Phylogenetic Evidence for Novel and Genetically Different Intestinal Spirochetes Resembling Brachyspira aalborgi in the Mucosa of the Human Colon as Revealed by 16S rDNA Analysis

SYSTEI\IL4TIC AND _htt-,-p_:llw_w_w_.ur_ba_n_fis_ch_er_.de--"/jo_u_rn_als_/s_am_ _ _ _ _ _ _ _ _ _ _ _ APPLIED MICROBIOLOGY System. Appl. Microbiol. 23, 355-363 (2000)

© Urban & Fischer Verlag

Phylogenetic Evidence for Novel and Genetically Different Intestinal Spirochetes Resembling Brachyspira aalborgi in the Mucosa of the Human Colon as Revealed by 16S rONA Analysis BERTIL PETTERSSONl, MEl WANG2, CLAES FELLSTROM 3, MATHIAS UHLEN\ GORAN MOLIN2, BENGT jEPPSSON\ and SIV AHRNE2

Department of Biotechnology, Royal Institute of Technology, Stockholm, Sweden Laboratory of Food Hygiene, P.O. Box 124, Lund University, Lund, Sweden 3 Department of Medicine and Surgery, Swedish University of Agricultural Sciences, Faculty of Veterinary Medicine, Uppsala, Sweden 4 Department of Surgery, Lund University Hospital, Lund and Department of Surgery Malmo University Hospital, Malmo, Sweden 1

2

Received May 31, 2000

Summary Intestinal spirochetes (Brachyspira spp.) are causative agents of intestinal disorders in animals and humans. Phylogenetic analysis of cloned 16S rRNA genes from biopsies of the intestinal mucosa of the colon from two Swedish 60-years old adults without clinical symptoms revealed the presence of intestinal spirochetes. Seventeen clones from two individuals and 11 reference strains were analyzed and the intestinal spirochetes could be divided into two lineages, the Brachyspira aalborgi and the Brachyspira hyodysenteriae lineages. All of the clones grouped in the B. aalborgi lineage. Moreover, the B. aalborgi lineage could be divided into three distinct phylogenetic clusters as confirmed by bootstrap and signature nucleotide analysis. The first cluster comprised 6 clones and the type strain B. aalborgi NCTC 11492T • The cluster 1 showed a 16S rRNA gene similarity of 99.4-99.9%. This cluster also harbored the only other strain of B. aalborgi isolated so far, namely strain WI, which was subjected to phylogenetic analysis in this work. The second cluster harbored 9 clones with a 98.7 to 99.5% range of 16S rDNA similarity to the B. aalborgi cluster 1. Two clones branched distinct and early of the B. aalborgi line forming the third cluster and was found to be 98.7% similar to cluster 1 and 98.3-99.1 % to cluster 2. Interestingly, this shows that considerable variation of intestinal spirochetes can be found as constituents of the colonic micro biota in humans, genetically resembling B. aalborgi. The presented data aid significantly to the diagnostic and taxonomic work on these organisms. Key words: Brachyspira aalborgi - Human colon - intestinal spirochetes - micro biota - Phylogeny - 16S rDNA

Introduction Gastrointestinal disorders such as chronic diarrhea and rectal bleeding, clinically defined as intestinal spirochetosis (IS), has been characterized for a variety of animal species as reviewed by DUHAMEL, 1997; SWAYNE and McLAREN, 1997; TAYLOR and TROIT, 1997. Diagnostically, the disease can be visualized by the appearance of a so-called "false brush border" originating from a mat formed by a large amount of spirochetes, which are attached by one end to the epithelial layer of the colonic mucosa (HARLAND and LEE, 1967). Spirochetes associated with IS are known as Brachyspira pilosicoli, formerly classified in the genus Serpulina (TROIT et aI., 1996) and Brachyspira aalborgi (HOVIND-HoUGEN et aI., 1982).

Evolutionary, these species have been shown to belong to a group of closely related organisms forming a distinct clade by bisecting the phylogenetic tree of the order Spirochaetales (PASTER et aI., 1991, PEITERSSON et a\., 1996). B. pilosicoli seem to be a frequently isolated organism in association with IS in pigs, dogs, chickens and humans (TAYLOR et aI., 1980; TROIT et aI., 1997; McLAREN et aI., 1997; TRIVETT-MoORE et a\., 1998). Moreover, a marked genetic diversity has been found among B. pilosicoli isolates as recovered from humans and animals (ATyEO et aI., 1996; LEE et aI., 1994). On the other hand, B. aalborgi has only been isolated twice (HOVIND-HoUGEN et aI., 1982; KRAAZ et aI., 2000) and 0723-2020/00/23/03-355 $ 15.00/0

356

B. PEITERSSON et aL

its role as an agent of human IS was confirmed recently (MIKOSZA et al., 1999; KRAAZ et al., 2000). In these studies, molecular biology techniques were applied on DNA extracts from paraffin embedded tissue samples from IS patients and short sequences from the 16S rRNA gene revealed involvement of B. aalborgi in human IS. Consequently, because only two 16S rRNA gene sequence of B. aalborgi exists (HOOKEY et al., 1994, KRAAZ et al., 2000), to be used for probe design, variants of this species might well have failed being detected so far. Despite the efforts in isolating B. aalborgi, this has only been repeated once when a second isolate of this species was recovered by FELLSTRGM and coworkers (KRAAZ et al., 2000). Especially, the slow growing feature (up to 2-3 weeks), the anaerobic requirement and the fastidious nature of B. aalborgi makes isolation of this species a difficult task. Therefore, the extent as to which this species occurs in humans has remained largely unknown. We are running a project aiming at phylogenetic description of the bacterial biodiversity in the human colonic mucosa and here we present the first ever study on the genetic variation in brachyspiras as based on almost full-length 16S rDNA clones in this actual habitat. The clones were obtained from libraries of two 60-years old healthy adults and suggested that a vast genetic variety of B. aalborgi can exist in humans without gastrointestinal symptoms of IS. Therefore, an important framework for the detection, the genetic diversity and the classificatory work on the intestinal spirochetes in the phylogenetic neighborhood of B. aalborgi is described in the present work.

