HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Supplementary Tables and Alignments Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Chantangsi, C. Articles by Ikonomi, P. Search for Related Content PubMed PubMed Citation Articles by Chantangsi, C. Articles by Ikonomi, P. Agricola Articles by Chantangsi, C. Articles by Ikonomi, P.

Barcoding ciliates: a comprehensive study of 75 isolates of the genus Tetrahymena

Chitchai Chantangsi^1,, Denis H. Lynn¹, Maria T. Brandl², Jeffrey C. Cole³, Neil Hetrick³ and Pranvera Ikonomi⁴

¹ Department of Integrative Biology, University of Guelph, Guelph, Ontario N1G 2W1, Canada
² United States Department of Agriculture, Agriculture Research Service, Western Regional Research Center, Food Safety and Health Unit, 800 Buchanan St, Albany, CA 94710, USA
³ Protistology Department, American Type Culture Collection, 10801 University Blvd, Manassas, VA 20110-2209, USA
⁴ Molecular Authentication Resource Center, American Type Culture Collection, 10801 University Blvd, Manassas, VA 20110-2209, USA

Correspondence
Denis H. Lynn
ddr{at}uoguelph.ca

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
The mitochondrial cytochrome-c oxidase subunit 1 (cox1) gene has been proposed as a DNA barcode to identify animal species. To test the applicability of the cox1 gene in identifying ciliates, 75 isolates of the genus Tetrahymena and three non-Tetrahymena ciliates that are close relatives of Tetrahymena, Colpidium campylum, Colpidium colpoda and Glaucoma chattoni, were selected. All tetrahymenines of unproblematic species could be identified to the species level using 689 bp of the cox1 sequence, with about 11 % interspecific sequence divergence. Intraspecific isolates of Tetrahymena borealis, Tetrahymena lwoffi, Tetrahymena patula and Tetrahymena thermophila could be identified by their cox1 sequences, showing <0.65 % intraspecific sequence divergence. In addition, isolates of these species were clustered together on a cox1 neighbour-joining (NJ) tree. However, strains identified as Tetrahymena pyriformis and Tetrahymena tropicalis showed high intraspecific sequence divergence values of 5.01 and 9.07 %, respectively, and did not cluster together on a cox1 NJ tree. This may indicate the presence of cryptic species. The mean interspecific sequence divergence of Tetrahymena was about 11 times greater than the mean intraspecific sequence divergence, and this increased to 58 times when all isolates of species with high intraspecific sequence divergence were excluded. This result is similar to DNA barcoding studies on animals, indicating that congeneric sequence divergences are an order of magnitude greater than conspecific sequence divergences. Our analysis also demonstrated low sequence divergences of <1.0 % between some isolates of T. pyriformis and Tetrahymena setosa on the one hand and some isolates of Tetrahymena furgasoni and T. lwoffi on the other, suggesting that the latter species in each pair is a junior synonym of the former. Overall, our study demonstrates the feasibility of using the mitochondrial cox1 gene as a taxonomic marker for ‘barcoding’ and identifying Tetrahymena species and some other ciliated protists.

Present address: Department of Zoology, University of British Columbia, Biological Sciences Bldg, 6270 University Blvd, Vancouver, British Columbia V6T 1Z4, Canada.

The GenBank/EMBL/DDBJ accession numbers for the cox1 and SSU rDNA sequences determined in this study are EF070242–EF070328, as detailed in Supplementary Table S1.

Details of the strains examined in this study, including sequence accession numbers, details of the nucleotide compositions of the cox1 and SSU rDNA sequences, values of overall, within-genus and between-genera divergence of datasets of the cox1 sequences and alignments of cox1 and SSU rDNA sequences are available as supplementary material with the online version of this paper.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
DNA-based identification has been proposed to serve as an alternative taxonomic approach for identifying the immense diversity of living organisms (Dawkins, 1998; Tautz et al., 2002, 2003). Hebert et al. (2003a, b) suggested a 650 bp 5' fragment of the mitochondrial cytochrome-c oxidase subunit 1 (cox1) gene as a universal marker or ‘DNA barcode’ for global biological identification of animal species. Use of this mitochondrial gene has a number of advantages. Firstly, among mitochondrial genes, the cox1 gene is one of only two protein-encoding genes [the other encoding cytochrome b (cytb)] that are present in all eukaryotes. In addition, it functions homologously in a wide range of eukaryotes, enabling its sequence to be compared across a diverse array of organisms. Secondly, cytochrome-encoding genes such as cox1 and cytb are more variable than other commonly studied mitochondrial genes, such as 12S RNA and tRNA genes (Janczewski et al., 1995; Kumazawa & Nishida, 1993). Therefore, these protein-encoding genes can be used to discriminate closely related species (Hebert et al., 2003b). Thirdly, the cox1 gene of various animal phyla is easily amplified using universal primers designed from conserved regions of the gene (Folmer et al., 1994). Fourthly, because mitochondria reproduce by binary fission and without sexual recombination, their genes are less subject to insertions, deletions or other large-scale rearrangements that introduce more ambiguous variation into the sequence. Finally, it has long been recognized that the mitochondrial genome evolves at a faster rate than the nuclear genome and therefore mitochondrial genomic sequences at a particular region will be more informative in differentiating or distinguishing closely related species (Hebert et al., 2003b). Preliminary studies have proven this gene to be a good taxonomic marker for discriminating species of animals such as gastropods (Remigio & Hebert, 2003), collembolans (Hogg & Hebert, 2004), mayflies (Ball et al., 2005), flies (Scheffer et al., 2006), moths (Brown et al., 2003), butterflies (Hajibabaei et al., 2006; Hebert et al., 2004a), beetles (Monaghan et al., 2005), ants (Smith et al., 2005), spiders (Barrett & Hebert, 2005), fishes (Ward et al., 2005) and birds (Hebert et al., 2004b).

Among living organisms, protists have long been recognized as an assemblage of organisms of complex forms and with polymorphic life histories. A majority of them are microscopic, and specific staining procedures and electron microscopy are often required in order to reveal key features for taxonomic identification (Corliss & Daggett, 1983). So far, a DNA barcoding approach using cox1 gene sequences has been applied to identify only a few groups of protists, such as red algae and some ciliate genera (Barth et al., 2006; Lynn & Strüder-Kypke, 2006; Saunders, 2005). However, among those ciliates examined, only a few species within two genera, Paramecium and Tetrahymena, were investigated.

To examine the usefulness of the cox1 barcode for ciliate species identification, species of the genus Tetrahymena were investigated. Tetrahymena includes a number of closely related species, both sexual outbreeders and asexual forms. Previous identification of Tetrahymena species has been based on several approaches such as morphology, a combination of ecology and life histories, mating tests, isozyme mobilities and PCR-RFLP (Corliss, 1973; Czapik, 1968; Holz & Corliss, 1956; Jerome & Lynn, 1996; Meyer & Nanney, 1987; Nanney & McCoy, 1976; Nyberg, 1981). However, because of the close relatedness of some members within the genus, species can often not be discriminated using morphological features, even at the ultrastructural level (Corliss & Daggett, 1983). Furthermore, several Tetrahymena isolates have long been known that share morphological similarities but are genetically isolated from each other and were initially named as different syngens, and they have proven to be difficult to discriminate without mating tests (Gruchy, 1955). In addition, some Tetrahymena species show phenotypic plasticity in response to different environmental conditions during their polymorphic life cycles, making them even more difficult to identify morphologically and ecologically (Corliss, 1973; Strüder-Kypke et al., 2001).

Among identification approaches, mating reactivity remains the ‘gold standard’ to determine conspecificity and to discover new species of Tetrahymena. Using this method, a large number of cryptic species were discovered within the Tetrahymena pyriformis species complex (Elliott, 1970; Gruchy, 1955; Nanney & McCoy, 1976). However, a complete set of living reference strains is not available and the approach is impossible for amicronucleate strains, which have been known to be common in Tetrahymena (Nanney & McCoy, 1976). With the advent of molecular techniques, several approaches, such as isozyme mobilities and RFLP, were used to discriminate Tetrahymena species without requiring mating tests. Borden et al. (1977) reported that the most closely related syngens shared an isozyme similarity coefficient of 67 %, while the distantly related ones had a coefficient of 0. This criterion was used to support the establishment of several new Tetrahymena species (Nanney & McCoy, 1976). However, these identification approaches are now known to have their own limitations: RFLP patterns are identical for several species pairs (Jerome & Lynn, 1996), while similar polymorphisms among species are shown for isozyme mobilities (Meyer & Nanney, 1987; Nanney et al., 1998).

