|
|
||||||||
School of Veterinary and Biomedical Sciences, Murdoch University, Perth, Western Australia 6150, Australia
Correspondence David J. Hampson d.hampson{at}murdoch.edu.au
Received August 15, 2003
Accepted January 12, 2004
The purpose of this study was to investigate genetic variation in the anaerobic intestinal spirochaete Brachyspira aalborgi by partial sequencing of the 16S rRNA and NADH oxidase genes. The spirochaete is poorly cultivable; hence, only six isolates were available for analysis. Additional sequences were amplified from DNA extracted from fixed colorectal biopsies from 26 patients with histological evidence of intestinal spirochaetosis, and from the faeces of six non-human primates (NHP). Multiple biopsies from sites along the large intestine were tested from three of the 26 patients. Sequences from two biopsies were closely related to those of the spirochaete Brachyspira pilosicoli. Eight B. aalborgi-like 16S rDNA sequences were generated from the biopsies from the other 24 patients, and four from the NHP faeces. The B. aalborgi 16S rDNA sequences were divided into three clusters, 1, 2 and 4, with individual sequence similarities to the type strain ranging from 97.49 to 100 %. All human isolates of B. aalborgi were located in cluster 1, as was the sequence of the so-called Brachyspira ibaraki'. All four 16S rDNA sequences from the NHP faeces and the two NHP isolates of B. aalborgi were located in cluster 4, which was distinct. Cluster 4 may represent a novel Brachyspira species. Evidence for multiple strains of B. aalborgi or other Brachyspira species was found in biopsies from two patients. In the three individuals from whom multiple biopsies were amplified, the sequences at each intestinal site were the same, indicating the presence of one dominant strain.
Abbreviations: IS, intestinal spirochaetosis; NHP, non-human primates; PET, paraffin-embedded tissue.
The GenBank accession numbers for the novel 16S rDNA sequences of Brachyspira aalborgi reported in this study are AY187050AY187061.
| INTRODUCTION |
|---|
|
|
|---|
Knowledge about strain diversity in B. aalborgi is currently limited, mainly because the spirochaete is difficult to isolate, and consequently very few isolates have been available for analysis. This lack of information has hampered attempts to find links between colonization with specific strains of B. aalborgi and the development of intestinal symptoms. To circumvent the problem of a lack of isolates, it is possible to use PCR to amplify B. aalborgi gene sequences directly from colonized tissue and/or faeces (Mikosza et al., 1999, 2001a, b; Kraaz et al., 2001) and then to examine these sequences for heterogeneity.
In an important study undertaken in Sweden, DNA was extracted from colonic biopsies from two 60-year-old patients and almost complete 16S rRNA genes were amplified using degenerate PCR primers. The DNA was cloned, the cloned genes sequenced and, from these data, 17 novel spirochaetal sequences were identified. Phylogenetic trees were constructed and all 17 sequences were shown to group in the B. aalborgi lineage rather than with the lineage containing the porcine intestinal spirochaete Brachyspira (formerly Serpulina) hyodysenteriae. The 17 sequences formed three clusters. The first cluster included sequences from the B. aalborgi type strain, whilst the other two clusters were distinct and new (Pettersson et al., 2000). This study caused some controversy, as it implied not just that there were multiple strains and genetic clusters of B. aalborgi, but also, for the first time, that individuals could harbour multiple different strains of B. aalborgi, as well as related but uncharacterized spirochaetes. In a subsequent study, partial (207 bp) sequences of the 16S rRNA gene were amplified from colonic biopsies taken from 33 patients (Kraaz et al., 2001). Although the small size of the sequence did not allow robust phylogenetic analysis, the results broadly supported the existence of strain variation in B. aalborgi, as well as the presence of several clusters of B. aalborgi-like spirochaetes, which were identified in different individuals. In another study, fluorescent in situ hybridization was used on IS biopsy samples from 40 patients (Jensen et al., 2001). Brachyspira genus-specific probes, but not probes for B. aalborgi, hybridized to 17 samples. Consequently, it was suggested that these B. aalborgi-like sequences belonged to a new intestinal spirochaete species with the provisional name Brachyspira christiani'. Japanese workers also identified what appeared to be novel 16S rRNA gene sequences amongst their isolates of B. aalborgi, and suggested that these came from a novel species proposed as Brachyspira ibaraki (Tachibana et al., 2002).
