S-layer glycan-specific loci on the chromosome of Geobacillus stearothermophilus NRS 2004/3a and dTDP-L-rhamnose biosynthesis potential of G. stearothermophilus strains

René Novotny1, Christina Schäffer1, Joseph Strauss2 and Paul Messner1

1 Center for NanoBiotechnology, University of Applied Life Sciences and Natural Resources, A-1180 Wien, Austria
2 Center of Applied Genetics, University of Applied Life Sciences and Natural Resources, A-1190 Wien, Austria

Correspondence
Christina Schäffer
christina.schaeffer{at}boku.ac.at


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The ~16·5 kb surface layer (S-layer) glycan biosynthesis (slg) gene cluster of the Gram-positive thermophile Geobacillus stearothermophilus NRS 2004/3a has been sequenced. The cluster is located immediately downstream of the S-layer structural gene sgsE and consists of 13 ORFs that have been identified by database sequence comparisons. The cluster encodes dTDP-L-rhamnose biosynthesis (rml operon), required for building up the polyrhamnan S-layer glycan, as well as for assembly and export of the elongated glycan chain, and its transfer to the S-layer protein. This is the first report of a gene cluster likely to be involved in the glycosylation of an S-layer protein. There is evidence that this cluster is transcribed as a polycistronic unit, whereas sgsE is transcribed monocistronically. To get insights into the regulatory mechanisms underlying glycosylation of the S-layer protein, the influence of growth temperature on the S-layer was investigated in seven closely related G. stearothermophilus strains, of which only strain NRS 2004/3a possessed a glycosylated S-layer. Chromosomal DNA preparations of these strains were screened for the presence of the rml operon, because L-rhamnose is a frequent constituent of S-layer glycans. From rml-positive strains, flanking regions of the operon were sequenced. Comparison with the slg gene cluster of G. stearothermophilus NRS 2004/3a revealed sequence homologies between adjacent genes. The temperature inducibility of S-layer protein glycosylation was investigated in those strains by raising the growth temperature from 55 °C to 67 °C; no change of either the protein banding pattern or the glycan staining behaviour was observed on SDS-PAGE gels, although the sgsE transcript was several-fold more abundant at 67 °C. Cell-free extracts of the strains were capable of converting dTDP-D-glucose to dtdp-L-rhamnose. Taken together, the results indicate that the rml locus is highly conserved among G. stearothermophilus strains, and that in the investigated rml-containing strains, dTDP-L-rhamnose is actively synthesized in vitro. However, in contrast to previous reports for G. stearothermophilus wild-type strains, an increase in growth temperature did not switch an S-layer protein phenotype to an S-layer glycoprotein phenotype, via the de novo generation of a new S-layer gene sequence.


Abbreviations: PAS, periodic acid-Schiff; TPR, tetratricopeptide repeat

The GenBank accession numbers for the sequences reported in this paper are: slg sequence, AF328862; rml operon-inclusive-flanking sequences, AY278518 and AY278519.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Prokaryotic cells are frequently covered by a regular, crystalline array of proteinaceous subunits, termed the ‘S-layer’ (Sleytr et al., 2002; Sleytr & Messner, 2003). In many organisms, the S-layer consists solely of proteins. However, in some organisms the proteins are modified, and glycosylation is the most common and most important modification. Originally, protein glycosylation was considered to be restricted to eukaryotic cells, but the existence of glycosylated prokaryotic proteins is now firmly established (Benz & Schmidt, 2001; Messner & Schäffer, 2000, 2003; Schäffer & Messner 2001; Power & Jennings, 2003). Among the approximately 60 S-layer protein structural genes from different organisms that have been sequenced up to now, only eight relate to glycosylated S-layers. Five of them are of archaeal origin: the S-layer structural genes of Halobacterium halobium, Haloferax volcanii, Methanothermus fervidus, Methanothermus sociabilis and Haloarcula japonica. Three S-layer structural genes are of bacterial origin, belonging to Thermoanaerobacter kivui, Geobacillus stearothermophilus NRS 2004/3a (Schäffer et al., 2002), and G. stearothermophilus ATCC 12980/G+ (Egelseer et al., 2001).

Previously, a set of 39 G. stearothermophilus strains was analysed for the occurrence of an S-layer (Messner et al., 1984). Most of the S-layer proteins, including that of the type culture strain G. stearothermophilus ATCC 12980T, revealed a single, non-glycosylated protein band on denaturing SDS-PAGE gels, with substantial differences in molecular mass among the strains. G. stearothermophilus NRS 2004/3a was the exception among the strains tested in yielding four distinct protein bands with apparent molecular masses of 93, 119, 147 and 170 kDa, respectively (Schäffer et al., 2002), with the three high molecular mass bands showing a periodic acid-Schiff (PAS) staining reaction for carbohydrates. The structure of the G. stearothermophilus NRS 2004/3a S-layer glycan has recently been elucidated by nuclear magnetic resonance (NMR) spectroscopy. It is a homopolymer, composed of, on average, 15 [->2)-{alpha}-L-rhap-(1->3)-{beta}-L-rhap-(1->2)-{alpha}-L-Rhap-(1->] trisaccharide repeating units. The polyrhamnan chains are O-glycosidically linked to the S-layer protein via a trisaccharide core region, consisting of two L-rhamnoses and a {beta}-D-galactose residue serving as linkage glycose. On each S-layer protomer of the oblique S-layer lattice, there are two potential glycosylation sites, at positions Thr-620 and Ser-794 of the S-layer protein precursor (Schäffer et al., 2002). Based on knowledge of the glycan primary structure and the S-layer structural gene sgsE, in combination with the localization of the glycosylation sites on the mature protein, we have chosen G. stearothermophilus NRS 2004/3a as a suitable candidate for studying details of the S-layer protein glycosylation process. Interestingly, another glycosylated S-layer protein has recently been reported for G. stearothermophilus, which has been designated ATCC 12980/G+. This organism possesses a rhamnose-containing S-layer glycoprotein and has been described to be a growth temperature-inducible variant of the non-glycosylated parent strain ATCC 12980T (Egelseer et al., 2001).

In general, L-rhamnose is a common constituent of S-layer glycoproteins (Messner & Schäffer, 2000, 2003), other bacterial surface polysaccharides (Raetz & Whitfield, 2002) and some spore glycoproteins (Fox et al., 2003). Four enzymes, encoded by the rml genes, act sequentially to synthesize dTDP-L-rhamnose, which serves as the nucleotide-activated sugar precursor required for the incorporation of L-rhamnose into growing glycan chains (Giraud & Naismith, 2000). The rmlABCD genes are often organized as an operon on the bacterial chromosome, allowing their coordinated expression. The biosynthetic pathway of dTDP-L-rhamnose has been described in detail for Gram-negative (Glaser & Kornfeld, 1961; Kornfeld & Glaser, 1961; Giraud & Naismith, 2000) as well as for Gram-positive bacteria (e.g. Tsukioka et al., 1997a, b; Graninger et al., 2002).

In this paper, we report the determination of the nucleotide sequence and the genetic assignment of the complete S-layer glycosylation (slg) gene cluster of G. stearothermophilus NRS 2004/3a. Based on the proposal of Egelseer et al. (2001), we have examined in this context growth temperature as a potential regulatory factor for S-layer protein glycosylation in different G. stearothermophilus strains. The selected strains have been screened for the presence of the rml genes, and the conversion of substrates specific for the dTDP-L-rhamnose pathway has been determined at different growth temperatures using cell-free extracts of the strains. The results are discussed in the context of a putative glycosylation pathway for the S-layer protein of G. stearothermophilus NRS 2004/3a.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and growth conditions.
G. stearothermophilus strains NRS 2004/3a and NRS 106/1b2 were obtained from the N. R. Smith Collection, US Department of Agriculture, Peoria, IL (Messner et al., 1984), and the G. stearothermophilus strain ATCC 12980T from the American Type Culture Collection, Manassas, VA. The S-layer-deficient variant ATCC 12980/S was from our own strain collection (Egelseer et al., 2000). G. stearothermophilus strains L32-65, E8-65, and S51-66 were obtained from F. Hollaus, Österreichisches Zuckerforschungs-Institut, Tulln, Austria (Messner et al., 1984). All strains were grown on modified S-VIII medium (1 % peptone, 0·5 % yeast extract, 0·5 % meat extract, 0·13 % K2HPO4, 0·01 % MgSO4, 0·06 % sucrose) to mid-exponential growth phase at 55 °C or 67 °C. Cells were separated from the culture broth by centrifugation for 15 min at 4300 g at a temperature of 4 °C. The biomass was stored at –20 °C. All strains were analysed for the presence of an S-layer and for glycoconjugates by Coomassie blue and PAS staining, respectively, after electrophoresis of biomass by SDS-PAGE (Altman et al., 1995).

