Size and refractive index determination of submicrometer particles by flow cytometry

Oral presentation
E. van der Pol, F.A. Coumans, L. de Rond, E.L. de Gool, A.N. Böing, A. Sturk, R. Nieuwland, and T.G. van Leeuwen
31st Congress of the International Society for Advancement of Cytometry, Seattle, United States
June 13, 2016


Introduction: Flow cytometers provide light scattering data in arbitrary units, which hampers data interpretation and comparison of results between instruments. Moreover, particle identification requires fluorescent labeling, which is expensive, laborious and prone to artifacts. Here we present a label-free method to obtain the diameter and refractive index (RI) of submicrometer particles from the light scattering signals of a flow cytometer. Besides the obvious advantage of relating scattering signals in arbitrary units to measurement units, we hypothesized that RI detection can discriminate between extracellular vesicles (EV; 1.36<RI<1.42) and lipoprotein particles (1.45<RI<1.60). This is clinically relevant, because EV and lipoproteins (1) both are submicrometer particles present in human blood plasma, and (2) EV change in composition and concentration with disease.

Methods: A flow cytometer (A50-Micro; Apogee, UK) with a 70 mW 405 nm laser was used to detect forward scattering (FSC) and side scattering (SSC). FSC and SSC of reference beads with known diameter and RI were measured and modeled with Mie theory, taking into account the optical properties of the beads and the optical configuration of the flow cytometer. To determine the diameter of unknown particles, we used the relation between the flow cytometry scatter ratio (Flow-SR=SSC/FSC) and the particle diameter. To determine the RI of unknown particles, we used the measured diameter and FSC or SSC to solve the inverse light scattering problem by Mie theory. We validated the method by determining the diameter and RI of a mixture with 250 and 390 nm silica beads and 250, 310, 380 nm polystyrene beads. We applied RI detection to differentiate between EV and lipoproteins from the platelet-depleted supernatant of an outdated platelet concentrate (centrifuged 3-fold, 1,550×g, 20 min).

Results: The obtained relation between Flow-SR and particle diameter is independent of the RI and has a unique solution for particles <500 nm. For each population in the bead mixture, the diameter and RI were determined with measurement errors on the mean <8% and <2%, respectively. For the platelet-depleted supernatant, two populations with different RI were clearly discernable. More than 86% of the population with RI<1.42 and less than 7% of the population with RI>1.42 was positive for CD41-FITC, which confirms the capability of RI detection to differentiate between EV and lipoproteins.

Conclusion: We have developed the first method that converts the scattering signals of a commercial flow cytometer to the comparable measurement units of dimension and RI. The method is based on the relation between Flow-SR and particle diameter, which is independent of the RI and has a unique solution for particles <500 nm. We have demonstrated that this method is accurate and that it allows label-free discrimination between EV and lipoprotein particles in plasma.