A collaborative project funded by the Landesstiftung Baden-Württemberg foundation, investigated the extent to which spectrally resolved Raman microscopy is able to represent specific enzymes in living cells without the use of markers.
Neuroblastoma cells of a different degree of differentiation were examined and Raman spectra acquired of any image pixel. The spectra were analysed using cluster analysis, which is a statistical method for allocating objects (in our case spectra) to different groups (clusters) based on their similarities or differences. The objects are clustered in such a way that the similarities between the objects are maximal within a cluster, and the differences between the objects are maximal outside the determined cluster. Our data were used for hierarchical cluster analysis, involving a spectral range of 500 to 3100 cm-1.
The above figures show a differentiated SHSY5Y neuroblastoma cell with the spectra obtained using cluster analysis and the representation of the distribution of these spectral characteristics within the cell. Cluster 1 (referred to as cl1 in the above figure) clearly emerged as a prominent cluster in the hierarchy very early on, and showing a distribution around the cell nucleus where the endoplasmic reticulum and mitochondria are normally located. The comparison of this spectrum with the cytochrome c spectrum showed it to be identical with the cluster 1 spectrum. The prominent appearance of cytochrome c bands can be explained by a resonant Raman effect since cytochrome absorbs at the wavelength of 532 nm that was used for excitation. The possibility of being able to identify cytochrome c without the use of markers in living cells might boost the future use of Raman microscopy since this molecule is a key enzyme in the respiratory chain and is also used as marker for mitochondrially induced apoptosis. Previously, the identification of cytochrome c was only possible through invasive methods using antibodies.