Discocelia Plataet Sp. n., a Small Incertae Sedis Cercozoan Flagellate

ABSTRACT

Cercozoa = Filosa (Rhizaria, SAR) is one of the largest rhizarian subgroups and consists of a diverse assemblage of amoeboid and flagellated protists. They are ecologically significant in microbial food webs, widely diverse, and even abundant in soils and deep marine sediments according to environmental sequencing. In spite of this, the cercozoan phylogeny remains poorly resolved by SSU rRNA gene analysis, and omics data are available for only a few well‐characterized species. Here, we have sequenced the transcriptomes of three new gliding monadofilosan strains: the glissomonad RAM19S6, the marimonad CRO19P5, and the discocelid GT001. Because of its unusual morphology, we performed a thorough morphological characterization of the strain GT001 using light and electron microscopy and described a new species, Discocelia plataet sp. n. Transmission electron microscopy and expansion microscopy revealed the structure of the flagellar apparatus, allowing us to identify cercozoan microtubular root homologies and supplement our knowledge of the discocelid cell structure with new details. Unique features of the new species are the absence of body tip and velum tip, discoidal mitochondrial cristae, and presence of an acronema on the posterior flagellum. We discuss the phylogenetic position of the three strains within Monadofilosa and the evolutionary context of the order Discocelida.

1. Introduction

Discocelis saleuta (Vørs 1988) Cavalier‐Smith 2013 was described from marine sites of Denmark based on morphological observations using light and electron microscopy (Vørs 1988). The genus name Discocelis, an invalid homonym of the flatworm genus Discocelis Ehrenberg, 1836, was later replaced with Discocelia by Cavalier‐Smith (2013): “we must distinguish flatworms from flat protozoa”. Ultrastructural studies have shown unique morphological features of D. saleuta that hampered it to be assigned to any known group of eukaryotes (Vørs 1988). Such characteristics included the pointed dorsal velum and cell body tip formed by underlying microtubules, the marginal row of round dense extrusomes, two naked heterodynamic flagella, a small and extremely dorsoventrally flattened cell body, and the characteristic structure of the flagellar apparatus. Vørs (1988) found three microtubular roots, a, b, and c, the latter consisting of two branches, c1 and c2. Two more elements of the cytoskeleton, the microtubules supporting the velum and the cell margins, were not included into the flagellar root system.

The only other described species, D. punctata (Larsen and Patterson 1990; Cavalier‐Smith 2013), from marine sites of Fiji and Brazil, was described based on light microscopy observations (Larsen and Patterson 1990), having larger cell size and a visible row of punctae (i.e., extrusomes) around the posterior and lateral cell margins. Since then, several observations of both D. saleuta and D. punctata were recorded from different marine samples around the globe (Aydin and Lee 2012; Ekebom et al. 1995/6; Lee 2002, 2006, 2015; Lee and Patterson 2000; Tong 1997; Tong et al. 1998; Vørs 1992). However, these identifications were vaguely made only with light microscopy and sometimes questionable: for example, D. saleuta was often described without the characteristic cell body tip and velum tip, and with no anterior flagellum. Indeed, Cavalier‐Smith (2013) noted that these findings likely represented undescribed species.

Up to date, there was no molecular data available for Discocelia, and all speculations on the evolutionary affinity of this lineage with any other eukaryotic group were based on the single ultrastructural analysis that was made based on a suboptimal fixation, as acknowledged by the author herself (Vørs 1988). This led to the classification of Discocelia in diverse lineages such as Bicosoecida (Karpov 2000), Cercomonadida (Cavalier‐Smith 1993), Apusozoa (Cavalier‐Smith 1997), and Glissodiscea as sister to Planomonadida (=ancyromonads) and Mantamonadida (Cavalier‐Smith 2013; Heiss et al. 2011). The latter position was proposed due to Discocelia's highly flattened cell body, emergence and direction of flagella, reduced acronematic anterior flagellum and trailing posterior flagellum, and inability to swim. Indeed, Mantamonas is superficially very similar to Discocelia, but it has never been studied ultrastructurally. Ancyromonas shares some similarities with Discocelia in the structure of the flagellar apparatus and flagellar transition zone (Cavalier‐Smith 2013; Heiss et al. 2011). However, these structures were studied insufficiently in both groups. In the latest revision of the classification of eukaryotes (Adl et al. 2019), Discocelia was placed in Cercozoa incertae sedis (Rhizaria, SAR) without any references or explanations, probably based on unpublished data.

Rhizaria is one major eukaryotic clade, part of the huge SAR (Stramenopiles, Alveolata, Rhizaria) supergroup, and consists mostly of heterotrophic single‐celled organisms (Adl et al. 2019). The Rhizaria clade was initially united based on molecular data; whereas morphological synapomorphies were identified not long ago, including the unique structure of the flagellar transition zone with a proximal hub‐lattice and distal hub‐spoke structures, and reticular or filose pseudopodia (Cavalier‐Smith et al. 2018; Bass et al. 2009; Cavalier‐Smith and Karpov 2012).

Cercozoa (=Filosa) is one of the largest rhizarian subgroups and consists of a diverse assemblage of amoeboid and flagellated protists. Besides the widely known filose thecate amoebae (e.g., Euglypha, Pseudodifflugia), it also contains naked and testate reticulose amoebae (e.g., Gromia, Filoreta), social amoebae with fruiting bodies (Guttulinopsis), amoeboflagellates (e.g., Cercomonas, Paracercomonas), diverse parasites (e.g., Mikrocytos, Plasmodiophora, Viridiraptor), primary (Paulinella) and secondary (Bigelowiella) algae, heliozoans (e.g., Clathrulina, Hedriocystis), colonial flagellates (e.g., Spongomonas, Rhipidodendron), as well as gliding or swimming heterotrophic flagellates (e.g., Mataza, Katabia, Sainouron, Allapsa). This huge morphological and lifestyle diversity explains cercozoan adaptability to a variety of habitats, but at the same time makes it difficult to identify synapomorphies for the whole group. Indeed, many protist species with previously unknown affinity were subsequently classified within Cercozoa thanks to molecular phylogenetics (Bass et al. 2009; Burki et al. 2002; Chantangsi et al. 2008; Chantangsi and Leander 2010; Hoppenrath and Leander 2006a, 2006b; Kühn et al. 2000; Longet et al. 2004; Vickerman et al. 2005).

Environmental SSU rRNA gene sequence analyses support that cercozoans are widely diverse and ecologically significant in microbial food webs (Cavalier‐Smith et al. 2018) and particularly abundant in soils (e.g., Harder et al. 2016; Dumack et al. 2020) and deep marine sediments (Cordier et al. 2022). However, the internal cercozoan phylogeny remains unresolved, with many incertae sedis lineages (Adl et al. 2019), mostly because the SSU rRNA gene is not informative enough (Cavalier‐Smith and Chao 2003; Howe et al. 2011; Scoble and Cavalier‐Smith 2014) and the lack of transcriptome or genome sequences available for extensive phylogenomic analyses (Cavalier‐Smith et al. 2018; Irwin et al. 2019). Only a few representative clades have sufficiently good omics data available: Aurigamonas in Pansomonadidae; Paracercomonas and Brevimastigomonas in Paracercomonadida; Mataza and Aulacantha in Thecofilosea; Guttulinopsis in Sainouroidea; and Abollifer, Paulinella, and Euglypha in Imbricatea (a gallimaufry group). Unfortunately, the available transcriptomic data from Minimassisteria (Granofilosea), Micrometopion (Metromonadea, Imbricatea), Helkesimastix (Sainouroidea), Thaumatomonas (Silicofilosa, Imbricatea), Nudifila (Imbricatea), Sandona (Glissomonadida), and Neocercomonas (Cercomonadida) are too partial and likely cross‐contaminated to be used further (Cavalier‐Smith et al. 2018). Therefore, high‐quality omics data are still lacking for Cercomonadida, Glissomonadida, Viridiraptoridae, and any yet‐to‐be‐discovered clades.

In an effort to improve the sampling of benthic heterotrophic flagellates (such as apusomonads (Torruella et al. 2023), ancyromonads (Yubuki et al. 2023), mantamonads (Blaz et al. 2023), and excavates (Galindo et al. 2023)), we have isolated and sequenced the transcriptomes of three cercozoan strains: the glissomonad strain RAM19S6, the marimonad strain CRO19P5 (both part of Imbricatea), and the novel species Discocelia plataet strain GT001, one of these incertae sedis lineages. Together with the detailed morphological characterization of Discocelia, we discuss its evolutionary position within Cercozoa.

2. Methods

2.1. Sample Collection and Culture Establishment

Three strains of benthic heterotrophic flagellates were isolated by single cell micromanipulation, resulting in monoeukaryotic cultures with unknown environmental bacteria as food. Two strains were isolated from marine coastal sediments: Discocelia plataet strain GT001 from Roc'h ar Bleïz, Roscoff, France, and Marimonad strain CRO19P5 from Malo jezero, Mljet Island, Croatia. The other strain, Glissomonad strain RAM19S6, was isolated from freshwater sediment from a stream at the Rambouillet forest, France. Strains were grown at room temperature in cell culture flasks, just covering the bottom with 1% yeast tryptone in 0.2 μm‐filtered seawater and Volvic water, respectively. Strains CRO19P5 and RAM19S6 were subsequently lost; only the culture of D. plataet is available in the DEEM team culture collection upon request.

