A New Dynamic in Mass Spectral Imaging of Single Biological Cells

Oct 24, 2008 - (31) The secondary ions undergo half a period of simple harmonic motion in the analyzer before impacting the detector with the same tim...
2 downloads 4 Views 382KB Size
Download PDF
Anal. Chem. 2008, 80, 9058–9064

A New Dynamic in Mass Spectral Imaging of Single Biological Cells John S. Fletcher,*,† Sadia Rabbani,† Alex Henderson,† Paul Blenkinsopp,‡ Steve P. Thompson,§ Nicholas P. Lockyer,† and John C. Vickerman† Manchester Interdisciplinary Biocentre, School of Chemical Engineering and Analytical Science, University of Manchester, Manchester, U.K., Ionoptika Ltd., Southampton. U.K., and Scientific Analysis Instruments, Manchester, U.K. Time-of-flight secondary ion mass spectrometry (TOFSIMS) has unique capabilities in the area of high-resolution mass spectrometric imaging of biological samples. The technique offers parallel detection of native and nonnative molecules at physiological concentrations with potentially submicrometer spatial resolution. Recent advances in SIMS technology have been focused on generating new ion sources that can in turn be used to eject more intact molecular and biological characteristic species from a sample. The introduction of polyatomic ion beams, particularly C60, for TOF-SIMS analysis has created a whole new application of molecular depth profiling and 3D molecular imaging. However, such analyses, particularly at high lateral resolution, are severely hampered by the accompanying mass spectrometry associated with current TOF-SIMS instruments. Hence, we have developed an instrument that overcomes many of the drawbacks of current TOF-SIMS spectrometers by removing the need to pulse the primary ion beam. The instrument samples the secondary ions using a buncher that feeds into a specially designed time-of-flight analyzer. We have validated this new instrumental concept by analyzing a number of biological samples generating 2D and 3D images showing molecular localization on a subcellular scale, over a practical time frame, while maintaining high mass resolution. We also demonstrate large area mapping and the MS/MS capability of the instrument. Time-of-flight secondary ion mass spectrometry (TOF-SIMS) has proved itself to be a very effective technique for surface chemical characterization of many different types of materials. The mass spectral capability, linked in recent years with high spatial resolution liquid metal ion sources, has enabled its application to many demanding analytical problems.1-3 The study of biological systems offers many analytical challenges that would benefit from * To whom correspondence should be addressed. E-mail: John.Fletcher@ manchester.ac.uk. † University of Manchester. ‡ Ionoptika Ltd. § Scientific Analysis Instruments. (1) Kettle, S.; Chater, R. J.; Graham, G. A.; McPhail, D. S.; Kearsley, A. T. Appl. Surf. Sci. 2004, 231, 893–898. (2) Vickerman, J. C.; Briggs, D. TOF-SIMS: Surface Analysis by Mass Spectrometry Surface Spectra; Manchester and IM Publications: Chichester, 2001. (3) Oakes, A. J.; Vickerman, J. C Surf. Interface Anal. 1996, 24, 695.

