Langmuir 2007, 23, 8035-8041
8035
Structural Effects in the Analysis of Supported Lipid Bilayers by Time-of-Flight Secondary Ion Mass Spectrometry Christelle Prinz and Fredrik Ho¨o¨k* DiVision of Solid State Physics, Department of Physics, Lund UniVersity, SE-221 00 Lund, Sweden
Jakob Malm and Peter Sjo¨vall* DiVision of Solid State Physics, Department of Physics, Lund UniVersity, SE-221 00 Lund, and Chemistry and Materials Technology, SP Technical Research Institute of Sweden, P.O. Box 857, SE-501 15 Borås, Sweden ReceiVed February 16, 2007. In Final Form: April 28, 2007 We contribute to the rapidly emerging interest in the application of time-of-flight secondary ion mass spectrometry (TOF-SIMS) for chemical analysis of biological materials by presenting a careful TOF-SIMS investigation of structurally different SiO2-supported phospholipid assemblies. Freeze-dried supported 1-oleoyl-2-palmitoyl-sn-glycero-3phosphocholine (POPC) bilayers, Langmuir-Blodgett POPC monolayers, and disordered thick POPC films were investigated. Compared with the two latter structures, the supported bilayer showed a strong (5-10 times) enhancement in the yield of both the molecular and the dimer ion peaks of POPC, suggesting that the molecular peak may be used as a sensitive indicator for changes in the membrane structure and, in particular, an indicator for the presence of bilayer structures in, e.g., cell and tissue samples. The detection efficiency and the useful lateral resolution indicate that a lateral resolution of around 100 nm can be obtained on all structures by imaging the phosphocholine ion at 184 u using Bi3+ primary ions. For the chemically specific molecular peak at 760 u, the measured detection efficiencies correspond to a useful lateral resolution of around 2 µm for the bilayer structure. The results are discussed in relation to recent dynamic SIMS (nano-SIMS) analysis of freeze-dried supported lipid bilayers, displaying similar or higher lateral resolution, but which in contrast to TOF-SIMS requires isotopic labeling of the analyzed lipids.
Introduction There is currently a strong interest in the application of timeof-flight secondary ion mass spectrometry (TOF-SIMS) to the chemical analysis of biological materials, such as cells and tissues.1-11 The capability to provide detailed chemical information at submicrometer lateral resolution of unlabeled substances comprises an exciting opportunity for realizing global, unprejudiced analysis of the subcellular localization of biomolecules in biological structures. To achieve this, however, a number of issues need to be carefully considered. One critical question is how to prevent the disruption of structures and migration of substances while subjecting the biological sample to the vacuum environment required by the TOF-SIMS analysis. Other important questions concern how to improve the detection efficiency to * To whom correspondence should be addressed. (P.S.) E-mail:
[email protected]. Phone: +46 33 165299. Fax: +46 33 103388. (F. H.) E-mail:
[email protected]. Phone: +46 46 2221494. Fax: +46 46 2223637. (1) Arlinghaus, H. F.; Kriegeskotte, C.; Fartmann, M.; Wittig, A.; Sauerwein, W.; Lipinsky, D. Appl. Surf. Sci. 2006, 252, 6941-6948. (2) Brunelle, A.; Touboul, D.; Laprevote, O. J. Mass Spectrom. 2005, 40, 985-999. (3) Ewing, A. G. Appl. Surf. Sci. 2006, 252, 6821-6826. (4) Monroe, E. B.; Jurchen, J. C.; Lee, J.; Rubakhin, S. S.; Sweedler, J. V. J. Am. Chem. Soc. 2005, 127, 12152-12153. (5) Nygren, H.; Borner, K.; Malmberg, P.; Hagenhoff, B. Appl. Surf. Sci. 2006, 252, 6975-6981. (6) Ostrowski, S. G.; Van Bell, C. T.; Winograd, N.; Ewing, A. G. Science 2004, 305, 71-73. (7) Parry, S.; Winograd, N. Anal. Chem. 2005, 77, 7950-7957. (8) Sjovall, P.; Johansson, B.; Lausmaa, J. Appl. Surf. Sci. 2006, 252, 69666974. (9) Sjovall, P.; Lausmaa, J.; Johansson, B. Anal. Chem. 2004, 76, 4271-4278. (10) Touboul, D.; Brunelle, A.; Halgand, F.; De La Porte, S.; Laprevote, O. J. Lipid Res. 2005, 46, 1388-1395. (11) Touboul, D.; Kollmer, F.; Niehuis, E.; Brunelle, A.; Laprevote, O. J. Am. Soc. Mass Spectrom. 2005, 16, 1608-1618.