2.2. Molecular Analyses: Cell Culture, RNA‐Seq, SSU rDNA Identification, and Phylogenomics

Cells from grown cultures were scratched from the culture flasks with a cell scraper, collected in 50 mL Falcon tubes, and concentrated by centrifugation for 15 min at 15,000 g and 10°C. For each strain, total RNA was extracted from high‐density cultures, using the RNeasy mini Kit (Qiagen) following the manufacturer's protocol, including DNAse treatment; and quantified using a Qubit fluorometer (Thermo Fisher Scientific). After polyA mRNA selection, cDNA Illumina libraries were constructed, tagged, and paired‐end (2 × 150 bp) sequenced with NovaSeq 6000 S2 (Eurofins Genomics, Germany) in the same sequencing run (NG‐25522). Raw reads were submitted to SRA NCBI database under the Bioproject PRJNA1162809.

High‐quality reads, checked with FastQC v0.11.8 (Andrews 2010), were used for de novo assembly using Spades v3.13.1 (Bankevich et al. 2012) in –rna mode. To identify the species taxonomic affiliation, ribosomal operon sequences were retrieved from the transcriptomes by BLASTn v2.2.26+ (Camacho et al. 2009), querying the OM966641 sequence from Podomonas capensis (Torruella et al. 2023). They were submitted to GenBank under accession numbers PQ362543 for CRO19P5, PQ362544 for RAM19S6, and PQ362545 for GT001. Barrnap v0.9 (Seemann 2013) was then used to annotate the three genes (Table S1) and each SSU rRNA sequence was searched by BLASTn against the core_nt database; the sequence similarity and distance trees were inspected.

For the phylogenomic analysis, transcripts were translated using TransDecoder v5.5 (https://github.com/TransDecoder/) with the LongOrfs option. Gene completeness was investigated with BUSCO v4 (Simão et al. 2015) using eukaryota_db10 (Table S2). The three sets of predicted peptides were included in a paneukaryotic phylogenomic dataset, as in a previous study (Torruella et al. 2024) using BLASTp. Briefly, homologs from 351 markers were identified using proteins from Homo, Saprolegnia, Spizellomyces, and Diphylleia as queries. Non‐orthologous sequences were removed after manually inspecting the single marker alignments with 230 taxa (with Aliview v1.26 (Larsson 2014)) and the corresponding trees (with Figtree v1.4.3 (http://tree.bio.ed.ac.uk/software/figtree/)), done with MAFFT v7.427 (Katoh and Standley 2013) L‐INS‐i with 1000 iterations, trimmed with Trimal v1.4.rev22 (Capella‐Gutiérrez et al. 2009) in automated1 mode, and inferred with FastTree v2.1.11 (Price et al. 2010). For the final dataset, most taxa were removed, keeping the rhizarian ingroup and closest outgroups (two alveolates and two stramenopiles). To maximize the representation of rhizarian data, markers with less than 80% of taxa (i.e., < 14 sequences) were removed, resulting in a dataset of 169 markers for 18 species. After realigning and trimming with the same software and parameters, the supermatrix was concatenated with alvert.py from the barrel‐o‐monkeys (http://rogerlab.biochemistryandmolecularbiology.dal.ca/Software/Software.htm#Monkeybarrel) and contained 56,888 amino acid positions (Table S3). The maximum likelihood (ML) phylogenetic tree was inferred using IQ‐TREE (Minh et al. 2020) v1.6 with the PMSF approximation (Wang et al. 2018) with a guide tree inferred with the LG + F + R5 + C60 mixture model, and statistical support was calculated with 1000 ultrafast bootstraps (UFBS). All data have been deposited in the figshare repository (https://doi.org/10.6084/m9.figshare.27074848). For the SSU rRNA gene phylogenetic analyses, all three new strains were placed in a dataset containing 108 rhizarian sequences identified from several previous datasets (Howe et al. 2011; Bass et al. 2016, 2018) selected by length (more than 1700 bp), representing all main groups in Cercozoa, and 5 sequences of Radiolaria (Retaria) as an outgroup. The multiple sequence alignment was constructed using the L‐INS‐i algorithm in MAFFT v.7.475 (Katoh and Standley 2013) and trimmed using trimAl v.1.2 (Capella‐Gutiérrez et al. 2009) with the automated1 method. The number of analyzed sites after trimming was 1819 bp. To infer Bayesian phylogenetic trees, the parallel MPI version of MrBayes v.3.2.7a (Ronquist et al. 2012) was used with four categories of gamma‐distributed among‐site rate variation under the GTR + I + GAMMA4 substitution model. To calculate posterior probabilities of individual nodes, four independent Metropolis‐coupled Markov chains were run for 5 million generations and summarized after a 50% burn‐in. The convergence of log‐likelihood values and model parameters for chains was verified using a plot and convergence diagnostics provided by the MrBayes sump utility. ML phylogenetic trees were inferred using IQ‐TREE v.2.2.2.6 (Nguyen et al. 2015) with 1000 non‐parametric bootstrap pseudoreplicates under the best‐fit model (TIM2 + F + I + R6 substitution model) determined by the in‐built ModelFinder (Kalyaanamoorthy et al. 2017). All trees were visualized using the cloud‐based platform iTOL (Letunic and Bork 2024).

2.3. Morphological Analyses: Microscopy and Imaging

Light microscopy (LM) observations were made using an upright Zeiss Axioplan 2 microscope equipped with an oil‐immersion DIC and phase contrast 100× objective. Videos were recorded with a Sony α9 digital camera. Images as still frames were captured using a Light Alloy video player (https://light‐alloy.com/).

For scanning electron microscopy (SEM), cells of D. plataet sp. n. strain GT001 attached to 12‐mm coverslips were fixed in 2% glutaraldehyde diluted in sterile seawater for 30 min. After fixation, the cells were washed with 0.2 M sodium cacodylate buffer pH 7.2 for 2 × 10 min, in distilled water for 2 × 10 min, and dehydrated in a graded series of ethanol baths (30%, 50%, 70%, 80%, 90%, 1 × 10 min each; 100% 3 × 10 min). Then, the cells were washed with 100% hexamethyldisiloxane for 3 × 10 min and subsequently dried at room temperature. Dry coverslips were mounted on aluminum stubs, coated with 6 nm platinum, and observed using a GeminiSEM 500 (Carl Zeiss, Germany) scanning electron microscope.

The ultrastructure expansion microscopy (U‐ExM) protocol was adapted from Gambarotto et al. (2019) and Mikus et al. (2024). The 8‐mm coverslips with attached cells of D. plataet sp. n. strain GT001 were fixed with 4% formaldehyde diluted in sterile seawater for 20 min at room temperature and subsequently washed with phosphate‐buffered saline (PBS) 2 × 10 min. Then, coverslips with cells were incubated in AA/FA solution (1% acrylamide, 0.7% formaldehyde, diluted in PBS) for 12 h at 37°C. Gelation and polymerization were performed with a mix of ammonium persulfate (APS), N,N,N′,N′‐tetramethylethylenediamine (TEMED), and monomer solution (19% sodium acrylate; 10% acrylamide; 0.1% N,N′‐methylenbisacrylamide in PBS) for 1 h at 37°C. Denaturation was done using denaturation buffer (50 mM tris(hydroxymethyl)aminomethane (Tris) pH 9.0, 200 mM NaCl, 200 mM sodium dodecyl sulfate (SDS), adjusted to pH 9.0 with HCl) at 95°C for 1.5 h. Subsequent expansion of the gel was performed by washing the gel with ddH2O 10 × 5 min. For α and β‐tubulin antibody staining, the gel was washed in PBS 3 × 5 min to remove excess water, and incubated in primary antibody solution (mouse IgG2a anti‐β‐tubulin ABCD_AA344 and anti‐α tubulin ABCD_AA345) 1/300 diluted in 0.1% PBST (0.1% triton in PBS) supplemented with 3% BSA for 12 h at 37°C on a rocking platform, washed in PBS buffer 3 × 5 min, and incubated with secondary antibody solution (goat anti‐mouse IgG (H + L) cross‐adsorbed secondary antibody, Alexa‐Fluor 488) 1/1000 diluted in 0.1% PBST supplemented with 3% BSA for 2 h at 37°C in the dark on a rocking platform. For DNA and protein staining, gels were washed in PBS buffer 3 × 5 min and incubated with Hoechst 33,342 nuclear stain (Thermo Fisher; 62,249) and Alexa‐Fluor 594 NHS Ester (Thermo Fisher Scientific) diluted directly in PBS for 2 h at room temperature in the dark on a rocking platform. For confocal microscope observations, the gels were expanded by washing in ddH2O 3 × 5 min, mounted on coverslips covered with poly‐D‐lysine attached to slides using iSpacer, and examined on a Leica SP8 confocal microscope with an HC PL APO × 40/1.25 glycerol objective. Images were analyzed, measured, and visualized with the Fiji ImageJ program (https://fiji.sc/).

For transmission electron microscopy (TEM), a cell pellet was obtained by centrifugation for 20 min at 3234 g, resuspended, and mixed 1:1 v/v with fixative solution, containing 2.5% glutaraldehyde and 2% osmium tetroxide diluted in 0.1 M cacodylate buffer pH 7.2 and fixed for 30 min on ice. After fixation, the cell pellet was collected by centrifugation for 10 min at 3234 g, washed with 0.1 M cacodylate buffer pH 7.2 1 × 10 min, and with distilled water 2 × 10 min. After dehydration in an alcohol series (30, 50, 70, 80, 90, 96, 1 × 10 min each; 100% 2 × 10 min) and propylene oxide (2 × 10 min), the pellets were embedded in Low Viscosity resin (EM 0300 Sigma‐Aldrich). Ultrathin sections (70 nm) were prepared with a Leica EM UC6 ultramicrotome (Leica Microsystems, Germany) and observed using a JEM 1400 transmission electron microscope (JEOL, Japan). Measurements were taken using Fiji (https://fiji.sc/) on 21 cells, measuring the structures twice to diminish biases.

