Documentation of a new hypotrich species in the family Amphisiellidae, Lamtostyla gui n. sp. (Protista, Ciliophora) using a multidisciplinary approach

An integrated approach considering both morphologic and molecular data is now required to improve biodiversity estimations and provide more robust systematics interpretations in hypotrichs, a highly differentiated group of ciliates. In present study, we document a new hypotrich species, Lamtostyla gui n. sp., collected from Chongming wetland, Shanghai, China, based on investigations using living observation, protargol staining, scanning and transmission electron microscopy, and gene sequencing. The new species is mainly recognized by having a short amphisiellid median cirral row composed of four cirri, three frontoventral cirri, three dorsal kinetids, four to eight macronuclear nodules, and small colorless cortical granules distributed as rosettes around dorsal bristles. Transmission electron microscope observation finds the associated microtubules of cirri and pharyngeal discs of L. gui are distinct from those in other hypotrichs. Morphogenesis of this species indicates that parental adoral membranelles retained intact or partial renewed is a potential feature to separate Lamtostyla granulifera-group and Lamtostyla lamottei-group. Phylogenetic analysis based on small subunit ribosomal RNA (rRNA) gene shows that this molecular marker is not useful to resolve phylogenetic relationships of the genus Lamtostyla, as well as many other hypotrichous taxa. We additionally characterize the internal transcribed spacers (ITS) region and the almost complete large subunit rRNA, which will be essential for future studies aimed at solving phylogenetic problems of Lamtostyla, or even the family Amphisiellidae. As a final remark, the critical screening of GenBank using ITS genes of our organism allows us to recognize a large amount of hypotrichous sequences have been misclassified as fungi. This observation suggests that hypotrichs could be frequently found in fungi-rich environment and overlooked by fungal specialists.

transverse cirri (TC), that is the common "1 + 4" pattern ( (Table 1; Figs. 1a,b, 2a,h,i, 3a,g,h). Right marginal row commencing at level of rightmost frontal cirrus or anteriormost cirrus of ACR, and terminating posterior to transverse cirri; left marginal row starting posterior to adoral zone and extending along margin of posterior body end, so that marginal rows distinctly separated posteriorly (Figs. 1a,b, 2a,h,i, 3a). In addition, two rows of small protrusions extending along the right sides of marginal rows shown on SEM (Fig. 3a,g,h). Three bipolar dorsal kineties with about 15, 17, 18 dikinetids in each rows; bristles located on anterior kinetosome of each dikinetid and about 3 μm long in vivo (  [4][5][6] In the opisthe, stomatogenesis commences with the formation of an oral primordium which originates from the dedifferentiation of the leftmost one or two transverse cirri (Figs. 4a-c, 5a). The oral primordium enlarges into a long field with broadened anterior end and then the new adoral membranelles begin to organize at the anterior end of the primordium (Figs. 4b-d, 5b-d). At the same time, the undulating membranes anlagen (UMA), i.e., the frontoventral transverse cirri (FVT) anlage I, is formed to the right of the oral primordium (Figs. 4d, 5d). Later, the leftmost frontal cirrus is generated from the anterior end of the UMA (Figs. 4d, 5d). In the later stage, the anterior end of the newly built adoral zone bends to the right and the differentiation of membranelles is almost completed; meanwhile, the UMA splits longitudinally into two streaks which give rise to the paroral and endoral membrane, respectively (Figs. 4f,h, 5e-g,i).   Table 1. Morphometric characteristics of Lamtostyla gui n. sp. All data measurements in micrometers. a Data based on living cells, b data based on protargol-stained specimens, c data based on protargol-stained specimens and scanning electron microscopy micrographs. ACR amphisiellid median cirral row, CV coefficient of variation in percentage, M median, Max maximum, Mean arithmetic mean, Min minimum, n number of individuals measured, No. number, SD standard deviation.
In the proter, parental adoral membranelles are mostly retained, with only those in the posterior end renewed in situ (Figs. 4c,d, 5c-f). The UMA is formed from the dedifferentiation of the parental undulating membranes (Figs. 4b-d, 5b,c). In subsequent stages, the development of the UMA follows a similar pattern to that in the opisthe (Figs. 4f,h, 5d-i).
