Surface Proteins with Noncovalent Association to the Cell Wall InlB, a Protein Loosely Associated to the Cell Wall by GW Modules

InlB is a surface protein required for L. monocytogenes entry into certain eukaryotic cell types (see Section 4.2). InlB is the only L. monocytogenes surface protein carrying a domain organization consisting of a N-terminal LRR domain and a C-terminal domain of three repetitions of 80 amino-acids, called GW modules (Braun et al. 1997). GW modules are also found in autolysins (see below). Domain swapping experiments revealed that the GW-modules mediate binding of InlB to lipoteichoic acids (LTA) (Jonquieres et al. 1999). This association also occur when purified InlB is added externally to intact cells of L. monocytogenes serotype 1/2a. InlB is efficiently extracted from the cell surface when bacteria are incubated in the presence of LTA, which indicates that the InlB-LTA association may be rather weak. Noteworthy, InlB does not bind to purified cell walls containing TA (Jonquieres et al. 1999), which supports the idea that the GW-modules specifically interacts with polymers of the LTA type. The strength of the GW domain-LTA association increases with the number of GW-modules. Thus, an InlB variant bearing the 8-GW module region of the autolysin Ami binds more efficiently to cell surface. It has been shown that purified In1B does not attach to either the L. innocua surface or that of S. pneumoniae, the latter being decorated with LTA devoid of polyglycerol-P. This observation leaves open the possibility that other cell wall components, as TA, may also modulate the association of InlB to the cell wall. This hypothesis could be tested by determining the capacity of purified InlB for binding to L. monocytogenes 4b cells, which have a TA structurally different to serotype 1/2a cells. The Autolysin Family

The peptidoglycan is a highly dynamic macromolecule whose structure is continuously modified by hydrolytic enzymes, also known as autolysins (Holtje 1998; Popowska 2004). These enzymes cleave preexisting linkages and act coordinately with biosynthetic activities during the incorporation of nascent peptidoglycan or the separation of daughter cells upon cell division. Hydrolytic enzymes are also responsible for the active release of cell-wall components (up to 30-50% per generation in some bacteria) and, as in the case of gram-negative bacteria, for the active recycling of peptidoglycan turnover products. Hydrolytic enzymes necessarily have to be subjected to tight temporal and spatial control since their indiscriminate activity may lead to cell lysis. The profile of autolytic enzymes in a given bacterium is assessed by zymogram assays using gels loaded with cell walls. Renaturated proteins displaying cell-wall degrading activities are visualized following gel staining. Zymogram assays performed with extracts of L. monocytogenes surface proteins have revealed a large number of hydrolytic enzymes (McLaughlan and Foster 1997, 1998; Popowska 2004). In some cases, these assays have proved to be very useful for assigning hydrolytic activity to novel proteins uncovered by genome data (Cabanes et al. 2004; Carroll et al. 2003; McLaughlan and Foster 1998; Milohanic et al. 2001). L. monocyto-genes contains several types of autolysins, all of them harbouring domains that promote attachment of the protein to the cell wall (Cabanes et al. 2002). These include the "amidase" domain, with similarity to the MurNAc-L-Ala amidase of the Atl autolysin of S. aureus; the "LysM" domain; the GW modules; and, the so-called "P60-domain". Some of these enzymes are required for L. monocyto-genes pathogenicity (see Section 5.4.2). Many of the autolysins containing an amidase domain carry a variable number of GW modules (Milohanic et al. 2001). Domains containing short repetitions are also present in autolysins of Streptococcus pneumoniae and Staphylococci. In S. pneumoniae these modules are responsible for binding of the protein to choline residues that decorate the TA and LTA. Seven proteins with this "amidase-GW" domain organization are known in the L. monocytogenes strain EGD-e (Cabanes et al. 2002). Two representative autolysins of this subfamily are Ami (Lmo2558) and Auto (Lmo1076), which contain 8- and 4-GW modules, respectively, in their C-half region (Cabanes et al. 2004; Milohanic et al. 2001). Interestingly, Ami from L. monocytogenes strains of serotypes 4b, 4d, and 4e carry only 6-GW modules and, in contrast to its N-half, the C-half region displays low identity (54% at the aminoacid level) to Ami of EGD-e (serotype 1/2a) (Milohanic et al. 2004). Purified Ami protein from serotype 4b binds less efficiently to 1/2a bacterial cells than to 4b cells (Milohanic et al. 2004), suggesting that in addition to LTA, other cell-wall elements may modulate cell wall-protein association mediated by the GW modules. Such an element could be the TA molecule, which is structurally different in serotypes 1/2a and 4b (see Section 5.2.2). This hypothesis fits to the fact that Ami of serotype 1/2a does not bind to the surface of L. innocua serotype 6a (Milohanic et al. 2001), which has a TA structure similar to L. monocytogenes serotype 4.

