Taxonomy and SSU rDNA-Based Phylogeny of Two Heterotrich Ciliates (Ciliophora, Heterotrichea) Collected From Subtropical Wetlands of China, Including the Description of a New Species, Linostomella pseudovorticella n. sp.

The Heterotrichea Stein, 1859 are a group of ciliated protists (single-celled eukaryotes) that occur in a wide variety of aquatic habitat where they play important roles in the flow of nutrients and energy within the microbial food web. Many species are model organisms for research in cytology and regenerative biology. In the present study, the morphology and phylogeny of two heterotrich ciliates, namely, Linostomella pseudovorticella n. sp. and Peritromus kahli Villeneuve-Brachon, 1940, collected from subtropical wetlands of China, were investigated using morphological and molecular methods. L. pseudovorticella n. sp. differs from its only known congener, Linostomella vorticella Ehrenberg, 1833 Aescht in Foissner et al., 1999, by having more ciliary rows (48–67, mean about 56 vs. 26–51, mean about 42) and its small-subunit (SSU) rDNA sequence, which shows a 15-bp divergence. Although P. kahli has been reported several times in recent decades, its infraciliature has yet to be described. A redescription and improved diagnosis of this species based on a combination of previous and present data are here supplied. Phylogenetic analyses based on SSU rDNA sequences revealed that the genus Linostomella is positioned within Condylostomatidae, and Peritromidae is sister to Climacostomidae with relatively low support, and the family Spirostomidae is the root branch of the class Heterotrichea.

The genus Peritromus is characterized by the strongly dorsoventrally flattened body and the ciliary pattern on dorsal and ventral sides being obviously differentiated (Song and Wilbert, 1997). Although 16 nominal species of Peritromus have been reported, detailed morphological information and molecular data are available for only two, i.e., Peritromus faurei Kahl, 1932and Peritromus kahli Villeneuve-Brachon, 1940(Song and Wilbert, 1997Rosati et al., 2004;Miao et al., 2009). It has been hypothesized that Peritromidae (Peritromus) is the ancestral taxon of Heterotrichea (Chi et al., 2021), although this finding is inconsistent with most other molecular phylogenetic trees (Yan et al., 2015;Fernandes et al., 2016;Chen et al., 2017). Therefore, the phylogenetic position of Peritromidae remains ambiguous.
In the present study, two heterotricheans, L. pseudovorticella n. sp. and P. kahli, were isolated from subtropical wetlands in Ningbo, China (Figure 1). Their taxonomy and phylogeny were investigated based on detailed morphological information and small-subunit (SSU) rDNA sequences.

Sample Collection, Observation, and Identification
Linostomella pseudovorticella n. sp. was collected from a freshwater pond near Yong River (N29 • 58 5.97 ; E121 • 30 26.05 ), Ningbo, China on March 6, 2018. The water temperature was about 12.5 • C. Samples were collected from the upper layer of water using a 20-µm mesh plankton net.
Peritromus kahli population-I was collected from a brackish lake near Meishan Island (N29 • 45 11.36 ; E121 • 54 16.85 ), Ningbo, China on May 20, 2020, when the water temperature was about 27.0 • C and salinity was about 14.1 PSU. Population-II was collected from the intertidal zone of a sandy beach at Xiangshan Bay (N29 • 38 28.07 ; E121 • 46 45.68 ), Ningbo, China on May 21, 2019, when the water temperature was about 24.0 • C and salinity was about 17.6 PSU. Samples were transferred to the laboratory with some pieces of aquatic plant stems and leaves collected from the same habitat.
The behavior of both species was studied in Petri dishes under a dissecting microscope. Their morphology in vivo was observed using bright field and differential interference contrast microscopy (Leica DM2500, Germany) at 100-1,000 × magnifications. The ciliary pattern and nuclear apparatus were revealed by protargol staining (Wilbert, 1975). Terminology followed Song and Wilbert (1997); Rosati et al. (2004), andChi et al. (2020).

