OUP user menu

The Corynebacterium glutamicum gene pmt encoding a glycosyltransferase related to eukaryotic protein-O-mannosyltransferases is essential for glycosylation of the resuscitation promoting factor (Rpf2) and other secreted proteins

Martina Mahne, Andreas Tauch, Alfred Pühler, Jörn Kalinowski
DOI: http://dx.doi.org/10.1111/j.1574-6968.2006.00269.x 226-233 First published online: 1 June 2006

Abstract

Two-dimensional gel electrophoresis and immunoassays revealed several proteins of the secretory subproteome of Corynebacterium glutamicum to be glycosylated. By genome-wide searches for genes involved in glycosylation, the C. glutamicum gene cg1014 was found to exhibit significant similarity to eukaryotic protein-O-mannosyltransferases (PMTs) and to a recently identified orthologue of Mycobacterium tuberculosis, Rv1002c, which is responsible for protein-O-mannosylation. The putative membrane protein Cg1014 showed the same predicted transmembrane topology as Saccharomyces cerevisiae PMT1 and M. tuberculosis Rv1002c along with conserved amino acid residues responsible for catalytic activity. Deletion of the C. glutamicum pmt gene (cg1014) caused a complete loss of glycosylation of secreted proteins including the resuscitation promoting factor 2 (Rpf2), which is involved in intercellular communication and growth stimulation of C. glutamicum. Because the gene pmt as well as rpf genes are present in the genomes of all actinobacteria sequenced so far, this work provides new insights into bacterial protein glycosylation and new opportunities to elucidate the molecular mechanisms of Rpf activity in pathogenic growth and infection.

Keywords
  • protein glycosylation
  • glycosyltransferase
  • PMT family
  • Corynebacterium glutamicum
  • resuscitation promoting factor

Introduction

For a long time, protein glycosylation has been considered to be a specific feature of eukaryotes. In fact, protein glycosylation is also widespread throughout the domains of Archaea and Bacteria and meanwhile documented by a number of studies, for the most part concentrating on bacterial pathogens. Protein-O-glycosylation has been studied extensively in the yeast Saccharomyces cerevisiae (Strahl-Bolsinger , 1993; Strahl-Bolsinger & Scheinost, 1999), where the initial sugar transfer to serine and threonine residues is catalyzed by protein-O-mannosyltransferases of the protein-O-mannosyltransferase (PMT) family. PMTs are highly conserved among eukaryotes from yeast to human, but experimental data about prokaryotic homologues are very scarce. Recently, a mycobacterial orthologue of S. cerevisiae PMT1, Rv1002c, was identified and shown to be responsible for protein-O-mannosylation in Mycobacterium tuberculosis (VanderVen , 2005). Owing to the fact that the pathogenic M. tuberculosis is taxonomically related to the nonpathogenic Corynebacterium glutamicum, it is of special interest to investigate protein glycosylation in this potential model organism. To date, only one glycosylated protein of C. glutamicum ATCC 13032 is known (Hartmann , 2004). The resuscitation promoting factor (Rpf2) has been demonstrated to exist in three glycosylated isoforms which contain at least mannose and galactose. The functional role of the Rpf2 glycosylation is still unclear. Rpf2 was shown to be involved in growth stimulation of long-stored C. glutamicum cells. Furthermore, Rpf proteins of M. tuberculosis are highly immunogenic (Yeremeev , 2003).

The aim of this work was the identification and characterization of C. glutamicum genes encoding key enzymes involved in protein glycosylation in general and of Rpf2 glycosylation in particular.

Materials and methods

Bacterial cultivation and cloning techniques

Escherichia coli DH5αMCR cells were used for routine cloning experiments and grown at 37°C in liquid or on solid Luria–Bertani complex medium (Sambrook , 1989). Corynebacterium glutamicum strains were cultivated in minimal medium MM1 (MMYE without yeast extract, Katsumata , 1984) or on Luria–Bertani agar plates at 30°C. Kanamycin was used (E. coli: 50 μg mL−1, C. glutamicum: 25 μg mL−1) for selection of recombinant clones. For protein analysis, C. glutamicum strains were grown aerobically in MM1 at 30°C with horizontal shaking at 150 r.p.m. and harvested during exponential growth phase. PCR techniques and isolation, manipulation and transformation of DNA were performed following standard procedures according to Horton (1995).