Development of the frontoventral transverse cirri. In the early stage, buccal cirrus (II/2), three frontoventral cirri (cirrus III/2, IV/2, IV/3) and the posterior two cirri of ACR (cirrus V/3, 4) dedifferentiate into small groups of kinetosomes when the oral primordium of the opisthe begins to elongate (Fig. 4b). Later, these long anlagen connect with the anterior end of the oral primordium, forming five thread-like primordia with anlage I of both proter and opisthe disconnected (Figs. 4c, 5b). Then, the primordia separate very soon, usually two sets of six FVT anlagen (I-VI) of the proter and opisthe are formed (Figs. 4d, 5c). Subsequently, these cirral anlagen develop independently in both dividing parts. Each anlage begins to segment into new cirri (Figs. 4f,h, 5d-j,m): as is usual, anlage I generates left frontal cirrus (cirrus I/1) and undulating membranes; anlage II produces buccal cirrus (cirrus II/2) and middle frontal cirrus (cirrus II/3); anlage III forms one transverse cirrus (cirrus III/1), one frontoventral cirrus (cirrus III/2), and rightmost frontal cirrus (cirrus III/3); anlage IV forms one transverse cirrus (cirrus IV/1) and two frontoventral cirri (cirrus IV/2, 3); anlage V produces one transverse cirrus (cirrus V/1) and the posterior two cirri of ACR (cirrus V/3, 4); anlage VI forms one transverse cirrus (cirrus VI/1), one pretransverse cirrus (cirrus VI/2) and the anterior two cirri of ACR (cirrus VI/3, 4) (Tables 2, 3, Fig. 4f,h). Occasionally, the anlagen III-VI form more or less cirri compared with most common pattern (Table 2), from which we can deduce reasonable explanations for some variations on the number of cirri in frontoventral and transverse regions in interphasic specimens (Table 3, Fig. 6a-h2). Besides,one additional anlage is sometimes present between IV and V (Fig. 6d,h2) which could contribute one more TC and one (Fig. 6h2) or two (Fig. 6d) more frontoventral cirri causing cells with five TC and four or five frontoventral cirri. As usual, the parental three frontal cirri and cirrus VI/3, 4 (anterior two cirri in ACR) are never involved in the FVT-anlagen formation and remain unchanged even in the very late dividers (Figs. 4f,h, 5g-j,m, 6a-h2), indicating that they may be resorbed after cell division.  The cortical granules (arrowheads) scatter near right marginal cirri on ventral side (f) and distribute as rosettes around dorsal bristles on dorsal side (g). (h,i) Ventral (h) and dorsal (i) views of the holotype specimen, showing the infraciliature and nuclear apparatus; arrow indicates the pretransverse cirrus, arrowheads show the digested scuticociliates and three frontoventral cirri are depicted by circle. (j,k) Frontoventral ciliatures of two individuals, where the individual (j) has five cirri in ACR and three frontoventral cirri (circled); the individual (k) has four cirri in ACR and five frontoventral cirri (circled). (l) Infraciliature of the digested scuticociliate. (m-q) Different transverse and pretransverse cirral patterns: one pretransverse cirrus and three transverse cirri in (m); two pretransverse cirri and three transverse cirri in (n); four transverse cirri in (o); two pretransverse cirri and four transverse cirri in (p); one pretransverse cirrus and five transverse cirri in (q). (r-v) Variations of macronuclear apparatuses: four binodal nodules in (r), three binodal and two ellipsoid nodules in (s), two binodal and four globular nodules in (t), one binodal and six ellipsoid nodules in (u), and eight globular nodules in (v). ACR amphisiellid median cirral row, AZM adoral zone of Development of marginal rows and dorsal kineties. The marginal anlagen are formed at two levels within the parental marginal rows, i.e., near the anterior end, and below the mid-body; they then stretch posteriad and gradually replace the parental rows (Figs. 4d, 5d-g, 6a). Unfortunately, the early and middle stages of dorsal kinety anlagen formation are not observed, and only a very late stage shows that three dorsal kineties are newly formed in both dividers (Fig. 4e). Dorsomarginal rows, dorsal kinety fragmentation and caudal cirri are lacking.
