Article pubs.acs.org/ac
Complete Protocol for Slow-Spinning High-Resolution Magic-Angle Spinning NMR Analysis of Fragile Tissues Marion André,† Jean-Nicolas Dumez,† Lamya Rezig,‡ Laetitia Shintu,‡ Martial Piotto,§ and Stefano Caldarelli*,†,‡ †
Institut de Chimie des Substances Naturelles, CNRS UPR 2301, Avenue de la Terrasse, 91190 Gif sur Yvette, France Aix Marseille Université, Centrale Marseille, CNRS, iSm2 UMR 7313, 13397, Marseille, France § Bruker Biospin, 67166 Wissembourg, France ‡
S Supporting Information *
ABSTRACT: High-resolution magic-angle spinning (HR-MAS) nuclear magnetic resonance (NMR) is an essential tool to characterize a variety of semisolid systems, including biological tissues, with virtually no sample preparation. The “non-destructive” nature of NMR is typically compromised, however, by the extreme centrifugal forces experienced under conventional HR-MAS frequencies of several kilohertz. These features limit the usefulness of current HR-MAS approaches for fragile samples. Here, we introduce a full protocol for acquiring highquality HR-MAS NMR spectra of biological tissues at low spinning rates (down to a few hundred hertz). The protocol first consists of a carefully designed sample preparation, which yields spectra without significant spinning sidebands at low spinning frequency for several types of sample holders, including the standard disposable inserts classically used in HR-MAS NMR-based metabolomics. Suppression of broad spectral features is then achieved using a modified version of the recently introduced PROJECT experiment with added water suppression and rotor synchronization, which deposits limited power in the sample and which can be suitably rotor-synchronized at low spinning rates. The performance of the slow HR-MAS NMR procedure is demonstrated on conventional (liver tissue) and very delicate (fish eggs) samples, for which the slow-spinning conditions are shown to preserve the structural integrity and to minimize intercompartmental leaks of metabolites. Taken together, these results expand the applicability and reliability of HR-MAS NMR spectroscopy. These results have been obtained at 400 and 600 MHz and suggest that high-quality slow HR-MAS spectra can be expected at higher magnetic fields using the described protocol.
H
was observed in erythrocytes exposed to a 4 kHz MAS rotation.12 The benefits of performing HR-MAS analysis at moderate rates (slow HR-MAS) to preserve sample integrity have been long recognized for many years. Indeed, the centrifugal forces are reduced by 2 orders of magnitude when the spinning rate is decreased by a factor of 10. Several pulse sequences have been developed to suppress spinning sidebands in MAS spectra of solid samples, for which numerous overlapping SSB arise in slow MAS spectra. Some, such as the twodimensional (2D) phase-adjusted spinning sidebands (PASS)13 and the 2D phase-corrected magic angle turning (PHORMAT),14 can be extended to biological, soft-solid samples to produce essentially sideband-free isotropic spectra. These approaches have led, inter alia, to the successful analysis with PASS of intact organs at spinning speeds as low as 43 Hz with a spectral resolution comparable to or better than the resolution obtained with fast MAS.15 This methodology has been recently
igh-resolution magic-angle spinning (HR-MAS) nuclear magnetic resonance (NMR) spectroscopy has become a widely used analytical tool for the study of biological systems, which provides metabolic information on intact tissues with no need for sample extraction.1 Proton HR-MAS NMR has been successfully applied to analyze intact cells and tissues from organs such as brain, lung, kidney, heart, and muscle, with a focus on mobile molecules.2−9 Typically, NMR spectra of these heterogeneous systems are recorded at MAS frequencies of a few kilohertz to effectively suppress spinning sidebands (SSB), that can overlap with metabolite signals and alter peak intensities, and approach the resolution obtained for isotropic liquid samples. A serious problem associated with conventional HR-MAS protocols is the large centrifugal forces experienced in the sample during the rotation, which can reach 100 000g and destroy tissue structures or even individual cells. For example, using rotors with an outer diameter of 4 mm (O.D.) with an inner diameter (I.D.) of 3 mm, in a sample of lipid-laden adipocytes, ∼20% of the cells were damaged after 2 h of MAS rotation at 3.5 kHz,10 serious visible damage was observed in human prostate tissues after spinning at 5 kHz,11 and cell lysis © 2014 American Chemical Society
Received: July 26, 2014 Accepted: October 6, 2014 Published: October 6, 2014 10749
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Analytical Chemistry
Article
applied in conjunction with microcoils.16,17 More spectacularly, PHORMAT has allowed the acquisition of 2D spectra on excised rat liver at a spinning frequency of 1 Hz.18 However, the very long acquisition times involved (up to 6 h for the PHORMAT experiment) as well as sensitivity and resolution issues further limit the application of these methods to the analysis of intact biological tissues.19 Perhaps most importantly, the intensity of the isotropic peak is affected by the extent of the SSB pattern, which may vary from sample to sample, and adds an element of variability to the quantitative analysis. As an alternative to the use of solid-state NMR pulse sequences, we have investigated the experimental conditions that would lead to high-quality HR-MAS spectra on slow spinning samples with solution-state NMR pulse sequences.20 These studies are motivated by the expected source of SSB in HRMAS for the mobile metabolites. In biological samples, which contain many intracellular and intercellular structures, metabolites are distributed heterogeneously, and often close to a boundary of a material with a different susceptibility than the compartment containing the compound under observation. The susceptibility differences between the various compartments induce magnetic field gradients in the sample. As a result, the resonance shifts become space dependent, resulting in a line broadening and in SSB in slow HR-MAS NMR spectra. In addition, SSB may be caused by disruptions of the liquid isotropic nature due to the presence of air bubbles and/or on the occurrence of inhomogeneities of the radio frequency (RF) field. The introduction of sidebands due to wobbling of the rotor around the magic angle during spinning has also been suggested.21 In our recent experiments, we have shown that, optimizing the sample preparation, particularly avoiding air bubbles and using reduced sample volumes with optimized geometry, led to well-resolved low-sideband TOCSY and HSQC spectra on tissues using HR-MAS at low spinning rates.20 The use of a reduced rotor volume of 12 μL led to a decrease of both the number and intensity of SSB in slow HR-MAS spectra, since RF field inhomogeneities and magnetic susceptibility line broadening effects can be minimized. The introduction of air bubbles increased the number of SSB for larger chambers and highlighted the role of magnetic susceptibility distributions in the presence of SSB at low spinning speeds (500 Hz). This collection of observation opens the way to standardized slow HR-MAS NMR experiments. In this work, we introduce a robust approach for the slow HR-MAS analysis of two representative tissue samples and exploit it to investigate the effect of fast and slow MAS on sample integrity. Using an optimized preparation protocol, we obtain slow HR-MAS spectra comparable to those obtained at conventional, high spinning rates. Furthermore, we adapted the classic relaxation filter used to remove broader signals to slow HR-MAS NMR experiments. We illustrate how the proposed approach yields high-quality spectra and preserve sample integrity for very fragile objects, with the example of fish eggs. These results are demonstrated at two different magnetic fields and with two types of HR-MAS rotors, including standard biopsy holding systems.