9058

Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

this capability, and reports of TOF-SIMS analysis of cell and tissue samples are becoming ever more frequent.4-10 However, because the high-energy atomic primary ions such as Ga+, In+, and Au+ generated severe subsurface chemical damage, static analysis conditions of less than 1013 ions cm-2 had to be strictly adhered to, limiting impacts to 1% of the surface and thus ensuring pristine material was impacted by each primary ion. This meant that even using the time-of-flight analysers introduced in the 1980s to maximize secondary ion collection efficiency, the number of ions collected per unit area limited the useful spatial resolution to ∼1 µm.11 For biological analysis, these limitations were compounded by the fact that the yield of bioorganic molecular ions above m/z 300 was very small. Around the turn of the century, practical metal cluster and polyatomic ion sources appeared that provided significantly greater ion yields, particularly of higher mass molecular ions.12-16 The metal cluster ions, Aux+ and Bix+ (x ) 1-7), also offered similar high spatial resolution capabilities to their atomic ions.17 However, because on impact they fragment into high-energy atomic ions, the subsurface chemical damage effects still occur and the static limit still has to be adhered to. The 1 µm useful spatial resolution is reduced a little, but not much.11 The polyatomic ions such as C60+ introduced at the same time (and to a lesser degree earlier work using SF5+) not only greatly increase higher mass ion yield, they also generate a great (4) Ostrowski, S. G.; Van Bell, C. T.; Winograd, N.; Ewing, A. G. Science 2004, 305, 71–73. (5) Sjovall, P.; Lausmaa, J.; Johansson, B. Anal. Chem. 2004, 76, 4271–4278. (6) Broersen, A.; van, L. R.; Altelaar, A. F. M.; Heeren, R. M. A.; McDonnell, L. A. J. Am. Soc. Mass Spectrom. 2008, 19, 823–832. (7) Heeren, R. M. A.; Ku ¨ krer-Kaletas¸, B.; Taban, I. M.; MacAleese, L.; McDonnell, L. A. Appl. Surf. Sci. In press. (8) Eriksson, C.; Maimberg, P.; Nygren, H. Rapid Commun. Mass Spectrom. 2008, 22, 943–949. (9) Magnusson, Y.; Friberg, P.; Sjovall, P.; Dangardt, F.; Malmberg, P.; Chen, Y. Clin. Physiol. Funct. Imaging 2008, 28, 202–209. (10) Lee, T. G.; Park, J. W.; Shon, H. K.; Moon, D. W.; Choi, W. W.; Li, K.; Chung, J. H. Appl. Surf. Sci. In press. (11) Jones, E. A.; Lockyer, N. P.; Vickerman, J. C. Int. J. Mass Spectrom. 2007, 260, 146–157. (12) Davies, N.; Weibel, D. E.; Blenkinsopp, P.; Lockyer, N.; Hill, R.; Vickerman, J. C. Appl. Surf. Sci. 2003, 203- (204,), 223–227. (13) Wong, S. C. C.; Hill, R.; Blenkinsopp, P.; Lockyer, N. P.; Weibel, D. E.; Vickerman, J. C. Appl. Surf. Sci. 2003, 203 (204), 219–222. (14) Weibel, D.; Wong, S.; Lockyer, N.; Blenkinsopp, P.; Hill, R.; Vickerman, J. C. Anal. Chem. 2003, 75, 1754–1764. (15) Kotter, F.; Benninghoven, A. Appl. Surf. Sci. 1998, 133 (1-2), 47–57. (16) Gillen, G.; Roberson, S. Rapid Commun. Mass Spectrom. 1998, 12 (19), 1303–1312. (17) Tomboul, D.; Kollmer, F.; Niehuis, E.; Brunelle, A.; Lapre´vote, O. J. Am. Soc. Mass Spectrom. 2005, 16, 1608. 10.1021/ac8015278 CCC: $40.75  2008 American Chemical Society Published on Web 10/25/2008