Thus, the cox1 DNA barcoding approach, which has proven to be useful for identifying species of animals, was chosen to identify Tetrahymena species. Since few DNA barcoding studies have been done on protists so far, it is uncertain whether the 650 bp 5' region of the cox1 gene, which works well taxonomically for the discrimination of animal species, would be suitable for differentiating ciliated protists. Therefore, in this study, we amplified and sequenced almost the entire length of the cox1 gene of representatives of 36 Tetrahymena species and six wild isolates of undescribed Tetrahymena species as well as three non-Tetrahymena ciliates that are close allies of Tetrahymena, Colpidium campylum, Colpidium colpoda and Glaucoma chattoni. This sampling covered all valid species of this ciliate genus that have been described so far and whose cultures are available. This enabled us to evaluate which part of the gene was the most appropriate diagnostic barcoding region for identifying species of ciliates. In addition, 33 strains of six different Tetrahymena species isolated from different geographical localities were examined to demonstrate that levels of genetic variation within species are sufficiently low to ensure unambiguous identification with the cox1 barcode.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Source of samples.
Seventy-eight isolates of tetrahymenine ciliates were examined in this study (see Supplementary Table S1 in IJSEM Online). Of 42 taxonomically valid species of Tetrahymena (Fenchel & Finlay, 2004), 36 are available in culture and at least one isolate of each of these species was selected as a species representative. These species were Tetrahymena americanis, T. asiatica, T. australis, T. bergeri, T. borealis, T. canadensis, T. capricornis, T. caudata, T. corlissi, T. cosmopolitanis, T. elliotti, T. empidokyrea, T. farleyi, T. furgasoni, T. hegewischi, T. hyperangularis, T. leucophrys, T. limacis, T. lwoffi, T. malaccensis, T. mimbres, T. mobilis, T. nanneyi, T. nipissingi, T. paravorax, T. patula, T. pigmentosa, T. pyriformis, T. rostrata, T. setosa, T. shanghaiensis, T. silvana, T. sonneborni, T. thermophila, T. tropicalis and T. vorax. The remaining six valid Tetrahymena species, Tetrahymena chironomi, T. dimorpha, T. edaphoni, T. rotunda, T. sialidos and T. stegomyiae, were not examined because of the unavailability of their cultures. In addition, six isolates of undescribed Tetrahymena species and three other tetrahymenine species, C. campylum, C. colpoda and G. chattoni, were also investigated. Furthermore, among the Tetrahymena species, additional isolates of T. borealis (4), T. lwoffi (1), T. patula (1), T. pyriformis (4), T. thermophila (3) and T. tropicalis (4) were examined to assess intraspecific sequence divergence. Moreover, the 16 cox1 sequences available from GenBank for one T. pyriformis and 15 T. thermophila isolates were also included. Detailed information is provided in Supplementary Table S1 for all isolates, including catalogue numbers, designations, localities and GenBank accession numbers.

Culture methods and maintenance.
All tetrahymenine ciliates except for the two species of Colpidium were cultured axenically in 10 ml sterile proteose peptone yeast extract (PPYE) medium (0.5 g glucose, 2.0 g proteose peptone, 2.0 g yeast extract and 400 ml distilled water). The two species of Colpidium were cultured in 10 ml bacterized dried cereal grass leaves (Cerophyl) medium, prepared as follows. Cerophyl (20 g) was added to 400 ml distilled water and boiled for 10 min. Distilled water was added to compensate for evaporation and the medium was filtered through Whatman no. 1 filter paper. An aliquot of 10 ml of this concentrated medium was transferred into culture test tubes and autoclaved for 20 min at 121 °C. Ten millilitres of the concentrated Cerophyl medium was diluted with 200 ml distilled water in a sterile flask. Using sterile technique, Enterobacter aerogenes was inoculated into the diluted Cerophyl medium. The bacterized Cerophyl medium was left at room temperature for 24 h prior to inoculation with Colpidium spp. All ciliate cultures were transferred biweekly and maintained at room temperature using sterile technique.

DNA extraction, amplification and sequencing.
In contrast to dead cells and cell debris, which settle on the bottom of the culture tube, healthy cells swim actively just below the surface of the medium. Thus, about 1 ml of this portion of the culture medium containing ciliates was placed in a 1.5 ml microcentrifuge tube. The cells were pelleted by centrifugation at 11 000 g for 5 min. Two DNA extraction protocols, using Chelex beads as described by Walsh et al. (1991) and total nucleic acid purification as specified by EPICENTRE, were performed to yield total genomic DNA of the ciliates. An appropriate amount of DNA template was used in amplification of the cox1 gene and the small-subunit (SSU) rDNA using puReTaq Ready-To-Go PCR beads (GE Healthcare).

cox1 gene.
Approximately 2000 bp of the mitochondrial cox1 gene were amplified by PCR using the forward primer 5'-ATGTGAGTTGATTTTATAGAGCAGA-3' and the reverse primer 5'-GGDATACCRTTCATTTT-3', which were newly designed in this study. The thermal cycler was programmed as follows: hold at 94 °C for 4 min; 5 cycles of denaturation at 94 °C for 30 s, annealing at 45 °C for 1 min and extension at 72 °C for 105 s; 35 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 1 min and extension at 72 °C for 105 s; and hold at 72 °C for 10 min. PCR products corresponding to the expected size were separated by agarose gel electrophoresis, purified using the GENECLEAN kit (Qbiogene) and sequenced in both the forward and reverse directions with an ABI 3730 DNA Analyser using the standard BigDye Terminator version 3.1 cycle-sequencing kit using amplification and internal primers. Almost the entire length, about 1821 bp, of the cox1 gene of 45 species representatives was sequenced, while at least 689 bp of the 5' region of the cox1 gene of 17 additional intraspecific isolates of six Tetrahymena species mentioned previously were sequenced to assess the diagnostic barcoding region.

SSU rDNA.
Approximately 2900 bp of an rDNA fragment were amplified using the forward primer A (5'-CAACCTGGTTGATCCTGCCAGT-3') and the reverse primer C (5'-TTGGTCCGTGTTTCAAGACG-3') (Jerome & Lynn, 1996; Medlin et al., 1988). The PCR product included the SSU rDNA, internal transcribed spacer (ITS) 1, the 5.8S rDNA, ITS2 and a portion of the large-subunit (LSU) rDNA. The thermal cycler was programmed as follows: hold at 94 °C for 4 min; 35 cycles of denaturation at 94 °C for 1 min, annealing at 55 °C for 90 s and extension at 72 °C for 3 min; and hold at 72 °C for 10 min. PCR products corresponding to the expected size were processed as described above. However, only a 1800 bp region of the SSU rDNA was sequenced and included in further analyses.

The SSU rDNA was amplified and sequenced for 19 tetrahymenine species that were chosen as species representatives and examined in this study but which have not yet been investigated in any previous studies. These species were T. americanis, T. asiatica, T. caudata, T. cosmopolitanis, T. elliotti, T. furgasoni, T. leucophrys, T. limacis, T. lwoffi, T. mimbres, T. nipissingi, T. paravorax, T. shanghaiensis, T. silvana, T. sonneborni, Tetrahymena sp. 1 (Foissner), Tetrahymena sp. 3 (RA9), Tetrahymena sp. 5 (NI) and C. colpoda. In addition, the SSU rDNA of two isolates of T. pyriformis and four isolates of T. tropicalis were also amplified and sequenced. Detailed information on these species and isolates is given in Supplementary Table S1.

Sequence analyses.
The cox1 and SSU rDNA sequences were first aligned automatically by CLUSTAL W (Thompson et al., 1994) using the MEGA program version 3.1 (Kumar et al., 2004) and then further refined by eye. The alignments for cox1 and SSU rDNA sequences of the examined tetrahymenine ciliates are available as supplementary material in IJSEM Online.

The cox1 gene sequences obtained from this study were 1821 bp in length except for that of G. chattoni, which was 1785 bp long. The 1821 nucleotides span positions 52–1872 with reference to the complete cox1 genes of T. pyriformis and T. thermophila published in GenBank (Brunk et al., 2003; Burger et al., 2000). These cox1 sequences were then divided into the following five datasets.

Dataset 1. This dataset comprised the 1821 bp sequences of the cox1 genes from a representative of each species, for a total of 45 sequences. Thirty-six positions (1786–1821) at the 3'-end of the cox1 gene sequence of G. chattoni were treated as missing data.

Dataset 2. This dataset comprised only the 5' half of the 1821 bp sequences of the cox1 genes from a representative of each species, for a total of 45 sequences. This region is 912 bp in length.

Dataset 3. This dataset comprised only the 3' half of the 1821 bp sequences of the cox1 genes from a representative of each species, for a total of 45 sequences. This region is 909 bp in length. Thirty-six positions (874–909) at the 3'-end of the cox1 gene sequence of G. chattoni were treated as missing data.

