Chapter 4. Lipid Detection, Identification, and Imaging Single Cells with SIMS

Transcription

1 Chapter 4 Lipid Detection, Identification, and Imaging Single Cells with SIMS Michael L. Heien, Paul D. Piehowski, Nicholas Winograd, and Andrew G. Ewing Abstract Time-of-flight secondary ion mass spectrometry (ToF-SIMS) can be utilized to map the distribution of various molecules on a surface with submicrometer resolution. Many of its biological applications have been in the study of membrane lipids, such as phospholipids and cholesterol. For these studies, the effectiveness of chemical mapping is limited by low signal intensity from various biomolecules. Because of the high-energy nature of the SIMS ionization process, many molecules are identified by detection of characteristic fragments. Cluster ion sources are able to increase ionization, leading to increased information collected from a surface. In this chapter, we highlight the utility of SIMS to image lipids at single cells. Particularly, we will describe sample preparation, data collection, and the analysis of lipids for two systems; rat oligodendrocytes and Tetrahymena thermophila. SIMS spectra yield information regarding lipid identity and concentration across cell surface. Key words: Time-of-flight (ToF), secondary ion mass spectrometry (SIMS), freeze-fracture, freezeetch, imaging, mass spectrometry, lipids, single cell. 1. Introduction Mass spectrometry imaging with time-of-flight secondary ion mass spectrometry (ToF-SIMS) has revealed the spatial distribution of chemicals on a surface (1). When applied to biological samples, this method offers spatial information on biologically relevant small molecules (<1,000 Da). Indeed, ToF-SIMS imaging has been shown to be a powerful analytical tool for mapping the distribution of lipids on a cell membrane at the single-cell S.S. Rubakhin, J.V. Sweedler (eds.), Mass Spectrometry Imaging, Methods in Molecular Biology 656, DOI / _4, Springer Science+Business Media, LLC

2 86 Heien et al. level (2 11). This method combines submicrometer spatial resolution, high chemical specificity, and surface sensitivity, making it a promising tool for the study of lipids in cellular membranes. Briefly, a pulsed beam of primary ions is directed at the analysis surface, causing the sputtering of ions and molecules. The ions are extracted into a ToF mass analyzer generating a mass spectrum. To obtain an image, the primary beam is raster-scanned across the surface, recording a mass spectrum for each pixel. A particular m/z is then selected from the spectrum, and its intensity is mapped across the imaging area. When combined with cryogenic sample preparation techniques, imaging ToF-SIMS permits the detailed study of membrane lipids during dynamic processes such as membrane fusion. Mass spectrometric imaging with freeze-fracture sample preparation allows the study of lipid domain formation in native membranes without chemical labels. This chapter outlines the methods necessary to perform ToF-SIMS imaging. We outline sample preparation methods for two types of cells, rat oligodentrocytes and T. thermophila, an adhering and non-adhering cell type, respectively. The cells are cultured, mounted onto a silicon substrate, frozen, fractured, and analyzed. An alternate sample treatment, freeze-etching (8), is briefly described in the Section 3. These methods can be adapted for a given system. Images are then generated, and major lipid fragments are identified during data analysis. 2. Materials 2.1. Cell Culture 1. Tetrahymena cultures were obtained from Craig T. VanBell at Edinboro University, Edinboro, PA, USA. 2. PPYGFe medium: 2% proteose peptone (Becton, Dickinson and Company, Sparks, MD, USA), 0.1% yeast extract (Becton, Dickinson and Company), 0.2% yeast extract (Becton, Dickinson and Company), and 0.003% ferric EDTA. Prepare immediately prior to use. 3. Tris buffer solution: 10 mm Tris HCl, ph adjusted to 7.4 using 1 M NaOH. Store at room temperature ml centrifuge tubes L baffled polycarbonate Erlenmeyer flasks with filtered cap (Sigma Aldrich, Allentown, PA, USA). 6. Primary Sprague Dawley rat pup oligodendrocytes were obtained from Gong Chen, The Pennsylvania State University.

