Differential Expression of the Bhmp39 Major Outer Membrane Proteins of Brachyspira hyodysenteriae
http://www.100md.com
感染与免疫杂志 2006年第6期
Australian Research Council Centre of Excellence in Structural and Functional Microbial Genomics, Department of Microbiology, Monash University, Melbourne, Victoria 3800, Australia
ABSTRACT
The enteric, anaerobic spirochete Brachyspira hyodysenteriae is the causative agent of swine dysentery, a severe mucohemorrhagic diarrheal disease of pigs that has economic significance in every major pork-producing country. Recent investigation into potential vaccine candidates has focused on the outer membrane proteins of B. hyodysenteriae. Bhmp39 (formerly Vsp39) is the most abundant surface-exposed outer membrane protein of B. hyodysenteriae; its predicted gene sequence has previously been shown to share sequence similarity to eight genes divided evenly between two paralogous loci. The peptide sequence suggested that Bhmp39 is encoded by one of these genes, bhmp39h. The biological significance of maintaining eight homologous bhmp39 genes is unclear, though it has been proposed that this may play a role in antigenic variation. In this study, real-time, reverse transcription-PCR was used to demonstrate that bhmp39f and bhmp39h were the transcripts most abundantly expressed by B. hyodysenteriae strain B204 cultured under in vitro growth conditions. Mass spectrometry data of the purified 39-kDa membrane protein showed that both Bhmp39f and Bhmp39h were present. Northern blot analysis across predicted Rho-independent terminators demonstrated that the genes of the bhmp39efgh locus result in monocistronic transcripts.
INTRODUCTION
The intestinal spirochete Brachyspira hyodysenteriae is the causative agent of swine dysentery, a severe mucohemorrhagic diarrheal disease of pigs. Previous vaccination trials using killed or attenuated live cells as vaccines have not stimulated immunoprotection against swine dysentery (15, 16), while commercially available bacterin vaccines fail to provide complete protection (8). An alternative approach may be to generate subunit vaccines that might be delivered by the expression of recombinant protein on the outer membrane of a bacterial delivery vector, such as Salmonella enterica serovar Typhimurium (9, 17).
Recent investigations into potential targets for novel vaccines against swine dysentery have focused on the outer membrane proteins of B. hyodysenteriae, though much is still unknown about these proteins. Hampson et al. have recently proposed that a new nomenclature be introduced for membrane proteins and lipoproteins of B. hyodysenteriae (13); we have used that nomenclature in this paper. It has been suggested that B. hyodysenteriae evades the host immune system through antigenic variation of the expression of tandemly arranged paralogous genes (6). Known examples are the bhlp29.7abcd (previously blpGFEA) locus (5) and the two bhmp39 (previously vsp) loci, containing bhmp39a to bhmp39d and bhmp39e to bhmp39h (12, 18, 19). Variations in the expression of key outer membrane antigens may be a factor in the failure of vaccines to provide full immunoprotection against swine dysentery. Mechanisms of antigenic variation have been characterized for other pathogenic spirochetes, including Treponema pallidum (22), Borrelia burgdorferi, and Borrelia hermsii (6), though such mechanisms have not yet been investigated in B. hyodysenteriae.
Bhmp39 is the most abundant outer membrane protein of B. hyodysenteriae and is known to be surface exposed based on surface iodination studies (12). Eight paralogous bhmp39 genes have been identified, and N-terminal sequencing of the 39-kDa protein indicated that the expressed protein was encoded by bhmp39h (18). In this study we sought to determine the transcriptional organization of the bhmp39 genes under in vitro conditions and to determine which protein(s) is expressed by the bhmp39 genes.
MATERIALS AND METHODS
Bacterial strains. B. hyodysenteriae strain B204 was kindly provided by P. Coloe, Royal Melbourne Institute of Technology, Melbourne, Australia. Cells were grown anaerobically (80% N2, 10% CO2, 10% H2) at 37°C to between 5 x 107 cells/ml and 2 x 108 cells/ml in Trypticase soya broth supplemented with 0.5% yeast extract, VPI salts, and 5% fetal calf serum (23). Broth cultures were subjected to continuous agitation with magnetic stirrers.
Oligonucleotides. Oligonucleotides used in this study are listed in Table 1. Gene-specific nucleotide sequences were identified by DNA sequence alignments of the bhmp39 gene sequences in GenBank (accession numbers AF012102 and AY027775), and where primers were to be used in real-time reverse transcription-PCR (RT-PCR), Primer Express 1.5a software (Applied Biosystems, Inc., Foster City, CA) was utilized in the design process.
Molecular biology methods. Chromosomal DNA was prepared from B. hyodysenteriae using the cetyltrimethylammonium bromide method (2). Standard methods in molecular biology were performed as described by Ausubel et al. (2). Annealing steps in PCR cycling were performed at 47°C to compensate for the low G+C content of B. hyodysenteriae genomic DNA. When PCR products were required for sequencing, they were purified with a QIAquick PCR purification kit (QIAGEN, Hilden, Germany). Nucleotide sequencing was performed with a BigDye Terminator v3.1 cycle sequencing kit (Perkin-Elmer Corp., Norwalk, CT) and an Applied Biosystems Inc. model 3730 automated sequencer.
RNA isolation and purification. RNA was isolated from B. hyodysenteriae B204 cultures that had cell densities between 5 x 107 and 2 x 108 cells/ml with TRIzol (Invitrogen, La Jolla, CA) at 4°C, according to the manufacturer's instructions. All work using RNA was performed in the presence of RNasin (Promega, Madison, WI). Purified RNA was treated with 20 U of DNase (Roche Molecular Biochemicals, Basel, Switzerland) for 20 min at 37°C, and the RNA was further purified by using RNeasy columns (QIAGEN).
Northern blotting. Northern blotting was carried out as described by Cullen et al. (5). Oligonucleotides containing the T7 promoter sequence were designed to amplify 300- to 321-bp templates for the synthesis of riboprobes complementary to portions of bhmp39e, bhmp39g, and bhmp39f/bhmp39h (single probe for two genes) (Table 1). A riboprobe was generated that hybridized to the transcripts of both bhmp39f and bhmp39h due to the high sequence identity between the two genes.
