Taxonomy and Phylogeny of Two Spirotrichean Ciliates (Alveolata, Protista, and Ciliophora) From a Plateau Saline–Alkaline Lake in Tibet, China, Including Description of a New Species

There are few researches concerning ciliates from “extreme” environments such as plateau salt lakes; nevertheless, such a faunistic survey is of great significance for understanding the diversity and biogeography of these microeukaryotes. In this study, two ciliates, namely, Hemiurosomoida tibetensis sp. nov. and Euplotes euryhalinus Valbonesi and Luporini, 1990, isolated from Kyêbxang Co, a Tibet Plateau saline-alkaline lake, China, were investigated based on observations of live and silver-stained specimens, and 18S rDNA sequences were analyzed. The new species can be characterized by having a size of 60–95 × 20–40 μm in vivo, an elongate elliptical body, colorless cortical granules mainly grouped around cirri and dorsal bristles, about 21 adoral membranelles, 20 cirri on each marginal row, and two macronuclear nodules and two micronuclei. The Chinese population of E. euryhalinus corresponds well with the original one. Phylogenetic analyses show that Hemiurosomoida is non-monophyletic, H. tibetensis sp. nov. has a closer relationship with H. longa within this genus, and that Euplotes euryhalinus presents a wide geographic distribution. This study first reveals the ciliature, morphogenesis, and rRNA gene sequence data for H. tibetensis sp. nov. and the Chinese population of E. euryhalinus, which thus increases our knowledge about the diversity of ciliates at high altitudes.

Euplotes Ehrenberg, 1830, is a speciose ciliate genus and several taxonomists have recognized more important characters for species delineation (Tuffrau, 1960;Borror, 1972;Carter, 1972). Curds (1975) provided the most recent taxonomic revision (Berger, 2011). With the addition of new members recently (Jiang et al., 2010a,b;Pan et al., 2012;Chen et al., 2013;Fotedar et al., 2016;Syberg-Olsen et al., 2016;Ivalji et al., 2020;Lian et al., 2020Lian et al., , 2021Abraham et al., 2021), so far more than 160 species and sub-species have been included in Euplotes s. l (Berger, 2001). However, the identification of Euplotes species remains difficult even though many species have been investigated using a range of modern methods for (1) remarkable intraspecific variation of diagnostic characters even within clonal cultures; (2) poor original report and self-contradiction of re-descriptions in certain common species (Kahl, 1930(Kahl, -1935Burkovsky, 1970;Carter, 1972;Agatha et al., 1993;Petz et al., 1995;Song and Wilbert, 2002). Euplotes euryhalinus was first reported by Valbonesi and Luporini (1990) from cold water in Antarctica, with a pattern of dorsal argyrome and ciliature under protargol-staining and electron microscopy supplied, but lacking morphological information in vivo and corresponding molecular data. In this study, it was rediscovered, which provided a good chance to supplement the description of the species.
The Kyêbxang Co is a plateau saline-alkaline lake at an altitude of approximately 4,660 m above sea level. The PFUs were placed in the lake at a depth of 50 cm for approximately 2 months (from August to October 2019) as artificial substrata to colonize ciliates. The water temperature, salinity, and pH are 8.81 • C, 41.39 , and 10.13, respectively, when the PFUs were collected. The sponges and habitat water were then transported to the laboratory for further treatment. Isolated specimens were maintained in the laboratory for approximately 1 week as raw or uni-protistan cultures.
Living cells were observed with a light microscope equipped with differential interference contrast (100-1,000 ×). The protargol staining method was used to reveal the ciliature and nuclear apparatus (Wilbert, 1975). The Chatton-Lwoff silver nitrate impregnation was applied to reveal the silverline system (Foissner, 2014). The DAPI silver samples were incubated with 50% formalin solution (1:1 v/v) at room temperature for 1 min and stained with DAPI solution at 1.25 µg/ml final concentration to visualize the nuclei (Gong et al., 2020). Counts and measurements of stained specimens were performed at a magnification of × 1,250. Drawings of living cells were produced based on freehand sketches and photomicrographs, and drawings of silver-stained specimens were made with the help of a camera lucida.