3 Lipid Detection, Identification, and Imaging Minimum Essential Medium (Invitrogen). 8. N1 supplement (Sigma Aldrich) ml polystyrene tissue culture flasks Sample Preparation 1. HPLC grade hexane (EMD chemicals, Gibbstown, NJ, USA). 2. Absolute ethanol (Pharmco-AAPER, Brookfield, CT, USA). 3. Deionized water (18 M ) mm silicon shards (Ted Pella, Redding, CA, USA) ml scintillation vials. 6. Ethane 99%. 7. Liquid nitrogen (LN 2 ). 8. Storage dewar, 50 l minimum. 9. London finder grid (Ted Pella, Redding, CA, USA). 10. Bath Sonicator. 11. Fluoroware wafer carrier box (Fluoroware Inc., Chaska, MN, USA). 12. Colloidal silver liquid (Ted Pella) Instrumentation 1. Kratos Prism ToF-SIMS spectrometer (Manchester, UK) or similarly designed static SIMS imaging instrument. 2. Indium liquid metal ion source (Beaverton, OR, USA). 3. Microchannel plate detector (Galileo Co., Sturbridge, MA, USA). 4. Channeltron detector (Burle, Lancaster, PA, USA). 5. Vertical illuminator microscope (Olympus, Melville, NY, USA). 6. Spot RT CCD Camera (Diagnostic Instruments, Sterling Heights, MI, USA). 7. Keithley 485 Auto Ranging Picoammeter (Keithley, Inc., Solon, OH, USA). 8. Custom-made freeze-fracture sample preparation chamber, Fig. 4.2c (12). 3. Methods As stated in the introduction, we outline sample preparation methods for two types of cells, rat oligodendrocytes and T. thermophila. The culture protocols for T. thermophila are discussed;

4 88 Heien et al. however, this does not limit the analysis to this type of cell. The cells are cultured, mounted onto a silicon substrate, frozen, fractured, and analyzed. An alternate sample treatment, freezeetching, is also described. These methods are transferable to imaging other single cells, and it is assumed that the reader has some familiarity with SIMS mass spectrometry Cell Culture 1. Autoclave PPYGFe medium. 2. The two complementary mating types of T. thermophila, B2086 and Cu428.1, are maintained separately in autoclaved PPYGFe medium. Cells are grown overnight in 2 L flasks at 30 C with shaking (3). 3. Prior to mating, the cells are washed by centrifugation, 400 g for 5 min, in 10 mm Tris buffer (ph 7.4). The cells are resuspended in the Tris buffer and shaken overnight at 30 C. 4. To initiate mating, the complementary strains are mixed in equal proportions and incubated, without shaking, at 30 C for 4 h. 5. Rat oligodendrocytes are maintained in MEM with N1 supplement and incubated at 37.4 C until maturation was observed under a microscope Preparation of Cells for Mass Spectrometry Imaging 1. Before use the silicon shards must be cleaned (see Note 1). To do this, the shards are placed in a glass scintillation vial and covered with approximately 10 ml of HPLC grade hexane (see Note 2). The vial is placed in a bath sonicator for 10 min. Drain the hexane and repeat. A third rinse using absolute ethanol should be conducted in the same manner. 2. After cleaning the shards are rinsed with DI water and blown dry with nitrogen (see Note 3). Shards are stored in airtight Fluoroware containers until use. 3. For adherent oligodendrocyte, the Si substrates must be added to the culture dish to allow adhesion. Rat oligodendrocytes are dislodged from the culture flask by agitation, most often by tapping gently against the side of the culture flask. The culture medium, containing dislodged cells, is drawn into a pipette. The solution is slowly added to a sterile culture dish containing the cleaned substrates. The dish is then incubated at 37.4 C for 30 min. 4. While the samples are in the incubator the liquid ethane for plunge-freezing is prepared. To collect liquid ethane, an ethane tank is fitted with a regulator and held in a ring stand. The regulator is attached to a length of flexible PVC tubing with a graduated pipette in the opposite end to create a small aperture for the hose. The pipette is placed in a LN 2 -cooled