Figure 1. Sample preparation protocol for slow high-resolution magicangle spinning (HR-MAS) analysis of soft tissues.
■
Figure 2. 1H HR-MAS spectra of a heifer liver sample spun at 400 Hz on a 600 MHz spectrometer: (a) water-presaturated CPMG and (b) PROJECT.
MATERIALS AND METHODS Samples. A box of lumpfish roe was purchased at the supermarket and kept at 4 °C until the experiments. Heifer liver was purchased fresh from the butcher shop and kept at −80 °C until analysis. Optimized Protocol for the Preparation of Soft Tissues. A six-step protocol is described in Figure 1, for two
different volume-holding systems: 12 μL rotors and disposable inserts. Investigation of the Effect of Magic-Angle Spinning on Lumpfish Eggs. Experiments were carried out with a 10750
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Analytical Chemistry
Article
Figure 3. Water-presaturated 1H HR-MAS PROJECT spectra of heifer liver samples prepared according to the optimized protocol in a 12-μL rotor spun at (a, b) 4 kHz and (c, d) 400 Hz and recorded at 600 MHz (panels a and c) or 400 MHz (panels b and d). Sidebands are indicated by asterisks (∗).
first bringing the rotor to a speed of ∼1000−1500 Hz before reaching the set speed. This procedure is clearly not adapted if the threshold spinning speed must be avoided; thus, we resorted to manual adjustment of the spinning speed for the 400 Hz experiments. Newer versions of the MAS unit software optimized for slow spinning do not present this problem. Sidebands were identified by comparing spectra recorded at 4 kHz and 400 Hz. If any ambiguity persisted, a third experiment was performed at 600 Hz in order to obtain better characterization of the SSB. The T2-weighted spectra were obtained using either the CPMG sequence from the Bruker pulse program library or the presaturated PROJECT sequence from Aguilar et al.;22 both were adapted to include a presaturation pulse for water suppression. One-dimensional (1D) water-presaturated proton spectra were recorded using a spectral width of 10 ppm for fish eggs and 18 ppm for heifer liver, 16K data points, 64 scans, a relaxation delay of 2.5 s, and acquisition times of 1.36 s for fish eggs and 0.75 s for heifer liver. The total echo time was 96 ms for both water-presaturated CPMG and PROJECT sequences and the interpulse delay was set equal to the rotor period, leading to a total acquisition time of ∼4 min. Post-processing of all of the spectra included Fourier transformation of the free induction decays (FIDs) after exponential multiplication with a factor of 0.30 Hz, followed by phase and baseline correction.
regular 4-mm rotor (volume of 96 μL) and a ZrO2 rotor equipped with a 12 μL Teflon insert. For the regular rotor, 5 or 6 eggs were rinsed with a solution of phosphate buffer (0.05M, pH 6.8), in H2O/D2O (75/25) and placed in the rotor. Five microliters (5 μL) of phosphate buffer were added before carefully closing the rotor. For rotors equipped with a 12-μL insert, one egg was rinsed with the phosphate buffer and placed in the rotor. Five microliters (5 μL) of phosphate buffer were added before the rotor was carefully closed. The samples were submitted either to the classic (4 kHz) or the slow HR-MAS rotation rate (400 Hz) for 15 min. The rotor was then opened and the eggs carefully removed to observe the effect of the rotation. At the end of the analysis, after 30 min of spinning, the egg was carefully removed from the rotor to observe its general aspect. The solution was kept in the rotor and the volume used was increased to 12 μL, using fresh buffer solution. The solution was subjected to HR-MAS NMR analysis at 4 kHz to evaluate its metabolic content after different spinning rates and determine an eventual leakage of metabolites due to sample spinning. HR-MAS NMR Spectroscopy. All HR-MAS NMR experiments were performed at room temperature on a Bruker Avance II spectrometer operating at a resonance frequency of 400.36 MHz and a Bruker Avance I spectrometer operating at a resonance frequency of 600.13 MHz. Both instruments are equipped with a 4 mm double resonance (1H, 13C) gradient HR-MAS probe. Experiments were conducted at either 400 or 4000 Hz MAS frequencies. The spinning rate was controlled using a commercial MAS speed controller with frequency stability better than ±3 Hz. For the 4000 Hz experiments, the automatic procedure available to us was employed. This involves
■
RESULTS AND DISCUSSION Relaxation Filters. The spin echo is perhaps the most important building block in HR-MAS NMR spectroscopy. It is used to suppress uninformative, broad spectral features, which
10751
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Analytical Chemistry
Article
Figure 4. Water-presaturated 1H HR-MAS PROJECT spectra of heifer liver samples prepared according to the optimized protocol in a disposable insert spun at (a, b) 4 kHz and (c, d) 400 Hz and recorded at 600 MHz (panels a and c) or 400 MHz (panels b and d). Sidebands are indicated by stars.
arise typically from macromolecules that are not commonly analyzed in metabolomics. One limitation of this relaxation filter is the echo modulation induced by scalar couplings. In the classical Carr−Purcell−Meiboom−Gill (CPMG) pulse sequence, this modulation is minimized if the condition J ≪ 1/τ is met, where τ is the interpulse delay and J is the magnitude of the scalar coupling.23,24 Short interpulse delays, repeated for several cycles sufficient to implement a filter of the correct length, are thus necessary and commonly used to refocus J-modulation. However, the resulting, large high-RF duty cycles cause sample heating that can affect tissue integrity. At relatively low spinning rates ( 1 ms). While this is beneficial in terms of power deposition in the sample, the CPMG sequence is then unsuitable to refocus the J-modulation properly. The spectrum of Figure 2a shows a 1D 1H spectrum obtained on a 600 MHz spectrometer by applying the water-presaturated CPMG sequence on a heifer liver sample spun at 400 Hz, using a rotor-synchronized interpulse delay of 2.5 ms. As expected, the presence of J-modulation leads to spectral distortions, which are well-illustrated by the apparition of negative signals and loss of intensity, thereby compromising data interpretation. Recently, the PROJECT sequence has been introduced in liquid-state NMR as an alternative to the CPMG sequence that suppresses J-modulation, even for long interpulse delays.22 However, the advantages of the PROJECT sequence have not been demonstrated yet for slow HR-MAS NMR. Figure 2b shows a 1D 1H spectrum obtained with the PROJECT sequence at a spinning rate of 400 Hz. The interferences observed on the CPMG spectrum are suppressed.
