Refining the “ Melting Pot ” Genus Holosticha s. l. (Protozoa, Ciliophora, Hypotrichia) Based on Multigene Datasets with Establishment of a New Species Caudikeronopsis monilata sp. nov.

The genus Holosticha s. l. is a typical “ melting pot ” group with an intricate history, and so far, it has been divided into eleven genera. Both newly obtained taxonomic and molecular data provide the opportunity to gain more insights to outline the taxa in it and to understand their systematic and evolutionary relationship. Here, we describe Caudikeronopsis monilata sp. nov. from intertidal sediment on the China coast of the Yellow Sea and analyze the phylogenetic relationships of Holosticha s. l. by obtaining a total of 16 new sequences of seven isolates. The results demonstrate that (1) the morphological features of Holosticha s. str. are outlined very well, but its systematic relationship with Uncinata is still puzzling; (2) based on both morphological and molecular databases, the genera Adumbratosticha , Arcuseries , Caudikeronopsis , Extraholosticha , and “ Holosticha + Uncinata ” complex are separated clearly from each other in the phylogenetic analyses; and (3) the Anteholosticha isolates are dispersed among the urostylids in the phylogenetic analyses, even though its generic diagnostic features are described very clearly. In the present work, however, the secondary structure predictions do not provide better resolutions for understanding the systematic and evolutionary relationships among the holostichids. And the genus Anteholosticha becomes a new “ melting pot ” taxon.


Introduction
As common members of the periphyton, benthos and psammon in marine, brackish, and freshwater habitats, the speciesrich hypotrichous ciliates are much more diverse than previously thought, and their supraspecific categories and systematic relationship have been drawn more and more attention. Holosticha Wrzesniowski, 1877, has ever been a large and puzzling genus, in which over 100 species were originally assigned to it, and many others were transferred to and excluded from it [1][2][3][4][5][6][7]. With the continuous discovery of new taxa, the phylogenetic analyses based on multigene data provide the opportunity to break the Holosticha s. l. "melting pot," to gain more insights to outline to the taxa in it and to better understand their systematic and evolutionary relationship [8].
Caudikeronopsis was a monotypic genus established by Li et al. [5] with C. marina [9] as the type. In this study, we describe a new species of Caudikeronopsis from intertidal sediment on the China coast of the Yellow Sea. Based on reliable identification, we also newly obtain 16 sequences (five of SSU rDNA, six of ITS1-5.8S-ITS2, and five of LSU rDNA) of seven holostichid isolates including Caudikeronopsis monilata sp. nov. In order to infer the evolutionary relationships within the confusing holostichids, the phylogenetic analyses were conducted using these multiple markers and new sequences, and the secondary structure predictions of SSU rRNA and ITS regions were also investigated. This paper attempts to improve our understanding of the diversity and phylogeny of hypotrichous ciliates through historical revision of the representative "melting pot" genus Holosticha.

Material and Methods
2.1. Ciliate Collection and Identification. Seven isolates of holostichids were collected from the coastal areas in Qingdao, China (Table 1), including the Caudikeronopsis monilata sp. nov., which was isolated from intertidal sediment in Qingdao Bay (36°03′39″N, 120°19′31″E), Qingdao, China, with salinity 30 psu and water temperature about 14°C. Living cells were maintained in the Petri dishes with habitat water at ambient temperature (ca. 20°C), and examined in vivo using a bright field and differential interference microscope at magnifications of 100-1000 ×. The infraciliature and nuclear structure were revealed using protargol impregnation according to Wilbert [10]. The species identification was achieved through both living observation and protargol staining based on raw culture.

Phylogenetic Analyses.
Using CLUSTAL W implemented in BioEdit 7.0 [15], the pairwise alignments of the sequences are conducted. The sites which are gaps in more than half of the sequences were excluded resulting in matrixes of 1785, 588, 1859, and 4158 characters for SSU rDNA, ITS1-5.8S-ITS2, LSU rDNA, and concatenated rDNA analyses, respectively. The pairwise distances between species/populations were calculated with MEGA 6.06 [16]. Modeltest 3.4 [17] and MrModeltest version 2.0 [18] were used for model selection of the maximum-likelihood (ML) and Bayesian inference (BI) analysis, respectively. According to the Akaike information criterion (AIC), the GTR+I+G model was selected as the optimal choice for SSU rDNA, LSU rDNA, and concatenated rDNA, and the GTR+G model was selected as the optimal choice for ITS1-5.8S-ITS2. The ML analyses were performed using RAxML-HPC2 on XSEDE 8.2.12 [19,20] on the CIPRES Science Gateway (http:// www.phylo.org/sub_sections/portal; [21]). With MrBayes 3.2.2 [22] via the CIPRES Science Gateway, Bayesian inference analyses were also conducted. The TreeView v1.6.6 [23] and MEGA 4.0 [24] were used to visualize tree topologies. The classification shown in the phylogenetic analyses is according to Berger [3], Huang et al. [4], and Luo et al. [6].