The purposes of the current study were: (i) to delineate the phylogeny of B. aalborgi, (ii) to determine the level of genetic variability within the species, (iii) to look for evidence of colonization of individuals with multiple strains and (iv) to determine whether novel Brachyspira species may colonize humans and/or NHP. To achieve these aims, the B. aalborgi 16S rRNA and NADH oxidase genes amplified from available strains of B. aalborgi, from colonic biopsies from a large number of human patients with IS and from the faeces of NHP were directly sequenced and compared. In order to obtain as much information as possible, the biopsies were selected from patients originating from a variety of geographical origins, who were either asymptomatic or had various non-specific intestinal symptoms.
| METHODS |
|---|
|
|
|---|
Biopsy samples.
Paraffin-embedded colorectal biopsy samples from 26 patients with histologically diagnosed IS were used for PCR amplification (Table 1). Previously, using diagnostic PCRs, B. aalborgi sequences had been identified in 24 of these samples, and B. pilosicoli in two (Mikosza et al., 1999, 2001a). Multiple biopsies were available from three of the patients: two patients had biopsies taken from the caecum, five sites along the length of the colon and the rectum, whilst the third patient had samples available from the appendix and colon (Table 1). Biopsies were from patients living in Australia (n = 22), Norway (n = 2), France (n = 1) and the USA (n = 1).
|
Faecal samples.
Six faecal samples from NHP from the Perth Zoological Gardens were analysed. These samples, which were designated Zoo 12C, 13, 17C, 18, 23 and 28, had been collected in a previous study, and were shown to contain B. aalborgi DNA (Munshi et al., 2003). B. aalborgi isolates Zoo 12 and Zoo 17 had been isolated from samples 12C and 17C, respectively.
DNA extraction.
DNA from paraffin-embedded tissue (PET) samples was extracted using a previously described method (Mikosza et al., 1999). DNA was extracted from the NHP faeces as previously described (Munshi et al., 2003).
PCR.
PCR amplification of part of the 16S rRNA gene (915 bp; Escherichia. coli numbering A55 to T998) was performed using three overlapping sets of primers specific for Brachyspira spp. (Table 2). For PCR set 16S/2, primer pair 16S/2aal was used for B. aalborgi and primer pair 16S/2pil for B. pilosicoli.
|
PCR amplification of a 483 bp section of the B. aalborgi NADH oxidase (nox) gene, utilizing primers 5'-AACTCTTCCTAACGGTGCT (F833 nox) and 5'-TGGTCTACAGTCATACCGT (R1279 nox), was also carried out on the same samples, except for PET samples Fr 6 and USA 1, where insufficient material was available.
All amplification mixtures consisted of a reaction mixture of 1 x PCR buffer, 1.0 U Tth Plus DNA polymerase (Biotech International), 1.5 mM MgCl2, 10 nmol each dNTP (Amersham Pharmacia Biotech) and 25 pmol each primer prepared to a total volume of 48 µl. Two microlitres of DNA PCR template prepared from PET or faeces was applied to each PCR. For bacterial isolates, cells were applied directly to a 50 µl PCR via a clean wooden toothpick that had been briefly dipped into the cells.
The thermocycling protocol consisted of an initial hold of 94 °C for 4.5 min, followed by 35 cycles of 94 °C for 30 s, 30 s of annealing (specific for each primer set; 16S rDNA; Table 2; nox PCR; 50 °C) and 72 °C for 30 s. A final 5 min holding period at 72 °C was performed.
Sequencing reaction.