Reverse transcription. RT reactions were performed as described by Boucher et al. (3). Negative-control reactions were performed by omitting the reverse transcriptase. The cDNA templates were generated by incubating the reaction mixes at 42°C for 2.5 h, followed by denaturation at 70°C for 15 min. cDNA samples were diluted 1:150 and stored at –20°C until required.
Real-time RT-PCR. Real-time RT-PCR was performed using an ABI PRISM model 7700 sequence detector, with product accumulation quantified by the incorporation of the fluorescent dye SYBR green. Primers for use in real-time RT-PCR were designed to amplify 86 to 162 bp of target bhmp39 genes. Reactions were performed in triplicate, with each 20-μl reaction mix containing 2.4 μl of the DNA template, 1.6 μl of primer (625 mM), and 10 μl of SYBR green 2x PCR master mix (Applied Biosystems, Inc.), in a total volume of 20 μl. Gene-specific standard curves were constructed from 20-fold dilutions of B. hyodysenteriae B204 genomic DNA and were used to determine relative cDNA template concentrations in each reaction mixture. Gene expression levels were normalized against gyrB (Table 1). All primer pairs for use in real-time RT-PCR were confirmed to amplify single gene products by melt curve analysis and by sequencing PCR products. Relative levels of gene expression were averaged from the three technical replicates for each of three biological replicates and compared using Welch-corrected, unpaired t tests.
Analysis of Bhmp39 proteins in the B. hyodysenteriae membrane. Urea-insoluble membrane proteins were purified from B. hyodysenteriae B204 by solubilization with 1% Tween 20 and ultracentrifugation, as described by Gabe et al. (11). Sample preparation and matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry were carried out as described previously (5). A list of monoisotopic peaks consisting of the bhmp39 gene products was used to search both local and online protein databases. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of protein samples was carried out using a Mini-PROTEAN 3 electrophoresis apparatus (Bio-Rad Laboratories, Hercules, CA) as described previously (5).
Bioinformatics analysis. BLAST programs (1) were used to perform similarity searches of the GenBank nucleotide and protein sequence databases, as appropriate. DNA sequences were aligned using the ClustalW algorithm (21). Transcript sizes from Northern hybridization were predicted using the Sequaid II software (University of Kansas, Kansas City, KS). Peptide masses from protease digestion of proteins were determined using PeptideMass (24). MASCOT (20) was used to identify proteins from peptide mass data by searching the "Other Bacteria" subset of the NCBI nonredundant database. The algorithm was instructed to exclude peptides with missed cleavages, and the peptide tolerance was set to ±0.2 Da. A protein score greater than 63 was considered to be significant (P < 0.05) when data were analyzed with MASCOT. Statistical analyses were conducted using the InStat program (Graphpad Software, Inc., San Diego, CA). The strength of transcriptional termination of Rho-independent terminators was predicted using RNAfold (14).
RESULTS
Prediction of Rho-independent transcriptional terminators. Inspection of the sequence downstream of each of the bhmp39 coding regions revealed the presence of putative Rho-independent terminators. RNAfold was used to predict the minimum free energy of each inverted repeat and thereby enable a comparison of the strength of each of the putative transcriptional terminators (Table 2). In each instance, the inverted repeat of each putative terminator begins 26 to 43 bp downstream of the stop codon, and the inverted repeats form the stem-loop structure, which is 9 to 14 bp in length. Each putative terminator contained a much higher G+C content (about 4 to 10 bp more) than the flanking nucleotide sequence, which resulted in highly negative G values for all of the predicted terminators, except for that of bhmp39g, with which a high number of mismatches resulted in a predicted free energy that was almost 0. We resequenced the bhmp39c terminator region in order to clarify the undefined bases in the available lodged sequence. The RNA secondary-structure prediction from the revised sequence suggested that it would form an inverted repeat of a length and strength similar to those of the other predicted terminators (Table 2).
Northern hybridization. Northern hybridization analysis was performed to determine the approximate size of the B. hyodysenteriae B204 bhmp39efgh transcripts (Fig. 1). Individual transcripts of the bhmp39abcd locus could not be investigated due to difficulty in designing riboprobes of appropriate lengths and with sufficient specificity to individual genes. The riboprobe specific for bhmp39e identified a single transcript of 1.1 kb (Fig. 1a), but no transcript was detected for bhmp39g (Fig. 1b). The bhmp39f/bhmp39h riboprobe hybridized to a product of approximately 1.2 kb (Fig. 1c).
Real-time RT-PCR quantification of bhmp39 mRNA. In order to ascertain which bhmp39 genes were expressed by cells grown under in vitro conditions, the transcription of seven of the eight bhmp39 genes of B. hyodysenteriae strain B204 was investigated by real-time RT-PCR (Fig. 2). Primers used for real-time RT-PCR were designed specifically to amplify an internal region of each of the bhmp39 genes from B. hyodysenteriae strain B204; this was not possible for bhmp39c because of its very high nucleotide sequence identity to bhmp39b and bhmp39d. The expression of gyrB was used as an internal control, as described previously (4), in order to compare the relative levels of expression of bhmp39 genes from different biological replicates. Averaged data from three technical replicates of three biological replicates indicated that bhmp39f was the most highly transcribed bhmp39 gene, followed by bhmp39h (Fig. 2).
Identification of the 39-kDa urea-insoluble membrane protein. Having observed that more than one bhmp39 gene was transcribed under the culture conditions used in this study, we set out to determine the identity of the 39-kDa outer membrane protein. We purified membrane proteins from B. hyodysenteriae B204 using the protein extraction method described by Gabe et al. (11). The urea-insoluble fraction pelleted by ultracentrifugation was visualized by SDS-PAGE, revealing a single band of approximately 39 kDa (Fig. 3). MALDI-TOF data from the trypsin-digested 39-kDa protein was analyzed by MASCOT, which determined that 11 peptide masses originated from Bhmp39f and Bhmp39h, corresponding to approximately 27% of the Bhmp39f sequence and 24% of the Bhmp39h sequence. Unique peptide fragments indicated the presence of both Bhmp39f and Bhmp39h (Fig. 4). No other significant matches were detected.
