Molecular characterisation of a type III restriction-modification system in Campylobacter upsaliensis

Two examples of Campylobacter upsaliensis RM3195 and JV21 strains are shown to carry putative type III restriction (res)-modification (mod) enzyme gene clusters, following genome sequence analyses. It is suggested that the cluster is composed of at least three structural genes, res, internal methylase gene and mod, in the strains, based on the nucleotide sequence information. A ribosome binding site, a putative promoter consisting of a consensus sequence at the -10-like structure and a semiconserved T-rich region and a putative intrinsic p-independent transcriptional terminator were identified for the gene cluster in the two strains. Using two primer pairs, f-/r-res and f-/r-mod, 34 of 41 C. upsaliensis isolates generated two expected amplicons of the res and mod gene segments, and using another primer pair, the same number of isolates also generated an amplicon of the res and mod gene segments cluster, including the third internal methylase gene. Thus, C. upsaliensis isolates frequently carried putative type III R-M gene clusters, encoding the three enzymes. Interestingly, two possible overlaps were identified within the three tandem structural genes. In addition, the type III R-M gene cluster loci appear to be very similar among the C. upsaliensis isolates and very different from other thermophilic campylobacters.

Construction, expression and characterisation of recombinant molecules of the urease gene operon from a urease-positive thermophilic Campylobacter (UPTC) isolate

A recombinant molecule of the full-length urease gene operon was constructed in vitro from the Japanese urease-positive thermophilic Campylobacter (UPTC) CF89-12 isolate and expressed in Escherichia coli cells. Several large deletion recombinant variants of urease subunit genes were also constructed and expressed in E. coli cells. A positive urease reaction with the log-phase cultured E. coli JM109 cells in the NiCl2-containing medium transformed with pGEM-T vector carrying the recombinant molecule of the full-length operon was detected with isopropyl-beta-D-thiogalactoside. Among the several deletion recombinant variants, each ureA-, ureB-, ureE-, ureF-, ureG- and ureH-large deficient, only ureE-large deletion variant (63% deficient) showed a positive urease reaction (approximately 15-fold). In addition, a ureE-complete deletion recombinant variant (100% deficient) constructed also showed a positive reaction of urease (approximately 18-fold). Recombinant urease subunits A and B were immunologically identified by Western blot analysis with anti-urease alpha (A) and beta (B) raised against Helicobacter pylori.

Expression and analysis of a cytolethal distending toxin (cdt) gene operon in Campylobacter lari

The present study examines the expression of cytolethal distending toxin (cdt) gene encoding a cytotoxin in Campylobacter lari (n=6 urease-negative [UN] C. lari; n=4 urease-positive thermophilic Campylobacter [UPTC]). When reverse transcription polymerase chain reaction (RT-PCR) was carried out with 10 C. lari isolates using a primer pair to amplify the cdtB gene transcript segment, an approximate 260 bp RT-PCR amplicon was generated with all the isolates. In addition, cdtA, cdtB and cdtC gene operon was identified to be polycistronicly transcribed in the C. lari cells. The cdtB gene translation in the C. lari cells was also confirmed by Western blot analysis. Thus, the cdt gene operon in C. lari organisms, including UN C. lari and UPTC, was expressed at the transcriptional and translational levels in the cells. The present results suggest that all three cdt genes may be functional in the cells.

Comparative analysis of Campylobacter lari cytolethal distending toxin (CDT) effect on HeLa cells

We aimed to clarify if Campylobacter lari exerts a cytolethal distending toxin (CDT) effect on HeLa cells. Campylobacter cell lysates (CCLys) from C. jejuni 81-176 and urease-positive thermophilic Campylobacter (UPTC) CF89-12 and UPTC NCTC12893 isolates were shown to exert a CDT effect on HeLa cells with morphological changes examined by Giemsa staining and microscopy. However, Campylobacter lari JCM2530(T) isolate showed no effect. In addition, Campylobacter cell culture supernatant wash gave low or absent toxic effects with both C. jejuni and C. lari organisms. When western blot analysis was carried out to clarify if there was a CDTB effect in the CCLys and soluble fractions from Campylobacter isolates, which had a CDT effect on HeLa cells or did not have any effect, anti-recombinant CjCDTB antibodies identified an immunoreactively positive signal at around approximately 25 kDa on all the C. lari isolates examined, as well as the C. jejuni 81116 strain. Thus, all the Campylobacter isolates including those without any CDT effect were shown to express CDTB at the translational level.