Pseudomonas aeruginosa PAO1 was grown in PPGAS medium (1 % peptone, 0·5 % glucose, 0·02 M NH4Cl, 0·02 M KCl, 0·0016 M MgSO4, 0·12 M Tris/HCl; pH 7·2) at 29 °C for 48 h (Rahim et al., 2001) and served as a control in the assay for lipophilic glycoconjugates.

DNA manipulation.
Chromosomal DNA from G. stearothermophilus strains was isolated using Genomic Tips 100/G (Qiagen). Generation of G. stearothermophilus genomic libraries was performed using the restriction endonuclease digestion and ligation protocols described by Sambrook et al. (1989). Restriction enzymes and T4 DNA ligase were obtained from Invitrogen.

PCR and DNA sequencing.
Primers for DNA sequencing were purchased from Invitrogen. For the identification of the genes responsible for dTDP-L-rhamnose biosynthesis, the highly conserved seven amino acid stretch YDKPMIY of RmlA, and the six amino acid stretch TDEVYG of RmlB, were used to design the degenerate oligonucleotide primers wRmlA/for2 (5'-TAYGAYAARCCNATGATHTAY-3') and wRmlB/rev2 (5'-CCRTANACYTCRTCNGT-3'), where N=A/C/G/T, Y=C/T, R=A/G and H=A/C/T. Based on the knowledge that in some bacterial strains the genes responsible for glycosylation are close to the gene encoding the specific target protein (Benz & Schmidt, 2001; Power & Jennings, 2003), we also used the S-layer gene specific primers sgsE/upstream (5'-TTCATTTGCGCTTTCGCTTG-3') and sgsE/downstream (5'-GTCTAAACCGATTACGCTGTCC-3') to sequence the flanking regions of the G. stearothermophilus NRS 2004/3a S-layer structural gene sgsE (GenBank accession no AF328862). For sequence analysis of the glycosylation cluster, chromosome walking was applied as previously described (Kneidinger et al., 2001), except that pBluescript II SK+ was used as the subcloning vector. Oligonucleotide primers used for chromosome walking of G. stearothermophilus NRS 2004/3a are not listed.

To screen for a conserved N-terminal region of the S-layer protein in the different G. stearothermophilus strains, the primer combinations S-layer_forward/S-layer_reverse (5'-TCGTCAGCCAAGCGAAAG-3'/5'-TTCGGCTTTGAACGCATC-3') and S-layer_forward2 (5'-GCGAAAGATCTGAAAAAAGCAG-3')/S-layer_reverse were chosen. The oligonucleotides were designed on the basis of a conserved region, consisting of approximately 800 bp, which is present in four of five known G. stearothermophilus S-layer genes (GenBank accession nos X71092, AF055578, AF228338, AF328862). sbsC and sbsD/sgsE were identified using the primer combinations S-layer_forward/S-layer (sbsC)_reverse (5'-TGGTACTGGCTTGTTGTAAGTG-3') and S-layer_forward/S-layer (sbsD)_reverse (5'-ATTTCATCGTCGCCAGCC-3'), respectively. PCR amplification was carried out with Taq-polymerase using a PCR Sprint thermocycler (Hybaid). Double-stranded DNA sequencing was performed by Agowa.

Sequence analysis.
Nucleotide and protein sequences were analysed using the BLASTN and BLASTP on-line sequence homology analysis tools (National Center for Biotechnology Information). The DNA sequence was translated in all six frames, and all ORFs of greater than 250 bp were examined. The TMHMM v2.0 transmembrane prediction program was used to identify putative protein transmembrane-spanning domains. The percentage G+C base composition of each gene in the slg gene cluster was determined using the Gene Runner program (Hastings Software).

Nucleotide sequence accession numbers.
The nucleotide sequence of the slg gene cluster of G. stearothermophilus NRS 2004/3a has been deposited in GenBank under the accession number AF328862. The GenBank accession numbers of the dTDP-L-rhamnose (rml) operon-inclusive-flanking regions of G. stearothermophilus L32-65 and G. stearothermophilus ATCC 12980T are AY278518 and AY278519, respectively.

Northern blot analysis.
Total RNA was extracted from G. stearothermophilus NRS 2004/3a grown at 55 °C or 67 °C using the RNeasy Midi kit (Qiagen) and subsequently treated with RNase-free DNase I (Promega) to exclude DNA contamination. Northern blotting and hybridization was carried out as described previously (Strauss et al., 1999). DNA probes were obtained by PCR for sgsE (primer pair, S1, 5'-CGACGATGAAATGACAATCAAC-3'/S2, 5'-GGACAGCGTAATCGGTTTAGAC-3'), ORFG102 (primer pair, P1, 5'-ATGTCTGATTGGTATAAGTATTTAAATATTG-3'/P2, 5'-TCAACTCACTTTTTCCTGCTTAT-3'), and ORFG113 (primer pair, P3, 5'-GTGGTTAAGGTGATTAGAGGAAGA-3'/P4, 5'-CTAATATGCATTTTTATTTACCAAACC-3'). The probes were labelled with [32P]dCTP by random priming using Ready-To-Go DNA labelling beads (Amersham). Linearized vectors containing sgsE, ORFG102 and ORFG113 were used as positive controls for probe specificity. As a control for RNA loading, 16S rRNA was used, and the probe was amplified by PCR using the primers 16S_forward (5'-AGAGTTTGATCCTGGCTCAG-3')/16S_reverse (5'-ACGGTTACCTTGTTACGAC-3'). Genomic DNA from G. stearothermophilus NRS 2004/3a served as a control for the blotting efficiency of large fragments.

RT-PCR.
First strand cDNA was synthesized from 1 µg DNase I-treated total RNA (extracted from G. stearothermophilus NRS 2004/3a grown at 55 °C) in the presence of 10 mM dNTPs and 1 µM of a reverse primer specific for sgsE (S2), ORFG102 (P2), or rmlA (B1, 5'-TCAATAATTTTCAATCAATTCACCTTCTT-3'), using M-MuLV reverse transcriptase (MBI Fermentas) according to the manufacturer's instructions. After heat-inactivation of the enzyme at 70 °C for 10 min, one tenth of each cDNA reaction mixture was used as template for PCR using the REDTaq ReadyMix PCR reaction mix (Sigma-Aldrich). PCR reactions were carried out with primer pairs specific for sgsE (S3, 5'-TCGTCAGCCAAGCGAAAG-3'/S4, 5'-ATTTCATCGTCGCCAGCC-3'), ORFG101–ORFG102 (A4, 5'-AGTATCAACAGGCGATTCAG-3'/A3, 5'-CGTAAGAACAGCCGAAATCCC-3'), sgsE–ORFG102 (A5, 5'-ACGCACGTCATTACGATCAG-3'/A3), and ORFG107–rmlA (B4, 5'-CACCACATCCAAGTTATCCACC-3'/B2, 5'-CTCTCCAAGTTCTAAATACGCC-3'). Each PCR reaction included genomic DNA as a positive control, RNA and DNase I-treated RNA (without the RT-PCR cDNA-generating step) as a control for contamination of total RNA with chromosomal DNA, which may result in false-positive amplification products in the final PCR amplification step. PCR products were analysed on agarose electrophoresis gels containing 0·9 % ethidium bromide.