Dataset 4. This dataset comprised the 689 bp sequences of the cox1 genes from a representative of each species, for a total of 45 sequences. This region starts at position 169 with reference to the 1821 sites of our cox1 gene sequences and at position 220 with reference to the 2067 sites of the complete cox1 gene sequence. Compared to animal barcoding regions, the starting position of this 689 bp sequence is in close proximity to the beginning site of the diagnostic barcoding region used for identifying animal species.

Dataset 5. This dataset comprised the 689 bp sequences of the cox1 genes from all 78 isolates that were included in this study. This is the same region as dataset 4.

The SSU rDNA of 19 tetrahymenine species and of six isolates was newly sequenced in this study. In addition, another 24 species whose SSU rDNA sequences were available at GenBank were also included in this study. These species are T. australis, T. bergeri, T. borealis, T. canadensis, T. capricornis, T. corlissi, T. empidokyrea, T. farleyi, T. hegewischi, T. hyperangularis, T. malaccensis, T. mobilis, T. nanneyi, T. patula, T. pigmentosa, T. pyriformis, T. rostrata, T. setosa, T. thermophila, T. tropicalis, T. vorax, Tetrahymena sp. 6 (Brandl), C. campylum and G. chattoni. Moreover, full sequences of the SSU rDNA of Tetrahymena sp. 2 (CO) and Tetrahymena sp. 4 (SIN), kindly provided by Dr Michaela C. Strüder-Kypke (Department of Integrative Biology, University of Guelph, Canada), were included in our study. Detailed information on these species and isolates is provided in Supplementary Table S1. Aligned SSU rDNA sequences of all 45 species representatives were constructed as a dataset for sequence divergence calculation. This nucleotide dataset was 1650 positions in length and included a few gaps and some ambiguous characters. Exclusion of these gaps and ambiguous positions gave a total of 1639 sites, which were used in sequence divergence analyses. In addition, SSU rDNA sequences of two T. pyriformis isolates and four T. tropicalis isolates were added to the 45-sequence tetrahymenine SSU rDNA alignment to illustrate phylogenetic relationships among these organisms.

Sequence divergences were calculated for each dataset using the Kimura two-parameter (K2P) distance model (Kimura, 1980). A neighbour-joining (NJ) phylogenetic tree was inferred from genetic distances of cox1 sequences calculated by DNADIST with the K2P model of sequence evolution using PHYLIP version 3.65 (Felsenstein, 2004; Saitou & Nei, 1987). Using G. chattoni as the outgroup species, this tree-building approach was performed on the dataset of 689 bp cox1 gene sequences of 78 tetrahymenine ciliates and the dataset of 1639 bp SSU rDNA sequences of 51 tetrahymenine ciliates to provide an illustration of the patterning of divergence between and within species. In addition, 1000 bootstrap resamplings were carried out using SEQBOOT to determine confidence levels of deduced relationships. CONSENSE within the PHYLIP package was used to construct a consensus tree.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
With tetrahymenine cox1-specific primers, the cox1 gene was amplified and recovered for all tetrahymenine ciliates that were examined in this study. However, because of considerable divergence in the cox1 gene of G. chattoni, as indicated by its high percentage sequence divergence, our internal primers that were used to sequence toward the 3' region of the gene could not sequence that area, leaving 36 bp unavailable. Therefore, only 1785 bp of the cox1 gene was obtained from G. chattoni. Alignment of the entire length of 1821 bp of cox1 sequences of all 45 species representatives was straightforward as there were no insertions and no deletions, corroborating research on other organisms (Hebert et al., 2003b; Mardulyn & Whitfield, 1999; Ward et al., 2005).

Nucleotide composition
Mean frequencies of thymine (T), cytosine (C), adenine (A) and guanine (G) and molar percentage G+C content for the SSU rDNA of examined tetrahymenines are in the same range as in other ciliates (Elwood et al., 1985; Sogin & Elwood, 1986; Schlegel et al., 1991). Mean base frequencies and G+C and A+T contents of the cox1 gene for each genus were similar in five different datasets of the gene, as listed in Supplementary Table S2 in IJSEM Online. Relative nucleotide frequencies for the cox1 genes of the examined tetrahymenine species were 38.4–43.1 mol% for T, 11.0–13.6 mol% for C, 30.5–33.6 mol% for A and 13.9–15.6 mol% for G, yielding mean G+C and A+T contents of 25.2–29.2 and 70.8–74.8 mol%, respectively. The G+C content of the 1821 bp cox1 gene sequences was similar among the three tetrahymenine genera, with a mean of 26.7 mol%, and lower than that of the cox1 gene of Paramecium aurelia, which is 41.8 mol% (Burger et al., 2000). The G+C content of cox1 genes of tetrahymenine ciliates is considerably different from that of animal cox1 genes: the G+C content of several orders of insects is about 35 mol% (Hebert et al., 2003b), while that of 207 species of Australian fish ranges from 42 to 47 mol% (Ward et al., 2005). This G+C content is also characteristic of the nucleotide composition of the entire mitochondrial genomes of T. pyriformis and T. thermophila, which show low G+C contents, of 21.3 and 20.7 mol%, respectively (Brunk et al., 2003; Burger et al., 2000), considerably lower than the 41.2 mol% G+C of the P. aurelia mitochondrial genome (Cummings, 1992). A search was also carried out for stop codons of the ciliate mitochondrial genomes, UAA and UAG, which encode glutamine in the ciliate nuclear genomes; UGA was excluded as it encodes tryptophan in ciliate mitochondrial genomes (Brunk et al., 2003). These stop codons were not observed in any of the amplified cox1 gene sequences, demonstrating that these were fully functional mitochondrial cox1 gene sequences.

Sequence analyses for cox1 and SSU rDNA
Sequence analyses for the cox1 gene.
Mean sequence divergences of all 45 species representatives analysed from three different cox1 datasets, including 1821 bp, the 912 bp 5' half and the 909 bp 3' half, were closely similar, ranging from 10.59±0.61 to 11.14±0.52 % (mean±SEM), as shown in Supplementary Table S3. Starting at position 169 with reference to 1821 sites, about 689 bp in the 5' half of the cox1 gene were selected as the diagnostic barcoding region. The first portion of about 240 bp and the last portion of about 90 bp in this 689 bp region are comparable to the beginning part and the middle part of 650 bp animal barcodes, respectively. However, the middle portion of about 360 bp of the 689 bp tetrahymenine barcode is an insert unique to the ciliate mitochondrial cox1 gene (Ziaie & Suyama, 1987). The mean sequence divergence value calculated from the 689 bp barcoding region between species pairs for the 45 species representatives was 11.13±0.78 % (Table 1). This divergence value was very similar to those of the three other regions, indicating that any parts of the cox1 gene can be used as a diagnostic barcode to discriminate between the tetrahymenine species examined. Furthermore, 983 of 990 interspecific pairwise comparisons based on the barcoding region showed more than 2 % sequence divergence (Fig. 1). Among these pairwise comparisons, the following seven species pairs showed apparently low values, ranging from 0 to 1.17 %: T. canadensis ATCC^® 30368^TM and T. rostrata ATCC^® 30770^TM (0.73 %), T. furgasoni ATCC^® 30006^TM and T. lwoffi (1630/1G) (0 %), T. nanneyi ATCC^® 50071^TM and T. nipissingi ATCC^® 30837^TM (0.29 %), T. pyriformis ATCC^® 30005^TM and T. setosa ATCC^® 30782^TM (0 %), Tetrahymena sp. 2 (CO) and Tetrahymena sp. 3 (RA9) (0.15 %), Tetrahymena sp. 2 (CO) and Tetrahymena sp. 5 (NI) (1.17 %) and Tetrahymena sp. 3 (RA9) and Tetrahymena sp. 5 (NI) (1.02 %) (Fig. 1). Analyses of percentage sequence divergences of 1821 bp, the 912 bp 5' half and the 909 bp 3' half for these species pairs showed similar results. However, there were no differences in sequence divergence values in the 5' and 3' halves between Tetrahymena sp. 2 (CO) and Tetrahymena sp. 3 (RA9) (both 0.22±0.16 %) and between T. furgasoni ATCC^® 30006^TM and T. lwoffi (1630/1G) (both 0 %).

View larger version (14K): [in this window] [in a new window]   Fig. 1. Distribution of interspecific pairwise sequence divergence based on the 689 bp cox1 barcoding region of 45 tetrahymenine species representatives, showing that a majority of pairwise comparisons have more than 2 % sequence divergence.

Calculated from the 689 bp barcoding region, mean values (±SEM) for sequence divergences among five isolates of T. borealis, two isolates of T. lwoffi, two isolates of T. patula, six isolates of T. pyriformis, 19 isolates of T. thermophila and five isolates of T. tropicalis ranged from 0 to 9.07±0.83 % (Fig. 2). Among these six species, conspecific isolates of T. borealis, T. lwoffi, T. patula and T. thermophila showed unproblematic results in that they showed <1 % intraspecific sequence divergence, whereas some conspecific isolates of T. pyriformis and T. tropicalis showed considerably higher sequence divergences, leading to suspicion of the validity of cultures and/or the status of the isolates of these species.