16S rDNA and nox PCR products to be sequenced were purified using a commercially available kit (QIAquick PCR purification kit; Qiagen), according to the manufacturer's specifications. Purified PCR product was sequenced using the appropriate PCR primer pairs (Table 2) with a commercially available cycle sequencing kit (ABI PRISM dye terminator cycle sequencing ready reaction kit; Applied Biosystems), according to the manufacturer's specifications.
Phylogenetic analysis.
The sequence data obtained were aligned and compared with previously published Brachyspira species 16S rDNA and nox sequences obtained from GenBank, using SeqEd. These included sequences from all currently reported isolates of B. aalborgi (Kraaz et al., 2000; Jensen et al., 2001), as well as the type strains of six other Brachyspira species. Seventeen B. aalborgi-like sequences from a previous study involving IS patients (Pettersson et al., 2000) and a single B. aalborgi-like sequence reported in a previous study and provisionally named B. ibaraki (Tachibana et al., 2002), were also included for further comparison. The 16S rDNA sequences of two distantly related species of spirochaetes, Borrelia burgdorferi B31T (accession no. X57404) (Marconi & Garon, 1992) and Leptonema illini 3055T (accession no. Z21632) (Hookey et al., 1994), were included as outgroups. Gaps that existed between these outgroup species and Brachyspira spp. were removed.
The outgroup for the nox sequences was designated Enterococcus faecalis 10C1 (accession no. X68847) (Ross & Claiborne, 1992). The previously published nox sequences of all type strains of species in the genus Brachyspira were also included. All previously published nox sequences were obtained from GenBank using NCBI BLAST. Partial nox sequences for B. pilosicoli PCR-positive samples were not obtained.
Neighbour-joining phylogenetic trees were inferred on the basis of a distance matrix with the one-parameter model of Jukes and Cantor (equal base frequencies) (Jukes & Cantor, 1969) using PAUP 4.0b4a (Swofford, 1999). Support for tree topologies was tested by bootstrap resampling with 1000 replicates and percentage values were placed at the major nodes.
For the neighbour-joining phylogenetic tree based upon partial sequences of the nox gene, PAUP 4.0b4a was used as previously mentioned except that ambiguity codes were allowed in samples NT 2, WA 11, WA 21, Zoo 13, Zoo 18 and Zoo 23.
Nucleotide accession numbers.
Twelve partial 16S rDNA sequences that were obtained from the fixed tissue samples, faeces and isolates were submitted to the GenBank database, with the accession numbers listed in Table 3.
|
| RESULTS |
|---|
|
|
|---|
The tree constructed from 915 bp of the 16S rRNA gene is presented as Fig. 1. The type strains of the six Brachyspira species (other than B. aalborgi) are clustered at the top of the tree, with the eight B. pilosicoli sequences, including two amplified from PET samples, all clustering around that of the type strain, P43/6/78T. These B. pilosicoli sequences were divided into five groups, differing by a maximum of five bases (GP 17 and HRM2B). The sequence of WA 9, positive for B. pilosicoli by diagnostic PCR, was identical to a sequence from an Omani Arab patient (N26), with both being closely related to B. pilosicoli P43/6/78T.
|
The sequences of the B. aalborgi strains appear in the lower half of the figure, and are divided into four clusters, marked 14. Cluster 1 sequences were the most commonly detected from biopsies in the current study (five sequences generated from 14 PET samples, including all four non-Australian PET samples). Cluster 1 sequences also included those from all four cultured human strains of B. aalborgi (513AT, W1, JLL and AL), the sequence of B. ibaraki and six of the 17 sequences generated from two individuals in the study of Pettersson et al. (2000). The latter authors also designated this cluster as cluster 1 B. aalborgi, containing the type strain.
Cluster 2 sequences included three generated from nine PET samples in this study, as well as nine from among the 17 clones sequenced by Pettersson et al. (2000). None of the samples in the current study generated a cluster 3 sequence, as described by Pettersson et al. (2000), but the two sequences from the former study were placed in cluster 3 in the figure. Cluster 3 was relatively distantly placed from clusters 1 and 2.
