Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382XBlackwell Publishing Ltd, 2004? 2004523677690Original ArticleAttachment of the S-layer to the cell wall of T. thermophilusF. Cava et al.
Molecular Microbiology (2004) 52(3), 677–690
doi:10.1111/j.1365-2958.2004.04011.x
Binding to pyruvylated compounds as an ancestral mechanism to anchor the outer envelope in primitive bacteria Felipe Cava,1 Miguel A. de Pedro,1 Heinz Schwarz,2 Anke Henne3 and José Berenguer1* 1 Centro de Biología Molecular Severo Ochoa, Consejo Superior de Investigaciones Científicas-Universidad Autónoma de Madrid, Campus de Cantoblanco, 28049 Madrid, Spain. 2 Max Plank Institut für Entwicklungsbiologie, D-72076 Tübingen, Germany. 3 Goettingen Genomics Laboratory, Institute for Microbiology and Genetics, Grisebachstr. 8., 37077 Göettingen, Germany. Summary Electron microscopy of isolated cell walls of the ancient bacterium Thermus thermophilus revealed that most of the peptidoglycan (PG) surface, apart from the septal region, was shielded against specific aPG antibodies. On the other hand, an antiserum raised against S-layer-attached cell wall fragments aSAC) bound to most of the surface except for the (a septal regions. Treatments with a-amylase and pronase E made the entire cell wall surface uniformly accessible to aPG and severely decreased the binding of aSAC. We concluded that a layer of strongly bound secondary cell wall polymers (SCWPs) covers most of the cell wall surface in this ancient bacterium. A preliminary analysis revealed that such SCWPs constitute 14% of the cell wall and are essentially composed of sugars. Enzyme treatments of the cell walls revealed that SCWP was required in vitro for the binding of the S-layer protein through the S-layer homology (SLH) motif. The csaB gene was necessary for the attachment of the S-layer–outer membrane (OM) complex to the cell wall in growing cells of T. thermophilus. In vitro experiments confirmed that cell walls from a csaB mutant bound to the S-layer with a ~1/10) than that of the wild type. much lower affinity (~ CsaB was found to be required for pyruvylation of components of the SCWP and for immunodetection
Accepted 7 January, 2004. *For correspondence. E mail:
[email protected]. Tel. (+34) 91 4978099
© 2004 Blackwell Publishing Ltd
with a-SAC antiserum. Therefore, the S-layer–OM complex of T. thermophilus binds to the cell wall through the SLH motif of the S-layer protein via a strong interaction with a highly immunogenic pyruvylated component of the SCWP. Immuno-crossreactive compounds were detected with aSAC on cell walls of other Thermus spp. and in the phylogenetically related microorganism Deinococcus radiodurans. These results imply that the interaction between the SLH motif and pyruvylated components of the cell wall arose early during bacterial evolution as an ancestral mechanism for anchoring proteins and outer membranes to the cell walls of primitive bacteria. Introduction A common trait of many archaeal and bacterial cell envelopes is the presence of a regularly ordered, planar array of proteinaceous subunits termed surface (S-) layers (Sleytr and Beveridge, 1999; Sára and Sleytr, 2000). As a result of their structural simplicity and widespread distribution among thermophilic microorganisms, it has been proposed that ancestral S-layer preceeded modern cell wall structures as morphogenetic elements (Sára and Sleytr, 2000). However, S-layers have evolved further to fulfil many and different roles within the cells, which must be of sufficient importance to be selected for despite their great metabolic cost. With very few exceptions S-layers are built of a single (glyco)protein that represents up to 15% of the total cell protein content. It has been calculated that for a normal-sized bacterial cell with a generation time of 20 min, around 500 subunits of the S-layer protein have to be synthesized, translocated across one or more envelopes to the cell surface, and incorporated into the preexisting S-layer lattice every second. The strong lateral interactions between S-layer subunits restrain lateral diffusion, and so this entire complex mechanism of synthesis has to be targeted to defined insertion sites on the growing cell surface, where specific S-layer binding molecules have to be located. Different mechanisms have been proposed to explain the attachment of the S-layer to the underlying envelope (Sára, 2001). In certain Archaea, the S-layers are so
678 F. Cava et al. closely associated with the cytoplasmic membrane as to be considered integral membrane proteins (Sleytr and Beveridge, 1999). By contrast, S-layer proteins in bacteria are usually soluble, and attach to the cells by a variety of mechanisms. In Proteobacteria strain-specific N-terminal domains of the S-layer protein usually bind to specific Ochain groups of the lipopolysaccharide (LPS) component of the outer membrane (OM). In Gram-positives, at least two evolutionarily independent kinds of binding domains have evolved for the attachment of the S-layer to the cell walls (Sára, 2001). In Geobacillus stearothermophilus and in Aneurinibacillus thermoaerophilus DSM 10155, the N-terminal region of the S-layer proteins, containing many basic amino acids, is responsible for the binding to specific (secondary) cell wall polysaccharides (SCWPs) by a lectin-like interaction (Sára, 2001; Steindl et al., 2002). However, in other Bacillaceae the presence of at least one copy of the so-called S-layer homology (SLH) motif at the N-terminal region of the S-layer proteins seems to be required for cell-wall attachment (Mesnage et al., 2000). S-layer homology motifs consist of an approximately 55amino acid-long sequence, from which about 15 residues are generally well conserved. Although SLH-like sequences were described as a putative new type of noncovalent peptidoglycan binding motif (Lupas et al., 1994; Olabarría et al., 1996), it was later found that, in a number of species, the SLH motifs bind to SCWPs instead (Ries et al., 1997; Sára et al., 1998; Chauvaux et al., 1999; Mesnage et al., 1999a; 2000; Sára, 2001). In most cases, these SLH-binding molecules are covalently bound to the peptidoglycan network, but an exception to the rule have been described for the OlpB protein from Clostridium thermocellum, which also