FEMS Microbiology Ecology 34 (2000) 17^26
www.fems-microbiology.org
Symbiotic spirochetes in the termite hindgut: phylogenetic
identi¢cation of ectosymbiotic spirochetes of oxymonad protists
Toshiya Iida
a
a;
*, Moriya Ohkuma
a;b
, Kuniyo Ohtoko
a;c
, Toshiaki Kudo
a;b
Microbiology Laboratory, RIKEN (The Institute of Physical and Chemical Research), Wako, Saitama 351-0198, Japan
b
Bio-Recycle Project, Japan Science and Technology Corporation (JST), Wako, Saitama 351-0198, Japan
c
Department of Applied Chemistry, Toyo University, Kawagoe, Saitama 350-8585, Japan
Received 3 February 2000; received in revised form 26 July 2000 ; accepted 28 July 2000
Some species of protists inhabiting the hindgut of lower-termites have a large number of ectosymbiotic spirochetes on the cell surface. The
phylogenetic positions of the ectosymbiotic spirochetes of three oxymonad protists, Dinenympha porteri in the gut of Reticulitermes speratus,
and Pyrsonympha sp. and Dinenympha sp. in Hodotermopsis sjoestedti, were investigated without cultivation of these organisms. Protist
fractions carefully collected with a micromanipulator were used as templates for the amplification of small subunit ribosomal RNA genes
(SSU rDNA). The phylogenetic tree inferred from the nucleotide sequences of the SSU rDNA showed that they were affiliated with the
Treponema cluster of spirochetes and they were divided into two clusters. One was grouped together with the spirochetal sequences reported
previously from the gut of termites and the other was related to the Treponema bryantii subgroup of treponemes (denoted as termite
Treponema clusters I and II, respectively). Whole-cell in situ hybridization using a fluorescent-labeled oligonucleotide probe specific for the
group of sequences in cluster II identified most of the ectosymbiotic spirochetes of the oxymonad protists in the gut of R. speratus and H.
sjoestedti. However, not all of the ectosymbiotic spirochetes could be detected by means of this cluster II group-specific probe and the
population of ectosymbiotic spirochetes of cluster II was different among the oxymonad species. In the case of D. porteri, an oligonucleotide
probe specific for one member of cluster II recognized a portion of the ectosymbiotic spirochetes of cluster II, and their population was also
different depending on the cell-type of D. porteri in terms of the attachment of ectosymbiotic spirochetes. The results indicate that the
spirochetes of cluster II and probably those of a part of cluster I can be assigned to ectosymbiotic species of oxymonad protists and that the
population of ectosymbiotic spirochetes associated with a single protist consists of at least three species of phylogenetically distinct
spirochetes. ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
Keywords : Termite; Ectosymbiosis; Spirochete; Phylogeny; Whole-cell in situ hybridization
1. Introduction
Spirochetes are unique, morphological bacteria that are
widespread in several environments, either as free-swimming cells, or associated with other organisms, and some
are pathogenic to animals and humans. One of the environments rich in spirochetes is the gut £uid of xylophagous insects such as termites and wood-feeding cockroaches [1]. The hindgut microbes of the lower-termites
include both amitochondriate protists and bacteria which
may support lignocellulose digestion and provide metabolites to the host termite [2,3]. A number of spirochete-like
bacteria of several sizes (3^100 Wm in length, 0.2^1.0 Wm in
* Corresponding author. Tel. : +81 (48) 462 1111 ext. 5724;
Fax: +81 (48) 462 4672; E-mail : tiida@postman.riken.go.jp
width) are conspicuously present in the gut £uid of several
termites [1]. These observations suggest that elucidation of
the spirochete's functions in the microbial community in
the gut of termites is important to understand this complex symbiotic system. However, our understanding of the
physiological ecology of spirochetes has been limited. Only
recently, pure cultures of termite gut spirochetes have been
obtained [4]. These cultivable termite spirochetes (Treponema sp. strains ZAS-1 and ZAS-2) display CO2 -reducing
acetogenesis activity. This might account for the importance of symbiotic spirochetes for survival of the host
termite [5].
It is well known that ectosymbiotic spirochetes are
present on the cell surface of some species of termite symbiotic protists [1,6^9]. Ultrastructural analysis of the protists has revealed that the ectosymbiotic spirochetes are
embedded in the protist's membrane via special structures
0168-6496 / 00 / $20.00 ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 8 - 6 4 9 6 ( 0 0 ) 0 0 0 7 0 - 2
FEMSEC 1166 18-10-00
Cyaan Magenta Geel Zwart
Downloaded from http://femsec.oxfordjournals.org/ by guest on March 5, 2016
Abstract
18
T. Iida et al. / FEMS Microbiology Ecology 34 (2000) 17^26
2. Materials and methods
2.1. Termites
Two lower-termites were used in this study. A subterranean termite, Reticulitermes speratus (Rhinotermitidae),
was collected in the vicinity of Ogose, Saitama Prefecture,
Japan. A damp-wood termite, Hodotermopsis sjoestedti
(Termopsidae), was collected on Yakushima Island, Kagoshima Prefecture, Japan. Termite-infested wood samples
moistened with distilled water were kept in plastic boxes at
room temperature (20^25³C) for rearing.
2.2. Preparation of ¢xed protists
The guts of pseudergate-termites were pulled out with
FEMSEC 1166 18-10-00
sterilized forceps, suspended in solution U (2.164 g NaCl,
0.773 g NaHCO3 , 1.509 g Na3 C6 H5 O7 W2H2 O (citrate),
1.784 g KH2 PO4 , 0.083 g CaCl2 and 0.048 g MgSO4 dissolved in 1 l of distilled water [24]), and squeezed gently.
The gut debris was removed by means of nylon-mesh
(mesh size = 160 Wm), and the cells were ¢xed in solution
U containing 4%-neutralized formaldehyde overnight at
4³C. After washing with solution U, the protists were suspended in the appropriate solution, as described below.
