The Comparison between Different Extracellular Vesicle Isolation Methods by AFM-IR Nanospectroscopy

Extracellular vesicles (EVs) contain cell-type-specific signatures and have been proposed as biomarkers in various diseases. However, due to their small dimensions, quantification and size distribution, the biophysical characterization of these particles is challenging and still controversial. The diffraction-limiting effect limits acquiring spectral data on EV samples in the mid-IR range, making it impossible to capture images and spectra at the nanoscale by conventional vibrational spectroscopy methods. Here, we employed atomic force microscopy-infrared nanospectroscopy (AFM-IR) technique to elucidate the molecular signatures of EVs isolated from blood (serum and plasma), as well as the direct correlation of their topography and biochemical signature with respect to the isolation method. EVs were extracted from serum or plasma from C57BL/6 male mice by Total Exosome Isolation Reagent (TEIR), ultracentrifugation, and size-exclusion chromatography (SEC), and evaluated by microreflectance in FTIR and AFM-IR analysis. The multivariate analysis by PCA and PLS-DA showed that phosphate groups could be significantly altered depending on the EV isolation/purification method, as well as glycosidic linkages. The SEC processing strongly depletes the amount of biomolecules whose bands fall in the region above 1150 cm^-1, which includes α-helix and fatty acid content of EVs, while the β-sheet and random coil/turn/loop conformational content of proteins, nucleic acids, phospholipids, carbohydrates, and glycoproteins is almost preserved for SEC. The TEIR method can sustain the diverse fractions of conformational content of proteins at the cost of damaging nucleic acids, phospholipids, carbohydrates, and glycoproteins. Still, the UC processing was able to preserve the largest amount of biomolecules present in EVs. However, structural damage on membranes was observed, e.g., smearing out of phospholipids related to membrane-packing bands (1720-1800 cm^-1 region) and an overall decrease in protein bands decreasing.