Cluster 4 was located at the greatest distance from clusters 1 and 2, containing sequences with between 97.49 and 98.36 % sequence similarity over 915 bp of the 16S rRNA gene with the B. aalborgi type strain, which was located in cluster 1. Cluster 4 was divided into two subclusters. Four sequences generated from the six NHP faecal samples were located in cluster 4, and these sequences could be delineated by the species of primate from which they originated. Zoo 28 was from a baboon, whilst Zoo 12C and 17C sequences were both from vervets. Zoo 13, 18 and 23 had identical sequences, with samples Zoo 18 and 23 being obtained from an enclosure containing Tonkean macaques, and Zoo 13 being obtained from a Japanese macaque in a separate enclosure.
The tree derived from 413 bp of the nox gene is presented as Fig. 2. The tree is similar in its general structure to that of the 16S rRNA gene sequence tree, with the B. aalborgi sequences being distinct from the sequences of the other Brachyspira species (the latter being located at the top of the tree). Sequences for the cluster 3 B. aalborgi samples were not available. The nox sequences from the NHP faeces again clustered together, and were distinct from the nox sequences from samples that were located in clusters 1 and 2 in the 16S rDNA sequence tree (Fig. 1). The latter two clusters were no longer well defined in the tree of partial nox sequences.
|
Two of the PET samples, WA 21 and NT 6, generated some mixed template in the 16S/1 PCR. All other 16S rDNA chromatograms showed unambiguous peaks. For WA 21, mixed template occurred at a region of sequence corresponding to E. coli region V2 (T186 to G199), indicating the presence of at least two strains of B. aalborgi. At positions 191, 193, 194 and 198 the dominant peaks were A, A, T and T, with smaller peaks on the chromatograms at these positions showing G, T, A and C, respectively. In addition, there was a dominant T at position 213 and a smaller C peak at this position. The dominant sequence corresponded to B. aalborgi clusters 24, and the smaller peaks corresponded to a cluster 1 B. aalborgi sequence. The predominant sequence for WA 21 was used in construction of Fig. 1.
For sample NT 6, additional small peaks were present on the chromatograms in the regions equivalent to the V2 and S1b regions of the E. coli 16S rRNA. In the V2 region the primary sequence corresponded to B. aalborgi cluster 1, with a second, minor sequence equivalent to B. aalborgi cluster 24, and a third corresponding to Brachyspira innocensBrachyspira murdochii. At S1b, the main sequence corresponded to B. aalborgi, with the minor sequence corresponding to B. hyodysenteriae, B. murdochii or B. innocens. A further mixed peak for NT 6 occurred at base position 239, where both C and T appeared. The NT 6 sequence was not used in constructing Fig. 1.
| DISCUSSION |
|---|
|
|
|---|
One of the most striking features of the dendrograms produced from both the 16S rDNA and nox sequences was the clear demarcation between the B. aalborgi sequences from those of the six other Brachyspira species that were analysed. These six species, formerly belonging to the genus Serpulina, were all closely clustered in a single lineage (called the B. hyodysenteriae lineage by Pettersson et al., 2000). In the case of B. pilosicoli, the various isolates and the two B. pilosicoli sequences amplified from PET were closely clustered near their type strain, P43/6/78T. The relatively large distance between B. aalborgi and these other six Brachyspira species, compared with the relatively small distances between the latter species, has been reported in a number of other studies (Lee & Hampson, 1994; Pettersson et al., 1996, 2000; Stanton et al., 1996). At the time that it was proposed that the former Serpulina species be included in the genus Brachyspira (Ochiai et al., 1997), this relatively large 16S rDNA sequence difference between the Brachyspira and Serpulina lineages caused disquiet about the proposal (Hampson, 2000).
