We have actually three major research fields: the heme acquisition systems requiring an extracellular hemophore, the protein secretion by ABC exporter across the two membranes which delimit the Gram negative cells, and the genetics of biofilm formation.
1) Hemophore dependent heme acquisition systems (Sylvie Létoffé, Laurent Debarbieux, Julie Deleule, Laurent Courtes, Sylvia Rossi, Philippe Delepelaire, Annick paquelin, Jean-Marc Ghigo)
Iron ions are essential for many metabolic pathways. The low solubility of iron (III) salts at physiological pH in the presence of oxygen is a serious obstacle to microorganisms that require iron. Microorganisms colonizing human hosts are confronted with another problem. Body iron is not readily available. It is bound to iron carrier proteins or to heme in hemoproteins.
Therefore, the inability of pathogens to assimilate these diverse sources of iron would be one of the factors limiting bacterial success in colonization and establishment in the mammalian host. Study of these iron acquisition systems is growing in many laboratories with the aim of designing new antimicrobial drugs.
It is actually well established that several iron/heme acquisition systems might coexist in the same species. Microbes produce low molecular mass substances called siderophores which chelate ferric ion with extremely high affinity allowing its solubilization and its extraction from most mineral or organic complexes. Iron siderophore complexes are recognized and bound by specific receptors at the cell surface. Bacteria have also specific receptors which recognize iron or heme containing proteins. Another heme acquisition system relies on the secretion of hemophores (extracellular bacterial hemoproteins) which interact in the extra cellular environment with the heme source and present it to specific receptors. In Gram negative bacteria, transport of most of these iron sources across the outer membrane is dependent on energy derived from the PMF and on a multiprotein complex of the inner membrane consisting of the ExbB, ExbD and TonB proteins. The periplasmic domain of TonB interacts with a conserved sequence located close to the N-terminus of the various outer membrane receptors.
Bacterial extra cellular hemophores also named HasA for Heme Acquisition System were discovered in our laboratory in 1994 and form an independent family of hemoproteins found in several Gram-negative bacteria (Serratia marcescens, Pseudomonas aeruginosa, Pseudomonas fluorescens and Yersinia pestis) that take up heme from host heme carriers and shuttle it to specific outer membrane receptors (HasR). Hemophore receptors are required for the hemophore-dependent heme acquisition pathway and alone allow free or hemoglobin-bound heme uptake, but the synergy between the hemophore and its receptor greatly facilitates this uptake.
The 3D structure of the S. marcescens holo-hemophore (HasASM) shows that heme iron atom is ligated by tyrosine 75 and histidine 32. The phenolate of tyrosine 75 is also tightly hydrogen bonded to the N d atom of histidine 83. Alanine mutagenesis of the heme ligands showed that one of the two axial iron ligand is sufficient to efficiently ligate heme and that H83 which is not heme-coordinated in the wild type protein may become an alternate iron ligand in the mutant protein.
HasA tridimensional structure
Both heme loaded and unloaded hemophores bind to a hemophore specific outer membrane receptor on the the same site (or overlapping site) with the same apparent affinity. This raises the question of how the exchange between loaded and unloaded hemophores occurs on the receptors. We propose a heme exchange mechanism instead of a hemophore exchange mechanism. This might represent a very general mechanism for membrane transport with siderophores and hemophores working as " shuttles " (Létoffé et al 2001 Mol. Microbiol. "in press"). The respective domains on HasA and HasR involved in their interactions are not yet characterized. We are seeking them by genetical and biochemical approaches. Large amounts of the HasA/HasR complex have already been purified using a His tagged HasA protein. Crystallisation of the complexes will be done in W. Welte laboratory in Constance.
Whereas hemophore interaction is TonB independent, heme internalization through the outer membrane is TonB dependent.
S. marcescens, like several other Gram-negative bacteria has two TonB proteins: the previously characterized TonBSM and HasB, coded by a gene located in the same operon as hasA and hasR
TonBSM and HasB have significant homology and can replace each other for heme acquisition. On the other hand, TonBSM but not HasB mediates iron acquisition, showing that HasB and TonBSM only display partial redundancy.