Division of nuclear apparatus. The nuclear apparatus divides in a conventional manner for hypotrichous ciliates 3 , that is, four to eight macronuclear nodules fuse to form a single mass during the mid-divisional stage (Figs. 4e, 5c-f) and subsequently divides into two ellipsoidal nodules (Figs. 4g, 5g,h). Then each nodule divides again before cytokinesis (Figs. 4i, 5i,j). Two ellipsoidal (Fig. 5k) to binodal (Fig. 5l) macronuclear nodules could often be found in the newly separated daughter cells. Hence, after the cytokinesis, the division of macronuclear nodules may still proceed in daughter cells to form four or more macronuclear nodules as shown in the trophic stage. Sometimes, three ellipsoidal macronuclear nodules occur in a later divider (Fig. 5m), which may result from the asynchronization of nuclear division in the proter and opisthe. Micronuclei were observed to divide mitotically (Figs. 4g, 5h). transmission electron microscope observations. (Figs. 7-9) Adoral membranelles were separated by intermembranellar ridges which measured approximately 2 μm at their highest (Fig. 7a). Each membranelle comprised four rows of kinetosomes, with row 4 containing only three kinetosomes in the available sections ( Fig. 7b), which is consistent with SEM observation. The left wall of buccal lip and buccal seal were found in a protrusion between paroral and endoral membranes; and they were either straight or curved at their distal end in different sections ( Fig. 7c-e). The pellicle consisted of plasma membrane and alveoli; the alveoli were relatively flat, and sometimes hard to be noticed ( Fig. 8a,b). A single layer of subpellicular microtubules was beneath the pellicle in most parts of cell (Fig. 8a,b); and sometimes thickened microtubular bundles were also present, especially in the vicinity of cirri, which were possibly the associated microtubules of cirri (Fig. 8b). The associated microtubules of left marginal cirri were observed in details. The anterior microtubular bundles (Amb) and the posterior microtubular bundles (Pmb) connected with the rampart in the two shorter sides of cirri base; the linear microtubular arrays (Lma) strengthened the rampart in the left longer side of cirri base; and the kinetodesmal fiber connected directly with kinetosomes of cirri (Fig. 8c,d). Each kinetosomes consisted of axosome, basal plate and basal granule (Fig. 8g).
The cortical granules were composed of a membrane and the internal inclusion (Figs. 7d,e, 8e,f,h,i, 9a,b,g-i). Most of them were round to oval and measured 0.3-0.6 μm at their widest (Figs. 7d,e, 8e,f,h, 9a,g-i); while they were elongated and teardrop-shaped (up to 1.2 μm long) in some sections and more often present near the buccal area (Figs. 7d,e, 8e,h, 9b). These granules were ordinarily in clusters (several in number) in the vicinity of the kinetosomes of cirri and dorsal bristles (Fig. 8e,f,h) and occasionally in the cytoplasm away from the cortex as well ( Fig. 9g-i). In particular, dozens of the granules assembled in a large vesicle in the front area of the buccal field near the endoral membrane (Figs. 7d, 9a), and occasionally near the paroral membrane (Fig. 7e). The inclusion of cortical granules could be divided into a small anterior part and a large body part; these two parts were more obvious when cortical granules located near the buccal area attaching to the pellicle (Fig. 7d,e). The small anterior part near the pellicle is often of high electron-dense while the main part is always of less electron-dense (Fig. 7d,e).
