Differential temporal release and lipoprotein loading in B. thetaiotaomicron bacterial extracellular vesicles

Abstract

Bacterial extracellular vesicles (BEVs) contribute to stress responses, quorum sensing, biofilm formation and interspecies and interkingdom communication. However, the factors that regulate their release and heterogeneity are not well understood. We set out to investigate these factors in the common gut commensal Bacteroides thetaiotaomicron by studying BEV release throughout their growth cycle. Utilising a range of methods, we demonstrate that vesicles released at different stages of growth have significantly different composition, with early vesicles enriched in specifically released outer membrane vesicles (OMVs) containing a larger proportion of lipoproteins, while late phase BEVs primarily contain lytic vesicles with enrichment of cytoplasmic proteins. Furthermore, we demonstrate that lipoproteins containing a negatively charged signal peptide are preferentially incorporated in OMVs. We use this observation to predict all Bacteroides thetaiotaomicron OMV enriched lipoproteins and analyse their function. Overall, our findings highlight the need to understand media composition and BEV release dynamics prior to functional characterisation and define the theoretical functional capacity of Bacteroides thetaiotaomicron OMVs.

Keywords: BEV; Bacteroides; OMV; bacterial extracellular vesicles; outer membrane vesicles; proteomics.

FIGURE 1

Particle and lipid analysis highlighting limitations of using complex media in BEV studies. (a) B. thetaiotaomicron growth analysis in commonly used Brain–Heart Infusion media supplemented with hemin (BHIH). FM4‐64 and particle count signal does not correlate with growth. Size of fixed non‐dividing cells measured using light microscopy; lipid content of supernatants measured using FM4‐64 lipophilic dye; OD600 – optical density of cultures determined at 600 nm; particle numbers determined using ZetaView nanoparticle tracking analysis instrument. Experiment carried out in triplicate. (b) Particle count and FM4‐64 analysis of (BHIH), tryptone (10 g/L), yeast extract (YE; 5 g/L) and our media (BDMr6). Red dotted line represents detection limit. Analysis carried out on three separately prepared aliquots from individual powder stocks, since these represent technical replicates no formal statistical analysis is performed. BHIH, Brain–Heart Infusion media supplemented with hemin.

FIGURE 2

Analysis of B. thetaiotomicron BEVs produced in chemically defined media. (a) B. thetaiotaomicron growth analysis in BDMr6. The size of fixed non‐dividing cell size was measured using light microscopy; The lipid content of the supernatant was measured using FM4‐64 lipophilic dye; Optical density of cultures was measured at 600 nm (OD600); Particle number was determined using the ZetaView nanoparticle tracking analysis instrument. (b) Analysis of BEV associated double stranded DNA adjusted to the amount of lipid. Differences between time points are not statistically significant except for vesicular DNA between 9 and 49 h (1‐way ANOVA following Dunnett correction excluding the 5 h time point p = 0.0023). (c) Total protein present in BEVs adjusted to the amount of lipid. (d) Localisation of BEV proteins at 4 different timepoints. Predicted protein localisation. (e) Principal component analysis analysis of BEV proteomes over time. BEV, bacterial extracellular vesicles; CYTO, cytoplasmic; IN.MEMB, inner membrane; LIPO, lipoproteins; O.MEMB, outer membrane; PERI, periplasmic.

FIGURE 3

Cryo‐TEM analysis of B. thetaiotaomicron BEVs. (a) Combined violin and box plots of BEV size distribution analysed using Cryo‐TEM. Cut‐off value = 100 nm; n > 2000. Dashed lines show peak values for 5‐ and 49‐h samples. (b) Representative Cryo‐TEM micrographs of different sized BEV particles observed. Particles shown in the far right column display increased electron density. BEV, bacterial extracellular vesicles.

FIGURE 4

B. thetaiotaomicron metabolite, BEV and CFU analysis. (a) Graph highlighting the major fermentation products produced by B. thetaiotaomicron in BDMr6 with glucose as the carbon source. Mixed acid fermentation at 2:2:1 ratios of succinate:acetate:formate released by the cells. OD600, optical density of culture at 600 nm. Error bars shows one standard deviation. Experiment carried out in triplicate. (b) Graph highlighting change in CFU in relation to OD600, vesicle lipid content (FM4‐64) and lactate. Lipid content of supernatant measured using FM4‐64 lipophilic dye. Early glucose timepoints, ethanol and propionate analysis can be seen in Figure S2. BEV, bacterial extracellular vesicles; CFU, colony forming unit.

FIGURE 5

BEV incorporation signal peptide analysis. (a) Schematic of Nanoluciferase reporter construct used to evaluate targeting peptide sequences. Upper section depicts the genetic constructs with lower section depicting the expected protein topology on BEVs. LipoSeq is BT_3742 Sec targeting sequence which is cleaved and lipidated at a cysteine (C) residue; TargetSeq (red) is the variable 8 amino acid region which defines the pI (see Table S2 for all sequences); TEV site is the TEV protease cleavage site. (b) Graphs show Nanoluciferase signal in the supernatant of cultures two different growth phases plotted against the isoelectric point (pI) of the TargetSeq. Error bars represent one standard deviation between triplicate samples. BEV, bacterial extracellular vesicles.

FIGURE 6

Proposed B. thetaiotaomicron BEV release dynamics. BEV release in BDMr6 occurs in three phases. The first phase is defined by an increase in cell size and release of specialised OMV and formate. During the second phase, lactate is released, cells reduce in size with a reduction in OMV production. The final phase is defined by a decrease in cell density and release of DNA and lytic BEVs. BEV, bacterial extracellular vesicles; OMV, outer membrane vesicles.