Preparation of cell-free extracts.
G. stearothermophilus strains were grown to the mid-exponential phase. Cells were harvested from 400 ml of culture broth by centrifugation at 4300 g for 20 min at 4 °C, followed by two washes with 10 mM Tris/HCl (pH 7·6), containing 5 mM MgCl2 (buffer A), recentrifugation, and storage at –20 °C until use. To analyse the G. stearothermophilus strains for the ability to produce dTDP-L-rhamnose, 0·7 g frozen cells was resuspended in 7·5 ml buffer A and broken by ultrasonication using a Branson model 450 Sonifier (5 min, intensity 6, 50 % duty cycle, chilled on ice). Large cellular debris was separated by centrifugation at 4300 g for 20 min at 4 °C in an Eppendorf 5804 R centrifuge. Subsequently, cell membranes were removed by ultracentrifugation at 331 000 g for 40 min at 4 °C in a Beckman LE-80 ultracentrifuge. The total protein concentration of the supernatant was determined by the method of Bradford (1976) using the Bio-Rad Protein Assay with BSA as a standard.

HPLC assay for formation of dTDP-L-rhamnose.
Starting with dTDP-D-glucose as a substrate, the combined activities of the rmlB, rmlC and rmlD gene products were confirmed by the detection of dTDP-L-rhamnose (Graninger et al., 2002). The reaction was performed using cell-free extracts of G. stearothermophilus strains as a source of enzymes. Each reaction mixture contained 15 nmol of dTDP-D-glucose, 15 nmol NAD+ and about 1 mg of total protein from strains grown at either 55 °C or 67 °C, in a final volume of 0·3 ml of buffer A. Incubation was performed at 37, 50 and 55 °C. After 1 h, 75 nmol NADPH was added and the reaction mixture was incubated for an additional 1 h at the respective temperature. After removal of the protein by centrifugation using Ultrafree-MC 10 000 ultracentrifugation cartridges (Millipore), 75 µl of reaction mixture was analysed by HPLC on a CarboPac PA-1 column (Dionex) to determine the extent of conversion of dTDP-D-glucose to dtdp-L-rhamnose (Graninger et al., 2002). dTDP-D-glucose, NADPH and NAD+ were obtained from Sigma.

Assay for lipophilic rhamnose-containing glycoconjugates.
The rml-positive G. stearothermophilus strains NRS 2004/3a, ATCC 12980T, ATCC 12980/S and L32-65 were grown to mid-exponential phase at 55 °C. P. aeruginosa PAO1 served as a positive control in the assay (Rahim et al., 2001). Cells were harvested by centrifugation, and the culture supernatants were passed through a 0·22 µm filter. To test for lipophilic rhamnose-containing glycoconjugates, the cell pellet was resuspended in 15 ml CHCl3/MeOH (1 : 1, v/v). The cells were broken by ultrasonication on ice (micro-tip, intensity 3, 50 % duty cycle, 5 min) and stirred for 30 min at 25 °C. After centrifugation for 20 min at 4300 g at 4 °C, the supernatant was taken off and stored at 4 °C. The lipid extraction of the pellet was repeated with 15 ml CHCl3/MeOH (1 : 2, v/v) and 15 ml CHCl3/MeOH/H2O (48 : 35 : 1, by volume). Supernatants were combined, filtered through a 0·22 µm filter, evaporated to dryness and resuspended in 15 ml CHCl3/MeOH/H2O (60 : 30 : 5, by volume). After phase separation by centrifugation at 4300 g for 20 min, the organic and the water phases were dried separately and then hydrolysed with 25 % trifluoroacetic acid (TFA) for 4 h at 110 °C. TFA was removed under a stream of nitrogen and the samples were resuspended in 1 ml CH2Cl2/H2O (1 : 1, v/v). The upper phase was separated from the lipid material by repeated extraction with 0·5 ml H2O, dried again and resuspended in 60 µl H2O. Samples of 1 µl of the lipid material as well as of the hydrolysed water phase were applied to precoated silica gel 60 TLC plates (Merck) and developed in a butanol/pyridine/0·1M HCl (50 : 30 : 20, by volume) solvent system. Sugars were detected down to the nanomolar range with thymol reagent (Winzler, 1955) and heating of the TLC plate to 110 °C for 10 min. A mixture of rhamnose, glucose and galactose served as the standard for TLC analysis.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Identification of the G. stearothermophilus NRS 2004/3a S-layer glycosylation (slg) gene cluster
Based on the finding that L-rhamnose represents the major constituent of the S-layer glycan of G. stearothermophilus NRS 2004/3a, we chose the genes responsible for dTDP-L-rhamnose biosynthesis as suitable candidates for the design of degenerate primers. Using genomic DNA of G. stearothermophilus NRS 2004/3a as a template for the PCR amplification reaction, a 1·75 kb DNA fragment was obtained. After confirmation of the presence of the rmlACB target genes, sequencing of upstream and downstream regions of the operon revealed the presence of genes which may code for additional components of the S-layer protein glycosylation machinery. In combination with the overlapping sequence data, obtained by chromosome walking starting from the S-layer structural gene sgsE, a DNA region of approximately 16·7 kb was obtained. Sequencing of the entire region identified 14 ORFs, all of which were transcribed in the same direction (Fig. 1). The nucleotide sequence has been deposited in GenBank under the accession number AF328862. For the opposite strand of the same DNA region, no ORFs of significant length and function were found. Most of the putative gene products encoded by the assigned ORFs showed high homology to proteins involved in the biosynthesis of bacterial surface polysaccharides. Based on these sequence similarities, putative biological functions could be assigned to nearly all of the genes of the gene cluster (Table 1). Interestingly, analysis of the G+C content of the individual ORFs revealed significantly lower values (mean 35 mol%) (Fig. 1) than for the rest of the G. stearothermophilus NRS 2004/3a genome (about 53 mol%).



View larger version (22K):
[in this window]
[in a new window]
 
Fig. 1. Genetic organization and restriction map of the slg gene cluster of G. stearothermophilus NRS 2004/3a. The locations and directions of gene transcription are represented by arrows; putative transcriptional terminator sequences are shown by a stem–loop symbol. Restriction enzyme sites: S, SacI; P, PstI; B, BamHI; X, XhoI; H, HindIII; E, EcoRI. The G+C content of each ORF is shown in the bar chart below. The dotted line shows the mean G+C content of the genome of G. stearothermophilus NRS 2004/3a.

 

View this table:
[in this window]
[in a new window]
 
Table 1. Predicted gene products encoded by the slg gene cluster of G. stearothermophilus NRS 2004/3a together with database homologies

 
Genetic characterization of the slg gene cluster
The S-layer gene sgsE (Schäffer et al., 2002), which encodes the G. stearothermophilus NRS 2004/3a S-layer structural protein, is located immediately upstream of the slg gene cluster with an intergenic region of 225 nt (Fig. 1). At a point 25 nt downstream of the stop codon of sgsE, a perfect palindromic stem–loop sequence of 17 nt was identified, which may terminate transcription of the gene (Schäffer et al., 2002). These data, derived from a nucleic acid motif search, suggest that the S-layer structural gene is transcribed monocistronically.

The intergenic region between sgsE and ORFG101 revealed no promoter-like sequence in computer analysis. Most of the ORFs, except ORFG107, showed putative ribosome binding sequences (RBS) directly in front of the initiation codons. ORFG101 through ORFG107 of the slg gene cluster are closely spaced one after another, suggesting that they are part of a polycistronic transcription unit.

ORFG101 displays only a low level of homology to proteins in sequence databases. The most significant match is with tetratricopeptide repeat (TPR)-containing proteins (Deckert et al., 1998). TPRs are generally believed to be involved in protein–protein interactions (Blatch & Lässle, 1999). A single transmembrane segment is predicted for the N-terminal region of the TPR-containing protein, suggesting an association with the cell membrane.

The deduced amino acid sequence of ORFG102, including 12 transmembrane segments distributed throughout the protein, shows similarity to both the O antigen ligase WaaL (Heinrichs et al., 1998) and the O-polysaccharide polymerase Wzy (Comstock et al., 1999).

The putative translation product of ORFG103 shows significant homology to several glycosyltransferases involved in cell wall biosynthesis, with the highest similarity to rhamnosyltransferases (Yamashita et al., 1998). At the N-terminus of the protein, a conserved sequence motif, specific for glycosyltransferases of family 2, is found. The hydropathy profile of the protein shows a single hydrophobic domain at the C-terminal region, indicating a transmembrane peptide structure.