Material and Methods Construction and screening of clone libraries Biopsies were recovered from 6O-years old adults, which were included in a pilot-study on the value of screening after malignancies by sigmoidoscopy. The sample taking was located to the sigmoidum and there were no lesions. The tissue samples were covered by lxTE, frozen in liquid nitrogen, and stored in -90°C until processed further. The DNA was prepared from 5 biopsies and the biopsies were incubated in an ultrasonic bath for 5 min (Millipore, Sundbyberg, Sweden) to enrich for bacteria. Thereafter vortex for two min (Chiltern, Therma-Glas, Gothenburg) and centrifugation at 1000 rpm for 2 min was performed. Bacteria in the supernatant were disintegrated by shaking with glass-beads (2 mm in diameter) for 30 min at 4 °C using an Eppendorf Mixer 5432. The shaking was followed by centrifugation and DNA was extracted by using the DNA DIRECT System I according to the manufacturer (Dynal AS, Oslo, Norway). Almost complete 16S rRNA genes were amplified using degenerated PCR primers targeting the termini of the gene. The primers were constructed as follows; 5'-agagtttgatiitggctcag-3' and 5'-cggitaccttgttacgac-3' where i denotes inosine and their 3'-ends targeting positions 27 and 1494 of the 16S rRNA gene of Escherichia coli (BROSIUS et aL, 1978), respectively. The temperature profile was 96 °C for 15 sec, 40°C for 30S, and 72 °C for 90 sec. A total of 30 cycles was performed followed by an extension step at 7 °C for 10 minutes. The fragments were checked on agarose gel and subsequently cloned into pCRII, a pGEM derivate for direct cloning of PCR products using AT-overhangs, according to the manufacturer

(Promega, Madison, USA). No purification of the PCR products prior to cloning was performed. White colonies were picked and screened by PCR using vector primers RIT28 and RIT29 (HULTMAN et aL, 1991).

DNA sequencing of clones The 5'- and the 3'-ends of the constructs were sequenced using universal sequencing primers flanking the cloning sites. These partial sequences were searched against GenBank using the Advanced BLAST similarity search option (ALTSCHUL et aL, 1997) accessible from the homepage at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.govl) . Novel clones, totaling 17, indicating spirochetal origin were sequenced completely using energy transfer dye terminator chemistry (Amersham Pharmacia Biotech, Uppsala, Sweden). The primers used for sequencing have been described elsewhere (PEITERSSON et aL, 1996). All nucleotide sequences were determined using the MegaBACE 96 capillary system by separating the dideoxy fragments at 5 kV for 120 minutes. The injection was performed for 50 s at 3 kV. Phylogenetic analysis The generated sequences were manually merged into an alignment using the GDE sequence editor, Genetic Data Environment software (SMITH et al., 1992). A secondary structure model served as guideline for proper identification of stems and loops and for the verification of sequence variability. Outfiles to be used for construction of phylogenetic trees were generated in the format suitable for PHYLIP, Phylogenetic Inference Package (FELSENSTEIN, 1993). The neighbor-joining method (NJ) described by SAITOU and NEI (1987) is hiding under the program name NEIGHBOR, which was used for construction of evolutionary distance trees. DNAML was used to compute trees with the maximum likelihood (ML) algorithm. Distances were corrected for multiple substitutions at single locations by the oneparameter model of JUKES and CANTOR (1969) and statistical evaluation was performed by re-sampling of the data 1000 times. The F84 evolutionary model with empirical nucleotide frequencies and a transition/transversion ratio set to 2.0 (FELSENSTEIN, 1993) was used to calculate evolutionary trees by ML. The option for global rearrangement was invoked to find the best tree. Accession numbers All sequences generated and used in this study can be retrieved from GenBank using the accession numbers listed in Table 1. The clone sequences have been deposited under, adhucornu, a name showing that the origin is from adult human colon mucosa. This clone designation was developed as to be in line with human fecal clones (SUAU et al., 1999). Thus, the clone library Hc, Tva, or Tvb, and a running number of the particular clone of the actual library follow adhucomu.