DNA Extraction, PCR Amplification, and Sequencing
For each species, a single cell was picked directly from the original sample and washed five times using autoclaved ultrapure water before DNA extraction Genomic DNA was extracted using DNeasy Blood and Tissue Kit (QIAGEN, Hilden, Germany), following the manufacturer's instructions. Primers 18S-F (5 -AAC CTG GTT GAT CCT GCC AGT-3 ) and 18S-R (5 -TGA  TCC TTC TGC AGG TTC ACC TAC-3 ) were used for SSU rDNA amplification (Medlin et al., 1988). To minimize the possibility of PCR amplification errors, Q5 Hot Start High-Fidelity DNA Polymerase (New England BioLabs, United States) was used. The PCR amplification protocol followed was as follows: 30 s initial denaturation (98 • C), 35 cycles of 10 s at 98 • C, 30 s at 69 • C, 1 min at 72 • C, and a final 2 min at 72 • C. Sequencing was performed bidirectionally by the Tsingke Biological Technology Company (Qingdao, China).

Phylogenetic Analyses
Besides the two newly obtained sequences, further 109 sequences used in the present phylogenetic analyses were downloaded from GenBank. Novistrombidium orientale (FJ422988) and Strombidium apolatum (DQ662848) were selected as outgroups. Sequence alignments were performed online using the MUSCLE software package on the European Bioinformatics Institute web server. 1 The resulting alignment was checked and edited manually, leading to a matrix with 2,224 positions.
Maximum likelihood (ML) analysis with 1,000 bootstrap replicates was performed to estimate the reliability of internal branches using RAxML-HPC2 on XSEDE v.8.2.10, with the GTR + I + G model provided on the online server CIPRES Science Gateway (Stamatakis, 2014). Bayesian inference (BI) analysis was run using MrBayes v.3.2.6 on XSEDE on the CIPRES Science Gateway with the GTR + I + G model (selected using MrModeltest v.2.2) (Nylander, 2004). Markov Chain Monte Carlo (MCMC) simulations were then run with two sets of four chains using the default settings. MCMC simulations were then run with two sets of four chains using the default settings. The chain length for the analysis was 10,000,000 generations with trees sampled every 100 generations. The first 10% of trees were discarded as burn-in. The remaining trees were used to calculate posterior probabilities by a majority rule consensus. MEGA v.6.0 (Kumar et al., 2016) was used to visualize tree topology. Systematic and terminology classification follow Curds (1975) and Shao et al. (2015).

Type Materials
Two protargol-stained slides including the holotype (Figures 1B,C Table 1). Cilia of apical membranelles up to 14 µm long in vivo. Undulating membranes in Stylonychia pattern, paroral membrane and endoral membrane straight and parallel to each other, with almost equal length; the endoral ahead of the paroral ( Figures 1B, 2H,I).
The cirral pattern constant, and the number of cirri almost invariable. Three frontal cirri slightly enlarged, the right one behind the distal end of the adoral zone. Buccal cirrus right of the anterior end of endoral membrane. Four frontoventral cirri arranged in a V-shaped pattern, cirrus III/2 at the same level of cirrus VI/3. Four transverse cirri J-shaped. Three postoral ventral cirri closely positioned, near the proximal end of the adoral zone. Two pretransverse cirri, the left one ahead of cirrus III/1, the right one just anterior to cirrus VI/1. One marginal row on each side of the cell, non-confluent posteriorly, almost equal in the number of cirri, the left row commencing at the level of left of the proximal portion of the buccal field and extending to the rear end of the cell, right one starting at right of cirrus IV/3 (Figures 1B, 2H). Transverse cirri about 14-16 µm long in vivo, all of them protruding beyond the rear end of the cell (Figures 1A, 2H). Invariably, four dorsal kineties including one dorsomarginal row (dorsal kinety 4) with conspicuously long dorsal cilia, approximately 4.5-5.0 µm long in vivo. Dorsal kineties 1-3 almost throughout the entire body length, composed of on average 9, 10, and 10 dikinetids, respectively. Dorsal kinety 4 with about 7 dikinetids, commencing at the anterior end of the cell and terminating at mid-body ( Figure 2C). Two caudal cirri, about 10-12 µm in vivo, one each at the posterior end of dorsal kineties 1 and 2 (Figures 1C, 2J).

