New Data Define the Molecular Phylogeny and Taxonomy of Four Freshwater Suctorian Ciliates With Redefinition of Two Families Heliophryidae and Cyclophryidae (Ciliophora, Phyllopharyngea, Suctoria)

Four suctorian ciliates, Cyclophrya magna Gönnert, 1935, Peridiscophrya florea (Kormos & Kormos, 1958) Dovgal, 2002, Heliophrya rotunda (Hentschel, 1916) Matthes, 1954 and Dendrosoma radians Ehrenberg, 1838, were collected from a freshwater lake in Ningbo, China. The morphological redescription and molecular phylogenetic analyses of these ciliates were investigated. Phylogenetic analyses inferred from SSU rDNA sequences show that all three suctorian orders, Endogenida, Evaginogenida, and Exogenida, are monophyletic and that the latter two clusters as sister clades. The newly sequenced P. florea forms sister branches with C. magna, while sequences of D. radians group with those from H. rotunda within Endogenida. The family Heliophryidae, which is comprised of only two genera, Heliophrya and Cyclophrya, was previously assigned to Evaginogenida. There is now sufficient evidence, however, that the type genus Heliophrya reproduces by endogenous budding, which corresponds to the definitive feature of Endogenida. In line with this and with the support of molecular phylogenetic analyses, we therefore transfer the family Heliophryidae with the type genus Heliophrya to Endogenida. The other genus, Cyclophrya, still remains in Evaginogenida because of its evaginative budding. Therefore, combined with morphological and phylogenetic analysis, Cyclophyidae are reactivated, and it belongs to Evaginogenida.


INTRODUCTION
Ciliates are complex and well-developed single-celled eukaryotes which are mainly characterized by having cilia in their life history (Corliss, 1979;Lynn, 2008;Zhang et al., 2020). Ciliates have been studied for over three centuries, and estimates of the number of free-living ciliate species vary from three thousand to thirty thousand (Finlay et al., 1996;Foissner et al., 1999). The fact that ciliates are highly diverse and omnivorous means that they are considered to be a major link in the microbial food web and to play an important role in energy flow and material circulation in aquatic environments (Chi et al., 2020;Wu et al., 2020).
The subclass Suctoria Claparede & Lachmann, 1858, is a special group of ciliates. While the asexual reproduction of most ciliates is achieved by binary fission, the reproduction mode of suctorians is budding. This means that suctorians are polymorphic, with two distinct stages in their life history. Specifically, the sessile trophonts are usually non-ciliated but possess tentacles, while the free-swimming swarmers (larval forms) are typically ciliated (Chen et al., 2005(Chen et al., , 2008aLynn, 2008;Song et al., 2009;Hu et al., 2019). Suctoria is divided into three orders based on their different modes of budding: Exogenida Collin, 1912, Endogenida Collin, 1912, and Evaginogenida Jankowski, 1978. This classification system is widely accepted by researchers, although some other more complex classifications have been proposed by protozoologists (Kormos and Kormos, 1958;Dovgal, 2002). There are about 560 suctorian ciliates widely distributed in various environments, such as marine, freshwater, and soil, as well as in the digestive tract of other organisms, as ectosymbionts on diverse invertebrates, or sometimes as endocommensals in hosts (Matthes, 1988;Foissner et al., 1999Foissner et al., , 2002Chen et al., 2008a,b;Marіño-Pérez et al., 2011;Hu et al., 2019). Most free-living suctorians are carnivorous, feeding primarily on other ciliates and flagellates, and thus, they are important components of the microbiological food web as predators (Lynn, 2008).
The characteristics of suctorian ciliates are mainly summarized as follows: (1) Body shapes are highly variable, from simple spheroid to flattened discs to complex branching forms; (2) tentacles are highly diverse, including prehensile, clavate, rod-like, and branched tentacles, which may be clustered in fascicles or scattered across the whole cell surface; (3) a lorica may be present or absent; (4) stalks are non-contractile, including both a real stalk and a stylotheca protuberance which is an extension of the posterior end of lorica; and (5) swarmer shape and infraciliature are important features for the identification of suctorians. Due to their highly diversified morphology and the fact that silver staining methods cannot be widely used for suctorians, there have historically been a mass of confusions and errors in the literature on suctorians. In recent years, however, research into suctorians is modernizing, and the more extensive application of staining, electron microscope, and molecular methods to their study has increased the availability of infraciliature, ultrastructure, and multi-gene sequence information. As a result, the taxonomic standard of suctorian ciliates is gradually improving (Batisse, 1994;Foissner et al., 1995;Chen et al., 2008a,b;Marіño-Pérez et al., 2011;Zhao et al., 2014).
In the present study, four suctorian ciliates, Cyclophrya magna Gönnert, 1935, Peridiscophrya florea (Kormos & Kormos, 1958) Dovgal, 2002, Heliophrya rotunda (Hentschel, 1916) Matthes, 1954, andDendrosoma radians Ehrenberg, 1838, were isolated from a freshwater lake in Ningbo, China. They were investigated both in vivo and by using staining methods. Molecular data were reported for the first time for the latter three species, and the phylogenetic relationships within Suctoria were also analyzed based on SSU rDNA sequences.