The deduced 268 and 409 amino acid proteins encoded by ORFG104 and ORFG105, respectively, reveal high similarity to proteins of the ABC-2 transporter family, involved in the transport of bacterial surface polysaccharides to the cell surface (Bronner et al., 1994). According to the common nomenclature, ORFG104 and ORFG105 were designated wzm and wzt, respectively (Reeves et al., 1996; Rocchetta & Lam, 1997). Six transmembrane domains, evenly distributed in ORFG104, suggest that this protein is the integral membrane component of the transporter.

The putative start codons of ORFG105 and ORFG106 overlap with the stop codons of the preceding genes. In its C-terminal region, the 1127 amino acid translation product of ORFG106 shows high similarity to the RfbC protein of other organisms (Guo et al., 1996; Xu et al., 1998). Two regions homologous to glycosyltransferase family 2 proteins were identified between amino acids 604 to 765 and 863 to 1043. Interestingly, the N-terminal region of the protein shows low homology to several methyltransferases (Becker et al., 1997).

The 413 amino acid gene product of ORFG107 shows only a few similarities to proteins in the databases. Low homologies were found to the 1275 amino acid RfbC protein of Myxococcus xanthus (Guo et al., 1996) and the 421 amino acid WbbX protein of Yersinia enterocolitica (Zhang et al., 1993). However, the C-terminal part of the protein contains a glycosyltransferase group 1 motif, indicating its participation in a yet-unknown transfer process for sugar residues. A 260 nt non-coding DNA region precedes the following ORF, but no promoter-like sequence was found.

The ORFG108, ORFG109, ORFG110 and ORFG111 gene products show a high degree of amino acid homology to the RmlACBD proteins involved in the biosynthesis of dTDP-L-rhamnose by Aneurinibacillus thermoaerophilus DSM 10155 (Graninger et al., 2002) and other micro-organisms (Xu et al., 1998; Jiang et al., 2001). The four genes were designated rmlA, rmlC, rmlB and rmlD, respectively, in accordance with the established nomenclature (Giraud & Naismith, 2000). The putative start codons of rmlC and rmlD overlap with the stop codons of rmlA and rmlB.

Comparison of the translation product of ORFG112 with proteins in the database suggested a homology to several rhamnosyltransferases (Nölling et al., 2001). A glycosyltransferase family 2 motif was identified at the N-terminus, while a single transmembrane-spanning domain was found at the C-terminal part of the protein. The stop codon of ORFG112 overlaps with the putative start codon of ORFG113, which represents the last gene of the slg gene cluster. The deduced amino acid sequence of ORFG113 shows homology to glycosyltransferases in the first steps of O-polysaccharide biosynthesis, which transfer UDP-sugars to a lipid carrier (Wang et al., 1996; Drummelsmith & Whitfield, 1999). This putative glycosyltransferase has five transmembrane segments at conserved positions, four in the N-terminal and one in the central region of the protein.

The slg gene cluster lies upstream of ORFG114, which encodes a fragmentary transposase for the insertion sequence element IS5376, which is similar to ORFA of G. stearothermophilus strain CU21 (Xu et al., 1993). IS5376 is a member of the IS21 family of bacterial insertion sequences (Mahillon & Chandler, 1998). The region downstream of ORFG114 contains an obvious stem–loop sequence, which consists of a 20 nt inverted repeat sequence with two mismatches, resembling a {rho}-independent termination signal. The presence of a transcriptional terminator downstream of the slg gene cluster suggests that the genes involved in the S-layer protein glycosylation of G. stearothermophilus NRS 2004/3a are part of a single, large transcription unit.

Using Northern analysis, we identified the sgsE gene as a monocistronic unit with a predicted mRNA size of about 3 kb. Interestingly, although the protein can be visualized in SDS-PAGE by Coomassie blue staining, the transcript is barely detectable under standard expression conditions of 55 °C, although it becomes abundant at the higher temperature of 67 °C (Fig. 2A). In addition, we used a set of probes derived from ORFG102, ORFG103, rmlA and ORFG113 to characterize the transcription of the putative polycistronic slg gene cluster at growth temperatures of 55 °C and 67 °C. The large size of the expected transcript, at least 16·5 kb, prompted us to use non-digested genomic DNA of G. stearothermophilus NRS 2004/3a, with comparable migration properties, as a control for blotting efficiency and to exclude negative results due to inefficient blotting. Despite the positive probing of the controls, no specific transcripts were detected (data not shown). At this point, we turned to the more sensitive RT-PCR method to identify specific mRNA(s) of the gene cluster (Fig. 2B). Using this technique, we were able to detect the sgsE transcript in cells grown at 55 °C, where Northern analysis failed to reveal a specific transcript. Furthermore, we confirmed that sgsE is a monocistronic unit, because the primer combination A5/A3, which should amplify the segment sgsE–ORFG101, failed to detect a transcript. Using primer combinations spanning the region ORFG101–ORFG102 in the 5' region and ORFG107–rmlA in the 3' region of the putative polycistronic mRNA, we obtained evidence that the slg gene cluster is transcribed as a polycistronic unit, as suggested above.



View larger version (23K):
[in this window]
[in a new window]
 
Fig. 2. Expression analysis of sgsE and the slg gene cluster of G. stearothermophilus NRS 2004/3a. (A) Northern blot analysis of 5 µg total RNA of G. stearothermophilus NRS 2004/3a grown at 55 °C or 67 °C. The blot was hybridized with a 32P-labelled probe specific for sgsE. Non-specific binding was observed for the 16S (~1·5 kb) and 23S (~2·9 kb) rRNAs. A 32P-labelled probe specific for the 16S rRNA gene was used as a control for RNA loading. (B) RT-PCR analysis of total RNA of G. stearothermophilus NRS 2004/3a grown at 55 °C. (1) Reverse transcription of 1 µg total RNA was performed with specific primers targeted to sgsE (S2), ORFG103 (P2) and rlmA (B1). (2) Subsequent cDNA amplification was performed with primer pairs S3/S4, targeted to sgsE (lanes 1–4); with primer pairs A4/A3, targeted to the overlapping transcriptional units ORFG101–ORFG102 (lanes 6-8); and with primer pairs B4/B2, targeted to transcripts rmlA–ORFG107 (lanes 12–14). The reaction with primer pairs A5/A3 was used to show that sgsE is not part of the polycistronic unit (lanes 9–11). Primer positions are depicted as arrows. Lanes (a) show the specific PCR amplification products, using RT single-strand cDNA as template; lanes (b) show control reactions, using DNase I-treated RNA without the RT step as PCR template; lanes (c) show the same as lanes (b), but without DNase I treatment; lanes (+) show positive controls using genomic DNA as template. 1 kb DNA Plus marker (Invitrogen) was used as DNA size marker (lanes 5 and 15).

 
Influence of growth temperature on S-layer protein glycosylation in G. stearothermophilus strains
Since G. stearothermophilus ATCC 12980T, which possesses a non-glycosylated S-layer protein, had been reported to change its phenotype to a variant with a rhamnose-containing S-layer glycoprotein on increasing the growth temperature from 55 °C to 67 °C (Egelseer et al., 2001), we investigated the regulatory effect of growth temperature on S-layer glycoprotein formation in different G. stearothermophilus strains available in our strain collection. This included six S-layer-carrying G. stearothermophilus strains (Messner et al., 1984), as well as the S-layer-deficient variant ATCC 12980/S (Egelseer et al., 2000). Unfortunately, the temperature-inducible variant ATCC 12980/G+ was not available for control purposes in the growth temperature experiments (Egelseer et al., 2001). After cultivation of the organisms using the published experimental procedure, SDS-PAGE analysis of whole-cell extracts showed no differences in the S-layer protein pattern between cultures grown at 55 °C (Fig. 3A) and 67 °C (results not shown). Identification of carbohydrates with PAS reagent gave a positive signal only for strain NRS 2004/3a, due to its glycosylated S-layer protein; all other strains showed no detectable PAS staining reaction (Fig. 3B). This observation also indicated that no other glycoconjugates were present in aqueous extracts of the cells.