binds to a non-covalently bound compound of the cell wall (Chauvaux et al., 1999). S-layer homology motifs are not exclusive to S-layer proteins, and they have been described in many cell-wallbound exoenzymes (Lupas et al., 1994). Furthermore, in some instances, chimaeric enzymes with a heterologous SLH motif were indeed attached to the cell wall (Mesnage et al., 1999b). S-layer homology domains have also been found in proteins of bacteria belonging to ancient phylogenetic groups unrelated to Gram-positives, such as Thermotogales and Deinococcales (Engelhardt and Peters, 1998). Interestingly, these bacteria actually have an outer membrane surrounding the peptidoglycan layer and therefore the SLH-wearing proteins could constitute a way of anchoring such OMs to the cell wall. The interaction between SLH-wearing proteins and cell wall fractions initially identified as peptidoglycan was clearly shown for the SlpA protein of Thermus thermophilus (Olabarría et al., 1996). Along with Deinococcus, the genus Thermus is grouped in one of oldest branches of the bacterial phylogeny (Hartmann et al., 1989). The cell
envelope of this bacterial group consists of a multilayered structure, in which a Gram-positive-like peptidoglycan network (Quintela et al., 1995) is surrounded by an outer membrane that generates a periplasmic space (Castán et al., 2002). In both genera, a hexagonally symmetric Slayer is present (Baumeister and Kübler, 1978; Castón et al., 1988), although the sequences of the respective protein subunits are unrelated. In fact, the SLH motif demonstrated in the S-layer protein (SlpA) of T. thermophilus is absent from the S-layer protein (HPI) of D. radiodurans (Engelhardt and Peters, 1998). As a consequence, whereas the HPI protein attaches to the surface of the outer membrane through a lipid-modified N-terminus (Peters et al., 1987; Engelhardt and Peters, 1998), SlpA is attached to the cell wall proper via SLH (Olabarría et al., 1996) and therefore some domains of SlpA have to cross the outer membrane to reach the underlying cell wall. Structural and biochemical evidence, as well as sequence analysis, indicate that SlpA represents a peculiar and, most probably, ancestral type of structural protein, with properties shared by modern S-layers and porins (Engelhardt and Peters, 1998). Indeed, under appropriate conditions SlpA is able to associate in regular structures that are remarkably similar to those built up by bacterial porins instead of the normal hexagonal array exhibited in vivo as S-layer (Castón et al., 1993). Mesnage and co-workers demonstrated that in Bacillus anthracis a pyruvylated form of an SCWP containing galactose, N-acetyl glucosamine, and N-acetyl mannosamine, constitutes the binding site for the SLH domain of its two S-layer proteins (EA1 and Sap) and for a number of exoenzymes, including those involved in cell autolysis (Mesnage et al., 2000). Pyruvylation of this SCWP was shown to be dependent on the activity of CsaB, the second product of the csaAB operon, located immediately upstream of the tandemly organized S-layer genes. CsaB homologues were identified in partial or complete genome sequences of a few bacterial species. The discovery of a perfect correlation between the presence of CsaB homologues and SLH-containing proteins led to the proposal that the interaction between a pyruvylated SCWP and SLH-bearing proteins might represent a new type of protein–cell wall anchoring mechanism (Mesnage et al., 2000). As T. thermophilus is among those bacteria in which the presence of a CsaB homologue was proposed (Mesnage et al., 2000), we wondered whether binding of the S-layer to the cell wall would be mediated by SCWP in this ancient bacteria. Our results indicate that this is indeed the case. Therefore, pyruvylated SCWP–SLH interaction is likely to represent an ancestral mechanism for the attachment of the outer membrane (OM)–S-layer complex envelope to the bacterial cell wall. © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
Attachment of the S-layer to the cell wall of T. thermophilus Results A covalently bound layer of SCWP covers the peptidoglycan surface in T. thermophilus Raw cell wall fractions from T. thermophilus were prepared as the material which remains insoluble after three consecutuive 8 h periods of boiling in 4% (w/v) SDS (Experimental procedures). It consists of the entire peptidoglycan network, which preserves the cell morphology, and any kind of material covalently attached to it. In fact, preliminary work from our laboratory suggests the existence of muramidase-resistant material in these fractions (Quintela et al., 1995). When raw cell walls from T. thermophilus were subjected to immunodetection with antiserum K84 (aPG), raised against the peptidoglycan of E. coli (Olabarría et al., 1996; de Pedro et al., 1997), only septal regions appeared intensely labelled (Fig. 1, inset A1). By contrast, E. coli cell walls used as internal controls in these experiments appeared uniformly decorated by the antibodies (inset A2). Calculation of the number of gold particles per unit (mm2) of surface area on aPG-immunolabelled raw cell walls of T. thermophilus (excluding division sites and cell poles) was around an order of magnitude lower (7.2%) than that of E. coli (Table 1). By contrast, when cell wall preparations were digested with a-amylase and pronase E before immunomicroscopy, a homogeneous labelling with aPG was observed (Fig. 1B). The number of gold particles on cell walls of T. thermophilus subjected to this treatment increased substantially, although it was still lower (~36%) than that found for a similar surface of E. coli sacculi (Table 1). This could be the consequence of differences in the PG composition or simply a result of the lower PG content per unit cell surface described for T. thermophilus (Quintela et al., 1995). Individual treatments with each enzyme partially increased the amount of gold particles bound to the cell
Table 1. Binding of aPG and aSAC to protein-free cell walls of T. thermophilus. Raw (Raw) or digested with a-amylase (A), pronase E (P), or both (A + P) cell walls of T. thermophilus, were immunolabelled with antiserum against peptidoglycan (aPG) or against Slayer-attached cell wall fragments (aSAC), and analysed by electron microscopy.