A lot of aggregations of single membrane bounded flattened saccules were found attached to the pellicle in the buccal area behind the cortical granules assemblage (Fig. 9a,b,e,f); such flattened saccules also existed in the cytoplasm near and away from the buccal area ( Fig. 9d,g,h). Those near the pellicle of the buccal area were usually elongated and parallel to each other ( Fig. 9a,b,e,f), while others in deeper cytoplasm were shorter or curled and sometimes encased in cytoplasmic vesicles with cortical granules and/or mitochondria inside ( Fig. 9d,g,h). Cytoplasm was also rich of widespread mitochondria, cytoplasmic vesicles of varied sized, and food vacuoles (Fig. 9c,i). Endoplasmic reticulum was found occasionally and closed to mitochondria (Fig. 9l). Macronucleus was composed of a discontinuous nuclear envelope, numerous irregular shaped chromatin bodies, and nucleoli (Fig. 9j). The nuclear pores embedded in the envelope could be clearly observed (Fig. 9k). In all reconstructed SSU rRNA phylogenetic trees, several poorly resolved regions could be observed. Using different alignment filters or treeing methods (i.e., maximum likelihood (ML) or bayesian inference (BI)) did not allow us to properly resolve them. In Fig. 10, one of the obtained BI trees constructed by a final 272 character matrix derived from the alignment filter 5%-98%, where columns with a similarity lower than 5% and higher than or equal to 98% had been removed and left only those informative columns, was reported as example. Although, L. gui apparently branched alone in the reported tree, it was indeed belonging to part of a large unresolved region with low posterior probability support. All species or subclades constituted the unresolved region were highlighted in red in Fig. 10. Within this region, three highly supported subclades of organisms emerged membranelle, BC buccal cirrus, DB dorsal bristle, DK dorsal kinety, EM endoral membrane, FC frontal cirrus, LMR left marginal row, Ma macronucleus, Mi micronucleus, PM paroral membrane, PTC pretransverse cirrus, RMC right marginal cirrus, RMR right marginal row, TC transverse cirrus. Scale bars 40 μm (a-e,h,i), 5 μm (l). Although we characterized the almost full ribosomal operon, only the SSU rRNA gene could be used for genus/family level phylogenetic reconstruction due to the lack or extreme limitations of ITS and LSU rRNA gene sequences of related organisms. But, the ITS and LSU rRNA gene sequences of our organism represented the first available data for the genus Lamtostyla and one of the more complete data after Sterkiella histriomuscorum Foissner et al., 1991, GenBank accession no. FJ545743 46 .

Molecular data and phylogenetic analyses.
A critical analysis of the ITS region revealed that a large number of sequences in GenBank, which shared up to 85% (usually above 90%) identity with our organism, had been annotated as uncultured fungi. With a similar level of identity, the available sequences of hypotrichs could also be retrieved. Using BLAST taxonomy reports, out of the best 1,092 hits, 748 were annotated as the phylum Ciliophora, 152 as fungi, 185 as unknown eukaryotes, 6 as unknown organisms, and 1 as moss. The information of those 152 fungal annotated hits can be found in Table S1.
Discussion comparison with closely related species. (Table 4) In the latest review of the family Amphisiellidae, Berger 3 revised 12 species in the genus Lamtostyla and further separated them into three groups based on the number of cirri in ACR, the number of dorsal kineties and the presence/absence of cortical granules, i.e., (1) Lamtostyla lamottei-group which comprises most Lamtostyla species, having an ACR shorter than 50% of body length, but composed of more than four cirri, and very likely lacking cortical granules; (2) Lamtostyla granulifera-group, where species are characterized by having four cirri in the ACR, three dorsal kineties, and cortical granules, including L. decorata and L. granulifera; and (3) Lamtostyla longa-group, where species have basically the same cirral pattern as those in the Lamtostyla granulifera-group, but have five (rather than three) dorsal kineties, including L. longa and L. raptans. After that, only two species, L. ovalis and L. salina, were described, and both of them should be assigned to the Lamtostyla lamottei-group since they have more than four cirri (8-16 Most common pattern 6 FVT anlagen plus an additional anlage www.nature.com/scientificreports www.nature.com/scientificreports/ cirri in the former, 5-13 cirri in the latter) in the ACR 17,45 . Morphologically, L. gui n. sp. should be assigned to the Lamtostyla granulifera-group by having the ACR composed of four cirri, three dorsal kineties and cortical granules 3 . Hence, in total, 15 Lamtostyla spp. are listed below, belonging to three subgroups (Table 4). www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ indicates the extra frontoventral cirrus produced by anlage IV; a proter with two pretransverse cirri (cirrus V/2 and VI/2) and four transverse cirri is shown on (c); (d) Ventral view of a late divider whose proter follows the normal six frontoventral transverse cirri anlagen pattern, whereas, having an additional anlage (arrow) between anlage IV and V in the opisthe. (e) Ventral view of a late divider where the proter