Aim of this study is to provide a suitable preparation routine for systematic ToF-SIMS analysis and imaging experiments of lipids in (bone) cells. For systematic investigations the routine has to be highly reproducible but ToF-SIMS has further specific requirements being a vacuum technique. In summary, a suitable preparation routine has to be reproducible, vacuum compatible, leave the sample in a native like state concerning structure and molecular distributions, does not remove or delocalize any lipids and ideally, makes the sample storable.

In SIMS literature, freeze-fracturing is regarded as the "gold standard"¹. However, it seems to lack reproducibility² and is a method of high complexity impeding its use for systematic investigations with large sample numbers.

Thus, in this study, different alternative sample preparation routines for lipid analysis of human bone marrow stromal cells were investigated and evaluated.

Several chemical fixations were compared to freeze-drying procedures and cryofixations, including freeze-fracturing. For comparison, positive and negative spectra were collected with a TOF.SIMS 5 instrument (ION-TOF, Münster, Germany) using a pulsed 25 keV Bi₃⁺ primary ion beam with a primary ion dose below the static limit. Data evaluation was done by principal component analysis (PCA) in order to reveal differences between the preparation routes concerning their lipid content.

For freeze-fracturing, these investigations showed disadvantages in terms of reproducibility. In comparison, "simple" chemical fixation techniques via glutar- and/or paraformaldehyde and a freeze-drying procedure without chemical fixation but after plunge freezing proved advantageous concerning reproducibility and lipid content over other preparations while alcohol and osmium tetroxide application should be avoided.

The above mentioned sample preparation routines were also investigated in storage experiments in order to reveal an ideal storage procedure. Storage is always a crucial factor for cell experiments where the cells have to be harvested at the same time but analysis at once is impossible. We herein compare dry and wet storage conditions. Drying was achieved via freeze-drying, critical-point-drying and air-drying prior to measurement. Wet storage was carried out in either phosphate buffered saline or Milli-Q water.

1 M. E. Kurczy, P. D. Piehowski, S.A. Parry. M. Jiang, G. Chen, A. G. Ewing, N. Winograd Applied Surface Science 2008, 255, 1298-1304.

2 S. Vaidyanathan, Surf. Interface Anal. 2013, 45, 255-259.

Phospholipids (PLs) are major lipids of living cells that play important roles in biological membranes, protein sorting, regulation of biophysical properties and signaling pathways. PLs are classified by their head groups into phosphatidic acid (PA), Phosphatidylcholine (PC), phosphatidylglycerol (PG), phosphatidylinositol (PI), phosphatidylserine (PS) and cardiolipin (CL). In MALDI-TOF analysis, all PLs are detected at the positive ion mode, but in ToF-SIMS analysis PLs are detected at positive or negative ion modes due to different ionization efficiency. For example, PC and PS are generally detected as positive ions, and PE and PI as negative ions. Thus, metabolite analyses in ToF-SIMS need to perform tandem mass spectrometry measurements at both ion modes to identify unknown PLs.

For tandem mass spectrometry measurements in ToF-SIMS, a post source decay (PSD)-like method was successfully applied to identify several lipids by using cholesterol as a model molecule at the positive ion mode [1]. In our study, we adapted 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphor-rac-(1-glycerol) ammonium salt with well-known fragmentation pathways as a model molecule at the negative ion mode to identify PI lipids. By using the PSD-like method at both ion modes, we successfully identified PC and PI from MCF-7 breast cancer cell lysates and showed that the PSD-like method could be a useful method to identify unknown lipids from biological samples in ToF-SIMS.