View larger version (53K):
[in this window]
[in a new window]
 
Fig. 3. (A) Coomassie blue-stained and (B) PAS-stained SDS-PAGE gels (7·5 %) of whole cells of G. stearothermophilus strains: Mark12 standard (lane 1), NRS 2004/3a (lane 2), ATCC 12980T (lane 3), ATCC 12980/S (lane 4), L32-65 (lane 5), E8-65 (lane 6), S51-66 (lane 7), and NRS 106/1b2 (lane 8). The amounts of total protein loaded were 10 µg (Coomassie blue staining) and 20 µg (PAS staining) per lane. Results are shown for strains grown at 55 °C; identical banding patterns were obtained at 67 °C (not shown).

 
Screening for the dTDP-L-rhamnose biosynthesis operon in G. stearothermophilus strains with non-glycosylated S-layer proteins
Because of the discrepancy between our results (Fig. 3) and the suggested appearance of a rhamnose-containing S-layer glycoprotein in the temperature-induced variant G. stearothermophilus ATCC 12980/G+ (Egelseer et al., 2001), the six strains whose S-layer glycan deficiency was documented were screened by PCR for the presence of the rml operon. For the amplification reaction, the same combination of degenerate primers (wRmlA_for2/wRmlB_rev2) was used as for the amplification of the rmlACB genes in G. stearothermophilus NRS 2004/3a. Chromosomal DNA, originating from strains grown at both 55 °C and 67 °C, served as the template for the PCR reactions.

DNA fragments of approximately 1·65 kb could be amplified from strains ATCC 12980T and ATCC 12980/S, whereas strain L32-65 yielded a 1·75 kb amplification product (not shown). The PCR products of strains ATCC 12980T and L32-65 were sequenced; both contained a 5'-incomplete rmlA gene, rmlC, and the 5'-portion of rmlB. The GenBank accession numbers of the sequences are AY278518 and AY278519, respectively. Further sequencing of the upstream and downstream regions confirmed the presence of the complete rml locus. In both strains, the rml operon is flanked by genes assigned to putative glycosyltransferases which may be involved in the assembly of S-layer glycan chains. In all G. stearothermophilus strains which contain the rml operon, the four genes are highly homologous and are arranged in the order rmlACBD. As ATCC 12980/S is an S-layer-deficient variant of the parent strain (Egelseer et al., 2000), sequencing of this strain was not performed.

Ability of G. stearothermophilus strains to synthesize dTDP-L-rhamnose from dtdp-D-glucose
Based on the detection of the rml operon in G. stearothermophilus strains NRS 2004/3a, ATCC 12980T, ATCC 12980/S and L32-65 by PCR methods, the biosynthetic activity of the proteins encoded by this gene locus was investigated. In addition, G. stearothermophilus strains E8-65, NRS 106/1b2 and S51-66, which gave no amplification products with rmlAB-specific primers, were included in the assay, to allow for the possibility of a different arrangement of the rml genes in the operon. To test their ability to produce dTDP-L-rhamnose in vitro, cell-free extracts of the strains, grown at 55 °C or 67 °C, were used. The combined enzymic activity of RmlB, RmlC and RmlD of each strain was monitored by HPLC at different turnover temperatures (37 °C, 50 °C and 55 °C). Conversion of dTDP-D-glucose to dtdp-L-rhamnose, in the presence of NADPH, was detected by quantitative HPLC (Marumo et al., 1992).

All strains containing the rml operon showed enzymic activity of their rmlB, rmlC and rmlD gene products, demonstrated by the synthesis of dTDP-L-rhamnose. The levels of dTDP-L-rhamnose synthesis of G. stearothermophilus NRS 2004/3a and L32-65 were about three and four times higher, respectively, than those of strains ATCC 12980T and ATCC 12980/S, when equal amounts of total protein were used in the enzyme assay (results not shown). As a representative example, the HPLC elution profiles of G. stearothermophilus NRS 2004/3a, possessing a glycosylated S-layer protein; of strain L32-65, possessing a non-glycosylated S-layer protein; and of the rml-negative strain E8-65, all cultivated at 55 °C, are shown in Fig. 4. As expected, strains lacking the rml genes did not produce dTDP-L-rhamnose, as shown for E8-65 in Fig. 4, profile (D).



View larger version (14K):
[in this window]
[in a new window]
 
Fig. 4. HPLC elution profiles showing the in vitro formation of dTDP-L-rhamnose by combined RmlBCD activities. (A) Cell-free extract of G. stearothermophilus NRS 2004/3a, before addition of dTDP-L-glucose (negative control); (B) assay for G. stearothermophilus NRS 2004/3a (rml-positive, glycosylated S-layer protein); (C) G. stearothermophilus L32-65 (rml-positive, non-glycosylated S-layer protein); (D) G. stearothermophilus E8-65 (rml-negative, non-glycosylated S-layer protein). Strains were grown at 55 °C or 67 °C (not shown), and the conversion of dTDP-D-glucose to dtdp-L-rhamnose was monitored at 50 °C.

 
Detection of lipophilic rhamnose-containing glycoconjugates in G. stearothermophilus strains with enzymically active Rml proteins
The four G. stearothermophilus strains NRS 2004/3a, ATCC 12980T, ATCC 12980/S and L32-65, which had been shown to possess active Rml proteins, were screened for the occurrence of rhamnose-containing glycoconjugates other than S-layer glycoproteins. Among such glycoconjugates, rhamnolipids have been reported to occur in bacteria (Maier & Soberón-Chávez, 2000). P. aeruginosa PAO1, whose rhamnolipid secretion has been well documented (Rahim et al., 2001), was used as a positive control in the assay. For the direct detection of lipophilic rhamnose-containing compounds, culture supernatants and lipid extracts of cell membranes were analysed by one-dimensional TLC. Rhamnose was visualized for P. aeruginosa at Rf=0·77 on thymol staining. No rhamnose was detected in either phase of membrane extracts of the four investigated G. stearothermophilus strains, indicating the absence of lipophilic rhamnose-containing glycoconjugates (results not shown).

Investigation of the influence of growth temperature on the S-layer genotype in G. stearothermophilus strains by PCR analysis
All seven G. stearothermophilus strains included in this study were investigated for the influence of growth temperature on S-layer genotype. Chromosomal DNA and whole cells from cultures grown at 55 °C or 67 °C, were used as templates for PCR. To screen for conserved N-terminal coding regions in the S-layer structural genes, the primers S-layer_forward and S-layer_reverse were used, which consist of conserved regions found in four out of five sequenced G. stearothermophilus S-layer gene sequences.

From six of the seven investigated G. stearothermophilus strains, NRS 2004/3a, ATCC 12980T, ATCC 12980/S, L32-65, S51-66 and NRS 106/1b2, a 0·62 kb DNA fragment could be amplified (Fig. 5A). No PCR product was obtained from G. stearothermophilus E8-65, suggesting that at least one primer sequence is not conserved in that strain. A second PCR screen with the primer pair S-layer_forward2/S-layer_reverse yielded an amplification product for E8-65 also (results not shown).The N-terminus of the S-layer protein therefore represents a conserved region in many G. stearothermophilus strains.



View larger version (28K):
[in this window]
[in a new window]
 
Fig. 5. (A) Analysis of a conserved N-terminal-coding region in the S-layer gene of different G. stearothermophilus strains, using the primer combination S-layer_forward/S-layer_reverse. G. stearothermophilus NRS 2004/3a (lane 1), ATCC 12980T (lane 2), ATCC 12980/S (lane 3), L32-65 (lane 4), E8-65 (lane 5), S51-66 (lane 6), NRS 106/1b2 (lane 7), 1 kb DNA ladder (M). The larger amplification product of strain ATCC 12980/S (lane 3) is due to the presence of an IS element (Egelseer et al., 2000). (B) PCR detection of sbsC- and sbsD, sgsE-specific sequences, respectively, using either a primer combination specific for the sbsC gene of ATCC 12980T (S-layer_forward/S-layer(sbsC)_reverse) or a primer pair specific for the S-layer gene sbsD of the glycosylated variant ATCC 12980/G+ and sgsE of NRS 2004/3a (S-layer_forward/S-layer(sbsD)_reverse). Lanes (a), G. stearothermophilus NRS 2004/3a; lanes (b), G. stearothermophilus ATCC 12980T; lanes (c), G. stearothermophilus ATCC 12980/S. Strains were grown at 55 °C or 67 °C.