follows the common morphogenesis pattern, but its opisthe has one extra anlage III-originated cirrus (arrow). (f) Ventral view of another late divider whose cirral alignment is normal in the proter but with one extra anlage VI-originated cirrus (arrow) in the opisthe, causing five cirri in ACR. (g) Ventral view of one proter with two pretransverse cirri and three transverse cirri. (h1, 2) Two interpretations on the origination of a supernumerary cirrus (arrow) in the same proter: it may come from anlage IV (shown in h1) or from one additional anlage (arrowhead) between anlage IV and V (shown in h2). Frontal cirrus highlights in purple, buccal cirrus in green, frontoventral cirrus in grey, ACR in red, pretransverse cirrus in yellow, and transverse cirrus in blue. Cirri originated from the same anlage are connected by dash lines. I-VI frontoventral transverse cirri anlage I-VI. Scale bars 40 μm. www.nature.com/scientificreports www.nature.com/scientificreports/ The other two species in the Lamtostyla granulifera-group, namely, L. decorata and L. granulifera, should be compared with our organism. Compared with L. decorata, L. gui has more (4-8, on average 5 vs. 2-4, on average 2.3) macronuclear nodules, about ten cortical granules distributed as a rosette (vs. dozens of cortical granules formed a conspicuous of plaque) around dorsal bristle, fewer pretransverse and transvers cirri (4-6 vs. 6-9), and a relatively plump body (the ratio of body length to width 3-5:1 vs. 4-9:1 in vivo) 47 . Lamtostyla gui differs from L. granulifera mainly in having (1)  Morphogenetic comparison with related taxa. The most important events in the morphogenetic processes of Lamtostyla gui n. sp. can be summarized as follows: (1) in the proter, the posterior part of parental adoral zone of membranelles is renewed in situ; in the opisthe, the oral primordium is formed parakinetally from the leftmost one or two transverse cirri; (2) two sets of six FVT anlagen are formed in the middle stage; (3) four cirri in ACR originate from anlage V (posterior part) and VI (anterior part); (4) the FVT anlagen III to VI (sometimes IV to VI) provide the rearmost cirri to form the transverse cirri, while anlage II does not form a transverse cirrus at all; (5) marginal row primordia and dorsal kinety anlagen develop intrakinetally; and (6) the macronuclear nodules fuse to form a single mass before dividing. The ontogenetic processes of L. gui confirm it to be an amphisiellid, since it shares the main apomorphy, namely the ACR comprises cirri originated from at least two anlagen 3,49 . Moreover, detailed morphogenetic processes have previously only been reported for L. australis and L. salina, both are representatives of the Lamtostyla lamottei-group 3,45 ; while our study on L. gui provides the first ontogenetic data for the Lamtostyla granulifera-group. A comparison of these two subgroups shows that there is only one difference between them, i.e., the parental adoral membranelles retain intact in the Lamtostyla lamottei-group (vs. renew partially in situ in the Lamtostyla granulifera-group) 3,45,50 . Nevertheless, since morphogenetic information are limited to few species, whether such morphogenesis difference stably supports the division of these two Lamtostyla subgroups or not still needs further confirmation. And the relationship of the three subgroups, or to be more precise, the inner-relationship of the genus Lamtostyla has not been resolved yet. www.nature.com/scientificreports www.nature.com/scientificreports/ Ultrastructure. No study about the ultrastructure of amphisiellids was previously available. Present study provides some first fine structure information for this hypotrichous group in terms of pellicle, cortical granules, ciliature, nuclear apparatus, and cytoplasm.
The pellicle is the "boundary" of the ciliates cortex, and it is sometimes covered by an additional membrane, the perilemma, and usually apposed to a layer of subpellicular microtubules 43 . Perilemma is lacking in Lamtostyla gui, which recalls the uncertainty of its presence in hypotrichs, i. e., it has been reported in some species of hypotrich ciliates 51-54 , whereas absent in others [55][56][57] . Berger 1 discussed the correlation relationship between the arrangement of subpellicular microtubules and body rigidity in hypotrichs, that is, species with a flexible body (such as Oxytricha fallax Stein, 1859) bear a single layer of subpellicular microtubules, whereas species with a rigid body (such as Stylonychia mytilus Ehrenberg, 1838) bear subpellicular microtubules arranged in crosswise layers 54,[58][59][60] . Such deduction is supported by present study, as L. gui has a soft body and meanwhile a single layer of subpellicular microtubules.  www.nature.com/scientificreports www.nature.com/scientificreports/ small subectoplasmic rootlets are not found, while serval microtubule bundles left of cirrus are observed, which indicates it shares more similarity with urostyloids than other hypotrichs, e.g., oxytrichids. However, in urostyloids, the linear microtubule arrays are present in both longer sides of cirri 57,63,64 , while they present only in the left side in L. gui. Whether the fine structure of associated microtubules reflecting systematic position or not may deserve further discussion.