DISCUSSION
McCaman et al. (18) noted that the bhmp39 genes share a sequence identity of between 56% and 97%, though the encoded Bhmp39 proteins do not share high sequence identity with any proteins in the current version of GenBank. The high level of similarity between these genes and the low G+C content of the B. hyodysenteriae genome presented significant design constraints that limited the range of gene-specific reagents available for the study. For instance, it was not possible to design suitable primers for the specific detection of bhmp39c by real-time RT-PCR due to its high sequence identity with both bhmp39b and bhmp39d.
The presence of inverted repeat sequences downstream of the bhmp39 genes suggests that the bhmp39 genes are expressed as monocistronic transcripts. Examination of the nucleotide sequence downstream of each bhmp39 gene revealed a base-pairing region that was high in G and C residues, with an A/T-rich loop. These regions were flanked upstream by an A-rich region and downstream by a T-rich region. All of these features are characteristic of typical Rho-independent terminators (10). Analysis of the proposed terminator regions using RNAfold (14) predicted that these structures would form long hairpin-loop structures.
The characteristics of the predicted transcriptional terminators of the bhmp39 genes resemble transcriptional attenuators of organisms like Escherichia coli (7), which could feasibly result in multicistronic transcripts. The previous identification of only one Bhmp39 protein suggests that these genes are individually expressed. Thus, it is feasible that each of the inverted repeats functions as a terminator in this organism, despite their lengths. We determined by Northern hybridization analysis that bhmp39e, bhmp39f, and/or bhmp39h results in individual transcripts. The sequence downstream of bhmp39g does not appear to contain an apparent Rho-independent terminator. However, the real-time RT-PCR data showed unequivocally that bhmp39f and bhmp39h were expressed strongly, while no bhmp39g transcript was detected. This finding is consistent with the hypothesis that each of these three genes is individually transcribed. Given that the predicted terminators of genes in the bhmp39abcd locus have G values comparable to those of the genes of the bhmp39efgh locus, we propose that these genes also result in monocistronic transcripts. The GenBank database sequence was found to contain two undefined bases within the bhmp39c terminator region. Interestingly, our sequencing of this region demonstrated that these bases are erroneous and that their subsequent removal resulted in a complete inverted repeat that was one of the strongest transcriptional terminators predicted in this study.
In order to confirm that the bhmp39 genes of the second locus are transcribed individually, Northern blot analysis was performed using riboprobes specific for bhmp39e, bhmp39g, and bhmp39f/bhmp39h. The degree of nucleotide sequence identity between bhmp39f and bhmp39h (97%) prevented the synthesis of probes specific for their individual transcripts. The riboprobes for bhmp39e and bhmp39f/bhmp39h hybridized specifically to single transcripts of approximately 1.1 kb and 1.2 kb, respectively. No bhmp39g transcript was detected, a finding consistent with the real-time RT-PCR result showing that bhmp39g is expressed at a very low level. Larger transcripts or breakdown products of transcripts were not detected by any of the probes, indicating that only single transcripts were produced.
Real-time RT-PCR was utilized to demonstrate that bhmp39f is the most highly expressed of the bhmp39 paralogs, with bhmp39h and bhmp39d transcripts being the next most abundant. Though these genes are found in the same locus, their different relative expression levels are consistent with their individual transcription. Interestingly, the purified 39-kDa membrane protein that was previously N-terminally sequenced by Gabe et al. (12) was found by McCaman et al. to have been a product of bhmp39h (18). It is possible that the N-terminal sequencing of the LysC-digested protein did not have the capacity to sequence a sufficient range of peptides, which would be required to differentiate between Bhmp39f and Bhmp39h, as they share 93% amino acid sequence identity. Furthermore, one of the most unique regions of Bhmp39f occurs in the last 10 amino acids of the C terminus of a 46-amino-acid peptide generated by cleavage with LysC (predicted using PeptideMass), which may have been difficult to sequence by sequential Edman degradation.
In order to confirm that both bhmp39f and bhmp39h can be identified from the urea-insoluble fraction of the membranes of B. hyodysenteriae grown in vitro, we repeated the protein isolation method employed by Gabe et al. (12), after which the purified protein(s) was digested with trypsin and MALDI-TOF data were obtained from the resulting peptides. This method enabled us to detect a wider range of peptides, and we were thus able to detect peptides that were unique for Bhmp39f and for Bhmp39h, as well as peptides common to both, indicating the presence of both proteins in the membrane preparation. Other Bhmp39 proteins, if present, were clearly below the very sensitive detection limits of the mass spectrometry method used.
We have therefore demonstrated that more than one of the bhmp39 genes are transcribed at a high level and have confirmed that proteins with a peptide mass fingerprint corresponding specifically to both Bhmp39f and Bhmp39h can be isolated from the B. hyodysenteriae membrane. We have also shown that the genes of the second locus are expressed as monocistronic transcripts. This work thus provides evidence that each bhmp39 gene is individually regulated, supporting the theory that they are involved in the antigenic variability of this pathogen.
Other pathogenic spirochetes are known to exhibit antigenic variation. The syphilis spirochete T. pallidum contains a family of 12 tpr genes with an upstream locus for expression, known as tprE. The expression locus undergoes recombination events with the silent tpr genes and thus is able to present many different surface antigens (22). The spirochetes causing Lyme disease, B. burgdorferi, and relapsing fever, Borrelia spp., including B. hermsii, recombine entire genes from silent loci on extrachromosomal elements with expression loci, with the occasional generation of gene hybrids (6). Genes up- and downstream of the bhmp39 genes do not closely resemble any of the bhmp39 genes, and there is currently no evidence for an expression locus for the bhmp39 genes in B. hyodysenteriae. However, as we have identified Rho-independent terminators for each bhmp39 gene and shown individual transcripts, we consider it likely that the bhmp39 genes are regulated individually by an as-yet-uncharacterized mechanism. Potential antigenic variation in B. hyodysenteriae should thus be taken into consideration in any vaccine development studies.