Reliability of a multiplex PCR assay for the identification of the major Campylobacter taxa

The primer pair (C412F/C1228R) constructed previously for the polymerase chain reaction (PCR) identification of the genus Campylobacter using an approximate 800 base pair (bp) 16S rRNA gene target segment proved to be useful for the identification of a total of 49 Campylobacter lari isolates including urease-positive thermophilic Campylobacter (UPTC) organisms (n=25). When the primer pair (CLF/R) developed previously for the PCR identification of C. lari species using an approximate 250 bp glyA segment was employed, 27 C. lari isolates, including all the UPTC isolates, were identified to be PCR-negative (55%). Therefore, this PCR procedure developed for the molecular identification of C. lari was shown to be unreliable for C. lari identification. Nucleotide sequencing analysis clarified the reason(s) why PCR-negative examples occurred in many C. lari isolates, including UPTC isolates. The primer pair target sequences in the C. lari-specific PCR-negative isolates apparently varied at the 3' end region, as compared with C. lari-specific PCR-positive isolates. Thus, the multiplex PCR assay developed previously was shown to be unreliable for the molecular identification of C. lari subspecies organisms.

Molecular identification and characterization of clustered regularly interspaced short palindromic repeats (CRISPRs) in a urease-positive thermophilic Campylobacter sp. (UPTC)

Novel clustered regularly-interspaced short palindromic repeats (CRISPRs) locus [7,500 base pairs (bp) in length] occurred in the urease-positive thermophilic Campylobacter (UPTC) Japanese isolate, CF89-12. The 7,500 bp gene loci consisted of the 5'-methylaminomethyl-2-thiouridylate methyltransferase gene, putative (P) CRISPR associated (p-Cas), putative open reading frames, Cas1 and Cas2, leader sequence region (146 bp), 12 CRISPRs consensus sequence repeats (each 36 bp) separated by a non-repetitive unique spacer region of similar length (26-31 bp) and the phosphatidyl glycerophosphatase A gene. When the CRISPRs loci in the UPTC CF89-12 and five C. jejuni isolates were compared with one another, these six isolates contained p-Cas, Cas1 and Cas2 within the loci. Four to 12 CRISPRs consensus sequence repeats separated by a non-repetitive unique spacer region occurred in six isolates and the nucleotide sequences of those repeats gave approximately 92-100% similarity with each other. However, no sequence similarity occurred in the unique spacer regions among these isolates. The putative σ(70) transcriptional promoter and the hypothetical ρ-independent terminator structures for the CRISPRs and Cas were detected. No in vivo transcription of p-Cas, Cas1 and Cas2 was confirmed in the UPTC cells.

Molecular and comparative analyses of the full-length cytolethal distending toxin (cdt) gene operon and its adjacent genetic loci from urease-positive thermophilic Campylobacter (UPTC) organisms

Molecular and comparative analyses of the full-length cytolethal distending toxin (cdt) gene operon and its adjacent genetic loci (2.7-9.4 kilo base pairs in length) are carried out with 12 urease-positive thermophilic Campylobacter (UPTC) isolates using several polymerase chain reaction (PCR) primer pairs. Three putative open reading frames (ORFs) for cdtA, cdtB and cdtC, two putative promoters and a hypothetically intrinsic rho-independent transcription terminator were identified in all the operons of the 12 UPTC isolates examined. Although the number of amino acid residues slightly varied for the putative cdtA and cdtC ORFs, those for the cdtB were similar among all the UPTC isolates, as well as the six urease-negative (UN) C. lari examined previously. Regarding the cdt genes in UPTC CF89-12, each ORF commenced with an ATG start codon and terminated with a TAG stop codon for cdtA and cdtB and a TAA for cdtC. Start and stop codons of the three ORFs for the other 11 UPTC isolates were identical to those from the UPTC CF89-12 isolate except for the TTG start codon for cdtC in the two isolates (NCTC12892 and 12893) and the TGA stop codon for cdtA in five isolates (A1, A2, A3, 89049 and 92251). Two putative promoter structures, consisting of sequences at the -35-like (TTAATA) and -10-like (TATTAA) regions, as well as the start codon (ATG), were identified for the transcriptional promoter, immediately upstream of the cdtA gene in all the 12 isolates, Although the genetic heterogeneity of the cdtB gene locus occurred in all 28 C. lari isolates (n = 16 UN C. lari; n = 12 UPTC) examined, all nine amino acid-specific DNase residues were completely conserved in all their cdtB genes. Variable gene insertions with heterogeneous order and combinations occurred between cdtC and lpxB genes in the all UPTC organisms examined.