Sample Collection, Observation, and Identification
Four species were collected from a subtropical freshwater lake, Rihu Lake (N29°53′32′′; E121°33′45′′), in Ningbo, China (Figure 1). C. magna is relatively common in summer when the water temperature is about 25°C. It was collected using artificial substrates (glass slides) which were immersed in water at a depth 0.5-1.0 m for 7 to 15 days during June 2016. Peridiscophrya florea was separated from the surface of fresh willow roots (Salix babylonica) immersed in water during May 2016 when the water temperature was about 20°C. Heliophrya rotunda was also collected using artificial substrates in January 2017 when the water temperature was about 9°C. D. radians was separated from the immersed surface of water hyacinth (Eichhornia crassipes) in February 2017 when the water temperature was about 11°C. All collected ciliates died within one or 2 days when maintained with habitat water at room conditions, regardless of being cultured with ciliates in situ or Paramecium sp. Thus, we could not culture either suctorian ciliate. It was possible, however, to separate enough individuals of the four species for morphological and molecular research.

DNA Extraction, PCR Amplification, and Gene Sequencing
For every species, cells were optically identified and peeled off the substrates using an anatomic needle. Single cells were isolated and washed four times in ultra-pure water and then placed in 1.5-ml microfuge tubes with 45 μl of buffer. Genomic DNA was extracted with the Dneasy Blood and Tissue Kit (Qiagen, Hilden, Germany) following the manufacturer's instructions. Polymerase chain reaction (PCR) amplification of the SSU rDNA was performed using Q5 ® Hot Start High-Fidelity DNA Polymerase (NEB Co., Ltd., M0493, Beijing) with the Frontiers in Microbiology | www.frontiersin.org universal eukaryotic primers 82F (5′ GAA ACT GCG AAT GGC TC 3′) and 18S-R (5′ TGA TCC TTC TGC AGG TTC  ACC TAC 3′) (Medlin et al., 1988). An E.Z.N.A.™Quik Gel Extraction Kit (OMEGA Bio-Tek, D2500-01, Guangzhou) was used to purify PCR products, and a pEASY ® -T1 Cloning kit (TransGen, CB101, Beijing, China) was used for cloning. Sequencing was performed bidirectionally (BGI Co., Ltd., Shanghai, China).

Phylogenetic Analyses
The newly characterized SSU rDNA sequences, and the sequences of another 39 species/populations obtained from the NCBI GenBank database, were used for phylogenetic analyses. Although C. magna (AY007445, AY007446, AY007447, AY007448, and AY007449) is referred to as Heliophrya erhardi in the NCBI database, H. erhardi is a synonym for C. magna, and therefore, the latter name was used in the phylogenetic analyses reported here. Sequences were aligned using the GUIDANCE algorithm (Penn et al., 2010a) with MUSCLE parameters in the GUIDANCE web server (Penn et al., 2010b). Ambiguously aligned sites were refined using Gblocks v.0.91b (Castresana, 2000), and ambiguous columns were removed based on confidence scores calculated by GUIDANCE. Bayesian inference (BI) and maximum likelihood (ML) analyses were carried out online on the CIPRES Science Gateway v 3.3). 1 The best fitting model for phylogenetic analyses was selected by MrModeltest v2.2 (Nylander, 2004). Bayesian inference (BI) analysis was performed with MrBayes 3.2.6 (Ronquist et al., 2012) using the GTR + I + G evolutionary model (Nylander, 2004). The program was run for 1,000,000 generations with a sample frequency of 100, and a burn-in of 2500. ML analysis was performed with RAxML-HPC2 on XSEDE v8 (Stamatakis, 2014) using the GTR + I + G model as selected by Modeltest v3.4 (Posada and Crandall, 1998). The reliability of the ML internal branches was assessed using a nonparametric bootstrap method with 1000 replicates. MEGA v5.0 (Tamura et al., 2011) was used to visualize tree topologies. 1 http://www.phylo.org/portal2   Gönnert, 1935 Cyclophrya magna is mainly characterized by the disc-like body and multiple tentacles in fascicles. This well-known species has been reported several times in the half century since the original description. Most of these reports, however, focused on the ultrastructure of the tentacles, rather than attempting to present accurate morphological data and taxonomic research (Hauser and Eys, 1976;Spoon et al., 1976;Hanke-Bücker et al., 2000). There is therefore a need to provide an improved diagnosis here.