Generation of C. glutamicum mutant strain MM113 lacking cg1014

The cg1014 gene was deleted by gene SOEing (Horton, 1995). For this purpose, up- and downstream regions of cg1014 spanning c. 700 bp each were amplified by two separate PCR reactions using the oligonucleotides 1014d1 to 1014d4 (Table 1). The resulting DNA fragments were fused by subsequent PCR using the products as templates along with the 5′-primer corresponding to the upstream fragment (1014d1) and the 3′-primer of the downstream fragment of cg1014 (1014d4). The purified fusion product was ligated with pK18mobsacB (Schäfer , 1994) using EcoRI and BamHI restriction sites and subsequently transformed into E. coli DH5αMCR using the CaCl2 method (Sambrook , 1989). The isolated and purified plasmid designated pMM113 was used to perform an allelic exchange in the C. glutamicum chromosome (Kirchner & Tauch, 2003). Corynebacterium glutamicum ATCC 13032 was transformed with plasmid DNA via electroporation (Tauch , 2002). Selection for gene replacement was carried out using the sacB system (Schäfer , 1994). Deletion of cg1014 was confirmed by PCR assays.

View this table:
Table 1

Strains, plasmids and oligonucleotides used in this study

NameDNA sequence or genetic characteristicsSource or reference
Oligonucleotides
1014d15′-GATCTAGAATTCGGTGAGGAGATTCGAAGAAA-3′This work
1014d25′-AGGTTAGCGCCAGCTTGGGACTCACTCGGTTGATTGTAGA-3′This work
1014d35′-TCCCAAGCTGGCGCTAACCT-3′This work
1014d45′-GCTTGTTCGGATCCTTGATG-3′This work
5′-1014ex5′-GATCTAGAATTCGGTGAGGAGATTCGAAGAAA-3′This work
3′-1014ex5′-GCTTGTTCGGATCCTTGATG-3′This work
Plasmids
pK18mobsacB E. coli cloning vector carrying the sacB selection system Schäfer (1994)
pBHK19MECA E. coliC. glutamicum shuttle vector Kirchner & Tauch (2003)
pMM113pK18mobsacB carrying a 1.10 kb fragment with deleted pmt geneThis work
pMM114pBHK19MECA carrying pmt with its putative native promoter siteThis work
pMH33 E. coliC. glutamicum shuttle expression vector pZ8-1 carrying rpf2 with 3′-fused His tag downstream of the Ptaq promoter Hartmann (2004)
C. glutamicum strains
ATCC 13032Wild-type strainATCC
C. glutamicum MM113ATCC 13032ΔpmtThis work
C. glutamicum MM114ATCC 13032Δpmt carrying pMM114 for expression of pmtThis work
C. glutamicum MM115ATCC 13032 carrying pMH33 for overexpression of His-tagged rpf2This work
C. glutamicum MM116ATCC 13032Δpmt carrying pMH33 for overexpression of His-tagged rpf2This work
  • * Restriction sites within oligonucleotides are shown in italics.

  • ATCC, American Type Culture Collection.

Construction of C. glutamicum mutant strain MM114

A DNA fragment spanning cg1014 and its putative promoter region was amplified by PCR using Pfx DNA polymerase along with the primer pair 5′-1014ex and 3′-1014ex (Table 1). The PCR product was cloned into pBHK19MECA (Kirchner & Tauch, 2003) by blunt end ligation at the EcoRV restriction site, followed by transformation of E. coli DH5α MCR with the resulting plasmid pMM114. Corynebacterium glutamicum MM113 was transformed with purified plasmid pMM114 by electroporation.

Extraction of secreted C. glutamicum proteins

To analyze secreted proteins, C. glutamicum strains were grown in MM1 at 30°C in Erlenmeyer flasks with horizontal shaking at 150 r.p.m. Corynebacterium glutamicum cells from 800 mL culture medium were harvested during exponential growth and pelleted at 3800 g and 4°C for 15 min. Supernatants were centrifuged for 45 min at 3800 g and 4°C to remove residual particles. After addition of one protease inhibitor tablet (Complete Mini, EDTA-free, Roche Molecular Biochemicals) per 400 mL, cell-free culture supernatants were frozen at −80°C and lyophilized to complete dryness. The lyophilisate corresponding to 400 mL of culture volume was rehydrated in 25 mL H2O, supplemented with 4 mL 0.5 M Tris-HCl (pH 6.8) and 200 μL 1 M dithiothreitol, and incubated at room temperature for 30 min. Protein was recovered by phenolic extraction. Then, 300 μL 8 M ammonium acetate and 200 μL 1 M dithiothreitol were added to the organic phase, followed by incubation at room temperature for 30 min. Proteins were precipitated with methanol (−20°C) and subsequently washed with −20°C cold ethanol (70%) and acetone. Pellets were dried at room temperature and rehydrated in 400 μL 0.1 M Tris-HCl, 1% (w/v) sodium dodecyl sulfate (SDS).