aPgTha % Th/Ecb aSACc %
Raw
A
P
A+P
24 ± 7 7.2 922 ± 111 100
49 ± 8 14.7 386 ± 41 41.8
67 ± 7 19.8 211 ± 13 22.8
122 ± 9 36.5 133 ± 25 14.4
a and c. The number of gold particles/mm2 of cell wall surface is shown. b. Percentage of labelling with respect to 1 mm2 of E. coli sacculi used as internal controls c. Untreated cell walls were taken as 100% because E. coli sacculi were not recognized by this antiserum. © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
679
wall surface, but never to the extent of labelling achieved with the double treatment (Table 1). These data strongly imply the existence of a layer of material that covers the surface of the cell walls of T. thermophilus and blocks the access of antibodies to the PG itself. In order to visualize this material directly, raw cell walls of T. thermophilus were extended onto microscope grids and treated in situ with muramidase (100 mg ml-1) for 30 min to digest the peptidoglycan before staining. As it is shown in Fig. 1C, this treatment allowed the detection of material which still kept the cell form, whereas no residues from E. coli cell walls could be detected after 10 min of treatment in parallel controls. From now on, we will call this protective coat the Secondary Cell Wall Polymer (SCWP) layer. A preliminary characterization of the muramidase-resistant material was carried out. The undigested material was also insoluble in boiling SDS, and represented around a 20% (w/w) of the total weight of the raw cell wall fraction. After acid hydrolysis, the content in ornithine (4.2% w/w) was analysed as an indicator of the amount of peptidoglycan still present (Quintela et al., 1995), which was a 6.7% (w/w) from that of raw cell walls, and a 32% of the content of the muramidase-resistant material. Therefore, the SCWP layer represents around a 14% (w/ w) of the cell walls from T. thermophilus. A composition of 74% (w/w) of reducing sugars (Ghuysen et al., 1966) and a 4% (w/w) of amino acids (Moore et al., 1958) was calculated for this SCWP. The SCWP-layer contains S-layer-associated polymers A rabbit antiserum against the S-layer protein (SlpA) from T. thermophilus HB8 was obtained after its isolation from lysozyme-digested cell envelope fractions (Experimental procedures). The specificity of the antiserum was checked by Western Blot on membrane fractions from an S-layer deletion (DslpA) mutant (Fernández-Herrero et al., 1995), showing that no other protein from this bacterium was recognized by the antiserum (Fig. 2). However, the antiserum was able to recognize raw cell wall fractions from this same DslpA mutant (Fig. 3C, inset C1). Thus, we concluded that an unknown cell wall component which remained strongly attached to SlpA during purification and SDS–PAGE elicited a strong immune response. We renamed this antiserum as S-layer and Cell wall (aSAC). When this antiserum was used in immunomicroscopy assays with raw cell walls, the entire surface of the T. thermophilus cell walls was uniformly and densely labelled except for the septal regions, where labelling was consistently weaker (Fig. 3A, inset A1). Interestingly, the aSAC was completely unable to bind to E. coli cell walls (Fig. 3A, inset A2), suggesting that the cell wall component recognized by aSAC in T. thermophilus
680 F. Cava et al.
Fig. 1. A layer of SCWP blocks the access of antibodies to the peptidoglycan. A. Raw cell walls of T. thermophilus (long forms) and E. coli (shorter forms) were immunolabelled with antiserum against peptidoglycan. Note the absence of labelling on most of the surface except for the septal region of T. thermophilus in the corresponding inset (A1), and the homogeneous labelling in E. coli (A2). B. Similar immunodetection was made after treatment of the sacculi with a-amylase and pronase E. Insets of T. thermophilus (B1) and E. coli (B2) are shown for detail. C. Raw cell walls of T. thermophilus were extended on the grid and digested in situ with muramidase (100 mg/ml for 30 min) before staining. Bars correspond to 1 mm.
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
Attachment of the S-layer to the cell wall of T. thermophilus
681
Fig. 2. Immunodetection of the SlpA protein. Membrane proteins from the wild-type strain T. thermophilus HB27C8 (27), and from its DSLH (X) and DslpA (D) derivatives, were subjected to SDS–PAGE along with two additional samples corresponding to the soluble (SB) and insoluble (IB) material obtained after digestion of raw cell walls from the wild-type strain (PGd). Figure shows the Coomassie blue stained gel and three parallel Western blots with aSAC rabbit antiserum, with a monoclonal antibody against the SLH domain (aSLHm), or with a mix of four monoclonal antibodies against different SlpA epitopes (aSlpAm) except the SLH domain. Note the absence of labelled proteins in the DslpA mutant, and the specificity of the aSLHm antibody. Protein markers (M): 115, 97, 54, 37 and 29 kDa.
shares little or no similarity with the peptidoglycan of E. coli. It is noteworthy that the labelling pattern obtained with aSAC on T. thermophilus raw cell walls was complementary to that obtained with the aPG (Fig. 1A). Therefore, it looks as if aSAC would recognize some of the component(s) of the SCWP-layer that block(s) access of aPG to the peptidoglycan. This assumption was further supported by the fact that digestion of T. thermophilus raw cell walls with a-amylase and pronase E, led to a substantial reduction in the level of aSAC binding (Fig. 3B). Quantification of binding (Table 1) showed that untreated raw cell walls bound around 930 gold particles per mm2, whereas around 130 were detected for an identical surface unit after digestion with both enzymes. Single digestions with each enzyme resulted in intermediate labelling (Table 1). Components of the SCWP-layer of T. thermophilus are required for the binding to the SLH domain of the S-layer The results above indicated that the PG of T. thermophilus is covered by an SCWP-layer (Fig. 1), and that some components of this SCWP-layer are strongly associated with the SlpA protein (Fig. 3). Therefore, we wondered whether the SCWP-layer could be the actual binding site for the S-layer protein in the intact cell, and if this binding was dependent on the presence of its SLH domain. As control of our binding experiments we used a T. thermophilus mutant strain in which the SLH domain was deleted (Olabarría et al., 1996). The absence of the SLH domain from the SlpA protein of this mutant was assessed by Western blot with an SLH-specific monoclonal antibody, and with a mix of four monoclonal antibodies against other © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
regions of SlpA (Fig. 2). The characterization of these monoclonal antibodies was previously described (Castón et al., 1996). To check this, membrane proteins were transferred to a nitrocellulose membrane, renatured and incubated with raw or enzyme-digested cell wall fragments of T. thermophilus HB8. Binding was detected by Western blot with aPG using non-radioactive methods. Samples from the T. thermophilus DSLH mutant were used as SLH-specificity controls. The results shown in Fig. 4B illustrate how wild-type SlpA proteins from both T. thermophilus strains (lanes 8 and 27), but not the DSLH derivative (lane X), bound specifically to T. thermophilus raw cell walls. Binding was also detected on lower-sized bands which corresponds to SLH-bearing fragments of SlpA, as it is shown with the monoclonal antibody against the SLH domain (Fig. 2). By contrast, binding was virtually abolished when a-amylaseand pronase E-treated cell walls were used (panel E). Single treatments with each of the enzymes denoted that pronase (D) had a greater effect than a-amylase (C) on the binding to wild-type SlpA proteins. The csaB gene is involved in pyruvylation of SCWP In the Gram-positive Bacillus anthracis, pyruvylation of a SCWP fraction was shown to be required for the binding of SLH-containing S-layer proteins to the cell walls (Mesnage et al., 2000). A gene, named csaB (cell surface anchoring) was shown to be required for such pyruvylation. Analysis of raw cell wall preparations of T. thermophilus revealed the presence of significant amounts of covalently bound pyruvic acid (see below). Indeed, a 330-amino-
682 F. Cava et al.
Fig. 3. The cell wall surface of T. thermophilus is covered by compounds that bind strongly to the S-layer protein. Immunodetection was carried out with aSAC, an antiserum raised against the S-layer and strongly bound cell wall fragments. Untreated (A, C) or a-amylase- and pronase Edigested (B) sacculi of the wild-type strain (A, B) and a DslpA mutant (C) of T. thermophilus are shown. The short forms correspond to E. coli sacculi used as internal controls in all the experiments. Details of figures A, B and C are shown in the insets. Bars correspond to 1 mm.