ACKNOWLEDGMENTS
We gratefully acknowledge the excellent assistance of Rebekah Henry, Vicki Vallance, and Ian McPherson.
This work was supported by research grants from Australian Pork Ltd. and from the Australian Research Council, Canberra, Australia.
REFERENCES
1. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25:3389-3402.
2. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 2000. Current protocols in molecular biology, vol. 1. John Wiley & Sons, Inc., New York, N.Y.
3. Boucher, D. J., B. Adler, and J. D. Boyce. 2005. The Pasteurella multocida nrfE gene is upregulated during infection and is essential for nitrite reduction but not for virulence. J. Bacteriol. 187:2278-2285.
4. Boyce, J. D., I. Wilkie, M. Harper, M. L. Paustian, V. Kapur, and B. Adler. 2002. Genomic scale analysis of Pasteurella multocida gene expression during growth within the natural chicken host. Infect. Immun. 70:6871-6879.
5. Cullen, P. A., S. A. Coutts, S. J. Cordwell, D. M. Bulach, and B. Adler. 2003. Characterization of a locus encoding four paralogous outer membrane lipoproteins of Brachyspira hyodysenteriae. Microbes Infect. 5:275-283.
6. Cullen, P. A., D. A. Haake, and B. Adler. 2004. Outer membrane proteins of pathogenic spirochetes. FEMS Microbiol. Rev. 28:291-318.
7. d'Aubenton Carafa, Y., E. Brody, and C. Thermes. 1990. Prediction of rho-independent Escherichia coli transcription terminators. A statistical analysis of their RNA stem-loop structures. J. Mol. Biol. 216:835-858.
8. Diego, R., I. Lanza, A. Carvajal, P. Rubio, and P. Carmenes. 1995. Serpulina hyodysenteriae challenge of fattening pigs vaccinated with an adjuvanted bivalent bacterin against swine dysentery. Vaccine 13:663-667.
9. Dunne, M., B. K. al-Ramadi, S. W. Barthold, R. A. Flavell, and E. Fikrig. 1995. Oral vaccination with an attenuated Salmonella typhimurium strain expressing Borrelia burgdorferi OspA prevents murine Lyme borreliosis. Infect. Immun. 63:1611-1614.
10. Farnham, P. J., and T. Platt. 1981. Rho-independent termination: dyad symmetry in DNA causes RNA polymerase to pause during transcription in vitro. Nucleic Acids Res. 9:563-577.
11. Gabe, J. D., R.-J. Chang, R. Slomiany, W. H. Andrews, and M. T. McCaman. 1995. Isolation of extracytoplasmic proteins from Serpulina hyodysenteriae B204 and molecular cloning of the flaB1 gene encoding a 38-kilodalton flagellar protein. Infect. Immun. 63:142-148.
12. Gabe, J. D., E. Dragon, R.-J. Chang, and M. T. McCaman. 1998. Identification of a linked set of genes in Serpulina hyodysenteriae (B204) predicted to encode closely related 39-kilodalton extracytoplasmic proteins. J. Bacteriol. 180:444-448.
13. Hampson, D. J., T. La, B. Adler, and D. J. Trott. 2006. Proposed revisions to the nomenclature for Brachyspira membrane proteins and lipoproteins. Microbiology 152:1-2.
14. Hofacker, I. L. 2003. Vienna RNA secondary structure server. Nucleic Acids Res. 31:3429-3431.
15. Hudson, M. J., T. J. Alexander, R. J. Lysons, and J. F. Prescott. 1976. Swine dysentery: protection of pigs by oral and parenteral immunisation with attenuated Treponema hyodysenteriae. Res. Vet. Sci. 21:366-367.
16. Hudson, M. J., T. J. Alexander, R. J. Lysons, and P. D. Wellstead. 1974. Swine dysentery: failure of an attenuated strain of spirochaete, given orally, to protect pigs against subsequent challenge. Br. Vet. J. 130:xxxvii-xl.
17. Kang, H. Y., and R. Curtiss III. 2003. Immune responses dependent on antigen location in recombinant attenuated Salmonella typhimurium vaccines following oral immunization. FEMS Immunol. Med. Microbiol. 37:99-104.
18. McCaman, M. T., K. Auer, W. Foley, and J. D. Gabe. 2003. Brachyspira hyodysenteriae contains eight linked gene copies related to an expressed 39-kDa surface protein. Microbes Infect. 5:1-6.
19. McCaman, M. T., K. Auer, W. Foley, and J. D. Gabe. 1999. Sequence characterization of two new members of a multi-gene family in Serpulina hyodysenteriae (B204) with homology to a 39 kDa surface exposed protein: vspC and D. Vet. Microbiol. 68:273-283.
20. Perkins, D. N., D. J. Pappin, D. M. Creasy, and J. S. Cottrell. 1999. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 20:3551-3567.
21. Stanton, T. B., M. G. Thompson, S. B. Humphrey, and R. L. Zuerner. 2003. Detection of bacteriophage VSH-1 svp38 gene in Brachyspira spirochetes. FEMS Microbiol. Lett. 224:225-229.
22. Sun, E. S., B. J. Molini, L. K. Barrett, A. Centurion-Lara, S. A. Lukehart, and W. C. Van Voorhis. 2004. Subfamily I Treponema pallidum repeat protein family: sequence variation and immunity. Microbes Infect. 6:725-737.
23. Wannemuehler, M. J., R. D. Hubbard, and J. M. Greer. 1988. Characterization of the major outer membrane antigens of Treponema hyodysenteriae. Infect. Immun. 56:3032-3039.