Topology
Testing. PAUP was used to build eight unrooted trees with enforced topological constraints. The site-wise likelihoods for the best-unconstrained ML tree and all constrained trees were calculated in PAUP under the GTR+I+G model (for SSU rDNA, LSU rDNA, and concatenated rDNA) and GTR +G model (for ITS1-5.8S-ITS2) with parameters as suggested by Modeltest 3.4 [17]. The statistical probability of the constrained trees was evaluated in the likelihood frameworks through the approximately unbiased (AU) test [25], as implemented in the CONSEL software package [26].
2.5. Secondary Structure Predictions. The secondary structures of the variable regions 4 and 9 of SSU rRNA and ITS regions were computed by submission of primary sequences to the RNA folding website supporting MFOLD version 3.6 (http://unafold.rna.albany.edu/?q=mfold/RNA-Folding-Form; [27]), using the default parameters for folding except T = 25°C. The secondary structures of variable regions 4 and 9 of SSU rDNA were predicted based on the models of the eukaryotic SSU rRNA presented by Wuyts et al. [28] and Wang et al. [29]. The sequences of ITS regions were constrained according to the rules suggested by Yi et al. [13]. The structures were edited with RNA VIZ 2.0 [30] to produce acceptable illustrations.     Most cirri with cilia are about 11 μm in length except for transverse cirri with cilia about 20 μm long. One buccal cirrus locating at the right of the paroral membrane and single parabuccal cirrus is behind the rightmost frontal cirrus. Consistently two frontoterminal cirri located left of anterior right marginal cirral row. Invariably three enlarged frontal cirri connecting to the midventral complex composed of 11-16 cirral pairs. Four to eight transverse cirri adjacent to the posterior part of the midventral complex, with two pretransverse ventral cirri, one left and one right marginal cirral row with 30-58 and 34-54 cirri, respectively. Two to four caudal cirri and four to eight dorsal kineties with cilia about 7 μm long, conspicuous in vivo ( Figure 1, J (arrows)).
The results showed that seven to 26 Anteholosticha isolates disperse in the urostylids. By contrast, the different isolates in both Arcuseries and Caudikeronopsis assemble very well with high support values, respectively (Figures 2, 4, and 5). In all the topologies, Anteholosticha pulchra nests, in the pseudokeronopsids, and Anteholosticha gracilis locate with Monocoronella carnea steadily.
Uncinata isolates keep inside the Holosticha isolates, and they cluster together with high support values in all the topologies (Figures 2, 4, and 5), except for the phylogenetic trees based on ITS1-5.8S-ITS2 ( Figure 3).

