, 76100, Israel.

cPacific Northwest National Laboratory,EMSL,3335 Q Ave,Richland, WA 99354, USA.

d Department of Chemistry; Department of Biochemistry, Cell Biology, and Molecular Biology; Neurobiology & Physiology; Department of Radiology, Northwestern University, 2145 Sheridan Road, Evanston, IL 60208,USA.

1. Schematic structures of compounds used in this work.

GdDOTA

3GD538 Double-labeled 2Gd (green balls)DNA

Details of synthesis of Gd-595, Gd-538, 3Gd-538 and double tagged DNA are published in:

[S1.1] Song, Y.; Kohlmeir, E.; Meade, T. J. Synthesis and Characterization of Multimeric Contrast Agents for Cell MR Imaging.J. Am. Chem. Soc.2008, 130, 6662–6663.

[S1.2] Song, Y.; Meade, T. J.; Astashkin, A. V.; Klein, E. L.; Enemark, J. H.; Raitsimring, A. Distance Measurements in Biomacromolecules Labeled by Gd3+ Markers Using Pulsed Dipolar Spectroscopy.J. Magn. Reson.2011, 210,59-68.

2. Shapes and integrals of refocused primary echogenerated by pulses of different durations.

Fig. S2(a): shapes of refocused primary echoes, V(t), calculated,using Eq.5 of text, for two different pulse durations.Fig S2(b):integrals of these shapes. As evident from Fig. S2 (b), when theintegrals are taken over wide limits,they are equalto each other,in accordance with Eq.6of text. The integralsbetween half-height pointsremain equal as well,as shown in Fig.S2 (c,d)

3. Examples of kinetics of phase relaxation and magnetization recovery for some Gd(III)-based tags.

Fig. S3_1. Phase relaxation kinetics of Gd538, (see structure above) collected in Ka-band (1 and 3) and W-band (2 and 4) at spectrum position “a” (1 and 2) and “b” (3 and 4), as shown in inset. Experimental conditions: two pulse ESE; Ka-band: mw frequency, 30.3 GHz, magnetic field: (1)-1.0845T and (3)- 1.0738 T, respectively. W-band: mw frequency, 94.9 GHz, magnetic field values: (2)- 3.4038 T, (4)-3.3938 T, respectively. Samples prepared in D2O/d8 glycerol solution (1:1, v:v);concentration of GdL = 30µM; Temperature of measurements-14K (Ka-band) and 10K (W-band). The durations of /2 and  pulse were 15ns and 30 ns, in W-band experiment. In Ka-band the primary echo was generated by pulses of equal duration of 20 ns.

Fig. S3_2. Ka-band. Logarithm of normalized magnetization recovery (inversion recovery technique),, presented in linear, (a) and non-linear, (b) time coordinate.Assignment of time coordinate axes for each curve indicated by arrow. The V(t) and are instant and asymptotic amplitude of echo signal, respectively.

Inversion pulse duration, 8 ns, observation pulses duration, 40 ns. The mw frequency: 35.35GHz, magnetic field: 1.2628T, corresponding tomaximum of the EPR spectrum. Temperature:7K. The sample: water-glycerol (1:1, v:v) solutions of Gd538 at concentration of ~ 50 µM.

Fig. S3_3. W-band (94.9GHz). Logarithm of normalized magnetization recovery,, presented in linear, (1) and non-linear, (2) time coordinate. Assignment of time coordinate axes for each curve indicated by arrow. The V(t) and are instant and asymptotic amplitude of echo signal, respectively. The data are collected for Gd-DOTA at maximum of EPR spectrum. Temperature: 10K.

4. Comparative example of long distance DEER measurements in W- and Ka-bands for DNA labeled by Gd(III) tag withlargeD.

The measurements were performed using available W-band instrument [D. Goldfarb,Ya. Lipkin,A. Potapov,Ye.Gorodetsky, B.Epel, A. Raitsimring, M. Radoul, I. Kaminker, J. Magn. Reson. 194, 8 (2008)] to determine accessible range in distance measurements for a worst case scenario, when Gd(III) tagsattached to macrobiomoleculehavelargeD. In the case presented below the tags were Gd538, for which D50mT. In DEER distance measurements the accessible distance, upper limit of concentration, time interval and flip probability (pumping pulse efficiency) are interrelated.

To obtain reliable information about the distance between spin labels, the time domain pattern must be recorded for at least a half-period of the dipolar modulation, t1/2,as defined by Eq. S1:

(1)

where the distance, r, is in Å. The time domain pattern inDEERis a product of the partial kinetics, Vir(t) and Via(t), due to inter- and intrapair dipolar interactions, respectively. While Via(t) contains the desired information about the intrapair distance, Vir(t) represents an unwanted contribution. For data acquisition performed in a limited time interval, the extraction of Via(t)from the collected time domain pattern V(t) is only reliable if the intrapair decay is comparable with or exceeds the interpair decay.The label concentration, [GdL], should not exceed the value given by Eq. 2:

[GdL](mM) = 1000/t1/2 (2)

wheret1/2 is defined by Eq.1. For a “benchmark” distance of ~100 Å, thetime base of measurements has to be, therefore,at least 10µs, and the concentration of the labels should not exceed ~0.1 mM.

