Internalins and the Invasion of Nonprofessional Phagocytic Cells

Several genes encoding internalins were identified in L. monocytogenes and L. iva-novii in the pregenomic era [6-8]. All of them belong to the superfamily of LRR (leucine-rich repeat) proteins, but there are two distinct classes of internalins: (a) those which are relatively large in size and cell-associated and (b) those which are considerably smaller and secreted. While L. monocytogenes possesses a large number of internalins belonging to class a and only one (InlC) belonging to class b, the opposite seems to be the case for L. ivanovii (see below).

The first internalin identified was InlA, an acidic protein of 800 amino acids which possesses two extended repeat domains, the first of which consists of 15

LRRs [6]. The InlA protein has a typical N-terminal transport signal sequence and a cell wall anchor in the C-terminal part comprising the sorting motif LPXTG followed by a hydrophobic membrane-spanning region. This distal LPXTG motif has been shown to be responsible for the covalent attachment of InlA to the bacterial cell envelope in a process mediated by the enzyme sortase [9].

InlB [6], a protein of 630 amino acids, also carries an N-terminal transport signal sequence, eight LRRs, and three C-terminal modules each beginning with the amino acids glycine (G) and tryptophan (W) (GW modules), but, in contrast to InlA, has no LPXTG motif and no cell-wall-spanning region. InlB is targeted to the bacterial surface via the noncovalent interaction ofthe GW modules with lipo-teichoic acid in the listerial cell wall [10].

The three-dimensional structures of both internalins were solved at the atomic level [11, 12]. It was found that in InlA and InlB, the N-terminal parts in each protein are combined to form a contiguous internalin domain with the LRR region as the central part. The extended b-sheet resulting from the LRRs constitutes an adaptable concave surface proposed to interact with the respective mammalian receptor molecules during infection.

Both the internalins, InlA and InlB, were shown to be involved in the internalization of L. monocytogenes by various nonphagocytic mammalian cells [6, 13]. Human E-cadherin was identified as the InlA receptor [14]. The highly specific interaction of human E-cadherin with InlA is based on the concave interaction domain which surrounds and specifically recognizes E-cadherin [12]. The cytoplasmic domain of E-cadherin interacts with a- and b-catenins, which on the other hand directly bind to actin filaments and hence complete the link between the listerial surface protein and the host cell cytoskeleton [15].

The receptor tyrosine kinase Met was identified as an InlB receptor required for InlB-dependent entry of L. monocytogenes into various cells [13], which induces rapid tyrosine phosphorylation of Met followed by a complex series of intracellular signal transduction events. The PI-3 kinase is in the center of this signal transduction pathway leading to actin rearrangement [16], and is most likely activated upon interaction with adaptor proteins recruited to the phosphorylated cytoplas-mic domain of Met [13, 17]. The generation of PIP3 by the PI-3 kinase [16] then leads to activation ofthe downstream kinases. It is currently believed that these events finally lead to the activation ofthe Arp2/3 complex (see below) which is necessary to induce the actin polymerization involved in the rearrangement ofthe actin cytoskeleton during the uptake process.

The complement receptor for the globular part ofthe C1q fragment (gC1q-R) is another InlB receptor. Direct interaction of the ubiquitously expressed gC1q-R and InlB has been demonstrated, but the role of the gC1q-R in InlB-mediated entry of L. monocytogenes is still unknown [18]. Finally, glucosaminoglycans (GAGs) were identified as a third type of InlB ligand [19]. GAGs are present on the surface of mammalian cells, where they decorate the proteoglycans. InlB binds to GAGs through its C-terminal GW repeats, which anchor the protein to the bacterial cell surface. GAGs are hence believed to detach InlB from the bacterial surface, allowing its interaction with the Met [20].

The significance of the listerial surface protein InlA in in vivo host cell invasion was for a long time less clear. To test whether the single nucleotide exchange in mouse E-cadherin compared to human E-cadherin (Pro16Glu), which renders the mouse homologue unable to bind InlA [21], is the reason for the missing role of InlA in virulence in the mouse model, Lecuit et al. [22] produced a transgenic mouse expressing human E-cadherin in its enterocytes. With this transgenic mouse the role of InlA in the penetration of the intestinal barrier was clearly demonstrated. L. monocytogenes mutants lacking InlA were significantly less able to cross the intestinal barrier since they were severely impaired in promoting their own uptake by the enterocytes forming the intestinal barrier [22].

The function of InlC, the prototype of the group of secreted small internalins, is less clear at the moment. Deletion of inlC reduces virulence in the mouse sepsis model, and participation of InlC in InlA-mediated internalization has been suggested [7, 23].

0 0

Post a comment