Figure 5. Photographs of a set of lumpfish eggs (a) before being subjected to MAS, (b) after 15 min at 4 kHz MAS, and (c) after 15 min at 400 Hz MAS.
In particular, the spectral regions at δ = 4.67 ppm, δ = 2.85 ppm, δ = 3.2 ppm show negative peaks in the CPMG spectrum but are totally refocused in the PROJECT spectrum. The PROJECT sequence thus provides for T2-filtered, slow HR-MAS spectra, which are essential for many applications. Sample Preparation Protocol. A robust sample preparation protocol is necessary for high-quality sideband-free spectra at low spinning rates on biological samples. Figure 1 shows a six-step protocol that builds upon observations made with SSBreduced slow-HRMAS experiments and introduces several new, essential steps.20,25,26 Figures 3 and 4 show the 1D 1H HR-MAS NMR spectra, obtained at 4000 Hz MAS, of a heifer liver sample contained either in a 12-μL rotor (Figures 3a and b) or in a disposable insert (Figures 4a and b). The spectrum consists of an envelope of broad signals that could arise either from macromolecules and from components with restricted molecular motions superimposed on well-resolved peaks from small molecules characterized by fast and isotropic mobility. Preparation of the same 10752
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Analytical Chemistry
Article
Figure 6. Water-presaturated 1H HR-MAS PROJECT spectra of an isolated lumpfish egg spun at (a) 4 kHz and (c) 400 Hz. Spectra shown in panels b and d correspond, respectively, to the residual solution in the rotor, enhanced by factors of (b) 4 and (d) 10 after 15 min of spinning at 4 kHz and 400 Hz. Photographs on the top left (inset, panel a) and top right (inset, panel c) show the aspect of the egg after fast and slow MAS, respectively.
Earlier studies have also shown that disruptions of the internal structure could cause the transfer of metabolites between the tissue and the solution. Freeze−thaw cycles of the sample, as well as prolonged high-frequency spinning, have been shown to induce changes in HR-MAS NMR spectra that compromise data interpretation.29,30 Figures 6a and 6c display the spectra of a single lumpfish egg analyzed at 4 kHz and 400 Hz, respectively. The two spectra are very similar and both of them display satisfactory intensity and resolution. The absence of spinning sidebands at 400 Hz MAS is likely due to the natural geometry of the sample (sphere), the minimized effect of B1 inhomogeneities for small-volume rotors and the isotropic composition of the egg close to liquid. Figures 6b and 6d show the spectra of the solution obtained after removing the egg that underwent MAS, and topping up the solution with fresh buffer to reach a volume of 12 μL. In Figure 6b, the spectrum acquired on the solution after the 4 kHz HR-MAS shows a metabolic profile very similar to that of the entire egg (Figure 6a). This suggests an important leakage of the inner content of the egg into the remaining solution at high spinning rates, because of the disruption of the surrounding membrane. The number and the intensity of the metabolites signals is much lower in the residual solution after slow HR-MAS, as shown in Figure 6d. The low intensity signals between 3.2 and 4.0 ppm could arise from the porous character of the membrane of the egg or from components external to the membrane as they can also be identified in the washing solution of the lumpfish row (see Figure S2 in the Supporting Information). These results clearly demonstrate the high potential of using slow MAS for HR-MAS studies on fragile samples as the integrity of the sample is preserved and the transfer of metabolites from the sample studied to the solution is minimized.
sample according to the optimized protocol leads to spectacular results at the low spinning rate of 400 Hz, as illustrated in Figures 3c and 3d for the 12-μL rotor and Figures 4c and 4d for the disposable insert. In particular, the centrifugation steps before and after addition of the phosphate buffer solution contribute to an optimal sample geometry in the rotor and help remove air bubbles from the tissue to get a relatively air-free, homogeneous sample. Note that the centrifugal forces induced by the centrifugation steps are of the order of 3000 g, which corresponds to the forces experienced by the sample during slow spinning. Faster centrifugation speeds result in the partial destruction of the sample (see Table S1 in the Supporting Information). Using the protocol proposed here, slow HR-MAS spectra are almost devoid of SSB and are comparable in both resolution and intensity to those obtained at higher spinning frequencies (4000 Hz). The different geometry of the two experimental setups is the likely cause for the differences observed in the SSB patterns. Importantly, disposable inserts are the standard tools in metabolomics applications of HR-MAS NMR spectroscopy. Magic Angle Spinning of a Fragile Sample. In order to explore the influence of the spinning frequency on the integrity of fragile samples, we recorded fast and slow HR-MAS spectra of fish eggs. Recent studies have illustrated the deleterious effects of high-frequency MAS on the morphology of human glioblastoma tumors27 or fish eggs.28 In this latter case, the globular and plump eggs became shriveled and bleached, because of the rupture of the egg membrane and the dehydration of egg cells. Figure 5 shows photographs of a set of lumpfish eggs before and after 15 min of MAS at either 4 kHz or 400 Hz. As expected, the eggs that have been subjected to faster MAS are characterized by severe structural damages due to intense centrifugal forces (Figure 5b). In contrast, the eggs submitted to slow MAS are still intact after 15 min of spinning at 400 Hz, where centrifugal forces remain below 1000 g (Figure 5c). These results show that decreasing centrifugal forces by 2 orders of magnitude could make it possible to perform HR-MAS experiments on intact fragile biological samples.