 
In order to explain the divergent results observed for the S-layer protein, SbsC, of G. stearothermophilus ATCC 12980T in SDS-PAGE of Coomassie blue and PAS-stained gels in our growth temperature experiments and those described in the literature (Egelseer et al., 2001), we used PCR to screen for the presence of sbsC and sbsD. sbsD, which was proposed to be generated de novo on raising the growth temperature from 55 °C to 67 °C, had been assigned to encode the glycosylated S-layer protein of the variant strain ATCC 12980/G+.

The sbsC-specific primer combination S-layer_forward/S-layer (sbsC)_reverse yielded a PCR product of the expected size for strains ATCC 12980T (Jarosch et al., 2000) and ATCC 12980/S (Egelseer et al., 2000), grown at 55 °C or 67 °C. In contrast, the primer combination S-layer_forward/S-layer (sbsD)_reverse, which is specific for G. stearothermophilus ATCC 12980/G+ (sbsD) and NRS 2004/3a (sgsE), showed a PCR product only for NRS 2004/3a, while no amplification product was obtained for ATCC 12980T (Fig. 5B).

This finding confirms the observation already made at the S-layer protein level in this study, suggesting that in the G. stearothermophilus strains NRS 2004/3a and ATCC 12980T an increase in growth temperature leads neither to de novo generation of a new S-layer DNA sequence, accompanied by a loss of the original sequence, nor to the onset of S-layer protein glycosylation.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
G. stearothermophilus NRS 2004/3a possesses S-layer glycoprotein glycans with the structure 2-O-methyl-{alpha}-L-Rhap-(1->3)-{beta}-L-rhap-(1->2)-{alpha}-L-rhap-(1->[2)-{alpha}-L-Rhap-(1->3)-{beta}-L-Rhap-(1->2)-{alpha}-L-rhap-(1]n~15->2)-{alpha}-L-Rhap-(1->3)-{alpha}-L-Rhap-(1->3)-{beta}-D-Galp-Thr620/Ser794 (Schäffer et al., 2002), which have been shown in this study to be the exclusive source of L-rhamnose in that organism. The molecular characterization of S-layer glycoprotein biosynthesis was previously initiated by sequencing the S-layer structural gene sgsE (Schäffer et al., 2002). In the present work, we have identified a gene cluster of about 16·5 kb, containing 13 polycistronically transcribed ORFs, which probably encode the glycosylation of SgsE. The S-layer structural gene of G. stearothermophilus NRS 2004/3a has been shown in this study to be transcribed monocistronically and independently of the slg gene cluster.

Our inability to transform G. stearothermophilus has so far precluded the detailed characterization of the slg genes and the regulatory factors involved in S-layer glycan biosynthesis. However, comparison with protein database sequences has allowed an almost complete assignment of putative biological function to most of the genes. Four ORFs are organized in the well-characterized bacterial rml operon (Giraud & Naismith, 2000), which codes for the biosynthesis of dTDP-L-rhamnose, the activated intermediate compound involved in the biosynthesis of the poly L-rhamnan chain. However, no genes for the biosynthesis of UDP-Gal, which is required as a donor substrate for the {beta}-D-galactose linkage sugar of the S-layer glycan (Schäffer et al., 2002), are located in the slg gene cluster. This activated precursor is assumed to be synthesized by the products of housekeeping genes outside the slg gene cluster. In LPS biosynthesis, such an interplay of the products of glycan-specific genes and housekeeping genes is well known (Whitfield, 1995).

The slg gene cluster also encodes gene products which are obviously involved in the assembly of the S-layer glycan chain, whose elongation may eventually be terminated by 2-O-methylation at the non-reducing end, as is suggested by the presence of a putative methyltransferase. The occurrence of an ABC-2-type transporter system allows us to assume that the glycan chain is synthesized in a process comparable to the wzy-independent pathway of the LPS O-polysaccharide assembly route, which has been described for the biosynthesis of several homopolymers (Whitfield, 1995; Raetz & Whitfield, 2002). This assumption seems to be supported by the homology of ORFG102 to the O antigen ligase WaaL, which may act as an oligosaccharyl : : S-layer-polypeptide transferase in G. stearothermophilus NRS 2004/3a, catalysing the transfer of the S-layer glycan chain to distinct glycosylation sites of SgsE as the final step in the S-layer glycoprotein biosynthesis pathway.

It is interesting to note that the slg gene cluster is flanked by a fragmentary transposase at its downstream end. This, together with the observation that the G+C content of the cluster is lower than the mean G+C content of the organism, may indicate that G. stearothermophilus NRS 2004/3a has acquired its S-layer glycosylation capability by lateral gene transfer (Keenleyside & Whitfield, 1999; Ochman et al., 2000).

In our attempt to understand the regulatory mechanisms underlying S-layer protein glycosylation in Gram-positive bacteria in general, and G. stearothermophilus NRS 2004/3a in particular, we were interested in the effects of altered culture conditions on the S-layer. Recently, S-layer variation has been reported in the closely related organism, G. stearothermophilus ATCC 12980T, at a growth temperature of 67 °C instead of 55 °C (Egelseer et al., 2001). The de novo-generated S-layer protein, SbsD, was proposed to be glycosylated, thus representing a glycosylated S-layer variant of the parent strain. Based on this observation, we investigated whether growth temperature in general plays a role in S-layer protein glycosylation. For that purpose, the effect of different growth temperatures was investigated under standard conditions in a set of G. stearothermophilus strains, including strain ATCC 12980T. No difference in the protein banding pattern was observed for strains grown at the different temperatures. Thus, the spontaneous development of a glycosylated S-layer protein in strain ATCC 12980T on raising the growth temperature was not confirmed in our experiment.

Since it had also been reported that the change from the non-glycosylated G. stearothermophilus ATCC 12980T S-layer protein SbsC to the glycosylated variant SbsD of ATCC 12980/G+ is based on the de novo generation of sbsD with concomitant loss of sbsC (Egelseer et al., 2001), we also investigated the effect of the temperature increase on the S-layer gene at the molecular level. Genomic DNA of G. stearothermophilus strains NRS 2004/3a and ATCC 12980T, grown at 55 °C or 67 °C, was analysed by PCR, using primer combinations specific either for sbsC of ATCC 12980T or for sbsD/sgsE of ATCC 12980/G+ and NRS 2004/3a, respectively. The results of PCR analysis for the presence of specific S-layer gene sequences were in accordance with observations already made at the protein level. It was concluded that a temperature increase to 67 °C did not lead to a change of the S-layer genotype of either strain, i.e. no S-layer glycoprotein-carrying variant of G. stearothermophilus ATCC 12980T was formed. However, it should be mentioned that in G. stearothermophilus NRS 2004/3a, raising the growth temperature does have an effect on sgsE at the transcriptional level, with the transcript being several-fold more abundant at 67 °C as compared to 55 °C (Fig. 2A). This may indicate the involvement of a functional heat-or stress-response element in the promoter of this gene.

As L-rhamnose has been shown to be a frequent constituent of many S-layer glycoproteins (Bock et al., 1994; Graninger et al., 2002; Schäffer et al., 2002; Sleytr et al., 2002), and the rml-encoded biosynthesis of its nucleotide-activated form is well characterized in prokaryotes (Schnaitman & Klena, 1993; Whitfield & Valvano, 1993; Rocchetta et al., 1999; Schäffer & Messner, 2001), we screened a set of glycosylated, non-glycosylated, and S-layer-deficient G. stearothermophilus strains (Messner et al., 1984; Egelseer et al., 2001) for the presence of the dTDP-L-rhamnose operon. PCR analysis with an rmlAB-specific primer pair yielded amplification products for the strains ATCC 12980T, ATCC 12980/S and L32-65, in addition to the glycosylated strain NRS 2004/3a. Sequence analysis of the amplification products confirmed the presence of the rml operon. These strains are able to convert dTDP-D-glucose to dtdp-L-rhamnose, which demonstrates an active enzymic machinery. As expected, in the rml-deficient strains G. stearothermophilus E8-65, NRS 106/1b2 and S51-66, no dTDP-L-rhamnose synthesis was observed.