obtain a sufficiently high lateral resolution and how to compensate for matrix effects in the obtained ion images, i.e., to determine whether observed signal variations depend on concentration variations along the surface or variations in the chemical environment. To be able to make proper interpretations of TOFSIMS data, it is important to have control over these factors and to be aware of the extent to which these factors limit the amount of available information. Different strategies have previously been used for the preparation of biological materials for TOF-SIMS analysis. In the analysis of atomic ions and small fragments using ion microscopy or laser postionization secondary neutral mass spectrometry (laser SNMS), retention of membrane integrity until fixation is critical to prevent migration across the cell membranes.12-16 For this purpose, plunge freezing followed by controlled freeze drying or analysis in the frozen hydrated state has proven to be a successful preparation strategy, demonstrated by observations of retained anatomical K+/Na+ concentration ratios inside and outside the plasma membrane.13,16,17 In the analysis of lipids (and other biomolecules) in cells and tissues, it is less clear as to what degree the integrity of the membrane structures needs to be retained up to the point of fixation, depending on the mobility of the specific molecules under study. (12) Chandra, S.; Morrison, G. H. Biol. Cell 1992, 74, 31-42. (13) Chandra, S.; Smith, D. R.; Morrison, G. H. Anal. Chem. 2000, 72, 104A114A. (14) Grovenor, C. R. M.; Smart, K. E.; Kilburn, M. R.; Shore, B.; Dilworth, J. R.; Martin, B.; Hawes, C.; Rickaby, R. E. M. Appl. Surf. Sci. 2006, 252, 6917-6924. (15) Marxer, C. G.; Kraft, M. L.; Weber, P. K.; Hutcheon, I. D.; Boxer, S. G. Biophys. J. 2005, 88, 2965-2975. (16) Wittig, A.; Wiemann, M.; Fartmann, M.; Kriegeskotte, C.; Arlinghaus, H. F.; Zierold, K.; Sauerwein, W. Microsc. Res. Tech. 2005, 66, 248-258. (17) Arlinghaus, H. F.; Fartmann, M.; Kriegeskotte, C.; Dambach, S.; Wittig, A.; Sauerwein, W.; Lipinsky, D. Surf. Interface Anal. 2004, 36, 698-701.
