Team projects
Our integrative approach to understanding the bacterial, host and environmental factors associated with LD acquisition/severity is organized in two synergistic projects: (1) Bacterial determinants of Legionella-host cell interaction, and (2) Host, microbiota and environmental determinants of LD severity.
Axis1- Bacterial determinants of Legionella-host cell interaction
Project1-1. A FOCUS on Dot/Icm secretion
The T4SS Dot/Icm is recognized as a major virulence factor, by translocating ~330 effectors into the host cell, enabling bacterial intracellular multiplication in phagocytic cells, amoeba and human alveolar macrophages. Among the regulatory elements that control the expression of the Dot/Icm secretion system, we demonstrated the role of the Nucleoid-Associated-Proteins (NAP) Fis. Notably, Lp encodes 3 fis genes rather than the usual one. Fis1 and Fis2 cooperate to regulate virulence gene expression by targeting similar AT-rich motifs, albeit with distinct affinity, and by being capable of forming heterodimers (Andréa, bioRxiv 2024). We are now deciphering the structural and functional molecular basis of this control.
In addition to studying the regulation of Dot/Icm machinery and effectors synthesis, we showed that effectors translocation is orchestrated in a cyclic-diGMP dependent manner, and this control is required for efficient infection (Allombert, Infect. Immun 2014). By developing an assay to monitor injection of T4SS effectors (Allombert, Methods Mol Biol 2024), we demonstrated that translocation of each effector is accurately set up in time and correlated to its function during infection (Figure 1). We also found that one cyclic-diGMP metabolizing enzyme (out of 22 synthesized by Lp) locally controls the Dot/Icm machinery for appropriate delivery of effectors (Allombert, Jaboulay, J. Mol. Biol. 2021). We are now investigating the molecular mechanism by which c-di-GMP controls the T4SS machinery.
Besides exploring regulation of Dot/Icm secretion, we elucidated the function of several Dot/Icm effectors, and more recently, investigated inter-effector relationship. We demonstrated that the protein kinase LegK2 controls ARP2/3-mediated actin polymerization at the surface of the Legionella containing vacuole (LCV) (Michard, mBio 2015). We also defined the functional interactions between VipA and LegK2. Specifically, we demonstrated that LegK2/VipA effector pair target different substrates, ARP2/3 for LegK2 and G-actin for VipA, to temporally control actin polymerization at the LCV and interfere with phagosome maturation and endosome recycling (Pillon, Cellular Microbiology 2024). Now, rather than assessing the role of individual effectors, we are addressing their collective impact on the regulation of specific cellular functions. For instance, we are comprehensively defining the role of the five Dot/Icm effectors (LegK2, VipA, Ceg14, RavK, WipA) that targets the cell actin cytoskeleton. Moreover, we are conducting innovative screens using a proximity-labelling approach to study the localization of effectors within various cellular organelles and elucidate their functional contributions.
Project 1.2. Without « a priori » approach - Experimental evolution
We sought to characterize virulence factors without a priori by conducting experimental evolution strategies (Figure 2). After hundreds of generations without host cells, we identified one mutation that confers a strong attenuated virulent phenotype compared to the ancestral Lp virulent isolate. This modification consists in the integration of the endogenous plasmid pLPP into the chromosome. We demonstrated that plasmid integration and associated phenotypes can be reversed when bacteria were re-evolved in presence of amoebae, suggesting that the episomal form and/or the copy number of the plasmid confer a strong fitness advantage to bacteria and is essential for successful infections (Carrillo, in preparation). We are now performing functional investigations to decipher this novel regulatory pathway of Lp virulence.
Figure 2. Experimental evolution of L. pneumophila under various conditions and/or hosts.
Project 1.3. Role of c-di-GMP in the formation of persisters
Regarding bacterial determinants associated with slow- or non-resolving LD, we investigated the hypothesis that therapeutic failures could be due to Lp resistance to macrolides, the first-line therapy for LD treatment. We selected lineages of high-level macrolide-resistant Lp with mutations in 23S rRNA, L4 and L22 proteins (Descours, Antimicrob Agents Chemother 2017), and in non‑coding sequences upstream TolC-dependent efflux proteins encoding genes (Massip, J Antimicrob Chemother. 2017; Vandewalle-Capo, Int J Antimicrob Agents 2017). Yet, although macrolide-resistant strains have been recently identified in the environment (Portal, J Antimicrob Chemother 2021; Ginevra, J Antimicrob Chemother 2022), no clinical resistance has been reported, suggesting that recalcitrant LD is most likely due to Lp persisters (Pouderoux, Clin Infect Dis. 2020; Adams-Ward, Front Cell Infect Microbiol 2023). Our preliminary data suggest that host cell-derived NO regulates development of intracellular persisters in relation with a two-component regulatory system and c-di-GMP second messenger concentration.