Our findings support the emerging hypothesis that TonB-homologues are widespread in bacteria where they may have specific functions in receptor-ligand uptake systems (Paquelin et al, submitted).
The has operon of S. marcescens comprizes the genes encoding the receptor (HasR), the hemophore (HasA), two of the three components of the ABC transporter (HasD, HasE) and HasB. The has operon is negatively regulated by the Fur repressor in the presence of iron. The operon contains a Fur operator, just upstream to hasR. DNA sequence indicated the existence of two open reading frames located upstream to hasR also preceeded by a Fur operator. DNA sequence homologies indicated that these two genes might code for an extracytoplasmic sigma factor (HasI) and its modulator (HasS). Heme transport through the outer membrane might change the conformation of HasS allowing the sigma factor activation. We constructed lacZ transcriptional fusions to the various has genes, introduced them in the S. marcescens has operon This allowed us to show that HasI is required for hasR-lacZ expression in S. marcescens. The inducer is the holo-HasA molecule. Our work demonstrates that the inducer of the has operon (holo-HasA) and the transported substrate are different molecules (S. Rossi et al manuscript in preparation).
All the hemophores so far characterized are cleaved at their C-terminus in their natural hosts The physiological significance of the C-terminal cleavage is unclear. In an heterologous assay using the S. marcescens receptor HasR, we found that only the cleaved P. aeruginosa and P. fluorescens and Y. pestis hemophores are active, stimulating heme uptake. However, the unprocessed and in vitro processed forms of HasASM appear to have the same biological activity . We hope to reconstitute very soon other hemophore dependent systems to determine whether cleavage has a physiological meaning.
The S. marcescens Has system
2) Protein secretion by Gram negativeABC transporters (Guillaume Sapriel, Laurent Debarbieux and Philippe Delepelaire)
Hemophores are our model system to study the secretion of proteins lacking an N-terminal signal peptide by ABC exporters.
Bacteria secrete many proteins which have diverse sequences and functions. However, only four secretion pathways (numbered from I to IV) have been actually described in Gram-negative bacteria. Type II pathway is the only secretion system that requires the universal N-terminal signal peptide and the general export pathway to the cross of the inner membrane. Type I and III pathways are independent of the signal peptide and sec system.
The Type I secretion pathway also called the ABC secretion pathway consists of three proteins located in the cell envelope: the ABC protein belonging to the ATP Binding Cassette class of proteins; another cytoplasmic membrane protein belongs to the "membrane fusion protein" class and an outer membrane component .
ATP Binding cassette (ABC) transporters are implicated in the vectorial movement of solutes across biological membranes. They constitute one of the most abundant family of proteins in the living organisms where they play essential functions as evidenced by the growing number of human health disorders linked to an ABC transporter deffect. The typical feature of these proteins is the presence of a large cytoplasmic 200 aminoacid conserved domain which contains the two conserved ATP binding motifs that form the ATP binding pocket.
The exoproteins secreted by type I pathway usually have an uncleaved C-terminal secretion signal. Secretion occurs in one step, from the cytoplasm to the extracellular medium with no periplasmic intermediate.
Our work has largely contributed to show that bacterial ABC protein exporters are widespread in Gram-negative organisms ; that they consist of three cell envelope proteins. We have shown that in several cases (such as the E. coli hemolysin transporter) the third outer membrane component belongs to the TolC family and is a multifunctional protein involved in drug efflux and colicin import. TolC has been very recently crystalized revealing a new fold with a very large periplasmic domain forming a tunnel. Binding of the exoprotein C-terminal secretion signal to the ABC protein modulates the ATPase activity and triggers the association of the three secretion proteins in a mutltiprotein complex. Chaperones such as SecB are required to target the N-terminus of the nascent polypeptide to the ABC protein (Sapriel et al 2001 submitted). In the absence of SecB or in the absence of the secretion apparatus the intracellular hemophore folds and is able to bind heme. This folded polypeptide is uncompetent for further secretion. Moreover, it still interacts with the transporter and inhibits the secretion of nascent unfolded new molecules. (Debarbieux et Wandersman 2001 submitted). Thus, our results lead to a paradoxal model in which a protein with a C-terminal secretion signal is secreted cotranslationnally.