10.1021/la7004634 CCC: $37.00 © 2007 American Chemical Society Published on Web 06/15/2007
8036 Langmuir, Vol. 23, No. 15, 2007
We have previously shown that migration of lipids can be significant in freeze-dried tissue sections. A massive migration of cholesterol to the tissue surface upon warming of a mouse brain section to around 0 °C in vacuum was observed.8 In addition to cryofixation, various chemical stabilization/fixation strategies have been shown to preserve the cellular structures for subsequent TOF-SIMS analysis. For example, Parry and Winograd7 obtained TOF-SIMS images of phosphatidylcholine and phosphatidylethanolamine distributions in lyophilized and fractured macrophages and glial cells after stabilization in a trehalose-glycerol matrix. Additional studies of the mobility of biomolecules in cells and tissues under various conditions are definitely needed to ensure the biological relevance of future TOF-SIMS results on cells and tissues. Phosphatidylcholine (PC) is one of the most frequently studied groups of biomolecules by TOF-SIMS, which is partly due to its biological relevance and relatively high abundance in cells and tissues. In addition, the characteristic phosphocholine ion at 184 u has a relatively high secondary ion yield, providing a strong and easily recognized signal for its detection. The phosphocholine peak has been used in the characterization of PC model systems, such as phase-separated Langmuir-Blodgett (LB) monolayers,18,19 mixed LB monolayers,20-23 and deposited PC films.21,24 Although often observed in tissue samples,8-10,25 in particular when cluster primary ions are used, the molecular peaks at 700-800 u have been significantly less characterized in TOF-SIMS studies of PC model systems, despite their significantly higher specificity (providing information about the entire molecule, including the fatty acid tails), as compared to the phosphocholine peak. An exception, however, is a TOFSIMS study by Luxembourg et al.,26 in which it was found that the secondary ion yields of the molecular peaks of PC were greatly enhanced if it was deposited in a MALDI matrix as compared to in a pure PC film. In this work, we use TOF-SIMS to investigate the appearance of the molecular peak and peaks of various fragments for different SiO2-supported lipid membrane assemblies. Inspired by the recent work by Boxer and co-workers,15,27 demonstrating that supported phospholipid bilayers (SPBs) on SiO2 can be successfully freezedried with preserved integrity for dynamic SIMS (nano-SIMS) analysis, a nondestructive cryofixation procedure was used to preserve the bilayer structure for the TOF-SIMS analysis. By comparing the TOF-SIMS results obtained for LB monolayers and SPBs with those obtained for disordered lipid films, we show that the bilayer structure is retained in vacuum. An advantage of TOF-SIMS over nano-SIMS is that spatially resolved chemical analysis can be made without introduction of isotopic labels. Supported by a careful analysis of the secondary ion yields, disappearance cross sections, efficiencies, and useful lateral (18) Bourdos, N.; Kollmer, F.; Benninghoven, A.; Sieber, M.; Galla, H. J. Langmuir 2000, 16, 1481-1484. (19) Leufgen, K. M.; Rulle, H.; Benninghoven, A.; Sieber, M.; Galla, H. J. Langmuir 1996, 12, 1708-1711. (20) McQuaw, C. M.; Sostarecz, A. G.; Zheng, L.; Ewing, A. G.; Winograd, N. Appl. Surf. Sci. 2006, 252, 6716-6718. (21) Pacholski, M. L.; Cannon, D. M.; Ewing, A. G.; Winograd, N. J. Am. Chem. Soc. 1999, 121, 4716-4717. (22) Ross, M.; Steinem, C.; Galla, H. J.; Janshoff, A. Langmuir 2001, 17, 2437-2445. (23) Sostarecz, A. G.; Cannon, D. M.; McQuaw, C. M.; Sun, S. X.; Ewing, A. G.; Winograd, N. Langmuir 2004, 20, 4926-4932. (24) Roddy, T. P.; Cannon, D. M.; Ostrowski, S. G.; Ewing, A. G.; Winograd, N. Anal. Chem. 2003, 75, 4087-4094. (25) Altelaar, A. F. M.; Klinkert, I.; Jalink, K.; de Lange, R. P. J.; Adan, R. A. H.; Heeren, R. M. A.; Piersma, S. R. Anal. Chem. 2006, 78, 734-742. (26) Luxembourg, S. L.; McDonnell, L. A.; Duursma, M. C.; Guo, X. H.; Heeren, R. M. A. Anal. Chem. 2003, 75, 2333-2341. (27) Kraft, M. L.; Weber, P. K.; Longo, M. L.; Hutcheon, I. D.; Boxer, S. G. Science 2006, 313, 1948-1951.