Since G. stearothermophilus strains possessing either a non-glycosylated S-layer protein (ATCC 12980T and L32-65) or no S-layer protein at all (ATCC 12980/S) were shown to have an active Rml pathway, we were interested to know if dTDP-L-rhamnose is incorporated in vivo into lipophilic rhamnose-containing glycoconjugates. This was because the presence of hydrophilic glycoconjugates could not be ruled out from the negative carbohydrate staining reaction of whole cells on SDS-PAGE . As no rhamnose could be detected in the CHCl3/CH3OH extract of any of these G. stearothermophilus strains, it is reasonable to assume that dTDP-L-rhamnose is synthesized in vivo to serve as a substrate exclusively for S-layer protein glycosylation, providing that the underlying biosynthesis machinery is functional.

Detailed studies are currently under way to elucidate the regulatory events governing the S-layer glycosylation process. The complete sequencing of the slg gene clusters of G. stearothermophilus ATCC 12980T and L32-65, which carry a non-glycosylated S-layer protein, but which have been shown in this study to synthesize dTDP-L-rhamnose in vitro, may unravel the mechanisms that prevent their S-layer proteins from becoming rhamnosylated. From the results presented here, it is possible that many members of the family Bacillaceae have the ability to produce dTDP-L-rhamnose, which suggests that they may be able to carry out S-layer glycosylation in their natural environment.


   ACKNOWLEDGEMENTS
 
We thank Sonja Zayni and Andrea Scheberl (Center for Ultrastructure Research) for their excellent technical assistance, and Harald Berger and Andreas Bernreiter (Center of Applied Genetics) for help with the RNA experiments. Partial 16S rRNA sequencing of G. stearothermophilus strains was performed by Dr Cathrin Spröer, Deutsche Sammlung für Mikroorganismen und Zellkulturen, Braunschweig, Germany. We thank Dr Joseph Lam, University of Guelph, Ontario, Canada, for providing the culture of Pseudomonas aeruginosa PAO1. This work was supported by the Austrian Science Fund, projects P14209-B07 and P15612-B07 (to P. M.) and Nestec Ltd, project RE-002804.05 (to C. S.) and Y114MOB (to J. S.).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Altman, E., Schäffer, C., Brisson, J.-R. & Messner, P. (1995). Characterization of the glycan structure of a major glycopeptide from the surface layer glycoprotein of Clostridium thermosaccharolyticum E207-71. Eur J Biochem 229, 308–315.[Abstract]

Becker, A., Rüberg, S., Küster, H., Roxlau, A. A., Keller, M., Ivashina, T., Cheng, H.-P., Walker, G. C. & Pühler, A. (1997). The 32-kilobase exp gene cluster of Rhizobium meliloti directing the biosynthesis of galactoglucan: genetic organization and properties of the encoded gene products. J Bacteriol 179, 1375–1384.[Abstract]

Benz, I. & Schmidt, M. A. (2001). Glycosylation with heptose residues mediated by the aah gene product is essential for adherence of the AIDA-I adhesin. Mol Microbiol 40, 1403–1413.[CrossRef][Medline]

Blatch, G. L. & Lässle, M. (1999). The tetratricopeptide repeat: a structural motif mediating protein-protein interactions. Bioessays 21, 932–939.[CrossRef][Medline]

Bock, K., Schuster-Kolbe, J., Altman, E., Allmaier, G., Stahl, B., Christian, R., Sleytr, U. B. & Messner, P. (1994). Primary structure of the O-glycosidically linked glycan chain of the crystalline surface layer glycoprotein of Thermoanaerobacter thermohydrosulfuricus L111-69. Galactosyl tyrosine as a novel linkage unit. J Biol Chem 269, 7137–7144.[Abstract/Free Full Text]

Bradford, M. M. (1976). A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72, 248–254.[CrossRef][Medline]

Bronner, D., Clarke, B. R. & Whitfield, C. (1994). Identification of an ATP-binding cassette transport system required for translocation of lipopolysaccharide O-antigen side chains across the cytoplasmic membrane of Klebsiella pneumoniae serotype O1. Mol Microbiol 14, 505–519.[Medline]

Comstock, L. E., Coyne, M. J., Tzianabos, A. O. & Kasper, D. L. (1999). Interstrain variation of the polysaccharide B biosynthesis locus of Bacteroides fragilis: characterization of the region from strain 638R. J Bacteriol 181, 6192–6196.[Abstract/Free Full Text]

Deckert, G., Warren, P., Gaasterland, T. & 12 other authors (1998). The complete genome of the hyperthermophilic bacterium Aquifex aeolicus. Nature 392, 353–358.[CrossRef][Medline]

Drummelsmith, J. & Whitfield, C. (1999). Gene products required for surface expression of the capsular form of the group 1 K antigen in Escherichia coli (O9a : K30). Mol Microbiol 31, 1321–1332.[CrossRef][Medline]

Egelseer, E. M., Idris, R., Jarosch, M., Danhorn, T., Sleytr, U. B. & Sára, M. (2000). ISBst12, a novel type of insertion-sequence element causing loss of S-layer-gene expression in Bacillus stearothermophilus ATCC 12980. Microbiology 146, 2175–2183.[Abstract/Free Full Text]

Egelseer, E. M., Danhorn, T., Pleschberger, M., Hotzy, C., Sleytr, U. B. & Sára, M. (2001). Characterization of an S-layer glycoprotein produced in the course of S-layer variation of Bacillus stearothermophilus ATCC 12980 and sequencing and cloning of the sbsD gene encoding the protein moiety. Arch Microbiol 177, 70–80.[CrossRef][Medline]

Fox, A., Stewart, G. C., Waller, L. N., Fox, K. F., Harley, W. M. & Price, R. L. (2003). Carbohydrates and glycoproteins of Bacillus anthracis and related bacilli: targets for biodetection. J Microbiol Methods 54, 143–152.[CrossRef][Medline]

Giraud, M.-F. & Naismith, J. H. (2000). The rhamnose pathway. Curr Opin Struct Biol 10, 687–696.[CrossRef][Medline]

Glaser, L. & Kornfeld, S. (1961). The enzymatic synthesis of thymidine-linked sugars. II. Thymidine diphosphate L-rhamnose. J Biol Chem 236, 1795–1799.[Medline]

Graninger, M., Kneidinger, B., Bruno, K., Scheberl, A. & Messner, P. (2002). Homologs of the Rml enzymes from Salmonella enterica are responsible for dTDP-{beta}-L-rhamnose biosynthesis in the Gram-positive thermophile Aneurinibacillus thermoaerophilus DSM 10155. Appl Environ Microbiol 68, 3708–3715.[Abstract/Free Full Text]

Guo, D., Bowden, M. G., Pershad, R. & Kaplan, H. B. (1996). The Myxococcus xanthus rfbABC operon encodes an ATP-binding cassette transporter homolog required for O-antigen biosynthesis and multicellular development. J Bacteriol 178, 1631–1639.[Abstract]

Heinrichs, D. E., Monteiro, M. A., Perry, M. B. & Whitfield, C. (1998). The assembly system for the lipopolysaccharide R2 core-type of Escherichia coli is a hybrid of those found in Escherichia coli K-12 and Salmonella enterica. Structure and function of the R2 WaaK and WaaL homologs. J Biol Chem 273, 8849–8859.[Abstract/Free Full Text]

Jarosch, M., Egelseer, E. M., Mattanovich, D., Sleytr, U. B. & Sára, M. (2000). S-layer gene sbsC of Bacillus stearothermophilus ATCC 12980: molecular characterization and heterologous expression in Escherichia coli. Microbiology 146, 273–281.[Abstract/Free Full Text]

Jiang, S.-M., Wang, L. & Reeves, P. R. (2001). Molecular characterization of Streptococcus pneumoniae type 4, 6B, 8, and 18C capsular polysaccharide gene clusters. Infect Immun 69, 1244–1255.[Abstract/Free Full Text]

Keenleyside, W. J. & Whitfield, C. (1999). Genetics and biosynthesis of lipopolysaccharide O-antigens. In Endotoxin in Health and Disease, pp. 331–358. Edited by H. Brade, S. M. Opal, S. N. Vogel & D. C. Morrison. New York & Basel: Marcel Dekker.