Affinity purification of Rpf2

His-tagged Rpf2 protein was isolated from cell culture supernatants by immobilized-metal affinity purification using Protino Ni prepacked columns (Macherey-Nagel). Eluates were concentrated using Amicon Microcon YM-30 centrifugal devices with a nominal molecular weight limit of 30 kDa (Millipore). Protein-containing fractions were analyzed by one-dimensional (1D) SDS-polyacrylamide gel electrophoresis (PAGE).

SDS-PAGE and protein identification

Protein extracts were separated by 1D and two-dimensional (2D) SDS-PAGE using experimental protocols from Amersham Pharmacia Biosciences. ReadyStrip IPG Strips (pH 3–6, 24 cm) and ampholytes (Bio-Lyte 3–10 Buffer) were purchased from Bio-Rad. Isoelectric focusing was performed on an IPGPhor (Amersham Pharmacia) according to the following protocol: rehydration at 20°C and 0 V for 1 h, focusing at 20°C, max. 50 μA per strip for 12 h at 30 V, 2 h at 60 V, 1 h at 500 V, 1 h at 1000 V, and 6–8 h at 8000 V. SDS-PAGE was performed at 25°C for 30 min at 3 W per gel and 4–6 h at 20 W per gel. Proteins in 2D gels were fixed and stained overnight in Coomassie staining solution [42.5% (v/v) ethanol, 5% (v/v) methanol, 10% (v/v) acetic acid, 0.2% (w/v) Coomassie Brilliant Blue G250, 0.05% (w/v) Coomassie Brilliant Blue R250] and destained with H2O. Protein spots cut off from 2D gels were treated with modified trypsin (Promega) and identified by peptide-mass fingerprinting using the MASCOT software (Matrix Science). Mass spectra were obtained on an Ultraflex MALDI-TOF mass spectrometer (Bruker), equipped with a nitrogen laser, in a positive ion reflector mode using α-cyano-4-hydroxy-cinnamic acid as matrix for ionization (Hermann , 2000).

Western blotting techniques

Western blotting of proteins to polyvinylidene fluoride transfer membranes (Millipore) was carried out by semi-dry electrotransfer following standard procedures (Towbin , 1979).

Glycoproteins were detected using the DIG glycan detection kit (Roche Molecular Biochemicals) according to the manufacturer's instruction. Digoxigenin-coupled proteins were detected in an enzyme immunoassay with a commercial digoxigenin specific antibody conjugated with alkaline phosphatase which converts 4-nitroblue tetrazolium chloride/5-bromo-4-chloro-3-indolyl-phosphate in 100 mM Tris-HCl, 50 mM MgCl2, 100 mM NaCl, pH 9.5.

Results and discussion

Several glycosylated proteins in the secretory subproteome of C. glutamicum ATCC 13032 were detected by 2D gel electrophoresis

To identify glycosylated proteins of Corynebacterium glutamicum, two standard methods of protein analysis were combined. Because it was expected that proteins are glycosylated on the extracytoplasmic side of the cell membrane, we analyzed the extracellular subproteome by 2D PAGE and subsequent Western blotting, inspecting different pI gradients between pH 3 and 10. Figure 1 shows a 2D gel of secreted proteins of C. glutamicum in the pI range of pH 3.5 to 5.5, and the corresponding Western blot with digoxigenin-labelled glycoproteins. Selected proteins were identified by MALDI-TOF-MS and peptide-mass fingerprinting. By this approach, 88 spots were identified, of which a total number of 78 spots represent proteins containing a signal peptide required for secretion to the extracellular space (data not shown).

Figure 1

Two-dimensional polyacrylamide gel electrophoresis (2D PAGE) and Western blot of secreted proteins of Corynebacterium glutamicum ATCC 13032. Secreted proteins purified from the culture supernatant of C. glutamicum ATCC 13032 were separated by 2D PAGE and either visualized by staining with Coomassie Blue (a) or transferred to polyvinylidene fluoride membranes and detected by Western blotting (b). Glycoproteins were detected by covalent modification of the glycan moieties with digoxigenin and a digoxigenin-specific antibody coupled with alkaline phosphatase. The prestained protein standard All Blue (Bio-Rad) was used as molecular weight marker. n.i., not identified.