Prinz et al.
resolutions for selected molecular fragments, the potential of TOF-SIMS for studies of SPBs, cells, and tissues is compared with that of nano-SIMS, which was recently proven compatible with the determination of the lipid composition in domains down to 100 nm in diameter and variations within domains of diameters down to 1-2 µm.27 Materials and Methods Lipids. Synthetic 1-oleoyl-2-palmitoyl-sn-glycero-3-phosphocholine (POPC), purity >99%, in powder form, was purchased from Avanti Polar Lipids, Inc. (Alabaster, AL). 2-[12-[(7-Nitrobenz-2oxa-1,3-diazol-4-yl)amino]dodecanoyl]-1-hexadecanoyl-sn-glycero3-phosphocholine (NBD-C12-HPC), which is a fluorescently labeled variant of phosphatidylcholine, was purchased from Molecular Probes (Leiden, The Netherlands). Surfaces for Supported Lipid Bilayers. The lipid bilayers were formed on a Si wafer onto which a 300 nm thick SiO2 layer had been grown thermally. To limit the defect propagation within the supported bilayer, the SiO2 surface was coated with a titanium oxide (TiO2) grid delimiting 500 × 500 µm SiO2 squares separated by 10 µm wide TiO2 lines. TiO2 acts as a barrier for the bilayer’s lipids confined in the SiO2 squares.28 The substrate had a TiO2 frame that remained hydrophilic after the bilayer formation on SiO2, keeping a thin film of water everywhere on the sample during the plunge freezing in liquid ethane, preventing any contact of the bilayer with air which would destroy the upper leaflet. The TiO2 pattern was made using the standard UV lithography technique, using a double-layer resist (LOR3A and S1813, MicroChem Corp., Newton, MA) and an MJB3 contact mask aligner (Karl Su¨ss, Mu¨nchen, Germany). The surface was cleaned in a UV ozone chamber (Plasma Preen, Terra Universal, Inc.) just before the bilayer formation. Preparation of Supported Lipid Bilayers. Supported lipid bilayers were formed by vesicle adsorption and rupture on the SiO2/ TiO2 surface according to procedures reported previously.29 Briefly, 40 nm diameter unilamellar vesicles at 5 mg/mL in 10 mM Tris, 100 mM NaCl, pH 8 buffer were obtained by two subsequent extrusion steps of multilamellar vesicles through 100 and 30 nm pore filters. The lipid molar composition was 99% POPC and 1% NBD-C12HPC, making the vesicles and bilayers fluorescent. Each vesicle batch was tested for bilayer formation on SiO2 surfaces using quartz crystal microbalance dissipation (QCM-D) measurements. The SiO2/TiO2 surface was immersed in a further diluted 0.1 mg/mL vesicle solution for 20 min to allow bilayer formation, which was verified by fluorescence recovery after photobleaching (FRAP). By bleaching a small spot on the surface and measuring the fluorescence recovery, we could tell if there was a bilayer (recovery) or simply vesicles adsorbed (no recovery) on the surface. After bilayer formation, the samples were rinsed in Milli-Q water. Within 2 h after formation of the bilayers, the samples were plunge frozen in liquid nitrogen-cooled ethane. The frozen samples were then transferred into liquid nitrogen (LN2), where they were stored until analysis (up to approximately two weeks). Immediately prior to TOF-SIMS analysis, the bilayer samples were transferred from the LN2 dewar to a precooled sample holder dedicated for TOF-SIMS analysis at controlled sample temperatures down to -130 °C. To minimize condensation on the sample surface and to keep the sample cold, the sample holder was placed inside a LN2-filled container just above the LN2 level during mounting. After mounting, the sample was immediately transferred into the vacuum chamber of the TOF-SIMS instrument. The maximum sample temperature was typically around -70 °C during sample transfer into the TOF-SIMS instrument. Inside the TOF-SIMS instrument, the sample was freeze-dried by keeping the sample temperature constant at -90 °C. After approximately 2 h at this temperature, all ice was removed and the TOF-SIMS analysis was initiated. (28) Groves, J. T.; Ulman, N.; Cremer, P. S.; Boxer, S. G. Langmuir 1998, 14, 3347-3350. (29) Reimhult, E.; Hook, F.; Kasemo, B. Langmuir 2003, 19, 1681-1691.
