Bacterial Extracellular Vesicles Exploit Filopodial Surfing and Retraction Mechanisms to Reach the Host Cell Body in an Actin-Dependent Manner

Abstract

Extracellular vesicles derived from gram-negative bacteria are nano-sized particles of different size and origin released by these microbes and are collectively called bacterial extracellular vesicles (BEVs). These BEVs may serve as vehicles for delivering bacterial molecules to eukaryotic host cells. Depending on the bacterial species, BEVs elicit various host cellular and immunomodulatory responses, often aiding bacterial survival and communication. Early events in the initial interaction between BEVs and the host cell, as well as how BEVs reach the cell body, remain unexplored. In this study, we describe the interaction of BEVs with actin-rich cellular extensions, including filopodia and retraction fibres, which extend from the host cell surface. Using microscopy-based tracking at the single cell level, BEVs were shown to exploit cellular extensions at the cell periphery to reach the main cell body, either by hijacking retracted extensions or by surfing along these extensions in an actin-dependent manner. BEVs bind to the outer surface of the extensions, but no internalization occurs at this stage. Instead, they serve as transport for BEVs to the main cell body, where endocytosis takes place. Importantly, this process appears to be a general phenomenon for BEVs across different bacterial species and cell origins.

Keywords: bacterial extracellular vesicles (BEVs); extracellular processing; filopodia and cellular extensions.

FIGURE 1

Purity and size of BEVs were confirmed using TEM and NTA. (a) Flowchart illustrating the isolation and purification of BEVs (Created with BioRender); (b) Density (blue) and protein concentration (red) of vesicle fractions measured after crude vesicle purification via density gradient centrifugation; (c) TEM images of various BEV fractions with corresponding scale bars; (d) TEM images of purified BEVs. Scale bars are included in all images; (e) NTA graph showing the concentration and size of purified BEVs. The data are representative of more than ten independent experiments. BEV, bacterial extracellular vesicles; TEM, transmission electron microscopy.

FIGURE 2

Cells are surrounded by actin rich cellular extensions. (a) SEM image of AGS cells showing cellular extensions (white arrows). Image is a representative of multiple images from three independent experiments; (b) Confocal microscopy image of AGS cells stained with Phalloidin‐iFluor 488 (green) to highlight actin and DAPI (blue) to visualize nuclei. A mock control is included. The image represents multiple images from over ten independent experiments; (c) AGS cells incubated with FM 4‐64FX‐labelled BEVs (red) for 2 h, then stained for F‐actin with Phalloidin‐iFluor 488 (green) and nuclei with DAPI (blue). BEVs bound to extensions (white arrows) and those on top of the (white arrowheads) are shown. Scale bar = 5 µm. The image is representative of over ten independent experiments. SEM, scanning electron microscopy.

FIGURE 3

BEVs associate with cellular extensions without endocytosis. (a) TEM analysis of gold‐labelled BEVs, with gold particles (arrowheads), confirmed BEVs immunogold‐labelling; (b) SEM image showing gold‐labelled BEVs bound to cellular extensions; (c) Zoomed‐in SEM view confirming BEVs association with cellular extensions; (d) BEVs attached to the mid‐sections of extensions; (e) Magnified SEM view further confirming BEV attachment; (f) BEVs observed on top of the cells. White dots indicate the gold‐labelled BEVs, confirming their bacterial origin. Scale bars = 0.5 µm. Images represent multiple images obtained from three independent experiments. BEV, bacterial extracellular vesicles; SEM, scanning electron microscopy; TEM, transmission electron microscopy.

FIGURE 4

Correlative microscopy further confirms that BEVs are associated with the outer surface of extension. (a) Overlay of SEM and confocal images, demonstrating alignment between different microscopy techniques; (b) SEM image of AGS cells incubated with BEVs. Black arrows indicate BEVs association with extensions. A zoomed‐in fluorescent inset highlights a potential BEV (white arrowhead); (c)–(d) SEM images show BEVs on cellular extensions; (e) TEM analysis displays multiple BEVs attached to extensions. Images represent multiple observations from two independent experiments. Scale bars are shown. BEV, bacterial extracellular vesicles; SEM, scanning electron microscopy; TEM, transmission electron microscopy.