[1] David Touboul et al. Rapid Commn. Mass Spectrom. 2006, 20, 703-709

Due to recent technical developments ToF-SIMS is more and more applied in the life sciences where sample preparation plays an eminent role for the quality of the analytical results.(1-3) This paper focusses on sample preparation of bone tissue and its impact on ToF-SIMS analysis. The analysis of bone is pivotal for the understanding of bone diseases and the development of replacement materials and new drugs for the cure of diseased bone. Main purpose of this paper is to find out which preparation process is best suited for ToF-SIMS analysis of bone tissue in order to obtain reliable and reproducible analytical results. Any embedding process has to ensure that interesting components are not rinsed out, remain at their original location and that no damage of the sample due to chemical reactions occurs. Nonetheless the embedding procedure will influence the sample, and it is important to evaluate this influence. It is major aim of this paper to judge this influence and to work out the optimal sample preparation route which allows combining the advantages of ToF-SIMS analysis with histological or histochemical investigations. Therefore we focus on the analysis of bone samples from human femoral heads which have been embedded in typical resins. All cut and ground sections of bone as also native samples without embedding for comparison were analysed by ToF-SIMS and evaluated using Principal Component Analysis (PCA). It is shown that epoxy resin as well as methacrylate based plastics (Epon and Technovit) as embedding materials do not infiltrate the mineralized tissue and that cut sections are better suited for the ToF-SIMS analysis than ground sections. In case of ground samples a resin layer is smeared over the sample surface due to the polishing step and overlap of peaks is found. Beside some signals of fatty acids in the negative ion mode, the analysis of native, not embedded samples does not provide any advantage. The influence of bismuth bombardment and O₂ flooding on the signal intensity of organic and inorganic fragments due to the variation of the ionization probability is additionally discussed. As C₆₀ sputtering has to be applied to remove the smeared resin layer, its effect especially on the organic fragments of the bone is analysed and described herein.

1. Fletcher JS. Latest applications of 3D ToF-SIMS bio-imaging. Biointerphases. 2015;10(1).

2. Vickerman JC, Briggs D. ToF-SIMS: Materials Analysis by Mass Spectrometry. 2nd ed: IM Publications LLP and Surface Spectra Limited; 2013.

3. Goodwin RJA. Sample preparation for mass spectrometry imaging: Small mistakes can lead to big consequences. Journal of Proteomics. 2012;75(16):4893-911.

Sample preparation for multimodal ToF-SIMS and NanoSIMS analysis of Drosophila brain tissue

Mai Hoang ¹; Nhu Phan², Per Malmberg^1. Andrew Ewing^1,2

¹ Department of Chemical and Biological Engineering, Chalmers University of Technology, Gothenburg, ²Department of Chemistry and Molecular Biology, University of Gothenburg, Gothenburg, Sweden

Secondary Imaging Mass Spectrometry (SIMS) is a powerful technique applied to biological imaging with high mass and lateral resolution. Because SIMS imaging requires ultrahigh vacuum conditions, sample preparation is a key issue to achieve reliable and reproducible analysis of biological materials. Ideally, sample preparation should preserve the structural, biological, and chemical distributions of the samples. Rapid fixation either by freezing or chemical fixation is needed to prevent loss and redistribution of compounds of interest. Chemical redistribution can also occur rapidly after sectioning and during the subsequent removal of water, dehydration step, before analysis in vacuum. The conventional methods for biological sample preparation recently used for analysis by mainly ToF-SIMS and MALDI imaging depend on cryosectioning and drying/freeze-drying while NanoSIMS analysis usually depends on fixation and embedding of tissue.

In this study we focus on developing sample preparation protocols that allow multimodal imaging of Drosophila brain tissue with both ToF-SIMS and NanoSIMS. Reproducible sectioning of the fly brain is obtained by using a fly collar, which keeps all the fly heads in the same orientation. Using aqueous chemical fixation and resin embedding to maintain tissue morphology produces samples that are suitable for analysis with both SIMS techniques. One of the advantages of the resin matrix embedding is to reduce the local charging effects which allows high spatial resolution imaging. In addition, the resin embedded technique renders a flatter sample surface, which results in the minimization of topological artifacts in the SIMS images. In the future, these fixation procedures can be improved not only for SIMS analysis but also for other imaging techniques.

Imaging of cells is important to be able to study different cell processes and structures such as migration, proliferation, differentiation and cell morphology. Chemical analysis of single cells can be used to detect molecular differences between cells in a population. Individual cells within a homogenous cell population have been shown to exhibit differences that can have a major impact on the function of the entire group of cells. Thus imaging of single cells has become of importance for the study of how small changes in one cell can effect entire populations of cells or even entire tissues/organs.

We have developed a preparation protocol in order to use a combination of different imaging techniques to provide both structural and targeted and non-targeted chemical information of the samples.

Here we have examined cells that have been transfected to overexpress a specific protein that has been shown to cause major morphological changes in cells. Fluorescence microscopy was used to identify transfected cells with the help of a GFP tag on the overexpressed protein. Mass spectrometry imaging (ToF-SIMS) was used to study changes in cell membrane composition after transfection on frozen hydrated and freeze-dried samples. Scanning electron microscopy was used to examine the change in cell morphology after transfection.

