|Home | About | Journals | Submit | Contact Us | Français|
Spin-lattice relaxation times (T1s) of both small water-soluble spin labels in the aqueous phase as well as lipid-type spin labels in membranes increase when the microwave frequency increases from 2 to 35 GHz (Hyde et al., J. Phys. Chem. B 108  9524–9529). The T1 measured at W-band (94 GHz) for the water-soluble spin labels CTPO and TEMPONE (Froncisz et al., J. Magn. Reson. 193  297–304) is, however, shorter than when measured at Q-band (35 GHz). In this paper, the decreasing trends at W-band have been confirmed for commonly used lipid-type spin labels in model membranes. It is concluded that the longest values of T1 will generally be found at Q-band, noting that long values are advantageous for measurement of bimolecular collisions with oxygen. The contribution of dissolved molecular oxygen to the relaxation rate was found to be independent of microwave frequency up to 94 GHz for lipid-type spin labels in membranes. This contribution is expressed in terms of the oxygen transport parameter W = T1−1(Air) − T1−1(N2), which is a function of both concentration and translational diffusion of oxygen in the local environment of a spin label. The new capabilities in measurement of the oxygen transport parameter using saturation-recovery (SR) EPR at Q- and W-band have been demonstrated in saturated (DMPC) and unsaturated (POPC) lipid bilayer membranes with the use of stearic acid (n-SASL) and phosphatidylcholine (n-PC) spin labels, and compared with results obtained earlier at X-band. SR EPR spin-label oximetry at Q- and W-band has the potential to be a powerful tool for studying samples of small volume, ~30 nL. These benefits, together with other factors such as a higher resonator efficiency parameter and a new technique for canceling free induction decay signals, are discussed.
The first reports that oxygen can affect EPR spectra of organic free radicals in solutions via liquid-phase Heisenberg exchange are dated about 50 years ago [1–3]. Later, Povich pointed out that oxygen-broadening of the EPR spectra of spin labels can be used to measure dissolved oxygen concentration and diffusion in homogeneous solvents [4,5]. Backer et al.  were the first who applied spin-label oximetry to biological systems. Shortly after, the water-soluble spin probe 3-carbamoyl-2,2,5,5-tetramethyl-3-pyrroline-1-yloxyl (CTPO), which is widely used for spin-label oximetry, was introduced (see Ref.  for additional historical details). The first rigorous spin-label oximetric study was carried out by Windrem and Plachy , who reported in 1980 profiles of the oxygen diffusion-concentration product across the lipid bilayer membrane obtained from line-broadening measurements of lipid-soluble spin labels. These first applications were based on the oxygen-induced line-broadening of spin-label EPR spectra.
In the beginning of the 1980s, T1-sensitive spin-label oximetry methods were developed at the National Biomedical EPR Center at the Medical College of Wisconsin for measurement of the product of the oxygen diffusion constant and the oxygen concentration in membranes [9–11]. T1-sensitive methods have significant advantages over T2-sensitive methods because T1 (usually 1 to 10 µs) is from one to three orders of magnitude longer than T2. Additionally, T1-sensitive methods can be applied to any system that can be spin-probed or spin-labeled, without the need for a narrow EPR line or the presence of a resolved superhyperfine structure. T1-sensitive oximetry methods include saturation recovery (SR) (the absolute T1 method [11–13]), CW saturation [10, 14], passage display , and the multiquantum approach [16,17]. Spin-label oximetry can be used as a quantitative method because every collision of oxygen with a spin label contributes to a change in the EPR spectrum in both T1 and T2 sensitive methods [10,18–20].
The first oximetry measurements using either T1- or T2-sensitive methods were made at X-band. The development of loop-gap resonators (LGRs)  permitted oximetry measurements to be made at lower microwave frequencies for both analytical samples and small laboratory animals [22,23]. This direction developed into a significant branch of in vivo oximetry and imaging in which oxygen concentration and the distribution of oxygen in laboratory animals [24–26] and also in humans [24,27] are measured. These measurements are done at L- and S-band using T2-sensitive methods. This direction of research has also stimulated development of new spin labels including deuterated CTPO  and trityls , as well as other microscopic particulate probes, lithium phtalocyanine , coal-derived particles (fusinite) , and India ink .
The development of new LGRs [33,34] has allowed transfer of the T1-sensitive SR oximetry method to higher microwave frequencies, including Q- and W-band, which permits oximetry measurements to be made for very small water-containing samples (~30 nL). In our previous work , we reported oximetric data acquired using SR at frequencies from 2 to 35 GHz. Here we present oximetry measurements at Q- and W-band and discuss advantages of spin-label oximetry at these frequencies, especially for biological samples that can only be obtained in limited quantity. In two previous papers [35,36], we showed that (1) the T1 of water-soluble spin labels depends on microwave frequency (being largest at Q-band), and that (2) the effect of collisions between oxygen and spin-labels on the measured T1 value is independent of frequency up to 35 GHz. Here we extend SR measurements to W-band, showing also that the T1 of lipid-type spin labels in membranes exhibits a maximum value at Q-band.