acid-long homologue of CsaB (CsaBth, from now on, for clarity) is encoded in the genome of T. thermophilus HB27 (http://www.g2l.bio.uni-goettingen.de) as previously suggested (Mesnage et al., 2000). To analyse the putative role of the csaB gene of T. thermophilus HB27 in pyruvylation, a null mutant (csaB::kat) was constructed (Experimental procedures). Analysis of raw cell wall fractions detected around 3.8 mg of pyruvic acid per mg of PG in the wild type strain, whereas less than 0.1 mg per mg of PG were detected in the csaB mutant. Therefore, the csaB gene found in the genome of T. thermophilus HB27 is required for pyruvylation of components of its cell wall. Although pyruvylation of PG has not been reported before, we checked whether CsaBth had any influence on the muropeptide composition of the PG by means of HPLC analysis (Quintela et al., 1995). Our results demonstrated that inactivation of csaB had no detectable
effect on PG composition (not shown), pointing to the SCWP-layer detected in Figs 1 and 3 as the likely target for CsaBth-dependent pyruvylation. Pyruvylation of the SCWP is required for attachment of the S-layer A highly significant characteristic of the csaB mutant was the formation of multicellular bodies (MBs) during growth (Fig. 5). These structures were also observed in mutants defective for the regulation of S-layer synthesis, in which they arise from detachment of the S-layer and associated outer membrane (S-layer–OM complex) from the cell wall, and from failure of this structure to invaginate concomitantly with cell division (Castán et al., 2001). As a consequence, cells divide but remain enclosed by the S-layer– OM complex, and generate a spherical body containing a © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
Attachment of the S-layer to the cell wall of T. thermophilus
683
The aSAC antiserum recognizes pyruvylated compounds of the cell wall Disruption of csaB did not affect the ability of the SCWPlayer to shield peptidoglycan against aPG. In fact, raw cell wall preparations from the wild type and the csaB mutant bound aPG to a similar extent and with similar patterns (Fig. 7). In both cases, labelling at the septal regions was more intense than at the rest of the surface (Fig. 7A and B). Therefore, csaB has no major effect on the presence of the SCWP-layer. By contrast, a clear difference between fluorescence intensity in cell walls of the wild type and the csaB mutant was found when the antiserum aSAC was used. An intense and apparently uniform signal on raw cell walls from the wild type was detected, whereas mutant cell walls appeared faintly labelled, and required prolonged exposure in order to be detected (Fig. 7D). Interestingly, when the cell walls were digested with a-amylase and pronase E before immunodetection with aSAC, the fluorescence signal was very weak, but similar for both strains (Fig. 7F). Thus, our data imply that the major antigen recognized by the aSAC antiserum is a pyruvylated component of the SCWP-layer, whose presence depends on the activity of CsaBth.
Fig. 4. In vitro binding of cell walls to the S-layer. Membrane proteins of the T. thermophilus strains HB27C8 (8), HB27 (27) and a DSLH mutant (X) were separated by SDS–PAGE and stained with Coomassie blue (A), or transferred to a nitrocellulose filter and subjected to binding assays with untreated (B), or digested with a-amylase (C), pronase E (D) or both (E), cell wall preparations.
large number of cells. Therefore, impairment of csaB apparently leads to detachment of the S-layer–OM complex from the cell walls during growth. These results imply that pyruvylation is required for normal S-layer-cell wall attachment in vivo. To confirm this point in vitro, binding to SlpA of cell wall preparations from the csaB mutant and the wild-type strains were compared by the assays described above. The results of the experiment (Fig. 6) showed that SlpA has a much lower affinity (~1/10) for binding to raw cell walls of the csaB mutant than for binding to cell walls of the wild-type strain. However, binding was not completely abolished by the csaB mutation, and at the highest concentration used, binding of SlpA to cell walls of the csaB mutant was still detected. Binding was in all instances dependent on the presence of the SLH domain in SlpA, as indicated by the total absence of binding to the DSLH-SlpA protein (lane X). Therefore, we concluded that the CsaBth protein was required for pyruvylation of the cell wall of T. thermophilus, making it proficient for high-affinity binding to the SLH domain of the S-layer protein. © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
aSAC recognizes the cell walls of phylogenetically related bacteria In order to check whether the SCWP-layer component recognized by aSAC is strain or species specific, or if it is a more conserved trait, cell walls from T. thermophilus HB8, T. thermophilus HB27, T. aquaticus YT1, and the phylogenetically related radiotolerant Deinococcus radiodurans, were subjected to immunolabelling with aSAC and observed by electron microscopy. As shown in Fig. 8, all the strains assayed were identified by the antiserum, which did not bind at all to the E. coli sacculi used as the internal control in the experiments.
Discussion The results discussed below indicate that binding of the SLH motifs to pyruvylated components of the bacterial cell wall represents an evolutionarily conserved proteinanchoring mechanism dating back at least to the Thermus–Deinococcus branching point in the bacterial phylogeny. A covalently bound SCWP-layer covers the peptidoglycan of T. thermophilus We have identified a new cell envelope layer (the SCWPlayer) which is covalently bound to the peptidoglycan in T.
684 F. Cava et al.
Fig. 5. The external envelopes detach from the cell surface in the csaB::kat mutant. Phase-contrast micrographs of samples from cultures of the wild type (A) and the csaB::kat mutant (B–G) grown up to different cell densities (OD550). A, Wild type; B, OD550=0.2; C, OD550= 0.5; D, OD550=0.8; E,Overnight cultures; F, A detail of panel E; G, A broken multicellular body corresponding to overnight cultures of the mutant.