DNA Extraction, Gene Amplification, and Sequencing
Clonal cultures of neither species could be established. Thus, single cells of each species were isolated from the original sample, washed three times with filtered habitat water (0.22-µm pore size membrane, Millipore, United States) and twice using ultrapure water, and placed in 1.5-ml microfuge tubes with a minimum volume of water. Genomic DNA was extracted using a DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The SSU rDNA was amplified with the universal eukaryotic primers 18SF (5 -AAC CTG GTT GAT CCT GCC AGT-3 ) and 18SR (5 -TGA TCC TTC TGC AGG TTC ACC TAC-3 ) (Medlin et al., 1988).
The polymerase chain reaction (PCR) conditions for the amplification of SSU rDNA sequences were as follows: a cycle of initial denaturation at 98 • C for 30 s, followed by 35 cycles of amplification (98 • C, 10 s; 56 • C, 20 s; 72 • C, 100 s), and a final extension at 72 • C for 120 s. Q5 Hot Start High-Fidelity DNA Polymerase (NEB Co., Ltd., M0493, Beijing, China) was used to minimize the possibility of PCR amplification errors. PCR products were checked using agarose gel and were sequenced in TSINGKE (Hangzhou, China). Sequence fragments were assembled into contigs using Seqman (DNAStar).

Phylogenetic Analyses
Phylogenetic analyses of SSU rDNA sequences were performed using an alignment comprising 69 representative sequences of Heterotrichea (Supplementary Table 1) and six sequences of Karyorelictea as the outgroup. All sequences were aligned with the MAFFT algorithm applying the default parameters provided on the GUIDANCE web server 1 (Penn et al., 2010a,b). The ends of alignments were trimmed by BioEdit v.7.1.3.0 (Hall, 1999). Hypervariable sites were removed using Gblocks version 0.91b with default setting 2 (Castresana, 2000;Talavera and Castresana, 2007), which resulted in a matrix of 1,552 characters.

Etymology
The species-group name pseudovorticella is a composite of the Greek adjective pseudo-(wrong, lying) and the species group name vorticella, referring to the similar morphology between L. vorticella and L. pseudovorticella.
Family Peritromidae Stein, 1867. Genus Peritromus Stein, 1863. Peritromus kahli Villeneuve-Brachon, 1940. Peritromus kahli had been reported several times but mostly in ecological studies or faunal surveys. Consequently, several morphological characters remain unknown. An improved diagnosis is here supplied based on the present and previous studies.

Morphological Description of Ningbo Population-I
Body size 100-135 µm × 65-75 µm in vivo, 80-145 µm × 60-95 µm after protargol staining (Table 1), length to width ratio approximately 1.5:1 (Figures 4A, 5A). Body outline generally reniform with both ends widely rounded; right cell margin slightly concave, left margin slightly distinctly convex ( Figure 5A). Cell strongly dorsoventrally flattened ( Figure 5C). With an irregular hump on dorsal side, edge of hump decorated with several wart-like prominences through which spine-like cilia (belonging to internal dorsal kinety) project (Figures 4E, 5F). Adoral zone of membranelles (AZM) commences in anterior quarter near the right margin of the cell, extends around anterior end, continues along the left margin, proximal portion bends toward the cytostome that is located in a narrow depression near the left side in mid-region of the body (Figures 4A,D, 5A  irregularly located on dorsal side (Figures 4C, 5J). Cytoplasm highly transparent and colorless ( Figure 5A). Middle part of the cell opaque and dark gray due to the presence of numerous small globules (about 1-2 µm in diameter) and food vacuoles (about 8-18 µm across, containing small algae) (Figures 4A,   5A). Contractile vacuole absent. Two ellipsoidal macronuclear nodules, on average 17 µm × 13 µm in size, one positioned in anterior right 1/3, the other positioned in posterior left 1/3 of the body (Figures 4E, 5B,G); two micronuclei, each closely associated with a macronuclear nodule ( Figure 6F). Locomotion usually by   crawling slowly on substrate. When stimulated or disturbed, cells contract and adhere firmly to the substrate. Infraciliature as shown in Figures 4, 6, including oral apparatus, three kinds of longitudinal ventral kineties, one caudal margin kinety and two circles of dorsal kineties ( Table 1). AZM and paroral membrane almost in parallel, terminating proximally near cytostome. AZM composed of about 63-84 membranelles with cilia about 15-20 µm long in vivo. Paroral membrane composed of numerous obliquely oriented rows of two or three kinetosomes arranged in a line with all kinetosomes ciliated. Eighteen to 23 fiber bundles, about 2-3 µm long in vivo, located between paroral membrane and cytostome (Figures 5E,  6H). Numerous short, rod-like structures, similar to extrusomes, scattered in cytoplasm ( Figure 6F). Ventral kineties composed of dikinetids with both kinetosomes ciliated (cilia about 8-10 µm long in vivo), including 3-8 short preoral kineties (PrK), 6-17 slightly shortened postoral kineties (PoK), 17-24 bipolar kineties (BK), and a caudal margin kinety surrounding the margin of posterior half of ventral side (Figures 4J, 6J). Preoral kineties located anterior of cytostome, postoral kineties located posterior of AZM. Bipolar kineties almost covering entire ventral surface. Caudal margin kinety located between somatic kineties and external dorsal kineties ( Figure 6J). Wide gap near posterior end of ventral surface of cell, probably location of cytoproct (Figures 4J, 6J).
Dorsal kineties comprise one external and one internal dorsal kinety ( Figure 4E). External dorsal kinety forming a complete circle around margin between ventral and dorsal sides, cilia about 6 µm long in vivo (Figures 4E, 5H). Internal dorsal kinety composed of 17-34 kinetosomes, forming a complete circle around margin of dorsal hump, cilia about 10 µm long in vivo (Figures 4E, 5F).