Out of the identified proteins, four turned out to be glycosylated (Fig. 1). The first one is the cell surface-associated Rpf2, occurring in three isoforms of different molecular masses in 2D PAGE, which were identified with MOWSE scores of 55 up to 64 and a peptide coverage of 16%. Migrating at c. 41, 46 and 49 kDa, these protein species are larger than the predicted processed protein which has a theoretical molecular mass of 35 kDa considering cleavage of the 4 kDa leader peptide. The detection of Rpf2 in these immunoassays confirmed former results, because Rpf2 was shown to be glycosylated before (Hartmann , 2004). The second glycoprotein, termed LppS, is a putative secreted lipoprotein of the ErfK/YbiS/YcfS/YnhG family (MOWSE score: 107, peptide coverage: 40%), encoded by cg2720. It migrates at c. 49 kDa in 2D electrophoresis, whereas the calculated molecular mass is 40 kDa. The third one is annotated as an unclassified secreted protein and encoded by cg1859 (MOWSE score: 48, peptide coverage: 50%). The Cg1859 protein migrates at c. 22 kDa in 2D PAGE. The fourth detected glycoprotein with an apparent molecular mass of 40 kDa could not be identified with significant MOWSE scores within the Mascot database search, probably due to its low abundance in the 2D gel. Therefore, it was not assigned to a protein predicted from the genome sequence. Other proteins may not appear as glycosylated due to their low abundance in 2D PAGE. Generally, only few bacterial proteins have been described to be glycosylated up to now, and most of them were found during examination of pathogenic species belonging to the genera Mycobacterium or Campylobacter (Benz & Schmidt, 2002; Schmidt , 2003).

The C. glutamicum gene cg1014 is conserved among actinobacteria and encodes a membrane protein homologous to eukaryotic protein-O-mannosyltransferases

To investigate genes responsible for protein glycosylation in C. glutamicum, the whole genome sequence (Kalinowski , 2003) was examined for genes with putative glycosyltransferase function. Focusing on enzymes that catalyze key steps in protein glycosylation, the predicted coding sequence cg1014 turned out to be an interesting target. Its deduced protein product revealed high similarity (40% identical amino acids) to the newly described protein-O-mannosyltransferase Rv1002c of Mycobacterium tuberculosis (VanderVen , 2005) and global similarity (15–18% identical amino acids) to eukaryotic protein-mannosyltransferases of Saccharomyces and Candida species, as well as of higher eukaryotes, for instance Mus musculus, Drosophila melanogaster and Homo sapiens. Therefore, we analyzed cg1014 in regard to its biological function using bioinformatic, genetic and proteomic methods.

The gene cg1014 consists of 1563 nucleotides encoding a protein of 520 amino acids with a theoretical molecular mass of 58.7 kDa and a calculated pI of 9.9. It is predicted to contain 11 transmembrane helices consisting of 18 or 19 amino acids (Tusnády & Simon, 2001). Interestingly, the hydropathy profile (Kyte & Doolittle, 1982) of Cg1014 is closely related to those of yeast protein-O-mannosyltransferases, as displayed for Saccharomyces cerevisiae PMT1 (Fig. 2a), a feature that was also described for M. tuberculosis Rv1002c (VanderVen , 2005). However, the yeast PMTs contain a long central hydrophilic loop after the seventh helix motif, which is missing in the protein sequence encoded by cg1014 (Fig. 2a). Nevertheless, these results strongly suggest that Cg1014 fulfils the same catalytic function as eukaryotic protein-O-mannosyltransferases and Rv1002c. Homologues of cg1014 were also found in other Actinomycetales, including Corynebacterium efficiens and Corynebacterium diphtheriae as well as Mycobacterium leprae and Mycobacterium bovis (data not shown). Functionally important amino acid residues that are conserved between members of the PMT family and actinomycetes were searched for by multiple local alignments performed with the software program dialign (Morgenstern, 1999) (Fig. 2b). Besides the Actinomycetales proteins Cg1014, CE0964, DIP0868, Mb1092c, Ml0192 and Rv1002c, Streptomyces coelicolor SCO3154, Streptomyces avermitilis SAV3592, and Propionibacterium acnes PPA0523, the analysis included the PMT family members of S. cerevisiae, PMT1 and PMT2, Candida albicans Pmt6, D. melanogaster POMT1 and POMT2, M. musculus POMT1 and POMT2, and Homo sapiens POMT1 and POMT2. The Cg1014 protein contains only two hydrophilic loops which are large enough to assemble a functionally active catalytic site. These are found between the first and the second helix motif, and between the seventh and eighth predicted transmembrane domain. Six residues of Cg1014 (YP_225182) are conserved among all sequences: Arg-49, Asp-65, Glu-66, Pro-100, Pro-101, and Arg-122. The amino acid residues Arg-64, Glu-78 and Arg-138 of S. cerevisiae PMT1 as well as Asp-55 and Glu-56 of M. tuberculosis Rv1002c have been shown to be essential for catalytic activity (Girrbach , 2000; VanderVen , 2005). Therefore, Cg1014 possesses essential features of protein-O-mannosyltransferases, indicating that Cg1014 is the protein-O-mannosyltransferase of C. glutamicum ATCC 13032. Accordingly, the gene cg1014 was designated pmt. These results imply that the Actinomycetales homologues of Pmt and Rv1002c encode a bacterial subgroup of the protein-mannosyltransferase family, which is displayed by the phylogenetic tree deduced from sequence alignments (Fig. 2c). Pmt of C. glutamicum is the first representative from nonpathogenic Corynebacterineae.