thermophilus. The first clues regarding the existence of such an additional envelope could be deduced from thin sections of entire cells, like that published by Lasa et al. (1992) who noted the presence of a thick layer of unstructured material between the PG and the S-layer–OM complex of wild-type cells of T. thermophilus. Interestingly, an ‘intermediate layer’ with similar location and thickness was also described in thin sections of the phylogenetically related bacterium, D. radiodurans (Baumeister et al., 1986). Therefore, we propose that the SCWP-layer identified in this article corresponds to such an ‘intermediate layer’, whose presence, and function in the attachment of an ancestral outer membrane, also present in D. radiodurans, constitutes a common trait among this primitive bacterial group. The main unexpected characteristic of the muramidase-resistant fraction containing the components of this SCWP-layer was its insolubility in SDS. The PG content of this fraction probably corresponds to SCWP-bound muropeptides and additional disaccharides whose proximity to these branching points made inaccessible their linking bonds to the muramidase. As a result of this probable partial digestion, and to the insolubility of this compounds in SDS, it was not possible to estimate a size for the SLH-binding SCWP,
neither by relation with the disaccharide-heptapeptide unit, nor by SDS–PAGE mobility. However, chemical analysis of its hydrolysis products revealed an essentially polysaccharidic nature, with a minor amino acid content. Whether both, sugars and amino acids, detected as components of the SCWP-layer belong to a single molecule, or to more than one SCWP can not be ruled out at present, but evidences are presented that supports the first hypothesis. First, treatment with a-amilase duplicates the accessibility to aPG, decreases by a 60% the binding of aSAC (Table 1), and decreases the intensity of binding in the affinity assays (Fig. 4).Therefore, there are aamilase sensitive bonds in the SCWP which binds to the SLH motif, despite the rare occurrence of a-1,4 bonds among structural polysaccharides. On the other hand, pronase E has even a greater effect on these three assays, supporting also the relevance of a peptidic moiety of the SLH-binding SCWP, either as part of the antigenic structure detected by aSAC and recognized by the SLH domain, or as components of the PG-anchoring bridges. Future analytical work will indicate if the SCWP from T. thermophilus is related to that of Gram-positive bacteria, of if it has any specific components for the attachment of the S-layer–OM envelope. © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
Attachment of the S-layer to the cell wall of T. thermophilus
685
CsaBth is implicated in pyruvylation of the SCWP-layer The csaB gene of T. thermophilus is clearly required for the pyruvylation of non-peptidoglycan components of raw cell wall preparations, and therefore belonging to the SCWP-layer. Concomitantly, the labelling with aSAC (Fig. 7D) suggested that the main SlpA-bound antigen was also CsaBth-dependent. On the other hand, as the access of aPG was not affected by this mutation (Fig. 7B), we must conclude that the SCWP was still in place in the csaB mutant, but with a change in its composition that affects the main SlpA-bound antigen. Taken together, these data imply that the main SlpA-bound antigen is a pyruvylated compound from the SCWP-layer, whose synthesis or modification depends on the activity of the CsaBth protein. Interestingly, the aSAC was able to bind to the cell wall not just of related Thermus spp. strains, as might be expected, but also to that of D. radiodurans. Although the S-layer subunit of this bacteria, the HPI protein, does not have an SLH domain, at least three SLH-containing proteins (Engelhardt and Peters, 1998) and a CsaB homologue (Mesnage et al., 2000) are encoded within its genome, implying a putative structural and functional relationship between the pyruvylated component of the cell walls of T. thermophilus and D. radiodurans. Binding of SlpA to pyruvylated cell walls The results discussed above suggest that SlpA strongly binds to an SCWP-layer component whose synthesis depends on CsaBth. Accordingly, SlpA protein from both HB27C8 and HB27 strains bound in vitro to cell wall fragments from the csaB mutant with greatly decreased affinity compared to binding to cell wall fragments of the parental strain (Fig. 6). Moreover, the csaB mutant was unable to maintain the attachment of the S-layer–OM complex to the underlying cell wall during growth. On the other hand, in vitro binding to wild-type cell wall fragments was not possible after digestion with a-amylase and pronase E, which, as shown in Figs 1 and 3, eliminates the SCWP-layer from the cell wall surface. Therefore, pyruvylation of the SCWP-layer was required both in vivo and in vitro for binding to the SlpA protein.
Fig. 6. The cell wall of the csaB::kat mutant has low affinity for binding to the SLH domain. The indicated amounts of cell wall preparations from the wild type (HB27) or from the csaB::kat mutant were used to assay their affinity for binding to the S-layer proteins of the HB27C8 (8), HB27 (27), csaB::kat (C) and DSLH (X) strains. A. Coomassie blue staining. B. Binding assays. C. Densitometry of the experiment showed in part B. Each point corresponds to the mean density for the three wild-type S-layer proteins assayed. Symbols: circles, wild-type strains; diamonds, csaB mutant. © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
686 F. Cava et al. bacteria or if it is specific to the S-layer protein from T. thermophilus. The SLH domain of SlpA has specific sequence characteristics (Gly for Pro33, and Tyr for Ala4o) that are common to SLH motifs from proteins identified as porins, and to that of an outer membrane spacer protein from an ‘old’ bacterium, the Thermotoga maritima Ompa protein (Engelhardt and Peters, 1998). Moreover, phylogenetic analysis of 60 SLH motif sequences placed that of the SlpA protein in a cluster near the tree root with those of Ompa from T. maritima and of two hypothetical proteins encoded in the genome of D. radiodurans (Q9RVA5 and Q9RRB6) (not shown). Also, consistent with an ‘old’ status for the SLH domain from the SlpA protein, phylogenetic analysis of its counterpart CsaBth implies a position near the tree root for that of Thermus, close to that of a hypothetical CsaB protein encoded in the genome of Deinococcus radiodurans (Fig. 9). Bearing in mind the ancient origin of the Thermus-Deinococcus phylogenetic group (Hartmann et al., 1989), our data suggest that the pair SLH-motif/CsaBth might represent one of the first systems that co-evolved to bind proteins and associated envelopes to the cell wall non-covalently. In this sense, a less stringent structural requirement to recognize SCWP, shown by the low affinity of SLH for unpyruvylated cell walls of the csaB mutant, may represent an evolutionary relic of such a co-evolutionary process.
Experimental procedures Strains and growth conditions
Fig. 7. Labelling of cell walls from the csaB::kat mutant with aPG and aSAC. aPG; confocal micrographs of purified cell wall fractions from the wild type (A) and the csaB::kat mutant (B). Note the intense labelling of division regions (arrows). aSAC; raw (C and D) or digested with a-amylase and pronase E (E and F) sacculi from the csaB::kat mutant were covalently labelled with succinimidyl ester of Oregon Green, and mixed with undigested (C and D) or digested (E and F) sacculi from the wild-type strain. The sacculi mix was immunolabelled with aSAC and analysed by confocal microscopy. D and F; immunodetection (red channel). C and E; detection of csaB::kat cell walls (green channel).