Pharyngeal discs existed in the oral cortex have been reported in many ciliate taxa and considered as food vacuole membrane precursors [65][66][67][68][69] . For ciliates in the subclass Hypotrichia, limited reports regarding the morphology of pharyngeal discs are available, i.e., only in the genus Stylonychia Ehrenberg, 1830 and Thigmokeronopsis Wicklow, 1981, where there are small vesicles-like structures distributed in the cytoplasm of oral area 59,70 . The flattened saccules existed in the buccal area of L. gui are pharyngeal discs considering their location and membrane nature. However, they represent a different form from those in Stylonychia and Thigmokeronopsis. This might be related with the systematic assignment of these three species, i.e., Stylonychia and Thigmokeronopsis belong to the family Oxytrichidae and Pseudokeronopsidae Borror & Wicklow, 1983, respectively, whereas, Lamtostyla represents the family Amphisiellidae 1-3 .

Marker gene sequences and phylogenetic analyses.
In the present study, we characterized the almost complete ribosomal operon of Lamtostyla gui n. sp. The available SSU rRNA data only allows us to observe that L. ovalis and L. salina (both belonging to the Lamtostyla lamottei-group) show an identity of 99.58% that is higher than 98.21% and 97.98% which is the identity they respectively present with L. gui (belonging to the Lamtostyla granulifera-group). Additionally, L. gui presents an apparently higher evolutionary rate compared with other two species in the phylogenetic tree. The present SSU rRNA phylogenetic tree does not support nor deny the possible monophyly of the genus Lamtostyla, which is in agreement with previous studies 17,45 and the phylogenetic position of Lamtostyla still remains undetermined, since the present topology is rather unstable with low support values across the tree and the lack of molecular phylogenetic information for the type species. Moreover, our species also shows high similarity in the SSU rRNA gene sequences of some familial unknown hypotrichs, namely, Bistichella cystiformans KJ509196 (98.03%), Bistichella variabilis HQ699895 (98.10%), Orthoamphisiella breviseries AY498654 (97.99%), Orthoamphisiella sp. JQ723974 (98.03%), Parabistichella variabilis JN008943 (98.15%), Uroleptoides longiseries MH143251 (97.68%) and Uroleptoides magnigranulosa AM412774 (98.27%). It is worth to reminding that the reciprocal positions of all these species or clades are rather unstable; indeed they emerge from the region which is highlighted in red (Fig. 10) where relative positions are not properly resolved using only SSU rRNA gene sequences (posterior probability values lower than 0.50). Consequently, any discussion about molecular similarity among these organisms/clades is, at present, only speculative and other molecular markers will be necessary in the future to properly resolve the issue. It is intriguing that, morphologically, Lamtostyla significantly differs from Bistichella Berger, 2008 in usually having one buccal cirrus (vs. more than one) and a short ACR originated from anlage V and anlage VI (vs. two separately long frontoventral cirral rows originated from V and VI, which are much longer than the ACR) 3,9,80 . The genus Orthoamphisiella Eigner & Foissner, 1991 differs from Lamtostyla in having a long frontoventral cirral row forming from a single anlage (vs. from at least two anlagen), more than one buccal cirrus (vs. usually one), and transverse cirri absent (vs. present) 3,4 . The genus Parabistichella Jiang et al., 2013 is a bakuellid-like hypotrichs with a midventral complex composed of pairs and row which is absent in Lamtostyla; beside, over six frontoventral-transverse cirri anlagen are present during the division of Parabistichella, whereas Lamtostyla usually have six frontoventral-transverse cirri anlagen 3,13 . The type species in both Lamtostyla and Uroleptoides Wenzel, 1953 are not well described morphologically and molecularly, causing the uncertain familial classifications of those two genera; now, they are preliminary recognized as amphisiellids with only one pragmatic solution to separate them, that is, by the length of ACR (more than 50% of body length in Uroleptoides against less than 50% in Lamtostyla) 3 .
