The Proteome of Extracellular Vesicles Produced by the Human Gut Bacteria Bacteroides thetaiotaomicron In Vivo Is Influenced by Environmental and Host-Derived Factors

Abstract

Bacterial extracellular vesicles (BEVs) released from both Gram-negative and Gram-positive bacteria provide an effective means of communication and trafficking of cell signaling molecules. In the gastrointestinal tract (GIT) BEVs produced by members of the intestinal microbiota can impact host health by mediating microbe-host cell interactions. A major unresolved question, however, is what factors influence the composition of BEV proteins and whether the host influences protein packaging into BEVs and secretion into the GIT. To address this, we have analyzed the proteome of BEVs produced by the major human gut symbiont Bacteroides thetaiotaomicron both in vitro and in vivo in the murine GIT in order to identify proteins specifically enriched in BEVs produced in vivo. We identified 113 proteins enriched in BEVs produced in vivo, the majority (62/113) of which accumulated in BEVs in the absence of any changes in their expression by the parental cells. Among these selectively enriched proteins, we identified dipeptidyl peptidases and an asparaginase and confirmed their increased activity in BEVs produced in vivo. We also showed that intact BEVs are capable of degrading bile acids via a bile salt hydrolase. Collectively these findings provide additional evidence for the dynamic interplay of host-microbe interactions in the GIT and the existence of an active mechanism to drive and enrich a selected group of proteins for secretion into BEVs in the GIT. IMPORTANCE The gastrointestinal tract (GIT) harbors a complex community of microbes termed the microbiota that plays a role in maintaining the host's health and wellbeing. How this comes about and the nature of microbe-host cell interactions in the GIT is still unclear. Recently, nanosized vesicles naturally produced by bacterial constituents of the microbiota have been shown to influence responses of different host cells although the molecular basis and identity of vesicle-born bacterial proteins that mediate these interactions is unclear. We show here that bacterial extracellular vesicles (BEVs) produced by the human symbiont Bacteroides thetaiotaomicron in the GIT are enriched in a set of proteins and enzymes, including dipeptidyl peptidases, an asparaginase and a bile salt hydrolase that can influence host cell biosynthetic pathways. Our results provide new insights into the molecular basis of microbiota-host interactions that are central to maintaining GIT homeostasis and health.

Keywords: Bacteroides thetaiotaomicron; bacterial extracellular vesicles; intestine; microbiota; proteome.

FIG 1

Structure, size, concentration, and protein content of BEVs produced in vitro and in vivo (A) Transmission electron microscopy generated images of BEVs derived from Bt cells grown in BHI media or from cecal contents of conventionalized germfree mice. (B) Nanoparticle tracking analysis of BEV suspensions. Points (black) represent the mean and the error bars (gray) represent the standard deviation (SD), n = 3. (C) Principal Component Analysis performed on normalized abundances of each protein under each condition. X and y axis show principal component 1 and principal component 2 explaining 93.6% and 4.1% of the total variance, respectively, for BEV proteins and 90.1% and 3.7%, respectively, for parent cell proteins. Prediction ellipses are such that with probability 0.95 a new observation from the same group will fall inside the ellipse.

FIG 2

Proteins enriched in BEVs produced in the mouse GI tract. The 113 proteins enriched in BEVs in vivo (fold change ≥ 15, Table S1) were divided into two groups based upon comparing their levels in BEVs versus parental cells. Results of the 51 proteins with a greater than 3-fold change in the parent cells in vivo (Table S1A) are combined in (A). The 62 proteins with a less than 3-fold increase in abundance in the parent cells in vivo (Table S1B) are combined in (B). For these two groups of proteins their expression is compared to that of the mRNA expression level of the corresponding gene in cells grown under similar conditions (28). (C) SignalP-5.0 Server at https://services.healthtech.dtu.dk/service.php?SignalP-5.0 was used to predict the presence of different types of signal peptides present among the two sets of enriched BEV proteins. The set of 51 proteins identified in (A) is represented by light blue bars and the set of 62 enriched proteins is represented by dark blue bars. Sec/SPI are secretory signal peptides transported by the Sec translocon and cleaved by Signal Peptidase I; Sec/SPII are lipoprotein signal peptides transported by the Sec translocon and cleaved by Signal Peptidase II; Tat/SPI are Tat signal peptides transported by the Tat translocon and cleaved by Signal Peptidase I; “Other” are predicted to be nonsecreted proteins and translocated by unknown secretion pathways.

FIG 3

Bile salt hydrolase activity in Bt and BEVs. Thin layer chromatography was used to identify BSH activity and substrate specificity of the BSH encoded by BT_2086 present in Bt cells and BEVs obtained after growth in BHI media. Cholic acid (CA), taurocholic acid (TCA) and glycocholic acid (GCA) standards were incubated with whole cells or BEVs from wild type (WT) or a Bt BSH1 deletion mutant (ΔBT_2086) for 24h at 37°C after which supernatants were spotted onto a silica gel plates. The plate was inserted into a TLC chamber, run for 40 min, and stained with phosphomolybdic acid. The data shown is representative of two experiments.

FIG 4

Enrichment of dipeptidyl peptidase activity in BEVs in vivo. (A) Nanoparticle tracking analysis was used to determine the size of extracellular vesicles produced by Bt grown in vitro or produced in the ceca of conventionalized germfree mice. Following size exclusion chromatography (SEC), fractions F1-F6 from 4 biological replicates were analyzed. The graph is representative of particle sizes measured in the 4 replicates. (B) DPP activity measured in SEC fractionated particles obtained under different conditions (n = 4 each). (C) DPP activity measured in protein cell extracts from Bt cells obtained under different conditions (n = 4 each). Activities were normalized according to CFU measured for each sample before extraction. In vitro WT: vesicles or cells from WT Bt strain grown in culture; in vivo GF: particles from germfree mice; in vivo WT: particles or cells from germfree mice conventionalized with WT Bt; + Vilda.: particles or cells for which 0.33 μM vildagliptin was added to the reaction mixture. *** = P ≤ 0.001.

FIG 5

Enrichment of asparaginase activity in BEVs in vivo. (A) Nanoparticle tracking was used to determine the mean size of the extracellular vesicle population released in the cecal lumen of germfree mice conventionalized with Bt WT (in vivo WT) and ΔBT_2757 (in vivo ASNase) strains or produced in the cecum of nonmanipulated germfree mice (GF). Following size exclusion chromatography (SEC) fractions F1-F6 from 4 biological replicates were analyzed. The graph is representative of particle sizes measured in all replicates. (B) Asparaginase activity measured in combined SEC fractions (F1-F6) of particles extracted under different conditions; n = 4 for activity measured in particles extracted from mouse cecum of conventionalized germfree mice, n = 3 for vesicles obtained in vitro and for EVs obtained from germfree mice. (C) Asparaginase activity measured in protein cell extracts from Bt cells extracted under different conditions, n = 4. Activities were normalized according to CFU measured for each sample before extraction. In vitro WT = vesicles from WT Bt strain grown in BHI media; In vitro ASNase = vesicles or cells from Bt strain BT_2757 deletion mutant grown in BHI media; in vivo GF = particles from nonmanipulated germfree mice; in vivo WT = particles or cells from germfree mice conventionalized with WT Bt; in vivo ASNase = particles or cells from germfree mice conventionalized with Bt ΔBT_2757 strain. Ns = not significant, * = P ≤ 0.05; ** = P ≤ 0.01; *** = P ≤ 0.001.