Fig. S4_1 (left panel) normalized W-band DEER kinetics of double tagged 2Gd538_DNA. Time base - 11µs; accumulation time,-15hours at repetition rate of 2kHz; p=94.78 GHz; o=94.9 GHz; duration of  observation pulses, 50 ns; duration of  pumping pulse, 15 ns; pumping pulse applied at maximum of GdL EPR spectrum at 3.405 T; temperature of measurements, 10K; concentration of double labeled DNA ~60mM. (right panel): (1), W-band DEER kinetics of double tagged 2Gd538DNA due to intrapair interactions (derived from kinetics presented in left panel after removal of background due to interpair interaction). For comparison, normalized Ka-band kinetics (2) is also presented. For Ka-band, o=35.163 GHz; p=35.043 GHz, B=1.2629T, corresponding to maximum of EPR spectrum at p;time base – 7µs. Other parameters are the same as in W-band.

Reliable information on the distance distribution function between tags can be obtained when the ratio of the normalized DEER effect-to-noise exceeds ~10. Obviously, measurements haveto be completed in some realistic acquisition time. The general consensus is that the time spent on measurements should be in the “hours” range and not exceed ‘overnight’, that is ~ 10 hours. The experiment forwhich data is shown in Fig S4_1 was performed before the analysis presented in the paper. The time spent to collect the data presented in Fig. S4_1 was about 15 hours. Underoptimal conditions similar quality data can be obtained innot more than 3-4 hours. Note, that the normalized DEER effect in W-band, due to the decrease of the linewidth of the -½ ½ transition, is larger than in Ka-band.

5. Potential improvements in W-band measurements using broadband W-band instrument of EMSL PNNL [S4.2] with non-resonant cavity. Some instrument characteristics and examples.

a. Acquisition set up and justification of pulse parameters.

To eliminate adverse effects of phase instability, in the PNNL ESML instrument for each step of measurements,the “echo” coordinate is digitized (step of 0.5 ns) using 2 channels of theU1080AAcqiris 8-bit High-Speed cPCI Digitizer for I/Q outputs of quadrature detection. As an example, the recorded trace along the spin echo coordinate for one of the channels in the DEER experiment is shown in Fig. S5_1.

Fig. S5_1 Fast digitizer output from one of quadrature channels during ‘four pulse’ DEER experiment. ‘op’ designates observation pulses; pp-pumping pulse; ‘pe’, ‘se’, and ‘rpe’ designate primary, stimulated and refocused primary echo, respectively. DEER kinetics is a variation of refocused primary echo (the value of gated integral of echo signal) upon position (t’) of pumping pulse. Sample: 3Gd-538, temperature 6K;Nominal duration of observation pulses for maximal RPE signal: 6ns; 8ns; 8 ns. Pumping () pulse duration -8 ns; observation andpumping frequencies were 94.704 and 95.104 GHz, respectively. Inset depicts shapes of refocused primary echo from I/Q channels after phase correction.

Commonly, in pulsed EPR the effective duration of pulses differs from the nominal one due to departure of pulse shape from rectangular and leakage of mw power through switches used for pulse preparation. In this particular case the experimental line shape was well reproduced by simulation, as shown in Fig S4_2, when rectangular 4ns (/2); 8ns(); 8 ns ()pulseswere used.

Fig S5_2. Experimental (1)(presented in Fig.S4_1 as I) and simulated(2)) shapes of refocused primary echo. Spectra slightly shiftedin vertical direction relative tooneanother to avoid complete overlap. In simulation the pulses were: 4ns (/2); 8ns(); 8 ns () .

b. Demonstration of efficiency of short pumping pulse to increase normalized DEER effect (spin flip probability).

Figure S5_3.Normalized intra-pair W-band DEER time domain patterns for double tagged 2Gd538 - DNA, collected using instruments [S5.1] (1) and [S5.2] (2). For (1),p=94.9 GHz, |o-p|=120 MHz;  -pumping pulse duration-15ns;  observation pulses, 50 ns,T=10K. For (2) p=95.104GHz, o=94.704 GHz (|o-p|=400);  -pumping and observation pulses -8ns,/2 observation pulse-6ns; temperature,7K; In both casespis in resonance with maximum of the EPR spectrum. Concentration of labeled DNA ~60µM.Accumulation time: ~2 and 12 hours;pulse train repetition rate 2kHz and 500Hz for 1 and 2, respectively. Arrows mark magnitude of the spin flip probability, which evidently increases by about 3.5 times when a shorter pumping pulse is used.

[S5.1] D. Goldfarb,Ya.Lipkin,A. Potapov,Ye.Gorodetsky, B.Epel, A. Raitsimring, M. Radoul, I. Kaminker, J. Magn. Reson.194, 8 (2008).

[S5.2]