■
CONCLUSION
In summary, we have introduced a robust, metabolomics-ready, approach for the characterization of fragile biological objects with slow high-resolution magic-angle spinning (HR-MAS) and demonstrated its potential for sample-preserving analyses. 10753
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Analytical Chemistry
Article
(14) Hu, J. Z.; Wang, W.; Liu, F.; Solum, M. S.; Alderman, D. W.; Pugmire, R. J.; Grant, D. M. J. Magn. Reson., Ser. A 1995, 113, 210− 222. (15) Hu, J. Z.; Rommereim, D. N.; Minard, K. R.; Woodstock, A.; Harrer, B. J.; Wind, R. A.; Phipps, R. P.; Sime, P. J. Toxicol. Mech. Methods 2008, 18, 385−398. (16) Wong, A.; Aguiar, P. M.; Sakellariou, D. Magn. Reson. Med. 2010, 63, 269−274. (17) Wong, A.; Li, X.; Molin, L.; Solari, F.; Elena-Herrmann, B.; Sakellariou, D. Anal. Chem. 2014, 86, 6064−6070. (18) Hu, J. Z.; Rommereim, D. N.; Wind, R. A. Magn. Reson. Med. 2002, 47, 829−836. (19) Wind, R. A.; Hu, J. Z. Prog. Nucl. Magn. Reson. Spectrosc. 2006, 49, 207−259. (20) Renault, M.; Shintu, L.; Piotto, M.; Caldarelli, S. Sci. Rep. 2013, 3, 5. (21) Avni, R.; Mangoubi, O.; Bhattacharyya, R.; Degani, H.; Frydman, L. J. Magn. Reson. 2009, 199, 1−9. (22) Aguilar, J. A.; Nilsson, M.; Bodenhausen, G.; Morris, G. A. Chem. Commun. 2012, 48, 811−813. (23) Carr, H. Y.; Purcell, E. M. Phys. Rev. 1954, 94, 630−638. (24) Meiboom, S.; Gill, D. Rev. Sci. Instrum. 1958, 29, 688−691. (25) Beckonert, O.; Coen, M.; Keun, H. C.; Wang, Y.; Ebbels, T. M. D.; Holmes, E.; Lindon, J. C.; Nicholson, J. K. Nat. Protoc. 2010, 5, 1019−1032. (26) Wind, R. A.; Hu, J. Z.; Rommereim, D. N. Magn. Reson. Med. 2001, 46, 213−218. (27) Martínez-Bisbal; Esteve; Martínez-Granados, B.; Celda. J. Biomed. Biotechnol. 2011, 2011, 8. (28) Cai, H. H.; Chen, Y. S.; Cui, X. H.; Cai, S. H.; Chen, Z. PLoS One 2014, 9 (1), e86422. (29) Bourne, R.; Dzendrowskyj, T.; Mountford, C. NMR Biomed. 2003, 16, 96−101. (30) Opstad, K. S.; Bell, B. A.; Griffiths, J. R.; Howe, F. A. NMR Biomed. 2008, 21, 1138−1147.
Several challenges had to be overcome to operate at lower magic-angle spinning (MAS) frequencies. We have proposed an optimized sample preparation protocol that leads to strongly reduced spinning sideband (SSB) spectra at 400 Hz MAS, using either a 12-μL rotor or a disposable 30-μL insert. Both signal intensity and resolution are comparable to those obtained at higher spinning rates. The classical CPMG sequence was replaced by the PROJECT sequence in order to obtain in-phase, T2-filtered spectra. By decreasing the spinning rate by 1 order of magnitude, centrifugal forces are reduced from 100 000 g at 4 kHz to 1000 g at 400 Hz. These results demonstrate that slow HR-MAS constitutes a suitable analytical tool for the characterization of fragile samples. Development of protocols including two-dimensional (2D) NMR techniques can be readily envisioned and are underway in our laboratory.
■
ASSOCIATED CONTENT
S Supporting Information *
The authors declare no competing financial interests. This material is available free of charge via the Internet at http:// pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mails: [email protected], [email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank Christina Sizun for experimental help, as well as Institut de Chimie des Substances Naturelles (ICSN), Bruker, and ANR (No. ANR-2011-JS08-014-01) for funding.
■
REFERENCES
(1) Lindon, J. C.; Beckonert, O. P.; Holmes, E.; Nicholson, J. K. Prog. Nucl. Magn. Reson. Spectrosc. 2009, 55, 79−100. (2) Cheng, L. L.; Ma, M. J.; Becerra, L.; Ptak, T.; Tracey, I.; Lackner, A.; Gonzalez, R. G. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 6408−6413. (3) Tate, A. R.; Foxall, P. J. D.; Holmes, E.; Moka, D.; Spraul, M.; Nicholson, J. K.; Lindon, J. C. NMR Biomed. 2000, 13, 64−71. (4) Bollard, M. E.; Murray, A. J.; Clarke, K.; Nicholson, J. K.; Griffin, J. L. FEBS Lett. 2003, 553, 73−78. (5) Righi, V.; Mucci, A.; Schenetti, L.; Tosi, M. R.; Grigioni, W. F.; Corti, B.; Bertaccini, A.; Franceschelli, A.; Sanguedolce, F.; Schiavina, R.; Martorana, G.; Tugnoli, V. Anticancer Res. 2007, 27, 3195−3204. (6) Chen, J. H.; Wu, Y. V.; DeCarolis, P.; O’Connor, R.; Somberg, C. J.; Singer, S. Magn. Reson. Med. 2008, 59, 1221−1224. (7) Bollard, M. E.; Garrod, S.; Holmes, E.; Lindon, J. C.; Humpfer, E.; Spraul, M.; Nicholson, J. K. Magn. Reson. Med. 2000, 44, 201−207. (8) Garrod, S.; Humpfer, E.; Spraul, M.; Connor, S. C.; Polley, S.; Connelly, J.; Lindon, J. C.; Nicholson, J. K.; Holmes, E. Magn. Reson. Med. 1999, 41, 1108−1118. (9) Garrod, S.; Humpher, E.; Connor, S. C.; Connelly, J. C.; Spraul, M.; Nicholson, J. K.; Holmes, E. Magn. Reson. Med. 2001, 45, 781− 790. (10) Weybright, P.; Millis, K.; Campbell, N.; Cory, D. G.; Singer, S. Magn. Reson. Med. 1998, 39, 337−345. (11) Taylor, J. L.; Wu, C.-L.; Cory, D.; Gonzalez, R. G.; Bielecki, A.; Cheng, L. L. Magn. Reson. Med. 2003, 50, 627−632. (12) Aime, S.; Bruno, E.; Cabella, C.; Colombatto, S.; Digilio, G.; Mainero, V. Magn. Reson. Med. 2005, 54, 1547−1552. (13) Antzutkin, O. N.; Shekar, S. C.; Levitt, M. H. J. Magn. Reson., Ser. A 1995, 115, 7−19. 10754
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Complete Protocol for Slow-Spinning High-Resolution Magic-Angle Spinning NMR Analysis of Fragile Tissues Marion André,† Jean-Nicolas Dumez,† Lamya Rezig,‡ Laetitia Shintu,‡ Martial Piotto,§ and Stefano Caldarelli*,†,‡ †
Institut de Chimie des Substances Naturelles, CNRS UPR 2301, Avenue de la Terrasse, 91190 Gif sur Yvette, France Aix Marseille Université, Centrale Marseille, CNRS, iSm2 UMR 7313, 13397, Marseille, France § Bruker Biospin, 67166 Wissembourg, France ‡
S Supporting Information *
ABSTRACT: High-resolution magic-angle spinning (HR-MAS) nuclear magnetic resonance (NMR) is an essential tool to characterize a variety of semisolid systems, including biological tissues, with virtually no sample preparation. The “non-destructive” nature of NMR is typically compromised, however, by the extreme centrifugal forces experienced under conventional HR-MAS frequencies of several kilohertz. These features limit the usefulness of current HR-MAS approaches for fragile samples. Here, we introduce a full protocol for acquiring highquality HR-MAS NMR spectra of biological tissues at low spinning rates (down to a few hundred hertz). The protocol first consists of a carefully designed sample preparation, which yields spectra without significant spinning sidebands at low spinning frequency for several types of sample holders, including the standard disposable inserts classically used in HR-MAS NMR-based metabolomics. Suppression of broad spectral features is then achieved using a modified version of the recently introduced PROJECT experiment with added water suppression and rotor synchronization, which deposits limited power in the sample and which can be suitably rotor-synchronized at low spinning rates. The performance of the slow HR-MAS NMR procedure is demonstrated on conventional (liver tissue) and very delicate (fish eggs) samples, for which the slow-spinning conditions are shown to preserve the structural integrity and to minimize intercompartmental leaks of metabolites. Taken together, these results expand the applicability and reliability of HR-MAS NMR spectroscopy. These results have been obtained at 400 and 600 MHz and suggest that high-quality slow HR-MAS spectra can be expected at higher magnetic fields using the described protocol.
