Multiple Promoter Inversions Generate Surface Antigenic Variation in Mycoplasma penetrans

Mycoplasma strains and culture conditions.

M. penetrans GTU-54 was the original isolate from a urine sample of an HIV-infected patient (21) and was kindly provided by S.-C. Lo (Armed Forces Institute of Pathology, Bethesda, Md.). M. penetrans HF-2 was isolated from an HIV-negative patient with primary antiphospholipid syndrome (47) and was kindly provided by L. Cedillo and A. Yáñez (Centro de Investigación Biomédica de oriente-IMSS and Benemérita Universidad Autónoma de Puebla, Puebla City, Mexico). M. penetrans strains were cultured in PPLO medium (2.1% PPLO broth [Difco Laboratories, Detroit, Mich.], 0.25% glucose, 0.002% phenol red, 5% yeast extract [Difco Laboratories], 10% horse serum [Gibco BRL, Rockville, Md.], 50 μg of ampicillin per ml) at 37°C.

Antibodies.

A murine monoclonal antibody (MAb) specific to P35 (MAb 7) was established by T. Sasaki and was previously shown to react specifically with the P35 protein (24, 33). MAb 7 was used in this study at a 1:5,000 dilution for immunoblot analysis. Serum 6 (anti-HF-2) was produced by immunization of mice with total cell lysate of M. penetrans HF-2 and was used at a 1:500 dilution for immunoblot analysis.

Protein analysis.

M. penetrans proteins were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) under the conditions described by Laemmli (19). In most cases, 12% gels were used. Fractionation of LAMPs was done by the Triton X-114 (TX-114) phase-partitioning method (9). Membrane and cytosolic protein fractionation and TX-100 partitioning were also performed using published methods (28). For immunoblot analysis, proteins were transferred to nitrocellulose membranes after electrophoresis (38) and were detected by antibodies.

Peptide sequencing of the 46-kDa protein.

The 46-kDa protein was extracted from strain HF-2 by TX-114 phase partitioning and was purified by SDS-PAGE. The 46-kDa protein band was excised from the gel and treated with Staphylococcus aureus V8 protease (Sigma-Aldrich, Steinheim, Germany). The digested peptide fragments were separated using SDS-15% PAGE gels, transferred to a polyvinylidene difluoride membrane (Bio-Rad, Hercules, Calif.), and stained with Coomassie blue. Two major peptide fragment bands (approximately 15 and 8 kDa) were excised from the membrane, and the N-terminal sequences were analyzed by Edman degradation. Peptide sequencing was performed at APRO Life Science Institute (Tokushima, Japan) using the Procise 494 HT protein-sequencing system (Applied Biosystems, Foster City, Calif.).

DNA-sequencing analysis of pepIMP13 gene region of strain GTU.

Genomic DNA was isolated from a 5-ml culture of M. penetrans strain GTU (QIAmp DNA Mini Kit; Qiagen, Hilden, Germany) and was used as a template for PCR amplification. The pepIMP13 gene region was amplified by PCR with the oligonucleotide primers IMP13-CF (GCAACTGCAGATGGCAACAA) and IMP13-CR (ATGGCACCGCCTGATAACAT) using a high-fidelity DNA polymerase, Pyrobest (Takara, Tokyo, Japan). The amplified fragments were ligated into the SmaI site of the pUC19 plasmid. Sequencing of cloned PCR fragments was performed by a primer-walking method with the Big Dye terminator cycle-sequencing kit and the DNA sequencer PRISM 310 (Applied Biosystems). To avoid the artificial factor of mutations generated in the course of PCR, three independent plasmid clones were sequenced, and the data were integrated.

RNA isolation and slot blot analysis.