24. Wilkins, M. R., I. Lindskog, E. Gasteiger, A. Bairoch, J. C. Sanchez, D. F. Hochstrasser, and R. D. Appel. 1997. Detailed peptide characterization using PEPTIDEMASS—a World-Wide-Web-accessible tool. Electrophoresis 18:403-408.(Differential Expression o)
ABSTRACT
The enteric, anaerobic spirochete Brachyspira hyodysenteriae is the causative agent of swine dysentery, a severe mucohemorrhagic diarrheal disease of pigs that has economic significance in every major pork-producing country. Recent investigation into potential vaccine candidates has focused on the outer membrane proteins of B. hyodysenteriae. Bhmp39 (formerly Vsp39) is the most abundant surface-exposed outer membrane protein of B. hyodysenteriae; its predicted gene sequence has previously been shown to share sequence similarity to eight genes divided evenly between two paralogous loci. The peptide sequence suggested that Bhmp39 is encoded by one of these genes, bhmp39h. The biological significance of maintaining eight homologous bhmp39 genes is unclear, though it has been proposed that this may play a role in antigenic variation. In this study, real-time, reverse transcription-PCR was used to demonstrate that bhmp39f and bhmp39h were the transcripts most abundantly expressed by B. hyodysenteriae strain B204 cultured under in vitro growth conditions. Mass spectrometry data of the purified 39-kDa membrane protein showed that both Bhmp39f and Bhmp39h were present. Northern blot analysis across predicted Rho-independent terminators demonstrated that the genes of the bhmp39efgh locus result in monocistronic transcripts.
INTRODUCTION
The intestinal spirochete Brachyspira hyodysenteriae is the causative agent of swine dysentery, a severe mucohemorrhagic diarrheal disease of pigs. Previous vaccination trials using killed or attenuated live cells as vaccines have not stimulated immunoprotection against swine dysentery (15, 16), while commercially available bacterin vaccines fail to provide complete protection (8). An alternative approach may be to generate subunit vaccines that might be delivered by the expression of recombinant protein on the outer membrane of a bacterial delivery vector, such as Salmonella enterica serovar Typhimurium (9, 17).
Recent investigations into potential targets for novel vaccines against swine dysentery have focused on the outer membrane proteins of B. hyodysenteriae, though much is still unknown about these proteins. Hampson et al. have recently proposed that a new nomenclature be introduced for membrane proteins and lipoproteins of B. hyodysenteriae (13); we have used that nomenclature in this paper. It has been suggested that B. hyodysenteriae evades the host immune system through antigenic variation of the expression of tandemly arranged paralogous genes (6). Known examples are the bhlp29.7abcd (previously blpGFEA) locus (5) and the two bhmp39 (previously vsp) loci, containing bhmp39a to bhmp39d and bhmp39e to bhmp39h (12, 18, 19). Variations in the expression of key outer membrane antigens may be a factor in the failure of vaccines to provide full immunoprotection against swine dysentery. Mechanisms of antigenic variation have been characterized for other pathogenic spirochetes, including Treponema pallidum (22), Borrelia burgdorferi, and Borrelia hermsii (6), though such mechanisms have not yet been investigated in B. hyodysenteriae.
Bhmp39 is the most abundant outer membrane protein of B. hyodysenteriae and is known to be surface exposed based on surface iodination studies (12). Eight paralogous bhmp39 genes have been identified, and N-terminal sequencing of the 39-kDa protein indicated that the expressed protein was encoded by bhmp39h (18). In this study we sought to determine the transcriptional organization of the bhmp39 genes under in vitro conditions and to determine which protein(s) is expressed by the bhmp39 genes.
MATERIALS AND METHODS
Bacterial strains. B. hyodysenteriae strain B204 was kindly provided by P. Coloe, Royal Melbourne Institute of Technology, Melbourne, Australia. Cells were grown anaerobically (80% N2, 10% CO2, 10% H2) at 37°C to between 5 x 107 cells/ml and 2 x 108 cells/ml in Trypticase soya broth supplemented with 0.5% yeast extract, VPI salts, and 5% fetal calf serum (23). Broth cultures were subjected to continuous agitation with magnetic stirrers.
Oligonucleotides. Oligonucleotides used in this study are listed in Table 1. Gene-specific nucleotide sequences were identified by DNA sequence alignments of the bhmp39 gene sequences in GenBank (accession numbers AF012102 and AY027775), and where primers were to be used in real-time reverse transcription-PCR (RT-PCR), Primer Express 1.5a software (Applied Biosystems, Inc., Foster City, CA) was utilized in the design process.
Molecular biology methods. Chromosomal DNA was prepared from B. hyodysenteriae using the cetyltrimethylammonium bromide method (2). Standard methods in molecular biology were performed as described by Ausubel et al. (2). Annealing steps in PCR cycling were performed at 47°C to compensate for the low G+C content of B. hyodysenteriae genomic DNA. When PCR products were required for sequencing, they were purified with a QIAquick PCR purification kit (QIAGEN, Hilden, Germany). Nucleotide sequencing was performed with a BigDye Terminator v3.1 cycle sequencing kit (Perkin-Elmer Corp., Norwalk, CT) and an Applied Biosystems Inc. model 3730 automated sequencer.
RNA isolation and purification. RNA was isolated from B. hyodysenteriae B204 cultures that had cell densities between 5 x 107 and 2 x 108 cells/ml with TRIzol (Invitrogen, La Jolla, CA) at 4°C, according to the manufacturer's instructions. All work using RNA was performed in the presence of RNasin (Promega, Madison, WI). Purified RNA was treated with 20 U of DNase (Roche Molecular Biochemicals, Basel, Switzerland) for 20 min at 37°C, and the RNA was further purified by using RNeasy columns (QIAGEN).
Northern blotting. Northern blotting was carried out as described by Cullen et al. (5). Oligonucleotides containing the T7 promoter sequence were designed to amplify 300- to 321-bp templates for the synthesis of riboprobes complementary to portions of bhmp39e, bhmp39g, and bhmp39f/bhmp39h (single probe for two genes) (Table 1). A riboprobe was generated that hybridized to the transcripts of both bhmp39f and bhmp39h due to the high sequence identity between the two genes.