View larger version (15K): [in this window] [in a new window]   Fig. 2. Intraspecific sequence divergence (%, mean±SEM) of T. borealis, T. lwoffi, T. patula, T. pyriformis, T. thermophila and T. tropicalis calculated from the 689 bp cox1 barcoding region. Intraspecific isolates of T. pyriformis and T. tropicalis showed considerably higher sequence divergence values when compared with those of the remaining four species, suggesting the presence of cryptic species or misidentification or contamination of cultures.

The within-species groupings showed a low degree of genetic distance, as represented by short branch length in the 689 bp cox1 NJ tree, and were generally supported by 100 % bootstrap values (Fig. 3). However, conspecific isolates of T. pyriformis and T. tropicalis showed high sequence divergence values, explaining the interspersion of isolates of these two species on the NJ tree (Fig. 3). Four of six isolates of T. pyriformis were clustered together with 70.6 % bootstrap support, but separately from the other two isolates (Fig. 3). In addition, T. setosa ATCC^® 30782^TM and T. tropicalis ATCC^® 205060^TM were also clustered with this T. pyriformis clade with 100 % bootstrap support in the NJ analyses (Fig. 3). The remaining two T. pyriformis isolates, T. pyriformis ATCC^® 205038^TM and T. pyriformis ATCC^® 205062^TM, were grouped with low bootstrap support with the main T. pyriformis clade, but on a separate well-supported clade with 79.8 % bootstrap support (Fig. 3). Moreover, the terminal branches of these two T. pyriformis isolates appeared to be long, showing larger genetic distances and suggesting the possibilities of their status as novel species.

View larger version (35K): [in this window] [in a new window]   Fig. 3. NJ tree inferred from 689 bp of the diagnostic barcoding region of the cox1 gene for all 78 tetrahymenine species isolates. Genetic distance calculation was based on the K2P model of sequence evolution. Branch lengths separating taxa represent genetic distances. Bar, 0.02 nucleotide substitutions per site. Percentage bootstrap support for taxa within the T. pyriformis and T. thermophila clades, which are not shown in the figure, were typically lower than 20 and 5 %, respectively. Values are percentages of bootstrap support from 1000 resamplings for the NJ analysis. Because of the interspersion of intraspecific isolates of T. pyriformis and T. tropicalis, those strains are labelled in bold. Amicronucleate species and strains (if known) are labelled with asterisks (*).

Sequence divergence analyses of different portions of the 1821 bp region showed that the 689 bp 5'-region could be used effectively as a DNA barcode for identifying Tetrahymena spp., as supported by very similar values of percentage sequence divergence calculated from the four regions of the cox1 gene. A comparable region of the cox1 gene has been used extensively as a diagnostic barcoding region in identifying species for several animal groups, as noted above.

Sequence analyses for SSU rDNA.
The 1639 bp of SSU rDNA were included in the sequence divergence analysis. Calculation of 1599 positions of the gene, excluding gaps and missing data, for 45 species representatives showed a mean sequence divergence of 1.56±0.16 %. However, the mean value of sequence divergence for the 1639 sites increased slightly to 1.71±0.16 % when gaps and missing data were deleted in a pairwise manner (Table 1). The sequence divergence values calculated from the SSU rDNA were considerably lower than those obtained from the cox1 gene in every analysis (Table 1). A majority of interspecific pairwise comparisons based on SSU rDNA showed 0–2 % sequence divergence (Fig. 4). Except for the two species pairs Tetrahymena sp. 2 (CO) and Tetrahymena sp. 3 (RA9) and Tetrahymena sp. 3 (RA9) and Tetrahymena sp. 5 (NI), which both showed 0.06 % sequence divergence in SSU rDNA, the other five species pairs that showed a low degree of sequence divergence based on cox1 gene sequences had 0 % sequence divergence between their SSU rDNAs.

View larger version (14K): [in this window] [in a new window]   Fig. 4. Relationship between interspecific sequence divergences of the 689 bp cox1 barcoding region and 1639 bp SSU rDNA for 45 tetrahymenine species representatives. Note that the cox1 gene provides better discrimination among these tetrahymenine species.

Comparison of intraspecific sequence divergence in Tetrahymena and other organisms
The low intraspecific sequence divergence values for all four unproblematic Tetrahymena species examined (T. borealis, T. lwoffi, T. patula and T. thermophila) are similar to those calculated for the barcoding region of various groups of animals. For example, mean values of within-species sequence divergence are 0.39 % in Australian fishes (Ward et al., 2005), 0.43 % in North American birds (Hebert et al., 2004b), 1 % in mayflies (Ball et al., 2005) and <1 % in springtails (Hogg & Hebert, 2004), lepidopterans (Hajibabaei et al., 2006; Hebert et al., 2003b) and several arachnid species (Barrett & Hebert, 2005). These results are consistent with intraspecific divergences of mitochondrial genes hardly greater than 2 % and most less than 1 % (Avise, 2000). However, several species of animals, as well as some protists, have been reported to have high degrees of intraspecific divergence. For example, values of 5 and 13 % within-species divergence were reported in the collembolans Sminthurides malmgreni and Folsomia quadrioculata, respectively (Hogg & Hebert, 2004). Two specimens of the shortnose chimaera Hydrolagus novaezelandiae showed 14.08 % intraspecific distance (Ward et al., 2005). A single individual of the mayfly Maccaffertium modestum showed 13.7 % intraspecific divergence from the other specimens examined, implying the presence of a cryptic species complex as suggested by some behavioural evidence (Ball et al., 2005). In addition, intraspecific divergence values of the ciliates Paramecium caudatum and Paramecium multimicronucleatum were 7 and 9.5 %, respectively (Barth et al., 2006). However, Barth et al. (2006) suspected that their intraspecific isolates of these nominate Paramecium species might belong to more than one syngen or biological species and suggested future investigation of these isolates by mating reactivities to determine whether they belong to known or novel syngens.

Intraspecific sequence divergence analyses for T. pyriformis.
When Nanney & McCoy (1976) established nominal species for the syngens of Tetrahymena, they did not at the same time establish holotype strains for each species. Therefore, we suggest that the T. pyriformis strain phenoset A ATCC^® 30327^TM be established as the holotype strain based on its original isolation in 1922 by Professor Dr André Lwoff (ATCC catalogue). With this isolate as the type for the species, four T. pyriformis isolates (an unnamed isolate represented by GenBank accession no. AF160864, ATCC^® 30327^TM, ATCC^® 30005^TM and ATCC^® 30202^TM), T. setosa HZ-1 (=ATCC^® 30782^TM) and T. tropicalis G1-R (=ATCC^® 205060^TM) clustered together with 100 % bootstrap support in the cox1 NJ tree: these six strains differed from each other by less than 0.2 % in their cox1 gene sequences. The relatedness between T. pyriformis and T. setosa has been suggested by several previous studies. For example, there are no nucleotide differences in the 5.8S, SSU rDNA or partial LSU rDNA sequences between T. pyriformis and T. setosa (Nanney et al., 1989; Preparata et al., 1989; Strüder-Kypke et al., 2001). However, this species pair showed an isozyme similarity coefficient of only 62 %, indicating their independent species status based on the isozyme species criterion (Meyer & Nanney, 1987; Nanney et al., 1989).

In contrast to our study, previous studies have not shown a close relationship between T. tropicalis G1-R (=ATCC^® 205060^TM) and strains of T. pyriformis. Referred to as strain G1-R, T. tropicalis ATCC^® 205060^TM showed isozyme mobilities distinct from those of other strains of Tetrahymena, including strains of T. pyriformis (Borden et al., 1973a). In addition, based on sequences of the D2 domain of the LSU rDNA, T. tropicalis ATCC^® 205060^TM was grouped together with other T. tropicalis isolates, but separately from the ‘T. pyriformis’ cluster (Nanney et al., 1998). However, our SSU rDNA sequence corroborated the result from the cox1 data, confirming the clustering of T. tropicalis ATCC^® 205060^TM with T. pyriformis (GenBank accession no. X56171): there was 0 % sequence divergence in the SSU rDNA (data not shown).

Based on cox1 gene sequences, deep intraspecific divergence was found between T. pyriformis ATCC^® 205038^TM and T. pyriformis ATCC^® 205062^TM, as represented by 8.46 % genetic distance between these two isolates and the remaining four T. pyriformis isolates. However, analysis of sequences of 190 bases of the LSU rDNA showed a close relationship between these two T. pyriformis isolates and other T. pyriformis isolates (Nanney et al., 1998). Given a mean intraspecific sequence divergence value of <1.0 %, these two T. pyriformis isolates showed more than eight times this conspecific divergence value. The incongruence between our study and previous ones may indicate either the presence of cryptic species within T. pyriformis or misidentification or contamination of the cultures.