Uneven distribution of the luxS gene within the genus Campylobacter

Polymerase chain reaction (PCR) amplification was performed on 20 isolates of five Campylobacter species using a degenerate primer pair designed in silico to generate a product of the luxS gene or its homologue from Campylobacter organisms. Although the primer pair successfully amplified products of approximately 500 base pairs (bp) with the eight isolates of C. jejuni and C. coli and some of C. upsaliensis and C. fetus, it failed to amplify fragments with all four isolates of C. lari (two urease-negative C. lari; two urease-positive thermophilic campylobacters). When Southern blot hybridisation analysis was carried using the mixed luxS gene fragments prepared from the C. jejuni, C. coli, C. upsaliensis and C. fetus strains as a probe, all C. jejuni, C. coli, C. upsaliensis and C. fetus isolates gave positive signals, but no positive signal was detected with any C. lari isolate. These results clearly indicate that C. jejuni, C. coli, C. upsaliensis and C. fetus carry the luxS gene or its homologue. However, no luxS gene or its homologue was identified to occur in the C. lari genome. Although autoinducer-2 assays were positive in C. jejuni, C. coli, C. upsaliensis and C. fetus isolates, it was negative with all the C. lari isolates examined. In addition, a biofilm formation assay demonstrated that biofilm formation in the C. lari species does not appear to correlate with the occurrence of the luxS gene because biofilm formation occurred among some isolates of C. lari.

Campylobacter lari: molecular and comparative analyses of the virulence-associated chromosome locus J (vacJ) gene homologue, including the promoter region

Following TA cloning and sequencing with a novel in silico-designed polymerase chain reaction (PCR) primer pair (f-ClvacJ/r-ClvacJ), approximately 750 base pairs (bp) of promoter and structural gene regions for vacJ and its adjacent genetic loci (approximately 1.14 kbp) were identified in 20 isolates of Campylobacter lari (urease-negative C. lari [n=7]; urease-positive thermophilic Campylobacter [n=13]). The nucleotide sequences of an approximately 70-bp non-coding region, including the typical promoter structure, showed sequence differences at 12 loci among 21 isolates including C. lari RM2100. The putative sigma70 promoter region upstream of the putative open reading frame (ORF), a start codon TTG and a probable ribosome binding site, AGGA, for the vacJ gene were also identified in all 21 C. lari isolates examined. Each ORF for the vacJ terminated with a TAA stop codon. No hypothetical transcriptional terminators were identified within the amplicons. The putative ORFs of the vacJ gene from 21 C. lari isolates consisted of 684 bases, similarly differing from those of the other thermophilic campylobacters (696 bases for C. jejuni RM1221 and NCTC11168 and C. coli RM2228; 690 for C. upsaliensis RM3195). Reverse transcription PCR analysis confirmed the transcription of the vacJ gene in the C. lari cells. A neighbour joining tree suggested a strong molecular discrimination efficacy between UPTC and UN C. lari employing vacJ nucleotide sequence information. The vacJ gene homologue from C. lari organisms appears not to be a lipoprotein signal peptide or a signal peptide in silico.

Molecular analysis and characterisation of the full-length flagellin C gene (flaC) from Campylobacter lari

A degenerate polymerase chain reaction (PCR) primer pair (f-ClflaC/r-ClflaC) was constructed in silico to amplify flaC and its adjacent genetic loci from Campylobacter lari isolates. Approximately 1.45 kbp amplicons, including the sequences encoding the flaC structural gene of 750 bp, putative promoter, rho-independent intrinsic terminator regions and partial sequences of two putative open reading frames (ORFs), immediately upstream and downstream of the gene, were identified in 16 C. lari isolates (four urease-negative [UN] C. lari; 12 urease-positive thermophilic campylobacters [UPTC)]). All 16 flaC structural genes commenced with an ATG start codon and terminated with a TAA stop codon and probable ribosome-binding sites were identified in all 16 isolates. These probably indicate a monocistronic operon structure for the flaC gene in C. lari isolates. In addition, the putative flaC gene ORFs were deduced to be similar in 747 bp among all 26 thermophilic Campylobacter isolates examined, resulting in a similar calculated molecular weight of approximately 26.6-26.9 kDa. The flaC from C. lari was different from the flaA-like sequence and the shorter flaA of UPTC isolates found previously. Reverse transcription PCR and Northern blot hybridisation analyses identified flaC transcription in C. lari cells. The transcription initiation site for the flaC gene was also determined by primer extension analysis. A dendrogram constructed, based on the nucleotide sequence information of flaC from 17 C. lari isolates, demonstrated that the C. lari isolates were genetically variable and formed two minor clusters for UN C. lari and UPTC.