Analysis of Supported Lipid Bilayers by TOF-SIMS
Langmuir, Vol. 23, No. 15, 2007 8037
Figure 1. Schematic drawing of the different lipid overlayer structures studied in the present work: (a) bilayer, (b) LangmuirBlodgett monolayer, (c) disordered layer. Preparation of Langmuir-Blodgett Monolayers. POPC monolayers were formed at the air/water interface in a Nima Langmuir trough (Coventry, England). A clean Si/SiO2 surface was attached to the dipper and immersed in the water subphase. A mixture of POPC (99%) and 1% M NBD-PC at 1 mg/mL in chloroform was spread at the air/water interface. The chloroform was allowed to evaporate for 10 min, after which the layer was compressed at a speed of 20 cm2/min. The compression was stopped at 32 mN/m corresponding to a molecular area of 68 Å2/molecule, which is expected to be the same as for vesicle adsorption bilayers (60-70 Å2/molecule). The SiO2 surface was then raised at a speed of 1 mm/min while the surface pressure was kept constant, transferring the monolayer onto the substrate. Preparation of Deposited POPC Films. Disordered POPC films were prepared by depositing approximately 100 µL of 0.5 mg/mL POPC/chloroform solution on a cleaned Si wafer and allowed to dry. Before deposition, the Si wafers were cleaned in 2% sodium dodecyl sulfate (SDS) in deionized water for 15 min, rinsed with deionized water, and exposed to UV ozone for 15 min. The thickness of the obtained POPC films was relatively inhomogeneous with thinner regions in the peripheral parts. All data were obtained from the thicker regions in the center of the POPC deposits. TOF-SIMS Analysis. The TOF-SIMS analysis was carried out in a TOF-SIMS IV instrument (ION-TOF GmbH, Germany) equipped with a Bin+ cluster liquid metal ion source. All data were obtained using 25 keV Bi3+ primary ions. Positive and negative spectra at high mass resolution were recorded with the instrument in the socalled bunched mode (m/∆m ≈ 6000, beam diameter ∼4-6 µm), while high spatial resolution images were obtained in the so-called burst alignment mode (m/∆m ≈ 500, beam diameter ∼250 nm). The pulsed Bi3+ primary ion current was typically 0.1 pA in the bunched mode (at a pulse width of 10 ns and a pulse frequency of 6667 Hz) and 0.04 pA in the burst alignment mode (pulse width 100 ns, frequency 8333 Hz). Charge compensation using lowenergy electron flooding was applied only in the case of deposited POPC films. The secondary ion yields were obtained from spectra recorded in the bunched mode for 50 s over an analysis area of 100 × 100 µm2 (accumulated primary ion dose density (PIDD) 3.1 × 1011 ions/cm2). The disappearance cross sections (see below) were calculated from time profiles extracted from data recorded for 600 s at 100 × 100 µm2 (PIDD ) 3.8 × 1012 ions/cm2), at which point the ion signals typically had decreased to approximately 40% of their maximum value. The acquisition time for the highresolution images was 72 s for the 500 × 500 µm2 image, which considering that the beam diameter (250 nm) is smaller than the pixel size (256 × 256 pixels) corresponds to a PIDD of 4.4 × 1011 ions/cm2. The acquisition time for the 50 × 50 µm2 image (256 × 256 pixels) was 216 s, corresponding to a PIDD of 2.16 × 1012 ions/cm2. Values of the secondary ion yield (Y), disappearance cross section (σ), efficiency (E), and useful lateral resolution (∆L) were calculated for the supported POPC bilayers and the LB monolayers according to the definitions and methodology described previously.30 Briefly, the following definitions were used: Y (secondary ion yield), number of detected secondary ions per incident primary ion; σ (disappearance cross section), average area damaged for subsequent (30) Kotter, F.; Benninghoven, A. Appl. Surf. Sci. 1998, 133, 47-57.