On the protein face, binding to the pyruvylated SCWP was in all instances dependent on the presence of an SLH motif, as both high- and low-affinity signals were completely abolished in a DSLH derivative. In fact, in the DSLH mutant, as in the csaB mutants, the S-layer–OM complex detaches from the cell during growth (Olabarría et al., 1996), as might be expected from their relationship. At present, we do not know if this residual binding to unpyruvylated SCWP is common to SLH domains of other
Thermus thermophilus HB8 was obtained from the American Type Culture Collection (Rockville, MD). Deinococcus radiodurans, T. aquaticus YT1 and T. thermophilus HB27 were generously given to us by W. Baumister, M. Bothe and Y. Koyama respectively. The T. thermophilus strains HB27C8 and SlpA-X have been described previously (Fernández-Herrero et al., 1995; Olabarría et al., 1996). The Escherichia coli strain DH5aF¢ [F¢, supE44, D(lacZYA-argF)U169 (F80 lacZDM15) hsdR17, recA1, endA1, gyrA96, thi1, relA1] (Bethesda Research Laboratories, Gaithersburg, MD) was used as the host for genetic manipulation of plasmids. Thermus thermophilus strains were grown under strong aeration in rich medium (TB) (Olabarría et al., 1996) at 70∞C. Plates containing agar (1.5%, w/v) were usually incubated in a moist chamber for 24–72 h. For mutant selection, kanamycin (30 mg ml-1) was added to the plates. Escherichia coli was grown in LB (Lennox, 1955) at 37∞C. When a plasmid was present, ampicillin (100 mg ml-1) and/or kanamycin (30 mg ml-1) were added to plates or liquid media. Cells were made competent as described elsewhere (Dagert and Ehrlich, 1979). Deinococcus radiodurans was grown under aeration at 30∞C in LB medium.
Plasmids Plasmid pKT1, used as a source of the kat gene (Lasa et al., © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
Attachment of the S-layer to the cell wall of T. thermophilus
687
Fig. 8. Immunodetection of sacculi from different bacteria with aSAC. Raw cell walls from T. thermophilus HB8 (A), T. thermophilus HB27 (B), T. aquaticus (C) and D. radiodurans (D), were immunolabelled with aSAC. Insets are shown for detail. Bars correspond to 1 mm.
1992), pUC18 (Viera and Messing, 1987) and pCR2.1-TA cloning vector (Invitrogen, USA) were used.
Isolation of csaB mutants A fragment of the csaB gene from T. thermophilus HB27 (Accession number AJ567754) was amplified by PCR with
the primers csaF2 (5¢-GATGTGGGCGAGAAGCC-3¢) and csaRint (5¢-GTGGAGGCGCATGGAGA-3¢), starting at positions -576 and +1313, respectively, with respect to its translational start codon. The amplified 1889 bp DNA fragment was cloned directly into the pCR2.1 vector (TA cloning Kit, Invitrogen), according to the manufacturer’s procedure. For mutant isolation, the fragment was cloned in pUC18 to give
Fig. 9. Comparison of the CsaB proteins of different origin. Phylogenetic tree with the available sequences of CsaB homologues. The CsaBth protein is grouped in the same cluster as that of D. radiodurans, near the tree root. © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
688 F. Cava et al. pUC18csaB, and the kat gene subsequently inserted into the XhoI site of the csaB gene (position +60), leading to the pUC18csaBkat. In this, the kat gene was oriented in the same transcriptional direction as csaB to allow any putative downstream gene to be expressed. Plasmid pUC18csaBkat was linearized by SphI digestion and used to transform the T. thermophilus HB27 strain by natural competence as described elsewhere (de Grado et al., 1999). Kanamycin-resistant colonies grown for 48 h were tested by PCR and Southern blot for the presence of the expected mutation.
Cell wall purification Murein was extracted as insoluble material after three 8 h periods of boiling in 3% (w/v) sodium dodecyl sulphate (SDS), followed each time by centrifugation (80 000 g, 15 min, 30∞C) (Pisabarro et al., 1985; Quintela et al., 1995; de Pedro et al., 1997). The murein was washed by centrifugation until it was free of SDS.
Determination of pyruvic acid in the cell wall of T. thermophilus Pyruvate was extracted from entire cell walls by acid hydrolysis. To do this, 7.75 mg of cell wall preparations from the wild type or the csaB::kat mutant were acid-hydrolysed either in 25 mM oxalic acid for 4 h at 100∞C or in 0.5 M HCl for 3 h at 100∞C, to ensure that all the pyruvate was hydrolysed (Duckworth and Madden, 1993). After neutralization with calcium carbonate and centrifugation, 0.5 ml of supernatant were tested for pyruvate content in a final reaction volume of 2 ml. Lactate dehydrogenase (LDH from rabbit muscle; Roche) was used to measure the pyruvate concentration of the cell walls. The reaction was initiated by the addition of the enzyme (55 units) into prewarmed tubes at 37∞C. NADH consumption over 10 min was monitored by absorbance at 340 nm in a Hitachi U-200 spectrophotometer. Finally, the percentage of pyruvic acid in the cell wall samples was calculated according to Duckworth and Madden (1993).
Antiserum preparation Peptidoglycan analysis Cell wall preparations were treated for 1 h at 37∞C with 100 mg ml-1 of a-amylase (Sigma), followed by 1 h at 60∞C with 100 mg ml-1 of pronase E (Merck). After a further step of SDS treatment to eliminate the enzymes and digestion fragments, the PG fraction was treated with muramidase (40 mg ml-1) from Chalaropsis spp. (Boehringer Mannheim) (Quintela et al., 1995). The muropeptides obtained were analysed by HPLC on a Hypersil RP18 column (250 ¥ 4 mm, 3 mm diameter; Teknochroma, Barcelona, Spain) according to the method of (Glauner, 1988) adapted for Thermus spp. (Quintela et al., 1995).
Quantification of ornithine, amino acids and reducing sugars Murein and muropeptide concentrations were routinely calculated from their content in diamino acid. Samples were hydrolysed in 6 N HCl (12 h, 105∞C), vacuum-dried, and resuspended in distilled water. The concentration of ornithine, the diamino acid component of the peptidoglycan of Thermus spp. (Quintela et al., 1995), was quantified by the ninhydrin assay as described previously (Work, 1957). For the analysis of the SCWP, 10 mg of raw cell walls were digested twice with muramidase from Chalaropsis spp. (40 mg ml-1, 8 h, 37∞C). The insoluble material was recovered by centrifugation and washed twice with distilled water followed by further centrifugation steps. Finally, the material was lyophilized and weighted (2 mg). This material was hydrolysed in 6 N ClH in N2 atmosphere for 14 h at 95∞C, vacuum-dried and resuspended in distilled water. This last washing step was repeated once more to eliminate the remaining HCl. The concentration of reducing sugars was analysed as described (Ghuysen et al., 1966), and that of amino acids with ninhydrin (Moore et al., 1958).