Results and Discussion Characterization of the 165 rRNA gene sequence libraries This study focused on the investigation of the spirochetal biodiversity in the mucosa of the human distal colon. Biopsies were collected by sigmoidoscopy and total DNA was prepared from a beating bead-mill procedure. Consequently, only a small fraction of bacterial DNA could be recovered and the number of cycles needed for generating sufficient amount of fragment for suc-

Intestinal Spirochetes in the Human Colonic Mucosa

cessful cloning had to be in the range of 25 to 30 cycles. In order to compensate for biased amplification, three independent PCR procedures were carried out and the amplicons were pooled prior to cloning. Screening of the resulting 16S rDNA clone libraries by DNA sequencing indicated the presence of intestinal spirochetes in two healthy adults. The fraction of spirochetal clones in the individuals Hc and Tv was 24 and 20% respectively. This is a pretty high number and one can speculate in that these organisms are weakly, if at all, pathogenic. Merely, these significant fractions are likely pointing at the fact that these organsims attach to the mucosa ousting other bacterial specimens (HARLAND and LEE, 1967). Thereby, the resulting brush-like formation by the spirochetes blocks out microbial association by other bacteria and hinder nutrient transport through the mucosa. Thus, the result would be a lowered functionality of the intestinal mucosa with an increased number of brachyspiras. Unfortunately, no pathology/histological studies were performed at the time when the biopsy material was recovered mostly because IS was not diagnosed. Seventeen clones were selected for more detailed analysis of their genetic relationships to previously described intestinal

357

spirochetes, i.e. members of the genus Brachyspira. All clones were found to be highly similar to the 16S rDNA sequences of the two hitherto characterized strains of B. aalborgi, namely strains NCTC 11492T (HOVINDHOUGEN et aI., 1982) and WI (KRAAZ et aI., 2000). The clones differed slightly from each other and ranged between 98.3 to 99.9% in 16S rDNA sequence similarity. The micro-variability between the 16S rDNA clones was striking and one could question the true spirochetal origin of the clone sequences or if they e.g., should be judged as being artifacts generated by the procedure from one single phylotype. However, both strands of the cloned spirochetal 16S rRNA genes were sequenced between the positions 27 and 1494 according to E. coli (BROSIUS et aI., 1978) minimizing the risk for inaccurate editing of the electropherograms due to the presence of e.g. compressions, false peaks, etc. Thus, the observed micro-variability might be due to differences between the ribosomal operons, incorporation errors by the Taq polymerase and/or simply, a result of strain diversity. The first scenario is not regarded as a plausible explanation because no polymorphisms due to rrn operonal differences were observed when scrutinizing the resulting elec-

Table 1. Clones and strains of the genus Brachyspira used in this study. Species / Clone

Strain

Origin

Ace. no

Reference

Brachyspira aalborgi

NCTC 11492 T

human

Z22781

Brachyspira aalborgi

WI

AF200693 AF228806 AF228807 AF228808 AF228809 AF228810 AF228811 AF228812 AF228813 AF228814 AF228815 AF228816 AF228817 AF228818 AF228819 AF228820 AF228821 AF228822 U23030

HOVIND-HOUGEN et ai., 1982; HOOKEY et ai., 1994 KRAAZ et ai., 2000; This study This study This study This study This study This study This study This study This study This study This study This study This study This study This study This study This study This study DAVELAAR et ai., 1986; STANTON et ai., 1998 HARRIS et ai., 1972; PETTERSSON et ai., 1996 KINYON and HARRIS 1979; PETTERSSON et ai., 1996 STANTON et ai., 1997 STANTON et ai., 1997 TAYLOR et ai., 1980; PETTERSSON et ai., 1996 FELLSTROM et ai., 1995; PETTERSSON et ai., 1996 PETTERSSON et ai., 1996 PETTERSSON et ai., 1996

Brachyspira alvinipulli

ATCC 51933 1

human human human human human human human human human human human human human human human human human human aVian

Brachyspira hyodysenteriae

ATCC 27164 T

porcine

014930

Brachyspira innocens

ATCC 29796 T

porcine

014920

Brachyspira intermedia Brachyspira murdochii Brachyspira pilosicoli

ATCC 511401 155-20 ATCC 51139 T

porcine porcine porcine

U23033 U22838 014927

Brachyspira sp

C378

porcine

014918

Brachyspira sp. Brachyspira sp.

C555 ANI023

porcine porcine

014926 UI49I5

Hca09 Hca20 Hca25 Hca41 Hca61 Hca76 Hcc07 Hcc33 Hcc35 Hcc36 Hcc39 Tva17 Tva35 Tvb03 Tvb21 Tvb40 Tvb72

358

B. PETTERSSON et al.

Table 2. Unique and signature nucleotide positions in the 165 rRNA molecule of the B. aalborgi lineage. Position of base or base pair2

Base or base pair in lineage, order or domain: 1 B. aalborgi

lineage

B. hyodysenteriae lineage

Spirochaetales 3

Bacteria

no signature no signature G

(183 - 194) 184 - 193 260

A-U C-G A

C-G U-A

GS

nd nd G

265 590 - 649

A A-U

G C - G, U - R

A,G,U Y-D

G Y-D

591 - 648 688 - 699 772 - 807

C-G G-U G-C

U-R G-C A-U

WeD G-Y B-D

no signature G-C W-W

822 - 878

C-G

U-A

U-A

D-B

834 - 852 835 - 851

U-C C-G

U-A U-G

D-B D-B

no signature D-N

843 1010 - 1019

A A-U

U U-A

A,C,U B-V

no signature B-N

1116 - 1184 1159 1168 1281

U-G C U U

C-G U

Y-G U U,C A,U

no signature no signature no signature no signature

CS A

Exceptions 4

A: Gram-positives low G+C (a few subgroups) A: Rare outside Spirochaetales A: Chlamydiales, y- and E-proteobacteria (few), Gram-positives low G+C (a few subgroups) U: Rare outside Spirochaetales G - C: Aquificales, Slackia spp., Ammonifex, Leptothrix buccalis, Haloanaerobium group C - G: Thermotogales, Chloroflexus subdivision (some), Borrelia, Thermus group (most), Verrucomicrobiales (most), Chlamydiales, Fibrobacter, Chryseobacterium (some), C: Thermus group (few), FCB (few), Fibrobacter, Proteobacteria (few)