Figure 2. Results from FRAP analysis of a supported POPC bilayer in an aqueous environment, showing (a) a fluorescence micrograph recorded after photobleaching (the dark line corresponds to the 10 µm thick TiO2 grid on the SiO2 substrate), (b) a fluorescence profile across the sample surface after bleaching, and (c) a graph of the fluorescence bleaching vs time indicating the diffusion-induced recovery within the bleached area of the bilayer. emission of secondary ions per incident primary ion, calculated from the time constant of the decaying signal intensity for the specific secondary ion in question, E (efficiency), maximum number of detectable secondary ions (SIs) per unit surface area on the sample, calculated by the ratio of the yield to the disappearance cross section, Y/σ; ∆L (useful lateral resolution), the side of a minimum square on the sample surface that is required to be analyzed to obtain a detectable signal, calculated by (ND/E)1/2, where ND is the number of detected ions required for producing a “significant signal”; in our calculations, ND ) 4 was used.
Results and Discussion Positive TOF-SIMS spectra were recorded on three different POPC overlayer structures, schematically represented in Figure 1. The diffusion coefficient, D, of POPC molecules in bilayers was measured using FRAP (see Figure 2). Its value (D ) 2 µm2/s) is in good agreement with previously reported values31 for POPC bilayers, thus verifying the formation of a bilayer structure. For disordered layers or unruptured adsorbed vesicles,
8038 Langmuir, Vol. 23, No. 15, 2007
Prinz et al.
Figure 3. (a) Chemical structure of POPC and (b) positive TOF-SIMS spectrum from a freeze-dried POPC bilayer. Table 1. Characteristic Peaks and Assignments in the Positive TOF-SIMS Spectra of POPC mass (u)
assignment
58.07 86.11 104.12 166.09 184.11 224.14 478.42 494.39 504.43 760.66 1520.32
(N(CH3)3 - H)+ (N(CH3)3(CH2)2 - H)+ (N(CH3)3(CH2)2OH)+ (N(CH3)3(CH2)2OPO2)+ (N(CH3)3(CH2)2OPO3H2)+ (N(CH3)3(CH2)2OPO3HC3H5)+ (N(CH3)3(CH2)2OPO3HC3H4(C16H31O2))+ (N(CH3)3(CH2)2OPO3HC3H4(C16H31O2)O)+ (N(CH3)3(CH2)2OPO3HC3H4(C18H33O2))+ (M + H)+ (2M + H)+
the expected diffusion constants are orders of magnitude lower. The bilayers were frozen and freeze-dried inside the TOF-SIMS instrument at -90 °C, according to the procedure described in the Materials and Methods. The TOF-SIMS analysis was carried out at -90 °C. Figure 3 shows a positive TOF-SIMS spectrum of a freezedried POPC bilayer, recorded with the instrument optimized for high mass resolution (m/∆m ≈ 6000). The chemical structure of POPC is shown in Figure 3a, and a list of the observed major fragment ion peaks and molecular ion peaks is provided in Table 1. In the low mass region, below 250 u, the spectrum is in good agreement with previously published spectra from different types of phosphatidylcholine overlayers.18,22 Starting with the peak at 58 u originating from the N(CH3)3 end group, a number of major peaks can be identified as ions originating from increasingly larger portions of the molecular head group, including the entire phosphocholine head group at 184 u. In the higher mass region, two peaks at 478 and 504 u can be identified as the entire phosphatidylcholine molecule after abstraction of either the oleoyl or palmitoyl group, respectively. The peaks at 760 and 1520 u can be identified as the protonated molecular ion (M + H)+ and the protonated dimer ion (2M + H)+, respectively. To investigate the integrity of the POPC bilayer structure after freeze drying, high-resolution images (∆l ≈ 300 nm) were recorded. Figure 4 shows ion images of phosphocholine (184 u) and total ion images in two different fields of view. In the larger field of view (Figure 4a,b), the images show a large homogeneous region on the right side of the analysis area and a region on the left side in which the bilayer structure has been distorted. In (31) Smith, E. A.; Coym, J. W.; Cowell, S. M.; Tokimoto, T.; Hruby, V. J.; Yamamura, H. I.; Wirth, M. J. Langmuir 2005, 21, 9644-9650.
note choline phosphocholine lyso-palmitoylphosphatidylcholine lyso-oleoylphosphatidylcholine molecular ion dimer ion
Figure 4c,d, magnified images from the homogeneous region show that no significant inhomogeneities could be detected at a length scale down to