Based on the success of the cox1 gene in identifying animal and other Tetrahymena species, our work implies that T. setosa HZ-1 (=ATCC^® 30782^TM) and T. tropicalis G1-R (=ATCC^® 205060^TM) should be assigned to T. pyriformis (see below). Analyses of the SSU rDNA to assess the status of T. pyriformis ATCC^® 205038^TM and T. pyriformis ATCC^® 205062^TM confirmed their placement, based on cox1 sequences, close to T. pyriformis (GenBank accession no. X56171), with a mean of 0.04 % sequence divergence in the SSU rDNA for these three strains (data not shown).

Intraspecific sequence divergence analyses for T. tropicalis.
As indicated by nearly complete interspersion of its conspecific isolates in the NJ tree, T. tropicalis showed a high degree of intraspecific sequence divergence. Again, since Nanney & McCoy (1976) did not designate a holotype strain for T. tropicalis, we suggest that strain TC-105 of T. tropicalis ATCC^® 30276^TM be chosen as the holotype strain on the basis of its origin from one of the type localities of the species, which are the American tropics and the Pacific islands (Nanney & McCoy, 1976), and its original isolation from Rio Martin Sanchez, Panama, by A. M. Elliott (Nanney et al., 1998). Given this isolate as the holotype strain, the other four isolates did not show close intraspecific relationships to it based on the cox1 gene sequence. This result is consistent with our analyses of SSU rDNA sequences but contradictory to that based on analysis of sequences of 190 bases of the LSU rDNA, which showed a grouping together of all four isolates, T. tropicalis ATCC^® 30276^TM, T. tropicalis ATCC^® 205097^TM, T. tropicalis ATCC^® 205060^TM and T. tropicalis ATCC^® 205083^TM (Nanney et al., 1998). In addition, T. tropicalis ATCC^® 205156^TM was placed distantly from these other T. tropicalis isolates in the 689 bp cox1 NJ tree and confirmed by the SSU rDNA tree, which placed this isolate as a sister taxon of T. empidokyrea ATCC^® 50595^TM (Fig. 5). These inconsistencies between our study and previous ones may indicate either the presence of cryptic species within T. tropicalis or mislabelling or contamination of the cultures. Amplification and sequencing of the cox1 gene of additional T. tropicalis isolates should be performed to assess its species status.

View larger version (40K): [in this window] [in a new window]   Fig. 5. NJ tree of 51 tetrahymenine species, including 45 species representatives, two additional intraspecific isolates of T. pyriformis and four additional intraspecific isolates of T. tropicalis, inferred from 1639 bp of the SSU rDNA. Percentage bootstrap values at nodes are based on 1000 resamplings of the data. Magnified clades show the topology more clearly, but percentage bootstrap values at the nodes are typically extremely low. Genetic distance calculation was based on the K2P model of sequence evolution. Branch lengths separating taxa represent genetic distance. Bar, 0.02 nucleotide substitutions per site.

In our study, there were several Tetrahymena species pairs that showed <1 % interspecific sequence divergence calculated from the 689 bp barcoding region. These species pairs are T. canadensis ATCC^® 30368^TM and T. rostrata ATCC^® 30770^TM, T. furgasoni ATCC^® 30006^TM and T. lwoffi (1630/1G), T. nanneyi ATCC^® 50071^TM and T. nipissingi ATCC^® 30837^TM, T. pyriformis ATCC^® 30005^TM and T. setosa ATCC^® 30782^TM, Tetrahymena sp. 2 (CO) and Tetrahymena sp. 3 (RA9), Tetrahymena sp. 2 (CO) and Tetrahymena sp. 5 (NI) and Tetrahymena sp. 3 (RA9) and Tetrahymena sp. 5 (NI).

Taxonomic relationship of T. canadensis and T. rostrata.
Formerly known as syngen 7 of the T. pyriformis species complex, T. canadensis is morphologically indistinguishable but reproductively isolated from other members of the T. pyriformis species complex and, based on isozyme data, the organism was established as a named species by Nanney & McCoy (1976). T. rostrata was first described as Paraglaucoma rostrata by Kahl (1926). Corliss (1952) transferred this species to the genus Tetrahymena as T. rostrata. On the basis of its capacity to form both reproductive and resting cysts, to exhibit edaphic and parasitic forms and to possess histophagous or parasitic habits, Corliss (1952) established it as the ‘type’ of the rostrata complex.

The low value of interspecific sequence divergence between T. canadensis ATCC^® 30368^TM and T. rostrata ATCC^® 30770^TM implies either convergent molecular evolution of these two species, phenotypic morphological variations within a single genetic species, contamination or misidentification of the cultures. Interspecific sequence divergence values inferred from 1821 bp and 689 bp of the cox1 gene were 1 % (18 nucleotide differences) and 0.73 % (5 nucleotide differences), respectively. In addition, the 1639 sites of the SSU rDNA sequences were identical. Close relationships between T. canadensis and T. rostrata were also revealed by several previous molecular studies. For example, based on analysis of 180 sites of 23S rRNA gene sequences, no differences were observed (Nanney et al., 1989). In addition, these two species were grouped within riboset A, since identical sequences were also found for 120 bp of the LSU rDNA and 154 bp of the 5.8S rDNA (Preparata et al., 1989). Furthermore, only 3 of 579 bp of the histone H3II/H4II region were different between these two species (Sadler & Brunk, 1992). Although analysis of isozyme mobility between T. canadensis and T. rostrata showed a similarity coefficient of 66 %, which suggested their discrete but very closely related species status (Borden et al., 1977; Meyer & Nanney, 1987), all of the other molecular evidence apparently indicates their identity as a single species.

Since we observed neither reproductive nor resting cysts in T. rostrata ATCC^® 30770^TM, misidentification of this culture is likely. New field isolates of T. rostrata demonstrating species-specific characters must be collected to confirm the taxonomic validity and barcode for this species. This will also help us determine whether T. canadensis is just a pyriform stage of T. rostrata, and the former should then be regarded as a junior synonym of the latter.

Taxonomic relationship of T. furgasoni and T. lwoffi.
T. furgasoni and T. lwoffi as well as T. elliotti are all amicronucleate and distinct from T. pyriformis and from each other based on heterogeneous isozyme patterns (Borden et al., 1973a, b). Nanney & McCoy (1976) established these three isozymically distinct groups as new species and named them after three pioneer tetrahymenologists, Waldo H. Furgason, André Lwoff and Alfred M. Elliott.

The failure to discriminate between T. furgasoni ATCC^® 30006^TM and T. lwoffi (1630/1G) was due to the 0 % sequence divergence in both the cox1 and SSU rDNA sequences. These two species were assigned to the pyriformis complex and, before formally receiving Latinized binomial names, T. furgasoni was known as two strains of T. pyriformis, GL-5 and GL-10, whereas T. lwoffi was recognized as four strains of T. pyriformis, PP, CH-S, GP and H-1 (Borden et al., 1973a; Meyer & Nanney, 1987). These two groups are isozymically different and were therefore later established as the new species T. furgasoni and T. lwoffi by Nanney & McCoy (1976). Re-examination of isozyme mobilities of T. furgasoni and T. lwoffi by Meyer & Nanney (1987) showed different results from the previous study by Borden et al. (1973a): these two species then had indistinguishable isozyme patterns. Meyer & Nanney (1987) suggested that they should be synonymized and suppressed the species name T. lwoffi. In addition, a high similarity in cytoskeleton proteins between T. furgasoni and T. lwoffi was observed by Williams et al. (1984). Furthermore, since both species lack a micronucleus, a mating test and the production of fertile F1 could not be performed. Thus, DNA-based identification using cox1 gene sequences has supported the synonymous status of these two species. Other isolates of T. furgasoni and T. lwoffi (if any) are required in order to support their conspecificity further.

Taxonomic relationship of T. pyriformis and T. setosa.
T. setosa, formerly known as T. setifera, was described by Holz & Corliss (1956) as a new species on the basis of having a caudal cilium at its posterior end. Possession of a caudal cilium by T. setosa gives darkly stained granules surrounded by a delicate fibril at the posterior pole of the cell, the so-called polar basal granule (PBG) complex. The species was also distinguished based on its contractile vacuole pores (CVP) compared with those of T. pyriformis: the two CVPs of T. setosa are placed parallel to each other, whereas those of T. pyriformis are positioned obliquely to each other.