Note on Morphogenesis
In the opisthe, oral primordium lengthens and differentiates into new membranelles in a posteriad direction (Figures 1F,H); at the same time, undulating membranes (UM) anlage (I) is formed as a streak at the right of the oral primordium, which will split into the paroral and endoral membranes and produce the left frontal cirrus at its anterior end later ( Figure 1H). In the proter, the parental adoral zone of membranelles remains entirely unchanged; however, the parental undulating membranes dedifferentiate from their anterior portion and gradually produce UM-anlage (Figures 1F, 2I), developing in the same way as in the opisthe. In a middle divider, five streaks of the frontal-ventral-transverse cirral anlagen (FVTA II-VI) form to the right of the UMA in both proter and opisthe, separately (Figures 1F, 2K,L); however, it remains obscure if these anlagens are derived from primary primordia or not. At this time, cirri III/2, IV/2, IV/3, V/3, and V/4 disappear. Later, these streaks fragment and form 16 new cirri in a 2: 3: 3: 4: 4 pattern from left to right (Figures 1H,J, 2M,N).
Both marginal row anlagens develop within the parental rows in both opisthe and proter, and later than the origin of FVTA ( Figure 1H). Then, all marginal cirral anlagens enlarge and generate new cirri while the parental structures gradually disintegrate and disappear ( Figure 1J). In the same way, two sets of dorsal kinety anlagens (DKA) 1-3 occur intrakinetally (Figures 1D,G-I). The DKA 4 is formed to the right of the right marginal row and then develops into the short dorsomarginal kinety ( Figure 1J). One caudal cirrus is formed at the posterior end of the new DKA 1 and 2, respectively ( Figure 1L). While the DKA 1-4 proliferate and extend at both ends, the parental bristles gradually become resorbed.
One replication band was seen in each macronuclear nodule (Figure 1G), then two nodules completely fuse into a single mass, which was then divided two times for both proter and opisthe in the late stage (Figures 1I,K). Valbonesi and Luporini, 1990Euplotia Jankowski, 1979Euplotida Small and Lynn, 1985Euplotidae Ehrenberg, 1838 Euplotes Ehrenberg, 1830 Euplotes euryhalinus Valbonesi and Luporini, 1990

Description
Cell size about 45-65 µm × 25-40 µm in vivo. Body asymmetrically oval with an anterior portion slightly wider than the posterior portion, and a distinct projection at the right side (Figures 3A, 4A-D,F). Buccal field approximately 75% of cell length. On the ventral side, three long ridges extending posteriorly to the transverse cirri accompanied by short ridges between; on the dorsal side, six conspicuous ridges almost extending the entire cell length (Figures 3A,D, 4A-C,E,G). Dorsal cilia about 1.5 µm long in vivo. Cytoplasm colorless, transparent at the marginal area, but opaque in the central part where numerous crystals (2-4 µm) are grouped. Contractile vacuole (?) located dorsally at the level of the rightmost transverse cirri, about 3-4 µm across; its pulsating not observed ( Figure 4F). Macronucleus consistently C-shaped, with one micronucleus attached to the macronucleus in the middle portion (Figures 3C, 4K). Locomotion typically by moderately fast crawling or incessant jerking on the substrate. Adoral zone composed of 24-36 membranelles (Figures 3A,B, 4H and Table 2). Paroral membrane small, lying under buccal lip (Figures 3B, 4H). Constantly ten frontoventral cirri with cilia about 15 µm long; five transverse cirri, cilia of which are about 20-22 µm long; two caudal cirri caudally positioned and two marginal cirri at the left posterior margin, cilia 11-15 µm in length (Figures 3A,B, 4F,H). A total of 10 or 11 dorsal kineties composed of dikinetids each; the leftmost one with 2-4 dikinetids, located ventrally near the left margin; both the middle and rightmost one with on average 11 dikinetids each (Figures 3B,C, 4I and Table 2). The dorsal silverline system double-eurystomus pattern (Figures 3E, 4J).  (Valbonesi and Luporini, 1990 All data are based on protargol-stained specimens. Measurements in µm. CV, coefficient of variation in%; DK, dorsal kinety; M, median; Max, maximum; Mean, arithmetic mean; Min, minimum; n, number of cells measured; SD, standard deviation.