Anti-S-layer and cell wall components (aSAC) was obtained by immunization of New Zealand rabbits with purified SlpA protein subjected to SDS–PAGE. The purification method included digestion of the peptidoglycan with muramidase before solubilization with neutral detergents and ionic exchange column purification (Faraldo et al., 1988). Samples of the purified fraction were further separated in denaturing SDS–PAGE (11% w/vol, 40:1 acrylamide–bisacrylamide), and gel slices containing the protein used for immunization as described elsewhere (Boulard and Lecroisey, 1982).
Cell wall-binding assays Total protein of T. thermophilus HB27C8, HB27, csaB::kat and SlpA-X were solubilized in disruption buffer, separated in an SDS-10% PAGE gel, and electrotransferred to a nitrocellulose membrane (NC45; Serva, Heidelberg) as described previously (Olabarría et al., 1996). The membranes were incubated in renaturation buffer [20 mM Tris-HCl, 150 mM NaCl, 2.5 mM dithiothreitol, 2.5% (v/v) Nonidet P-40, 10% (w/ v) glycerol, and 5% skimmed milk, pH 7.5] for 16 h at 4∞C to allow the proteins to renature, before appropriate concentrations of sonicated (1 ml samples; 0.5 s pulses for 1 min, maximum power in a labsonic sonicator; B. Brawn) cell wall fractions were added in binding buffer [10 mM Tris-HCl, 50 mM, 1 mM dithiotreitol, 1 mM EDTA, 5% (w/v) glycerol, and 0.125% (w/v) skimmed milk, pH 8]. Incubation was maintained for 3 h at room temperature. Finally, protein-attached cell wall fragments were identified with a specific antipeptidoglycan (aPG) antiserum (Olabarría et al., 1996; de Pedro et al., 1997) and revealed by the ECL method (Amersham).
Optical, confocal and electron microscopy Thermus thermophilus HB27 and csaB::kat were grown at 70∞C under mild stirring conditions. Samples were fixed with formaldehyde (0.25% v/v) and immobilized inside a 1% (w/v) © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
Attachment of the S-layer to the cell wall of T. thermophilus agarose gel-matrix before observation by phase-contrast optical microscopy. For confocal microscopy, immunodetection on purified cell walls was performed as described (de Pedro et al., 1997), with adequately diluted aPG or aSAC antisera. Bound antibodies were detected with Alexa 594-conjugated goat antirabbit antibodies. For identification in mixed samples, cell walls from the mutant or wild-type strains were first labelled with the NH2-specific reagent Oregon Green 488 carboxylic acid succinimidyl ester (Molecular Probes Europe BV, Leiden, the Netherlands), before immunolabelling. Microscopy was carried out under a Leica TCS-NT confocal microscope. For electron microscopy the method described by (de Pedro et al., 1997) was followed. In essence, purified cell walls, treated or not with enzymes as required, were adhered to carbon-pioloform-coated grids and immunolabelled with diluted aPG or aSAC antisera for 60 min. After washing, bound antibodies were detected with 5- to 10-nm gold-protein A conjugates, and the presence of the cell walls further contrasted by staining in 1% (wt/vol) uranyl acetate. Microscopic observations were performed on a Philips CM10 transmission electron microscope at an acceleration voltage of 40 or 60 kV.
Phylogenetic analysis Sequence comparisons of the CsaB proteins and SLH domains were carried out at the website http:// www.ncbi.nlm.nih.gov/entrez/ using the BLAST program (Basic Local Alignment Search Tool): http:// www.ncbi.nlm.nih.gov/BLAST/. Multiple alignment was carried out using the CLUSTALW program (Thompson et al., 1994). Phylogenetic trees were constructed in accordance with the maximum probability, maximum parsimony, and long distance methods with the Prodom service at the remote URL: http://protein.toulouse.inra.fr/cgi-bin/ (Gasteiger et al., 2001), and the GeneBee TreeTop program for phylogenetic tree prediction at the remote URL: http://www.genebee.msu.su/ services/phtree_reduced.html.
Acknowledgements This work was supported by grants number BIO2001-1627 to J. Berenguer and BMC2001-2264 to M. A. de Pedro, and by an institutional grant from the Fundación Ramón Areces to CBMSO. F. Cava holds a fellowship from the Ministerio de Educación, Cultura, y Deporte, Spain.
References Baumeister, W., and Kübler, O. (1978) Topographic study of the cell surface of Micrococcus radiodurans. Proc Natl Acad Sci USA 75: 5525–5528. Baumeister, W., Barth, M., Hegerl, R., Guckenberger, R., Hahn, M., and Saxton, W.O. (1986) Three-dimensional structure of the regular surface layer (HPI layer) of Deinococcus radiodurans. J Mol Biol 187: 241–250. © 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
689
Boulard, C., and Lecroisey, A. (1982) Specific antisera produced by direct immunization with slices of polyacrilamide gel containing small amounts of proteins. J Immunol Meth 50: 221–226. Castán, P., de Pedro, M.A., Risco, C., Valles, C., Fernandez, L.A., Schwarz, H., and Berenguer, J. (2001) Multiple regulatory mechanisms act on the 5¢ untranslated region of the S-layer gene from Thermus thermophilus HB8. J Bacteriol 183: 1491–1494. Castán, P., Zafra, O., Moreno, R., de Pedro, M.A., Valles, C., Cava, F., et al. (2002) The periplasmic space in Thermus thermophilus: evidence from a regulation-defective S-layer mutant overexpressing an alkaline phosphatase. Extremophiles 6: 225–232. Castón, J., Carrascosa, J., de Pedro, M.A., and Berenguer, J. (1988) Identification of a crystalline layer on the cell envelope of the thermophilic eubacterium Thermus thermophilus. FEMS Lett 51: 225–230. Castón, J.R., Berenguer, J., de Pedro, M.A., and Carrascosa, J.L. (1993) The S-layer protein from Thermus thermophilus HB8 assembles into porin-like structures. Mol Microbiol 9: 65–75. Castón, J.R., Olabarría, G., Lasa, I., Carrascosa, J.L., and Berenguer, J. (1996) Differential domain accessibility to monoclonal antibodies in three different morphological assemblies built up by the S-layer protein of Thermus thermophilus HB8. J Bacteriol 178: 