| Popis: |
Time-of-flight secondary ion mass spectrometry (ToF-SIMS) has proven its ability to characterise (in)organic surfaces, and is increasingly used for the characterisation of biological samples such as single cells. By combining ion imaging and molecular depth profiling it is possible to render 3D chemical images, which provides a novel, label-free way to investigate biological systems. Major challenges lie, however, in the development of data analysis tools and protocols that preserve the cell morphology. Here, we develop and employ such tools and protocols for the investigation of neuronal networks. One of the reasons 3D ToF-SIMS imaging of cells is underused is the lack of powerful data analysis tools as 3D ToF-SIMS measurements generate very large data sets. To address this issue, we developed a method that allows the application of principal component analysis (PCA) to be expanded to large 3D images making 3D ToF-SIMS image processing of whole, intact cells and cellular networks with multivariate analysis now accessible on a routine basis. Using this method, we are able to separate cellular material from the substrate and can then correct z-offsets due to the cells' topography resulting in a more accurate surface heightmap. The method also facilitates differentiation between cellular components such as lipids and amino acids allowing the cell membrane, the cytoplasm and the extracellular matrix (ECM) to be easily distinguished from one another. These developments permit us to investigate the intracellular localisation of specific native and non-native compounds label-free, not just in single cells but also in larger cellular networks. The visualisation of the cellular uptake of non-native compounds, namely fluorescent dyes, in primary rat cortical neurons and the chemical differentiation between cell types, namely primary rat cortical neurons and retinal pigment epithelium (RPE) cells, are presented as applications. Even though the dyes have distinct fragment ions in the high mass range, it was not possible to detect the fluorophores by 3D ToF-SIMS imaging of freeze-dried cells. However, it was possible to detect distinct differences in the kind of ions detected for freeze-dried primary rat cortical neurons and RPE cells albeit in the low mass range. To obtain meaningful results, however, it is paramount that sample preparation does not induce significant physical or chemical changes. We present the first comprehensive comparison between large 3D ToF-SIMS images of freeze-dried and frozen-hydrated cells using PCA to facilitate the data analysis of these large data sets. A higher degree of colocalisation of the K+ signal with cell regions is observed for frozen-hydrated cells, which indicates a lower degree of membrane damage and migration of diffusible chemical species. Frozen-hydrated cell samples are therefore considered to best reflect the native cell state, but freeze-dried cell samples allow far easier sample handling. The mass spectrum of frozen-hydrated cellular material also has increased ion intensities for higher-mass fragments, which is an additional advantage, because the poor signal-to-noise ratio of molecular species with m/z > 200 is a major bottleneck in the advancement of ToF-SIMS imaging as a diagnostic tool. |