Reverse transcription. RT reactions were performed as described by Boucher et al. (3). Negative-control reactions were performed by omitting the reverse transcriptase. The cDNA templates were generated by incubating the reaction mixes at 42°C for 2.5 h, followed by denaturation at 70°C for 15 min. cDNA samples were diluted 1:150 and stored at –20°C until required.
Real-time RT-PCR. Real-time RT-PCR was performed using an ABI PRISM model 7700 sequence detector, with product accumulation quantified by the incorporation of the fluorescent dye SYBR green. Primers for use in real-time RT-PCR were designed to amplify 86 to 162 bp of target bhmp39 genes. Reactions were performed in triplicate, with each 20-μl reaction mix containing 2.4 μl of the DNA template, 1.6 μl of primer (625 mM), and 10 μl of SYBR green 2x PCR master mix (Applied Biosystems, Inc.), in a total volume of 20 μl. Gene-specific standard curves were constructed from 20-fold dilutions of B. hyodysenteriae B204 genomic DNA and were used to determine relative cDNA template concentrations in each reaction mixture. Gene expression levels were normalized against gyrB (Table 1). All primer pairs for use in real-time RT-PCR were confirmed to amplify single gene products by melt curve analysis and by sequencing PCR products. Relative levels of gene expression were averaged from the three technical replicates for each of three biological replicates and compared using Welch-corrected, unpaired t tests.
Analysis of Bhmp39 proteins in the B. hyodysenteriae membrane. Urea-insoluble membrane proteins were purified from B. hyodysenteriae B204 by solubilization with 1% Tween 20 and ultracentrifugation, as described by Gabe et al. (11). Sample preparation and matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry were carried out as described previously (5). A list of monoisotopic peaks consisting of the bhmp39 gene products was used to search both local and online protein databases. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of protein samples was carried out using a Mini-PROTEAN 3 electrophoresis apparatus (Bio-Rad Laboratories, Hercules, CA) as described previously (5).
Bioinformatics analysis. BLAST programs (1) were used to perform similarity searches of the GenBank nucleotide and protein sequence databases, as appropriate. DNA sequences were aligned using the ClustalW algorithm (21). Transcript sizes from Northern hybridization were predicted using the Sequaid II software (University of Kansas, Kansas City, KS). Peptide masses from protease digestion of proteins were determined using PeptideMass (24). MASCOT (20) was used to identify proteins from peptide mass data by searching the "Other Bacteria" subset of the NCBI nonredundant database. The algorithm was instructed to exclude peptides with missed cleavages, and the peptide tolerance was set to ±0.2 Da. A protein score greater than 63 was considered to be significant (P < 0.05) when data were analyzed with MASCOT. Statistical analyses were conducted using the InStat program (Graphpad Software, Inc., San Diego, CA). The strength of transcriptional termination of Rho-independent terminators was predicted using RNAfold (14).
RESULTS
Prediction of Rho-independent transcriptional terminators. Inspection of the sequence downstream of each of the bhmp39 coding regions revealed the presence of putative Rho-independent terminators. RNAfold was used to predict the minimum free energy of each inverted repeat and thereby enable a comparison of the strength of each of the putative transcriptional terminators (Table 2). In each instance, the inverted repeat of each putative terminator begins 26 to 43 bp downstream of the stop codon, and the inverted repeats form the stem-loop structure, which is 9 to 14 bp in length. Each putative terminator contained a much higher G+C content (about 4 to 10 bp more) than the flanking nucleotide sequence, which resulted in highly negative G values for all of the predicted terminators, except for that of bhmp39g, with which a high number of mismatches resulted in a predicted free energy that was almost 0. We resequenced the bhmp39c terminator region in order to clarify the undefined bases in the available lodged sequence. The RNA secondary-structure prediction from the revised sequence suggested that it would form an inverted repeat of a length and strength similar to those of the other predicted terminators (Table 2).
Northern hybridization. Northern hybridization analysis was performed to determine the approximate size of the B. hyodysenteriae B204 bhmp39efgh transcripts (Fig. 1). Individual transcripts of the bhmp39abcd locus could not be investigated due to difficulty in designing riboprobes of appropriate lengths and with sufficient specificity to individual genes. The riboprobe specific for bhmp39e identified a single transcript of 1.1 kb (Fig. 1a), but no transcript was detected for bhmp39g (Fig. 1b). The bhmp39f/bhmp39h riboprobe hybridized to a product of approximately 1.2 kb (Fig. 1c).
Real-time RT-PCR quantification of bhmp39 mRNA. In order to ascertain which bhmp39 genes were expressed by cells grown under in vitro conditions, the transcription of seven of the eight bhmp39 genes of B. hyodysenteriae strain B204 was investigated by real-time RT-PCR (Fig. 2). Primers used for real-time RT-PCR were designed specifically to amplify an internal region of each of the bhmp39 genes from B. hyodysenteriae strain B204; this was not possible for bhmp39c because of its very high nucleotide sequence identity to bhmp39b and bhmp39d. The expression of gyrB was used as an internal control, as described previously (4), in order to compare the relative levels of expression of bhmp39 genes from different biological replicates. Averaged data from three technical replicates of three biological replicates indicated that bhmp39f was the most highly transcribed bhmp39 gene, followed by bhmp39h (Fig. 2).
Identification of the 39-kDa urea-insoluble membrane protein. Having observed that more than one bhmp39 gene was transcribed under the culture conditions used in this study, we set out to determine the identity of the 39-kDa outer membrane protein. We purified membrane proteins from B. hyodysenteriae B204 using the protein extraction method described by Gabe et al. (11). The urea-insoluble fraction pelleted by ultracentrifugation was visualized by SDS-PAGE, revealing a single band of approximately 39 kDa (Fig. 3). MALDI-TOF data from the trypsin-digested 39-kDa protein was analyzed by MASCOT, which determined that 11 peptide masses originated from Bhmp39f and Bhmp39h, corresponding to approximately 27% of the Bhmp39f sequence and 24% of the Bhmp39h sequence. Unique peptide fragments indicated the presence of both Bhmp39f and Bhmp39h (Fig. 4). No other significant matches were detected.