Improved Diagnosis
Disc-like body about 50-190 μm in diameter, some of them are oval. Transparent adhesive disc obvious, with a ring width of about 3-12 μm. Capitate tentacles clustered in numerous fascicles along the body margin, mostly in 3-9 fascicles. Contractile vacuoles about 3-14. Macronucleus branched.

Morphological Description of Ningbo Population
Trophont body flat disc-shaped, without lorica or stalk, directly attached to substrates using a transparent adhesive disc (Figures 2A, 3A,E). Body size 70-160 μm × 60-160 μm in vivo, usually about 90 μm × 100 μm, adhesive disc width about 3-11 μm. Capitate tentacles straight and clustered in 3-9 fascicles (usually in four fascicles; Table 1) Swarmer formed by evaginative pattern (Figure 3K). Newly born swarmer swimming freely in water, slender ellipsoid, or finger-shaped, about 100 μm × 35 μm in vivo, 80 μm × 30 μm after protargol impregnation. Body surface densely covered with cilia and arranged in nine longitudinal ciliary rows (Figures 2M,N The taxonomic position of Peridiscophrya florea has been changed many times since its first discovery, and an accurate morphological description is also still lacking. It was discovered in Hungary and reported by Kormos and Kormos (1958) who described the type species of their new genus Catharina, named Catharina florea. Two years later, Catharina was substituted as Caracatharina Kormos, 1960in Corliss (1960. Matthes (1988) transferred it to Discophrya Claparede & Lachmann, 1859. Dovgal (2002) moved it to Peridiscophrya without any description or illustration. Zharikov et al. (2005) only supplied a short redescription based on a Russian population. Thus, an improved diagnosis is needed here based on detailed morphological characters.

Improved Diagnosis
Ellipsoidal or finger-shaped body enclosed in the distal region cup-like lorica, about 55-200 μm × 15-75 μm in vivo. Lorica colorless and transparent, apical cup-shaped, antapical prolonged like a stalk (stylotheca), about 1-2.5 times length of upper part. About 100-200 capitated tentacles concentrated on the apical surface of the body, up to 200 μm in length. Contractile vacuoles present, about 1-4 in number. Macronucleus filiform, sometimes with short branches.

Morphological Description of Ningbo Population
Cell body ellipsoidal or finger-shaped, about 55-200 μm × 15-75 μm in size, usually about 120 μm × 40 μm (Table 1)  Free-swimming swarmer not observed. Newly attached individual spherical, exposed at the top of lorica, cytoplasm colorless with many granular inclusions scattered in upper part of cell. Capitate tentacles scattered across the whole cell surface (Figures 4B,C, 5J). Transparent stylotheca, without cup-shaped structure, with longitudinal stripes on the surface and many gray particles arranged in a line along the inner axis (particles absent in trophont individuals, function unknown; Figures 4B-D, 5E), terminal adhesive disc obvious. Macronucleus curved sausage shape ( Figure 5I). Contractile vacuoles not confined to apical region (Figures 4B,C, 5J,K). Subsequently, upper part of lorica gradually expanded to form a cocktail glass-shaped structure (Figures 4D, 5K). Then, lorica cup growing bigger and longer, gradually enclosing cell body which is compressed to become ellipsoidal or fingershaped (Figures 4E, 5L), contractile vacuoles and tentacles also moved to apical region (Figures 4E, 5L) (Hentschel, 1916) Matthes, 1954 Heliophrya rotunda is a common freshwater species which has been found several times since its discovery (Hentschel, 1916;Saedeleer and Tellier, 1930;Gönnert, 1935;Rieder, 1936Rieder, , 1988Matthes, 1954Matthes, , 1988Mogensen and Butler, 1982; Foissner  , 1995). A clear definition has never been provided, however, and its classification is still controversial. Accurate taxonomic identification and improved diagnosis are therefore needed based on well characterized morphological features.
Transparent adhesive disc about 6 μm in width. Capitate tentacles up to 190 μm long and arranged in many fascicles along cell margin. About 3-22 contractile vacuoles. Single macronucleus, oval or kidney-shaped.