In order to verify the monophyly of Lamtostyla as well as many other genera of hypotrichs, additional molecular markers showing a higher evolutionary rate should be used, such as ITS or mitochondrial cytochrome oxidase subunit 1 (COI) gene. Indeed, in the genera where SSU rRNA presents similar features, like Tetrahymena Furgason, 1940 and Paramecium Müller, 1773, these additional markers are routinely characterized and used for fast identification [81][82][83][84][85][86] . Additionally, the combined phylogenetic analyses of multiple genes (e. g., SSU rRNA, ITS1-5.8S-ITS2 and LSU rRNA or phylogenomic data) with increased sampling are now becoming popular within ciliates [87][88][89][90][91] , also in the Hypotrichia 31,32,34,35,38 , to study the evolutionary relationships, and are revealing to provide more robust interpretations. Hence, it becomes reasonable that as many as possible molecular marker genes of hypotrichous species should be characterized and added to species description or re-description to provide more information in solving complex systematics problems.
Concerning ITS and LSU rRNA data, due to the present lack of sequences from related species, we could not use these data for phylogenetic analyses, and especially, to test the monophyly of Lamtostyla. As an additional note, it is worth mentioning that the properly taxonomically annotated complete ribosomal operon sequence we Scientific RepoRtS | (2020) 10:3763 | https://doi.org/10.1038/s41598-020-60327-5 www.nature.com/scientificreports www.nature.com/scientificreports/ produced, in particular the ITS region, will help correcting mis-annotations present in GenBank. ITS region is commonly used as the phylogenetic marker to identify fungi in culture independent studies [92][93][94][95] . The performed GenBank survey using ITS1-5.8S-ITS2 sequencing data of L. gui, highlights the presence of up to 152 sequences attributed to fungi, which are likely misannotated deriving either from the phylum Ciliophora or being of the chimeric origin. In addition to reveal probable mis-annotations, our survey also highlights the presence of ciliates, especially hypotrichs, in the environment/matrices generally studied for the presence of fungi. Intriguingly, most of these environment/matrices represent human byproducts, e.g., compost 96 , forest soil 97,98 , or even extreme environment, e. g., crater lake 99

Material and Methods
Sample collection, cultivation and identification. The mixtures of water and sediments were collected with clean jars on 14 November, 2015 from the surface of a marsh wetland in North Beach (31°36′50.22″N, 121°49′5.27″E), Chongming Island, Shanghai, China. The water temperature was 12 °C, salinity was 10‰ and pH was 8.2. Samples were transported to laboratory, then, poured out in Petri dishes and maintained at room temperature (about 24 °C). Pre-sterilized wheat grains were added to the Petri dishes to enrich the growth of bacterial food for ciliates. Lamtostyla gui n. sp. appeared one week later in the raw culture, together with Urosoma salmastra (Dragesco & Dragesco-Kernéis, 1986) Berger, 1999. The feeding of L. gui on U. salmastra caused the acute decline of the latter and consequently the decline of itself within a short period. Later, L. gui throve again with the explosion of an unidentified scuticociliate. Two uniprotistan cultures were set up, each starting with about ten isolated cells from the raw sampling. In one culture, a wheat grain was added to enrich bacterial food, whereas, in the other, L. gui was feed with pure culture of the unidentified scuticociliate.
Living cells were directly isolated from the Petri dishes using micropipettes and observed with bright field and differential interference contrast microscope (Olympus BX 51) at a magnification of 100-1,000×. Commercial protargol powder was used following staining protocol 101 to reveal the nuclear apparatus and the infraciliature. Stained specimens were counted and measured at a magnification of 400-1,000×, while drawings were made with the aid of a camera lucida at a magnification of 1,250×. In illustrations of morphogenetic processes, old (parental) ciliary structures were depicted by contours whereas new structures were shaded black. Systematics and terminology are mainly according to Berger' s monograph 3 . electron microscopy. For SEM, the specimens from one poly-clonal culture using bacteria as food source were fixed in a 1:6 mixture with 1% O S O 4 in 0.1 M cacodylate buffer (pH 7.2) and saturated solution of HgCl 2 in distilled water for 10 mins at room temperature; then cells were washed three times in 0.1 M phosphate buffer (pH 7.0) to remove fixation solution; after alcohol dehydrations and critical point drying by CO 2 , cells were finally coated with platinum; observation was conducted under a Hitachi S-4800 scanning electron microscope with accelerating voltage of 5.0 kV.