DISCUSSION
McCaman et al. (18) noted that the bhmp39 genes share a sequence identity of between 56% and 97%, though the encoded Bhmp39 proteins do not share high sequence identity with any proteins in the current version of GenBank. The high level of similarity between these genes and the low G+C content of the B. hyodysenteriae genome presented significant design constraints that limited the range of gene-specific reagents available for the study. For instance, it was not possible to design suitable primers for the specific detection of bhmp39c by real-time RT-PCR due to its high sequence identity with both bhmp39b and bhmp39d.
The presence of inverted repeat sequences downstream of the bhmp39 genes suggests that the bhmp39 genes are expressed as monocistronic transcripts. Examination of the nucleotide sequence downstream of each bhmp39 gene revealed a base-pairing region that was high in G and C residues, with an A/T-rich loop. These regions were flanked upstream by an A-rich region and downstream by a T-rich region. All of these features are characteristic of typical Rho-independent terminators (10). Analysis of the proposed terminator regions using RNAfold (14) predicted that these structures would form long hairpin-loop structures.
The characteristics of the predicted transcriptional terminators of the bhmp39 genes resemble transcriptional attenuators of organisms like Escherichia coli (7), which could feasibly result in multicistronic transcripts. The previous identification of only one Bhmp39 protein suggests that these genes are individually expressed. Thus, it is feasible that each of the inverted repeats functions as a terminator in this organism, despite their lengths. We determined by Northern hybridization analysis that bhmp39e, bhmp39f, and/or bhmp39h results in individual transcripts. The sequence downstream of bhmp39g does not appear to contain an apparent Rho-independent terminator. However, the real-time RT-PCR data showed unequivocally that bhmp39f and bhmp39h were expressed strongly, while no bhmp39g transcript was detected. This finding is consistent with the hypothesis that each of these three genes is individually transcribed. Given that the predicted terminators of genes in the bhmp39abcd locus have G values comparable to those of the genes of the bhmp39efgh locus, we propose that these genes also result in monocistronic transcripts. The GenBank database sequence was found to contain two undefined bases within the bhmp39c terminator region. Interestingly, our sequencing of this region demonstrated that these bases are erroneous and that their subsequent removal resulted in a complete inverted repeat that was one of the strongest transcriptional terminators predicted in this study.
In order to confirm that the bhmp39 genes of the second locus are transcribed individually, Northern blot analysis was performed using riboprobes specific for bhmp39e, bhmp39g, and bhmp39f/bhmp39h. The degree of nucleotide sequence identity between bhmp39f and bhmp39h (97%) prevented the synthesis of probes specific for their individual transcripts. The riboprobes for bhmp39e and bhmp39f/bhmp39h hybridized specifically to single transcripts of approximately 1.1 kb and 1.2 kb, respectively. No bhmp39g transcript was detected, a finding consistent with the real-time RT-PCR result showing that bhmp39g is expressed at a very low level. Larger transcripts or breakdown products of transcripts were not detected by any of the probes, indicating that only single transcripts were produced.
Real-time RT-PCR was utilized to demonstrate that bhmp39f is the most highly expressed of the bhmp39 paralogs, with bhmp39h and bhmp39d transcripts being the next most abundant. Though these genes are found in the same locus, their different relative expression levels are consistent with their individual transcription. Interestingly, the purified 39-kDa membrane protein that was previously N-terminally sequenced by Gabe et al. (12) was found by McCaman et al. to have been a product of bhmp39h (18). It is possible that the N-terminal sequencing of the LysC-digested protein did not have the capacity to sequence a sufficient range of peptides, which would be required to differentiate between Bhmp39f and Bhmp39h, as they share 93% amino acid sequence identity. Furthermore, one of the most unique regions of Bhmp39f occurs in the last 10 amino acids of the C terminus of a 46-amino-acid peptide generated by cleavage with LysC (predicted using PeptideMass), which may have been difficult to sequence by sequential Edman degradation.
In order to confirm that both bhmp39f and bhmp39h can be identified from the urea-insoluble fraction of the membranes of B. hyodysenteriae grown in vitro, we repeated the protein isolation method employed by Gabe et al. (12), after which the purified protein(s) was digested with trypsin and MALDI-TOF data were obtained from the resulting peptides. This method enabled us to detect a wider range of peptides, and we were thus able to detect peptides that were unique for Bhmp39f and for Bhmp39h, as well as peptides common to both, indicating the presence of both proteins in the membrane preparation. Other Bhmp39 proteins, if present, were clearly below the very sensitive detection limits of the mass spectrometry method used.
We have therefore demonstrated that more than one of the bhmp39 genes are transcribed at a high level and have confirmed that proteins with a peptide mass fingerprint corresponding specifically to both Bhmp39f and Bhmp39h can be isolated from the B. hyodysenteriae membrane. We have also shown that the genes of the second locus are expressed as monocistronic transcripts. This work thus provides evidence that each bhmp39 gene is individually regulated, supporting the theory that they are involved in the antigenic variability of this pathogen.
Other pathogenic spirochetes are known to exhibit antigenic variation. The syphilis spirochete T. pallidum contains a family of 12 tpr genes with an upstream locus for expression, known as tprE. The expression locus undergoes recombination events with the silent tpr genes and thus is able to present many different surface antigens (22). The spirochetes causing Lyme disease, B. burgdorferi, and relapsing fever, Borrelia spp., including B. hermsii, recombine entire genes from silent loci on extrachromosomal elements with expression loci, with the occasional generation of gene hybrids (6). Genes up- and downstream of the bhmp39 genes do not closely resemble any of the bhmp39 genes, and there is currently no evidence for an expression locus for the bhmp39 genes in B. hyodysenteriae. However, as we have identified Rho-independent terminators for each bhmp39 gene and shown individual transcripts, we consider it likely that the bhmp39 genes are regulated individually by an as-yet-uncharacterized mechanism. Potential antigenic variation in B. hyodysenteriae should thus be taken into consideration in any vaccine development studies.