Figure 2

Features of the Corynebacterium glutamicum protein PMT. (a) Hydropathy profiles of the amino acid sequences of PMT and Saccharomyces cerevisiae PMT1 were calculated according to the Kyte–Doolittle method (Kyte & Doolittle, 1982) with a window size of 19 using the Protein Hydrophilicity/Hydrophobicity Search and Comparison Server and the FASTA programs (http://fasta.bioch.virginia.edu/o:fasta/home.html; http://bip.weizmann.ac.il/index.html). Hydrophobic domains are characterized by values greater than 0; hydrophilic regions by values smaller than 0. Both sequences contain eleven putative transmembrane helices (filled in grey). Predicted transmembrane helices are indicated with roman numbers (I–XI). The hydrophilic loop between helices I and II, which is supposed to contain the catalytic domain of PMT, is boxed. (b) The multiple sequence alignment was calculated with DIALGN 2.2.1 (Morgenstern, 1999). An excerpt of this alignment which displays the hydrophilic loop between helices I and II is shown. Fully conserved amino acid residues are marked with asterisks. Dots indicate strong (:) and weak (•) functional conservation of amino acids. Residues which are putatively involved in catalysis and complex formation are shaded grey. Cglu1014, C. glutamicum putative dolichyl-phosphate-mannose: protein-mannosyltransferase (YP_225182); Cdip0868, C. diphtheriae putative membrane protein (NP_939232); Ceff0964, C. efficiens hypothetical protein (NP_737574); Mlep0192, Mycobacterium leprae membrane protein (H86932); Mbov1029, M. bovis conserved membrane protein (NP_854686); Mtub 1002c, M. tuberculosis hypothetical protein (O05586); CalbPMT6, Candida albicans protein-mannosyltransferase 6 (AAF16867); ScerPMT1 and PMT2, S. cerevisiae protein-O-mannosyltransferase 1 (NP_010188) and 2 (NP_009379); MmusPOMT1 and POMT2, Mus musculus protein-O-mannosyltransferase 1 (NP_660127) and 2 (NP_700464); HsapPOMT1 and POMT2, Homo sapiens protein-O-mannosyltransferase 1 (NP_009102) and 2 (NP_037514); DmelPOMT1 and POMT2, Drosophila melanogaster protein-O-mannosyltransferase 1 (Q9VTK2) and 2 (Q9W5D4); Scoe3154, Streptomyces coelicolor putative integral membrane protein (NP_627370); Save3592, S. avermitilis putative integral membrane protein (NP_824769); Pacn0523, Propionibacterium acnes putative dolichyl-phosphate-mannose: protein-mannosyltransferase (YP_055234). (c) The phylogenetic tree was deduced from a global multiple sequence alignment of eukaryotic and Actinomycetales homologues of C. glutamicum PMT.

The C. glutamicum pmt deletion mutant is unable to glycosylate secreted proteins including Rpf2

To examine the effect of the pmt deletion, the C. glutamicum mutant MM113 and the genetically complemented strain C. glutamicum MM114 were analyzed by the same proteomic approach as shown for the wild-type strain. Western blotting of secreted proteins isolated from C. glutamicum MM113 culture supernatants showed that a complete loss of protein glycosylation occurred (Fig. 3). After transformation of the deletion mutant with plasmid pMM114, protein glycosylation was restored and glycoproteins were detectable in immunoassays (Fig. 3), displaying one additional glycoprotein spot compared with the wild-type strain. However, this spot could not be assigned to a predicted protein of the C. glutamicum genome sequence. Nevertheless, these findings clearly demonstrated that Pmt is essential for protein glycosylation in C. glutamicum. Owing to the fact that the mycobacterial orthologue Rv1002c functions as protein-O-mannosyltransferase (VanderVen , 2005), it can be concluded that Pmt also catalyzes the initial transfer of a mannosyl residue to the polypeptide, in C. glutamicum.