Note on Morphogenesis
Only two stages were observed. In an early middle divider, two groups of frontal-ventral-transverse cirral anlagens, each composed of four streaks of basal bodies, appear to the right of the buccal area and between frontoventral cirri and transverse cirri ( Figure 4M); meanwhile, the opisthe's oral primordium is formed in a subpellicular pouch posterior to the proximal membranelles ( Figure 4L). In a middle divider, the oral primordium develops into a new adoral zone of membranelles and paroral membrane in the opisthe; five cirral anlagens (II-VI) in either proter or opisthe give rise to cirri in a 3: 3: 3: 3: 2 pattern from left to right; two left marginal cirri and one frontoventral cirrus are formed de novo; new anlagens for dorsal kineties occur intrakinetally at two levels ( Figures 3F,G, 4N,O). The parental adoral zone of membranelles remains unchanged at these two stages.

18S rRNA Gene Sequence and Phylogeny
The two new SSU rRNA gene sequences obtained in this study were deposited in GenBank. Their lengths, GC contents, and accession numbers are as follows: Hemiurosomoida tibetensis sp. nov., 1,730 bp, 45.55%, MZ457562; Euplotes euryhalinus, 1,825 bp, 43.71%, MZ457561. The ML and BI analyses of the SSU rRNA gene sequences generated phylogenetic trees with similar topologies. Therefore, only the ML tree is shown here with support values from both algorithms (Figure 5). The new species groups with six populations of Hemiurosomaida longa with strong support (ML 91%, BI 1.00), forming a sister clade to the clade formed by Parakahliella macrostoma (MH393767) and Hemiurosomaida warreni (MT448250) with weak support in ML analysis (60%) and high support in BI analysis (0.99). Hemiurosomoida tibetensis sp. nov. differs from Hemiurosomoida longa and Hemiurosomoida warreni by 46-89 and 42 unmatched nucleotides (94.6-97.2 and 97.4% in sequence identity), respectively. and it differs from Parakahliella macrostoma by 46 nucleotides (97.2% in sequence identity) (Figure 6). Furthermore, the monophyly of the genus Hemiurosomaida is not rejected by the AU test (loglikelihood: -21292.517698, AU value: 0.253). The Tibet population of Euplotes euryhalinus clusters into a clade including four other populations of this species (MW369533, EF094967, KP297368, JF903799, JF903800), Eulpotes indica (MN038061) and Euplotes magnicirratus (AJ549209) with strong to full support, being sister to the Euplotes cf. antarcticus (FJ998023) + Euplotes trisulcatus (EF690810) clade with low to nearly full support (ML 68%, BI 0.99).

Comparison of Hemiurosomoida tibetensis sp. nov. With Related Species
Hemiurosomoida Singh and Kamra, 2015, was a recently erected genus with reduced transverse cirri, the lack of dorsal kinety 3 fragmentation, and two caudal cirri at the posterior end of dorsal kineties 1 and 2. The newly isolated form matches diagnosed characteristics of the genus but cannot be assigned to any one of two known congeners, namely, H. longa and H. warreni , so we named it H. tibetensis sp. nov.
Hemiurosomoida longa was first described as Oxytricha longa Gelei and Szabados, 1950, and then redefined by Singh and Kamra (2015) based on a few descriptions by Gelei and Szabados (1950) and Ganner et al. (1986Ganner et al. ( , 1987 as well as an Indian population by themselves. It is very similar to Hemiurosomoida tibetensis sp. nov., especially in body size and most aspects of ciliature; however, it differs from the latter mainly by the lack of cortical granules (vs. apparently present). The SSU rRNA gene sequence of Hemiurosomoida longa isolate (JX893371) with morphological information supplied differs from that of Hemiurosomoida tibetensis sp. nov. by 46 nucleotides, which also support the separation of both forms at the species level (Figure 6). The other five populations of H. longa (MG994998, KY852451, KY912033, AF164125, and AF508763) lack morphological data; of them, the Indian population differs from the remaining populations by 10-51 nucleotides. Compared with other oxytrichs, H. longa indeed presents relatively high SSU rDNA sequence divergence among populations (Li et al., 2018;Luo et al., 2021;Wang et al., 2021), which may infer that cryptic species exist.