A - U: Leptospirillum subgroup, a-proteobacteria (few), Fusobacteria (few), Gram-positives low G+C (most Mollicutes and a few others)

The specific dominating base or pair is noted when present in >10% of the bacterial taxa, otherwise the residues as found in each position are given according to an ambiguity code. All of the possibly combinations of base pairings as suggested by the ambiguity codes, do not necessarily mean that they exist. Merely, this listing denotes the actual residues, which can be observed in the actual positions even if both canonical and non-canonical are formed in some of them. The base composition at every position follows the suggested letter code by the Nomenclature Committee of the International Union of Biochemistry where R means A/G, W is AJU, Y is CIU, B is C/GIU not A, D is A/GIU not C, B is A/C/T not G, V is A/C/G not U, and N denotes A/C/GIU. 2 Positions given are according to E. coli (BROSIUS et aI., 1978) and those within parenthesis are lacking a corresponding pair in E. coli. 3 Positions that were too difficult to give a composition due to hyper variability are denoted with nd (not determined). 4 Exceptions to the signatures shown in bold face have been listed in a general way. A more detailed description is given in the Results and Discussion. S Brachyspira alvinipulli ATCC 51933 T of the B. hyodysenteriae lineage has an adenosine and a uridine in the positions 260 and 1168, respectively. 1

tropherograms from the sequencing procedure of B. aalborgi strain Wi. Moreover, we performed cloning of a peR product of the 16S rRNA gene from B. aalborgi Wi and sequencing was performed between the Vi and V3 regions. Twelve clones were sequenced, but only the nucleotide composition for sub-cluster la (Table 3) was found and none for the other clusters and subclusters of the B. aalborgi lineage (Fig. 1, Table 3) were revealed from these clones (not shown). Actually all but one were identical and with the twelfth having a mutation in a position different from any of those variable positions as found in the Hc and Tv clones (biopsy material from the human intestinal mucosa). Moreover, PETIERSSON et aI., (1996) published a phylogenetic analysis of 26 porcine

intestinal spirochetes without the notification of the presence of so-called polymorphisms (micro-heterogeneities) between different operons. In that study, it was concluded that the intestinal spirochetes are characterized by having identical 16S rRNA genes or that these spirochetes only have one rrs gene, which has been shown for the species Brachyspira hyodysenteriae (ZUERNER and STANTON, 1994). However, the complete picture on this issue is obtained only by mapping members of the clusters and sub-clusters of the B. aalborgi lineage. By studying two individuals, Hc and Tv, of which one, Tv, was represented by two separate libraries Tva and Tvb as constructed from two independent amplifications from DNA preparations from two different biopsies, identical

Intestinal Spirochetes in the Human Colonic Mucosa

359

Table 3. Differentiating nucleotide positions in the 16S rRNA molecule modeled for the clusters of the B. aalborgi lineage. Position of base or base pair!

Base or base pair in sub-cluster, cluster or lineage:

B. aalborgi

B. aalborgi

B. aalborgi

B. aalborgi

B. hyodysenteriae Exceptions

sub-cluster la

sub-cluster 1 b

cluster 2

cluster 3

lineage

81 - 88 144 - 178 185 - 192

G-U G-C G-C

A-U G-C G-C

G-U G-U A-U

G-U G-C A-U

G-U G-C A-U

187 188 406 - 436 416 - 427 610 612 - 628 630 677 - 713 700 701 1000 - 1040 1246 - 1291

U A G-U G-U U U-G G U-A A U A-U C-G

U A G-U G-U U U-G G U-A A U A-U U-G

A U G-U G-U U U-G A U-A A U A-U C-G

A U G-C U-U G C-G A U-G G C G-U C-G

A C,U G-C G-U G C-G A,G U-G G C A-U G-G

1

G - U: B. intermedia Be130 (PETTERSSON unpublished)

A: B. aalborgi NCTC 11492T A: B. pilosicoli WesB

U - G: B. aalborgi NCTC 11492 T

The number of an actual position corresponds to that in the 16S rRNA molecule of E. coli (BROSIUS et al., 1978). Positions within parenthesis are located in region, which is truncated in the 16S rRNA molecule of E. coli and the positions are those that are closest.

Fig. 1. An evolutionary distance tree computed by neighbor joining (SAITOU and NEI, 1987) from a matrix corrected for multiple substitutions at single locations by the one-parameter model of JUKES and CANTOR (1969). The tree shows the phylogenetic positions of the spirochetal clones as obtained from the intestinal mucosa of the human distal colon. The B. hyodysenteriae and the B. aalborgi lineages are marked in the tree as well as the defined clusters and sub-clusters as defined in this work. Bootstrap values are shown as percentage of times out of 1000 replicates that a clade to the right of the actual node occurred. The scale bar indicates nucleotide substitutions per nucleotide position. The tree was outgrouped to the distantly Borrelia related spirochetes burgdorferi B31 T, Treponema pallidum Nichols strain T and Leptonema illini 3055 T •

. - - - - - - Brachyspira alvinipulli ATCC 51933 T Brachyspira sp. strain C555 Brachyspira sp. strain AN1023 Brachyspira sp. strain C378 Brachyspira murdochii 155-20 100 Brachyspira innocens ATCC 29796 T Brachyspira intermedia ATCC 51140 T Brachyspira hyodysenteriae ATCC 27164 T 1.-_ _ _ _ _ _ _ Brachyspira pilosicoli ATCC 51139 T

Hcc33

]

Brachyspira aalborgi W1 1a Brachyspira aalborgi NCTC 11492 T

Tvb72 Tvb21 Tva35 Tva17 Tvb40

1b

2

100

0.01 100

Hca25 Tvb03

360

B.