Auto is the only autolysin of this subfamily that is absent in the nonpathogenic species L. innocua. The aut gene is flanked by lmo1077, a gene with function related to TA synthesis and also absent in L. innocua. Whether Lmo1077 is required for association of Auto to the cell wall is at present unknown.

Another important group of L. monocytogenes autolysins is formed by the P60-subfamily. All the members of this subfamily share an NPLC/P60 domain in their C-terminal region. Four proteins of the EGD-e strain have been classified in this subfamily: Spl (P45), Lmo0394, Lmo1104, and P60 itself. Lmo1104 is the only autolysin of the group that is absent in L. innocua. In addition to the P60 domain, the P60 protein carries two LysM domains and a bacterial Src-homology 3 (SH3) domain which promote protein association to the cell wall (Cabanes et al. 2002). P60-defective mutants display abnormal morphology, characterized by the presence of filamented cells containing fully formed septa (Gutekunst et al. 1992; Wuenscher et al. 1993). This phenotype links the function of P60 to cell division.

A third type of domain organization is found in a recently characterized autolysin named MurA (Lmo2691). This protein contains an amidase domain in the N-half of the protein followed by 4 LysM domains (Carroll et al. 2003). As in P60-deficient mutants, strains lacking MurA display elongated morphology. The defect in MurA also correlates with increased resistance to detergent-mediated lysis (Carroll et al. 2003), which suggests that MurA could be involved in generalized remodelling of the peptidoglycan. MurA and P60 are secreted, together with other proteins, by a specialized secretion machinery dependent on the SecA2 protein (Lenz et al. 2003). Defects in the secretion of these two autolysins have been linked to the transition to a rough phenotype observed in some L. monocytogenes isolates (Machata et al. 2005).

The L. monocytogenes EGD-e strain has another two putative autolysins with a unique domain organization. The first, Lmo0849, contains an amidase domain in the middle part of the protein and a transmembrane domain in its C-terminal region. This protein, which would be the only L. monocytogenes autolysin embedded in the plasma membrane, has not been characterized yet. The second is Lmo0327, a LPXTG protein containing five LRR domains in the N-half of the protein and 15 repeat regions specific of this protein in its C-half region (Popowska and Markiewicz 2006). The autolytic activity of Lmo0327 was inferred in a screening of autolytic activity using a library of L. monocytogenes

EGD-e cloned in E.coli. The L. monocytogenes lmo0327 mutant lacks several bands in zymogram assays, one of them with the expected molecular weight of Lmo0327. This mutant displays an elongated shape, defects in cell separation, and slightly higher resistance to Triton X100-stimulated lysis. Further work is required to confirm whether Lmo0327, which does not contain any domain related to peptidoglycan hydrolysis, is a bona-fide autolysin. In vitro assays with purified protein could provide such evidence. It is worth to mention that to date no study has provided a direct proof of the specific linkage(s) of the peptidoglycan cleaved by any autolysin of L. monocytogenes. Attempts made with purified P60 on peptidoglycan were unsuccessful due to the inherent property of this protein to aggregate (Wuenscher et al. 1993).

Was this article helpful?

0 0

Post a comment