Estimation of spectra sample size for characterizing single cells using micro-Raman spectroscopy

Single-cell micro-Raman spectroscopy has the potential to become a powerful, new cytometric approach for discriminating between cell types and identifying subpopulations of cells based on differences in the intrinsic, molecular content of the cells. Despite the considerable progress in demonstrating many biomedical applications of single-cell Raman spectroscopy, an unresolved issue with this method is the inconsistent manner in which the technique has been applied experimentally to acquire spectra from a cell, which can potentially lead to irreproducible and inconsistent results from study to study. Specifically, the problem pertains to the validity of using micro-Raman spectroscopy to sample an arbitrary fraction of the cell volume and to claim that the resulting spectrum represents the entire cell volume. In this study, we investigate the effect of sample size (i.e. the number of Raman spectra acquired from distinct locations in a cell) on the ability to generate a Raman spectrum that accurately describes the total molecular content of the cell. Furthermore, we demonstrate that a minimum sample size of Raman spectra acquired from a cell can be used in place of a full hyperspectral Raman image to achieve the same degree of discrimination between different cell populations. Lymphocytes exposed to the chemotherapy drug, doxorubicin, at different concentrations and exposure times are used as a model biological system in this study. This work demonstrates the importance of adequate spectral sampling and presents an approach for determining the minimum sample size needed to reproduce a Raman spectrum of a whole cell, which are expected to impact future single-cell Raman spectroscopy studies.