FIGURE 5

Live cell imaging reveals BEV transport via retraction and surfing. (a) Confocal microscopy of LifeAct‐BFP expressing AGS cells (grey) incubated with FM 4‐64FX‐labelled BEVs (red). Selected frames from Movie‐2 (Supplementary Materials) show BEVs reaching cell body via retraction (yellow line, arrow, and arrowhead). Scale bar: 5 µm. For control including only LifeAct‐BFP expressing AGS cells, see Movie‐1 (Supplementary Materials). (b) Filopodial surfing as a transport mechanism. Selected frames from Movie‐2 (Supplementary Materials) show BEVs moving along filopodia towards the cell body. Yellow (1), red (2), green (3) and pink (4) arrows track individual BEVs. Scale bar: 10 Scale bar: 10 m. Images represent observations from more than ten independent experiments. (c) Trajectories of three BEVs, tracked using the Fiji plug‐in in ImageJ (check materials and methods for details), are shown. (d) Individual BEV velocity along extensions varies, sometimes rapid (sliding/collapsing), sometimes stationary (n = 12 individual BEVs, selected from 10 independent experiments). Each BEVs was monitored with up to 31 frames until they reached the cell body. (e) Retraction is generally faster than surfing. n = 30 observations for each of retractions and surfing, obtained from 10 independent experiments. Error bars show mean ± SD. ^** p ≤ 0.01 (non‐parametric two‐tailed t‐test). BEV, bacterial extracellular vesicles.

FIGURE 6

Surfing and retraction movement of BEVs along cellular extensions is actin dependent. (a) Schematic of retrograde flow of actin filament (created with BioRender); (b) Confocal microscopy images of untreated (control) and Jas‐treated AGS cells incubated with FM™ 4‐64FX labelled BEVs for 2 h. Representative of multiple images from three independent experiments. For a control with only LifeAct‐BFP expressing AGS cells, please see Movie‐1 (Supplementary Materials). (c) Quantification of BEV association with the cell body in untreated (control), Jas‐ and CytD‐treated cells: 30 cells per condition, three independent experiments. (d) Comparison of BEV velocity in control versus Jas‐ or CytD‐ and Noc‐treated AGS cells (15 BEVs per condition) from more than 10 independent experiments. (e) Overall average BEV velocity in untreated (control) and treated (Jas, CytD or Noc‐treated) cells. (f) Distance travelled by individual BEVs. (g) Average distance travelled by BEVs in untreated and treated cells. The x‐axis represents the number of frames counted. (h) TEM images showing endocytosis follows after BEVs move along extensions. AGS cells were incubated with gold‐labelled BEVs (10 nm or 5 nm + 10 nm) for 2 h, fixed and analysed via TEM. Scale bar = 0.5 m. Images represent over 50 observations from four independent experiments. Error bars show mean ± SD. Two (**) and three (***) asterisks indicate p ≤ 0.01, p ≤ 0.001, respectively and “ns” indicates a non‐significant difference (non‐parametric two‐tailed t‐test).

FIGURE 7

BEV binding and movement along extensions is a general mechanism across cell lines and gram‐negative species. (a) Images are from live cell imaging of HeLa cells expressing LifeAct‐BFP (grey) incubated with FM 4‐64FX‐labelled BEVs (red) (Supplementary Movie‐8). (b) CHO cells incubated with FM 4‐64FX‐labelled BEVs(red) for 2 h. BEVs attached to extensions (black arrowheads) are shown in the overlay. Cell body outlined in yellow dotted line (overlay). Scale bar = 10 m. (c) BEVs derived from E. coli (MC1061) exhibit similar attachment and transport mechanism(s), suggesting a conserved process. Data are representative of three independent experiments. Scale bar: 5 µm.