Molecular analysis and characterization of a urease gene operon from Campylobacter sputorum biovar paraureolyticus

When recombinant plasmid DNA from a genomic DNA library and inverse PCR products of Campylobacter sputorum biovar paraureolyticus LMG17591 strain were analyzed, an approximate 6.5-kb pair region, encoding a urease gene operon, was identified. Within the operon, seven closely spaced and putative open reading frames for ureG, ureH(D), ureA, ureB, ureC, ureE, and ureF were detected in order. A possible overlap was detected between ureG and ureH(D), ureH(D) and ureA, and ureE and ureF. In addition, two putative promoter structures, probable ribosome-binding sites and a putative ρ-independent transcriptional terminator structure were identified. The urease gene operon transcription in the cells was confirmed by the reverse transcription-PCR analysis. A neighbor-joining tree constructed based on the nucleotide sequence information of urease genes showed that C. sputorum biovar paraureolyticus formed a cluster with Arcobacter butzleri, urease-positive thermophilic Campylobacter and some Helicobacter spp., separating those from the other urease-producing bacteria, suggesting a commonly shared ancestry among these organisms.

Structural analysis and expression of the full-length cytochrome P450 gene operon in Campylobacter lari

Two sets of PCR primers are constructed to clone the cytochrome P450 structural gene, including putative promoter and terminator structures, and its adjacent genetic loci in Campylobacter lari isolates. The putative open reading frames (ORFs) of the P450 genes from 11 C. lari isolates (n=5 for urease-negative (UN) C. lari; n=6 urease-positive thermophilic campylobacters [UPTC]) examined consisted of 1365 or 1371 bases (455 or 457 amino acid residues), differing from those of the other thermophilic campylobacters (1359 [453] for C. jejuni and C. upsaliensis; 1368 [456] for C. coli). Each of the putative ORFs from the 11 isolates examined was also shown to carry start and stop codons and ribosome binding sites. Two putative promoter structures, consisting of sequences at the -35- and -10-like regions were also identified upstream of the ORFs. A single copy of the P450 gene in the genome was identified with UN C. lari JCM2530(T) and UPTC CF89-12, based on Southern blot hybridisation analysis. In addition, when reverse transcription polymerase chain reaction (RT-PCR) analyses were carried out, the transcription of the P450 structural gene in C. lari organisms in vivo was confirmed. The transcription initiation site for the gene was also determined. High nucleotide sequence similarities (95.2-98.8%) of the full-length P450 structural gene were shown with each of the 12 C. lari isolates. The UN C. lari and UPTC organisms showed similar findings with the neighbour-joining method, based on the sequence information of the P450 structural gene.

Cloning and structural analysis of the full-length cytolethal distending toxin (cdt) gene operon from Campylobacter lari

Polymerase chain reaction (PCR) amplicons (approximately 2.5 kbp) encoding a cdt gene operon and two partial and putative open reading frames (ORFs) were identified in six urease-negative (UN) Campylobacter lari isolates using a new PCR primer pair constructed in silico. Three closely spaced and putative ORFs for cdtA, cdtB and cdtC, two putative promoters and a hypothetically intrinsic p-independent transcription terminator were found in the operon. Each ORF commenced with an ATG start codon and terminated with a TGA stop codon for cdtA and cdtB and a TAA for cdtC. Interestingly, an overlap of four nucleotides was detected between cdtA and cdtB and the non-coding region of six base pairs occurring between cdtB and cdtC. The start codons for the three cdt genes were preceded by Shine-Dalgarno sequences. Although nucleotide sequence differences were identified at seven loci in the cdtA gene, six in cdtB and two in cdtC among the seven isolates (including C. lari RM2100), no polymorphic sites occurred in the putative promoters, hypothetically intrinsic transcription terminator and the three ribosome binding sites among the seven isolates. All nine amino acid residues specific for both Escherichia coli cdtB and mammalian DNase I were completely conserved in the cdtB gene locus in the 26 C. lari isolates, as well as in C. jejuni and C. coli. No PCR amplicons were generated with urease-positive thermophilic campylobacters (UPTC; n=10) using the primer pair.