SSU rDNA Sequences and Phylogenetic Analyses
The SSU rDNA sequences are deposited in the GenBank database. The lengths, GC contents, and accession numbers are as follows: L. pseudovorticella n. sp. (1,582 bp, 47.16%, MZ092860), P. kahli population-I (1,602 bp, 44.57%, MZ092861), and P. kahli population-II (1,602 bp, 44.63%, MZ092862). The SSU rDNA sequences of the two populations of P. kahli differ from each other by three nucleotides. The ML and BI analyses based on the SSU rDNA sequences generated phylogenetic trees with nearly identical topologies; therefore, only the ML tree is shown here with support values from both algorithms (Figure 7).
The genus Linostomella is monophyletic as all members grouped into one clade with full support (ML/BI, 100%/1.00). It is sister to the genus Condylostomides with full support (ML/BI, 100%/1.00) within the family Condylostomatidae. In the Linostomella clade, L. pseudovorticella n. sp. clusters with Linostomella sp. (MT175516) with moderate support (ML/BI, 67%/0.85), forming a subclade that is sister to the other subclade that consists of two unidentified Linostomella sequences. These four sequences form a maximally supported group that is sister to L. vorticella (MN783328).
The family Peritromidae forms a maximally supported clade that is sister to the family Climacostomidae although with only poor to moderate support (39%/0.91). The internal relationships within the family Peritromidae remain unresolved as indicated by the low support values of the branches. P. kahli (MT175520) is sister to P. faurei (EU583993) (ML/BI, 57%/0.50), followed by P. kahli (AJ537427) and the newly obtained P. kahli population-II (MZ092862) (ML/BI, 51%/0.67). This cluster forms a polytomy with two populations of P. kahli (KP970237 and population-I MZ092861) and two unclassified Peritromus species (KJ651830 and GQ926915).