Morphological Description of Ningbo Population
Stalkless body flat disc-shaped, without lorica, attached to the substrates by transparent adhesive disc (Figures 4F, 6A). Body diameter 40-80 μm in vivo, usually about 60 μm (  (Figures 4H-J, 6A,D). Cytoplasm colorless but usually greenish or brownish due to mass of green granular inclusions (Figures 6B-E). Macronucleus shape slightly variable, usually oval, sometimes kidney-shaped, always located in the center of cell (Figures 6E-H Dendrosoma radians is a well-known suctorian ciliate with a ramified body. Since the original report by Ehrenberg (1838), it has been reinvestigated many times in the last two centuries (Kent, 1882;Hickson and Wadsworth, 1909;Penard, 1920;Gönnert, 1935;Bick, 1972;Matthes, 1988;Foissner et al., 1995;Dovgal, 1996). Here, we combine all the historic descriptions and present data to provide an improved diagnosis.

Improved Diagnosis
Ramified body in the form of an individual tree or colony. Huge differences in cell size, from 150 to 5000 μm in height, 5-70 μm in width of stem. Each end of branch with one fascicle of capitate tentacles, fully extended tentacles about 250 μm in length. Contractile vacuoles numerous, randomly distributed along the stem and branches. Macronucleus filiform and expanded in stem and branches. Swarmer ellipsoid.

Description of Ningbo Population
Ramified body stalkless, without lorica, attached to substrates by basal body surface. Younger individuals colorless to light gray, straight or curved stick-shaped, not branched, usually about 200 μm tall (Figures 7C, 8F). Developing individuals gray or brownish, main body with several branches, body size about 1,000 μm (Figures 7D,E, 8B,D). Developed individuals brownish, ramified body composed of several stems and numerous branches, stems connected to each other by irregularly shaped baseplate; ramified body size up to 2,270 μm tall (Figures 7A, 8C and Table 1). Distal portion of branch colorless and translucent, with slight contractility; contracted branch spring-like with many folds; end of branch flat or slightly protuberated, covered with numerous capitate tentacles (Figures 7B, 8A,G,H). Highly contractile capitate tentacles; fully extended ones about 50-250 μm in length, contracted ones spring-like with expansion of spherical ends ( Figure 8H). Under unfavorable conditions, younger individuals gradually melted and disappeared ( Figure 8I). Contractile vacuoles numerous but highly variable in number (4-70), irregularly distributed along the body stem and branches, and even the irregular shaped baseplate (Figures 7A, 8J). Macronucleus filiform and interspersed in body stems and branches, including baseplate, usually in elongated belt, some parts broken or fused into expanded nodes (Figures 7F-L).