Preliminary data show that the sample preparation procedure was gentle enough to keep the cells intact while still removing salts and media residue. Transfected cells were successfully identified by correlating the ToF-SIMS images with the fluorescence images with the help of a localization grid on the samples.

Two different primary ion beams were used for analysis, 40 keV C₆₀ and 40 keV Ar₄₀₀₀. The Ar₄₀₀₀ showed increased signal for higher mass secondary ions compared to the C₆₀ but with reduced spatial resolution. Multivariate analysis was used to separate the single cells based on their chemical differences.

RNA is extremely unstable and prone to enzymatic degradation [Anal. Chem. 85 (2013) 2269]. To preserve the integrity of RNA for SIMS analysis we use two methods. First we isolate RNA from the environment by encapsulation in a “graphene sandwich” where the analyte is enclosed in graphene on both sides. In another method RNA is encapsulated into graphene flakes attached to graphene film.

For characterization of RNA, we use a custom-built cluster-SIMS instrument consisting of a source and linear time-of-flight, ToF, mass spectrometer operating in the event-by-event bombardment-detection mode (Suppl. Fig. 1). The experiments were run at the level of individual impacting at 50 keV with separate recording of SIs and electrons emitted in transmission from each collision. The event-by-event bombardment-detection mode allows to select specific impacts, in the present case those involving RNA nano-agglomerates, at the exclusion of signals from the target holder and support.

In the case of the graphene sandwich encapsulation, the procedure consists of drop casting transfer RNA (tRNA) dissolved in deuterated water onto a graphene sheet with subsequent attachment of another graphene sheet on top of the assembly. The graphene sandwich, after 10 minutes of drying, collapses with the generation of nano aqua cells in the interface area. Once inserted into the vacuum, the aqua cells dry within 30 minutes with production of tRNA nano agglomerates.

The resulting mass spectrum of ions emitted from the graphene sandwich (Suppl. Fig. 2a) consists of carbon clusters along with tRNA fragments such as the deprotonated adenine. The coincidence spectrum (Suppl. Fig. 2a) in which the projectile impacts on tRNA were selected (coincidence with tRNA phosphate groups) shows that the tRNA molecules are agglomerated into small structures with size >10 nm which are hydrated as shown by coincidence with D^- (Suppl. Table 1). Estimated coverage of graphene by tRNA agglomerates is less than 30% (Suppl. Equation 1). TEM micrographs (Suppl. Fig. 2b) verify the presence of round-shaped tRNA agglomerates ranging up to 60 nm.

In the second method, where tRNA was encapsulated in graphene flakes, the samples were prepared by drop-casting a solution of tRNA and single layer graphene flakes (0.5 – 5 µm in diameter) onto graphene film allowing it to dry before mounting. Similar tRNA agglomerates were detected as in the first method except they exhibited a more branched structure and their size ranged up to 500 nm.

Work supported by NSF grant CHE-1308312

The hair cuticle which covers hair fibers is easily damaged by external stimuli. The oxidative cleavage of the fatty acids and the thioester linkages of the proteins are known as the main damage of hair, additionally it was reported that the cleavage of fatty acids is the major change caused by bleaching processes [1]. However, the hair surface is damaged by many factors, such as shampooing, blow drying, combing, UV radiation and other chemical treatments, in daily hair care routines. In this study, the damage on the hair surface occurred in daily life was investigated using time-of-flight secondary ion mass spectrometry (ToF-SIMS).

Japanese female hair samples were prepared for the analysis. The structural changes of the proteins were examined by comparing the peak intensity of the amino acid fragment ions. It was confirmed that the intensity change of the cysteine / cystine fragment ion (m/z=76) was prominent . The good relationship between the peak ( m/z=76 ) intensity and that of 18-methyleicosanoic acid (18-MEA), the main component of the surface fatty acids, was shown. Furthermore, the cysteine / cysteine peak intensity decreased even in the part of hair from which the 18-MEA was not detected. The changes were consistent with the model damaged hair by a bleaching process . Thus it is suggested that the oxidative cleavage of 18-MEA is the initial damage and disulfide bonds are subsequently cleaved in daily life.

[1] Surface and Interdace Analysis, Vol. 44 (6), pp 736-739, 2012.