ACKNOWLEDGMENTS
We gratefully acknowledge the excellent assistance of Rebekah Henry, Vicki Vallance, and Ian McPherson.
This work was supported by research grants from Australian Pork Ltd. and from the Australian Research Council, Canberra, Australia.
REFERENCES
1. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25:3389-3402.
2. Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 2000. Current protocols in molecular biology, vol. 1. John Wiley & Sons, Inc., New York, N.Y.
3. Boucher, D. J., B. Adler, and J. D. Boyce. 2005. The Pasteurella multocida nrfE gene is upregulated during infection and is essential for nitrite reduction but not for virulence. J. Bacteriol. 187:2278-2285.
4. Boyce, J. D., I. Wilkie, M. Harper, M. L. Paustian, V. Kapur, and B. Adler. 2002. Genomic scale analysis of Pasteurella multocida gene expression during growth within the natural chicken host. Infect. Immun. 70:6871-6879.
5. Cullen, P. A., S. A. Coutts, S. J. Cordwell, D. M. Bulach, and B. Adler. 2003. Characterization of a locus encoding four paralogous outer membrane lipoproteins of Brachyspira hyodysenteriae. Microbes Infect. 5:275-283.
6. Cullen, P. A., D. A. Haake, and B. Adler. 2004. Outer membrane proteins of pathogenic spirochetes. FEMS Microbiol. Rev. 28:291-318.
7. d'Aubenton Carafa, Y., E. Brody, and C. Thermes. 1990. Prediction of rho-independent Escherichia coli transcription terminators. A statistical analysis of their RNA stem-loop structures. J. Mol. Biol. 216:835-858.
8. Diego, R., I. Lanza, A. Carvajal, P. Rubio, and P. Carmenes. 1995. Serpulina hyodysenteriae challenge of fattening pigs vaccinated with an adjuvanted bivalent bacterin against swine dysentery. Vaccine 13:663-667.
9. Dunne, M., B. K. al-Ramadi, S. W. Barthold, R. A. Flavell, and E. Fikrig. 1995. Oral vaccination with an attenuated Salmonella typhimurium strain expressing Borrelia burgdorferi OspA prevents murine Lyme borreliosis. Infect. Immun. 63:1611-1614.
10. Farnham, P. J., and T. Platt. 1981. Rho-independent termination: dyad symmetry in DNA causes RNA polymerase to pause during transcription in vitro. Nucleic Acids Res. 9:563-577.
11. Gabe, J. D., R.-J. Chang, R. Slomiany, W. H. Andrews, and M. T. McCaman. 1995. Isolation of extracytoplasmic proteins from Serpulina hyodysenteriae B204 and molecular cloning of the flaB1 gene encoding a 38-kilodalton flagellar protein. Infect. Immun. 63:142-148.
12. Gabe, J. D., E. Dragon, R.-J. Chang, and M. T. McCaman. 1998. Identification of a linked set of genes in Serpulina hyodysenteriae (B204) predicted to encode closely related 39-kilodalton extracytoplasmic proteins. J. Bacteriol. 180:444-448.
13. Hampson, D. J., T. La, B. Adler, and D. J. Trott. 2006. Proposed revisions to the nomenclature for Brachyspira membrane proteins and lipoproteins. Microbiology 152:1-2.
14. Hofacker, I. L. 2003. Vienna RNA secondary structure server. Nucleic Acids Res. 31:3429-3431.
15. Hudson, M. J., T. J. Alexander, R. J. Lysons, and J. F. Prescott. 1976. Swine dysentery: protection of pigs by oral and parenteral immunisation with attenuated Treponema hyodysenteriae. Res. Vet. Sci. 21:366-367.
16. Hudson, M. J., T. J. Alexander, R. J. Lysons, and P. D. Wellstead. 1974. Swine dysentery: failure of an attenuated strain of spirochaete, given orally, to protect pigs against subsequent challenge. Br. Vet. J. 130:xxxvii-xl.
17. Kang, H. Y., and R. Curtiss III. 2003. Immune responses dependent on antigen location in recombinant attenuated Salmonella typhimurium vaccines following oral immunization. FEMS Immunol. Med. Microbiol. 37:99-104.
18. McCaman, M. T., K. Auer, W. Foley, and J. D. Gabe. 2003. Brachyspira hyodysenteriae contains eight linked gene copies related to an expressed 39-kDa surface protein. Microbes Infect. 5:1-6.
19. McCaman, M. T., K. Auer, W. Foley, and J. D. Gabe. 1999. Sequence characterization of two new members of a multi-gene family in Serpulina hyodysenteriae (B204) with homology to a 39 kDa surface exposed protein: vspC and D. Vet. Microbiol. 68:273-283.
20. Perkins, D. N., D. J. Pappin, D. M. Creasy, and J. S. Cottrell. 1999. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 20:3551-3567.
21. Stanton, T. B., M. G. Thompson, S. B. Humphrey, and R. L. Zuerner. 2003. Detection of bacteriophage VSH-1 svp38 gene in Brachyspira spirochetes. FEMS Microbiol. Lett. 224:225-229.
22. Sun, E. S., B. J. Molini, L. K. Barrett, A. Centurion-Lara, S. A. Lukehart, and W. C. Van Voorhis. 2004. Subfamily I Treponema pallidum repeat protein family: sequence variation and immunity. Microbes Infect. 6:725-737.
23. Wannemuehler, M. J., R. D. Hubbard, and J. M. Greer. 1988. Characterization of the major outer membrane antigens of Treponema hyodysenteriae. Infect. Immun. 56:3032-3039.
24. Wilkins, M. R., I. Lindskog, E. Gasteiger, A. Bairoch, J. C. Sanchez, D. F. Hochstrasser, and R. D. Appel. 1997. Detailed peptide characterization using PEPTIDEMASS—a World-Wide-Web-accessible tool. Electrophoresis 18:403-408.(Differential Expression o)