Total cellular RNA was isolated from mid-logarithmic-phase cultures of both M. penetrans GTU and HF-2 (RNeasy Mini Kit; Qiagen). Twenty micrograms of total RNA were treated with 10 U of RNase-free DNase (Takara) for 1 h at 37°C. One microgram of RNA was diluted in RNase-free denaturation solution containing formamide (66%), formaldehyde (8%), and MOPS (morpholinepropanesulfonic acid) buffer and heated at 65°C for 5 min. The RNAs were blotted onto a GeneScreen Plus membrane (NEN, Boston, Mass.) by vacuum, using the Convertible Filtration Manifold System (Invitrogen, Carlsbad, Calif.). The oligonucleotide probes P35-P (CCCTTAATTGCAGCAGAATCACC) for the p35 gene transcript and P42-P (TTAAATCTGTTTCAGCTGTAATTT) for the p42 gene transcript were enzymatically labeled with digoxigenin (DIG)-labeled ddUTP by using the DIG oligonucleotide 3′-end labeling kit (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's instructions. The blots were hybridized with the probes at 37°C for 16 h in DIG Easy Hyb solution (Roche Diagnostics). After incubation, the hybridization signals were detected with the DIG luminescence detection kit (Roche Diagnostics) and visualized by exposure to medical X-ray film (Fuji Film, Tokyo, Japan).

Primer extension analysis.

Two oligonucleotide primers, P42-EXT3 (GCAACAATCCCAAAAGCT) and P42-EXT4 (TCCATTTCCATTATTGTTAT), were used for primer extension analysis to identify the 5′ end of the p42 gene transcript. The primer extension reaction was performed using the primer extension system-avian myeloblastosis virus (AMV) reverse transcriptase (Promega, Madison, Wis.). Briefly, the primers were end labeled using T4 polynucleotide kinase and [γ-³²P]ATP. Ten micrograms of total RNA prepared from M. penetrans HF-2 was mixed with each of the labeled primers in the AMV primer extension buffer and heated at 58°C for 20 min. The mixtures were cooled at room temperature for 10 min, and the AMV reverse transcriptase mix solution, containing 1 U of enzyme, was added. The extension reaction mixtures were incubated at 42°C for 30 min, 20 μl of loading dye was added, and the extension was terminated by heat inactivation at 90°C for 10 min.

The p42 gene region was cloned from M. penetrans HF-2 genomic DNA by PCR with primers IMP13-CF and IMP13-CR. The amplified p42 gene fragment was purified with the QIAquick PCR purification kit (Qiagen) for use as a template for DNA-sequencing reactions. The same primers that were used for primer extension were also used for sequencing. The fmol DNA-sequencing system (Promega) was used for DNA-sequencing reactions. The products of primer extension and sequencing reactions were analyzed by electrophoresis, using a 5% polyacrylamide sequencing gel. After electrophoresis, the gel was dried and exposed to X-ray film overnight to visualize the products.

Nucleotide sequence accession numbers.

The complete genome sequence of M. penetrans strain HF-2 was recently determined by our group in collaboration with Kitasato University. The detail of the genome sequence analysis will be described in another paper (33a). The whole-genome sequence data for M. penetrans HF-2 will appear in the DDBJ, EMBL, and GenBank databases under accession numbers AP004170, AP004171, AP004172, AP004173, and AP004174. The nucleotide sequence data for the p42 gene region of M. penetrans GTU will also appear in the DDBJ, EMBL, and GenBank nucleotide sequence databases under accession number AB084070.

Comparison of protein profiles between M. penetrans strains.

The protein electrophoresis patterns of the M. penetrans type strain, GTU, and isolate HF-2 were compared to identify molecular polymorphisms between M. penetrans strains (Fig. 1A). The major surface lipoprotein P35 of M. penetrans was present in both strains. However, a 34-kDa protein was present only in GTU, and a 46-kDa protein was found only in HF-2. The electrophoresis patterns were reproducible when the protein samples were prepared from cultures of other single colonies of each strain. Immunoblotting analysis gave a similar result (Fig. 1B and C). MAb 7 detected the P35 protein in both strains (Fig. 1B), while serum 6 (anti-HF-2) reacted with many proteins of both strains (Fig. 1C). Although the strongest reaction was observed against the P35 protein, serum 6 clearly detected a 46-kDa protein only in HF-2.

Characterization of 34- and 46-kDa proteins.