3) Genetic of biofilm formation in Gram-negative bacteria (Jean-Marc Ghigo)
Populations of surface-attached microorganisms in natural and artificial environments are referred to as biofilms. Using E. coli as a model system and molecular tools, we focused on the identification of cell factors involved in mature biofilm formation and maintenance.The characterization of such factors may lead to a better understanding of the bacterial biology in biofilm as well as to the identification of potential strategies to control biofilm formation.
a) Development of the experimental model
We designed a continuous flow culture system where biofilm formation can be reproducibly monitored and controlled in micro-fermentors developed in the Fermentation Laboratory of the Pasteur Institute. In this system, mature E. coli biofilms as thick as 1-2 mm form in 24/72 H on different solid surfaces. The amount of material that can be recovered from the surface allows further microscopy (Confocal microscopy) genetic or biochemistry analysis (i.e. Protein profiling, RNA extraction etc
b) Role of conjugative pili in biofilm formation
The comparison of the biofilm-forming capacity of different E. coli strains in the micro-fermentor system revealed that strains carrying the F conjugative plasmid rapidly form thick biofilms. We identified by mutagenesis the conjugative factor involve in biofilm formation as to be the expression of the unique F pilus expressed by cells harboring the F factor.
We screened in the micro-fermentor model, after introduction in the same E. coli background strain, more than 30 natural conjugative plasmids of all known incompatibility groups. This analysis showed that this capacity is not restricted to the F factor but, rather, is a general feature of Gram-negative bacterial conjugative plasmids. Furthermore, the contact of planktonic populations that repressed their conjugative functions with a biofilm induces a general de-repression leading to the horizontal transfer of the plasmids while biofilm formation takes place. Conjugation was kown to occur in biofilm. This work revealed the functional relationships between conjugation and biofilm. This has profound ecological consequences. It suggests in particular that conjugative plasmids express adhesion factors that drive their host cells into highly heterogeneous community where horizontal transfer of genetic material of diverse origins can be exchange at high rates through conjugation.
c) Identification of specific physiological responses in mature E. coli biofilm (Compared study of the biofilm and planktonic transcriptome)
Evidence exits that different genes are transcribed in the planktonic and Biofilm-associated phases of the bacterial life cycle. Moreover, biofilm bacteria display physiological properties that are distinct from their planktonic equivalents (e.g. antibiotic resistance). My objective is to characterize on a genomic scale, bacterial key cellular factors involved in biofilm formation by identifying genes that are specifically expressed inside a biofilm. Our model is the Gram-negative bacterium E. coli where a genome-wide analysis is made possible through expression-profiling using DNA arrays. Seeking for such genes may provide insight on still unknown developmental aspect of biofilm formation.
The production of large amounts of E. coli biofilm now allows the use of commercially available E. coli DNA arrays that are designed to identified genes whose expression varies between two cognate environmental conditions. We extracted messenger RNAs from 8 days old E. coli biofilms as well as from E. coli planktonic culture in exponential phases. These RNA's were retro-transcribed and the labeled cDNAs hybridized on duplicate membrane arrays (Genosys Panorama™) from where the E. coli 4,290 PCR-amplified ORF (representing all E. coli protein coding genes) have been spotted.
This analysis led to the identification of a small number of genes overexpressed in the biofilm (including 47% of unknown " Y " genes as well as genes that are repressed in biofilm conditions. We begun to characterize these candidates using reverse genetic/ molecular biology as well as biofilm-developmental assay in micro fermentors.
Photo 1 : HasA tridimensional structure
Photo 2 : The S. Marcescens Has system