Topology
Testing. The monophyly of the genus Holosticha is not rejected based on the ITS1-5.8S-ITS2 and LSU rDNA dataset but is rejected based on the SSU rDNA and rDNA datasets by the AU test. The statistical tests carried out on single gene do not reject the possibility of Uncinata and Holosticha forming a monophyletic clade.  Figure 6). The partial putative secondary structures (helix E23-1/2 and helix E23-4/7) of the V4 regions of 58 holostichid isolates are investigated ( Figure 6(a)). The structures of the V4 region are not conserved, that is, the number and position of bulges in helices E23-1/2 and E23-4/7 are not stable. The isolates in the gen-era Arcuseries and Holosticha have their own special motifs, respectively. There are many GC pairs in both the base part of helix E23-1/2 and around the middle bulge in helix E23-4/ 7 in the Arcuseries isolates. And some Anteholosticha isolates, the genus Caudikeronopsis, and Extraholosticha sylvatica isolates also have five to eight GC pairs in the base part of helix E23-1/2 ( Figure 6(a), yellow markers). The   Figure 6(a), blue markers). The partial putative secondary structures of the V9 regions of 44 holostichid isolates are investigated (Figure 6(b)). The structures of the V9 regions are much more conserved than those of the V4 region, that is, the number and position of bulges are stable. The motifs 5 ′ -GAGUG versus CACUU-3 ′ and 5′-UU versus AA-3′ (Figure 6(b), grey markers) appear in all the 44 holostichid and many other urostylid. Uncinata bradburyae isolates have one UA-pair and one UG-pair astride the motif 5 ′ -GAGUG versus CACUU-3 ′ , beside which Holosticha isolates have one UA-pair or one UG-pair  Figure 6(b), purple markers). Eight Arcuseries isolates have the motif 5′-GCGC versus GCGC-3′, which also appears in Anteholosticha multistylata and Holosticha heterofoissneri ( Figure 6(b), yellow markers). The motifs 5′-CC versus GG-3′ and 5′-GG versus CC-3′ exist together in the three Extraholosticha sylvatica isolates and Anteholosticha monilata and present alone in several Anteholosticha, Holosticha, and Uncinata bradburyae isolates (Figure 6(b), green markers).
3.6. Putative Secondary Structures of ITS1 and ITS2 Transcripts (Figure 7). The putative structures of the ITS1 and ITS2 transcripts of the 27 and 28 holostichid isolates are predicted, and the general models based on these structures are proposed, respectively (Figure 7). The ITS1 has three helices, which exhibit distinct size classes. As the most conserved one, helix I comprise only UA pairs in most holostichid isolates. As special cases, helix I in Caudikeronopsis isolates is the motifs 5′-CUC versus GAG-3′ (Figure 7(a), green markers); and there is an unpaired CC (Figure 7(a), purple markers) among the UA pairs in helix I of Uncinata bradburyae isolates. Helix II is special and variable, which can separate the genera Arcuseries, Caudikeronopsis, Holosticha, and Uncinata clearly from other holostichid isolates. Helix II in Arcuseries isolates is 48 nt and has fifteen GC pairs (Figure 7(a), yellow markers). Caudikeronopsis isolates have the motifs 5 ′ -GUGC versus GCGC-3 ′ and 5 ′ -GC versus GC-3 ′ in the middle and at the terminal end of Helix II, respectively. These two motifs are also present in some other holostichid isolates together or partly (Figure 7(a), green markers). Helix II in Holosticha and Uncinata isolates both have the motifs 5 ′ -GCG versus CGC-3 ′ (Figure 7(a), purple markers) and 5 ′ -GUGC versus  Journal of Zoological Systematics and Evolutionary Research GCGC-3 ′ at the base and in the middle of helix II, respectively. Helix III can separate Holosticha and Uncinata isolates from other holostichid isolates, with the motifs 5 ′ -UAAA versus UUUA-3 ′ and 5 ′ -CU versus AG-3 ′ at the base and terminal end (Figure 7(a), blue markers), which only two Anteholosticha species have. There are two helices in ITS2, that is, helix A 10-16 nt long and helix B 72-112 nt long. In helix A, all the holostichid isolates have the motif 5 ′ -GA/GGA versus UCUC-3 ′ near the terminal loop. Except for Caudikeronopsis, all the other holostichid isolates have three branches in the helix B, in which the right one has the same motif 5 ′ -CA/GG versus CUG-3 ′ (Figure 7(b), orange markers). All the holostichid isolates are very similar to each other in the middle branch (Figure 7(b), grey markers). Caudikeronopsis isolates have the motifs 5 ′ -GCCUCUGC versus GU/CAGAGGU/C and 5 ′ -GCGGG versus CCCGC-3 ′ in helix B (Figure 7(b), green markers), and the first one also occurs in Adumbratosticha tetracirrata, some Holosticha, and Uncinata bradburyae isolates. Holosticha isolates have the motif 5 ′ -AAG versus CUU-3 ′ at the base of helix B, which some other holostichid isolates also have (Figure 7(b), blue markers).

Historical Revision of the Genus Holosticha
Since molecular phylogeny largely focused on limited sampling, the incongruence between morphological data and gene sequences has been arising. By collecting the three-gene data (SSU rDNA, ITS1-5.8S-ITS2, and LSU rDNA) comprising 30 species in the "core urostyloids," Huang et al. [4] established Arcuseries containing three distinctly deviating Anteholosticha species, with the character of roughly U-shaped arranged transverse cirri (Figure 8, yellow arrow).
Combining the morphological, morphogenetic, and SSU rDNA-based phylogenetic analyses, Luo et al. [6] transferred Holosticha bradburyae Gong et al., 2001 to Uncinata, due to its possession of a characteristically prominent beak-like, leftwards curved projection and the developmental mode of the dorsal kineties (Figure 8, purple arrows).
However, due to a lack of detailed information on the ciliary pattern, morphogenetic process and molecular data, the taxonomy and systematics of Limnoholosticha viridis (Kahl, 1932) [5] were poorly known. Based on a Chinese population of this uncertain species, Song et al. [8]