H
was observed in erythrocytes exposed to a 4 kHz MAS rotation.12 The benefits of performing HR-MAS analysis at moderate rates (slow HR-MAS) to preserve sample integrity have been long recognized for many years. Indeed, the centrifugal forces are reduced by 2 orders of magnitude when the spinning rate is decreased by a factor of 10. Several pulse sequences have been developed to suppress spinning sidebands in MAS spectra of solid samples, for which numerous overlapping SSB arise in slow MAS spectra. Some, such as the twodimensional (2D) phase-adjusted spinning sidebands (PASS)13 and the 2D phase-corrected magic angle turning (PHORMAT),14 can be extended to biological, soft-solid samples to produce essentially sideband-free isotropic spectra. These approaches have led, inter alia, to the successful analysis with PASS of intact organs at spinning speeds as low as 43 Hz with a spectral resolution comparable to or better than the resolution obtained with fast MAS.15 This methodology has been recently
igh-resolution magic-angle spinning (HR-MAS) nuclear magnetic resonance (NMR) spectroscopy has become a widely used analytical tool for the study of biological systems, which provides metabolic information on intact tissues with no need for sample extraction.1 Proton HR-MAS NMR has been successfully applied to analyze intact cells and tissues from organs such as brain, lung, kidney, heart, and muscle, with a focus on mobile molecules.2−9 Typically, NMR spectra of these heterogeneous systems are recorded at MAS frequencies of a few kilohertz to effectively suppress spinning sidebands (SSB), that can overlap with metabolite signals and alter peak intensities, and approach the resolution obtained for isotropic liquid samples. A serious problem associated with conventional HR-MAS protocols is the large centrifugal forces experienced in the sample during the rotation, which can reach 100 000g and destroy tissue structures or even individual cells. For example, using rotors with an outer diameter of 4 mm (O.D.) with an inner diameter (I.D.) of 3 mm, in a sample of lipid-laden adipocytes, ∼20% of the cells were damaged after 2 h of MAS rotation at 3.5 kHz,10 serious visible damage was observed in human prostate tissues after spinning at 5 kHz,11 and cell lysis © 2014 American Chemical Society
Received: July 26, 2014 Accepted: October 6, 2014 Published: October 6, 2014 10749
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Analytical Chemistry
Article
applied in conjunction with microcoils.16,17 More spectacularly, PHORMAT has allowed the acquisition of 2D spectra on excised rat liver at a spinning frequency of 1 Hz.18 However, the very long acquisition times involved (up to 6 h for the PHORMAT experiment) as well as sensitivity and resolution issues further limit the application of these methods to the analysis of intact biological tissues.19 Perhaps most importantly, the intensity of the isotropic peak is affected by the extent of the SSB pattern, which may vary from sample to sample, and adds an element of variability to the quantitative analysis. As an alternative to the use of solid-state NMR pulse sequences, we have investigated the experimental conditions that would lead to high-quality HR-MAS spectra on slow spinning samples with solution-state NMR pulse sequences.20 These studies are motivated by the expected source of SSB in HRMAS for the mobile metabolites. In biological samples, which contain many intracellular and intercellular structures, metabolites are distributed heterogeneously, and often close to a boundary of a material with a different susceptibility than the compartment containing the compound under observation. The susceptibility differences between the various compartments induce magnetic field gradients in the sample. As a result, the resonance shifts become space dependent, resulting in a line broadening and in SSB in slow HR-MAS NMR spectra. In addition, SSB may be caused by disruptions of the liquid isotropic nature due to the presence of air bubbles and/or on the occurrence of inhomogeneities of the radio frequency (RF) field. The introduction of sidebands due to wobbling of the rotor around the magic angle during spinning has also been suggested.21 In our recent experiments, we have shown that, optimizing the sample preparation, particularly avoiding air bubbles and using reduced sample volumes with optimized geometry, led to well-resolved low-sideband TOCSY and HSQC spectra on tissues using HR-MAS at low spinning rates.20 The use of a reduced rotor volume of 12 μL led to a decrease of both the number and intensity of SSB in slow HR-MAS spectra, since RF field inhomogeneities and magnetic susceptibility line broadening effects can be minimized. The introduction of air bubbles increased the number of SSB for larger chambers and highlighted the role of magnetic susceptibility distributions in the presence of SSB at low spinning speeds (500 Hz). This collection of observation opens the way to standardized slow HR-MAS NMR experiments. In this work, we introduce a robust approach for the slow HR-MAS analysis of two representative tissue samples and exploit it to investigate the effect of fast and slow MAS on sample integrity. Using an optimized preparation protocol, we obtain slow HR-MAS spectra comparable to those obtained at conventional, high spinning rates. Furthermore, we adapted the classic relaxation filter used to remove broader signals to slow HR-MAS NMR experiments. We illustrate how the proposed approach yields high-quality spectra and preserve sample integrity for very fragile objects, with the example of fish eggs. These results are demonstrated at two different magnetic fields and with two types of HR-MAS rotors, including standard biopsy holding systems.