Phylogenies Inferred From the SSU rDNA
The SSU rDNA-based tree was constructed as shown in Figure 9. Since the topologies of the ML and BI trees were basically concordant just the topology of the ML tree is presented with support values from both algorithms indicated on the branches. The analysis includes all species of Suctoria for which SSU rDNA sequence data are available and five species of Cyrtophoria as the outgroup. It reveals that all three orders within Suctoria are monophyletic; that Evaginogenida and Exogenida are sister taxa; that Heliophrya rotunda and D. radians fall within Endogenida; and that C. magna and Peridiscophrya florea cluster together in Evaginogenida. Corliss, 1979 andCyclophryidae Jankowski, 2007 Cyclophrya magna was found by Gönnert (1935) and described as the type species of his newly established genus Cyclophrya. This genus was originally attributed to the family Dendrosomatidae Fraipont, 1878. At that time, Dendrosomatidae included another discoid genus, Heliophrya Saedeleer & Tellier, 1930. Most species in Dendrosomatidae, however, possess actinophores bearing tentacles. Since Cyclophrya and Heliophrya do not have these structures, Corliss (1979) established a new family for these two genera, Heliophryidae. Two years later, Jankowski put forward a point that Solenophrya crassa Claparede & Lachmann, 1859 is an older synonym for the name C. magna. Accordingly, he proposed to rename the family Heliophryidae Corliss, 1979 (which included the genus Cyclophrya) as Solenophryidae Jankowski, (1981). Seven years later, Rieder (1988) Rieder, 1936, Heliophrya erhardi (Rieder, 1936) Matthes, 1954and Trichophrya maxima Oppenheim, 1957 are older synonyms for Cyclophrya magna Gönnert, 1935. In the present study, we verify that the reproductive modes of Heliophrya and Cyclophrya are indeed as described in Rieder (1988); that is, the budding mode of Heliophrya is endogenous, while Cyclophrya is evaginative. Since within Suctoria, the reproductive mode is the sole basis for the classification of order taxa, Heliophryidae/Heliophrya must indeed be transferred from Evaginogenida to Endogenida, while the genus Cyclophrya remains in Evaginogenida and still in Cyclophryidae Jankowski, 2007. The present phylogenetic analyses support this morphologically based reassignment: that is, Heliophrya is completely aggregated in Endogenida, and Cyclophrya is aggregated in Evaginogenida. In other words, our research fully supports the establishment of Cyclophryidae Jankowski, 2007 based on morphological and molecular data. Hitherto, apart from Cyclophrya Gönnert, 1935 ( Figure 10A) and Heliophrya Saedeleer & Tellier, 1930, three other genera possess a disc-shaped body, all of which belong to the order Evaginogenida, namely Discosomatella Corliss, 1960 (Figure 10D), Dendrocometes Stein, 1852 (Figure 10B), and Niscometes Jankowski, 1987 ( Figure 10C). Cyclophrya is easily distinguished from Discosomatella by the arrangement of tentacles in each group (disordered clustering vs. arranged radially in a line; Dovgal, 2002). Cyclophrya differs from Dendrocometes and Niscometes in the shape of its tentacles (straight and capitate vs. branched and terminal tapering; Dovgal, 2002). Gönnert, 1935 Cyclophrya magna was first reported by Gönnert in 1935. Matthes (1988 provided a short description of this species and considered two other species as synonymous with it; that is, Craspedophrya erhardi Rieder, 1960 and H. erhardi (Rieder, 1936) Matthes, 1954. Matthes (1954 mentioned that this species had both evaginative and endogenous types of budding, which did not meet the classification standard for Suctoria. Furthermore, Matthes (1988) showed that C. magna had between 1 and 14 tentacle fascicles, and the range is too wide for a suctorian species. It is therefore suspected that other species have been confused in Matthes' description of C. magna. In Dovgal (2013), the main differences between C. magna and C. katharinae are arrangement of tentacles and the body size. It can be assumed that these are the manifestation of intraspecific variability. Therefore, C. katharinae is regarded as a junior synonym for C. magna. Hence, Cyclophrya is a monotypic genus.