Kneidinger, B., Graninger, M. & Messner, P. (2001). Chromosome walking by cloning of distinct PCR fragments. Biotechniques 30, 248–249.[Medline]

Kornfeld, S. & Glaser, L. (1961). The enzymatic synthesis of thymidine-linked sugars. I. Thymidine diphosphate glucose. J Biol Chem 236, 1791–1794.[Medline]

Mahillon, J. & Chandler, M. (1998). Insertion sequences. Microbiol Mol Biol Rev 62, 725–744.[Abstract/Free Full Text]

Maier, R. M. & Soberón-Chávez, G. (2000). Pseudomonas aeruginosa rhamnolipids: biosynthesis and potential applications. Appl Microbiol Biotechnol 54, 625–633.[CrossRef][Medline]

Marumo, K., Lindqvist, L., Verma, N., Weintraub, A., Reeves, P. R. & Lindberg, A. A. (1992). Enzymatic synthesis and isolation of thymidine diphosphate-6-deoxy-D-xylo-4-hexulose and thymidine diphosphate-L-rhamnose. Production using cloned gene products and separation by HPLC. Eur J Biochem 204, 539–545.[Abstract]

Messner, P. & Schäffer, C. (2000). Surface layer glycoproteins of Bacteria and Archaea. In Glycomicrobiology, pp. 93–125. Edited by R. J. Doyle. New York: Kluwer Academic/Plenum Publishers.

Messner, P. & Schäffer, C. (2003). Prokaryotic glycoproteins. In Progress in the Chemistry of Organic Natural Products, vol. 85, pp. 51–124. Edited by W. Herz, H. Falk & G. W. Kirby. Wien: Springer.

Messner, P., Hollaus, F. & Sleytr, U. B. (1984). Paracrystalline cell wall surface layers of different Bacillus stearothermophilus strains. Int J Syst Bacteriol 34, 202–210.

Nölling, J., Breton, G., Omelchenko, M. V. & 16 other authors (2001). Genome sequence and comparative analysis of the solvent-producing bacterium Clostridium acetobutylicum. J Bacteriol 183, 4823–4838.[Abstract/Free Full Text]

Ochman, H., Lawrence, J. G. & Groisman, E. A. (2000). Lateral gene transfer and the nature of bacterial innovation. Nature 405, 299–304.[CrossRef][Medline]

Power, P. M. & Jennings, M. P. (2003). The genetics of glycosylation in Gram-negative bacteria. FEMS Microbiol Lett 218, 211–222.[CrossRef][Medline]

Raetz, C. R. H. & Whitfield, C. (2002). Lipopolysaccharide endotoxins. Annu Rev Biochem 71, 635–700.[CrossRef][Medline]

Rahim, R., Ochsner, U. A., Olvera, C., Graninger, M., Messner, P., Lam, J. S. & Soberón-Chávez, G. (2001). Cloning and functional characterization of the Pseudomonas aeruginosa rhlC gene that encodes rhamnosyltransferase 2, an enzyme responsible for di-rhamnolipid biosynthesis. Mol Microbiol 40, 708–718.[CrossRef][Medline]

Reeves, P. R., Hobbs, M., Valvano, M. A. & 8 other authors (1996). Bacterial polysaccharide synthesis and gene nomenclature. Trends Microbiol 4, 495–503.[CrossRef][Medline]

Rocchetta, H. L. & Lam, J. S. (1997). Identification and functional characterization of an ABC transport system involved in polysaccharide export of A-band lipopolysaccharide in Pseudomonas aeruginosa. J Bacteriol 179, 4713–4724.[Abstract]

Rocchetta, H. L., Burrows, L. L. & Lam, J. S. (1999). Genetics of O-antigen biosynthesis in Pseudomonas aeruginosa. Microbiol Mol Biol Rev 63, 523–553.[Abstract/Free Full Text]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Schäffer, C. & Messner, P. (2001). Glycobiology of surface layer proteins. Biochimie 83, 591–599.[CrossRef][Medline]

Schäffer, C., Wugeditsch, T., Kählig, H., Scheberl, A., Zayni, S. & Messner, P. (2002). The surface layer (S-layer) glycoprotein of Geobacillus stearothermophilus NRS 2004/3a. Analysis of its glycosylation. J Biol Chem 277, 6230–6239.[Abstract/Free Full Text]

Schnaitman, C. A. & Klena, J. D. (1993). Genetics of lipopolysaccharide biosynthesis in enteric bacteria. Microbiol Rev 57, 655–682.[Medline]

Sleytr, U. B. & Messner, P. (2003). Crystalline bacterial cell surface layers (S layers). In Desk Encyclopedia of Microbiology, pp. 284–291. Edited by M. Schaechter. San Diego: Elsevier Science.

Sleytr, U. B., Sára, M., Pum, D., Schuster, B., Messner, P. & Schäffer, C. (2002). Self-assembly protein systems: microbial S-layers. In Biopolymers, vol. 7, Polyamides and Complex Proteinaceous Matrices I, pp. 285–338. Edited by A. Steinbüchel & S. R. Fahnestock. Weinheim: Wiley-VCH.

Strauss, J., Horvath, H. K., Abdallah, B. M., Kindermann, J., Mach, R. L. & Kubicek, C. P. (1999). The function of CreA, the carbon catabolite repressor of Aspergillus nidulans, is regulated at the transcriptional and post-transcriptional level. Mol Microbiol 32, 169–178.[CrossRef][Medline]

Tsukioka, Y., Yamashita, Y., Oho, T., Nakano, Y. & Koga, T. (1997a). Biological function of the dTDP-rhamnose synthesis pathway in Streptococcus mutans. J Bacteriol 179, 1126–1134.[Abstract]

Tsukioka, Y., Yamashita, Y., Nakano, Y., Oho, T. & Koga, T. (1997b). Identification of a fourth gene involved in dTDP-rhamnose synthesis in Streptococcus mutans. J Bacteriol 179, 4411–4414.[Abstract]

Wang, L., Liu, D. & Reeves, P. R. (1996). C-terminal half of Salmonella enterica WbaP (RfbP) is the galactosyl-1-phosphate transferase domain catalyzing the first step of O-antigen synthesis. J Bacteriol 178, 2598–2604.[Abstract]

Whitfield, C. (1995). Biosynthesis of lipopolysaccharide O-antigens. Trends Microbiol 3, 178–185.[CrossRef][Medline]

Whitfield, C. & Valvano, M. A. (1993). Biosynthesis and expression of cell-surface polysaccharides in Gram-negative bacteria. Adv Microb Physiol 35, 135–246.[Medline]

Winzler, R. J. (1955). Determination of serum glycoproteins. Methods Biochem Anal 11, 279–311.

Xu, K., He, Z.-Q., Mao, Y.-M., Sheng, R.-Q. & Sheng, Z.-J. (1993). On two transposable elements from Bacillus stearothermophilus. Plasmid 29, 1–9.[CrossRef][Medline]

Xu, Y., Murray, B. E. & Weinstock, G. M. (1998). A cluster of genes involved in polysaccharide biosynthesis from Enterococcus faecalis OG1RF. Infect Immun 66, 4313–4323.[Abstract/Free Full Text]

Yamashita, Y., Tsukioka, Y., Tomihisa, K., Nakano, Y. & Koga, T. (1998). Genes involved in cell wall localization and side chain formation of rhamnose-glucose polysaccharide in Streptococcus mutans. J Bacteriol 180, 5803–5807.[Abstract/Free Full Text]

Zhang, L., al-Hendy, A., Toivanen, P. & Skurnik, M. (1993). Genetic organization and sequence of the rfb gene cluster of Yersinia enterocolitica serotype O : 3: similarities to the dTDP-L-rhamnose biosynthesis pathway of Salmonella and to the bacterial polysaccharide transport systems. Mol Microbiol 9, 309–321.[Medline]

Received 10 December 2003; accepted 22 December 2003.