We thought that the 34- and 46-kDa proteins were variable surface lipoproteins of the P35 family (products of mpl genes) because of their different expression patterns among M. penetrans strains. To confirm this, we characterized the properties of the 34- and 46-kDa proteins. First, we fractionated the proteins of M. penetrans strains GTU and HF-2 by a TX-114 extraction procedure that is used to separate LAMPs of M. penetrans (9, 33). The 34- and 46-kDa proteins were found in the TX-114 detergent phase together with the P35 protein (Fig. 2). Differences between the TX-114 extraction profiles of strains GTU and HF-2, in addition to the 34- and 46-kDa proteins, were also observed (Fig. 2). The GTU strain had at least two distinct bands, between 36 and 40 kDa, that were missing in the HF-2 profiles. One of these bands might be the P38 lipoprotein previously reported to be present in strain GTU (24, 33). We then analyzed the GTU and HF-2 proteins by membrane and cytosolic separation or by the TX-100 extraction method (28). In these analyses, the 34- and 46-kDa proteins were separated into membrane or TX-100-soluble fractions, along with the P35 protein (data not shown). These results suggested that the 34- and 46-kDa proteins are hydrophobic proteins with properties similar to those of the P35 lipoprotein.

The 46-kDa protein was chosen for amino acid-sequencing analysis. Initial attempts to analyze the intact 46-kDa protein by Edman degradation were unsuccessful, probably because of N-terminal blockage. Therefore, the 46-kDa protein was digested with protease V8 and peptide fragments of ∼15 and 8 kDa were sequenced by Edman degradation. The sequence LNDKVSLAGS was obtained from the N terminus of the 15-kDa fragment. Two sequences, TDLKITVDGG and FNFNIGIDST, were obtained from the analysis of the 8-kDa fragment, suggesting that it was a mixture of two peptides. These partial amino acid sequences of the 46-kDa protein matched the deduced amino acid sequence of a previously reported putative mpl gene, pepIMP13 (24). The previously reported pepIMP13 sequence is derived from strain GTU (GenBank accession no. AJ006698) and is truncated at the N terminus. To obtain the full-length pepIMP13 gene sequence, the complete genome sequence of strain HF-2 (see Materials and Methods) was searched. The pepIMP13 gene was found in the M. penetrans HF-2 genome as an open reading frame (ORF), designated MYPE6630 in accordance with the nomenclature system of the genome project. The ORF MYPE6630 was located at nucleotide positions 850970 to 852142 of the genome. The amino acid sequence of the 46-kDa protein was deduced from the MYPE6630 sequence and was compared to that of the P35 protein (Fig. 3). The full-length 46-kDa protein (the pepIMP13 protein) consists of 390 amino acids (aa) and possesses an N-terminal sequence identical to that of the P35 protein. This 30-aa N-terminal sequence is thought to be the signal peptide of the P35 family lipoproteins (24). The calculated molecular mass of the 46-kDa protein was 41,814 Da, somewhat smaller than that estimated by SDS-PAGE. We designated the 46-kDa protein P42, based on the calculated molecular mass.

Analysis of the nucleotide sequence of the p42 gene and flanking region.

To investigate the mechanism that underlies the antigenic variation of the 46-kDa protein (P42), we analyzed the nucleotide sequences of the p42 gene of strains HF-2 and GTU. The nucleotide sequence containing the p42 gene and flanking regions of HF-2 was obtained from complete genome sequence data. Using this sequence, we designed PCR primers to amplify the corresponding region from strain GTU. The PCR-cloned p42 gene region from GTU was sequenced by the primer-walking method and was compared to the sequence from HF-2. There were no differences in either the p42 structural gene itself or the downstream region between the two strains. However, sequence differences were found in the upstream region of the p42 gene (Fig. 4). The difference began 238 bp upstream from the ATG start codon of the p42 gene and was 135 bp in length (Fig. 4A). It was found that the 135-bp DNA sequence was inverted between the GTU and HF-2 strains and was flanked by 12-bp inverted-repeat sequences (Fig. 4A), suggesting that the inversion might be produced by site-specific recombination between 12-bp inverted-repeat sequences. We hypothesized that this inversion was the cause of the antigenic variation of the P42 protein and would affect the structure of promoter or other regulatory sequences. As expected, inspection of the 135-bp inverted DNA sequences revealed the presence of a promoter-like sequence in this region. The promoter-like sequence was very similar to the experimentally characterized promoter of the p35 gene (24). In the region containing a −10-like consensus sequence and a +1 transcription start site, the p42 promoter-like sequence and the p35 promoter were identical (Fig. 4B). In strain HF-2, the promoter-like sequence was oriented to transcribe the p42 gene, but this orientation was inverted in strain GTU (Fig. 4A). The directions of the promoter-like sequence (toward the p42 structural gene or not) were consistent with the expression patterns of the P42 protein in strains GTU and HF-2 (Fig. 1 and 2).