For physical separation of a certain protist cell from
the suspension, a combination of a microscope (Leica,
DMIRB) and a micromanipulator (Eppendorf, TransferMan) was used, together with a handmade microcapillary
which ¢t the target cells. The separated cells were suspended in fresh solution and the micromanipulation was
repeated three times to remove contaminating cells.
For identi¢cation of the protists which have associated
spirochetes, the ¢xed protists were stained with 0.2 Wg
ml31 4P,6-diamidino-2-phenylindole (DAPI) and examined
using an epi£uorescence microscope (Olympus, AX-70)
equipped with a UV ¢lter.
2.3. Polymerase chain reaction (PCR) ampli¢cation and
cloning
Small subunit ribosomal RNA (SSU rRNA) genes were
ampli¢ed directly from the protists isolated by micromanipulation, by PCR using ExTaq DNA polymerase (Takara) according to the manufacturer's instructions. As
PCR primers, the forward primer (27F ; 5P-AGAGTTTGATCCTGGCTCAG-3P) corresponded to nucleotide positions 8 to 27 of Escherichia coli SSU rRNA and the
reverse primer (1492R; 5P-GGCTACCTTGTTACGACTT-3P) corresponded to E. coli SSU rRNA positions
1510 to 1492 [25]. The reaction conditions were as follows :
25 (Dinenympha porteri and Pyrsonympha sp.) or 35 cycles
(Dinenympha sp.) after 8 min incubation at 94³C; each
cycle consisted of 94³C for 20 s, 54³C for 20 s, and
72³C for 60 s, then the last cycle was followed by incubation at 72³C for 5 min. PCR products corresponding to
bacterial rDNA of the expected size were puri¢ed by electrophoresis on an agarose gel and cloned into the pGEMT vector (Promega).
2.4. Nucleotide sequencing and phylogenetic analysis
Plasmid DNA was isolated as described by Sambrook et
al. [26] and used as a template for sequencing by means of
the ABI PRISM BigDye Terminator Cycle Sequencing
Ready Reaction Kit (Applied Biosystems) with primers
for SSU rDNA [21]. The sequencing reaction was analyzed
using an automatic sequence analyzer (ABI sequencer
model 377). GENETYX software (Software Development)
was used for general analysis of the nucleotide sequences.
Sequence data were aligned using the CLUSTAL W package [27], then corrected by manual inspection and the nu-
Cyaan Magenta Geel Zwart
Downloaded from http://femsec.oxfordjournals.org/ by guest on March 5, 2016
[10^13]. Speculation regarding the functions of these ectosymbiotic spirochetes has been made but no clear evidence
has been published except for one report [14]. Mixotricha
paradoxa, a symbiotic protist of the termite Mastotermes
darwiniensis, is a large-protist (500 Wm in length and 250
Wm in diameter) and it has a large number of ectosymbiotic spirochetes (10 Wm in length and 0.15 Wm in diameter) attached over the entire surface of the cell body. They
suggested that the ectosymbiotic spirochetes might contribute to the motility of the host cell, a relationship
known as `motility symbiosis'. However, most of the ectosymbionts observed so far might not be linked with protist
motility [1,11,15,16].
Although many protist species with ectosymbiotic microbes have been found, the molecular phylogenetic positions of the ectosymbionts have not yet been determined
due to di¤culty in separation of the organisms or due to
di¤culty in cultivation of the host-protists in pure form.
Recent studies focusing on molecular phylogenetic analysis of the termite gut microbes have shown that the symbiotic spirochetes are grouped in the Treponema species
but none are closely related to any known species of Treponema [17^23]. Some spirochete sequences were identi¢ed
by in situ hybridization as being derived from large spirochetes existing freely in the gut £uid [18,23]. Recently,
detailed examination of the symbiotic spirochetes of Reticulitermes £avipes has been reported and at least 21 novel
phylotypes of Treponema species were found within the
gut community of this termite alone [19]. These sequences
obtained from lower-termites might be a mixture of sequences derived from either free-swimming or ectosymbiotic spirochetes, but the species could not be discerned
from the nucleotide sequences. In this study, we analyzed
16S rRNA sequences of ectosymbiotic spirochetes on the
oxymonad protists isolated by micromanipulation, and
then used oligonucleotide probes in whole-cell in situ hybridization to identify the ectosymbiotic spirochetal a¤liations.
T. Iida et al. / FEMS Microbiology Ecology 34 (2000) 17^26
cleotide positions of ambiguous alignment were omitted
from subsequent phylogenetic analyses. The programs
used to infer phylogenetic trees were those in the PHYLIP
package version 3.5c [28]. DNADIST was used to calculate evolutionary distances. Phylogenetic trees were reconstructed from evolutionary distance data by the neighborjoining method, implemented through the program
NEIGHBOR. A total of 100 bootstrapped replicate resampling data sets for DNADIST were generated with
the program SEQBOOT, to provide con¢dence estimates
for tree topologies.
2.5. Whole-cell in situ hybridization
b
2.6. Nucleotide sequence accession numbers
The sequence data reported in this paper will appear in
the DDBJ, EMBL and GenBank nucleotide sequence databases under the accession numbers from AB031997 to
AB032009.
FEMSEC 1166 18-10-00
3. Results
3.1. Identi¢cation of the protists with attached spirochetes
The protists harboring ectosymbiotic spirochetes were
identi¢ed on the basis of morphological characteristics
by epi£uorescence microscopy after staining the DNA
with DAPI. Among the oxymonad protists of R. speratus,
several Dinenympha species including D. porteri, Dinenympha leidyi, and Dinenympha parva [32] have ectosymbiotic
spirochetes on the posterior end of the cell, and some of
them also have a few spirochetes on the anterior end of
the cell. A large number of spirochetes was observed over
the entire body of D. porteri. These ectosymbiotic spirochetes might be classi¢ed as type III or type IV, which
di¡er in their patterns of attachment. In the case of the
former (type III), the ectosymbionts are di¡usely distributed over the entire body, being especially dense on the
posterior end of the cell and, normally, large quantities of
wood particles are incorporated into the cell. In the latter
case (type IV), the ectosymbiotic spirochetes are not distributed di¡usely on the cell surface but dozens of the
spirochetes are aggregately attached and display a bundled
morphology [32]. In the hindgut community of H. sjoestedti, the presence of at least six species of oxymonad
protists has been reported, although speci¢c epithets for
them have not yet been given [33]. Some of them harbor
ectosymbiotic spirochetes at the posterior end of the cell.
A large number of ectosymbiotic spirochetes cover the
entire cell surface of one species of Dinenympha (50^70
Wm in length, 10 Wm in width) and that of the relatively
large-sized Pyrsonympha sp. (100^150 Wm in length, 30^40
Wm in width). We chose three species of oxymonad protists on which dense populations of attached spirochetes
over the entire cell body were consistently observed (D.
porteri of R. speratus, and Dinenympha sp. and Pyrsonympha sp. of H. sjoestedti) for molecular phylogenetic analysis of the ectosymbiotic microbes. Because two types of D.
porteri (type III and IV) were usually indistinguishable in
the ¢xed cell preparations [32], a mixture of the two types
was used for analysis of the D. porteri ectosymbionts.
3.2. Molecular phylogenetic analysis of the ectosymbiotic
spirochetes
Approximately 30 of each protist carefully isolated by
micromanipulation were used directly for preparation of
PCR templates and the SSU rDNA of symbiotic bacteria
were ampli¢ed by means of eubacterial universal primers
(27F and 1492R). The clones of ampli¢ed DNA fragments, randomly selected, were classi¢ed on the basis of
restriction fragment length polymorphism (RFLP) using
Sau3AI and HaeIII (D. porteri and Pyrsonympha sp.) or
Sau3AI and A£I (Dinenympha sp.). We classi¢ed the
clones into ten, seven and seven RFLP groups in the
case of those derived from D. porteri, Pyrsonympha sp.
Cyaan Magenta Geel Zwart
Downloaded from http://femsec.oxfordjournals.org/ by guest on March 5, 2016
Fixed hindgut contents (from 20^40 guts of R. speratus
and from three of H. sjoestedti) were incubated in 1 ml of
0.25 M HCl solution for 30 min at room-temperature, and
then washed with phosphate-bu¡ered saline (PBS) (136.89
mM NaCl, 2.68 mM KCl, 8.1 mM Na2 HPO4 12H2 O, 1.47
mM KH2 PO4 ) and with hybridization solution (20 mM
Tris^HCl (pH 7.2), 0.9 M NaCl, 0.01% SDS) without
SDS. The cells were resuspended in 50 Wl of hybridization
solution and incubated for 30 min at 37³C with £uorescent-labeled probes (0.5^2.0 pmol Wl31 , described below).
After washing with PBS twice, the cells were resuspended
in anti-fading solution (0.5% triethylenediamine in 90%
glycerol in PBS) and mounted on a glass-slide. Fluorescence signals were detected using an Olympus epi£uorescence microscope (AX-70) ¢tted with ¢lter sets for £uorescein (WIB) and Texas Red (WIY). The hybridization
signals were also detected by confocal laser scanning microscopy (Leica). The sequences of the £uorescent-labeled
probes used in this study were: EUBAC (5P-GCTGCCTCCCGTAGGAGT-3P), TT-484V3 (5P-TTGCTTATTCAAACCCTACC-3P), TT-1248V8 (5P -CTGCTTCGCWTCGCTCTGT-3P), RsDiSp1-638V4 (5P-ATTCAAGTATGAAAGTTCCC-3P) and RsDiSp3-638V4 (5P-CTCAAGTCACATAGTTCTC-3P). These probes recognize sequences within helices 15, V3, V8, V4 and V4, respectively,
as determined by reference to a model of the secondary
structure of prokaryotic SSU rRNA [29]. EUBAC binds
to most eubacterial cells [30]. The other four probes are
speci¢c for the clones obtained in this study and no completely matching sequences were found on the RDP database by the PROBE MATCH program [31]. The probes
were labeled in the 5P position with either FAM (6-carboxy£uorescein) or Texas Red.
19
20
T. Iida et al. / FEMS Microbiology Ecology 34 (2000) 17^26
Downloaded from http://femsec.oxfordjournals.org/ by guest on March 5, 2016
Fig. 1. Neighbor-joining tree of the SSU rDNA sequences of spirochetes from several termites including the sequences of ectosymbionts of oxymonad
protists. Bootstrap values above 50 from 100 resamplings are shown for each of the nodes. The scale bar indicates 0.01 nucleotide substitution per position. The sequences of clone names highlighted on the gray background were ampli¢ed by PCR in this study. Asterisks denote the sequences for which
the cells of origin were identi¢ed in situ by whole-cell hybridization in this study and those described previously (NL1 [23]; Mdmpsp15 [18]).
FEMSEC 1166 18-10-00
Cyaan Magenta Geel Zwart
T. Iida et al. / FEMS Microbiology Ecology 34 (2000) 17^26
Table 1
Populations of PCR clones as judged on the basis of RFLP and partial
nucleotide sequences of representative clones
Target protist
Number
of clones
Cluster I
D. Porteri
Pyrsonympha sp.
Dinenympha sp.
31
18
21
8
5
16
Cluster II
IIA
IIB
14
5
0
9
8
5
FEMSEC 1166 18-10-00
tence of two subclusters of cluster II, supported by high
bootstrap values (100%), was evident, and these were designated as `Termite Treponema clusters IIA and IIB' (Fig.
1). The sequences in both subclusters were those obtained
from micromanipulated fractions of D. porteri and Pyrsonympha sp. but only the sequences in cluster IIB could be
detected in the case of Dinenympha species. The average
interspecies similarity among the members of clusters IIA
and IIB was 96.8 and 97.3%, respectively, and the average
similarity between these subclusters was 91.7%. The signature bases of the T. bryantii subgroup and RFS9 or RFS12
(cluster II members of R. £avipes symbionts) have been
reported [19,34]. These signature nucleotides were well
conserved among the sequences of the subclusters. The
sequence of the Mastotermes darwiniensis clone sp40-12
was grouped with cluster II, as tenuously supported by
bootstrap analysis (79%). We excluded the sequence
sp40-12 from cluster II in this report, because it represented a lineage distinct from subclusters IIA and IIB,
and the average interspecies similarity among the members
of these subclusters was 85.9%. The signature bases of
sp40-12 were also conserved except for position 1038.
Some of the spirochetal sequences ampli¢ed from the
mixed population DNA of R. speratus [20] and H. sjoestedti [22] were closely similar to the sequences cloned in
this study (RsDiSp1 and Rs100 ; 99%, RsDiSp3 and Rs2 ;
99%, RsDiSp8 and Rs21; 96%, and HsPySp1 and Hs33;
99%).
3.3. In situ identi¢cation of the ectosymbiotic spirochetes
We designed £uorescent-labeled oligonucleotide probes
for in situ identi¢cation of strains possessing the SSU
rRNA sequences. Both NL1 of Nasutitermes lujae and
mpsp15 of Mastotermes darwiniensis in cluster I have
been identi¢ed as sequences derived from free-swimming,
large spirochetes in the gut £uid by whole-cell in situ hybridization [18,23], but none of the sequences belonging to
cluster II have been identi¢ed. We designed consensus
probes targeting members of cluster II, which corresponded to variable regions 3 (TT-484V3) and 8 (TT1248V8) of the SSU rDNA sequences. These probes completely matched the sequences of cluster II obtained in this
study, but not any of the sequences in cluster I. In wholecell in situ hybridization of gut £uid, weak hybridization
signals were apparent in the case of ectosymbiotic spirochetes of D. porteri and Pyrsonympha sp. using each of the
cluster II consensus probes. We used both probes simultaneously for further analysis to obtain stronger signal intensity. Fig. 2 shows the results of whole-cell in situ hybridization using the cluster II consensus probes labeled
with FAM, and a eubacterial universal probe labeled with
Texas Red. The Texas Red signals were observed inside or
on the periphery of the cell body of several oxymonad
protists, probably derived from symbiotic or incorporated
bacteria. Simultaneously, green FAM signals were ob-
Cyaan Magenta Geel Zwart
Downloaded from http://femsec.oxfordjournals.org/ by guest on March 5, 2016
and Dinenympha sp., respectively, and partial nucleotide
sequences (ca. 600 bp) of at least two-clones in the major
RFLP groups (six, four and four major RFLP groups,
respectively) were determined. Clones in minor RFLP
groups that consisted of only a single clone were excluded
from further nucleotide sequence analysis. Based on sequence similarity, we chose 13 representative clones and
the complete nucleotide sequences (ca. 1.5 kb) of these
clones were determined.
Fig. 1 shows a phylogenetic tree constructed by the
neighbor-joining method, including the sequences previously reported from several termites [4,17^19,23]. All of
the sequences obtained from the micromanipulated fractions were a¤liated with the Treponema cluster of spirochetes [34], together with those previously reported from
termites, and this was supported by high bootstrap probability (100%). In the phylogenetic tree, the sequences
from termites formed two large clusters, designated as
`termite Treponema clusters I and II' (hereafter written
as clusters I and II). Cluster I which was slightly supported
by bootstrap analysis (77%) included sequences of DNA
isolated previously from termite guts, Spirochaeta caldaria,
and S. stenostrepta, with an average interspecies similarity
of 90.2%. Cluster I also includes sequences obtained from
pure cultures of termite gut Treponema strains (ZAS-1 and
ZAS-2) [4]. The other cluster (cluster II) was closely related to rumen spirochetes (T. bryantii, Treponema saccharophilum and Treponema sp. strain CA), Treponema maltophilum, and Treponema pectinovorum, and this was
supported by high bootstrap probability (97%). The cluster of rumen spirochetes has been classi¢ed as a second
subgroup (T. bryantii subgroup) of the Treponema family
[34].
The sequences from the micromanipulated fractions
were found in either of the clusters. Approximately,
three-quarters of the sequences ampli¢ed from the micromanipulated protist fractions of D. porteri (74%) and Pyrsonympha sp. (72%) were grouped in cluster II (Table 1).
In the case of Dinenympha sp. of H. sjoestedti, 76% of the
clones constituted a part of cluster I. The cluster I sequences of the major RFLP groups ampli¢ed from each of the
protist fractions (RsDiSp8, HsPySp4 and HsDiSp320)
were grouped with the sequence RFS3, from R. £avipes,
and this was supported by high bootstrap probability
(100%). The sequence HsDiSp319 of the minor RFLP
group of cluster I (4/21) was not included in the group
containing RsDiSp8, HsPySp4 and HsDiSp320. The exis-
21
22
T. Iida et al. / FEMS Microbiology Ecology 34 (2000) 17^26
served over large portions of the ectosymbiotic spirochetes
of D. porteri (Fig. 2a,b), Pyrsonympha sp. (Fig. 2d) and
Dinenympha sp. (data not shown), and other oxymonad
protists in R. speratus (such as Fig. 2c) and H. sjoestedti
(data not shown). However, none of the unattached spirochetes in the gut £uid gave FAM signals (such as those
shown by arrowheads in panel a). These results suggested
that the spirochetes possessing the sequences grouped in
cluster II were ectosymbiotic spirochetes.
Confocal laser scanning microscopy was used for detailed examination of the patterns of association of the
phylogenetically discrete ectosymbiotic spirochetes (Fig.
3). The two types of D. porteri, type III (Fig. 3, panel 1)
and type IV (Fig. 3, panel 2), both covered with ectosymbiotic spirochetes, could be distinguished. The Texas Red
signals derived from the eubacterial universal probe were
apparent on the cell surface of the protists that were covered with the ectosymbiotic spirochetes (Fig. 3, panel b of
1 and 2), and in the case of these cells, hybridization sig-
nals were apparent also using the cluster II consensus
probes labeled with FAM (Fig. 3, panel a of 1 and 2).
The population of the members of each cluster was estimated (Table 2). The mean percentage of the ectosymbionts of D. porteri type III belonging to cluster II was
52.7 þ 16.5%, while that of D. porteri type IV was
78.3 þ 4.2%. The results concerning the symbionts of H.
sjoestedti are shown in Fig. 3, panels 3 and 4. In the case
of Pyrsonympha sp. (Fig. 3, panel 3), the Texas Red-labeled eubacterial universal probe hybridized with densely
attached spirochetes on the periphery of the cell and with
endosymbiotic or incorporated bacteria. At the same time,
not all, but a large portion of the ectosymbiotic spirochetes showed hybridization signals with the cluster II
consensus probes. On the contrary, the population of spirochetes belonging to cluster II on Dinenympha sp. showing hybridization signals was relatively small (Fig. 3, panel
4). The ratio of ectosymbionts belonging to cluster II in
the case of those of Pyrsonympha sp. and Dinenympha sp.
C
Fig. 3. Whole-cell in situ hybridization analysis visualized by confocal laser scanning microscopy. Cluster II consensus probes (TT-484V3 and TT1248V8, labeled with FAM) and a eubacterial universal probe labeled with Texas Red (panel 1^4) were used. In panels 5 and 6, a FAM-labeled eubacterial universal probe and the RsDiSp3-638V4 probe labeled with Texas Red were used. (a) FAM-signal, (b) Texas Red-signal, (c) di¡erential interference and (d) merging picture of (a) and (b) are shown. In panel (d), the doubly stained spirochetes appear orange, and wood particles incorporated by
these protists also gave an orange color. The scale bar in panel (d) represents 10 Wm. D. porteri type III (1 and 5), D. porteri type IV (2 and 6) in R.
speratus, and Pyrsonympha sp. (3), Dinenympha sp. (4) in H. sjoestedti are shown.
FEMSEC 1166 18-10-00
Cyaan Magenta Geel Zwart
Downloaded from http://femsec.oxfordjournals.org/ by guest on March 5, 2016
Fig. 2. Whole-cell in situ hybridization with cells doubly stained with the cluster II consensus probes (TT-484V3 and TT-1248V8, FAM) and the eubacterial universal probe (Texas Red), visualized by epi£uorescence microscopy. Phase-contrast (top), FAM-signal (middle), and Texas Red signal (bottom)
are shown. D. porteri (a and b) and an unidenti¢ed oxymonad protist (c) in R. speratus, and Pyrsonympha sp. (d) in H. sjoestedti were observed. The
scale bar represents 50 Wm. Arrowheads in panel (a) denote unattached spirochetes ; note that they were stained with the eubacterial universal probe
but not with the cluster II consensus probes. In contrast to the FAM-green signals, amorphous yellow in the middle panels and the corresponding red
signals in the bottom panels might be derived from ingested wood particles. Arrowheads in panel (c) denote the ectosymbiotic spirochetes associated
with the posterior end of the cell.
T. Iida et al. / FEMS Microbiology Ecology 34 (2000) 17^26
23
Downloaded from http://femsec.oxfordjournals.org/ by guest on March 5, 2016
FEMSEC 1166 18-10-00
Cyaan Magenta Geel Zwart
24
T. Iida et al. / FEMS Microbiology Ecology 34 (2000) 17^26
Table 2
Populations of ectosymbiotic spirochetes of protists as judged on the basis of whole-cell in situ hybridization
Target protist
Cluster II/Eubacteria
Cluster IIB/Cluster II
D. porteri type III
D. porteri type IV
Pyrsonympha sp.
Dinenympha sp.
52.7%
78.3%
72.6%
21.7%
55.1% ( þ 10.1)
74.9% ( þ 7.7%)
^
^
( þ 16.5%)
( þ 4.2%)
( þ 11.4%)
( þ 9.7%)
The numbers of ectosymbiotic spirochetes associated with D. porteri
type III, D. porteri type IV, Pyrsonympha sp., and Dinenympha sp. cells
were counted in 8, 8, 4 and 4 images, respectively, of several individuals
at several depths of focus.
4. Discussion
The molecular phylogenetic tree inferred from SSU
rDNA sequences ampli¢ed from micromanipulated protists revealed that these spirochetes were grouped in the
Treponema species, but none of them are closely related to
any known species of Treponema. The phylogenetic tree
FEMSEC 1166 18-10-00
Cyaan Magenta Geel Zwart
Downloaded from http://femsec.oxfordjournals.org/ by guest on March 5, 2016
was 72.6 þ 11.4% and 21.7 þ 9.7%, respectively (Table 2).
These ratios were similar to those of the clones isolated
from the micromanipulation fractions (Table 1).
As cluster II was divided into two subclusters, we prepared a probe speci¢c for each subcluster and used these
for whole-cell in situ hybridization. These probes are speci¢c for the SSU rDNA sequence of D. porteri in R. speratus. Although the ectosymbiotic spirochetes showed hybridization signals when treated with the eubacterial
universal probe, no £uorescent signal was observed when
they were treated with the probe speci¢c for subcluster
IIA. On the other hand, some portion of the ectosymbiotic
spirochetes of D. porteri showed hybridization signals
when treated with the probe speci¢c for subcluster IIB.
Fig. 3, panels 5 and 6 show the results of confocal laser
scanning microscopic analysis of the two types of D. porteri stained doubly with the eubacterial universal probe
(panel a, FAM-labeled) and the probe RsDiSp3-638V4
speci¢c for subcluster IIB (panel b, Texas Red). We also
observed hybridization signals using probes targeting the
cluster II consensus sequence and using RsDiSp3-638V4,
by confocal laser scanning microscopy (data not shown)
and the ratio of the members of the subclusters II among
the members of cluster II was estimated. The population
of the cluster IIB species was di¡erent comparing the two
types of D. porteri. Compared to the population of spirochetes on type IV cells of D. porteri, the population on
type III cells was smaller (panels 5 and 6). About
74.9 þ 7.7% of the members of cluster II associated with
D. porteri type IV were assigned to cluster IIB, but only
half (55.1 þ 10.1%) of the cluster II spirochetes on the
other type of D. porteri were assigned to cluster IIB. These
¢ndings suggest that the population of ectosymbiotic spirochetes on a single D. porteri cell consists of a mixture of
more than three species.
also revealed that two large clusters of termite symbiotic
spirochetes were formed, designated as termite Treponema
clusters I and II. Similar results have been reported for the
spirochete population of the whole-gut microbial community of several termites [17^23]. The results of wholecell in situ hybridization showed that spirochetal members
of cluster II were ectosymbiotic species of the oxymonad
protists. On the other hand, we could not identify the
members of cluster I by whole-cell in situ hybridization
as either ectosymbiotic or unattached species, although
consensus sequence-speci¢c probes and probes speci¢c
for the cluster I sequences were used. The higher-termites
harbor no symbiotic protists, therefore the spirochetes in
their hindgut should be free-swimming species in the gut
£uid. All spirochete sequences cloned from several highertermites are grouped in cluster I [22], including sequences
from two free-swimming spirochetes identi¢ed in situ by
whole-cell hybridization. These results suggest that at least
some of the spirochetal members of cluster I are freeswimming species in the gut £uid. However, we were
able to amplify several cluster I sequences from the micromanipulated protist fractions reproducibly. In the case of
Dinenympha sp., the most abundant ampli¢ed clone
(HsDiSp320; 57% of the clones) belonged to cluster I.
Closely related sequences were also ampli¢ed from Pyrsonympha sp. (HsPySp4) and D. porteri (RsDiSp8). These
results suggest that some of the cluster I species are also
ectosymbiotic spirochetes of the protists although we
could not completely deny the possibility that these cluster
I species corresponded to unattached spirochetes entangled among the ectosymbionts. Although we failed to
detect cells hybridized with cluster I consensus probes, a
group-speci¢c probe for Treponema species may clarify the
a¤liation of the bacterial species which did not hybridize
with cluster II consensus probes.
The relationships between oxymonad protists and their
ectosymbiotic spirochetes do not seem to be simple. At
least three species of phylogenetically distinct spirochetes
were found to be present on the cell surface of D. porteri
type III and type IV, which might include species of cluster I or presently unknown clusters (showing spirochete
morphology, and showing hybridization signals with the
eubacterial universal probe but no signals with the cluster
II consensus probes), species of the cluster IIA (showing
hybridization signals with the cluster II consensus probes
but not with the cluster IIB-speci¢c probe), and species of
cluster IIB (showing hybridization signals with the cluster
IIB-speci¢c probe). The populations of the three species of
ectosymbiotic spirochetes were di¡erent comparing those
associated with the type III and type IV D. porteri cells.
The population of the cluster IIB spirochetes associated
with the type IV cells was signi¢cantly larger than the
population associated with the type III cells, and the estimated population size of cluster I species was the reverse.
On the posterior end of the D. porteri type III cells, a
dense population of ectosymbiotic spirochetes was ob-
T. Iida et al. / FEMS Microbiology Ecology 34 (2000) 17^26
FEMSEC 1166 18-10-00
One of the proposed functions of the ectosymbiotic spirochetes is `motility symbiosis'. However, no clear relationship between the movements of the spirochetes and
the protists was observed in any spirochetes-associated
protists [1,11] except for Mixotricha paradoxa [14]. Recently, cultivation of symbiotic spirochetes (Treponema
sp. strains ZAS-1 and ZAS-2) associated with Zootermopsis angusticollis was reported [4]. The results indicated that
one of the spirochetes in the termite hindgut is a CO2 reducing acetogenic species ; they utilize H2 and CO2 to
produce acetate. Interestingly, the sequences from the cultivable spirochetes strains ZAS-1 and ZAS-2 are clustered
with the major cluster I sequences ampli¢ed from the micromanipulated protists, as supported by high-bootstrap
probability (96%). As discussed also by Leadbetter et al.
[4], the ectosymbiotic spirochetes may function as H2 and
CO2 consumers, absorbing these compounds from their
environment inside of the protist, and the driving force
for the symbiosis between the spirochetes and the protists
may be a £ow of reducing equivalents, as in the case of the
relationship between anaerobic ciliates and endosymbiotic
methanogenic archaea (reviewed in Embley and Finlay
[36]). With respect to the phylogenetic positions of the
other major ectosymbiotic spirochetes (cluster II), they
belong to the T. bryantii subgroup, and T. bryantii has
the characteristic ability to enhance the cellulolytic activity
of other microbes [37,38]. Although this aspect has not yet
been fully investigated, the ectosymbiotic spirochetes of
cluster II may function as such. To con¢rm these speculations, pure cultivation and biochemical analysis of the
protists which have associated spirochetes will be required.
Other than oxymonad protists, we found several interactions between protists and spirochete-like microbes in
the hindgut of diverse types of termites by DAPI staining
(our unpublished data). For example, a hypermastigote
protist, H. mirabile, in C. formosanus and a trichomonad,
Devescovina sp., in the Neotermes termite also each harbor
a dense population of spirochete-like ectosymbionts. In
order to understand the real nature of the symbiotic relationships between protists and spirochetes, molecular phylogenetic identi¢cation of the ectosymbiotic spirochetes
associated with these hypermastigotes and trichomonads
is necessary. Probably, the micromanipulation technique
used in this study is the only way to isolate speci¢c protists
from a mixed population, since pure cultures of these protists have not yet been obtained. The PCR-based cultureindependent approach and in situ identi¢cation of speci¢c
cells described here would be a powerful strategy to investigate the diversity and phylogenetic identity of the members of the termite gut microbial community.
Acknowledgements
We thank Yasue Ichikawa and the Biodesign DNA sequencing facility at RIKEN for assistance with nucleotide
sequencing. This work was partially supported by grants
Cyaan Magenta Geel Zwart
Downloaded from http://femsec.oxfordjournals.org/ by guest on March 5, 2016
served in some cases. Most of these posterior ectosymbionts showed no hybridization signal with the cluster II
consensus probes, suggesting that they were spirochete
species belonging to cluster I. However, ectosymbiotic
populations on other parts of the cell were dominated
by spirochete species of cluster II. For this reason, it
was evident that the spirochetal population of cluster I
on the type III cells was larger than that on the type IV
cells. The number of the posterior ectosymbionts on individual cells re£ects the large di¡erence in the populations
of spirochetes of cluster I and II associated with the D.
porteri type IV cells. Other oxymonad protists such as D.
porteri type I (which have associated ectosymbiotic spirochetes only on the posterior end of the cell), and Pyrsonympha sp. and Dinenympha sp. in H. sjoestedti also had
multiple species of ectosymbiotic spirochetes that di¡ered
in phylogenetic positions as determined through analysis
of SSU rRNA sequences. The factors determining the ectosymbiotic population present and their selection of association sites remain to be clari¢ed.
The unique ultrastructure of spirochete^protist attachment sites has been reported for several protists found in
termites and wood-feeding cockroaches. Two types of attachment structures have been identi¢ed so far in the symbiotic protists of R. £avipes, R. tivialis, and Cryptocercus
punctulatus by electron microscopic analysis [11^13,35].
One is a narrow nose-like appendage that makes contact
directly with the plasma membrane of the host cell, and
the other is a £attened end of the spirochete in contact
with the protistan membrane and it contains a thick layer
of electron dense material. It is of interest to analyze the
relationship between the attachment structures and the
phylogenetic positions proposed in this report.
Recently, sequences of SSU rRNA genes from spirochetes in the gut of lower-termites (R. £avipes and C. formosanus) have been published [19]. The results showed
that about half of the spirochetal clones ampli¢ed from
the gut of R. £avipes fell into three groups with the sequences RFS9, RFS3 and RFS12, respectively. Sequences
related to these groups were also ampli¢ed abundantly
from the micromanipulated oxymonad fractions of R.
speratus and H. sjoestedti. In the hindgut community of
R. £avipes, several oxymonad protists were also found to
harbor ectosymbiotic spirochetes. Our results suggest that
the spirochete species belonging to the RFS9 and RFS12
groups (and potentially the RFS3 group) should be classi¢ed as ectosymbiotic species of oxymonad protists and
these spirochetal populations in the hindgut community
might be especially abundant. These ¢ndings also suggest
to us that the ectosymbiotic spirochetes species may considerably in£uence the termite hindgut ecosystem. We are
also interested in the sequences from spirochetes in the
hindgut community of C. formosanus [19], as none of these
spirochetal sequences fell within cluster II (data not
shown) even though an interaction between spirochetes
and a hypermastigote protist (Holomastigotoides mirabile)
was observed (our unpublished data).
25
26
T. Iida et al. / FEMS Microbiology Ecology 34 (2000) 17^26
for the Biodesign Research Program, the Genome Research Program, and the Eco Molecular Science Research
Program from RIKEN, and by a grant for the International Cooperative Research Project (Bio-Recycle Project)
from the Japan Science and Technology Corporation. One
of us (T.I.) is a recipient of a Special Postdoctoral Research Fellowship from RIKEN, and (K.O.) was supported by a grant for the Junior Research Associate Program from RIKEN.
[18]
[19]
[20]
[21]
[22]
References
FEMSEC 1166 18-10-00
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
Cyaan Magenta Geel Zwart
Downloaded from http://femsec.oxfordjournals.org/ by guest on March 5, 2016
[1] Breznak, J.A. (1984) In: Bergey's Manual of Systematic Bacteriology
(Krieg, N.R. and Holt, J.G., Eds.), vol. 1, pp. 67^70. Williams and
Wilkins, Baltimore, MD.
[2] Breznak, J.A. (1982) Intestinal microbiota of termites and other xylophagous insects. Annu. Rev. Microbiol. 36, 323^343.
[3] Breznak, J.A. and Brune, A. (1994) Role of microorganisms in the
digestion of lignocellulose by termites. Annu. Rev. Entomol. 39, 453^
487.
[4] Leadbetter, J.R., Schmidt, T.M., Graber, J.R. and Breznak, J.A.
(1999) Acetogenesis from H2 plus CO2 by spirochetes from termite
guts. Science 283, 686^689.
[5] Eutick, M.L., Veivers, P., O'Brien, R.W. and Slator, M. (1978) Dependence of the higher termite Nasutitermes exitiosus and the lower
termite, Coptotermes lacteus on their gut £ora. J. Insect Physiol. 24,
363^368.
[6] Margulis, L., Chase, D. and To, L.P. (1979) Possible evolutionary
signi¢cance of spirochaetes. Proc. R. Soc. Lond. B Biol. Sci. 204,
189^198.
[7] Margulis, L. and Hinkle, G. (1992) In: The Prokaryotes (Balows, A.,
Trupper, H.G., Dworkin, M., Harder, W., Schleifer, K.H., Eds.), pp.
3965^3978. Springer, New York.
[8] Radek, R., Ro«sel, J. and Hausmann, K. (1996) Light and electron
microscopic study of the bacterial adhesion to termite £agellates applying lectin cytochemistry. Protoplasma 193, 105^122.
[9] To, L.P., Margulis, L., Chase, D. and Nutting, W.L. (1980) The
symbiotic microbial community of the Sonoran desert termite: Pterotermes occidentis. Biosystems 13, 109^137.
[10] Bloodgood, R.A., Miller, K.R., Fitzharris, T.P. and Mcintosh, J.R.
(1974) The ultrastructure of Pyrsonympha and its associated microorganisms. J. Morph. 143, 77^106.
[11] Bloodgood, R.A. and Fitzharris, T.P. (1976) Speci¢c associations of
prokaryotes with symbiotic £agellate protozoa from the hindgut of
the termite Reticulitermes and the wood-eating roach Cryptocercus.