PETTERSSON

et al.

clones (not included) justified the existence of the actual phylotype. The significance of a low performance by the polymerase is thus not believed to be the detrimental effect to cause the observed clone sequence variation, which is also ruled out by the presence of signature nucleotide features (below). Previously, a micro-variability in the range of 99.699.9% was observed between strains within individual species belonging to the B. hyodysenteriae line of descent (PEITERSSON et al., 1996). Moreover, new strains of this lineage, which have been isolated from other hosts than humans and pigs rather show a slight variation than sharing identical 16S rDNA sequences (PEITERSSON and FELL STROM unpublished). Therefore, most of the non-redundant clones can be judged as representatives of brachyspiral speciemens that can be found in the mucosa of the human colon.

Phylogenetic analysis At present, we maintain an alignment consisting of more than 80 different 16S rDNA sequences, which belongs to Brachyspira spp. All sequences are characterized by showing strikingly high inter-species similarity values between their primary structures of the 16S rRNA genes (>95.5%), which is congruent with previously published data (PEITTERSSON et al., 1996). The evolutionary distance tree presented in Fig. 1 was computed from an alignment comprising 1433 positions without the removal of gapped positions and the overall topology was not altered when the ML method was used for phylogenetic calculations. The tree was outgrouped to Borrelia burgdorferi B31 T, Treponema pallidum Nichols strainT and Leptonema illini 3055 T • Bootstrap percentage values as obtained from 1000 resamplings of the data set are given at the nodes that were found to be statistically significant at the level of 85% or more. Also, all of the 17 clones showed the signature nucleotide features, which were previously defined for the intestinal spirochete lineage within the order Spirochaetales (PEITERSSON et al., 1996). Thus, these previously published signatures of the brachyspiras still hold. The cladistic investigation of the evolutionary relationships between the clones generated in this work and 16S rRNA gene sequences from type strains and selected reference strains of the genus Brachyspira showed that the intestinal spirochete line of descent bifurcated into two distinct sub-lines within the order Spirochaetales (Fig. 1). These have been named the B. hyodysenteriae lineage and the B. aalborgi lineage borrowing their binomial epithets from representative organisms belonging to the respective clade.

The Brachyspira aalborgi lineage The B. hyodysenteriae lineage was subjected to signature nucleotide analysis with special emphasis on the latter lineage. The informative positions are shown in Table 2 and those regarded as being signatory are shown in bold. In this context, a signature nucleotide was defined

as a nucleotide residue explicitly found in a certain position within the sequences of the B. aalborgi lineage where the base that is present differs from that found in the majority of the other taxa in the domain Bacteria. Interestingly, several signature positions could be found defining the B. aalborgi line of descent supporting the distinct branching from the other intestinal spirochetes in the tree and adding to the monophyly of the B. aalborgi lineage. In addition, unique nucleotides were found that differentiated this lineage from other spirochetes and members of the B. hyodysenteriae lineage. Typically, the 16S rRNA similarities between members of the B. aalborgi lineage and the B. hyodysenteriae lineage were in the range of 96.0-97.2%. Interestingly, all trees revealed the formation of three clusters of species and clones within the B. aalborgi lineage and they have been labelled 1,2, and 3 in Fig. 1. These clusters were supported by high bootstrap values as well as by the presence of discriminative nucleotide patterns (Table 3). Like the B. hyodysenteriae lineage, the B. aalborgi lineage has most of its differences in the variable regions in the vicinity of the 5'-end of the 16S rRNA gene. Because of the high similarity between the OTUs, the exclusion of gaps resulted in that some of the internal nodes within the clusters and sub-clusters collapsed (not shown), but the lineages discussed herein remained stable. At present the B. aalborgi lineage contain only two cultured strains of the type species and a set of more or less divergent clones from the human colonic mucosa.

Clusters within the Brachyspira aalborgi lineage Cluster 1 harbored the two strains isolated so far of the type species of the genus Brachyspira (strains NCTC 11492 T and WI), one clone from the Hc library (Hcc33), and five Tv clones. The 16S rDNA similarity ranged between 99.4 to 99.9%, which is almost within the limits for strain variation of the species of the B. hyodysenteriae lineage. Because this cluster branched into two subclusters in both the NJ and ML tree, these were marked out as 1a and 1b of the cluster 1 (Fig. 1). No nucleotide motifs were found to support the sub-cluster la, while two unique sequence features in the resulting 16S rRNA molecule, namely A8! and U!246 (E. coli numbering), were restricted to sub-cluster 1b (Table 3). Statistically, there was weak support for sub-cluster 1a and we suggest these sub-clusters to be considered as tentative until investigated further. On the other hand, because of the unusually high 16S rDNA inter-species similarities between members of the B. hyodysenteriae lineage (HOOKEY et al., 1994; PASTER et al., 1991; PEITERSSON et al., 1996; STANTON et al., 1991; 1996; 1997; 1998), the Tv clones cannot be identified as originating from the species B. aalborgi without having the phylotype Tv isolated. Surprisingly, only the Hc library was found to harbor a clone that was identified as B. aalborgi. Cluster 2 contained clones of the Hc library, solely. Internally, the cluster showed a 16S rDNA similarity of 99.5-99.9%, while being 98.7-99.5% similar to cluster 1. Its intermediate position between the clusters 1 and 3

Intestinal Spirochetes in the Human Colonic Mucosa was also supported by the nucleotide features listed in Table 3. Cluster 3 consisted of only two clones, one from each library (Fig. 1). This suggest this genotype to be in minority in the two libraries or being differentially amplified in the PCR. The 16S rDNA similarities were 98.7% to cluster 1 and 98.3-99.1 to cluster 2. Hca25 and Tvb03 were 99.8% similar and formed an early branch of the B. aalborgi lineage.