3. Results

3.1. Phylogenetic Position of Three New Strains of Monadofilosa

All three strains were taxonomically affiliated to the Monadofilosa within Cercozoa using the SSU rRNA sequences (Table S1; Figure S1). Strain CRO19P5 probably belongs to the order Marimonadida (part of Imbricatea), as it always branched with moderate statistical support in a clade containing Pseudopirsonia, Auranticordis, and Cyranomonas. Strain RAM19S6 is likely placed in the order Glissomonadida, as it robustly branched with Dujardina stenomorpha. The position of strain GT001, morphologically identified as a Discocelia species, was unresolved and, depending on the taxon sampling, it branched with low statistical support either as sister to Metromonas, Spongomonas, or Verrucomonas (data not shown). Although the cultures of the first two small strains (< 5 μm large) did not survive to perform a detailed morphological characterization, few light micrographs supported that the marimonad strain CRO19P5 might be similar to Cyranomonas (Lee and Park 2016), being small round biflagellates < 5 μm; and that the glissomonad strain RAM19S6 was a globular gliding biflagellate (Figure 1).

FIGURE 1.

Maximum likelihood phylogenomic analysis of Rhizaria including the new glissomonad RAM19S6, marimonad CRO19P5, and Discocelia plataet sp. n. GT001. The tree was inferred from 56,888 amino acid positions and 18 taxa with the LG + F + R5 + C60 model, and rooted with Stramenopiles and Alveolata. The numbers at branches represent ultrafast bootstrap values (1000 replicates). Black dots indicate maximum support, and the scale bar indicates the number of expected substitutions per unit branch length. Under the tree, differential interference contrast micrographs of the three new strains, with scale bars indicating 5 μm.

The position of the three strains within Monadofilosa was further confirmed by the Maximum Likelihood (ML) phylogenomic analysis of a dataset of 169 paneukaryotic markers including our highly complete transcriptomes (Tables S2, S3). Using two stramenopiles and two alveolates as outgroups, the monophyly of Rhizaria was fully supported. Two foraminiferan representatives of Retaria form the earliest‐branching rhizarian lineage (98% UFBS), followed by the monophyletic Endomyxa + Aquavolonida on one side (98% UFBS) and Cercozoa = Filosa on the other one (100% UFBS). Within Filosa, Chlorarachnea branched from Monadofilosa with maximum support. Finally, within Monadofilosa, the new marimonad strain CRO19P5 was sister to Paulinella (Imbricatea, Silicofilosa) and the new glissomonad strain RAM19S6 was sister to Aurigamonas (another glissomonad). Sister to the previous monophyletic group, the soil social amoeba Guttulinopsis and the new strain of marine benthic Discocelia branched together with moderate support (95% UFBS) (Figure 1).

3.2. External Morphology of Discocelia Plataet Sp. n

Light (LM) and scanning electron microscopy (SEM) observations of the new strain GT001 revealed the main morphological traits of Discocelia plataet sp. n. (Figure 2). Morphological characteristics and measurements of our new species and previously described Discocelia species are provided in Table 1. Living cell movements are shown in Videos S1–, S4. The cell body is dorsoventrally flattened, almost circular in outline, slightly broader laterally, measuring on average 3.10 × 3.55 μm (n = 21). Fixed cells (according to SEM) are smaller, 2.68 × 3.24 μm (n = 18). Flagella emerge subapically. The anterior flagellum (AF) is short and poorly visible in LM (Figure 2B). According to SEM, the AF is 0.46 μm in length, emerges on the dorsal side of the cell, and is directed posteriorly and to the left (Figure 2K–M). The posterior flagellum (PF) is 6.02 μm in length (2 times longer than the cell body), emerges from the ventral side of the cell, trails behind the cell, and bends gently when the cells stop (Figure 2B, Video S1, S2). The AF has a pointed tip (acronema) of about half of its length (Figure 2K–M), and the PF bears an acronema of 0.55 μm in length according to SEM (Figure 2A,K–M). The dorsal left side of the cell extends to a flat, rounded, and sometimes elongated velum (Figure 2A,B,K). A row of extrusomes along the entire cell margin is visible on some cells in SEM (Figure 2M). Fixed cells have rounded depressions on the dorsal side of the cell (Figure 2K–M), which probably represent vacuoles visible in both LM (Figure 2C,I) and U‐ExM (Figures 3O, S1–S6, see below). A single filopodium has been seen extending from the cell body (Figure 2I, Video S3). Cells slowly glide on the substrate and capture bacteria using a small pseudopodium (Figure 2F–H, Video S4). Cell division is longitudinal (Figure 2J). Cysts were not observed.

FIGURE 2.

External morphology of Discocelia plataet sp. n. cells, viewed from the dorsal side: (A–J) Living cells, visualized by light microscopy with differential interference contrast (A–C) and phase contrast (D–J). Stopped cells (A–C) with coiled posterior flagellum, and active gliding cells (D–J). Three instances of a cell with a captured bacterium (F–H). A cell before division (I) and the same cell during division (J). (K–M) Fixed cells visualized by scanning electron microscopy. Abbreviations: Ac, acronema; af, anterior flagellum; b, bacterium; ex, extrusomes; pf, posterior flagellum; ps, pseudopodium; v, vacuole; vl, velum. Scale bars: 5 μm for A–J; 2 μm for K–M.

TABLE 1.

Comparative morphology and measurements of Discocelia species.

References	Cell length, μm (min‐max)	Cell width, μm (min‐max)	AF, μm (min‐max)	PF, μm (min‐max)	Acronema on AF, μm (min‐max)	Acronema on PF, μm (min‐max)	Velum	Extrusomes (punctae)	Habitat
Discocelia plataet sp. n.
This study, living cells, n = 21	3.10 (2.16–4.41)	3.55 (2.44–5.04)	barely visible, inactive	6.02 (3.67–10.32)	not visible	not visible	Elongated rounded	not visible	Roc'h ar Bleïz, Roscoff, France
This study, SEM, n = 18	2.68 (2.12–3.43)	3.24 (2.47–3.88)	0.46 (0.35–0.60)	5.94 (4.91–7.45)	0.23 (0.16–0.3)	0.55 (0.36–1.03)		visible
This study, expansion (ExF = 5), n = 7	13.11 (10.18–16.84)	16.30 (10.63–20.98)	2.31 (1.83–2.67)	25.44 (16.7–31.44)	8.20 (5.32–10.45)	2.72 (2.22–3.56)		not visible
Discocelia saleuta
Vørs 1988	4–5	5–6	0.5‐1.5 μm, beats rapidly	5–8 μm	1/3 of the AF length*	n/d	pointed	not visible on LM	Limfjorden, Denmark
Aydin and Lee 2012	4–5	n/d	barely visible	1.2–1.5 × CL	n/d	n/d	n/d	not visible on LM	Aegean Sea, Turkey
Ekebom et al. 1995/6	4–4.75	n/d	not visible	~2 × CL*	n/d	n/d	n/d	n/d	Queensland, Australia
Larsen and Patterson 1990	3–6 × 5	n/d	very short, absent on small cells	1.5–2 × CL*	n/d	n/d	present on big cells, absent on small cells*	visible on LM	Queensland, Australia; Rio de Janeiro, Brazil; Panama
Lee 2006	~4	n/d	not visible*	1.5–2 × CL*	n/d	n/d	n/d	visible on LM	Queensland, Australia
Lee 2015	3.5–6.5	n/d	< 1 μm	slightly longer than the cell	n/d	n/d	n/d	visible on LM	South‐eastern Australia
Tong 1997	2.5–4.5	n/d	visible on TEM whole mount	~2.5 × CL	n/d	n/d	n/d	n/d	Southampton Water, U.K.
Vørs 1992	2–5	n/d	n/d	n/d	n/d	n/d	n/d	n/d	Gulf of Finland, Baltic sea
Discocelia punctata
Larsen and Patterson 1990	6.5–9	n/d	up to 3 μm, relatively inactive	up to 10 μm	n/d	n/d	n/d	visible on LM	Fiji; Rio de Janeiro, Brazil
Lee 2002	6.5–9	n/d	< 1.5 μm, inactive	~1.4 × CL	n/d	n/d	n/d	visible on LM	South Korea
Lee and Patterson 2000	~6	n/d	< 1 μm, inactive	slightly longer than the CL	n/d	n/d	n/d	visible on LM	Sydney, Australia
Tong et al. 1998	5	n/d	0.5 × CL	~2 × CL	n/d	n/d	n/d	visible on LM	Sydney, Australia

Note: The type descriptions are indicated in bold; *according to illustration; n/d – no data.

FIGURE 3.