Within the B. aalborgi lineage, four clusters of sequences were evident. The existence of these clusters confirms the previous findings of Pettersson et al. (2000), who identified three of these clusters when sequencing cloned 16S rDNA from just two patients. The current findings therefore also provide indirect support for the contention of Pettersson et al. (2000) that individuals may be colonized with multiple strains of B. aalborgi and B. aalborgi-like spirochaetes. Whether colonization with multiple strains occurs frequently, however, remains unclear. In the current study, biopsies from different sites along the large intestine all generated the same sequence for both genes in all three patients tested. This finding suggests that, if there were multiple strains present in these three individuals, one predominant strain colonized along the length of the large intestine and was preferentially amplified in the PCRs. In biopsies from two other patients, WA 21 and NT 6, analysis of sequence chromatograms suggested that there was a dominant B. aalborgi strain present, with either another strain of B. aalborgi (WA 21), or other Brachyspira species (NT 6) also being present. Both these PET samples were from the rectum, but whether this is significant is unclear. Previously, using diagnostic PCRs, the authors identified two colorectal biopsies containing DNA from both B. aalborgi and B. pilosicoli (Mikosza & Hampson, 2001).
Considerable strain diversity was identified within the B. aalborgi lineage. Clusters 1 and 2 predominated, with no sequence being generated from the cluster 3 of Pettersson et al. (2000). Hence, cluster 3 strains appear to be uncommon. Sequences in clusters 1 and 2 came from individuals from different geographical origins, of different ages and genders, and were from different sites along the large intestine. Similarly, it was not possible to link specific strains or clusters of strains to specific symptoms in the patients from whom the biopsies were obtained.
An interesting feature of the study was that the type strain (513AT) and the three other cultured isolates from humans all belonged to B. aalborgi cluster 1. Another strain that was recently isolated from faeces in Australia (Brooke et al., 2003) also has a cluster 1 sequence (unpublished data). Cluster 1 therefore is best thought of as containing typical cultivable B. aalborgi strains. It is possible that strains of clusters 2 and 3 are even more fastidious in culture than those of cluster 1, and that is why they have not been isolated to date. Analysis of published 16S rDNA sequence of the so-called B. ibaraki', from a cultured isolate (Tachibana et al., 2002), suggested that it also is a cluster 1 B. aalborgi isolate (Fig. 1; Table 4).
|
Cluster 4 isolates all originated from NHP faeces, and were themselves divided into at least two subclusters. The relatively large distance between B. aalborgi clusters 1 and 4, and the fact that all cluster 4 sequences came from NHPs, suggests that the cluster 4 spirochaetes may represent one or more novel species in the genus Brachyspira. This is particularly evident when one considers the relatively small 16S rDNA sequence differences between most of the other Brachyspira species. Previously, these isolates had been misidentified as B. aalborgi based on amplification with a B. aalborgi PCR (Munshi et al., 2003). Samples 12C and 17C came from two vervet monkeys, and previously spirochaetes were cultured from both samples (Munshi et al., 2003). Further work is required to characterize these isolates and determine whether they represent a distinct novel Brachyspira species. Whether or not the so-far uncultivated human spirochaetes in clusters 2 and 3 may also represent novel species, perhaps even the B. christiani of Jensen et al. (2001), cannot be answered until isolates are available for detailed study. New PCR tests need to be developed for specific detection of and rapid differentiation between spirochaetes from the four B. aalborgi-like clusters.
| ACKNOWLEDGEMENTS |
|---|
|
|
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
J.-C. Lin, S.-J. Hou, L.-U. Huang, J.-R. Sun, W.-K. Chang, and J.-J. Lu Acute Cholecystitis Accompanied by Acute Pancreatitis Potentially Caused by Dolosigranulum pigrum. J. Clin. Microbiol., June 1, 2006; 44(6): 2298 - 2299. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. K. Jensen, P. S. Teglbjaerg, C. F. Lindboe, and M. Boye Demonstration of Brachyspira aalborgi lineages 2 and 3 in human colonic biopsies with intestinal spirochaetosis by specific fluorescent in situ hybridization J. Med. Microbiol., April 1, 2004; 53(4): 341 - 343. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |