Autofluorescence signatures for classifying lung cells during epithelial mesenchymal transition

Abstract

The cellular mechanism of epithelial mesenchymal transition (EMT) has been observed to play a pivotal role in embryogenesis, wound healing and cancer metastasis. This fundamentally dynamic phenomenon requires a multifaceted appreciation to evaluate its role in cancer progression. The present study documents alterations in the endogenous fluorescence signatures of lung cells (normal and cancer) to classify them during the progression of EMT and reports their association with cytomorphological and cytoskeletal attributes. Cellular endogenous red and green fluorescence showed a gradual increase during the progression of EMT while the blue component showed the reverse effect. Furthermore, randomness in F-actin fibrillar arrangement was notably elevated and its expression increased. Altered cell shape with heightened vimentin expression was also documented. Principal component analysis (PCA), an unsupervised classifier, was effectively employed endorsing these multi-level attributes to visualize EMT progression. Dimensionality reduction was further carried out to elucidate the significance of autofluorescence. Red fluorescence had the greatest contribution in differentiating cells during the transition. The features were re-evaluated using another in-house built binary classifier, namely vector valued regularized kernel approximation (VVRKFA), in order to understand EMT progression. Gradual increase in classification accuracy with EMT progression indicated cellular changes in the above-mentioned features. Logical corroboration of the findings suggested strong connections between cellular shape and cytoskeletal attributes with autofluorescence phenomena during EMT.