Transmission electron microscopy (TEM) preparation was obtained by fixing specimens from one poly-clonal culture using bacteria as food source in a 1:1 mixture of 2% OsO 4  www.nature.com/scientificreports www.nature.com/scientificreports/ polymerized at 37 °C for 16 h, 45 °C for 24 h and 60 °C for 48 h. Thin sections were cut with a diamond knife and then placed on copper grids using uranyl acetate and lead citrate for staining; finally, the sections were observed under a Hitachi HT7700 transmission electron microscope with accelerating voltage of 100 kV.

DNA extraction, polymerase chain reaction (PCR) amplification and sequencing. Two cells
form one poly-clonal culture using bacterial food source were washed three times in distilled water and then conducted total genomic DNA extraction using the DNeasy Tissue Kit (Qiagen, Hilden, Germany), following the manufacture's instruction. The SSU rRNA gene sequence was amplified by touchdown PCR, using Q5 ® Hot Start High-Fidelity DNA Polymerase (New England BioLabs version cat. no. M0494S) and the universal primers 102 . Cycling parameters for PCR amplifications were as follows: 98 °C for 30 s, followed by 17 cycles of 98 °C for 10 s, 67 °C (decreasing by 1 °C per cycle) for 30 s, and 72 °C for 1 min; and followed by 18 cycles of 98 °C for 10 s, 50 °C for 30 s, and 72 °C for 1 min; then 72 °C for 5 mins for final extension. Cloning was performed using the pEASY ® -Blunt Cloning Kit (TransGen Biotech, Beijing, China) following the manufacture's instruction. Bidirectional sequencing was performed by the Thermo Fisher Scientific China Co. Ltd. (Shanghai, China) using the M13-47 and M13-48 primers. Furthermore, in order to avoid mismatches in the sequencing results obtained from cloning product, direct sequencings of a new purified PCR product (NucleoSpin ® Extract II Kit, Macherey-Nagel, Germany) were also made (GATC Biotech AG, European Custom Sequencing Centre, Germany) using the internal primers R536, F783, and R1052 103 . Results were 100% identical.
phylogenetic analyses based on SSU rRnA. The obtained SSU rRNA sequences were assembled using Chromas Lite 2.1 software and compared with the non-redundant sequence database using NCBI-BLAST 104 , then aligned using the editor and alignment tools from the ARB program package 105 together with related sequences contained in the SSU rRNA SILVA 102 database 106,107 and some latest released amphisiellids sequences on GenBank database. The alignment was then corrected taking into account the base-pairing scheme in rRNA secondary structure. Similarity values among sequences were calculated using the appropriate tool from the ARB software package, after trimming the PCR primer region. Phylogenetic analyses were performed on a selection of 104 Fig. 10. Out of the species selection and alignment, we produced different databases for later analysis. These databases were produced applying filters to retain or exclude columns whose overall similarity was either, 5% or below, or 98% or above (gap was counted as a fifth character for similarity calculation). The rationale for excluding low similarity region (5% or below) was to reduce possible "noise" (saturated signal) from hypervariable region. The rationale to exclude columns with a similarity of 98% or higher was to exclude columns containing identical characters (100% identity, uninformative) and columns in which only one sequence out of 104 was different from the others (99% identity). Indeed, in the last case, the single difference in the highly conserved region could be the result of a sequencing error. ML trees were calculated with the PHYML software version 2.4 108 from the ARB package, performing 100 pseudo-replicates. BI analyses were performed with MrBayes 3.2 109 using three runs each with one cold and three heated Monte Carlo Markov chains, with a burn-in of 25%, iterating for 1,000,000 generations. Analyses were performed with the different generated filters and topologies of the trees compared. (2020) 10:3763 | https://doi.org/10.1038/s41598-020-60327-5 www.nature.com/scientificreports www.nature.com/scientificreports/