In collaboration with N. Personnic, we are deciphering the molecular mechanism involved.
Axis 2- Host, microbiota and environmental determinants of LD severity
Projet 2.1. Proteomic interrogation of the lung-bacteria interface
This project aims to dissect the molecular dialogue occurring at the host-Lp interface. Its objective is to explore how the lung proteome is altered during LD, how it translates to the patient’s response, and to identify the role of active enzymes from host and pathogen in Lp infection. The bacterial surface is a critical interface for interactions between host and pathogen proteins, significantly impacting infection outcomes (Figure 3). A major goal is to identify host-derived bacterial binding proteins (HBBPs) that interact with Lp to uncover novel components of the host innate immune surveillance system. Notably, we identified the host lectin SP-D as an HBBP, highlighting its impact on Lp infection. Notably, we identified the host lectin SP-D as an HBBP, highlighting its impact on Lp infection (Dubois et al., in preparation)
Figure 3.
Projet 2.2. Proteomic interrogation of the lung-bacteria interface
In a simple human macrophages cell line model, we have demonstrated that clinically relevant Lp isolates displayed TNF secretion and TNF-induced cell death according to their genotypes (Guillemot, Virulence 2022). We are now developing more complex and physiologically relevant models to recreate the alveolar-capillary barrier of the lung, which is the main infection site of Legionella pneumophila. To do so, we are using Transwell® and lung-on-chip models established by Huh (Science 2010) and now commercialized by Emulate©. Briefly, miniaturized cell culture devices (PDMS chips) are composed of one upper channel seeded with epithelial cells, separated from the lower channel seeded with endothelial cells by a porous membrane. Macrophages could be added in the upper channel after epithelial cell differentiation, to mimick resident alveolar macrophages (Figure 4). Chips are incubated into a cell culture module that controls microfluidics and mechanical stretches, to better mimick the alveolar-capillary barrier of the lung. With this innovative model, we aim at better understanding the cellular and immunological pathways involved during infections with relevant clinical strains, as well as identifying associated bacterial determinants.
Figure 4. The upper compartment of chips is seeded with epithelial cells (in dark blue), cultured in liquid then differentiated in air condition. The lower channel is separated by a porous membrane and seeded with endothelial cells (in red). Macrophages could be added in the “air” compartment, as well as monocytes (light blue) in the blood compartment to better mimick to alveolar-capillary barrier. Once the human cells differenciated, bacteria are added to the air compartment and their ability to establish an infection, induce a immunological response and translocate into the blood compatment are evaluated. Adapted from Huh, 2010, Science.
Projet 2.3. Global understanding of pathophysiology of severe LD
This project aims to identify bacterial and host markers associated with the adverse outcome of LD, we showed that severe LD is associated (i) with higher local (pulmonary) and systemic (serum) Lp DNA load, and a highly disturbed bacterial lung microbiota (Pérez-Cobas et al., Cell. Rep. Med. 2023 & mBio 2020); (ii) impairments immune response with a hyper-inflammatory phase in the early phase of Lp pneumonia and a leukocyte immunoparalysis for a large number of cytokines (Allam et al., Front. Cell Infect Microbiol. 2023; Allam & Ibranosyan, in preparation). We are also focusing on the systemic phase of Lp infection and the gut/lung axis during LD using a multi omics approach (metatranscriptomics/metagenomics/metabolomics) to decipher the complex interactions between the digestive, blood and lung compartments and to identify potential metabolite biomarkers associated with severity.
In the same way, in collaboration with a large consortium in the VORTEX project (PEPR MIE 2023) on emerging respiratory infection, we are analyzing the Volatile Organic Compounds (VOC) using exhaled breath during LD as potential biomarkers of infection in combination with a in vivo / in vitro approach in order to better explain the VOC profile detected in patient with LD.
Projet 2.4. Environmental determinants of human infection and severity of LD
Micropollutants are widespread at very low concentrations in inhaled air, water, food, but also in human fluids because of everyday multi-exposure. We aim to integrate human co-exposure to such chemical agents and microbiological ones on the LD model, based on a One Health approach. In this project, 3D lung models will be exposed to pollutant molecules and submitted to bacterial challenges; cellular and immunological responses will be measured in presence or absence of Legionella, in order to explore how the exposition to pollution could sensitize to LD.
In addition, we aim to characterize the susceptibility and adaptive mechanisms of Lp to endocrine disruptors found at very low concentrations in natural water systems. The project EPiLEG aims to decipher the impact of triclosan and phthalates on Legionella in its environmental habitat (planktonic and amoebic environmental proliferation, biofilm formation, evolution of Lp genomics under chemical pressure), but also over the course of LD (intramacrophage multiplication, antimicrobial susceptibility, immune response).
Figure 5.