Molecular epidemiology of clinical isolates of Pseudomonas aeruginosa isolated from horses in Ireland

Clinical isolates (n = 63) of Pseudomonas aeruginosa obtained from various sites in 63 horses were compared using ERIC2 RAPD PCR to determine their genetic relatedness. Resulting banding patterns (n = 24 genotypes) showed a high degree of genetic heterogeneity amongst all isolates examined, indicating a relative non-clonal relationship between isolates from these patients, employing this genotyping technique. This study characterised 63 clinical isolates into 24 distinct genotypes, with the largest cluster (genotype E) accounting for 10/63 (15.9%) of the isolates. ERIC2 RAPD PCR proved to be a highly discriminatory molecular typing tool of P. aeruginosa in isolates recovered from horses. With the adoption of several controls to aid reproducibility, this technique may be useful as an alternative to PFGE, particularly in epidemiological investigations of outbreaks where speed may be a significant parameter. This is the first report of clonal heterogeneity amongst P. aeruginosa from horses and demonstrated that ERIC RAPD PCR is a rapid method for the examination of this species in horses, which may be useful in outbreak analysis.

Molecular characterization of the non-coding promoter and leader regions and full-length 16S ribosomal RNA (rRNA) gene ofTaylorella asinigenitalis

The 3,339 base pair (bp) sequences encoding a putative open reading frame (ORF), non-coding promoter and leader regions (approximately 320 bp), full-length 16S ribosomal RNA (rRNA) gene (approximate 1,540 bp) and part of the 16S-23S rDNA internal spacer region (ISR) were determined from genome DNA libraries of the Taylorella asinigenitalis (UK-1) isolate. The non-coding promoter and leader regions included antiterminators (boxB, boxA and boxC) immediately upstream of the 16S rRNA gene sequence. An approximately 680 bp region upstream of the non-coding promoter region appears to contain a putative ORF with high sequence similarity to GTP cyclohydrolase I. In addition, a typical order of intercistronic tRNA genes with the 48 nucleotide spacer of 5'-16S rDNA-tRNA(Ile)-tRNA(Ala)-23S rDNA-3' was demonstrated in a part of the 16S-23S rDNA ISR. The antiterminators of boxB and boxA were also identified in the ISR.A phylogenetic analysis based on the 16S rRNA gene sequence information clearly demonstrated that the five T. asinigenitalis isolates formed a cluster together with the three T. equigenitalis strains, more similar to Pelistega europaea than the other beta-Proteobacteria from the 13 species of 11 genera.

Nucleotide Sequencing and Analysis of 16S rDNA and 16S-23S rDNA Internal Spacer Region (ISR) of Taylorella equigenitalis, as an Important Pathogen for Contagious Equine Metritis (CEM)

The primer set for 16S rDNA amplified an amplicon of about 1500 bp in length for three strains of Taylorella equigenitalis (NCTC11184(T), Kentucky188 and EQ59). Sequence differences of the 16S rDNA among the six sequences, including three reference sequences, occurred at only a few nucleotide positions and thus, an extremely high sequence similarity of the 16S rDNA was first demonstrated among the six sequences. In addition, the primer set for 16S-23S rDNA internal spacer region (ISR) amplified two amplicons about 1300 bp and 1200 bp in length for the three strains. The ISRs were estimated to be about 920 bp in length for large ISR-A and about 830 bp for small ISR-B. Sequence alignment of the ISR-A and ISR-B demonstrated about 10 base differences between NCTC11184(T) and EQ59 and between Kentucky188 and EQ59. However, only minor sequence differences were demonstrated between the ISR-A and ISR-B from NCTC11184(T) and Kentucky188, respectively. A typical order of the intercistronic tRNAs with the 29 nucleotide spacer of 5'-16S rDNA-tRNA(Ile)-tRNA(Ala)-23S rDNA-3' was demonstrated in the all ISRs. The ISRs may be useful for the discrimination amongst isolates of T. equigenitalis if sequencing is employed.

Homogeneity of the 16S rDNA sequence among geographically disparate isolates of Taylorella equigenitalis

Background

At present, six accessible sequences of 16S rDNA from Taylorella equigenitalis (T. equigenitalis) are available, whose sequence differences occur at a few nucleotide positions. Thus it is important to determine these sequences from additional strains in other countries, if possible, in order to clarify any anomalies regarding 16S rDNA sequence heterogeneity. Here, we clone and sequence the approximate full-length 16S rDNA from additional strains of T. equigenitalis isolated in Japan, Australia and France and compare these sequences to the existing published sequences.

Results

Clarification of any anomalies regarding 16S rDNA sequence heterogeneity of T. equigenitalis was carried out. When cloning, sequencing and comparison of the approximate full-length 16S rDNA from 17 strains of T. equigenitalis isolated in Japan, Australia and France, nucleotide sequence differences were demonstrated at the six loci in the 1,469 nucleotide sequence. Moreover, 12 polymorphic sites occurred among 23 sequences of the 16S rDNA, including the six reference sequences.

Conclusion

High sequence similarity (99.5% or more) was observed throughout, except from nucleotide positions 138 to 501 where substitutions and deletions were noted.