One nucleotide difference in the cox1 gene sequences between T. pyriformis ATCC^® 30005^TM and T. setosa ATCC^® 30782^TM led to sequence divergence values of only 0.11 and 0 % calculated from the 3'-region and the barcoding region, respectively. In addition, the 1639 sites of their SSU rDNA sequences were identical. This gave rise to the suspicion of two possible scenarios, as discussed below. Firstly, the cultures may have been either mislabelled or contaminated. Secondly, these two tetrahymenines are actually the same genetic species. Observation and staining of starved live cells of T. setosa showed that none had a caudal cilium (data not shown). Cells from the T. setosa culture were stained using the Chatton–Lwoff procedure (Corliss, 1953). Fifty silver-stained cells were examined to determine the position of CVPs and the existence of a PBG complex. The parallel CVP type was found in 14 cells (n=50) of the T. setosa culture. The remaining 36 cells had an oblique pattern of CVPs (i.e. T. pyriformis-type CVPs). In addition, when the posterior ends of cells were observed, only two cells (n=12) had both a PBG complex and parallel CVPs. The remaining 10 cells had no PBG complex and oblique CVPs (i.e. T. pyriformis-specific characteristics). Although these microscopic observations suggest either that T. setosa is polymorphic for these traits or that the culture of T. setosa had been contaminated with T. pyriformis either prior to submission to the ATCC or at some time during its history, the possibility of them being the same genetic species cannot be ruled out.

As previously mentioned, T. setosa was described on the basis of possessing a caudal cilium, arrangement of CVPs and other morphological features. However, a study by Nelsen & Debault (1978) showed that a caudal cilium could be induced in T. pyriformis syngen 1, later named T. thermophila. Therefore, the use of this feature for taxonomic diagnosis seems uncertain.

Single-cell isolation should be performed to establish a pure T. setosa culture. Observation and staining of live cells should be then carried out to ensure that they truly manifest the T. setosa taxonomic characters described by Holz & Corliss (1956). Once the pure culture is established, both cox1 and SSU rDNA should be reamplified and sequenced to compare with the results presented here. If this yields results identical to ours, T. setosa should be considered a junior synonym of T. pyriformis, confirming that T. setosa might be only a strain of T. pyriformis, as stated once by Corliss (1972).

Taxonomic relationship of T. nanneyi and T. nipissingi.
T. nanneyi and T. nipissingi were described as new biological species on the basis of mating reactivities and isozyme mobilities by Simon et al. (1985) and Nyberg (1981), respectively. Isozyme similarity coefficient values between them are >62 % (Meyer & Nanney, 1987; Simon et al., 1985). In addition, Nyberg (1981) found a correlation between temperature tolerance and Tetrahymena species and used this species-specific correlation as a supporting criterion to describe new Tetrahymena species such as T. hegewischi, T. nipissingi and T. sonneborni. D. Nyberg (personal communication) has suggested that the following strains be established as holotypes: T. hegewischi KP7 (=ATCC^® 30832^TM), T. nanneyi LB2 (=ATCC^® 50071^TM), T. nipissingi X2-AM (=ATCC^® 30837^TM) and T. sonneborni EA2 (=ATCC^® 30834^TM). In this study, holotype cultures of T. nanneyi and T. nipissingi were examined.

Interspecific sequence divergences lower than 1 % for the cox1 gene have rarely been found. This phenomenon was observed in our study of T. nanneyi and T. nipissingi. The genetic distance between these two species is <0.33 % based on cox1 gene sequences, demonstrating their close relationship. However, nuclear gene sequences of T. nanneyi and T. nipissingi are identical for several genes, for example a partial sequence of the 23S rRNA gene (Preparata et al., 1989) and the entire SSU rDNA. Although molecular data showed great similarities between these species, reproductive isolation from each other and other species of Tetrahymena confirms their status as true biological species (Nyberg, 1981; Simon et al., 1985). However, this close relationship between T. nanneyi and T. nipissingi remains suspicious in light of several lines of evidence. When Nyberg (1981) described T. nipissingi as a new biological species, he also found two other Tetrahymena isolates of a new mating group numbered 15: WX0 and XQ5. In his mating reactivity experiment, Nyberg (1981) refrained from establishment of these two isolates as a new species since crossing between them failed to produce viable progeny. In addition, Nyberg (1981) reported the observation of mating reactivity between strains of group 15 and T. nipissingi by E. Simon, although mating reactivity between them had not been repeatable. Subsequently, viable immature progeny were produced by crossing Tetrahymena strain LB2 and isolates of Nyberg's group 15, providing corroborative evidence of a new biological species. This led Simon et al. (1985) to establish a new species, T. nanneyi. Given the close molecular relatedness between T. nanneyi and T. nipissingi from previous studies and our own and the evidence mentioned above, the close relationships between these two species may indicate either a recent divergence between T. nanneyi and T. nipissingi or that these two ciliates belong to the same species.

In order to validate the biological species status of T. nanneyi and T. nipissingi, additional strains from culture collections and nature should be thoroughly examined using the cox1 gene and SSU rDNA. Since the isolate of T. nipissingi X2-AM (=ATCC^® 30837^TM) that was examined in this study was only one representative of the species, other isolates, such as T. nipissingi XK 2 (=ATCC^® 205160^TM), X13-R1 (=ATCC^® 30838^TM), X32H3 (=ATCC^® 205029^TM) and X7E3 (=ATCC^® 205091^TM), should particularly be examined. In addition, other isolates of T. nanneyi [i.e. the isolate strains XQ5 (=ATCC^® 50067^TM) and WX0 (=ATCC^® 205090^TM)] that were originally used in describing this species should also be investigated.

Taxonomic relationships among undescribed Tetrahymena species.
In our study, six undescribed species, Tetrahymena sp. 1 (Foissner), Tetrahymena sp. 2 (CO), Tetrahymena sp. 3 (RA9), Tetrahymena sp. 4 (SIN), Tetrahymena sp. 5 (NI) and Tetrahymena sp. 6 (Brandl), were examined with the aim of identifying and assigning them to a nominal species using the DNA-based identification criteria of the 689 bp cox1 barcoding region.

Based on the 689 bp cox1 gene sequences, Tetrahymena sp. 1 (Foissner), Tetrahymena sp. 4 (SIN), and Tetrahymena sp. 6 (Brandl) showed considerable genetic distances of 11.47, 9.71 and 9.16 %, respectively, from other Tetrahymena species, while, based on SSU rDNA sequences, Tetrahymena sp. 1 (Foissner), Tetrahymena sp. 4 (SIN) and Tetrahymena sp. 6 (Brandl) showed a lower evolutionary distance, 1.18, 1.35 and 1.11 %, respectively, from other Tetrahymena species. Furthermore, these three undescribed Tetrahymena showed considerable cox1 sequence divergence from each other: 11.60 % between Tetrahymena sp. 1 (Foissner) and Tetrahymena sp. 4 (SIN), 9.76 % between Tetrahymena sp. 1 (Foissner) and Tetrahymena sp. 6 (Brandl) and 8.77 % between Tetrahymena sp. 4 (SIN) and Tetrahymena sp. 6 (Brandl). In addition, these three species did not cluster together in the cox1 NJ tree. Tetrahymena sp. 1 (Foissner) and Tetrahymena sp. 6 (Brandl) were found free-living in natural habitats, whereas Tetrahymena sp. 4 (SIN) is a parasite of the guppy Poecilia reticulata. Wilhelm Foissner (personal communication) suggested that Tetrahymena sp. 1 (Foissner) could be a new species because of its capacity to form a macrostome. A study by Brandl et al. (2005) reported a close relationship of Tetrahymena sp. 6 (Brandl) to T. mobilis and T. tropicalis based on SSU rDNA sequences. Similar relationships were found in our study, in which T. mobilis and some T. tropicalis isolates grouped with this isolate using both the cox1 gene and SSU rDNA sequences. Although Tetrahymena sp. 4 (SIN) as well as Tetrahymena sp. 2 (CO), Tetrahymena sp. 3 (RA9) and Tetrahymena sp. 5 (NI) were found as parasites of the guppy Poecilia reticulata, Tetrahymena sp. 4 (SIN) was not closely related to the other three isolates, showing a considerable degree of genetic distance, 7.79 % based on the cox1 sequence. Thus, Tetrahymena sp. 1 (Foissner), Tetrahymena sp. 4 (SIN) and Tetrahymena sp. 6 (Brandl) likely represent new species of Tetrahymena on the basis of cox1 gene sequences.

Interspecific sequence divergence among Tetrahymena sp. 2 (CO), Tetrahymena sp. 3 (RA9) and Tetrahymena sp. 5 (NI) ranged from 0.22 to 1.17 % based on cox1 gene sequences. Furthermore, these three new isolates showed a sequence divergence of 9.08 % from other Tetrahymena species. Given that the range of intraspecific divergence among animals and Tetrahymena species is generally around 1–2 %, these three isolates, together with two T. tropicalis strains (ATCC^® 205083^TM and ATCC^® 205097^TM), likely belong to the same new species, as indicated by less than 0.73 % cox1 sequence divergence among them. Additionally, this is supported by the ecological fact that the three new isolates are all apparently parasites of the guppy Poecilia reticulata, although a parasitic nature has not yet been demonstrated for T. tropicalis.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Our study illustrates the potential utility of cox1 gene sequences in identifying the closely allied species of the ciliate genus Tetrahymena as well as other tetrahymenine species examined. This study extends the spectrum of applicability of universal DNA barcodes for biological identification to ciliophoran protists. In addition, the study showed that the nucleotide length of 689 bp of the cox1 gene is sufficient to provide a diagnostic signal for the identification of tetrahymenine ciliates. Furthermore, this 689 bp cox1 barcoding region gives a unique diagnostic sequence for each taxonomic species, with a few exceptions that may be explained by cultures being misidentified or contaminated. Two noteworthy exceptions, T. pyriformis and T. tropicalis, will require additional taxonomic re-examination, as they may represent new cryptic species complexes.

Our results indicated that the cox1-based identification system should be applied as a tool by culture collections and research laboratories to authenticate and verify their collections, since unintentional mix-ups or losses of cell lines can happen. On one hand, our study indicated that the cox1 barcode will be a powerful alternative in the biological identification of ciliates, especially those whose SSU rDNAs, which have long been used as putative ‘barcodes’, are identical or very similar. This will speed up the discovery of cryptic ciliate species and also provide new insights into ciliate biodiversity. On the other hand, several scientific communities, such as ecologists, population biologists, parasitologists and evolutionary biologists, whose work requires the identification of the organisms, will be able to use our cox1-based identification approach without needing taxonomic expertise on a particular group.

	ACKNOWLEDGEMENTS

 
We are grateful to the ATCC for kindly providing cultures of Tetrahymena. In addition, our sincere gratitude is extended to all scientists who generously provided us with wild isolated cultures: Professor Dr Wilhelm Foissner, Dr Dina Zilberg and Miss Marcia P. Leibowitz. We wish to express our gratitude to Professor Clifford F. Brunk for kindly providing sequences of the cox1 gene for Tetrahymena paravorax and other Tetrahymena species to facilitate our primer design and to Dr Michaela C. Strüder-Kypke for kindly providing sequences of the cox1 gene for some Tetrahymena species for our analyses. This research was supported through funding to the Canadian Barcode of Life Network from Genome Canada through the Genomics Institute, a NSERC of Canada Discovery Grant awarded to D. H. L. and other sponsors (listed at http://www.BOLNET.ca). C. C. was supported by a national scholarship awarded by the Cooperative Research Network (CRN), Government of Thailand. Finally, we are thankful to three anonymous reviewers for their constructive and helpful comments on earlier drafts of the manuscript.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Avise, J. C. (2000). Phylogeography: the History and Formation of Species. Cambridge, MA: Harvard University Press.

Ball, S. L., Hebert, P. D. N., Burian, S. K. & Webb, J. M. (2005). Biological identifications of mayflies (Ephemeroptera) using DNA barcodes. J N Am Benthol Soc 24, 508–524.

Barrett, R. D. H. & Hebert, P. D. N. (2005). Identifying spiders through DNA barcodes. Can J Zool 83, 481–491.[CrossRef]

Barth, D., Krenek, S., Fokin, S. I. & Berendonk, T. U. (2006). Intraspecific genetic variation in Paramecium revealed by mitochondrial cytochrome c oxidase I sequences. J Eukaryot Microbiol 53, 20–25.[CrossRef][Medline]

Borden, D., Whitt, G. S. & Nanney, D. L. (1973a). Electrophoretic characterization of classical Tetrahymena pyriformis strains. J Protozool 20, 693–700.[Medline]

Borden, D., Whitt, G. S. & Nanney, D. L. (1973b). Isozymic heterogeneity in Tetrahymena strains. Science 181, 279–280.[Abstract/Free Full Text]

Borden, D., Miller, E. T., Whitt, G. S. & Nanney, D. L. (1977). Electrophoretic analysis of evolutionary relationships in Tetrahymena. Evolution 31, 91–102.[CrossRef]

Brandl, M. T., Rosenthal, B. M., Haxo, A. F. & Berk, S. G. (2005). Enhanced survival of Salmonella enterica in vesicles released by a soilborne Tetrahymena species. Appl Environ Microbiol 71, 1562–1569.[Abstract/Free Full Text]

Brown, J. W., Miller, S. E. & Horak, M. (2003). Studies on New Guinea moths. 2. Description of a new species of Xenothictis Meyrick (Lepidoptera: Tortricidae: Archipini). Proc Entomol Soc Wash 105, 1043–1050.

Brunk, C. F., Lee, C. L., Tran, A. B. & Li, J. (2003). Complete sequence of the mitochondrial genome of Tetrahymena thermophila and comparative methods for identifying highly divergent genes. Nucleic Acids Res 31, 1673–1682.[Abstract/Free Full Text]

Burger, G., Zhu, Y., Littlejohn, T. G., Greenwood, S. J., Schnare, M. N., Lang, B. F. & Gray, M. W. (2000). Complete sequence of the mitochondrial genome of Tetrahymena pyriformis and comparison with Paramecium aurelia mitochondrial DNA. J Mol Biol 297, 365–380.[CrossRef][Medline]

Corliss, J. O. (1952). Le cycle autogamique de Tetrahymena rostrata. C R Acad Sci 235, 399–402 (in French).

Corliss, J. O. (1953). Silver impregnation of ciliated protozoa by the Chatton-Lwoff technic. Stain Technol 28, 97–100.[Medline]

Corliss, J. O. (1972). Tetrahymena and some thoughts on the evolutionary origin of endoparasitism. Trans Am Microsc Soc 91, 566–573.[Medline]

Corliss, J. O. (1973). History, taxonomy, ecology, and evolution of species of Tetrahymena. In Biology of Tetrahymena, pp. 1–55. Edited by A. M. Elliott. Stroudsburg, PA: Dowden, Hutchinson & Ross.

Corliss, J. O. & Daggett, P.-M. (1983). "Paramecium aurelia" and "Tetrahymena pyriformis": current status of the taxonomy and nomenclature of these popularly known and widely used ciliates. Protistologica 19, 307–322.

Cummings, D. J. (1992). Mitochondrial genomes of the ciliates. Int Rev Cytol 141, 1–64.[Medline]

Czapik, A. (1968). La famille Tetrahymenidae et son importance dans la systématique et l'évolution des ciliés. Acta Protozool 5, 315–357 (in French).

Dawkins, R. (1998). Unweaving the Rainbow: Science, Delusion and the Appetite for Wonder. Boston: Houghton Mifflin.

Elliott, A. M. (1970). The distribution of Tetrahymena pyriformis. J Protozool 17, 162–168.

Elwood, H. J., Olsen, G. J. & Sogin, M. L. (1985). The small-subunit ribosomal RNA gene sequences from the hypotrichous ciliates Oxytricha nova and Stylonychia pustulata. Mol Biol Evol 2, 399–410.[Abstract]

Felsenstein, J. (2004). PHYLIP (phylogeny inference package), version 3.6. Distributed by the author. Department of Genome Sciences, University of Washington, Seattle, USA.

Fenchel, T. & Finlay, B. J. (2004). The ubiquity of small species: patterns of local and global diversity. Bioscience 54, 777–784.[CrossRef]

Folmer, O., Black, M., Hoeh, W., Lutz, R. & Vrijenhoek, R. (1994). DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotechnol 3, 294–297.[Medline]

Gruchy, D. F. (1955). The breeding system and distribution of Tetrahymena pyriformis. J Protozool 2, 178–185.

Hajibabaei, M., Janzen, D. H., Burns, J. M., Hallwachs, W. & Hebert, P. D. N. (2006). DNA barcodes distinguish species of tropical Lepidoptera. Proc Natl Acad Sci U S A 103, 968–971.[Abstract/Free Full Text]

Hebert, P. D. N., Ratnasingham, S. & deWaard, J. R. (2003a). Barcoding animal life: cytochrome c oxidase subunit 1 divergences among closely related species. Proc Biol Sci 270 (Suppl. 1), S96–S99.[CrossRef]

Hebert, P. D. N., Cywinska, A., Ball, S. L. & deWaard, J. R. (2003b). Biological identifications through DNA barcodes. Proc Biol Sci 270, 313–321.[Abstract/Free Full Text]

Hebert, P. D. N., Penton, E. H., Burns, J. M., Janzen, D. H. & Hallwachs, W. (2004a). Ten species in one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly Astraptes fulgerator. Proc Natl Acad Sci U S A 101, 14812–14817.[Abstract/Free Full Text]

Hebert, P. D. N., Stoeckle, M. Y., Zemlak, T. S. & Francis, C. M. (2004b). Identification of birds through DNA barcodes. PLoS Biol 2, e312[CrossRef][Medline]

Hogg, I. D. & Hebert, P. D. N. (2004). Biological identification of springtails (Hexapoda: Collembola) from the Canadian Arctic, using mitochondrial DNA barcodes. Can J Zool 82, 749–754.[CrossRef]

Holz, G. G. & Corliss, J. O. (1956). Tetrahymena setifera n. sp., a member of the genus Tetrahymena with a caudal cilium. J Protozool 3, 112–118.