Additionally, we showed that the measured collision rate between the nitroxide moiety and oxygen is the same from 2 to 94 GHz. Although the sensitivity in terms of number of spins required for good SR signals increases markedly with increase in microwave frequency, the concentration sensitivity of SR of spin labels has been found to be substantially independent of microwave frequency. It follows that the best frequency for spin-label oximetry in terms of sensitivity to bimolecular collisions with molecular oxygen is 35 GHz (Q-band). However, W-band has its own advantages over other frequencies, which are discussed.
One-palmitoyl-2-(n-doxylstearoyl)phosphatidylcholine spin labels (n-PC, n = 5, 12, or 16), tempocholine-1-palmitoyl-2-oleoylphosphatidic acid ester (T-PC), dimyristoylphosphatidylcholine (DMPC), and 1-palmitoyl-2-oleoylphosphatidylcholine (POPC) were obtained from Avanti Polar Lipids, Inc. (Alabaster, AL). The n-doxylstearic acid spin label (n-SASL, n= 5, 10, or 16) and the cholesterol analog, cholestane spin label (CSL), were purchased from Molecular Probes (Eugene, OR). Other chemicals, of at least reagent grade, were purchased from Sigma-Aldrich (St. Louis, MO).
The membranes used in this work were multilamellar dispersions of lipids (liposomes) containing 0.5 mol% of n-SASL or CSL (in DMPC) or containing 1 mol% n-PC or T-PC (in POPC), and were prepared as described previously . The buffer used for measurements with n-SASL was 0.1 M sodium borate at pH 9.5. To ensure that all carboxyl groups of SASL were ionized in DMPC membranes, a rather high pH was chosen [38,39]. For other measurements, 10 mM PIPES (piperazine-N,N′-bis(2-ethanesulfonic acid)) and 150 mM NaCl (pH 7.0) were used. The dense liposome dispersion (the loose pellet after centrifugation) with a final spin-label concentration of 1 to 3 mM was used. The structures of water-soluble and lipid-type spin labels used in this work and in our previous Q- and W-band SR measurements [35,36], together with structures of phospholipids, saturated DMPC, and unsaturated POPC, are shown in Figs. 1a and b.
For EPR measurements, samples were transferred either to a capillary made from the gas-permeable methylpenetene plastic known as TPX (i.d. 0.6 mm; X-band) or to a Teflon tube (i.d. 0.2 mm; Q-band). Quartz capillaries (i.d. 0.15 mm) were used at W-band. At X- and Q-band, samples were equilibrated directly in the resonator with nitrogen or a mixture of nitrogen and air adjusted with a flowmeter (Matheson Gas Products, model No. 7631 H-604) . The same gas was used for temperature control. At W-band, samples were equilibrated with either nitrogen or air at room temperature outside the resonator, transferred to a quartz capillary, and positioned in the resonator of the W-band spectrometer, which is equipped with a temperature-control system. At W-band, we also occasionally used the gas exchange method of removing oxygen, placing the sample into a Teflon sample tube and equilibrating it with the gas directly in the resonator. Results were indistinguishable.
Measurements were made using SR capabilities of X-, Q-, and W-band EPR spectrometers. Spectrometers and LGRs used for Q- and W-band measurements were described earlier [35,36]. For water-soluble spin labels, the duration of the microwave pulse was 0.3 µs. For lipid spin labels in membrane suspensions, the pulse duration was 0.3 to 1 µs. All recovery times acquired in these measurements were single exponential in character. For membrane spin labels, motion was sufficiently slow such that the nitrogen nuclear relaxation times were shorter than the electron T1 values, resulting in strong coupling of the three hyperfine lines. The statement that nuclear relaxation times are sufficiently short arises from the prediction of Yin et al.  that if this were not true, multi-exponential recovery times would be observed, contrary to the experiment. Experimental knowledge of nitrogen nuclear relaxation times under the conditions of this experiment does not exist to the best of our knowledge.