General considerations Human intestinal spirochetosis is a defined clinical syndrom (BARRETT, 1997), however the knowledge regarding which species that are involved in IS has remained largely unknown. The species reported in human IS are B. pilosicoli and B. aalborgi. While B. pilosicoli has been frequently isolated from IS patients, B. aalborgi has not. Several attempts failing to isolate B. aalborgi from clinical specimens have suggested that the recovery of this spirochete as performed by HOVIND-HOUGEN et ai. (1982), was a somewhat obscure event. This may be due to the longer time of incubation that is needed to cultivate B. aalborgi as compared to B. pilosicoli and perhaps that inappropriate media have been used for isolating B. aalborgi. Also, further prerequisites for isolation and growth of the phylotypes found in this study might be needed. Therefore, PCR based strategies for detecting intestinal spirochetes are of high importance and a suitable target is the primary structure of the 16S rRNA molecule. Recently, a set of paraffin embedded tissue samples from human mucosa was subjected to PCR amplification with primers specific for the rrs and nox genes of B. aalborgi and B. pilosicoli. The presence of 16S rDNA originating from intestinal spirochetes could be confirmed from 11 out of 17 biopsies with histological evidence for IS (MIKOSZA et ai., 1999). Three of these amplicons were sequenced revealing a 16S rRNA gene identity to B. aalborgi. However, the PCR assay did not generate amplicons from six of these samples, despite clear signs of IS. It was believed that the 16S rDNA information for primer design was insufficient having only one 16S rDNA sequence at hand, thus hampering the detection of B. aalborgi. In this study we have found that a vast variety of brachyspiral phylotypes might be present in the human intestinal mucosa supporting that some phylotypes might escape being detected when using PCR primers that targets a region in which sequence variation is pronounced. However, in order to reveal whether the phylotype of the B. aalborgi sub-cluster la (Fig. 1) is dominating or not, this has to be shown by e.g. using probe-based assays. The diagnostic problems using cultivation are of course most pronounced when two or more species are represented in a clinical sample, thus enriching for and falsely detecting the species B. pilosicoli only. This supports the recent findings that B. aalborgi and spirochetes other than B. pilosicoli are likely to be far more common in the intestinal mucosa than cultivation efforts previous-

361

ly have been able to prove (HENRIK-NIELSEN et ai., 1983; MIKOSZA et ai., 1999; KRAAZ et ai., 2000). Improved media for cultivation, if not being selective, might not necessarily solve this problem. Despite knowing whether the plausible organisms as phylotyped in this study are able to cause IS, the data presented will nevertheless facilitate the use of molecular biology techniques to verify the presence of B. aalborgi-like organisms in mucosal biopsies and in fecal samples. Nevertheless, they are believed to attach to the mucosa, because the biopsies were thoroughly washed prior to DNA preparation. Interestingly, FELLSTROM et ai. (unpublished) recently obtained 16S rRNA gene fragments about 200 nt in length from paraffinated samples, suggesting the presence of phylotypes representing all clusters as defined in this study. Evolutionary, the members of the B. hyodysenteriae and the B. aalborgi lineages are closely related (Fig. 1, PETTERSSON et ai., 1996, STANTON et ai., 1996). Considering the species of the B. hyodysenteriae lineage, there are six validly described species namely B. hyodysenteriae, B. intermedia, B. innocens, B. alvinipulli, B. murdochii and B. pilosicoli (Table 1, Fig. 1). These were regarded as separate species as judged from e.g. multi locus enzyme electrophoresis and DNA-DNA reassociation experiments, despite showing inter-species 16S rDNA similarity values of 97.8% (B. pilosicoli to Brachyspira sp. C555) and up to 99.5% (B. hyodysenteriae to B. intermedia and to B. innocens). This is clearly above the generallimit of 97% in 16S rDNA similarity for species definition, as outlined by STACKEBRANDT and GOEBEL (1994). The clonal variation between the clusters 1, 2 and 3 as observed in this study is in parity with the 16S rRNA gene diversity found between the species and strains of the B. hyodysenteriae lineage. Because the B. hyodysenteriae lineage is relatively closely related to B. aalborgi (HOOKEY et ai., 1994, PETTERSSON et ai., 1996), one could speculate in that the tempo and mode of evolution of species between and within these two lines of descent would not differ too much from each other. Comparing 16S rDNA distance values and assuming that the evolutionary rates do not differ widely between the species of the B. hyodysenteriae lineage to those between the clusters within the B. aalborgi lineage, we have faced 16S rRNA gene data of at least two new species of the genus Brachyspira to be isolated and characterized in the future. Furthermore, these data bring fuel into the debate of retaining the genus name Serpulina for members belonging to the B. hyodysenteriae lineage (to be discussed elsewhere).