deal less observable subsurface chemical damage, so that for many materials the static limit can be dispensed with.13,18 This is an enormous benefit. In principle, the whole of a sample can be used for analysis, greatly increasing sensitivity. Alternatively, molecular depth profiling can be carried out. Particular success using SF5+ has been demonstrated on biopolymers with drug delivery applications19-21 including the 3D imaging of pharmaceutical distribution within the polymer.22 C60+ depth profiling has been demonstrated on a wide range of samples including polymers (particularly PMMA),23 chemically alternating Langmuir-Blodgett films,24 biopolymer films,25,26 amino acids in ice,27,28 and peptides in trehalose.29 Dual-beam analysis using a well-focused liquid metal ion probe interleaved with polyatomic etch cycles has allowed 3D analysis with spatial resolution beyond 1 µm.30 New Instrumental Concept. To fully exploit the properties of these ion beams, our group in Manchester and the Winograd group at Penn State together concluded that radical changes must be made to the conventional TOF analyzer used in SIMS. For this reason at Penn State an AB Sciex QStar ortho-TOF instrument has been adapted with the incorporation of an Ionoptika C60 ion beam while we in Manchester have developed a radically new TOF-SIMS instrument using novel mass spectrometry. The principal drawback of conventional TOF-SIMS analyzers is the requirement to use short (ns) primary ion beam pulses. The ion beam is swept across a small aperture to generate the required pulse and may also be bunched. The generation of the very short pulses required for high mass resolution is incompatible with attaining the high spatial resolution of which the ion beam is capable. The very low duty cycle of the pulsed ion beam also means that imaging and depth profiling experiments can take an inordinately long time. This may be compounded by the reduction in ion beam current consequent upon aperturing and focusing the beam to a fine spot size. Thus, an analysis using a primary ion beam dose density of 1 × 1013 ions cm-2 (the static limit) from a 100 × 100 µm2 area with a focused ion beam providing 10 pA of target current using 50-ns primary ion pulses on an instrument running at 10 kHz would take almost 9 h. This problem is particularly severe if 3D molecular imaging is to be performed as is now possible using polyatomic beams. A sequence of images with an accumulated ion dose of 1 × 1015 ions cm-2 using the same parameters as above would require over a month! If the (18) Kotter, F.; Benninghoven, A. Appl. Surf. Sci. 1998, 133, 47–57. (19) Mahoney, C. M.; Roberson, S. V.; Gillen, G. Anal. Chem. 2004, 76, 3199– 3207. (20) Wagner, M. S.; Castner, D. G. Appl. Surf. Sci. 2004, 231- (2,), 366–376. (21) Mahoney, C. M.; Roberson, S.; Gillen, G. Appl. Surf. Sci. 2004, 231 (2), 174–178. (22) Gillen, G.; Fahey, A.; Wagner, M. S.; Mahoney, C. M. Appl. Surf. Sci. 2006, 252, 6537. (23) Mollers, R.; Tuccitto, N.; Torrisi, V.; Niehuis, E.; Licciardello, A. Appl. Surf. Sci. 2006, 252, 6509–6512. (24) Zheng, L. L.; Wucher, A.; Winograd, N. J. Am. Soc. Mass Spectrom. 2008, 19, 96–102. (25) Fletcher, J. S.; Conlan, X. A.; Jones, E. A.; Biddulph, G.; Lockyer, N. P.; Vickerman, J. C. Anal. Chem. 2006, 78, 1827–1831. (26) Shard, A. G.; Brewer, P. J.; Green, F. M.; Gilmore, I. S. Surf. Interface Anal. 2007, 39, 294–298. (27) Wucher, A.; Sun, S. X.; Szakal, C.; Winograd, N. Anal. Chem. 2004, 76, 7234–7242. (28) Conlan, X. A.; Lockyer, N. P.; Vickerman, J. C. Rapid Commun. Mass Spectrom. 2006, 20, 1327–1334. (29) Cheng, J.; Winograd, N. Anal. Chem. 2005, 77, 3651–3659. (30) Breitenstein, D.; Rommel, C. E.; Mo ¨llers, R.; Wegener, J.; Hagenhoff, B. Angew. Chem., Int. Ed. 2007, 46, 5332–5335.

Figure 1. Illustration of the Ionoptika J105 3D Chemical Imager (a). Sample insertion can be performed under an inert atmosphere using the glovebox to prevent frosting of frozen hydrated samples. Schematic of the mass spectrometry showing the coupling of the buncher to the harmonic reflectron (b). A section (∼0.3 m) of the continuous secondary ion stream is bunched to a time focus and accelerated into the reflectron. The collision cell for dissociation during MS/MS experiments is also labeled.