5 Lipid Detection, Identification, and Imaging 89 beaker. The ethane is leaked into the beaker. The flow rate is adjusted to maximize flow without letting the ethane escape. It takes a few minutes to collect a sufficient amount of liquid ethane. When the desired amount is acquired, the beaker is removed from the liquid nitrogen. The ethane will last about 10 min. 5. After incubation the substrates are removed from the medium and quickly (3 s or less), rinsed in DI water. A top shard is placed directly on the substrate to create a sandwich. The arrangement is shown in Fig The excess water is wicked away using a Kimwipe (see Note 4). The sandwiches are then plunged in liquid ethane for approximately 5 s, or until the ethane stops boiling. Once the shard is frozen it is placed under LN 2 for storage until use. 6. T. thermophila are non-adherent cells. 200 μl of suspended cell solution is dropped onto a clean substrate. The sample is covered with a top shard by placing it directly on the sample. The excess water is wicked away using a Kimwipe. The sample is then plunge-frozen in liquid ethane for 5 s, or until the ethane stops boiling. The sample sandwich is then placed under LN 2 for storage until use (3). Fig. 4.1 Sample on silicon shards. (a) Silicon substrate with sample. (b) Silicon substrate with sample after freeze-fracture, showing the exposed area. (c) Schematic diagram of the freeze-fracture chamber attached to the TOF-SIMS instrument. Thermocouple attached to the bottom of the vertical transfer rod is not shown. ((c) Reproduced from (12) with permission from the American Chemical Society.) 3.3. Sample Transfer into the Vacuum Environment 1. Before entering the sample into the mass spectrometer, all of the components that contact the sample must be cooled (see Note 5). The instrument has five dewars connected to it. The instrument is cooled by flowing gaseous N 2, at a pressure of 40 psi, through copper tubing that is immersed in the dewars. All five dewars are filled with LN 2 and the temperature is monitored using thermocouple feedthroughs attached

6 90 Heien et al. to the cooled components. In addition, the instrument has a sorption pump that evacuates the fast entry port. This pump must be cooled to LN 2 temperature prior to introducing the sample. 2. The sample must be mounted onto the analysis block under LN 2. A London finder grid is first attached to the sample block using conductive silver paste. Then, a Styrofoam cooler is filled with LN 2 and the sample block, samples, and fast entry apparatus are placed in the cooler. When the LN 2 stops boiling, the sample can be mounted. 3. To mount the sample, a shard is fit in the recessed sample area with the top shard facing out. Two screws with washers are used to affix the sample to the block (see Note 6). The sample block is mounted on the entry port post and fit inside the metal sheath. The sheath is then screwed onto the fast entry port arm, while keeping the sample assembly under nitrogen. 4. Introducing the sample must be done as quickly as possible to minimize warming as well as condensation. The arm assembly is quickly inserted into the entry port, and the port is evacuated using the sorption pump for 30 s. After 30 s, the sample is placed into the instrument and quickly transferred to the pre-cooled sample transfer arm (Fig. 4.2c) In Vacuo Freeze-Fracture 1. The sample is then transferred to the freeze-fracture stage. The temperature and pressure are crucial during the fracture process (13). It is usually necessary to wait several minutes for the vacuum to recover after introducing the sample. 2. To be able to fracture the sample, it is important that the block is mounted with the fracture shard extending above the substrate on the side of the fracture knife. The knife blade is positioned above the top shard. In the z-direction, the blade is positioned such that it will fall between the top shard and the sample block when brought down. 3. When the fracture stage reaches the desired temperature of C, the sample is fractured by turning the blade counterclockwise with a quick flick of the wrist (see Note 7). This should result in the top shard snapping off. It is important to design a cold stage to catch the shard to prevent water desorption changing the vacuum and depositing ice on the sample. 4. After the fracture, the blade is withdrawn from the sample approximately 5 mm and positioned directly in front of the sample. The knife blade acts as a cold trap, because it is colder than the sample, catching excess water vapor caused from the fracture.

7 Lipid Detection, Identification, and Imaging A B C D E F 91 Fig. 4.2 (a) Brightfield image of rat oligodendrocyte in culture. (b) Brightfield image of an oligodendrocyte cultured on silicon and cryogenically preserved using the described sample preparation method. (c) SIM (secondary ion microscopy) image of rat oligodendrocyte. (d) SIMS image of a preserved oligodendrocyte, m/z 184 (PC) green, m/z 28 (Si) blue, intensity scale 0 4. (e) Brightfield image of oligodendrocyte that has been warmed to allow crystallization of water. (f) SIMS image of a corresponding area shown is in (e), m/z 184 (PC) green, m/z 28 (Si) blue, intensity scale 0 2. Images were obtained using an In+ primary ion beam; all scale bars are 20 μm. (Reproduced from (8) with permission from the American Chemical Society.) 5. When the preparation chamber pressure has recovered from the fracture, the sample can then be transferred to the sample analysis chamber Preparation for Imaging 1. In the case of adherent cells it is useful to do a freezeetch step prior to imaging (8). This is achieved by removing the dewar from the sample stage cooling line. This will result in the sample stage warming at a rate of approximately 5 C/min. The stage is warmed to 80 C to remove

8 92 Heien et al. excess surface water and then quickly returned to LN 2 temperature. 2. The sample beam is focused. A standard sample, a London finder grid, attached to the sample block is used to obtain the desired focus. The primary ion beam is focused to a point by two electrostatic lenses in series. The focal point of these lenses can be manipulated by increasing and decreasing the potential that is applied to them. The London finder grids have distinct features of various sizes and are made to high specifications. The focal point of the beam is manipulated until the user can visually distinguish features of the same size scale as they intend to measure. The interface width (the width measured at which the signal decreases from 85 to 15%) of a sharp interface is a good estimate of the primary ion beam spot diameter. For optimal imaging, this width should be less than or equal to the width of the pixels to be used for imaging. 3. Before cell imaging, the SIMS primary ion beam must be aligned with the microscope mounted on the instrument so that cells to be imaged can be identified and positioned in front of the primary ion beam. To do this, a digital micrograph of the grid is taken using the camera on the microscope. Then a SIMS image of the grid is obtained. The two images are overlaid to determine the imaging area in the eyepiece. 4. Mass calibration of the mass spectrum is important to obtain the proper peak assignments. Calibration can be achieved by using the spectrum obtained from the SIMS image of the grid. A good three-point calibration can be obtained, from Na +,Cu +,andag +.TheAg + signal comes from the conductive silver paste that is used to mount the grid to sample block Collecting Images 1. When collecting SIMS images the static limit should be used to insure a reliable representation of the surface. The static limit is defined as ions/cm 2. To calculate the limit you must know the pulse width, primary ion current, field-of view, and the pixel size. A discussion of each of these is found below. 2. Primary ion current to measure the primary ion current, a picoammeter is connected to the analysis stage through the stage voltage feedthrough while the beam is in the DC mode. Higher primary ion currents result in faster image collection times, as the static limit is reached faster. However, higher currents are obtained at the expense of increase in spot size, which leads to decreased spatial resolution. 3. Spot size this is the diameter of the primary ion beam at the focal point; it is the ultimate limit to spatial resolution.

9 Lipid Detection, Identification, and Imaging 93 Smaller spot sizes can be obtained through rigorous beam alignment and generally result in lower primary ion currents. 4. Pixel size the pixel size is the field-of-view (FOV) divided by the amount of pixels chosen for the image (see Note 8). Choosing fewer pixels results in quicker acquisition times at the expense of spatial resolution. 5. Mass range the mass range that is collected for an experiment depends on what the analyst is looking for. Traditionally the SIMS analyst collects from 10 to 1,000 amu Data Analysis 1. The first step in data analysis is producing mass-specific images. Images are produced by selecting the mass of interest from the mass spectrum and plotting the intensity, the number of counts per pixel, versus the spatial position. This produces an x y plot of pixels, which corresponds to the distribution of particular atoms or molecules in the specimen. Signal intensity is conveyed through the use of a false-color scale. A table of characteristic fragments for common lipids is given in Section 3.8. Because of the high-energy nature of the SIMS ionization process, these fragments are generally of low mass < 200 Da. In addition, lipid species, with the exception of cholesterol, are identified by the lipid headgroup fragments. 2. Localization because of the complexity and difficulty of the sample preparation, images must be analyzed to verify that the cells are properly preserved. Figure 4.2 contains both optical and SIMS images of cells which have been freeze-etched. These contain well-prepared cells (Fig. 4.2a d) and cells that have been damaged during the process (Fig. 4.2e f). The first check to perform on cells is to examine the localization of cellular fragments, i.e., phosphocholine (PC), phosphoethanolamine (PE), cholesterol, sphingomyelin SM, and m/z These fragments should be found in the cell region and only in the cell region. Cell size and shape can be gleaned from brightfield images taken before and after analysis. Anti-localization is also important in verifying cell preservation. Sodium and silicon should be anti-localized to the cell, meaning that these ions should only be found surrounding the cell and not in it. Anti-localization of potassium, however, is indicative of a ruptured cell. 3. Line scans are used for the relative quantification of the distribution of lipids. A line scan is a plot of the signal intensity for a given ion, as a function of its lateral position on a line, drawn by the analyzer. The width and length of the line are chosen based on the feature to be displayed. This yields a method that can be used to quantify data. Figure 4.3 contains line scans taken on a pair of mating

10 94 Heien et al. Fig Line scans of the molecule-specific images of Tetrahymena graphically support the idea that PC decreases at the conjugation junction between mating cells. Data points were collected every 120 nm. (a) Line scan for m/z 69 through the conjugation junction, illustrating that the total lipid content is relatively constant across the mating cells. The inset shows the SIMS image for m/z 69, highlighting the pixels used for the line scan. (b) Line scan for m/z 184 through the junction, demonstrating a sharp decrease in signal at the conjugation junction. The inset shows the SIMS image for m/z 184, highlighting the pixels used for the line scan. (Reproduced from (3) with permission from the Association for the Advancement of Science.) cells (T. thermophila). They demonstrate the formation of a PC-depleted domain in the region where the two cells join (Fig. 4.3b). The absence of this domain in the line scan of the ubiquitous hydrocarbon fragment (m/z 69, Fig. 4.3a) indicates that the cell membranes are continuous in this region. 4. The signal intensity of a given ion is affected by many factors including surface concentration, topography, sample matrix, and primary ion beam current fluctuations. To account for these factors a pseudo-internal standard is used to standardize line scans, C 5 H + 9 or m/z 69. This ion is derived from the acyl chains of phospholipids and thus is constant across the cell. The intensity distribution can be used to demonstrate that intensity fluctuations are due to surface concentration as opposed to the other factors mentioned above.

11 Lipid Detection, Identification, and Imaging 95 Standardization is achieved by dividing the intensity of the ion being analyzed by the intensity of m/z 69, for each pixel Lipid Identification 1. Owing to the high-energy nature of SIMS, molecular ions for lipids have been less often observed to date. Lipids can often be identified, however, by identification of specific fragment ions. Fragment ions are determined by generating standard films of the lipid of interest, and then collecting the resultant mass spectra (5). Mass spectra collected from a biological specimen can then be compared to the mass spectra for the fragmentation of a lipid standard for identification. 2. Table 4.1 (5) contains common fragment ions for some lipids. These can be used to identify lipids present on the cell surface (see Note 9). Table 4.1 Fragment ions for common lipids Lipid Fragment Mass (m/z) Phosphatidylcholine C 5 H 12 N Phosphatidylcholine C 5 H 15 NPO Phosphatidylcholine C 8 H 19 NPO Phosphatidylethanolamine C 2 H 7 NPO Phosphatidylethanolamine C 2 H 9 NPO Phosphatidylethanolamine C 2 H 7 NPO Phosphatidylglycerol C 3 H 9 PO 6 Na Phosphatidylinositol C 6 H 10 PO Phosphatidylinositol C 6 H 12 PO Phosphatidylinositol C 9 H 16 PO Phosphatidylinositol C 3 H 8 NPO 6 Na Cholesterol C 27 H Cholesterol C 27 H 46 O Sulfatide C 6 H 9 SO 8 Na Notes 1. Cleaning the substrates is crucial, as SIMS is very sensitive to contaminants, which can interfere with signals of interest as well as internal standard calibration.

12 96 Heien et al. 2. Glass Scintillation vials should not be capped as solvent vapor can extract polymers from the cap, contaminating the solvent. 3. Once the shards have been cleaned they should only be handled by the edges using solvent-rinsed (acetone) tweezers. 4. Wicking away the excess water after the top shard is applied, greatly increases the likelihood of a successful fracture. 5. It is good to cover the styrofoam cooler containing the sample block, samples, and fast entry apparatus whenever you are not working with the sample to prevent condensation, which can damage the vacuum chamber. In addition, it is important to insure that the LN 2 completely covers all cooled parts; otherwise any exposed metal will collect ice quickly. 6. When tightening the screws holding the silicon substrate, it is important to stop as soon as resistance is met. Over tightening will result in broken substrates. For screwing the sample to block, we acquired a small spring-loaded grabber, used by computer technicians, for holding the screw and washer together under LN The fracture temperature is critical to obtaining a good fracture. If the sample is too cold, condensation dominates, leading to an ice-covered sample; too warm and sublimation dominates. If a sample has been fractured too warm, there will be little or no visible ice on the surface and SIMS images display significant smearing of lipids. In the case of a cold fracture, there is extensive ice visible on the surface and the SIMS spectrum will be dominated by water clusters. 8. To maximize lateral resolution and obtain the most signal, it is good to match the spot size to the pixel size. The data can later be down-binned to obtain the necessary signal intensity. This is done off-line by combing adjacent pixels and summing their mass spectra. 9. Some contaminants may interfere with the signals listed in the table. You must ensure these are not present in the sample. Acknowledgments This work was supported by the National Institutes of Health (R01EB002016). A.G.E. is supported as a Marie Curie Chair from the European Union 6th Framework.

13 Lipid Detection, Identification, and Imaging 97 References 1. Johansson, B. (2006) ToF-SIMS imaging of lipids in cell membranes. Surf Interface Anal, 38, Roddy, T. P., Donald M. Cannon, J., Meserole, C. A., Winograd, N., Ewing, A. G. (2002) Imaging of freeze-fractured cells with in situ fluorescence and time-of-flight secondary ion mass spectrometry. Anal Chem, 74, Ostrowski, S. G., Bell, C. T. V., Winograd, N., Ewing, A. G. (2004) Mass spectrometric imaging of highly curved membranes during tetrahymena mating. Science, 305, Sostarecz, A. G., McQuaw, C. M., Ewing, A. G., Winograd, N. (2004) Phosphatidylethanolamine-induced cholesterol domains chemically identified with mass spectrometric imaging. J Am Chem Soc, 126, Ostrowski, S. G., Szakal, C., Kozole, J., Roddy, T. P., Xu, J. Y., Ewing, A. G., Winograd, N. (2005) Secondary ion MS imaging of lipids in picoliter vials with a buckminsterfullerene ion source. Anal Chem, 77, Ostrowski, S. G., Kurczy, M. E., Roddy, T. P., Winograd, N., Ewing, A. G. (2007) Secondary ion MS imaging to relatively quantify cholesterol in the membranes of individual cells from differentially treated populations. Anal Chem, 79, McQuaw, C. M., Zheng, L. L., Ewing, A. G., Winograd, N. (2007) Localization of sphingomyelin in cholesterol domains by imaging mass spectrometry. Langmuir, 23, Piehowski, P. D., Kurczy, M. E., Willingham, D., Parry, S., Heien, M. L., Winograd, N., Ewing, A. G. (2008) Freeze-etching and vapor matrix deposition for ToF-SIMS imaging of single cells. Langmuir, 24, Piehowski, P. D., Carado, A. J., Kurczy, M. E., Ostrowski, S. G., Heien, M. L., Winograd, N., Ewing, A. G. (2008) MS/MS methodology to improve subcellular mapping of cholesterol using TOF-SIMS. Anal Chem, 80, Kurczy, M. E., Piehowski, P. D., Parry, S., Jiang, M., Chen, G. A., Ewing, A. G., Winograd, N. (2008) Which is more important in bioimaging SIMS experiments- the sample preparation or the nature of the projectile? Appl Surf Sci, 255, Fletcher, J. S., Rabbani, S., Henderson, A., Blenkinsopp, P., Thompson, S. P., Lockyer, N. P., Vickerman, J. C. (2008) A new dynamic in mass spectral imaging of single biological cells. Anal Chem, 80, Colliver, T. L., Brummel, C. L., Pacholski, M. L., Swanek, F. D., Ewing, A. G., Winograd, N. (1997) Atomic and molecular imaging at the single cell level with ToF-SIMS. Anal Chem, 69, Cannon, D. M. J., Pacholski, M. L., Winograd, N., Ewing, A. G. (2000) Molecule specific imaging of freezefractured, frozen hydrated model membrane systems using mass spectrometry. JAmChem Soc, 122,

14