INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Mycoplasmas are the smallest living prokaryotes. Because of their small genome, they do not have the capacity to synthesize molecules such as cholesterol, fatty acids, some amino acids, purines, and pyrimidines. The absence of such synthesis pathways determines the parasitic behavior of mycoplasmas, which is characterized by the uptake of essential products from the host cell (4, 34).

The transport of proteins through the cell membrane is mediated by the secretory pathway (46). To date, a small number of systems for the active transport of sugars, amino acids, K⁺, and Na⁺ through the membrane in mycoplasmas have been studied in detail, but the molecular basis of such transport to a large extent remains to be determined (33). In both prokaryotes and eukaryotes, the superfamily of ABC (ATP-binding cassette) transporters has been found to be responsible for the export of proteases, hemolysin, and polysaccharides and the import of sugars, inorganic ions, and di-, tri-, and oligopeptides (19, 20, 26, 32, 45).

For Mycoplasma hyorhinis, a binding-protein component of a putative binding-protein-dependent transport system belonging to the ABC superfamily (9, 13) was characterized as a 37-kDa protein that was found to influence tumor cell invasiveness in vitro (42). Sequencing of the genomes of Mycoplasma pneumoniae (24) and Mycoplasma genitalium (12) indicated the presence in mycoplasmas of a further ABC transporter that has been proposed to be responsible for the import of oligopeptides. This oligopeptide permease (Opp) system has been extensively characterized for Salmonella typhimurium (22, 23, 36, 44) and Bacillus subtilis (30, 37).

For Mycoplasma hominis, a facultative urogenital tract pathogen (31), a number of membrane proteins have been described (7, 16, 35); however, little is known about their function. Two proteins (P50 and P100) were identified to be associated with mycoplasmal adherence to host cells (15). In this paper, we show the peptide-binding character of the P100 lipoprotein and demonstrate that the P100 gene is organized within an operon structure containing genes putatively encoding the core domains of an ABC transporter. These findings suggest that P100 functions as the substrate-binding domain OppA of an oligopeptide permease of M. hominis.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Mycoplasma culture conditions. M. hominis FBG was grown in PPLO broth base medium containing arginine as described previously (3, 11). Stocks of isolate FBG were prepared from a mid-logarithmic-phase broth culture and stored at 70°C.

Bacterial strains and plasmids. pT7T3-19U (Pharmacia Biotech, Freiburg, Germany) was used as a cloning vector for the construction of recombinant plasmids with different genomic opp fragments. The plasmids were propagated in Escherichia coli DH5F' (Gibco BRL, Gaithersburg, Md.).

Oligonucleotides. Oligonucleotides were synthesized with a solid carrier on an Applied Biosystems 381A machine by the phosphoamidite method (5). The antisense oligonucleotide P100-1 [5'-CAT(A/G)TA(T/A/C)A(A/G)TTG(A/T)GG(A/T)GC(G/A)TTTA-3'] and the sense oligonucleotide P100-2 [5'-GTTGG(A/ T ) T T T (A/C)GA T A(C/ T ) T T AAA T GC(A/ T )CC(A/ T )CAA T T ATATATG - 3'] were used as probes in Southern blot analysis. The primers P100-pe1 (5'-GGA CCC AAG ACC AAG TAA TCA TAA-3'), P100-pe2 (5'-TAC ATG AAG CGG CTA CTA ATC-3'), and oppB-pe1 (5'-ACG CTA TTC TTT GTA ATA TAT ATT TTG TCA-3') were used in primer extension analysis.

DNA and RNA manipulations. Genomic mycoplasmal DNA was isolated by use of a QIAamp Tissue Kit (Qiagen, Hilden, Germany) as described previously (17). All further DNA techniques were carried out by standard procedures (38, 39, 43) or by following the instructions of the commercial suppliers of materials. RNA from exponential-phase cultures of M. hominis FBG was prepared by the single-step method of Chomczynski and Sacchi (6).

Reverse transcription-PCR. Before M. hominis RNA was used as a template in reverse transcription-PCR, contaminating traces of DNA were digested with 0.6 U of DNase I (Roche Molecular Biochemicals) per µg of nucleic acid in 50 mM Tris-HCl (pH 8.3)-75 mM KCl-3 mM MgCl₂ for 15 min at 37°C. RNA was reverse transcribed by use of a 1st-Strand cDNA Synthesis Kit (Clontech Laboratories, Inc.). With this cDNA as a template, gene-overlapping opp regions were amplified by standard PCR conditions (initial cycle of 2 min at 94°C and then 1 min at 94°C, 1.5 min at T_m  2°C, and 1 min at 72°C, for a total of 34 cycles) with PCR reagents from Perkin-Elmer Cetus (Überlingen, Germany) or AGS (Heidelberg, Germany).

Primer extension analysis. Primers P100-pe1, P100-pe2, and oppB-pe1 were labeled with [³²P]dATP, annealed to 10 µg of M. hominis RNA, and extended with avian myeloblastoma virus reverse transcriptase as described by Ayer and Dynan (2). With the same primers, genomic DNA sequencing reactions were performed, and samples were separated on an 8% polyacrylamide-8 M urea gel next to the primer extension products.

Sequence analysis. The analysis of DNA and protein sequences and the design of oligonucleotides were facilitated by use of the computer program Lasergene (DNASTAR Inc., Madison, Wis.). Distant relationships to other proteins were determined by use of Psi-blast (34a). The protein sequences of OppABCDF of other species were drawn from the Swissprot database of EMBL: M. genitalium, P47563, P47323, P47324, P47325, and P47326; M. pneumoniae, P75327, P75554, P75553, P75552, and P75551; B. subtilis, P24141, P24138, P24139, P24136, and P24137; S. typhimurium, P06202, P08006, P08066, P04285, and P08007; and Haemophilus influenzae, P44572, P45054, P45053, P45052, and P45051.

Fluorescence spectroscopy. The P100 protein was solubilized from the membrane with 0.5% n-dodecyl--D-maltoside (ICN Biomedical, Eschwege, Germany) in phosphate-buffered saline (pH 7.3) and purified by affinity chromatography by a procedure described earlier (15). Fluorescence spectra were determined with a Shimadzu RF-5001PC spectrofluorometer. An excitation slit of 3 nm and an emission slit of 5 nm were used for the emission spectra. All measurements were obtained with 0.15 µM P100 (0.15 mg/ml) after dialysis against 10 mM imidazole acetate (pH 7.3)-100 mM NaCl-0.03% n-dodecyl--D-maltoside (for analyzing Lys₃ or Ala₃) or 10 mM sodium phosphate (pH 7.0)-0.03% n-dodecyl--D-maltoside (for analyzing Ala₅ or Ala₃) in the presence or absence of the respective peptide at a final concentration of 100 µM. With an excitation wavelength of 290 nm, fluorescence emission was monitored from 300 to 430 nm at room temperature.

Nucleotide sequence accession number. The P100- and OppBCDF-encoding sequence of M. hominis FBG has been assigned EMBL accession no. X99740.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

opp gene sequences. A 100-kDa membrane protein of M. hominis (P100) was shown to be species specific and involved in cytoadherence of the organism (15). To characterize the P100-encoding gene, the P100 protein of M. hominis FBG was purified by affinity chromatography, and the N-terminal amino acids of a V8 protease-generated P100 peptide were determined by an Edman reaction (10). From this peptide, P100-1 (V-G-F-R-Y-L-N-A-P-Q-L-Y-M), the oligonucleotides P100-1 and P100-2 were deduced, characterized by low degeneration and a preference for AT-enriched codon usage (see Materials and Methods), and used as probes in Southern blot analysis. Both probes detected a 3-kb HindIII DNA fragment of M. hominis FBG. Cloning and sequencing of this fragment revealed that it encodes the C-terminal part of P100, including the 13 amino acids (aa) of peptide P100-1. This 3-kb fragment and overlapping DNA fragments were hybridized to restricted genomic DNA, followed by cloning and sequencing of the detected DNA fragments. Finally, a 10.4-kb genomic region of M. hominis FBG was sequenced (Fig. 1); this region contains the entire P100 gene sequence and a putative ribosomal binding site (AAGGA) 10 bp upstream of the translational start codon ATG. The deduced polypeptide chain of P100 starts with an N-terminal signal sequence of 28 aa which meets all the requirements of a transmembrane helix (from aa 7 to aa 26) and of a signal peptidase II recognition site [(4)-VAASC-(+1)] with a lipoprotein attachment site at position 28 (in bold) (LVAASCKIDPA). Thus, P100 seems to be a cysteine-anchored lipoprotein of M. hominis which is expressed as a precursor polypeptide.

View larger version (37K): [in this window] [in a new window]   FIG. 1.   Physical map of the opp operon in M. hominis FBG and its similarity to those of other permeases. Genomic DNA (10.4 kb) of M. hominis FBG is represented schematically, with characteristic restriction sites shown above the line (E, EcoRI; H, HindIII; R, RsaI) and the positions of the ORF flanking nucleotides shown below the line. The deduced polypeptide chains of P100, OppBCD, and OppF are shown below the ORFs, with an indication of their lengths. Their similarities to the respective hydrophobic integral membrane proteins OppB and OppC and the ATP-binding proteins OppD and OppF of other, partly speculative (asterisk), permeases were determined with respect to identical amino acids by use of the MegAlign module of the computer program Lasergene. The substrate-binding protein OppA is shown in homology to P100. Since the OppA-encoding genes of M. pneumoniae and M. genitalium are not polycistronically organized with the other opp genes but are located in distant regions in the genomes, they are not depicted on line.