3654– 3657. Chauvaux, S., Matuschek, M., and Beguin, P. (1999) Distinct affinity of binding sites for S-layer homologous domains in Clostridium thermocellum and Bacillus anthracis cell envelopes. J Bacteriol 181: 2455–2458. Dagert, M., and Ehrlich, S. (1979) Prolonged incubation in calcium chloride improves the competence of Escherichia coli cells. Gene 6: 23–28. Duckworth, M., and Madden, J.K. (1993) Determination of pyruvic acid in complex polysaccharides. In Methods in Carbohydrate Chemistry, Vol. IX. Whistler, R.L., (ed.). New York: John Wiley & Sons, pp. 123–127. Engelhardt, H., and Peters, J. (1998) Structural research on surface layers: a focus on stability, surface layer homology domains, and surface layer–cell wall interactions. J Struct Biol 124: 276–302. Faraldo, M.L., de Pedro, M.A., and Berenguer, J. (1988) Purification, composition and Ca2+-binding properties of the monomeric protein of the S-layer of Thermus thermophilus. FEBS Lett 235: 117–121. Fernández-Herrero, L.A., Olabarría, G., Castón, J.R., Lasa, I., and Berenguer, J. (1995) Horizontal transference of Slayer genes within Thermus thermophilus. J Bacteriol 177: 5460–5466. Gasteiger, E., Jung, E., and Bairoch, A. (2001) SWISS-PROT: connecting biomolecular knowledge via a protein database. Curr Issues Mol Biol 3: 47–55. Ghuysen, J.-M., Tripper, D.J., and Strominger, J.L. (1966) Enzymes that degrade cell walls. Methods Enzymol 8: 685–699. Glauner, B. (1988) Separation and quantification of muropeptides with high-performance liquid chromatography. Anal Biochem 172: 451–464. de Grado, M., Castán, P., and Berenguer, J. (1999) A high-
690 F. Cava et al. transformation efficiency cloning vector for Thermus. Plasmid 42: 241–245. Hartmann, R.K., Wolters, B., Kroger, B., Shultze, S., Spetch, T., and Erdmann, V.A. (1989) Does Thermus represent another deep eubacterial branching? Syst Appl Microbiol 11: 243–249. Lasa, I., Castón, J.R., Fernandez-Herrero, L.A., de Pedro, M.A., and Berenguer, J. (1992) Insertional mutagenesis in the extreme thermophilic eubacteria Thermus thermophilus. Mol Microbiol 11: 1555–1564. Lennox, E.X. (1955) Transduction of linked genetic characters of the host by bacteriophage P1. Virology 1: 190–206. Lupas, A., Engelhardt, H., Peters, J., Santarius, U., Volker, S., and Baumeister, W. (1994) Domain structure of the Acetogenium kivui surface layer revealed by electron crystallography and sequence analysis. J Bacteriol 176: 1224– 1233. Mesnage, S., Tosi-Couture, E., and Fouet, A. (1999a) Production and cell surface anchoring of functional fusions between the SLH motifs of the Bacillus anthracis S-layer proteins and the Bacillus subtilis levansucrase. Mol Microbiol 31: 927–936. Mesnage, S., Tosi-Couture, E., Mock, M., and Fouet, A. (1999b) The S-layer homology domain as a means for anchoring heterologous proteins on the cell surface of Bacillus anthracis. J Appl Microbiol 87: 256–260. Mesnage, S., Fontaine, T., Mignot, T., Delepierre, M., Mock, M., and Fouet, A. (2000) Bacterial SLH domain proteins are non-covalently anchored to the cell surface via a conserved mechanism involving wall polysaccharide pyruvylation. EMBO J 19: 4473–4484. Moore, S., Spackman, D.H., and Stein, W.H. (1958) Chromatography of amino acids on sulphonated polysterene resins. Anal Chem 30: 1185–1190. Olabarría, G., Carrascosa, J.L., de Pedro, M.A., and Berenguer, J. (1996) A conserved motif in S-layer proteins is involved in peptidioglycan binding in Thermus thermophilus. J Bacteriol 178: 4765–4772. de Pedro, M.A., Quintela, J.C., Holtje, J.V., and Schwarz, H. (1997) Murein segregation in Escherichia coli. J Bacteriol 179: 2823–2834. Peters, J., Peters, M., Lottspeich, F., Shafer, W., and Baumeister, W. (1987) Nucleotide sequence analysis of the gene encoding the Deinococcus radiodurans surface
protein, derived amino acids sequence and complementary protein chemical studies. J Bacteriol 169: 5216– 5223. Pisabarro, A.G., de Pedro, M.A., and Vazquez, D. (1985) Structural modifications in the peptidoglycan of Escherichia coli associated with changes in the state of growth of the culture. J Bacteriol 161: 238–242. Quintela, J.C., Pittenauer, E., Allmaier, G., Arán, V., and de Pedro, M.A. (1995) Structure of peptidoglycan from Thermus thermophilus HB8. J Bacteriol 177: 4947–4962. Ries, W., Hotzy, C., Schocher, I., Sleytr, U.B., and Sára, M. (1997) Evidence that the N-terminal part of the S-layer protein from Bacillus stearothermophilus PV72/p2 recognizes a secondary cell wall polymer. J Bacteriol 179: 3892– 3898. Sára, M. (2001) Conserved anchoring mechanisms between crystalline cell surface S-layer proteins and secondary cell wall polymers in Gram-positive bacteria? Trends Microbiol 9: 47–49. Sára, M., and Sleytr, U.B. (2000) S-layer proteins. J Bacteriol 182: 859–868. Sára, M., Egelseer, E.M., Dekitsch, C., and Sleytr, U.B. (1998) Identification of two binding domains, one for peptidoglycan and another for a secondary cell wall polymer, on the N-terminal part of the S-layer protein SbsB from Bacillus stearothermophilus PV72/p2. J Bacteriol 180: 6780–6783. Sleytr, U.B., and Beveridge, T.J. (1999) Bacterial S-layers. Trends Microbiol 7: 253–260. Steindl, C., Schaffer, C., Wugeditsch, T., Graninger, M., Matecko, I., Muller, N., and Messner, P. (2002) The first biantennary bacterial secondary cell wall polymer and its influence on S-layer glycoprotein assembly. Biochem J 368: 483–494. Thompson, J.D., Higgins, D.G., and Gibson, T.J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positionspecific gap penalties and weight matrix choice. Nucleic Acids Res 22: 4673–4680. Viera, J., and Messing, J. (1987) Production of single stranded plasmid DNA. Methods Enzymol 153: 3–11. Work, E. (1957) Reaction of ninhydrin in acid solution with straight-chain amino acids containing two amino groups and its application to the estimation of ae-diaminopimelic acid. Biochemistry 67: 416–423.
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 677–690