Typically, 106 decays were averaged, half of which were off-line and differenced for baseline correction, with 2048 (X- and Q-band) and 1024 (W-band) data points per decay. Sampling intervals depended on sample, temperature, oxygen tension, and microwave frequency, and were from 1 to 40 ns. Total accumulation time was about 2 to 5 min. Recovery curves were fitted by single, double, and triple exponentials, and compared. Results indicated that for all of the recovery curves obtained in this work, no substantial improvement in the fitting was observed when the number of exponentials was above one, establishing that recovery curves can be analyzed as single exponentials. Fits were based on a damped linear least-squares method, which utilizes the Gauss-Newton minimization procedure to which a scalar factor is included. This helps to “damp” the process toward the minimum and assures convergence in the iterative steps. The damped least-squares method has proven successful for fitting exponential decay curves  and EPR spectra . Figure 3 shows typical SR curves for 5-PC in POPC membranes from Q- and W-band measurements in the presence and absence of oxygen. Fits to single exponentials are excellent. Decay time constants were determined with accuracy better than +/−3%.
In Fig. 4, results obtained in the present work at W-band for lipid-type spin labels n-SASL and CSL in saturated DMPC membranes (see Fig. 1a for their structures) are combined with previously obtained measurements from the microwave frequency range 2.5 to 35 GHz . All spin-lattice relaxation times measured at W-band are shorter than when measured at Q-band. These data confirm again the remarkable fact that the values obtained at 94 GHz depart significantly from the trend observed using data obtained at lower frequencies. Plotted data result in smooth curves, lending credence to the reliability of the surprisingly short times obtained at 94 GHz. It can be concluded that the longest obtainable values for T1 will generally be found at Q-band.
Spin labels CTPO, TEMPONE, and n-SASL, together with saturated DMPC membranes, were used in our previous multifrequency measurements for “historical” reasons. Most earlier SR measurements were obtained from these spin labels in DMPC membranes [11,13,43], which provides reference data. However, the more biologically relevant phospholipid is unsaturated POPC, and the much better lipid-type spin-label analogues are the phospholipid analogues n-PC and T-PC, which can be obtained from Avanti Polar Lipids, Inc. (Alabaster, AL). Because of these circumstances, we show in Fig. 5 new data obtained at X-, Q-, and W-band for the phospholipid-type spin labels (5-, 16-, and T-PC) in unsaturated POPC membranes (see Fig. 1a for these structures). As expected, the longest T1 values were measured at Q-band. Values at X- and W-band were significantly shorter, confirming trends observed for other spin labels in water and membrane environments. The data of Figs. 4 and and55 indicate in all cases that T1 values are shorter at 94 GHz than at 35 GHz. These data indicate that the observed T1 dependence on microwave frequency is independent of the structure of the nitroxide moiety, the structure of the environment of the spin label, the polarity of the local nitroxide environment (which changes with the depth in the membrane), the rate and anisotropy of motion, and the temperature. This experimental finding complements our earlier paper , showing data only for CTPO and TEMPONE (see Fig. 1b for these structures). The break in the trend of relaxation rate versus microwave frequency (Figs. 4 and and5;5; see also Fig. 8 in Ref. ) requires further investigation. Hofbauer et al.  report unexpectedly short T1e values for 1 mM solution of TEMPO in a 10% glycerol-water mixture at W-band and suggest a mechanism involving dynamic modulation of the g-value.
Bimolecular collision rates with oxygen as a function of microwave frequency from 2.54 to 35 GHz using 16-SASL in DMPC bilayers at 27°C were reported by Hyde et al. . Data were expressed in terms of the oxygen transport parameter defined by Kusumi et al.  by Eq. (1):
where T1 values are the spin-lattice relaxation times of the nitroxide in samples equilibrated with atmospheric air and nitrogen, respectively. W(x) is proportional to the product of the local translational diffusion coefficient D(x) and the local concentration C(x) of oxygen at a “depth” x in the lipid bilayer that is in equilibrium with atmospheric oxygen.
Measurement of the oxygen transport parameter in the same system has been extended to 94 GHz, and all results are presented in Table 1. A new value of 3.02 µs−1 obtained at 94 GHz is in satisfactory agreement with values reported at lower frequencies. It is concluded that the effect of bimolecular collisions of oxygen on observed spin-label relaxation rates is independent of microwave frequency even at W-band, where spin-lattice relaxation times behave anomalously. To emphasize this independence, we added in Table 1 the values of T1s. These values change drastically with microwave frequency, increasing more than five times between 2.54 and 34.6 GHz before decreasing at 94 GHz. These results, together with the fact that the long values of the spin-lattice relaxation times are advantageous for measurement of bimolecular collisions with oxygen, indicate that the best frequency for spin-label oximetry measurements using SR is Q-band.
It is notable that for such a large change of microwave frequency (from 2 to 94 GHz), which causes a five-time increase in the spin-lattice relaxation times of lipid-type spin labels, the oxygen transport parameter measured for the same spin label at the same temperature is invariant of frequency. Thus, the measured spin-lattice relaxation rate (T1−1) can be related to the oxygen transport parameter, or expressed in frequency-independent oxygen transport parameter units (see Eq. (1)), which we call a frequency-independent ruler or a golden standard for T1 measurements.
One of the achievements of spin-label oximetry is the capability to acquire profiles of the oxygen transport parameter across lipid bilayer membranes. These profiles contain significant structural information that is sensed by the solubility and movement of the small hydrophobic molecule, molecular oxygen [11,43,45–47]. They also permit calculation of a significant membrane parameter—namely, the membrane oxygen permeability coefficient, which connects oxygen flux across the membrane with the difference in oxygen concentration in water on both sides of the bilayer [13,48,49]. In Fig. 6, we compared profiles of the oxygen transport parameter across the unsaturated POPC membrane obtained at X-, Q-, and W-band.
Based on the profiles obtained from X-, Q-, and W-band measurements (Fig. 6), values of the oxygen permeability coefficient across the POPC bilayer were evaluated using methods previously described . These values are listed in Table 2. They are independent of microwave frequency within an error of 30% (see Ref.  for the minimal and maximal evaluations of the oxygen permeability coefficient across lipid bilayer membranes). These data confirm our conclusion that the results of the oximetry measurements are independent of microwave frequency.
The W-band capabilities for studying oxygen transport within and across biological membranes at Q-band have been discussed previously . In the present paper, we demonstrate measurement of the oxygen transport parameter in model membranes as a function of the depth in the lipid bilayer (Fig. 6). Suspensions of POPC liposomes for Q-band measurements were prepared in the same way as for X-band measurements. The time of observation at Q-band was similar to that at X-band (~3 min) as was the signal-to-noise ratio (compare SR signals in Fig. 3a and b with those in Ref. ). An LGR, with a sample volume of 30 nL, was used for Q-band measurements [35,50], while the sample volume of the LGR used for X-band was 3 µL . Feasibility is therefore established for experiments on samples with limited sample volumes. For example, all profiles obtained at X-band for calf-  and pig-lens  lipid membranes were obtained for samples prepared from 50 eyes. Similarly, measurements for cortical and nuclear bovine-lens lipid membranes were based on samples prepared from 100 eyes . Because the sample volume at Q-band is 100 times smaller, reliable profiles can, in principle, be obtained based on samples prepared from one eye.
Some recommendations for handling, degassing, and equilibrating with gas-mixture samples of small size are given in Ref. . Yin and Hyde  showed that the use of high observing power in SR EPR experiments does not affect the ability to extract bimolecular collision rates and can increase the signal-to-noise ratio up to ten times. Thus, this approach can be used if the signal-to-noise ratio in samples prepared from biological materials is too low. Another way to increase the signal-to-noise ratio is to increase the repetition rate during data acquisition. This approach should be beneficial for measurements at microwave frequencies at which the spin-lattice relaxation time is shorter—for example, at W-band compared with Q-band.
The results of this paper also demonstrate the ability to make oximetry measurements at W-band. Saturation recovery “concentration sensitivity” has not previously been discussed in the literature. The general conclusion of the present work as well as in previous papers in this series is that SR concentration sensitivity is experimentally found to be independent of microwave frequency. This is one of the significant findings of the present study. However, when sensitivity issues are considered based on the number of spins measured, there is very significant increase in sensitivity as the microwave frequency increases. In the present study, the volume of sample was 20 times less at W-band compared with Q-band, but the SR signal intensity at a fixed accumulation time remained about the same.
A new technique for canceling free induction decay (FID) contamination was introduced for Q-band SR  and has been extended to W-band. In X-band, as practiced in this laboratory, the pump and observing power are derived from the same source, and signals are a superposition of FID and SR responses. The technique of 180° pump-phase modulation, which modulates the signal of the FID response but not that of the SR response, permits separation of the superimposed signals. In Q- and W-band spectrometers, the observing and pump powers are derived from separate synthesizers. The synthesizers are phase-locked, but the pump synthesizer is offset from the observe by a small frequency difference, typically 1 kHz. A benefit of this method is that the two frequencies are well within the spin packet width, but cannot drift with respect to each other. With the pump and observed frequencies separated by this difference, FID signals are effectively averaged away, leaving only the SR signal.
Other significant factors in the experiments presented here are the high resonator efficiency parameter Λ (the field produced by 1 Watt of incident power ) and the low resonator quality factor Q. The Λ is between 10 and 20 for Q- and W-band LGRs. The LGR with a water sample has a loaded Q factor of approximately 100. One significant consequence of high Λ and low resonator Q is that the dead time between the end of the pulse and the start of data collection can be reduced because there is less stored energy in the resonator when the pulse ends. Shortening of the dead time is crucial for the discrimination by oxygen transport method, which should allow discrimination of domains with high lipid exchange rates .
This work was supported by grants EY015526, EB002052, and EB001980 of the National Institutes of Health.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.