Ultrastructure expansion microscopy (U‐ExM) of Discocelia plataet sp. n. cells, viewed from the ventral side: (A–E) cell #1, (F–M) cell #3, (N–P) cell #7, (Q–U) cell #2. For detailed illustrations of all 7 cells, see Figures S2–, S7. Black and white pictures represent inverted images with merged stacks from 9 to 25 (A, N, Q). Cyan represents Hoechst (DNA staining), yellow represents NHS ester (protein density), magenta represents α‐ and β‐tubulin (microtubules). Abbreviations: Ac, acronema; af, anterior flagellum; da, dorsal anterior root of AB (=R3 of eukaryotes = velum microtubule D. saleuta); dp1, dorsal posterior root of PB (=R1 root in eukaryotes = c2 root in D. saleuta); dp2, dorsal posterior root of AB (=c2 root in D. saleuta); fr, fiber associated with PB; fs, fiber associated with AB; lr, left anterior root of AB (=R4 root in eukaryotes = marginal microtubules in D. saleuta); m, mitochondrion; mb, microbody; n, nucleus; pf, posterior flagellum; sm, secondary microtubules; v, vacuole; vp1, ventral posterior root of PB (=R2 root in eukaryotes = b root in D. saleuta); vp2, ventral posterior root of AB (=c1 root in D. saleuta). The stack number is indicated on the top left corner. Scale bars: 10 μm. Expansion factor 5.

3.3. Ultrastructure via Expansion Microscopy (U‐ExM) of Discocelia plataet Sp. n

Illustrations of seven expanded cells are provided in Figures 3 and S2–, S7. Measurements of the expanded gels allowed us to determine an expansion factor (ExF) of 5. Thus, we calculated and report here corrected measurements in accordance with that ExF (see Table 1).

The nucleus, visualized by Hoechst staining, is located in close connection with the flagella in the anterior part of the cell (Figures 3, S2–S7, cyan). The nucleus is dorsoventrally flattened and slightly elongated in the longitudinal direction, usually with a small concavity at the posterior end. The diameter is 1.09 μm (5.43 μm expanded). Using NHS ester, a known protein pan‐labelling dye (M'Saad and Bewersdorf 2020; Sheard et al. 2023), we were able to see a rounded spherical organelle associated with the nucleus from its posterodorsal side (Figure 3I,Q, S3–S6). This structure corresponds to the microbody‐like organelle described by Vørs (1988), with a diameter of 0.53 μm (2.65 μm expanded). Two elongated organelles, associated with the nucleus on its dorsolateral sides, are presumably mitochondria (Figure 3I, S3, S4, S6), 0.85 μm long and 0.52 μm wide (=4.23 × 2.59 μm expanded). NHS staining shows several round ‘empty’ vacuoles located inside the cell, in the middle dorsal parts of the cell body (Figures 3G–I,O, S2,T–S7), about 0.45 μm in diameter (2.25 μm expanded). We observed a dividing cell in the last stage of cytoplasmic division, with some strands of fibers (NHS stained) extending to the dorsal part of the cell body (Figure S5).

The flagellar apparatus was visualized by α‐ and β‐tubulin antibody staining. The AF is very short, 0.46 μm in length (2.31 μm expanded), with a gradually tapering end (acronema) (Figure 3A). The PF is 5.09 μm in length (25.44 μm expanded) and bears an acronema 0.55 μm long (2.72 μm non‐expanded). The flagellar apparatus includes six microtubular roots, four of which are described by Vørs (1988) as roots a, b, c1, and c2, and two other roots, the velum and marginal microtubules. The names of the cercozoan flagellar roots (vp1, vp2, dp1, dp1, da, lr) and fibers (fr, fs) are given here in accordance with Karpov et al. (2006) and Cavalier‐Smith and Karpov (2012). The names of the eukaryotic flagellar roots (R1, R2, R3, R4) are given here in accordance with Moestrup (2000), Yubuki and Leander (2013).

The prominent posterior ventral root b (=vp1 = R2) originates from the posterior end of the base of the posterior basal body (PB) (Figures 3, S2–, S6). It runs posteriorly and to the left along the PF for a short distance, curving smoothly toward the posterior end of the cell. Around the midpoint of the cell, it bends to the right, perpendicularly to the longitudinal axis of the cell, eventually reaching the rightmost cell margin, where it meets the marginal microtubules and the distal end of root a (=dp2) (see below). Another posterior ventral root, c1 (=vp2), originates from the posterior end of the base of the anterior basal body (AB) (Figures 3, S2–, S6). It runs near the base of the PB, initially parallel to root b (=vp1 = R2), before crossing it and turning in the opposite direction, eventually terminating at the leftmost cell margin. Somewhat distal to the PB, the posterior root c2 (=dp1 = R1) branches perpendicularly from root c1 (=vp2) to the left, running in parallel to the distal part of root c1. The distal end of root c2 terminates at the left edge of the cell, at some distance from the distal end of root c1 (=vp2). At the origin of root c2 (=dp1 = R1), immediately posterior to the PB, a dense fiber, fr, is visible with NHS staining (Figures 3, S2–, S5). On the right side of the base of the AB, there is another large dense fiber, fs (Figures 3, S2–, S5). It nucleates microtubules of root a (=dp2) directed to the right along the anterior right edge of the cell and ends in the middle part of the cell near the distal end of root b (=vp1 = R2). According to Vørs (1988), root a (=dp2) consists of three microtubules, one of which is half shorter than the others. We observed one shorter microtubule that extends slightly to the side from this root (Figures 3F,G,J,K, S4).

The marginal microtubules (=lr = R4) originate from the space between the AB and PB and extend along the entire cell periphery (Figures 3, S2–S6). An additional line of microtubule(s) may extend at some distance, more dorsally from the main line (Figures 3A–E, S2) in the posterior region of the cell. The marginal microtubules terminate at the anterior end of the cell, although one may end at the right side of the cell margin, near the distal ends of roots a (=dp2) and b (=vp1 = R2). The microtubule that supports the velum (=da = R3) originates from the space between the basal bodies, slightly distal to the origin of the marginal microtubules (=lr = R4), and runs to the left, in parallel to the marginal microtubules (Figures 3, S2–S4, S6). At its proximal end, this microtubule is associated with fibrous material and nucleates some additional single microtubules directed posteriorly on the dorsal side of the cell (Figures 3, S2, S4). The velum microtubule (=da) terminates at the anterior left edge of the velum (Figures 3, S2–S4).

3.4. Ultrastructure via Transmission Electron Microscopy of Discocelia plataet Sp. n

Ultrastructural features and organelle orientations observed using transmission electron microscopy (TEM) were in good agreement with those observed using U‐ExM. The nucleus is located in the anterior half of the cell (Figure 4A,B), slightly dorsoventrally flattened, 935 nm (820–1044 nm, n = 15) in length and 620 nm (558–744 nm, n = 12) in width in sagittal section. The subcentral nucleolus is elongated, 287 nm (194–349 nm, n = 28) in length and 216 nm (181–295 nm, n = 17) in width (length to width ratio is 1.4). The basal bodies are located close to the nucleus (Figure 4A,E). We did not observe any structures connecting the basal bodies with each other or to the nucleus. The Golgi apparatus, with three cisternae, is located anterio‐laterally to the nucleus (not shown).

FIGURE 4.

Transmission electron microscopy of Discocelia plataet sp. n.: (A‐B) whole cell, longitudinal section, frontal plane (A), and sagittal plane (B); (C) cut through the para‐crystalline striated layer underneath the plasmalemma; (D) mitochondrion with discoidal cristae and associated dense body; (E) longitudinal section, (F,G) contiguous longitudinal sections, and (H‐K) serial longitudinal sections of the anterior part of the cell showing the basal bodies and flagellar roots; (L) microbody associated with the nucleus; (M) longitudinal section of the PB and transverse section of the AB. Abbreviations: Ab, anterior basal body; ab', additional anterior basal body in dividing cell; af, anterior flagellum; da, dorsal anterior root of AB (=R3 of eukaryotes = velum microtubule in D. saleuta); db, electron‐dense body associated with mitochondria; dp1, dorsal posterior root of PB (=R1 root in eukaryotes = c2 root in D. saleuta); dp2, dorsal posterior root of AB (=c2 root in D. saleuta); ex, extrusomes; fr, fiber associated with PB; fr', electron‐dense sheet‐like fiber associated with PB; lr, left anterior root of AB (=R4 root in eukaryotes = marginal microtubules in D. saleuta); lr', additional pair of marginal microtubules; mb, microbody; mt, mitochondria; n, nucleus; n, nucleolus; pb, posterior basal body; pf, posterior flagellum; v, vacuole; vp1, ventral posterior root of PB (=R2 root in eukaryotes = b root in D. saleuta); vp2, ventral posterior root of AB (=c1 root in D. saleuta). The number of sections is indicated in the upper right corner. Scale bars: 500 nm for A, B, E‐K; 200 nm for C, D, L, M.

Two elongated mitochondria are adjacent to the posterior lateral sides of the nucleus (Figure 4A). Mitochondria are 896 nm (542–1081 nm, n = 19) in length and 535 nm (301–794 nm, n = 34) in width. Mitochondrial cristae are discoidal (Figure 4A,D), with each crista about 140 nm in diameter and about 15 nm wide. Often, mitochondria have a closely associated small electron‐dense body (Figure 4D), 205 nm (163–261 nm, n = 13) in length and 141 nm (115–199 nm, n = 13) in width. A rounded microbody is adjacent to the posterodorsal side of the nucleus (Figure 4A,B,L); it is 389 nm (205–530 nm, n = 42) in diameter. Many cells contained large ‘empty’ vacuoles, measuring 604 nm (353–1107 nm, n = 32) in diameter (Figure 4B). We observed a para‐crystalline striated layer underneath the plasmalemma (Figure 4C), about 10 nm thick. Rounded extrusomes are located along the cell periphery (Figure 4B,E), 111 nm (81–144 nm, n = 40) in diameter.