Figure 1. Sample preparation protocol for slow high-resolution magicangle spinning (HR-MAS) analysis of soft tissues.
■
Figure 2. 1H HR-MAS spectra of a heifer liver sample spun at 400 Hz on a 600 MHz spectrometer: (a) water-presaturated CPMG and (b) PROJECT.
MATERIALS AND METHODS Samples. A box of lumpfish roe was purchased at the supermarket and kept at 4 °C until the experiments. Heifer liver was purchased fresh from the butcher shop and kept at −80 °C until analysis. Optimized Protocol for the Preparation of Soft Tissues. A six-step protocol is described in Figure 1, for two
different volume-holding systems: 12 μL rotors and disposable inserts. Investigation of the Effect of Magic-Angle Spinning on Lumpfish Eggs. Experiments were carried out with a 10750
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Analytical Chemistry
Article
Figure 3. Water-presaturated 1H HR-MAS PROJECT spectra of heifer liver samples prepared according to the optimized protocol in a 12-μL rotor spun at (a, b) 4 kHz and (c, d) 400 Hz and recorded at 600 MHz (panels a and c) or 400 MHz (panels b and d). Sidebands are indicated by asterisks (∗).
first bringing the rotor to a speed of ∼1000−1500 Hz before reaching the set speed. This procedure is clearly not adapted if the threshold spinning speed must be avoided; thus, we resorted to manual adjustment of the spinning speed for the 400 Hz experiments. Newer versions of the MAS unit software optimized for slow spinning do not present this problem. Sidebands were identified by comparing spectra recorded at 4 kHz and 400 Hz. If any ambiguity persisted, a third experiment was performed at 600 Hz in order to obtain better characterization of the SSB. The T2-weighted spectra were obtained using either the CPMG sequence from the Bruker pulse program library or the presaturated PROJECT sequence from Aguilar et al.;22 both were adapted to include a presaturation pulse for water suppression. One-dimensional (1D) water-presaturated proton spectra were recorded using a spectral width of 10 ppm for fish eggs and 18 ppm for heifer liver, 16K data points, 64 scans, a relaxation delay of 2.5 s, and acquisition times of 1.36 s for fish eggs and 0.75 s for heifer liver. The total echo time was 96 ms for both water-presaturated CPMG and PROJECT sequences and the interpulse delay was set equal to the rotor period, leading to a total acquisition time of ∼4 min. Post-processing of all of the spectra included Fourier transformation of the free induction decays (FIDs) after exponential multiplication with a factor of 0.30 Hz, followed by phase and baseline correction.
regular 4-mm rotor (volume of 96 μL) and a ZrO2 rotor equipped with a 12 μL Teflon insert. For the regular rotor, 5 or 6 eggs were rinsed with a solution of phosphate buffer (0.05M, pH 6.8), in H2O/D2O (75/25) and placed in the rotor. Five microliters (5 μL) of phosphate buffer were added before carefully closing the rotor. For rotors equipped with a 12-μL insert, one egg was rinsed with the phosphate buffer and placed in the rotor. Five microliters (5 μL) of phosphate buffer were added before the rotor was carefully closed. The samples were submitted either to the classic (4 kHz) or the slow HR-MAS rotation rate (400 Hz) for 15 min. The rotor was then opened and the eggs carefully removed to observe the effect of the rotation. At the end of the analysis, after 30 min of spinning, the egg was carefully removed from the rotor to observe its general aspect. The solution was kept in the rotor and the volume used was increased to 12 μL, using fresh buffer solution. The solution was subjected to HR-MAS NMR analysis at 4 kHz to evaluate its metabolic content after different spinning rates and determine an eventual leakage of metabolites due to sample spinning. HR-MAS NMR Spectroscopy. All HR-MAS NMR experiments were performed at room temperature on a Bruker Avance II spectrometer operating at a resonance frequency of 400.36 MHz and a Bruker Avance I spectrometer operating at a resonance frequency of 600.13 MHz. Both instruments are equipped with a 4 mm double resonance (1H, 13C) gradient HR-MAS probe. Experiments were conducted at either 400 or 4000 Hz MAS frequencies. The spinning rate was controlled using a commercial MAS speed controller with frequency stability better than ±3 Hz. For the 4000 Hz experiments, the automatic procedure available to us was employed. This involves
■
RESULTS AND DISCUSSION Relaxation Filters. The spin echo is perhaps the most important building block in HR-MAS NMR spectroscopy. It is used to suppress uninformative, broad spectral features, which
10751
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Analytical Chemistry
Article
Figure 4. Water-presaturated 1H HR-MAS PROJECT spectra of heifer liver samples prepared according to the optimized protocol in a disposable insert spun at (a, b) 4 kHz and (c, d) 400 Hz and recorded at 600 MHz (panels a and c) or 400 MHz (panels b and d). Sidebands are indicated by stars.
arise typically from macromolecules that are not commonly analyzed in metabolomics. One limitation of this relaxation filter is the echo modulation induced by scalar couplings. In the classical Carr−Purcell−Meiboom−Gill (CPMG) pulse sequence, this modulation is minimized if the condition J ≪ 1/τ is met, where τ is the interpulse delay and J is the magnitude of the scalar coupling.23,24 Short interpulse delays, repeated for several cycles sufficient to implement a filter of the correct length, are thus necessary and commonly used to refocus J-modulation. However, the resulting, large high-RF duty cycles cause sample heating that can affect tissue integrity. At relatively low spinning rates ( 1 ms). While this is beneficial in terms of power deposition in the sample, the CPMG sequence is then unsuitable to refocus the J-modulation properly. The spectrum of Figure 2a shows a 1D 1H spectrum obtained on a 600 MHz spectrometer by applying the water-presaturated CPMG sequence on a heifer liver sample spun at 400 Hz, using a rotor-synchronized interpulse delay of 2.5 ms. As expected, the presence of J-modulation leads to spectral distortions, which are well-illustrated by the apparition of negative signals and loss of intensity, thereby compromising data interpretation. Recently, the PROJECT sequence has been introduced in liquid-state NMR as an alternative to the CPMG sequence that suppresses J-modulation, even for long interpulse delays.22 However, the advantages of the PROJECT sequence have not been demonstrated yet for slow HR-MAS NMR. Figure 2b shows a 1D 1H spectrum obtained with the PROJECT sequence at a spinning rate of 400 Hz. The interferences observed on the CPMG spectrum are suppressed.
Figure 5. Photographs of a set of lumpfish eggs (a) before being subjected to MAS, (b) after 15 min at 4 kHz MAS, and (c) after 15 min at 400 Hz MAS.
In particular, the spectral regions at δ = 4.67 ppm, δ = 2.85 ppm, δ = 3.2 ppm show negative peaks in the CPMG spectrum but are totally refocused in the PROJECT spectrum. The PROJECT sequence thus provides for T2-filtered, slow HR-MAS spectra, which are essential for many applications. Sample Preparation Protocol. A robust sample preparation protocol is necessary for high-quality sideband-free spectra at low spinning rates on biological samples. Figure 1 shows a six-step protocol that builds upon observations made with SSBreduced slow-HRMAS experiments and introduces several new, essential steps.20,25,26 Figures 3 and 4 show the 1D 1H HR-MAS NMR spectra, obtained at 4000 Hz MAS, of a heifer liver sample contained either in a 12-μL rotor (Figures 3a and b) or in a disposable insert (Figures 4a and b). The spectrum consists of an envelope of broad signals that could arise either from macromolecules and from components with restricted molecular motions superimposed on well-resolved peaks from small molecules characterized by fast and isotropic mobility. Preparation of the same 10752
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Analytical Chemistry
Article
Figure 6. Water-presaturated 1H HR-MAS PROJECT spectra of an isolated lumpfish egg spun at (a) 4 kHz and (c) 400 Hz. Spectra shown in panels b and d correspond, respectively, to the residual solution in the rotor, enhanced by factors of (b) 4 and (d) 10 after 15 min of spinning at 4 kHz and 400 Hz. Photographs on the top left (inset, panel a) and top right (inset, panel c) show the aspect of the egg after fast and slow MAS, respectively.
Earlier studies have also shown that disruptions of the internal structure could cause the transfer of metabolites between the tissue and the solution. Freeze−thaw cycles of the sample, as well as prolonged high-frequency spinning, have been shown to induce changes in HR-MAS NMR spectra that compromise data interpretation.29,30 Figures 6a and 6c display the spectra of a single lumpfish egg analyzed at 4 kHz and 400 Hz, respectively. The two spectra are very similar and both of them display satisfactory intensity and resolution. The absence of spinning sidebands at 400 Hz MAS is likely due to the natural geometry of the sample (sphere), the minimized effect of B1 inhomogeneities for small-volume rotors and the isotropic composition of the egg close to liquid. Figures 6b and 6d show the spectra of the solution obtained after removing the egg that underwent MAS, and topping up the solution with fresh buffer to reach a volume of 12 μL. In Figure 6b, the spectrum acquired on the solution after the 4 kHz HR-MAS shows a metabolic profile very similar to that of the entire egg (Figure 6a). This suggests an important leakage of the inner content of the egg into the remaining solution at high spinning rates, because of the disruption of the surrounding membrane. The number and the intensity of the metabolites signals is much lower in the residual solution after slow HR-MAS, as shown in Figure 6d. The low intensity signals between 3.2 and 4.0 ppm could arise from the porous character of the membrane of the egg or from components external to the membrane as they can also be identified in the washing solution of the lumpfish row (see Figure S2 in the Supporting Information). These results clearly demonstrate the high potential of using slow MAS for HR-MAS studies on fragile samples as the integrity of the sample is preserved and the transfer of metabolites from the sample studied to the solution is minimized.
sample according to the optimized protocol leads to spectacular results at the low spinning rate of 400 Hz, as illustrated in Figures 3c and 3d for the 12-μL rotor and Figures 4c and 4d for the disposable insert. In particular, the centrifugation steps before and after addition of the phosphate buffer solution contribute to an optimal sample geometry in the rotor and help remove air bubbles from the tissue to get a relatively air-free, homogeneous sample. Note that the centrifugal forces induced by the centrifugation steps are of the order of 3000 g, which corresponds to the forces experienced by the sample during slow spinning. Faster centrifugation speeds result in the partial destruction of the sample (see Table S1 in the Supporting Information). Using the protocol proposed here, slow HR-MAS spectra are almost devoid of SSB and are comparable in both resolution and intensity to those obtained at higher spinning frequencies (4000 Hz). The different geometry of the two experimental setups is the likely cause for the differences observed in the SSB patterns. Importantly, disposable inserts are the standard tools in metabolomics applications of HR-MAS NMR spectroscopy. Magic Angle Spinning of a Fragile Sample. In order to explore the influence of the spinning frequency on the integrity of fragile samples, we recorded fast and slow HR-MAS spectra of fish eggs. Recent studies have illustrated the deleterious effects of high-frequency MAS on the morphology of human glioblastoma tumors27 or fish eggs.28 In this latter case, the globular and plump eggs became shriveled and bleached, because of the rupture of the egg membrane and the dehydration of egg cells. Figure 5 shows photographs of a set of lumpfish eggs before and after 15 min of MAS at either 4 kHz or 400 Hz. As expected, the eggs that have been subjected to faster MAS are characterized by severe structural damages due to intense centrifugal forces (Figure 5b). In contrast, the eggs submitted to slow MAS are still intact after 15 min of spinning at 400 Hz, where centrifugal forces remain below 1000 g (Figure 5c). These results show that decreasing centrifugal forces by 2 orders of magnitude could make it possible to perform HR-MAS experiments on intact fragile biological samples.