Janczewski, D. N., Modi, W. S., Stephens, J. C. & O'Brien, S. J. (1995). Molecular evolution of mitochondrial 12S RNA and cytochrome b sequences in the pantherine lineage of Felidae. Mol Biol Evol 12, 690–707.[Abstract]

Jerome, C. A. & Lynn, D. H. (1996). Identifying and distinguishing sibling species in the Tetrahymena pyriformis complex (Ciliophora, Oligohymenophorea) using PCR/RFLP analysis of nuclear ribosomal DNA. J Eukaryot Microbiol 43, 492–497.[CrossRef][Medline]

Kahl, A. (1926). Neue und wenig bekannte Formen der holotrichen und heterotrichen Ciliaten. Arch Protistenkd 55, 197–438 (in German).

Kimura, M. (1980). A simple method of estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J Mol Evol 16, 111–120.[CrossRef][Medline]

Kumar, S., Tamura, K. & Nei, M. (2004). MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief Bioinform 5, 150–163.[Abstract/Free Full Text]

Kumazawa, Y. & Nishida, M. (1993). Sequence evolution of mitochondrial tRNA genes and deep-branch animal phylogenetics. J Mol Evol 37, 380–398.[Medline]

Lynn, D. H. & Strüder-Kypke, M. C. (2006). Species of Tetrahymena identical by small subunit rRNA gene sequences are discriminated by mitochondrial cytochrome c oxidase I gene sequences. J Eukaryot Microbiol 53, 385–387.[CrossRef][Medline]

Mardulyn, P. & Whitfield, J. B. (1999). Phylogenetic signal in the COI, 16S, and 28S genes for inferring relationships among genera of Microgastrinae (Hymenoptera; Braconidae): evidence of a high diversification rate in this group of parasitoids. Mol Phylogenet Evol 12, 282–294.[CrossRef][Medline]

Medlin, L., Elwood, H. J., Stickel, S. & Sogin, M. L. (1988). The characterization of enzymatically amplified eukaryotic 16S-like rRNA-coding regions. Gene 71, 491–499.[CrossRef][Medline]

Meyer, E. B. & Nanney, D. L. (1987). Isozymes in the ciliated protozoa. In Isozymes (Current Topics in Biological and Medical Research, vol. 13), pp. 61–101. Edited by M. C. Rattazzi. New York: Alan R. Liss.

Monaghan, M. T., Balke, M., Gregory, T. R. & Vogler, A. P. (2005). DNA-based species delineation in tropical beetles using mitochondrial and nuclear markers. Philos Trans R Soc Lond B Biol Sci 360, 1925–1933.[Abstract/Free Full Text]

Nanney, D. L. & McCoy, J. W. (1976). Characterization of the species of the Tetrahymena pyriformis complex. Trans Am Microsc Soc 95, 664–682.[Medline]

Nanney, D. L., Meyer, E. B., Simon, E. M. & Preparata, R.-M. (1989). Comparison of ribosomal and isozymic phylogenies of tetrahymenine ciliates. J Protozool 36, 1–8.[Medline]

Nanney, D. L., Park, C., Preparata, R. & Simon, E. M. (1998). Comparison of sequence differences in a variable 23S rRNA domain among sets of cryptic species of ciliated protozoa. J Eukaryot Microbiol 45, 91–100.[Medline]

Nelsen, E. M. & Debault, L. E. (1978). Transformation in Tetrahymena pyriformis: description of an inducible phenotype. J Protozool 25, 113–119.[Medline]

Nyberg, D. (1981). Three new "biological" species of Tetrahymena (T. hegewischi n. sp., T. sonneborni n. sp., T. nipissingi n. sp.) and temperature tolerance of members of the "pyriformis" complex. J Protozool 28, 65–69.

Preparata, R. M., Meyer, E. B., Preparata, F. P., Simon, E. M., Vossbrinck, C. R. & Nanney, D. L. (1989). Ciliate evolution: the ribosomal phylogenies of the tetrahymenine ciliates. J Mol Evol 28, 427–441.[CrossRef][Medline]

Remigio, E. A. & Hebert, P. D. N. (2003). Testing the utility of partial COI sequences for phylogenetic estimates of gastropod relationships. Mol Phylogenet Evol 29, 641–647.[CrossRef][Medline]

Sadler, L. A. & Brunk, C. F. (1992). Phylogenetic relationships and unusual diversity in histone H4 proteins within the Tetrahymena pyriformis complex. Mol Biol Evol 9, 70–84.[Abstract]

Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4, 406–425.[Abstract]

Saunders, G. W. (2005). Applying DNA barcoding to red macroalgae: a preliminary appraisal holds promise for future applications. Philos Trans R Soc Lond B Biol Sci 360, 1879–1888.[Abstract/Free Full Text]

Scheffer, S. J., Lewis, M. L. & Joshi, R. C. (2006). DNA barcoding applied to invasive leafminers (Diptera: Agromyzidae) in the Philippines. Ann Entomol Soc Am 99, 204–210.[CrossRef]

Schlegel, M., Elwood, H. J. & Sogin, M. L. (1991). Molecular evolution in hypotrichous ciliates: sequence of the small subunit RNA genes from Onychodromus quadricornutus and Oxytricha granulifera (Oxytrichidae, Hypotrichida, Ciliophora). J Mol Evol 32, 64–69.[CrossRef][Medline]

Simon, E. M., Meyer, E. B. & Preparata, R. M. (1985). New wild Tetrahymena from Southeast Asia, China, and North America, including Tetrahymena malaccensis, Tetrahymena asiatica, Tetrahymena nanneyi, Tetrahymena caudata, and Tetrahymena silvana n. spp. J Protozool 32, 183–189.[Medline]

Smith, M. A., Fisher, B. L. & Hebert, P. D. N. (2005). DNA barcoding for effective biodiversity assessment of a hyperdiverse arthropod group: the ants of Madagascar. Philos Trans R Soc Lond B Biol Sci 360, 1825–1834.[Abstract/Free Full Text]

Sogin, M. L. & Elwood, H. J. (1986). Primary structure of the Paramecium tetraurelia small-subunit rRNA coding region: phylogenetic relationships within the Ciliophora. J Mol Evol 23, 53–60.[CrossRef][Medline]

Strüder-Kypke, M. C., Wright, A.-D. G., Jerome, C. A. & Lynn, D. H. (2001). Parallel evolution of histophagy in ciliates of the genus Tetrahymena. BMC Evol Biol 1, 5[CrossRef][Medline]

Tautz, D., Arctander, P., Minelli, A., Thomas, R. H. & Vogler, A. P. (2002). DNA points the way ahead in taxonomy. Nature 418, 479[Medline]

Tautz, D., Arctander, P., Minelli, A., Thomas, R. H. & Vogler, A. P. (2003). A plea for DNA taxonomy. Trends Ecol Evol 18, 70–74.[CrossRef]

Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 4673–4680.[Abstract/Free Full Text]

Walsh, P. S., Metzger, D. A. & Higuchi, R. (1991). Chelex 100 as a medium for simple extraction of DNA for PCR-based typing from forensic material. Biotechniques 10, 506–513.[Medline]

Ward, R. D., Zemlak, T. S., Innes, B. H., Last, P. R. & Hebert, P. D. N. (2005). DNA barcoding Australia's fish species. Philos Trans R Soc Lond B Biol Sci 360, 1847–1857.[Abstract/Free Full Text]

Williams, N. E., Buhse, H. E., Jr & Smith, M. G. (1984). Protein similarities in the genus Tetrahymena and a description of Tetrahymena leucophrys n. sp. J Protozool 31, 313–321.

Ziaie, Z. & Suyama, Y. (1987). The cytochrome oxidase subunit I gene of Tetrahymena: a 57 amino acid NH₂-terminal extension and a 108 amino acid insert. Curr Genet 12, 357–368.[CrossRef][Medline]

This article has been cited by other articles:

				E. Gentekaki and D. H. Lynn High-Level Genetic Diversity but No Population Structure Inferred from Nuclear and Mitochondrial Markers of the Peritrichous Ciliate Carchesium polypinum in the Grand River Basin (North America) Appl. Envir. Microbiol., May 15, 2009; 75(10): 3187 - 3195. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Supplementary Tables and Alignments Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Chantangsi, C. Articles by Ikonomi, P. Search for Related Content PubMed PubMed Citation Articles by Chantangsi, C. Articles by Ikonomi, P. Agricola Articles by Chantangsi, C. Articles by Ikonomi, P.