mass spectrometry can be decoupled from the sputtering event, then dc beams could be used without any loss in mass resolution and would reduce the above analysis times to 16 s and 30 min, respectively. The principal aim of the development was to enable a continuous primary ion beam to be used while maintaining the advantages offered by time-of-flight analysers. With this in mind, our group at the University of Manchester entered into collaboration with Ionoptika Ltd. (Southampton, UK) and SAI Ltd. (Manchester, UK). The resulting instrument, the Ionoptika J105 3D Chemical Imager, combines our extensive experience in TOFSIMS along with technological expertise in sample handling, the ion beam systems of Ionoptika, and the mass spectrometric knowhow of SAI. It exploits a unique linear buncher in conjunction with a harmonic reflectron TOF analyzer (Figure 1). A continuous supply of secondary ions is generated from a sample using a 40 keV C60 primary ion beam operated in dc mode. The secondary ions are collisionally cooled in an rf-only quadrupole filled with a suitable gas (e.g., N2) and are then energy filtered by an electrostatic analyzer. This process provides secondary ions with a 1 eV energy spread for injection into a linear buncher. The buncher is filled with a portion of secondary ions ∼0.3 m long. Next, the buncher fires by suddenly applying an accelerating field that varies from 7 kV at the entrance of the buncher to 1 kV at the exit. This creates a time focus at the entrance of the TOF analyzer. The ultimate mass resolution is now dependent on the quality of this focus and independent of the sputtering event or sample topography. Due to the acceleration in the buncher, the ions now have a 6 keV energy spread. A harmonic field TOF reflectron is required and employed such that the path of the ions Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

9059

Figure 2. 3D reconstruction of image data from BPH cells. Isosurface rendering shows the distribution of signal in 3D through the entire sample while orthogonal slices through the data set facilitate visualization of the chemical distribution within selected cells. Data shown for the m/z 136 ion (left), the protonated molecular ion from adenine, and m/z 184 (right) originating from phosphocholine-containing lipids. The adenine signal is localized to the center of the cells (A) as it arises from the nuclear DNA while the m/z 184 signal is observed almost as rings around the edge of the cell (B) as it arises from the lipid membrane. A larger view of the orthogonal slice through a single cell is also shown for clarity. Field of view for the analysis was approximately 180 × 180 µm2, 128 × 128 pixels; 16 images were acquired as the analysis progressed through the cells.

is dependent only on the mass and charge, not the energy, of the secondary ions.31 The secondary ions undergo half a period of simple harmonic motion in the analyzer before impacting the detector with the same time spread as the focus from the buncher. The result is an imaging SIMS instrument offering the parallel mass detection and mass range of a TOF analyzer with all the advantages of using continuous primary ion beams. Further, for the first time on a TOF-SIMS instrument, nanoscale imaging is possible without sacrificing mass resolution. The mass spectrometry is decoupled from the sputtering of the sample; consequently mass calibration is not required for every spectrum and mass accuracy of 5 ppm is readily observed. The instrument is currently fitted with a 40-kV C60 ion gun capable of delivering a wien filtered beam of C60 with energy up to 120 keV by selecting the C603+ ion and can be focused to deliver an ultimate spot size of 200 nm. Analysis of biological samples including cells and tissue is of great interest and an expanding area in SIMS. One of the principal difficulties encountered is that of sample handling within the vacuum required for SIMS analysis. The J105 has been designed to allow easy sample insertion and transfer of biological samples that have been flash frozen to maintain their biological integrity. Frozen samples are transferred to a purged glovebox mounted over the sample insertion port. There is a sample stage in the preparation chamber of the instrument that during sample insertion rises to a load lock. The samples are placed directly onto this stage during insertion, and in the case of frozen samples, this stage can be precooled to temperatures as low as 100 K. The samples can be rapidly transferred to the sample stage in the main analysis chamber that can also be heated and cooled. A further design remit of the instrument was that it must be able to perform MS/MS analysis for both fundamental studies related to sputtering and secondary ion formation and also for molecular identification. In the J105, the MS/MS is performed in a TOF-TOF configuration. Ions exiting the buncher pass through a cell containing a suitable collision gas (e.g., helium, nitrogen, or argon). Because the collisions take place following the buncher, collision energies are in the range of 1-7 kV. All the species become fragmented, but since the collisions take place in a fieldfree region of the instrument, the parent ion and associated daughter ions continue to travel with the same velocity. Following 9060

Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

a short TOF region, a timed ion gate is used to select the parent-daughter combination of interest and admit these ions into the TOF analyzer. In this paper, we present preliminary data that have been generated during the continued development of the instrument. Despite the work in progress nature of the instrument development, the proof of concept is clear and the initial results show the clear benefits associated with the new instrumental style. RESULTS AND DISCUSSION Cell Analysis. One of the major goals in the development of this instrument is to fully realize the potential of SIMS in cellular imaging. Although this is a very active area of research with TOFSIMS, the majority of reports to date suffer from the limitations of pulsed primary ion beam modality and seldom show any subcellular localization although sample modification methods including addition of silver have shown potential benefits.32,33 The samples presented in this paper have been selected to explore the potential of this new instrumental paradigm: benign prostatic hyperplasia (BPH) cells, HeLa cells, and human cheek cells. The BPH cells clearly demonstrate the ability to generate 3D molecular images of cells with subcellular localization of characteristic fragments. Analysis was performed using a 40 kV primary ion beam providing 3 pA of continuous current onto the sample; 128 × 128 pixel images were acquired over an ∼180 × 180 µm2 area. Each image was generated in just under 6 min with a primary ion dose density of 2 × 1013 ions/cm2 with 16 images being generated in total. Figure 2 shows isosurface rendering of the BPH cell data, illustrating the distribution of the cells on the substrate, along with orthogonal sections through the data (generated using AVS Express, Advanced Visualization Systems Inc.). The adenine, m/z 136, is located within the center of the cell, i.e., the nucleus, whereas the lipid signal from the phosphocholine headgroup, m/z 184, is present in higher intensities in pseudospheres around a volume of low intensity where the maximum adenine signal from the nucleus is observed. Such (31) Makarov, A. A.; Raptakis, E. N.; Derrick, P. J. Int. J. Mass Spectrom. Ion Processes 1995, 146, 165–182. (32) Parry, S.; Winograd, N. Anal. Chem. 2005, 77, 7950–7957. (33) Nygren, H.; Malmberg, P. J. Microsc. 2004, 215 pt2, 156–161.