Cytobios 17, 103^122.
[12] Smith, H.E. and Arnott, H.J. (1974) Epi- and endobiotic bacteria
associated with Pyrsonympha vertens, a symbiotic protozoon of the
termite Reticulitermes £avipes. Trans. Am. Microsc. Soc. 93, 180^
194.
[13] Smith, H.E., Buhse Jr., H.E. and Stamler, S.J. (1975) Possible formation and development of spirochaete attachment sites found on the
surface of symbiotic polymastigote £agellates of the termite Reticulitermes £avipes. Biosystems 7, 374^379.
[14] Cleveland, L.R. and Grimstone, A.V. (1964) The ¢ne structure of the
£agellate Mixotricha paradoxa and its associated micro-organisms.
Proc. R. Soc. Lond. B Biol. Sci. 157, 668^683.
[15] Kirby, H. (1941) Devescovinid £agellates of termites. I. The genus
Devescovina. Univ. Calif. Pub. Zool. 45, 1^91.
[16] Kirby, H. (1945) Devescovinid £agellates. IV. The genera Metadevescovina and Pseudodevescovina. Univ. Calif. Pub. Zool. 45, 247^318.
[17] Berchtold, M., Ludwig, W. and Konig, H. (1994) 16S rDNA sequence and phylogenetic position of an uncultivated spirochete
from the hindgut of the termite Mastotermes darwiniensis Froggatt.
FEMS Microbiol. Lett. 123, 269^273.
Berchtold, M. and Konig, H. (1996) Phylogenetic analysis and in situ
identi¢cation of uncultivated spirochetes from the hindgut of the
termite Mastotermes darwiniensis. Syst. Appl. Microbiol. 19, 66^73.
Lilburn, T.G., Schmidt, T.M. and Breznak, J.A. (1999) Phylogenetic
diversity of termite gut spirochaetes. Environ. Microbiol. 1, 331^345.
Ohkuma, M. and Kudo, T. (1996) Phylogenetic diversity of the intestinal bacterial community in the termite Reticulitermes speratus.
Appl. Environ. Microbiol. 62, 461^468.
Ohkuma, M. and Kudo, T. (1998) Phylogenetic analysis of the symbiotic intestinal micro£ora of the termite Cryptotermes domesticus.
FEMS Microbiol. Lett. 164, 389^395.
Ohkuma, M., Iida, T. and Kudo, T. (1999) Phylogenetic relationships
of symbiotic spirochetes in the gut of diverse termites. FEMS Microbiol. Lett. 164, 389^395.
Paster, B.J., Dewhirst, F.E., Cooke, S.M., Fussing, V., Poulsen, L.K.
and Breznak, J.A. (1996) Phylogeny of not-yet-cultured spirochetes
from termite guts. Appl. Environ. Microbiol. 62, 347^352.
Trager, W. (1934) The cultivation of a cellulose-digesting £agellate
Trichomonas termopsidis and of certain other termite protozoa. Biol.
Bull. 66, 182^190.
Weisburg, W.G., Barns, S.M., Pelletier, D.A. and Lane, D.J. (1991)
16S ribosomal DNA ampli¢cation for phylogenetic study. J. Bacteriol. 173, 697^703.
Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory, New York.
Thompson, J.D., Higgins, D.G. and Gibson, T.J. (1994) CLUSTAL
W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positions-speci¢c gap penalties and
weight matrix choice. Nucleic Acids Res. 22, 4673^4680.
Felsenstein, J. (1989) PHYLIP-phylogeny inference package. Version
3.5. Cladistics 39, 783^791.
Van de Peer, Y., Robbrecht, E., de Hoog, S., Caers, A., De Rijk, P.
and De Wachter, R. (1999) Database on the structure of small subunit ribosomal RNA. Nucleic Acids Res. 27, 179^183.
Amann, R.I., Ludwig, W. and Schleifer, K.H. (1995) Phylogenetic
identi¢cation and in situ detection of individual microbial cells without cultivation. Microbiol. Rev. 59, 143^169.
Maidak, B.L., Cole, J.R., Parker Jr., C.T., Garrity, G.M., Larsen,
N., Li, B., Lilburn, T.G., McCaughey, M.J., Olsen, G.J., Overbeek,
R., Pramanik, S., Schmidt, T.M., Tiedje, J.M. and Woese, C.R.
(1999) A new version of the RDP (Ribosomal Database Project).
Nucleic Acids Res. 27, 171^173.
Koidzumi, M. (1921) Studies on the intestinal protozoa found in the
termites of Japan. Parasitology 13, 235^309.
Kitade, O., Maeyama, T. and Matsumoto, T. (1997) Establishment of
symbiotic £agellate fauna of Hodotermopsis japonica (Isoptera: Termopsidae). Sociobiology 30, 161^167.
Paster, B.J., Dewhirst, F.E., Weisburg, W.G., Tordo¡, L.A., Fraser,
G.J., Hespell, R.B., Stanton, T.B., Zablen, L., Mandelco, L. and
Woese, C.R. (1991) Phylogenetic analysis of the spirochetes. J. Bacteriol. 173, 6101^6109.
Smith, H.E., Stamler, S.J. and Buhse, H.E. (1975) A scanning electron microscope surbey of the surface features of polymastigote £agellates from Reticulitermes £avipes. Trans. Am. Micros. Soc. 94, 401^
410.
Embley, T.M. and Finlay, B.J. (1994) The use of small subunit rRNA
sequences to unravel the relationships between anaerobic ciliates and
their methanogen endosymbionts. Microbiology 140, 225^235.
Kudo, H., Cheng, K.J. and Costerton, J.W. (1987) Interactions between Treponema bryantii and cellulolytic bacteria in the in vitro
degradation of straw cellulose. Can. J. Microbiol. 33, 244^248.
Stanton, T.B. and Canale Parola, E. (1980) Treponema bryantii newspecies a rumen spirochete that interacts with cellulolytic bacteria.
Arch. Microbiol. 127, 145^156.