Modulation of intestinal immune responses by microbiota-derived extracellular vesicles

Abstract

The gut microbiota establishes dynamic and complex interactions with the intestinal epithelium and the immune system. In this regard, commensal microbiota plays essential role in the development and maturation of the host immune system, which must learn to discriminate between commensal and pathogenic bacteria. Appropriate development of the immune system is crucial to maintain intestinal homeostasis. Microbiota-host communication does not involve direct intercellular contacts since the intestinal epithelium is protected by a mucus layer that segregates luminal microbes from host cells. Therefore, the gut microbiota mainly communicates with the host through the secretion of bacterial factors which can diffuse through the mucus layer and reach epithelial and immune cells of the intestinal mucosa. In addition to metabolites and soluble proteins, microbiota releases bacterial extracellular vesicles (BEVs) as a mechanism of intercellular communication. BEVs are membranous structures that act as a crucial mechanism for intra- and inter-kingdom communication, allowing long distance delivery of bacterial effectors. In particular, our group has reported numerous beneficial properties associated to microbiota-derived BEVs. Our model of study is Escherichia coli, a natural gut colonizer species, and particularly the probiotic E. coli Nissle 1917 (EcN) and commensal E. coli strains. However, deep knowledge of the immune regulatory mechanisms involved in the mediated effects has been poorly understood. Therefore, the main objective of this thesis was to study the molecular mechanisms used by the gut microbiota to modulate innate and adaptative immune responses and safeguard intestinal homeostasis.

In our first study, we evaluated the ability of BEVs released by the probiotic EcN and commensal E. coli strains to modulate dendritic cell (DC) function and the subsequent Th cell effector responses (Chapter 1). The experimental approach consisted in an in vitro model of human monocyte-derived DCs (mo-DCs) co-cultivated

with total human CD4+ T cells isolated from peripheral blood mononuclear cells (PBMCs) from healthy donors. The effects were evaluated by cytokine quantification and flow cytometry analysis of specific markers. Results showed that BEVs represent a mechanism used by gut resident microbiota to prime the innate immune system, activating DCs in a strain-specific manner. BEVs from commensal and probiotic E. coli strains elicit similar Th17/Th17/Th22 responses and mainly differ in their ability to induce Th1 and Treg responses. BEVs from the probiotic EcN exert complex immunomodulatory Th/Treg responses, consistent with the beneficial effects of this probiotic. In this context, EcN BEVs promote a strong Th1 response, a mechanism that may account for the protective effects of this probiotic against enteric virus. In contrast, the commensal ECOR12 (group A) secretes BEVs that mainly contribute to maintaining host tolerance against gut microbes by increasing Treg/Th17 balance and promoting Th22 response. However, vesicles from this commensal do not trigger proper Th1 responses. These results provide evidence on the specific immunomodulatory properties of BEVs from probiotic and commensal strains, and their collaborative role in the maintenance of intestinal homeostasis.

Since gut-derived BEVs activate DCs in a strain-specific manner, we next wanted to investigate the regulatory mechanisms involved in these effects. To this end, we performed a miRNA deep sequencing approach trying to identify miRNAs differentially expressed in mo-DCs in response to BEVs from the probiotic EcN and the commensal ECOR12 strain (Chapter 2). Results indicated that the immunomodulatory properties of BEVs released by the probiotic EcN and the commensal ECOR12 in DCs are in part mediated through the regulation of miRNAs. The analysis revealed a common set of miRNAs modulated by BEVs from both strains. These miRNAs modulate basic biological processes, including metabolism, cell growth, and development. Importantly, some miRNAs such as miR-155, miR-let7i and miR-146a are related to DC activation and maturation, being involved in DC antigen-presenting functions, TLR signaling pathways, and release of inflammatory mediators. In addition, we identified

a number of miRNAs that showed differential expression depending on the bacterial strain. Some of them were related with immune function and their differential expression was in accordance with the cytokine profile and derived specific Th responses described in the previous chapter. In this context, BEVs from the probiotic EcN trigger upregulation of miR-29a, and consequently reduce the expression of its target IDO2. This regulatory mechanism is compatible with the weaker tolerogenic responses induced by EcN BEVs in comparison to ECOR12 BEVs. In contrast, ECOR12- derived BEVs trigger upregulation of miR-146b, miR-125a, and miR-125b-99b-let7e cluster. The differential upregulation of these miRNAs that attenuate the inflammatory response by targeting IL-12, IFN-γ, and TNF-α and proteins of TLR signaling pathways may contribute to the anti-inflammatory/tolerogenic action of ECOR12 BEVs. Taken together, these results indicate that differential regulation of miRNAs by BEVs is one of the mechanisms used by the gut microbiota to train the immune system and help to preserve intestinal homeostasis.

In addition to physical cell-to-cell contacts, DCs communicate with neighbouring T cells through secreted mediators that include cytokines and extracellular vesicles (EVs) such as exosomes. To get new insights into the mechanisms that mediate activation of T cell responses by BEV-stimulated DCs, in the third section we focused on DC-secreted factors. In particular, we evaluated the ability of DCs activated by EcN and ECOR12 BEVs to release cytokines and exosomes that distinctly influence derived CD4+ T cell responses (Chapter 3). We set up two cellular models to mimic the communication between DCs and T cells (i) Direct model, in which activated mo-DCs and naïve CD4+ T cells are co-cultured together in the same plate, and therefore can stablish physical interaction (ii) Indirect model using Transwell permeable supports, in which DC and T cells only communicate through released factors. Quantification of secreted cytokines revealed indirect communication between activated DCs and naïve CD4+ T cells despite the absence of direct cellular contacts. Similar T effector responses were observed in both models, although levels of secreted cytokines were lower in the

indirect co-culture model. The main differences between bacterial strains also affected the Th1 and Treg responses. Moreover, characterization of DC-derived exosomes by Cryo-transmission electron microscopy (TEM) and nanoparticle tracking analysis (NTA) revealed no significant differences in the amount, size, or structure between EVs produced by immature DCs or BEV-stimulated DCs. Next, we analysed whether microbiota vesicles could influence the functional properties and cargo of DC-derived exosomes. The results revealed that activation of mo-DCs by microbiota BEVs leads to significant changes in the exosomal content of costimulatory molecules and miRNAs compared to non-treated immature DCs. In this context, exosomes derived from BEV- activated DCs contain the typical exosome markers and are enriched in costimulatory molecules CD86, CD44, CD40, and MHC-II that are involved in antigen presentation, T cell activation or proliferation. Concerning their miRNA cargo, quantification of immune-related miRNAs in exosomes derived from BEV-stimulated DCs evidenced distinct miRNA profile depending whether DCs were stimulated with EcN or ECOR12 BEVs. These differences showed good correlation with their differential expression in DCs (Chapter 2) and are in accordance with the elicited Th responses elicited. Particularly, miRNAs that drive tolerogenic and anti-inflammatory profiles (miR-146b, miR-125a, miR-125b and miR-24) are increased in exosomes derived from DCs stimulated with ECOR12.

In conclusion, data presented in this thesis show that BEVs released by microbiota E. coli strains modulate host immune responses to preserve the intestinal balance. BEVs deliver bacterial effectors that drive DCs to coordinate appropriate T cell responses through several mechanisms that include regulation of miRNA expression and differential release of immune mediators through exosomes.