DISCUSSION
Comments on Linostomella pseudovorticella n. sp.
The genus Linostomella was erected for L. vorticella, in 1999, the type species by monotypy. The taxonomy and nomenclature of this special species have long been ambiguous due to its unusual morphological characteristics (Ehrenberg, 1833;Wrześniowski, 1870;Penard, 1922;Jankowski, 1978). Over the past two decades, the phylogenetic position of this species was gradually revealed (Foissner et al., 1992(Foissner et al., , 1999Lynn, 2008). Chi et al. (2020) redescribed L. vorticella based on both morphological and molecular information and supplied a taxonomic revision of the genus Linostomella.
Three populations of L. vorticella have previously been reported with data on the infraciliature, and all three overlap in terms of the number of somatic kineties, i.e., 39-45 in a Rheinland-Pfalz population, 26-45 in a Salzburg population, and 37-51 in a Qingdao population (Packroff and Wilbert, 1991;Foissner et al., 1999;Chi et al., 2020). Among these three populations, the morphological characters of the Qingdao population were described in the most detail, and the molecular phylogeny of this population was also analyzed based on SSU rDNA sequence data. We therefore compare L. pseudovorticella (2) the arrangement of the paroral membrane (two parallel rows of kinetosomes in L. pseudovorticella vs. two rows of kinetosomes arranged in "zig-zag" pattern in L. vorticella); (3) the presence of a distinct glabrous area at the posterior end of the cell in L. pseudovorticella (vs. posterior glabrous area lacking in L. vorticella); and (4) the SSU rDNA sequence of L. pseudovorticella (MZ092860) having a 15-bp difference from that of L. vorticella (Figure 8). These differences clearly support the validity of L. pseudovorticella as a separate species. Gelei (1954) reported a species named Condylostoma vorticella, which resembles L. vorticella in all key characters except the number of somatic kineties (60-70 vs. 26-51 in L. vorticella). Although the original description was very brief, we believe that this species is conspecific to the new species because they share the same ciliary pattern. In addition, SSU rDNA sequences of three unidentified Linostomella species (LN870136, MT175516, and LN869952) in the GenBank database differ from our new species at three sites (Figure 8), indicating that these three Linostomella spp. are probably conspecific with L. pseudovorticella n. sp.
The four best-documented populations of P. kahli exhibit small differences in their morphology, e.g., body shape, number of ventral kineties, and number of adoral membranelles (see Table 2). In addition, the differences in the SSU rDNA sequences among the two present populations and other populations with molecular data range from 0 to 3 nucleotides (except for P. kahli KP970237) (Figure 8). All the nucleotide differences between P. kahli (KP970237) and other populations are located in the conserved region for the genus Peritromus, so it is likely that these differences are due to low quality sequencing. Unfortunately, lack of morphological data for P. kahli (KP970237) means that it is not possible to compare it with other populations.

Phylogenetic Analyses Based on SSU rDNA Sequences
Phylogenetic relationships among families in the class Heterotrichea remain unresolved, as previous studies have concluded that either Peritromidae or Spirostomidae could be the basal group in SSU rDNA trees (Shazib et al., 2014;Fernandes et al., 2016;Yan et al., 2016;Chen et al., 2017;Chi et al., 2021). In the present study, Spirostomidae branches before Peritromidae, which is supported by the morphological data and is consistent with the findings of Fernandes et al. (2016) and Chi et al. (2020). Peritromidae is more morphologically complex than Spirostomidae, as the ciliary pattern on both the ventral and dorsal sides of the former family is conspicuously differentiated, and the paroral membrane is prominent and well-developed. However, some molecular studies did not support the early divergence of Spirostomidae (Rosati et al., 2004;Shazib et al., 2014;Yan et al., 2016;Chen et al., 2017). The grouping of Peritromidae and Climacostomidae is questionable, as the tree topology is far from stable. Furthermore, these two families have relatively few morphological similarities with each other. Therefore, more data from additional species are needed to determine the evolutionary relationships among families of the class Heterotrichea. Stentoridae, Blepharismidae, Fabreidae, Folliculinidae, Maristentoridae, and Gruberiidae are closely related in our SSU rDNA tree, which is consistent with previous analyses (Shazib et al., 2014(Shazib et al., , 2016(Shazib et al., , 2019Fernandes et al., 2016;Chen et al., 2017Chen et al., , 2018Luo et al., 2019;Chi et al., 2020Chi et al., , 2021. The monophyletic family Condylostomatidae is divided into two subclades, Condylostomides + Linostomella (freshwater habitat) and Chattonidium + Condylostentor + Condylostoma (marine water habitat). The genus Linostomella is monophyletic and is closely related with Condylostomides, which is consistent with previous studies (Rossi et al., 2016;Chi et al., 2020Chi et al., , 2021. The findings of the present study support the assertion that habitat preference is a phylogenetically informative character among these taxa (Chi et al., 2021). However, the genus Condylostoma failed to form a monophyletic group in many phylogenetic analyses (Miao et al., 2009;Shazib et al., 2014;Yan et al., 2015;Fernandes et al., 2016;Chen et al., 2020;Chi et al., 2021). The result in the present study is the same as before and supports the view in Chen et al. (2020) that Condylostoma is likely a paraphyletic group.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

AUTHOR CONTRIBUTIONS
DJ performed the experiments and drafted the manuscript. JH performed the phylogenetic section. TY and XZ helped to collect the samples. AW and SA-F checked all the taxonomic works and helped to writing the manuscript. XC supervised and coordinated the work. All authors read and approved the final manuscript.