Basal bodies are approximately 340–420 nm in length and oriented at 90° to each other; the proximal end of the PB is directed toward the side of the AB (Figure 4A,E–K,M). The transition zone is indistinct, with a poorly visible thin transverse plate (Figure 4A,E,G,H,M). Two small electron‐dense transverse discs are located proximally to the transverse plate and situated at a distance of about 60 nm from each other (Figure 4G,H,M).

We observed all six flagellar roots known for Discocelia. Root b (=vp1 = R2) contains two microtubules. It originates from the posterior side of the base of the PB and initially runs parallel to the PB; then bends posteriorly toward the nucleus, curving along its ventral side while turning to the right (Figure 4E–K). Root c1 (=vp2) contains one microtubule and originates from the posterior side of the base of the AB. It runs backward, initially in parallel with root b (=vp1 = R2), but then curves to the left, crossing root b (=vp1 = R2) (Figure 4E–K). Between the roots b (=vp1 = R2) and c1 (=vp2), near both basal bodies, the fr fiber can be observed (Figure 4E,J). We also observed a small sheet of electron‐dense material closely associated with the PB, which we call fr' (Figure 4E,M). Root c2 (=dp1 = R1) emerges from the fr fiber. It contains two microtubules and is directed to the left, almost immediately crossing root b (=vp1 = R2) (Figure 4F,G,K). Root a (=dp2) contains two microtubules and arises from the right side of the base of the AB. It is directed backward and to the right. We did not observe fs fiber or any other fibers associated with root a (=dp2). The two marginal microtubules (=lr = R4) arise in the space between the basal bodies, initially in parallel with the AB and almost immediately bending and running along the periphery of the cell (Figure 4F,G,I,M). Marginal microtubules can often be seen in association with extrusomes (Figure 4B). We also observed a second pair of microtubules (designated as lr') at the posterior end of the cell body, situated dorsally to the marginal microtubules (Figure 4B). The origin of lr' was not identified. The velum microtubule (=da = R3) is often visible at the anterior of the cell, in parallel with marginal microtubules (=lr = R4) but we were unable to trace its exact origin (Figure 4B,I).

4. Discussion

4.1. External Morphology of Known Discocelia Species

Here we compare the three known Discocelia species (Table 1). Cells of D. punctata are larger and have an apparent row of marginal extrusomes (Larsen and Patterson 1990). D. saleuta has less distinct, delicate punctae along the cell margin (Larsen and Patterson 1990; Lee 2015), which are not visible in LM in D. plataet (Figure 2A–J). However, electron microscopy observations of D. plataet (Figures 2M, 4B,E) and D. saleuta (Vørs 1988) confirm that the row of marginal extrusomes is a defining characteristic of the genus Discocelia.

One of the main morphological characteristics of D. saleuta is the velum with a pointed tip on the anterior left side of the cell and a pointed cell body tip on the anterior right side of the cell (Vørs 1988). The presence of these structures in D. saleuta and their absence in D. plataet is explained by differences in the location of the microtubules of the flagellar apparatus of these species (see below). Both, the velum tip and the cell body tip, are indistinct on some D. saleuta cells (Vørs 1988: Figures1–6), and some authors illustrate D. saleuta cells without them (Aydin and Lee 2012; Ekebom et al. 1995/6; Lee 2006, 2015), which makes this species morphologically similar to our new species. Therefore, careful observations of all possible morphological variations of cells in clonal cultures are necessary for confident species identification. Likewise, all known findings of D. saleuta without indications of cell body tip and velum tip cannot be confidently recognized as this species (Aydin and Lee 2012; Ekebom et al. 1995/6; Larsen and Patterson 1990; Lee 2006, 2015; Tong 1997; Vørs 1992).

The other differences of known Discocelia species concern the length and the movement of the AF (Table 1). In the original description of D. saleuta, the AF is 0.5–1.5 μm long, directed forward and beating rapidly while the cell is gliding (Vørs 1988). The AF of D. punctata is even longer (up to 3 μm), directed anteriorly and to the left, and relatively inactive during the cell movement (Larsen and Patterson 1990). The AF of D. plataet is very short (0.46 μm), poorly visible in LM, and motionless. Thus, most known findings of D. saleuta in the literature with poorly visible AF may actually refer to D. plataet (Aydin and Lee 2012; Ekebom et al. 1995/6; Larsen and Patterson 1990; Lee 2006, 2015; Tong 1997; Vørs 1992). D. plataet also differs by the presence of an acronema on the PF, which was not described in both known Discocelia species, although it might be present in some cells of D. saleuta according to illustrations (Vørs 1988: Figures 7, 9, 10, 13). While both D. saleuta and D. plataet have an acronema on the AF, we do not have any information about it for D. punctata (Larsen and Patterson 1990).

The prey capture in our new species is probably occurring at any part of the cell body surface, as we detect this process at the posterior part of the cell (Figure 1F–H; Video S4, a small, round bacterium was attached to the edge of the D. plataet cell and engulfed by a pseudopodium). Therefore, the velum is not involved in food uptake, as it was proposed before (Vørs 1988), but marginal extrusomes most probably are.

Although many authors expressed doubts about the correctness of the separation of the species D. punctata from D. saleuta because of cell size overlap and the presence of delicate punctae on some D. saleuta cells (Lee 2002, 2006, 2015; Lee and Patterson 2000), we assume that these species differ significantly. D. punctata is characterized by a longer and motionless AF, the absence of the velum tip and the cell body tip, clearly visible marginal extrusomes, and a bigger cell size on average (Larsen and Patterson 1990). On the other hand, D. saleuta is characterized by a visible and actively beating AF, the presence of the pointed velum tip and the cell body tip, both directed anteriorly, the absence of the acronema on the PF in some cells, and marginal extrusomes not visible in LM (Vørs 1992). The new species, D. plataet, has clear morphological differences, distinguishing it from the two other known species, including the acronema on the PF, a short and motionless AF, poorly visible in LM, the absence of pointed protrusions on the velum and the cell body, invisible marginal extrusomes in LM, and a smaller cell size on average.

4.2. Discocelia General Ultrastructure and Cell Body Plan

In general, the cell structure and the arrangement of organelles in Discocelia correspond to those of other small cercozoans gliding on their PF. Discocelia has a determined placement of most organelles in close association with the nucleus: basal bodies and flagellar roots, Golgi bodies, mitochondria, and microbody. We observed in D. plataet a slightly longitudinally elongated nucleus (Figures 3, S2–, S6) that becomes more circular after cell division (Figure S7). The nucleus of D. saleuta is round and shifted to the right side of the cell (Vørs 1988). Its size was not indicated (Vørs 1988), but judging by the illustrations (despite the absence of scale bars), the dimensions of the nucleus relative to the cell body size are approximately similar in both species.

A microbody‐like organelle (Figures 2I,Q, 4A,B,L, S3–, S5) is present in both species (Vørs 1992). We identified it by its location in the cell (posterior and slightly dorsal to the nucleus) and approximate dimensions (about 0.5 μm). We assume that it is a microbody, as microbodies in a similar position have been found in cercomonads, paracercomonads, eocercomonads, metromonads, sainouroids, glissomonads, and cryomonads (Cavalier‐Smith and Karpov 2012; Cavalier‐Smith and Oates 2012; Cavalier‐Smith et al. 2008, 2009; Karpov et al. 2006; Shiratori et al. 2020). Interestingly, we observed small (150–200 nm in diameter) oval electron‐dense bodies associated with mitochondria in D. plataet (Figure 4D), which have not been described in D. saleuta.

We observed two mitochondria in close association to the nucleus. Vørs (1988) also pointed out several mitochondria in the cells. The relative sizes of mitochondria in D. saleuta (according to illustrations), as well as their position in the cell, correspond to those we observed in D. plataet. Vørs (1988) described short tubular, and “often more or less bleb‐like” cristae. Similar tubular ampulliform cristae have also been described in glissomonads (Cavalier‐Smith and Oates 2012). Unexpectedly, we found discoidal mitochondrial cristae in D. plataet, while most cercozoans have tubular cristae, rarely flattened or vesicular (Adl et al. 2019). The most similar type of cristae, plate‐like (flattened), is present in Sainouron, Helkesimastix, Limnofila, and Kraken (Cavalier‐Smith et al. 2008, 2009; Dumack et al. 2017).

Rounded ‘empty’ vacuoles in the posterior half of the cell are visible by NHS staining in U‐ExM (Figures 3, S2–, S7) and in TEM sections (Figure 4B). These vacuoles are probably contributing to the sagging of the membrane during fixation for SEM, which leads to the formation of the observed dorsal depressions. Such vacuoles were not described by Vørs, nor by other authors.

The basal bodies of D. plataet are orthogonal to each other (Figure 4) in contrast with the acute angle observed in D. saleuta (Vørs 1988). Moreover, the proximal end of the PB of D. plataet is directed to the side of the AB, as described in many cercozoans (Cavalier‐Smith and Karpov 2012; Karpov et al. 2006; Hess and Melkonian 2014; Shiratori et al. 2020). Although the structure of the flagellar transition zone of D. saleuta was not well described in TEM (Vørs 1988), it seems to be very similar to what we observed in D. plataet—an indistinct transitional plate and two central discs beneath it (Figure 31 in Vørs 1988). These structures probably represent the proximal hub‐lattice and distal hub‐spoke structures described by Cavalier‐Smith et al. (2008, 2009). No additional structures such as the peripheral collar of coupling, as in some other cercozoans (see below), were found.