Conclusion This is the first study describing that different phylotypes of B. aalborgi-like organisms are present in the distal colonic mucosa within and between individuals without clinical symptoms of IS. Moreover, the description of the phylogenetic neighborhood of B. aalborgi was significantly improved aiding as useful guidelines in the taxonomic and diagnostic work on the putatively clinic im-

362

B. PETTERSSON et a!.

portant organisms to be discovered and which belong to the B. aalborgi lineage. Acknowledgement B.P. is indebted to the Swedish Foundation for Strategic Research. We thank Ausi Boholm and Helena Ronning for excellent technical assistance.

References ALTSCHUL, S. E, MADDEN, T. L., SCHAFFER, A. A., ZHANG, J., ZHANG, Z., MILLER, W., LIPMAN, D. J.: Gapped BLAST and PSI-BLAST: a new generation of protein database searchprograms. Nucleic Acids Res. 25, 3389-3402 (1997). ATYEO, R. E, OXBERRY, S. L., HAMPSON, D. ].: Pulse-field gel electrophoresis for sub-specific differentiation of Serpulina pilosicoli (formely "Angulina coli"). FEMS Microbio!. Lett. 141,77-81 (1996). BARRETT, S. P. Human intestinal spirochaetosis, pp. 243-266. In: Intestinal spirochaetes in domestic animals and humans (D. ]. Hampson, T. B. Stanton, eds.) Wallingford, England, CAB International 1997. BROSIUS, J., PALMER, ]. L., KENNEDY, J. P. & NOLLER, H. E: Complete nucleotide sequence of a 16S ribosomal RNA gene from Escherichia coli. Proc. Nat!' Acad. Sci. USA 75, 4801-4805 (1978). DAVELAAR, E G., SMIT, H. E, HOVIND-HoUGEN, K., DWARS, R. M., VAN DER VALK, P. c.: Infectious typhlitits in chickens caused by spirochetes. Avian Patho!' 15,247-258 (1986). DUHAMEL, G. E.: Intestinal spirochaetes in non-production animals, pp. 301-320. In: Intestinal spirochaetes in domestic animals and humans (D. ]. Hampson, T. B. Stanton, eds.) Wallingford, England, CAB International 1997. FELLSTROM, c., PETTERSSON, B., UHLEN, M., GUNNARSSON, A., JOHANSSON, K.-E.: Phylogeny of Serpulina based on sequence analyses of the 16S rRNA gene and comparison with a scheme involving biochemical classification. Res. Vet. Sci. 59,5-9 (1995). FELSENSTEIN, J.: PHYLlP: Phylogeny inference package (version 3.572). Department of Genetics, University of Washington, Seattle, WA, USA (1993) . HARLAND, W. A., LEE, E D.: Intestinal spirochaetosis. Br. Med. J. 3, 718-719 (1967). HARRIS, D. L., GLOCK, R. D., CHRISTENSEN, C. R., KINYON, J. M.: Swine dysentery I. Inoculation of pigs with Treponema hyodysenteriae (new species) and reproduction of the disease. Vet. Med. Small Anim. Clin. 67, 61-64 (1972). HENRIK-NIELSEN, R., ORHOLM, M., PEDERSEN, J. 0., HOVINGHOUGEN, K., TEGLBjAORG, THAYSEN, E. H. Colorectal spirochetosis: clinical significance of the infestation. Gastroenterology 85, 62-67 (1983). HOOKEY,]. v., BARRETT, S. P., REED, C. S., BARBER, P.: Phylogeny of human intestinal spirochaetes inferred from 16S rDNA sequence comparison. FEMS Microbio!. Lett. 117, 345-350 (1994). HOVIND-HoUGEN, K., BIRCH-ANDERSEN, A., HENRIK-NIELSEN, R., ORHOLM, M., PEDERSEN, ]. U., TEGLBjAERG, P. S., THAYSEN, E. H.: Intestinal spirochetosis: morphological characterization and cultivation of the spirochete Barchyspira aalborgi gen. nov., sp. nov. J. Clin. Microbio!. 16, 1127-1136 (1982). HULTMAN, T., BERGH, S., MOKS., T., UHLEN, M.: Bidirectional solid phase sequencing of in vitro-amplified plasmid DNA. BioTechniques 10, 84-93 (1991). JUKES, T. H., CANTOR, C. R.: Evolution of protein molecules. In