The existence of the 12-bp inverted repeat that flanked the promoter-like sequence raised the possibility that inversion of the promoter-like sequence would be reversible. Therefore, we analyzed the shotgun clones that were used in the whole-genome sequencing of strain HF-2 and found a clone that carried the p42 promoter region in the inverted orientation that was identical to the promoter region found in strain GTU (data not shown). This finding strongly suggests that the p42 promoter region is inverted in the strain HF-2 population during culture.

A characteristic 16-bp inverted repeat that partly overlapped one of the 12-bp inverted-repeat sequences adjacent to the 135-bp DNA was also found (Fig. 4A and 5A). This 16-bp inverted repeat seems to form a hairpin structure that resembles a terminator sequence (Fig. 5B). The calculated ΔG of this structure is −22.5 kcal/mol. The 16-bp inverted repeat can be formed only in the promoter orientation that is characteristic of GTU. In HF-2, DNA inversion between 12-bp inverted-repeat sequences disrupts the formation of the 16-bp inverted repeat and the hairpin structure (Fig. 5).

Transcriptional analysis of p42 gene.

To confirm whether the inversion of promoter-like sequence actually affects the expression of the p42 gene, we analyzed the transcription of the p42 gene. Total RNAs were isolated from strains GTU and HF-2 and analyzed by slot blot hybridization (Fig. 6). Using the oligonucleotide probe for the p35 gene transcript, the hybridization signals from both strains were observed. In contrast, a hybridization signal was obtained only from HF-2 with the probe for the p42 transcript. These results indicate that p42 transcription occurs in HF-2 but not in GTU. To further characterize the p42 promoter-like region, we performed primer extension analysis. Using the oligonucleotide primer P42-EXT3, the extension product was obtained with RNA from the HF-2 strain (Fig. 7A). The probable transcriptional start site was identified 137 bp upstream from the ATG start codon. This position corresponds to 3 bp upstream of the previously reported transcriptional start site of the p35 gene (24) (Fig. 7B). The primer extension product was also obtained with the other oligonucleotide primer, P42-EXT4, and the same start site was identified (data not shown). These results indicated that the p42 promoter-like sequence is an active promoter in M. penetrans cells.

Organization of mpl genes in M. penetrans HF-2 genome.

It was demonstrated that the promoter inversion was the cause of antigenic variation of P42 protein. The involvement of a promoter inversion mechanism in p42 gene expression raised the question of whether other mpl gene expression is regulated by a similar mechanism. To answer this question, we searched for all of the mpl genes in the whole-genome sequence of M. penetrans HF-2, using the p35 gene sequence as a query. Homology search revealed that the M. penetrans HF-2 genome contains at least 38 mpl genes, including p35 and p42. The deduced amino acid sequences of these mpl genes showed homology to that of the P35 protein (34 to 70% identity) and had almost identical mpl signal sequences in the N termini (data not shown). The signal sequences contain one cysteine residue (Fig. 3) that is thought to be the site which is modified with fatty acids, as are other known surface lipoproteins (5, 45). In contrast to the signal peptide sequences, the rest of the amino acid sequences of the mpl genes were of low homology. The homology search also detected six other genes (ORFs MYPE7020, -7030, -7040, -7050, -7060, and -7070) that were similar to the p35 gene, with homology ranging from 28 to 44% identity (data not shown). However, the deduced amino acid sequences of these six ORFs did not have the P35 signal peptide sequence in their N termini, so we did not include these genes among the members of the mpl gene family.

The 38 mpl genes clustered at three positions of the genome (Fig. 8). The largest cluster was located at nucleotide positions 830000 to 882000 of the genome (Fig. 8A). This 50-kb region contained 30 mpl genes and 6 non-mpl genes. The p35 and p42 genes were found in this cluster (MYPE6810 and -6630). The second cluster was found in a 20-kb DNA region at nucleotide positions 335500 to 355500 (Fig. 8B). In this region, two pairs of mpl genes are separated by five non-mpl genes. The last mpl gene cluster was located at nucleotide positions 966000 to 975000 (Fig. 8C). In this region, four mpl genes were found within 9 kb of DNA. In all three clusters, all the identified mpl genes were oriented in the same direction. The previously identified mpl genes, p34A (pepIMP14), p30 (p33), and p38 from strain GTU (24, 33), were also found in the largest cluster (Fig. 8). However, the previously reported pepIMP12 gene of GTU (24) (GenBank accession no. AJ006697) was not found in the HF-2 genome sequence. To determine whether the missing pepIMP12 gene in the HF-2 genome was the result of genomic polymorphism between strains GTU and HF-2, we used PCR to amplify the pepIMP12 gene sequence with primers IMP12-F (TAATATTAAATCTTTAGATG) and IMP12-R (AATTAAATGATAAAGTTAGC). Unexpectedly, the pepIMP12 sequence was not amplified from our GTU and HF-2 strains (data not shown). The reason is as yet unknown. It was also found that three mpl genes (MYPE6520, -6500, and -7380) possessed frameshift mutations in their sequences and were disrupted by internal stop codons. Of 14 non-mpl genes that exist in the three mpl gene clusters, MYPE6600, -6610, and -6620 showed relatively high homology to the transport system permease protein P69 (Mycoplasma hyorhinis), the ABC transporter ATP-binding protein (Mycoplasma pulmonis), and the high-affinity transport system protein P37 (M. hyorhinis), respectively. However, the other non-mpl genes did not show any significant homology to other known proteins (data not shown).

To investigate the mechanism involved in the expression of these mpl genes, we analyzed the intergenic sequences of these clusters. This showed that the intergenic sequences are well conserved and have structures similar to that of the p42 promoter region, namely, 133- to 138-bp DNA sequences were flanked by 12- to 14-bp inverted-repeat sequences, suggesting that these regions are also invertible DNA (data not shown). Specifically, it was also noted that the 133- to 138-bp DNA sequences were fairly well conserved but the sequences of the 12- to 14-bp inverted repeats were not conserved between intergenic sequences (data not shown). In most cases, the 133- to 138-bp DNA regions contained a highly conserved promoter-like sequence that was similar to those of the p35 and p42 promoters. Although the −35-like consensus regions of these promoter-like sequences were not conserved, the −10-like consensus and transcriptional-start-site regions were almost identical to those of the p35 and p42 promoters (data not shown). As the lack of a −35-like consensus sequence in mycoplasmal promoters has been reported in Mycoplasma pneumoniae (44), these promoter-like sequences may be active. These results suggested that most of the mpl genes possessed independent invertible promoter-like sequences in their upstream regions (Fig. 8). (See also the genome sequence data for M. penetrans HF-2 and their annotation. This will appear in databases.)

In summary, we have identified the invertible promoter-like sequences for 31 mpl genes (including p35 and p42) but not those for 7 other mpl genes (MYPE 6790, -6590, -2620, -2630, -7380, -7370, and -7330) (Fig. 8). Interestingly, of these 31 invertible promoter-like sequences, only two (for the p35 and p42 genes) were in the ON orientation. The other promoter-like sequences were in the OFF orientation (Fig. 8). These patterns of the promoter-like sequences of mpl genes are consistent with the expression patterns of P35 family lipoproteins of M. penetrans HF-2 (i.e., P35 and P42 are major lipoproteins expressed in strain HF-2) (Fig. 2).

We have also identified 16-bp inverted-repeat sequences upstream of all invertible promoter-like sequences, with the exception of MYPE7400 (Fig. 8). These 16-bp inverted-repeat sequences were homologous to those found upstream of the p42 gene of GTU (Fig. 4 and 5).