4.3. Discocelia Flagellar Root Homologies and Comparison

Combination of our U‐ExM data with both TEM ultrastructural studies (Vørs 1992 and present study) allowed us to analyze the Discocelia cytoskeleton and propose flagellar root homologies among the studied cercozoans and other eukaryotes. A schematic illustration of the flagellar apparatus is presented in Figure 5. We are adapting the terminology used for cercomonads and paracercomonads and their homology among roots of other eukaryotes (Cavalier‐Smith and Karpov 2012; Karpov et al. 2006). In her ultrastructural study, Vørs (1988) did not consider the marginal microtubules and the velum microtubules as part of the flagellar apparatus. However, following Cavalier‐Smith (2013), we consider them as flagellar roots, since they originate from the basal bodies and have homologies among the roots of studied cercozoans.

FIGURE 5.

Scheme of the flagellar apparatus of Discocelia plataet sp. n. viewed from the ventral side. For abbreviations see Figures 3 and 4.

Root b originates from the base of the PB, directed posteriorly and passing on the ventral side of the cell. Therefore, we propose its homology with the posterior ventral root vp1 of cercozoans, or root R2 of eukaryotes. It is important to note here the difference in the number of microtubules between the two Discocelia species: two in D. plataet and four in D. saleuta (Vørs 1988). The posterior ventral root c1 originating from the base of the AB and passing along the vp1 = b root clearly corresponds to the cercozoan root vp2 and has no homology among other eukaryotes.

The posterior ventral roots vp1 and vp2 are widely present in cercozoans, where they are usually directed posteriorly along the PF and support the sides of the ventral groove or the flagellar pit (Cavalier‐Smith and Karpov 2012; Karpov 2010/11; Karpov et al. 2006; Shiratori et al. 2014). Discocelia lacks a ventral groove, and the cell body is wide and strongly dorsoventrally flattened. This is probably a reason why both posterior ventral roots are directed sideways instead of following the cell posterior. The root vp1 of glissomonads crosses the root vp2 as in Discocelia, but has a fan‐like structure (Cavalier‐Smith and Oates 2012). The posterior ventral roots in Cercozoa usually have more microtubules in their proximal part, and the number of microtubules usually increases even more distally (see Table 2 in Cavalier‐Smith and Karpov 2012), which can be explained by the larger cell size of most cercozoans compared to the minuscule Discocelia (2.5–6 μm). An exception is Sainouron and Helkesimastix, whose microtubular roots are generally strongly reduced (Cavalier‐Smith et al. 2008, 2009). In contrast, the flagellar apparatus of Discocelia, despite its small size, is very well developed, with all the known cercozoan roots preserved.

Additional confirmation of root vp1 and root vp2 homology is that in its proximal part, root vp2 = c1 is associated with a dense fiber (‘dense material’ according to Vørs (1988)). We consider this fiber to be a homologue of the fr fiber of cercomonads and paracercomonads (Cavalier‐Smith and Karpov 2012), since it passes close to both ventral posterior roots and the PB, although it is not closely connected to the PB. As in Paracercomonas and Neocercomonas, this dense material nucleates secondary individual microtubules (Vørs 1988), which fan out posteriorly, and the third posterior root dp1 = c2. If truly homologous, the fr fiber in D. saleuta is shifted distally along the vp1/vp2 on a distance of one basal body width (see Figures 42 and 43 in Vørs 1988). However, in D. plataet we observe a closer location of the fr fiber to both basal bodies (Figure 4E,J). We assume that this feature may differ in both Discocelia species, as it is the case within paracercomonads: in P. virgaria, the fr fiber is located between the two basal bodies, whereas in P. metabolica it is connected only to the PB (Cavalier‐Smith and Karpov 2012). In cercomonads, this fiber extends from the AB (Karpov et al. 2006) and divides into two branches, one of which extends to the nucleus, and the other, as in Discocelia, goes along root vp2. In Discocelia, this fiber is connected to both basal bodies, the root vp2, and the nucleus. It is noteworthy that in paracercomonads this root is also associated with the nucleus in at least one species (Cavalier‐Smith and Karpov 2012).

We consider the root c2 as homologous to the root dp1 of paracercomonads, or R1 in eukaryotes. In P. virgaria, it arises from the proximal end of the PB, but is also closely associated with the fr fiber and the stem vp1/vp2 (Cavalier‐Smith and Karpov 2012). It is noteworthy that in P. metabolica this root is also slightly shifted distally from the base of the PB. In both studied paracercomonad species, the root dp1 is also located at a right or almost right angle to the vp1/vp2 stem, but is directed toward the posterior end of the cell body or slightly to the right. This displacement is explained by the very wide angle between the basal bodies in paracercomonads, in contrast with the acute/right angle in Discocelia species. The root dp1 is a posterior dorsal root, directed posteriorly along the dorsal side of the body. We also observed a more dorsal location of this root, although the difference between the dorsal and ventral position is not so clear in the extremely flattened Discocelia cells. In cercomonads, thaumatomonads, sainouroids, and glissomonads, the root dp1 is absent, and Cavalier‐Smith and Karpov (2012) suggested that the root dp1 serves in paracercomonads to support the highly mobile anterior end of the body when stretching forward. In Discocelia, this root could have been retained from the last common ancestor of Discocelia and paracercomonads to enhance the support of the cell body, since very wide cells gliding on the PF require a strong cytoskeleton. The root ar in Katabia (glissomonad), a non‐gliding cercozoan flagellate with uncertain position, is probably a homologue of dp1, as most elements of the flagellar apparatus of Paracercomonas and Katabia are unexpectedly similar (Cavalier‐Smith and Karpov 2012).

We interpret the prominent fiber at the right of the AB, visible with U‐ExM and also reported by Vørs (1988), as a homologue of the fs fiber of paracercomonads (Cavalier‐Smith and Karpov 2012). This is additionally supported by secondary microtubules nucleating from this fiber, as well as the backward‐directed microtubular root a, which we consider to be a homologue of the cercozoan root dp2. In addition, no other fibrous structures in a similar position are known in Monadofilosa. However, we could not determine the fs fiber in our TEM thin sections. As described by Vørs (1988), this fiber is not directly connected to the AB, but is located at some distance from it and is connected to it by some fibrous threads, which we did not detect due to insufficient resolution of the U‐ExM. Considering that Discocelia is quite distant from the known monadofilosans, this may simply be a characteristic feature of this lineage.

Apparently, both fs and fr fibers give rise to a fan of individual cytoplasmic microtubules in Discocelia, as it was observed in Paracercomonas, in contrast with Cercomonas and Neocercomonas, where only the fs or fr fiber, respectively, gives rise to individual cytoplasmic microtubules. However, these microtubules may not always be easy to notice, as, for example, we did not find a fan of microtubules radiating from the fr fiber, while Vørs (1988) did not observe secondary microtubules radiating from the fs fiber. Cavalier‐Smith and Karpov (2012) considered the diverging secondary microtubules from both fibers in Paracercomonas as an ancestral state. These authors also proposed that fibers fs and fr are serially homologous and fs develops into fr in each cell cycle during centriolar transformation.

We consider the root a as a homologue of the cercozoan root dp2, since it extends from the fs fiber toward the posterior end of the cell. According to Vørs (1988), this root has three microtubules at its base, but one of them is shorter and ends in the middle part of the root. Using U‐ExM, we also observed one short microtubule that radiates aside from the root and is directed a little more posteriorly. It could be the same short microtubule, or one of the individual cytoplasmic microtubules that extend from fiber fs. According to our observations, the root dp2 = a is directed to the right along the marginal edge of the cell. Vørs (1988) described this root as more inclined toward the longitudinal axis of the cell along the ventral side of the nucleus. Considering that root dp2 = a locates close to the edge of the cell, it is unclear whether it is dorsal or ventral in D. plataet.

Paracercomonads do not have a root dp2, which was originally found in Eocercomonas (Karpov et al. 2006). However, the authors pointed out that two of the single microtubules radiating from fiber fs in P. metabolica run parallel with each other in the posterior direction, as do the two microtubules of root dp2 of cercomonads (Cavalier‐Smith and Karpov 2012). In P. metabolica, these microtubules also run ventrally and to the left, as in Discocelia. In cercomonads, dp2 runs toward the vp1/vp2 stem (Cavalier‐Smith and Karpov 2012), which may provide additional support for the ventral groove. Another confirmation of the homology of the roots a and dp2 is that root dp2 in cercomonads, as well as two parallel microtubules in paracercomonads, are antiparallel (i.e., directed oppositely along the longitudinal line) to the AB and the velum microtubule (which we consider to be a homologue of the root da). Shiratori et al. (2014) showed the presence of the root dp2 in Abollifer (Imbricatea). It consists of six microtubules and arises on the left side of the AB, slightly above the origin of da, and runs along the left side of the flagellar notch. The origin and direction of this root differ from those in cercomonads and Discocelia, and, if the homology is correctly determined, may result from the apical position of the basal bodies in Abollifer.