Five genes and one mRNA. Genes encoding the domains of an active transport system are polycistronically transcribed in B. subtilis (30) and S. typhimurium (1, 23, 36). As described above, the three regions between the P100 gene and oppD contained only a few nucleotides, and the oppD and oppF genes even overlapped by 4 bp. Purine-enriched regions, which are characteristic for ribosomal binding sites, were detected 10 bp upstream of the P100 gene, 14 bp upstream of the oppC gene, 16 bp upstream of oppD, and 10 bp upstream of oppF. No oppB-specific ribosomal binding site was detected.

View larger version (43K): [in this window] [in a new window]   FIG. 2.   Northern blot analysis. The digoxigenin-labeled probes for the P100 gene (p100), oppB, oppC, oppD, and oppF were hybridized to 20 µg of electrophoretically separated RNA from M. hominis FBG. The marker was a 0.24- to 9.5-kb RNA ladder (Gibco BRL).

5' Mapping of the opp mRNA. The transcriptional initiation site for the 9.5-kb mRNA was determined by primer extension. Three antisense oligonucleotide primers were used: P100-pe1 and P100-pe2, annealing to the N-terminus-encoding region of the P100 gene, and oppB-pe1, annealing to the 5' end of oppB. No distinct product was obtained by use of oppB-pe1, whereas with both P100 primers, the 5' end of the mRNA was mapped to a guanosine at nucleotide 307 (Fig. 3), thus indicating that the polycistronic mRNA initiates 323 nucleotides upstream of the translational start codon of the P100 gene.

View larger version (66K): [in this window] [in a new window]   FIG. 3.   Primer extension analysis of the opp operon. The transcriptional initiation site for the 9.5-kb RNA was determined by primer extension with two antisense oligonucleotides of the P100 gene, P100-pe1 and P100-pe2, and an oppB-specific oligonucleotide, oppB-pe1, as described in Materials and Methods. The corresponding DNA sequences (TGCA) were generated by the dideoxy chain termination method with the same primers. The primer extension products (lanes P) are marked by arrows. As P100-pe1 (p100-1) binds at a site distal to P100-pe2 (p100-2), the sequence and transcription initiation site with this probe are seen higher up but are at the same residue. The nucleotide sequence of the region of the transcriptional start site is indicated, and the transcriptional start site itself is underlined.

Characterization of P100 as the oligopeptide-binding protein OppA. After the demonstration that the P100 gene and oppBCDF of M. hominis are organized within a transcriptionally active operon structure, we wanted to analyze whether P100 is part of the proposed oligopeptide permease of M. hominis. Like the experiments of Guyer and coworkers in characterizing the binding specificity of OppA of E. coli (14), we tested the peptide-binding capacity of P100 by spectrofluorometry. With 16 internal tryptophan residues, P100 exhibited excellent native tryptophan fluorescence when excited at 290 nm, with an emission maximum at 330 nm (Fig. 4). The addition of Lys₃ resulted in a substantial increase in tryptophan fluorescence and in a red shift of 1.6 nm of the emission maximum caused by conformational changes when the peptide was bound. The addition of the same volume of buffer resulted in a decrease in tryptophan fluorescence due to a dilution effect. Thus, the fluorescence increased by about 7% after peptide binding, confirming that P100 is an oligopeptide-binding protein. The fluorescence emission spectra of bovine serum albumin and carbonic anhydrase, used as controls, were not increased after the addition of Lys₃; the spectra remained unchanged for carbonic anhydrase and decreased for bovine serum albumin, as was observed after the addition of buffer alone. As has been shown for OppA of E. coli, we also did not observe a reproducible increase in the emission spectrum with Ala₃ as a substrate; however, with the pentapeptide Ala₅, we observed the same increase in tryptophan fluorescence as that depicted with Lys₃ in Fig. 4.

View larger version (13K): [in this window] [in a new window]   FIG. 4.   Fluorescence spectra of P100 affected by peptides. The relative fluorescence of lipoprotein P100 at a concentration of 1.5 µM (0.15 mg/ml) in 10 mM imidazole acetate (pH 7.3)-100 mM NaCl-0.03% n-dodecyl--D-maltoside was determined as a function of wavelength, with excitation at 290 nm. Symbols: , fluorescence of the protein alone; ---, fluorescence of the protein in the presence of 100 µM Lys3 (1 µl of peptide solution per 100 µl of probe); ···, fluorescence of the protein after the addition of 1 µl of buffer per 100 µl of probe. All spectra were obtained 5 min after the peptide was mixed with the P100 protein and are depicted as smooth curves.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Mycoplasmas are restricted to parasitic behavior because of their small genome size. Thus, ABC transport systems play an important role in the uptake of essential substrates, including sugars, proteins, and peptides.