Figure 3

2D PAGE and Western blot of secreted proteins of Corynebacterium glutamicum ATCC 13032 (a), MM113 (b), and MM114 (c). Secreted proteins purified from the culture supernatants of all three strains were separated by 2D PAGE and visualized by Coomassie Blue staining (upper row). Glycoproteins on polyvinylidene fluoride membranes were detected after labelling with digoxigenin using a digoxigenin-specific antibody coupled with alkaline phosphatase (bottom row). n.i., not identified.

It was shown in an earlier study that the Rpf2 glycoprotein of C. glutamicum ATCC 13032 is involved in growth stimulation and intercellular communication (Hartmann , 2004). Therefore, we focused on Rpf2 as test system to study protein glycosylation in more detail. To elucidate whether glycosylation is responsible for the appearance of several protein isoforms with different molecular mass, Rpf2 was purified from cell culture supernatants of C. glutamicum. For this purpose, the plasmid pMH33, which provides the enhanced expression of Rpf2 along with a His-tag fusion for affinity purification (Hartmann , 2004), was used. Both, C. glutamicum ATCC 13032 and the pmt deletion mutant MM113 were transformed with pMH33, resulting in the corresponding strains C. glutamicum MM115 and MM116 (Table 1). After affinity purification, Rpf2 protein samples from C. glutamicum MM115 and MM116 were analyzed by SDS-PAGE and subsequent Western blotting. Here, we detected three glycoprotein isoforms with apparent molecular masses of 35, 40 and 42 kDa (Fig. 4). All three protein isoforms turned out to be nonglycosylated if pmt was deleted. Consequently, glycosylation of Rpf2 is not the reason for the occurrence of isoforms with different molecular masses, as these are still detectable and have the same apparent molecular masses as glycosylated Rpf2, even though glycosylation is abolished. Furthermore, only mannose and galactose were detected in the glycan moiety of Rpf2 (Hartmann , 2004). These findings indicate that it is a rather small glycan that does not contribute to the remarkable mass shift in SDS gel electrophoresis.

Figure 4

Coomassie-stained one-dimensional (1D) SDS-PAGE gel and Western blot of affinity-purified Rpf2 protein. His-tagged Rpf2 proteins from Corynebacterium glutamicum MM115 (pmt) and MM116 (Δpmt), which both harbour plasmid pMH33 (Table 1) were isolated by affinity purification and subjected to (a) SDS-PAGE and (b) Western blotting. Rpf2 isoforms are marked with arrows. Glycosylated Rpf on the polyvinylidene fluoride membrane was detected after labelling with digoxigenin using a commercial digoxigenin-specific antibody conjugated with alkaline phosphatase. A protein marker containing α2-macroglobulin and avidin served as positive control. The prestained protein standard All Blue (Bio-Rad) was used as molecular weight marker.

Taken together, this work provides new insights into posttranslational modification of C. glutamicum proteins and bacterial protein glycosylation in general. Protein glycosylation in C. glutamicum was studied on genetic and proteomic levels, using Rpf2 as target system for identification and characterization of genes involved in this process. This work is also another step towards elucidation of the biological functions of Rpf2 glycosylation in vivo. Rpf2 was shown to be involved in bacterial growth stimulation and intercellular communication (Mukamolova , 1998, 2002; Hartmann , 2004). Vaccination studies furthermore showed that Rpf of Micrococcus luteus induces a protective immune response in mice vaccinated with this nonmycobacterial Rpf protein (Yeremeev , 2003). For the reason that pmt homologues are found in all actinobacteria including M. tuberculosis, and because of the high interest in growth stimulating capability and immunogenicity of Rpf proteins in mycobacteria, our work could provide new opportunities to clarify the molecular mechanisms of Rpf activity in mycobacterial growth and infection.

Acknowledgements

We thank Carola Eck for assistance in MALDI-TOF mass spectometry. This work was supported by the German Federal Ministry of Education and Research (grant BMBF 0312843E).

References

View Abstract