According to Vørs (1988) and our TEM observations, two marginal microtubules originate from the space between the basal bodies (Figure 4G,I,M). Accordingly, we assume that they represent the cercozoan root lr or root R4 of eukaryotes. It extends along the entire periphery of the cell and terminates at the right anterior part of the cell at the body tip in D. saleuta (Vørs 1988). D. plataet differs from D. saleuta in that it lacks a body tip, as its marginal microtubules instead encircle the cell body along the marginal edges, turning left and overlapping with the proximal end of these microtubules. Sometimes we observed a second pair of marginal microtubules (lr'), located dorsally at some distance from the main ones (lr) (Figures 3A–E, 4B). Although we could not determine its origin using either U‐ExM or TEM, and Vørs (1988) did not observe it at all, we speculate that lr’ may be a continuation of lr, which makes a full circle around the periphery of the cell and overlaps with itself.

Root lr is widely present in Cercozoa, where it is also connected longitudinally with the fibrous material and usually runs along root da (Cavalier‐Smith and Karpov 2012; Karpov et al. 2006; Karpov 2010/11; Shiratori et al. 2014, 2020; Cavalier‐Smith et al. 2008, 2009). In Cercomonas, Neocercomonas, and Katabia (=root ur), as in Discocelia, it consists of two microtubules and gives rise to secondary single microtubules. In Paracercomonas and Eocercomonas, this root was not always observed, consists of only one microtubule, and does not give rise to secondary microtubules (Cavalier‐Smith and Karpov 2012).

The microtubule(s) supporting the anterior‐dorsal side of the velum of Discocelia is therefore most likely homologous to the cercozoan root da, or R3 root of eukaryotes (according to Cavalier‐Smith and Karpov 2012; Karpov et al. 2006). In cercomonads, root da originates from the right side of the AB, while in paracercomonads it originates on the ventral side of the AB (Cavalier‐Smith and Karpov 2012). In Discocelia, if the homology is correct, this root is shifted to the left side and, thus, closer to the arrangement that was found in Abollifer (Shiratori et al. 2014). However, Vørs (1988) indicated that the proximal ends of the velum microtubules and, apparently, marginal microtubules are associated with connective fibers. In general, Vørs' fixation was suboptimal; we could not determine the origin of this root on thin sections, and the resolution of U‐ExM is insufficient to be sure of the correctness of the root homology determination. It is possible that the velum microtubule(s) start on the ventral side or even closer to the right side of the AB. Otherwise, such a shift in the place of origin of root da can be explained by an extremely flattened cell body that needs to support the left side of the cell.

In general, the U‐ExM methods allowed us to study all expected elements of the flagellar apparatus of D. plataet and made our subsequent TEM ultrastructural analysis easier to interpret. With U‐ExM, we were able to observe some additional ultrastructural details that were not found before, such as the proximal orientation of roots vp1 and vp2, cytoplasmic microtubules originating from fiber fs, the separation of marginal microtubules, as well as characteristic details of a new species, such as the proximal positioning of marginal and velum microtubules, which explain the absence of a pointed velum and cell body tip in D. plataet. Using TEM, we were able to follow the finer details of the cell structure of D. plataet and find significant differences from D. saleuta. These include a different type of mitochondrial cristae (discoidal), a different orientation of the basal bodies relative to each other (orthogonal), the presence of an elongated subcentral nucleolus, the presence of only two microtubules in the root vp1, the presence of an additional electron‐dense sheet (fr') associated with the base of the PB, and the presence of a second, more dorsally located pair of marginal microtubules (lr'). All these significant differences indicate that the two species studied ultrastructurally, D. saleuta and D. plataet, are probably distantly related despite their superficial similarity. Re‐isolation of D. saleuta and molecular phylogenetic analysis will help to test this assumption.

4.4. Evolutionary Context of Discocelia

Recent mentions of the taxonomy of Discocelidia provided no clear support or explanation, although we assume the authors had unpublished molecular evidence to claim that Discocelidia should be part of Imbricatea (Cavalier‐Smith et al. 2018) or simply an incertae sedis cercozoan (Adl et al. 2019). Based on our morphological and molecular data, Discocelia plataet sp. n. should still be considered an orphan lineage within Cercozoa = Filosa, likely within Monadofilosa. Although we reconstructed a phylogenomic tree with high statistical support, including it and the two newly sequenced Marimonad CRO19P5 and Glissomonad RAM19S6, it did not have enough taxon sampling to properly solve the internal relationships of the group. The phylogenetic relationship between the soil social amoeba Guttulinopsis (Brown et al. 2012) and the marine benthic Discocelia is likely to change in the future when additional taxa are included. Both taxa have higher evolutionary rates compared to other rhizarians, and together with more taxa, they will probably split into different clades. Abollifer (Marimonad) and eight other transcriptomes (Irwin et al. 2019; Cavalier‐Smith et al. 2018) were not included in our analyses due to the poor quality of the data. From a morphological point of view, Discocelia is characterized by a set of unique characters that, on the one hand, explain the separate phylogenetic position of this lineage, but on the other hand, complicate the comparison with known monadofilosan groups. Identifying the phylogenetic position of Discocelia within Monadofilosa in the future will be important for understanding the evolution and early radiation of this group.

The flagellar apparatus of cercomonads and paracercomonads is the most thoroughly studied among Cercozoa (Cavalier‐Smith and Karpov 2012; Karpov et al. 2006). Both groups are superficially similar by having a strongly amoeboid flattened body with a cytoplasmic contractile ‘tail’ temporarily adhering to the PF. Molecular phylogeny shows that these groups are likely paraphyletic (Bass et al. 2009; Cavalier‐Smith and Chao 2003; Howe et al. 2011), which allowed speculation about the conservatism of the flagellar apparatus features that these groups share (Cavalier‐Smith and Karpov 2012). Despite their very small cell size, Discocelia species also have a conserved cytoskeleton, with all six microtubular roots (vp1, vp2, dp1, dp2, lr, da) preserved from their common ancestor with cercomonads/paracercomonads (Cavalier‐Smith and Karpov 2012). A significant difference between discocelids and cercomonads/paracercomonads is the absence of a cytoplasmic tail adhering to the PF, which is maintained by the posterior ventral roots vp1 and vp2. The absence of such a tail in Discocelia is probably related to the absence of the vp1/vp2 association in the distal parts of these roots. Other differences include the position of the basal bodies relative to each other. In cercomonads and paracercomonads, basal bodies are located at a very wide angle with each other, and the proximal end of the PB is directed to the side of the AB. This affects the location of all roots extending from the basal bodies, as described in more detail in the previous section. Cercomonads and paracercomonads are similar to Discocelia in having an indistinct and short flagellar transition zone, unlike most other cercozoans, which possess structures such as peripheral dense collars or cylinders (Hess and Melkonian 2014).

According to our phylogenomic analysis, Discocelia is distantly related to the soil social amoeba Guttulinopsis. It is difficult to compare non‐flagellated species ultrastructurally, but Guttulinopsis is closely related to the heterotrophic flagellated sainouroid genera Helkesimastix, Sainouron, and Cholamonas based on SSU rRNA gene trees (Cavalier‐Smith et al. 2008, 2009; Bass et al. 2016). Like Discocelia, sainouroids have very small cells with a fixed position of most organelles and a relatively simple and short flagellar transition zone. However, they have a greatly reduced cytoskeleton (loss of roots dp1, dp2, and da), and their AF is reduced to an immobile stub in Helkesimastix or a papilla in Sainouron and Cholamonas.

Glissomonada is another monadofilosan subgroup of small gliding heterotrophic flagellates. They are similar to Discocelia and sainouroids in having a small semi‐rigid cell body, a narrow angle between the basal bodies, and the absence of a ventral groove (Cavalier‐Smith and Oates 2012; Cavalier‐Smith et al. 2008, 2009). Glissomonads have only four flagellar roots (lr, da, vp1, and vp2) and their ventral posterior root microtubules are numerous and mostly located in the form of a cone. The flagellar transition zone is longer than that of Discocelia, cercomonads, paracercomonads, and sainouronids, and generally differs from them in the presence of a thick dense peripheral collar and a distal dense collar‐plate structure.

The amoeboflagellated Viridiraptoridea (Viridiraptor, Orciraptor), like Discocelia, sainouroids, and almost all glissomonads, have a microbody, Golgi bodies and mitochondria closely associated with the nucleus (Hess and Melkonian 2014). Viridiraptor has round extrusomes along the entire cell periphery of similar shape to those found in Discocelia. However, the flagellar transition zone of Viridiraptor is more similar with the one of glissomonads, with a conspicuous ‘dense distal plate/collar complex’. Their flagellar apparatus of viridiraptorids contains two additional probasal bodies and two large rhizoplasts. In addition, they are a highly specialized freshwater algivorous group that occupies distinct ecological niches (Hess and Melkonian 2014).

Just a few known representatives of Pansomonadida (Agitata, Aurigamonas) have been poorly studied in terms of the structure of the flagellar apparatus (Vickerman et al. 2005). They are heterotrophic flagellates with an alternated sedentary amoeboid life stage (Adl et al. 2019). Aurigamonas differs from Discocelia in having numerous unique radiating haptopodia, used for active hunting, and mastigonemes on both flagella (Vickerman et al. 2005).

Spongomonads (Spongomonas, Rhipidodendron) are sessile biflagellated protists. Most known representatives form tree‐like colonies immersed in a matrix of dense non‐organic spherules (Hibberd 1976; Struder‐Kypke and Hausmann 1998). They differ significantly from many cercozoans in having a reduced flagellar apparatus, the absence of the vp1/vp2 association, the presence of characteristic fibrillar caps at the bases of the basal bodies, and an exceptionally long flagellar transition zone. The probably related gliding flagellated thaumatomonads have a deep flagellar pocket and ventral groove, from which branching filose pseudopodia emerge (Karpov, 2010/11). Most thaumatomonads are covered with species‐specific siliceous scales. They also have a minimal set of microtubular roots compared to discocelids, cercomonads, and paracercomonads. Thaumatomonads also differ from discocelids in the nearly parallel arrangement of the basal bodies (Hibberd 1976; Karpov 2010/11; Struder‐Kypke and Hausmann 1998).