■
CONCLUSION
In summary, we have introduced a robust, metabolomics-ready, approach for the characterization of fragile biological objects with slow high-resolution magic-angle spinning (HR-MAS) and demonstrated its potential for sample-preserving analyses. 10753
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
Analytical Chemistry
Article
(14) Hu, J. Z.; Wang, W.; Liu, F.; Solum, M. S.; Alderman, D. W.; Pugmire, R. J.; Grant, D. M. J. Magn. Reson., Ser. A 1995, 113, 210− 222. (15) Hu, J. Z.; Rommereim, D. N.; Minard, K. R.; Woodstock, A.; Harrer, B. J.; Wind, R. A.; Phipps, R. P.; Sime, P. J. Toxicol. Mech. Methods 2008, 18, 385−398. (16) Wong, A.; Aguiar, P. M.; Sakellariou, D. Magn. Reson. Med. 2010, 63, 269−274. (17) Wong, A.; Li, X.; Molin, L.; Solari, F.; Elena-Herrmann, B.; Sakellariou, D. Anal. Chem. 2014, 86, 6064−6070. (18) Hu, J. Z.; Rommereim, D. N.; Wind, R. A. Magn. Reson. Med. 2002, 47, 829−836. (19) Wind, R. A.; Hu, J. Z. Prog. Nucl. Magn. Reson. Spectrosc. 2006, 49, 207−259. (20) Renault, M.; Shintu, L.; Piotto, M.; Caldarelli, S. Sci. Rep. 2013, 3, 5. (21) Avni, R.; Mangoubi, O.; Bhattacharyya, R.; Degani, H.; Frydman, L. J. Magn. Reson. 2009, 199, 1−9. (22) Aguilar, J. A.; Nilsson, M.; Bodenhausen, G.; Morris, G. A. Chem. Commun. 2012, 48, 811−813. (23) Carr, H. Y.; Purcell, E. M. Phys. Rev. 1954, 94, 630−638. (24) Meiboom, S.; Gill, D. Rev. Sci. Instrum. 1958, 29, 688−691. (25) Beckonert, O.; Coen, M.; Keun, H. C.; Wang, Y.; Ebbels, T. M. D.; Holmes, E.; Lindon, J. C.; Nicholson, J. K. Nat. Protoc. 2010, 5, 1019−1032. (26) Wind, R. A.; Hu, J. Z.; Rommereim, D. N. Magn. Reson. Med. 2001, 46, 213−218. (27) Martínez-Bisbal; Esteve; Martínez-Granados, B.; Celda. J. Biomed. Biotechnol. 2011, 2011, 8. (28) Cai, H. H.; Chen, Y. S.; Cui, X. H.; Cai, S. H.; Chen, Z. PLoS One 2014, 9 (1), e86422. (29) Bourne, R.; Dzendrowskyj, T.; Mountford, C. NMR Biomed. 2003, 16, 96−101. (30) Opstad, K. S.; Bell, B. A.; Griffiths, J. R.; Howe, F. A. NMR Biomed. 2008, 21, 1138−1147.
Several challenges had to be overcome to operate at lower magic-angle spinning (MAS) frequencies. We have proposed an optimized sample preparation protocol that leads to strongly reduced spinning sideband (SSB) spectra at 400 Hz MAS, using either a 12-μL rotor or a disposable 30-μL insert. Both signal intensity and resolution are comparable to those obtained at higher spinning rates. The classical CPMG sequence was replaced by the PROJECT sequence in order to obtain in-phase, T2-filtered spectra. By decreasing the spinning rate by 1 order of magnitude, centrifugal forces are reduced from 100 000 g at 4 kHz to 1000 g at 400 Hz. These results demonstrate that slow HR-MAS constitutes a suitable analytical tool for the characterization of fragile samples. Development of protocols including two-dimensional (2D) NMR techniques can be readily envisioned and are underway in our laboratory.
■
ASSOCIATED CONTENT
S Supporting Information *
The authors declare no competing financial interests. This material is available free of charge via the Internet at http:// pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mails: [email protected], [email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank Christina Sizun for experimental help, as well as Institut de Chimie des Substances Naturelles (ICSN), Bruker, and ANR (No. ANR-2011-JS08-014-01) for funding.
■
REFERENCES
(1) Lindon, J. C.; Beckonert, O. P.; Holmes, E.; Nicholson, J. K. Prog. Nucl. Magn. Reson. Spectrosc. 2009, 55, 79−100. (2) Cheng, L. L.; Ma, M. J.; Becerra, L.; Ptak, T.; Tracey, I.; Lackner, A.; Gonzalez, R. G. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 6408−6413. (3) Tate, A. R.; Foxall, P. J. D.; Holmes, E.; Moka, D.; Spraul, M.; Nicholson, J. K.; Lindon, J. C. NMR Biomed. 2000, 13, 64−71. (4) Bollard, M. E.; Murray, A. J.; Clarke, K.; Nicholson, J. K.; Griffin, J. L. FEBS Lett. 2003, 553, 73−78. (5) Righi, V.; Mucci, A.; Schenetti, L.; Tosi, M. R.; Grigioni, W. F.; Corti, B.; Bertaccini, A.; Franceschelli, A.; Sanguedolce, F.; Schiavina, R.; Martorana, G.; Tugnoli, V. Anticancer Res. 2007, 27, 3195−3204. (6) Chen, J. H.; Wu, Y. V.; DeCarolis, P.; O’Connor, R.; Somberg, C. J.; Singer, S. Magn. Reson. Med. 2008, 59, 1221−1224. (7) Bollard, M. E.; Garrod, S.; Holmes, E.; Lindon, J. C.; Humpfer, E.; Spraul, M.; Nicholson, J. K. Magn. Reson. Med. 2000, 44, 201−207. (8) Garrod, S.; Humpfer, E.; Spraul, M.; Connor, S. C.; Polley, S.; Connelly, J.; Lindon, J. C.; Nicholson, J. K.; Holmes, E. Magn. Reson. Med. 1999, 41, 1108−1118. (9) Garrod, S.; Humpher, E.; Connor, S. C.; Connelly, J. C.; Spraul, M.; Nicholson, J. K.; Holmes, E. Magn. Reson. Med. 2001, 45, 781− 790. (10) Weybright, P.; Millis, K.; Campbell, N.; Cory, D. G.; Singer, S. Magn. Reson. Med. 1998, 39, 337−345. (11) Taylor, J. L.; Wu, C.-L.; Cory, D.; Gonzalez, R. G.; Bielecki, A.; Cheng, L. L. Magn. Reson. Med. 2003, 50, 627−632. (12) Aime, S.; Bruno, E.; Cabella, C.; Colombatto, S.; Digilio, G.; Mainero, V. Magn. Reson. Med. 2005, 54, 1547−1552. (13) Antzutkin, O. N.; Shekar, S. C.; Levitt, M. H. J. Magn. Reson., Ser. A 1995, 115, 7−19. 10754
dx.doi.org/10.1021/ac502792u | Anal. Chem. 2014, 86, 10749−10754