Mammalian protein metabolism, pp. 21-132. Edited by H. N. Munro. New York: Academic Press (1969). KINYON, J. M., HARRIS, D. L.: Treponema innocens, a new species of intestinal bacteria, and emended description of the type strain Treponema hyodysenteriae HARRIS et a!. Int ]. Syst. Bacterio!' 29,102-109 (1979). KRAAZ, W., PETTERSSON, B., THUNBERG, U., ENGSTRAND, L., FELLSTROM, C.: Unpublished LEE,]. I., HAMPSON, D. ].: Genetic characterisation of intestinal spirochaetes and their association with disease. ]. Med. Microbio!' 40, 365-371 (1994). McLAREN, A. J., TROTT, D. J., SWAYNE, D. E., OXBERRY, S. L., HAMPSON, D. J.: Genetic and phenotypic characterization of intestinal spirochetes colonizing chickens and allocation of known pathogenic isolates of three distinct genetic groups. ]. Clin. Microbio!. 35, 412-417 (1997). MIKOSZA, A. S. J., LA, T., BROOKE, J., LINDBOE, C. E, WARD, P. B., HEINE, R. G., GUCCION, ]. G., BASTIAAN DE BOER, W., HAMPSON, D. J.: PCR amplification from fixed tissue indicates frequent involvement of Brachyspira aalborgi in human intestinal spirochetosis. ]. Clin. Microbio!. 37, 2093-2098 (1999). PASTER, B. J., DEWHIRST, E E., WEISBURG, W. G., TORDOFF, L. A., FRASER, G. ]., HESPELL, R. B., STANTON, T. B., ZABLEN, L., MANDELCO, L., WOESE, C. R.: Phylogenetic analysis of the spirochetes. ]. Bacterio!' 173,6101-6109 (1991). PETTERSSON, B., FELLSTROM, c., ANDERSSON, A., UHLEN, M., GUNNARSSON, A., JOHANSSON, K.-E.: The phylogeny of intestinal porcine spirochetes (Serpulina species), based on sequence analysis of the 16S rRNA gene. ]. Bacterio!' 178, 4189-4199 (1996). SAITOU, N., NEI M.: The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mo!. Bio!. Evo!. 4, 406-425 (1987). SMITH, S.: GDE: Genetic data environment, Version 2.2. Millipore Imaging systems, Ann Arbor, MI (1992). STACKEBRANDT, E., GOEBEL, B. M.: Taxonomic note: a place for DNA-DNA reassociation and 16S rRNA sequence analysis in the present definition in bacteriology. Int. J. Syst. Bacterio!' 44, 846-849. STANTON, T. B., FOURNIE-AMAZOUZ, E., POSTIC, D, TROTT, D. J., GRIMONT, P. A. D., BARANTON, G., HAMPSON, D. J., SAINT GIRONS, I.: Recognition of two new species of intestinal spirochetes Serpulina intermedia sp. nov. and Serpulina murdochii sp. nov. Int. J. Syst. Bacteriol. 47,1007-1012 (1997). STANTON, T. B., JENSEN, N. 5., CASEY, T. A., TORDOFF, L. A., DEWHIRST, E E., PASTER, B. J.: Reclassification of Treponema hyodysenteriae and Treponema innocens in a new genus, Serpula gen. nov., as Serpula hyodysenteriae comb. nov. and Serpula innocens comb. nov. Int. J. Syst. Bacteriol. 41, 50-58 (1991 ). STANTON, T. B., POSTIC, D, JENSEN, N. S.: Serpulina alvinipulli sp. nov., a new Serpulina species that is enteropathogenic for chickens. Int.]. Syst. Bacteriol. 48, 669-676 (1998). STANTON, T. B., TROTT, D. J., LEE, ]. I., McLAREN, A. J., HAMPSON, D. J., PASTER, B. ]., JENSEN, N. S.: Differentiation of intestinal spirochaetes by multilocus enzyme electrophoresis analysis and 16S rRNA sequence comparisons. FEMS MicrobioI. Lett. 136, 181-186 (1996). SUAU, A., BONNET, R., SUTREN, M., GOD ON, ].-M., GIBSON, G. R., COLLINS, M. D., DORE,]. Direct analysis of genes encoding 16S rRNA from complex communities reveals many novel molecular species within the human gut. App!. Environ. 65,4799-4807 (1999). SWAYNE, D. E., McLAREN A. ].: Avian intestinal spirochaetes and avian intestinal spirochaetosis, pp. 267-300. In: Intestinal spirochaetes in domestic animals and humans (D. ].

Intestinal Spirochetes in the Human Colonic Mucosa Hampson, T. B. Stanton, eds.) Wallingford, England, CAB International 1997. TAYLOR, D. ]., SIMMONS, j. R., LAIRD, H. M.: Production of diarrhoea and dysenteriae in pigs by feeding pure cultures of a spirochaete differeing from Treponema hyodysenteriae. Vet. Rec. 106,326-332 (1980). TAYLOR, D. ]., TROTI, D. J.: Porcine intestinal spirochaetosis and spirochaetal colitis, pp. 211-242. In: Intestinal spirochaetes in domestic animals and humans (D. J. Hampson, T. B. Stanton, eds.) Wallingford, England, CAB Internationa I 1997. TRIVETI-MoORE, N. L., GILBERT, G. L., LAW, C. L. H., TROTI, D. ]., HAMPSON, D. J.: Isolation of Serpulina pilosicoli from rectal biopsy speciemens showing evidence of intestinal spirochetosis. J. Clin. Microbiol. 36, 261-265 (1998). TROTI, D. J., COMBS, B. G., MIKOSZA, A. S. ]., OXBERRY, S. L., ROBERTSON, I. D., PASEY, M., TAIME, j., SEHUKO, R., ALPERS,

363

M. P., HAMPSON, D. j.: The prevalence of Serpulina pilosicoli in humans and domestic animals living in the Eastern Highlands of Papua Guinea. Epidemiol. Infect. 119, 369-379 (1997). TROTI, D. ]., STANTON, T. B., JENSEN, N. S., DUHAMEL, G. E., JOHNSON, J. L., HAMPSON, D. J.: Serpulina pilosicoli sp. nov., the agent of porcine intestinal spirochetosis. Int. J. Syst. Bacteriol. 46, 206-215 (1996) . ZUERNER, R. L., STANTON, T. B.: Physical and genetic map of the Serpulina hyodysenteriae B78 T chromosome. j. Bact. 176, 1087-1092 (1994). Corresponding author: BERTIL PETIERSSON, Department of Biotechnology, Royal Institute of Technology, S-1004 Stockholm, Sweden Tel.: +44 8 790 82 87; Fax: +44 8 24 54 52 e-mail: [email protected]