analysis provides data that can clearly complement that obtained using confocal microscopy techniques. The analysis of the HeLa cells demonstrates a second, unique mode of operation where this time the entire cell is consumed during a single image analysis. The images here are more comparable to a bright-field microscope image except with the label-free chemical specificity associated with TOF-SIMS. The HeLa cells show a morphology similar to the BPH cells with the characteristic lipid species surrounding fragments from nuclear DNA. This data set however was acquired over a much smaller field of view (88 × 108 µm2) again using a 40 kV C60+ ion beam scanned over 128 × 128 pixels; each cell is ∼25-30 µm in diameter. Previous studies have demonstrated that multivariate analysis methods can provide a useful tool for discriminating between TOF-SIMS data for either classification of proteins based on amino acid fragment variation or identification of microorganisms.34,35 In the case of TOF-SIMS image analysis, such analyses can be used to improve image contrast, to identify regions of interest, and in each case to then output the associated variables (m/z values) for chemical identification.36 Comparative studies of various preprocessing procedures for principal components analysis along with maximum autocorrelation factor (MAF) analysis have indicated that for SIMS imaging MAF is potentially the best for reducing the number of variables required to describe the image, enhancing image contrast, and recovering key spectral features, although is more computationally intense.37 Panels a and b in Figure 3 show a result of MAF analysis of the TOF-SIMS image data HeLa cells. Factor 4 (displayed) captures the variation in signal between the nuclear region of the cell and the extranuclear material particularly the cell membrane. Inspection of the loadings shows clear contributions from different peaks that can then be imaged individually from the raw data (Figure 3c and d). The key new instrumental capabilities highlighted by these data are first that we are now imaging using a C60+ ion beam to detect intact molecular species on a subcellular scale. Second, unlike much of the previous work on this lateral scale, we are not simply detecting a lipid signal. This is because in this case the 2D image has been generated by consuming the entire cell. Normal static TOF-SIMS would not detect the nuclear material beneath the surface. Finally, the need for high mass resolution during imaging is highlighted using the HeLa cell data; three different species producing three clearly contrasting images can be generated from nominal m/z 86 (Figure 3, e-g). The low-mass peak is assigned as an isotope of Si3+ (m/z 85.93, the major peak appearing at m/z 83.93) from the silicon substrate while the highest mass peak of the three is attributed to C5H12N+ (m/z 86.10), a characteristic fragment of phosphocholine lipids and also the immonium ion of leucine. The m/z 86 spectral region is displayed in Figure 3g with ion arrival time as the abscissa; the 20-ns scale bar indicates clearly that to achieve the same mass resolution in conventional TOF-SIMS primary ion pulses shorter than this value would be required. (34) Fletcher, J. S.; Henderson, A.; Jarvis, R. M.; Lockyer, N. P.; Vickerman, J. C.; Goodacre, R. Appl. Surf. Sci. 2006, 252 (19), 6869–6874. (35) Wagner, M. S.; Castner, D. G. Appl. Surf. Sci. 2004, 231, 366–376. (36) Lee, J. L. S.; Gilmore, I. S.; Seah, M. P. Surf. Interface Anal. 2008, 40 (1), 1–14. (37) Tyler, B. J.; Rayal, G.; Castner, D. G. Biomaterials 2007, 28 (15), 2412– 2423.

Figure 3. TOF-SIMS analysis of HeLa cells cultured onto poly(Llysine)-coated silicon wafers. Images are 88 × 108 µm2. Result of MAF analysis highlights differences in chemistry within the image. Scores and associated loadings from MAF 4 (a and b) highlight two distinct chemistries within the cells. Single ions of interest can then be selected and imaged for example m/z 136 (assigned to adenine) (c) and 184 (the phosphocholine headgroup from membrane lipids) (d). The three images, (e-g) demonstrate the necessity of maintaining high mass resolution while imaging as all three are from nominal mass to charge value 86 with distinct differences in localization. Scale bar on the (time) spectrum relates to the arrival time of the ions and indicates that similar resolution in a standard TOF-SIMS instrument would require 10 keV) C60 is known to deposit on top of “hard” samples, thus rendering the depth profile experiment unfeasible although in some cases depth profiling begins successfully but the sputtering “turns off” at some point and then carbon deposition begins. Shard et al.39 have shown that increasing the energy of the C60 beam extends the time (or depth) before the deposition begins although at higher energy a concomitant decrease in depth resolution is noted. Gillen observed the formation of up to 6 µm-scale topographical features when etching silicon with 10 keV C60.40 The Winograd group have reported more successful profiling of a nickel/ chromium multilayer NIST standard sample where the sample was successfully profiled using 10, 15, and 20 keV C60+ compared to Ga+ where a profile could not be obtained.41 Similar to Shard et al., a reduction in depth resolution was observed with increasing beam energy and also as a function of etching time. We have performed depth profiles on the same NIST standard using our 40 keV C60+ ion beam; the result of one such analysis in shown in Figure 5, with very pleasing results. The interface width based on the first falling edge of the m/z 58 signal is measured (84-16%) between 3 and 4 nm. This depth is close to the predicted crater depth from the C60 impact based on molecular dynamic simulations.42 As long as the crater walls are avoided in the profile analysis there is little broadening of the interface width; the interface width at the end of the profile is ∼5 nm. Further work is now required to establish depth resolution on organic and biological samples. The sputter rate is expected, based on molecular dynamic simulations, to be much greater when probing organic surfaces and would therefore be expected to reduce the ultimate depth resolution under equivalent sputtering conditions simply based on the crater size from the C60 impact.43,44 Manufacturing reliable and reproducible organic depth profiling standards has proved difficult although success has recently been reported by the National Physical Laboratory (NPL) for a multilayer system comprising alternating layers of Irganox 1010 and Irganox 3114.39 The advent of such standards will allow optimization of parameters for molecular depth profiling as well as interlaboratory comparison and standardization. CONCLUSION We have developed a potentially revolutionary new instrument for secondary ion mass spectrometry. The instrument capitalizes on the advances in ion beam technology, particularly the advent of polyatomic ion beams such as C60. The instrument offers higher throughput with much reduced analysis times, which allow 3D images to be generated on a practical time scale. Due to the (39) Shard, A. G.; Green, F. M.; Brewer, P. J.; Seah, M. P.; Gilmore, I. S. J. Phys. Chem. B 2008, 112, 2596–2605. (40) Gillen, G.; Batteas, J.; Michaels, C. A.; Chi, P.; Small, J.; Windsor, E.; Fahey, A.; Verkouteren, J.; Kim, K. J. Appl. Surf. Sci. 2006, 252 (19), 6521–6525. (41) Sun, S.; Szakal, C.; Roll, T.; Mazarov, P.; Wucher, A.; Winograd, N. Surf. Interface Anal. 2004, 36 (10),), 1367–1372. (42) Postawa, Z.; Czerwinski, B.; Winograd, N.; Garrison, B. J. J. Phys. Chem. B 2005, 109 (24), 11973–11979. (43) Delcorte, A.; Garrison, B. J. Nucl. Instrum. Methods Phys. Res. B 2007, 255, 223–228. (44) Smiley, E. J.; Winograd, N.; Garrison, B. J. Anal. Chem. 2007, 79 (2), 494– 499.

Figure 5. Large -area imaging of a Xenopus blastomer following embedding in fish agar and cryosectioning. Images comprise an 8 × 8 array of 32 × 32 pixel images covering a total area of 896 × 1096 µm2. Cholesterol [M + H - H2O]+, m/z 369, and a range of peaks considered to arise from various phospholipids, m/z 313, 549, 576, and 601 are observed in the blastomer section. The spectra show the total ion signal from the analysis and the MS/MS analysis results for an intense peak at m/z 578 (inset) the daughter ions have been multiplied by 20 for clarity.

Figure 6. Results from the depth profiling of a Ni/Cr multilayer standard. AFM image of the etch crater following the analysis and a plot of the variation in signal from m/z 52 (52Cr+), 58 (58Ni+), and 28 (Si+). Erosion depth is calculated from the AFM analysis and assumes a constant sputter rate through the sample.

decoupling of the mass spectrometry from the ion generation process, mass resolution is no longer dependent on sample topography and is maintained during high-resolution imaging. The evidence from the cellular imaging confirms the success of this instrumental concept with high mass resolution subcellular imaging being demonstrated in two and three dimensions. MS/

MS capability is tested and provides a further tool for species identification. The instrument is still in the final development stages, which will significantly extend the routine mass range and increase the mass resolution, but these preliminary data show considerable promise. The potential applications for the instrument are vast Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

9063

with the label-free localization of non-native species such as pharmaceuticals within cellular samples being one of many targets. ACKNOWLEDGMENT We gratefully acknowledge Dr. Peter Gardner and Prof. Nancy Papalopulu, both of The University of Manchester, for providing the BPH cells and the blastomer samples. respectively; also Prof. Nicholas Winograd of Penn State University for providing the

9064

Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

NIST Ni/Cr standard sample. The work was funded by the Life Sciences Interface Initiative of the Engineering and Physical Sciences Research Council (EPSRC), UK.

Received for review July 21, 2008. Accepted September 29, 2008. AC8015278