Assessing the role of surface glycans of extracellular vesicles on cellular uptake

Abstract

Extracellular vesicles (EVs) are important mediators of cell-cell communication in a broad variety of physiological contexts. However, there is ambiguity around the fundamental mechanisms by which these effects are transduced, particularly in relation to their uptake by recipient cells. Multiple modes of cellular entry have been suggested and we have further explored the role of glycans as potential determinants of uptake, using EVs from the murine hepatic cell lines AML12 and MLP29 as independent yet comparable models. Lectin microarray technology was employed to define the surface glycosylation patterns of EVs. Glycosidases PNGase F and neuraminidase which cleave N-glycans and terminal sialic acids, respectively, were used to analyze the relevance of these modifications to EV surface glycans on the uptake of fluorescently labelled EVs by a panel of cells representing a variety of tissues. Flow cytometry revealed an increase in affinity for EVs modified by both glycosidase treatments. High-content screening exhibited a broader range of responses with different cell types preferring different vesicle glycosylation states. We also found differences in vesicle charge after treatment with glycosidases. We conclude that glycans are key players in the tuning of EV uptake, through charge-based effects, direct glycan recognition or both, supporting glycoengineering as a toolkit for therapy development.

Figure 1

Lectin array analysis of native EVs from MLP29 and AML12. Fluorescence for individual lectin spots was normalised to the highest signal for each samples before combining six lectin spot replicates for average values and standard deviations.

Figure 2

Heat map visualisation for lectin array analysis of EVs after neuraminidase (Neu) or Peptide-N-Glycosidase F (PNG) treatments. The normalised data represent the fold change in fluorescence signal from array binding, relative to EVs of native glycosylation.

Figure 3

Western blot analysis of native and glycosidase-treated EVs. Cell extracts and EVs were analysed by Western-blotting using antibodies against non-EV (Grp78), EV (Tsg101, Hsp70, LimpII, Lamp1) proteins. Note changes in the detection of glycosylated membrane proteins (LimpII and Lamp1) due to glycosylation treatments. This figure has been composited from a single membrane stained multiple times, with individual images available in Supplementary Information.

Figure 4

Physical characterisation of EV models. (a) Nanoparticle tracking analysis (NTA) of vesicle diameter sizes. Box plots show D10, mean and D90 parameters whilst whiskers exhibit minimum and maximum particle diameters observed. (b) Zeta potential determination of particle charges. Significance is calculated by Student’s t test from technical replicates (*p < 0.05, **p < 0.01, error bars represent S.D.). (c) Cryo-electron microscopy (cryo-EM) images.

Figure 5

Flow cytometry uptake analysis of glycosidase treated EVs by recipient cells. Uptake index is determined from the percentage of cell populations exhibiting fluorescence after 16 hours, normalised to MLP EVs. Significance is shown relative to native EVs and was calculated by one-way ANOVA, with error bars representing S.E.M. of 10 replicates over two experiments (*p < 0.05, **p < 0.01, ***p < 0.001).

Figure 6

Confocal microscopy uptake analysis of glycosidase-treated AML12 EVs. Cells were incubated in the presence of DiI-labelled AML12 EVs for 16 hours. Confocal images of DiI (red) and DAPI (blue) dyes were taken at 20x magnification (20 µm bar for scale).

Figure 7

High-content uptake analysis of native EVs incubated with a panel of 28 different human cell lines. Uptake index is determined as the fold-change in fluorescence compared to a DiI background control after a 16 hour incubation, with uptake index of the DiI background thus set to ‘1’ for each cell line. Cell lines are broadly clustered by organ type and lines marked with * are primary cultures.

Figure 8

High-content uptake analysis of EVs with modified glycosylation pattern. A panel of 28 human cell lines were incubated with native EVs or with glycosidase-treated EVs during 16 hours. Confocal images were acquired and quantified as indicated in Material and Methods section. The heat map represents the fold change in uptake relative to the uptake of unmodified EVs.