The Molecular Genetics Unit studies fundamental aspects of transcription activation and protein targeting in prokaryotes. The model systems studied are the E. coli maltose regulon, in which the transcription of genes involved in maltodextrin transport and metabolism is activated by the protein MalT and the specific effector maltotriose, the type II secretion system that transports folded proteins though the outer membrane of Gram-negative bacteria, and lipoprotein localization in the E. coli cell envelope.

Transcription activation and protein targeting are basic properties of all living cells that are most easily characterized in simple model organisms such as the bacterium Escherichia coli. Genetic studies aimed at understanding how transcription of E. coli genes involved in maltodextrin transport and metabolism (the maltose regulon) is regulated were initiated in our group over 20 years ago. Expression of maltose regulon genes requires the transcription activator protein MalT, a 103 kDa protein that binds to specific sites in the promoters it controls. Its ability to activate transcription of these promoters is controlled by the specific positive effector, maltotriose, and by ATP, as well as by three negative effector proteins, the MalK component of the maltodextrin permease, MalY and Aes. In addition, the mechanism by which MalT, alone or in synergy with CRP (Catabolite Response Proteins), activates transcription depends on the particular promoter and the configuration of the MalT and CRP binding sites therein. Biochemical and structural studies have begun to reveal details of what appears to be one of the most complex gene regulatory systems known in prokaryotes. The negative effectors stabilize the monomeric, inactive form of MalT while maltotriose induces the formation of high-order multimers that are able to activate transcription. The protein is made up of four structural domains. The C-terminal domain binds to the target sites in the maltose regulon promoters while the three N-terminal domains, whose interactions with the two positive effectors and one of the negative effectors have been characterized, might represent a new signal integration module. These domains have now been produced separately and in large amounts using recombinant DNA technology. They are being characterized biochemically and genetically and their structure is being analyzed by X-ray crystallography. In a complementary approach, the oligomeric form of MalT and its interaction with target DNA are being investigated by cryoelectron microscopy.

The type II secretion pathway or secreton is used by a wide range of Gram-negative bacteria to secrete proteins into the external environment. The first detailed characterization of this system, performed in our laboratory using the pullulanase secreton from Klebsiella oxytoca, indicated substantial genetic complexity (11 proteins are required for pullulanase transport across the outer membrane, together with the complete Sec pathway that is required for translocation across the inner (cytoplasmic) membrane. The products of these genes were localized in the cell and biochemical studies of their interactions were begun. A second type II secreton, this time from E. coli K-12, has also been studied recently. This system, which is normally cryptic, secretes a chitinase.

Several years ago, we noted the substantial similarity between the components of type II secretons and components of the type IV pilus assembly machinery found in many Gram-negative bacteria. Pili are surface-anchored appendages composed of subunits (pilins) that can be considered as secreted proteins. The first 30-40 amino acids of 5 secreton components (proteins G, H, I, J and K) are almost identical to the N-terminal regions of type IV pilins, leading us to propose that proteins G though K might be assembled into a pilus like structure. Liquid-grown cells do not produce pili but cells grown on plates produce pili composed exclusively of G protein. Thus, both pullulanase and protein G are secreted by the secreton. Pullulanase secretion and pilus formation require the same secreton components but do not interfere with each other; indeed, G protein is absolutely required for pullulanase secretion. Our aim now is to determine the structure of the G protein and the G pilus and to study the structural requirements for secretion and assembly. In parallel, other components of the secreton are being characterized in greater detail, most notably the secretin protein (D) and its partner (protein S) that form the multimeric channel in the outer membrane that is proposed to allow passage of both pullulanase and the assembled pilus. We are also undertaking a more detailed analysis of two proteins, C and E, that are thought to be involved in energy coupling for secretion and pilus assembly, respectively. Our long-term goal is to characterize their interactions with other secreton components and with pullulanase and to isolate the complete secreton complex (or subcomplexes thereof) for structural analysis and for reconstitution in vitro.

In contrast to most proteins secreted by the secreton, including chitinase, pullulanase is a lipoprotein whose lipid moiety at its N-terminus remains anchored to the cell surface after the polypeptide has been secreted. We have recently become interested in the ways in which lipoproteins are sorted to different sites in the cell envelope. Most lipoproteins are anchored in either the inner or the outer membrane, facing the periplasm. The signal that determines the site to which a lipoprotein will be targeted is the amino acid located immediately after the N-terminal fatty acylated cysteine residue (the +2 sorting rule). We showed that if this residue is an aspartate, a proline or an aromatic amino acid, the protein will be anchored in the inner membrane whereas if it is any other amino acid, it will be transported by the Lol system (characterized in Tokuda's laboratory) to the outer membrane. We are currently designing systems to study retention of lipoproteins in the inner membrane and to look for alternative systems that allow lipoproteins to reach the outer membrane. One particular lipoprotein, Llp encoded by bacteriophage T5, provide access to a genetic selection allowing the isolation of mutants affected in inner membrane sorting as well as to a system that allows us to study ways in which proteins can disobey the +2 sorting rule.

Armelle Lavenir

Olivera Francetic, Chargé de recherche IP

Nicolas Joly, stagiaire DEA

Maria Reyngoud, Agent de laboratoire