Cyclophrya magna
Peridiscophrya florea (Kormos & Kormos, 1958) Dovgal, 2002 Peridiscophrya florea was first reported as Catharina florea Kormos & Kormos, 1958. Two years later, it was transferred to Caracatharina Kormos, 1960. Matthes (1988 thought that Caracatharina was a synonym of Discophrya Claparede & Lachmann, 1859 and moved this species to Discophrya. Recently, Dovgal (2002) transferred it to Peridiscophrya Nozawa, 1938 which was mainly characterized by the cylindrical or finger-like body covered with lorica, a sturdy stylotheca, capitate tentacles in a single fascicle, and a ramified macronucleus. Peridiscophrya differs from Discophrya mainly in the stylotheca (Figures 10H,I). Based on these descriptions and comparisons, we agree with Dovgal's classification of this species.
There are two other nominal species in this genus, that is, Peridiscophrya japonica Nozawa, 1938 ( Figure 10K) and P. crassipes (Rieder, 1936) Dovgal, 2002 (Figure 10J). The type species, P. japonica, was collected by Nozawa from Kyoto, Japan, and differs from P. florea in three aspects of the lorica. Namely, in P. japonica, the lorica enveloped the whole cell body (vs. the anterior part of body leaked outside of the lorica), is smooth (vs. lorica horizontally striped in P. florea), and possesses a short and smooth stylotheca that extends up to one-third of the apical part of the lorica (vs. slender and vertically striped, about 1-2.5 times the upper part of lorica in P. florea).
Peridiscophrya crassipes was collected from a fishpond in Switzerland by Rieder (1936) (Figure 10J). It also differs from P. florea in the character of its lorica (covering one-third to half of body length vs. covering four-fifths or whole of body length), as well as in the location of its contractile vacuoles (end of the body vs. top of the body).
Heliophrya rotunda (Hentschel, 1916) Matthes, 1954 Heliophrya rotunda was first reported by Hentschel (1916) and named as Trichophrya rotunda. This was an obvious misidentification, however, because this species lacks actinophores to support the fascicle of tentacles and thus should not be classified into the genus Trichophrya. Saedeleer and Tellier (1930) found a similar organism, but they missed Hentschel's publication and described it as a new species, that is, Heliophrya collini Saedeleer & Tellier, 1930. Matthes (1954 combined Trichophrya rotunda with the genus Heliophrya, and treated  Gönnert, 1935(after Dovgal, 2002.
The original description of H. sinuosa provided by Rieder (1936) was very short and simple ( Figure 10F). It was named as H. rotunda var. sinuosa, and Jankowski (1981) treated it as a separated species without giving any reason. The main difference between H. sinuosa and H. rotunda (Figure 10G) is that the former has a regular wavy outline. Considering that the abundance of food will affect the morphology of suctorians, we deduce that these two species may be synonymous.
Heliophrya minima is a small species and only about 35 μm in diameter. It is characterized by its unfascicled tentacles ( Figure 10E): There are up to 22 single tentacles distributed along the cell margin. According to the above characters, it can be easily distinguished from H. rotunda. (Matthes, 1988;Foissner et al., 1995).

Dendrosoma radians Ehrenberg, 1838
Dendrosoma was established by Ehrenberg (1838) for the multibranched species, D. radians Ehrenberg, 1838. In the following century, several similar genera were erected, but they were all considered to be synonyms of the genus Dendrosoma. For example, Perez (1903) erected genus Lernaeophrya due to the presence of numerous short actinophores; Astrophrya Awerintzew, 1904 was established because of long actinophores and rather short tentacles; Swarczewsky (1928) reported three genera based on organisms found in Lake Baikal: Baikalophrya, Baikalodendron, and Gorgonosoma, which were characterized by the possession of ramified actinophores or a flattened body sprawled over the substrates. In fact, several authoritative protozoologists have considered all these genera to be synonyms of Dendrosoma (Hickson and Wadsworth, 1909;Foissner et al., 1995;Dovgal, 2002;Lynn, 2008).
Apart from Dendrosoma, Dendrosomides Collin, 1906, also has a branched body and numerous actinophores. The latter, however, belongs to the family Dendrosomididae Jankowski, 1978 and exhibits a reproduction mode of exogenous budding. In terms of morphological characteristics, Dendrosomides can be easily distinguished from Dendrosoma ( Figure 10M) by the possession of a stalk; that is, the branched body of Dendrosomides is connected to the substrates by a real stalk (Foissner et al., 1995; Figure 10L).
There are only two nominal species in the genus Dendrosoma, D. radians and D. capitata (Perez, 1903) Dovgal, 2002. D. capitata was originally reported as Lernaeophrya capitata by Perez (1903), and Dovgal combined it into Dendrosoma and regarded Lernaeophrya as a synonym of Dendrosoma. In fact, this species differs from D. radians mainly in the number of actinophores (more than 12 vs. about 10 in D. radians).
Considering that the number of branches and actinophores of D. radians are highly variable, however, we deduce that D. capitata is a synonym of D. radians.

