Rapid recognition of drug-resistance/sensitivity in leukemic cells by Fourier transform InfraRed microspectroscopy and unsupervised Hierarchical Cluster Analysis.

We tested the ability of Fourier Transform (FT) InfraRed (IR) microspectroscopy (microFTIR) in combination with unsupervised Hierarchical Cluster Analysis (HCA) in identifying drug resistance/sensitivity in leukemic cells exposed to tyrosine kinase inhibitors (TKIs). Experiments were carried out in a well established mouse model of human Chronic Myelogenous Leukemia (CML). Mousederived pro-B Ba/F3 cells transfected with and stably expressing the human p210BCR-ABL drug-sensitive wild-type BCR-ABL or the V299L or T315I p210BCR-ABL drug-resistant BCR-ABL mutants were exposed to imatinib-mesylate (IMA) or dasatinib (DAS). MicroFTIR was carried out at the Diamond IR beamline MIRIAM where the mid-IR absorbance spectra of individual Ba/F3 cells were acquired using the high brilliance IR synchrotron radiation (SR) via aperture of 15x15 μm2 in sizes. Conventional IR source (globar) was used to compare average spectra over 15 cells or more. IR signatures of drug actions were identified by supervised analyses in the spectra of TKIs-sensitive cells. Unsupervised HCA applied to selected intervals of wavenumber allowed to classify the IR patterns of viable (drug-resistant) and apoptotic (drug-sensitive) cells with an accuracy >95%. The results from microFTIR+HCA analysis were cross-validated with those obtained via immunochemical methods, i.e. immunoblotting and flow cytometry (FC) that resulted directly and significantly correlated. We conclude that this combined microFTIR+HCA method potentially represents a rapid, convenient and robust screening approach to study the impact of drugs in leukemic cells included peripheral blasts from patients in clinical trials with new anti-leukemic drugs.