Identification of Tibet Population of Euplotes euryhalinus
Euplotes euryhalinus was first reported based on an Antarctic population by Valbonesi and Luporini (1990). Our isolate corresponds well with the original description by the oval body shape, C-shaped macronucleus with one micronucleus situated anteriorly, basic ciliature (constantly ten frontoventral, and five transverse cirri as well as 11 dorsal kineties), and the dorsal silverline system of the "double" type. In the original description, Valbonesi and Luporini (1990) mentioned the presence of four (more occasionally five) caudal cirri, some of which are marginal cirri according to their position. One different characteristic that should be noticed is that there are six prominent ridges on the ventral surface and inconspicuous ridges on dorsal sides in the original population (vs. six conspicuous ridges visible on dorsal sides in our population). According to our observation, they are inconspicuous in bright field microscopy but distinct in differential interference contrast microscopy, and Valbonesi and Luporini (1990) probably overlooked them. Thus, we consider the two populations to be conspecific. Berger (2006) established the Dorsomarginalia for hypotrichs with dorsomarginal kinety. As a representative, Oxytrichidae is often divided into two groups based primarily on morphological and morphogenetic characteristics, namely, the "flexible" nonstylonychine Oxytrichidae and the "rigid" Stylonychinae (Berger and Foissner, 1997;Berger, 1999). Bernhard et al. (2001) provided molecular evidence for the monophyly of Stylonychinae, which is not supported by further analyses (Foissner et al., 2004;Schmidt et al., 2007) and this study. The grouping of Hemiurosomoida tibetensis sp. nov. with its congeners corroborates the generic assignment of the species. The clustering of Parakahliella macrostoma into Hemiurosomoida clade suggests that the Hemiurosomoida is non-monophyletic . The first grouping of P. macrostoma with H. warreni corresponds to some morphological similarities (e.g., dorsal ciliary pattern and multiple macronuclear nodules) (Ning et al., 2018;Chen et al., 2021).

Phylogenetic Analyses
The present isolate of Euplotes euryhalinus first forms a clade with three populations of this species (EF094967, KP297368, MW369533), all sampled from same Antarctica as the original population and differing by 0-2 nucleotides from our population, which confirms our morphological identification. However, the Indian population with two sequences (JF903799, JF903800) shows (28, 44 different nucleotides, respectively) from our Chinese population, such great genetic divergence in SSU rRNA gene, which may infer that their species entities need reconfirmation as morphological data are unavailable for it. Euplotes indica and E. magnicirratus are within the clade of Euplotes euryhalinus, and this phylogenetic grouping is in accordance with some morphological similarities among these isolates, such as (1) C-shaped macronucleus and one spherical micronucleus; (2) ten frontoventral, five transverse, two left marginal, and two caudal cirri; (3) dorsal silverline system of double-eurystomus type, but Euplotes euryhalinus morphologically differs from E. indica by its larger ratio of buccal field length to body length (75 vs. 68%), more dorsal kineties (10 or 11 vs. 7), and different habitat (saline water vs. freshwater), from Euplotes magnicirratus by more dorsal kineties (10 or 11 vs. 8) and less dorsal ridges (6 vs. 8) (Gelei and Szabados, 1950;Carter, 1972;Valbonesi and Luporini, 1990;Abraham et al., 2021).

AUTHOR CONTRIBUTIONS
XH designed the study. FK and WS identified species and drafted the manuscript. LJ and WZ performed the experiments and constructed molecular trees. CW and TW analyzed data and made figures. All authors contributed to the proofreading of the manuscript and approved the final version.

FUNDING
This study was supported by the National Natural Science Foundation of China (project number: 41976086) and the China Scholarship Council, both awarded to XH.

ACKNOWLEDGMENTS
We thank Weibo Song, the Ocean University of China, for his long-term support and concern for the FK.