The proposed Opp transport system of M. hominis described here displayed little overall sequence similarity with the respective domains of other species but shared their typical features: it is composed of the four core domains OppBCDF and the cytoadherence-associated lipoprotein P100 as the substrate-binding domain OppA. The homologies of the hydrophobic integral domains OppB and OppC were greater when the regions which are thought to interact with the ATP-binding domains were compared. Both domains were predicted from their sequences to cross the membrane six times, forming a pore which has 12 transmembrane segments and which has been predicted to be generally required for the transport of substrates (oligopeptides) through the membrane (21, 23). In addition, they carry in their C-terminal portions a conserved hydrophilic segment which is found in bacterial binding-protein-dependent permeases (40). The conserved nature of these functionally important structures in the hydrophobic Opp domains of M. hominis is a characteristic feature of active genes.

The two ATP-binding domains, OppD and OppF, showed homologies to the respective domains of the other species of up to 41.9%. This result is in good agreement with the finding that, in general, the ATP-binding domains show sequence identity of 30 to 50% and are thus more homologous than the respective integral domains of the transport system (18, 21). The greater homology detected for the ATP-binding sites indicates once more the selection pressure on functionally important structures of the permease system. A common feature of mycoplasma OppF domains seems to be their larger size (Fig. 1): in comparison with nonmycoplasma OppF domains, which have molecular masses of about 36 kDa, they have calculated molecular masses of about 98 kDa. What consequences this size difference may have on the function of the transporter remains to be elucidated.

In gram-negative bacteria, three distinct permease systems have been described for the separate transport of di-, tri- and oligopeptides through the membrane (19, 22, 25), whereas only two distinct systems have been detected for translocating tetra- and tri- or pentapeptides through the membrane in B. subtilis (29, 30). For the Mollicutes class, an ABC transport system that was well characterized at the protein level for M. hyorhinis was similar to that analyzed for gram-positive and -negative bacteria (9, 13). Based on genomic analyses, the presence of an ABC transporter for oligopeptides was proposed for M. pneumoniae (24) and M. genitalium (12), and this transporter should consist of the four core domains OppBCDF. However, the oppA gene has not been described for these two species.

As a P100-deficient mutant has never been described in the literature, has not been detected among our collection of isolates, and could not be generated by incubating M. hominis with P100-specific monoclonal antibodies, we speculate that P100 is essential for M. hominis. Thus, the question arises as to whether OppA is generally dispensable in mycoplasmas or is simply less conserved, thus rendering its identification more difficult (23, 41). The findings presented in this paper may support the latter notion, as we (i) identified the surface-localized lipoprotein P100 as the oligopeptide-binding protein OppA in M. hominis, (ii) demonstrated that the P100 gene is polycistronically organized with oppBCDF, and (iii) detected the OppA-encoding genes of M. pneumoniae and M. genitalium in homology searches with a P100 region conserved in many peptide-binding proteins. The presence of P100 as part of the Opp transporter was supported by Northern blot analysis, in which a P100 gene-less mRNA of the opp operon was never detected.

To prove the hypothesis that the cytoadhesive lipoprotein P100 functions as the substrate-binding domain OppA in the proposed oligopeptide permease of M. hominis, we tested the peptide-binding capability of solubilized P100 by spectrofluorometry by the procedure of Guyer and coworkers (14). In contrast to OppA from E. coli, which they studied, we had to deal with the disadvantage of P100 carrying a lipid moiety. Thus, it was difficult to isolate functionally active and detergent-free P100, and nearly every third preparation failed to react. Nevertheless, we were able to demonstrate the binding of tri- and pentapeptides to P100 by spectrofluorometry and to confirm the interaction of P100 with different peptides by bimolecular interaction analysis. The fact that P100 binds peptides of different lengths could endow the transporter with a less restrictive substrate specificity.

That the cytoadherence-mediating P100 protein also functions as an oligopeptide-binding domain of M. hominis corresponds to findings that the adherence of other organisms can be affected by mutations in distinct domains of the permease complex (8, 27, 28). Cundell and coworkers found that mutations in the peptide-binding proteins of the transport system of Streptococcus pneumoniae led to a reduction of bacterial affinity for the host (8). The adherence of Streptococcus gordonii was affected by the peptide-binding protein SarA as well as by mutations in the oppC gene (27, 28). Thus, both the oligopeptide-binding proteins and the entire oligopeptide transport system can be involved in bacterial adhesion, which we will now analyze at the protein level for M. hominis.