Is Holosticha Outlined Well?
After over decades' modification (Figure 2), the genus Holosticha s. str. is morphologically defined by a combination of features, that is, adoral zone of membranelles bipartite and rearmost ones wider, buccal cirrus distinctly ahead of paroral, number of transverse cirri nearly number of midventral pairs, anterior end of left marginal row composed of narrowly spaced cirri and distinctly curved rightwards, caudal cirri lacking [2]. Luo et al. [6] transferred Holosticha bradburyae (Gong et al. 2001) to Uncinata, due to its possession of a  Journal of Zoological Systematics and Evolutionary Research characteristically prominent beak-like, leftwards curved projection and the developmental mode of the dorsal kineties. And the phylogenetic analyses revealed Holosticha to be monophyletic and sister to Uncinata [6]. However, the phylogenetic analyses by Zhang et al. [7] indicated that Uncinata is a subgenus of Holosticha. And in the present work, by adding new sequences of Holosticha sp. (MT476978 for SSU rDNA, MT476975 for ITS1-5.8S-ITS2 regions, and MT476979 for LSU rDNA), not only Uncinata isolates always nest in Holosticha isolates in the phylogenetic tree but also the monophyly of the genus Holosticha is not supported well. The genera Holosticha and Uncinata have a very close relationship, and their putative secondary structures of V4 and V9 in SSU rRNA, ITS1 and ITS2 share similar motifs with each other. Even though the possibility of Uncinata and Holosticha forming a   Journal of Zoological Systematics and Evolutionary Research monophyletic clade is not all rejected, we cannot confirm that Uncinata isolates belong to the genus Holosticha.
Up to now, the morphological features of Holosticha and Uncinata are both outlined very well, but their systematic relationship with Uncinata is still puzzling. At least for a while, it is better to regard them as two separated genera.

Is Anteholosticha a New "Melting
Pot" Genus? Because of its intricate history, Holosticha had been a typical "melting pot" genus for many years until recently, most holostichid isolates divide into eleven genera (Figure 8). Based on both morphological and molecular databases, the genera Adumbratosticha, Arcuseries, Caudikeronopsis, Extraholosticha,  Helix II   II   II   I   I  I  I  I   ITS1  5ʹ   3ʹ   I   I   I   I  I  I  I  I   I  I   I  I   II  II  II  II   II   II   II  II  II  II   II   II  II   II  II   III   III   III  III  III  III   III  III   III  III  III  III  III  III  III  III   10 Journal of Zoological Systematics and Evolutionary Research and "Holosticha+Uncinata" complex are outlined quite well, and they are separated clearly from each other in the phylogenetic analyses. In contrast, the genus Anteholosticha isolates disperse among the urostylids in the phylogenetic analyses, even though its morphologically diagnostic features are described very clearly. To a certain extent, Anteholosticha is now a new "melting pot" genus.

Do Secondary Structure Predictions Always Provide
Better Resolutions for Phylogenetic Analyses? By predicting the secondary structures of four variable regions (V2, V4, V7, and V9) in the SSU rRNA of 45 urostylids, Wang et al. [29] considered the V4 region as the most effective in reveal-ing interspecific relationships, while V9 appeared suitable at the family level or higher; V2 was too conserved to reflect phylogenetic relationships at the family or lower lever, and V7 was the least informative. Hence, in the present work, we focus on the relatively conserved E23-1/2 and E23-4/7 parts of the V4 region and V9 region. Both the shapes and conserved motifs of the V4 and V9 of SSU rRNA correspond well with the phylogenetic trees' topology with a visual manner. With relatively high divergence, the secondary structure predictions of the internal transcribed spacer (ITS) regions are widely used for phylogenetic reconstructions below the family level [14,[32][33][34][35]. Among the holostichid isolates, the general secondary structure models are relatively stable,

11
Journal of Zoological Systematics and Evolutionary Research and there are some compensatory base changes (CBCs) to maintain their reliability. Although less stable than the Watson-Crick complementarities, the GU appositions often occur in their ITS2 regions.
In the present work, however, the secondary structure predictions do not provide better resolutions for understanding the systematic and evolutionary relationships among the holostichids, as did in previous studies [14,32]. The genus Anteholosticha becomes a new "melting pot," which is still puzzling.

Data Availability
The holotype slide (registration number: LJ14021502-1) and two paratype slides (LJ14021502-2 and LJ14021502-3) with protargol-impregnated specimens have been deposited in the Marine Biological Museum, Chinese Academy of Sciences, Qingdao. The holotype and paratype specimens are marked with black ink circles on the cover glass. The new sequences of SSU rDNA, ITS1-5.8S-ITS2, and LSU rDNA that support the findings of this study have been deposited in NCBI GenBank ( Table 1). The registration of Caudikeronopsis monilata sp. nov. and the publication of the present work in ZooBank are as follows: urn:lsid:zoobank.org:act:F1E4D812-379D-4F96-AA1E-F2AD7BC50CF1 and LSID:urn:lsid:zoobank.org:pub:78F09A0B-0BB0-457A-965D-214DDAEE1655, respectively.

Conflicts of Interest
The authors declare that they have no conflicts of interest.