Phylogenetic Position of Four Species and Phylogenetic Relationships in Suctoria
In the ciliate classification system, Suctoria is a special subclass in the phylum Ciliophora. Unlike other ciliates, the suctorians are characterized by a lack of cilia in the trophont, dense cilia covering the body in swarmers, and a polymorphic life cycle. In the system sorted by Corliss (1979), the subclass Suctoria was divided into three suborders: Endogenina Collin, 1912), Exogenina Collin, 1912, and Evaginogenina Jankowski, 1979. This classification system has been widely accepted by subsequent researchers. The Suctoria was divided into four subclasses in the classification system provided by Dovgal (2002): Evaginogenia Jankowski, 1978, Endogenia Collin, 1912, Vermigenia Jankowski, 1978, and Exogenia Collin, 1912. Apart from the three germination modes mentioned in Corliss (1979) (evaginative, endogenous, and exogenous), the fourth mode suggested by Jankowski (1978) was discussed in detail for the first time, namely, vermigemmy. The swarmers of Vermigenia are devoid of ciliature and crawl onto the surface of hosts using a special larval adhesive organelle (tentacle). In view of the lack of cilia in their swarmers, Jankowski established the fourth subclass, Vermigenia. However, the budding modes of these organisms are same as subclass Exogenia, and Dovgal deduced that Vermigemmins should be derived from exogenous ancestors. Lynn (2008) did not accept the fourth subclass and prefer to use the traditional three-group system. We will continue to follow Lynn (2008) classification system until there is new sufficient morphological and molecular evidence of vermigemmy species.
There are few phylogenetic studies on Suctoria due to the lack of molecular information (Snoeyenbos-West et al., 2004). A recent molecular phylogenetic study indicated that the three suctorian orders were divided into three distinct clusters, which is relatively consistent with the classification based on the three reproductive modes of budding (Zhao et al., 2014). In our present research, however, Evaginogenida and Exogenida clustered together and then form sister branches with Endogenida. This clustering also has a good correlation with the budding mode, that is, endogenous budding starts with an invaginated part of the cortex, the brood pouch, and free moving swarmers are formed in the mother cell. Exogenous budding, meanwhile, takes place essentially on the surface of the trophont, and the swarmer pinches off the surface of the mother cell directly to the environment. Evaginative budding involves the formation of a temporary brood pouch, and the swarmer is not freed within the mother cell, but the entire wall of the pouch evaginates out to the cell surface; thus, the process of cytokinesis is similar to exogenous budding (Dovgal, 2002;Lynn, 2008). In other words, exogenous and evaginative budding are ultimately formed on the surface of mother cells, and the free moving swarmer detaches off directly to the environment. Our present phylogenetic tree therefore more reasonably reveals the genetic relationships of the three orders, although the confidence value is low (54% ML, 0.86 BI).
The present study provides the SSU rDNA sequences of four genera/species. Three of these are reported for the first time, namely, D. radians, Heliophrya rotunda, and Peridiscophrya florea. In the phylogenetic tree, D. radians, C. magna, and P. florea fall in the clade as expected. The position of H. rotunda is controversial to Lynn (2008); that is, it falls within Endogenida instead of Evaginogenida. Although H. rotunda has been reinvestigated several times, the mode of reproduction had never been mentioned until Matthes (1954) reported its endogenous reproduction. Corliss (1979) established Heliophryidae for this genus, but classified it into Evaginogenida. Mogensen and Butler (1982) also observed that the reproduction mode of H. rotunda was endogenous, instead of evaginative. Jankowski (2007) transferred Heliophryidae from Evaginogenia to Endogenia. Based on the above discussion, the family Heliophryidae and genus Heliophrya should be transferred to Endogenida.
Until now, there has been no report of the phylogenetic relationship within suctorian order taxa due to the lack of sequence information. The new sequences presented here, however, mean that there are now six genera with sequence information in Endogenida, which gives us the opportunity to explore the family/genus-level relationships. In the current phylogenetic tree, endogenid ciliates are divided into two distinct clusters which are completely consistent with whether or not they have lorica, but the other characters are not well represented. It is impossible to extend the discussion to the relationships within Evaginogenida and Exogenida, however, due to the continuing lack of related sequences.

DATA AVAILABILITY STATEMENT
The data presented in the study are deposited in the NCBI database repository, accession number(s) MZ912680-MZ912683.

AUTHOR CONTRIBUTIONS
MM carried out the experiments, phylogenetic analyses, and drafted the manuscript. YL, QY, and XZ helped to collect samples and performed some experiments. JH, HM, and KA-R helped to write the manuscript, and XC conceived and designed the paper. All authors read and approved the final manuscript.