Cryomonads, like thaumatomonads, have a ventral groove, from which branching filopodia emerge. They are also characterized by a single‐ or multilayered extracellular organic theca covering almost all the cell surface, except the ventral groove and flagellar pits (Adl et al. 2019). In contrast, Discocelia has a para‐crystalline striated layer underneath the plasmalemma, only on the dorsal cell surface (Vørs 1988). Cryomonads have a complex structure of the flagellar transition region with a transitional cylinder that consists of a stack of 2–7 subrings (Cryothecomonas), or 4–5 dense transitional distal plates, and an electron‐dense disk sandwiched between less dense proximal plates (Ventrifissura) (Drebes et al. 1996; Shiratori et al. 2020). Flagellated cryomonads have a simple flagellar apparatus with a single microtubular root (vp1) preserved in Protaspa and Cryothecomonas (Drebes et al. 1996; Schnepf and Kühn 2000), and three (vp1, lr and ar) in Ventrifissura (Shiratori et al. 2020).

Non‐amoeboid obligate eukaryotrophs, like the marine flagellated metromonads (Imbricatea), differ greatly from Discocelia in having ‘coupling’ in the transition zone and at least three visible distal plates (Mylnikov et al. 2020). Furthermore, metromonads are characterized by the parallel arrangement of basal bodies, mastigonemes on both flagella, and rod‐shaped extrusomes with a complex wheel‐shaped structure (Mylnikov et al. 2020). Similar to cryomonads, metromonads have a thick fibrous layer above the plasmalemma that covers almost the entire cell surface.

In general, Discocelia species have a combination of features more or less similar to different cercozoan groups. Their flagellar apparatus likely represents an ancestral state due to the preservation of all microtubular roots and fibers known in Cercozoa, and the association of vp1 and vp2. At the same time, Discocelia has unique features, including extremely flattened cells and associated reorganizations in the structure of the flagellar apparatus, a thin para‐crystalline layer under the plasmalemma covering only the dorsal side of the cell, the presence of marginal microtubules, sometimes forming a cell body tip (in D. saleuta), and a dorsal anterior left velum supported by microtubules. Also, D. plataet has unique discoidal mitochondrial cristae. However, it should be taken into account that the ultrastructural studies of cercozoans are still fragmentary and completely absent for some groups, not even mentioning the undescribed diversity. The more diversity of heterotrophic flagellates gets studied, the more chances we will have to understand the underlying evolutionary forces that shaped this part of the eukaryotic tree of life.

4.5. Taxonomy Review

Rhizaria: Phylum Cercozoa Cavalier‐Smith 1997: Subphylum Monadofilosa incertae sedis.

Order Discocelida Cavalier‐Smith 1997.

Family Discoceliidae Cavalier‐Smith 2013.

Discocelia (Vørs 1992) Cavalier‐Smith 2013.

Discocelia plataet sp. n. Prokina, Torruella et al., sp. n.

Description: Cell is dorsoventrally flattened, circular in outline, 3.1 μm long, 3.6 μm wide. Discoidal mitochondrial cristae. Basal bodies are orthogonally oriented with each other; the proximal end of the posterior basal body is directed to the side of the anterior basal body. The posterior basal body has an additional sheet‐like fiber. The velum is rounded, without a pointed tip. Marginal microtubules do not protrude into the cell anterior region and, consequently, do not form a sharp body tip. The ventral posterior root vp1 contains two microtubules. The anterior flagellum is barely visible on LM, 0.46 μm long in fixed cells, directed posteriorly and to the left. The posterior flagellum is 6 μm long. Both flagella have acronema. The row of extrusomes along the entire cell margin is not visible in LM. Cysts are absent.

Type figure: Figure 2A illustrates a live cell of strain GT001.

Type culture: Strain GT001 is deposited in the DEEM culture collection, CNRS, and Université Paris‐Saclay (France).

Type material: A block of chemically fixed resin‐embedded cells of the type strain GT001 is deposited in the DEEM collection of type materials, CNRS and Université Paris‐Saclay (France). This constitutes the name bearing type of the new species (a hapantotype).

Type locality: Marine shore of Roc'h ar Bleïz, Roscoff, France.

Etymology: Plataet means flattened in Breton.

Gene sequence: SSU rRNA gene sequence of strain GT001 has the GenBank accession number PQ362545.

Zoobank Registration for this publication: urn:lsid:zoobank.org:pub:CD249290‐26A4‐4AF6‐A521‐3EF83A536CC5.

Zoobank Registration for genus Discocelia: urn:lsid:zoobank.org:act:6E8FCEE2‐70A4‐4FBD‐B0A8‐17F2BE21ADBC.

Zoobank Registration for Discocelia plataet sp. n.: urn:lsid:zoobank.org:act:99C38B48‐E421‐4044‐B1FE‐14397D27755D.

Supporting information

Figure S1: Phylogenetic tree generated from Bayesian analysis of SSU rRNA gene sequences. The tree contains 111 rhizarian sequences, including 5 outgroup sequences (Radiolaria, Retaria). Bayesian posterior probabilities (BPP, GTR + I + GAMMA4) and ML (TIM2 + F + I + R6 model) bootstrap values are indicated at branches; black dots indicate values of BPP = 1.00 and ML bootstrap = 100%; dt: different topology between trees. Sequences generated in this study are shown in bold red.

Figure S2: U‐ExM of Discocelia plataet sp. n. cell #1, ventral view. Black and white pictures represent an inverted image with merged channels. Cyan represents Hoechst (DNA staining), yellow represents NHS ester (protein density), magenta represents α‐ and β‐tubulin (microtubules). Abbreviations: ac—acronema; af—anterior flagellum; da—dorsal anterior root of AB (=R3 of eukaryotes; =velum microtubule in D. saleuta); dp1—dorsal posterior root of PB (=R1 root in eukaryotes; =c2 root in D. saleuta); dp2—dorsal posterior root of AB (=c2 root in D. saleuta); fr—fiber associated with PB; fs—fiber associated with AB; fv—food vacuole; lr—left anterior root of AB (=R4 root in eukaryotes; =marginal microtubules in D. saleuta); m—mitochondrion; pf—posterior flagellum; sm—secondary microtubules; v—vacuole; vp1—ventral posterior root of PB (=R2 root in eukaryotes; =b root in D. saleuta); vp2—ventral posterior root of AB (=c1 root in D. saleuta). The stack number is indicated on the top left corner. Scale bars: 10 μm. Expansion factor 5.

Figure S3: U‐ExM of Discocelia plataet sp. n. cell #2, ventral view. Black and white pictures represent an inverted image with merged channels. Cyan represents Hoechst (DNA staining), yellow represents NHS ester (protein density), magenta represents α‐ and β‐tubulin (microtubules). Abbreviations: mb—microbody, other abbreviations as in Figure S2. The stack number is indicated on the top left corner. Scale bars: 10 μm. Expansion factor 5.

Figure S4: U‐ExM of Discocelia plataet sp. n. cell #3, ventral view. Black and white pictures represent an inverted image with merged channels. Cyan represents Hoechst (DNA staining), yellow represents NHS ester (protein density), magenta represents α‐ and β‐tubulin (microtubules). Abbreviations: as in Figures S2, S3. The stack number is indicated on the top left corner. Scale bars: 10 μm. Expansion factor 5.

Figure S5: U‐ExM of Discocelia plataet sp. n. cell #6, ventral view. Black and white pictures represent an inverted image with merged channels. Cyan represents Hoechst (DNA staining), yellow represents NHS ester (protein density), magenta represents α‐ and β‐tubulin (microtubules). Abbreviations: as in Figures S2, S3. The stack number is indicated on the top left corner. Scale bars: 10 μm. Expansion factor 5.

Figure S6: U‐ExM of Discocelia plataet sp. n. cell #7, ventral view. Black and white pictures represent an inverted image with merged channels. Cyan represents Hoechst (DNA staining), yellow represents NHS ester (protein density), magenta represents α‐ and β‐tubulin (microtubules). Abbreviations: as in Figures S2, S3. The stack number is indicated on the top left corner. Scale bars: 10 μm. Expansion factor 5.

Figure S7: U‐ExM of Discocelia plataet sp. n. cells #4 and #5 (dividing cell), ventral view. Black and white pictures represent an inverted image with merged channels. Cyan represents Hoechst (DNA staining), yellow represents NHS ester (protein density), magenta represents α‐ and β‐tubulin (microtubules). Abbreviations: as in Figures S2, S3. The stack number is indicated on the top left corner. Scale bars: 10 μm. Expansion factor 5.

Table S1: Annotation of the ribosomal operon sequences for the three novel strains obtained with Barrnap.

Table S2: Transcriptome completeness values based on BUSCO with eukaryota_odb10, 255 markers.

Table S3: Concatenated phylogenomics dataset of 16 taxa and 169 protein markers.

Video S1: Live cell of Discocelia plataet sp. n. in DIC contrast.

Video S2: Live cell of Discocelia plataet sp. n. in phase contrast.

Video S3: Live cell of Discocelia plataet sp. n. with extending filamentous pseudopodium.

Video S4: Live cell of Discocelia plataet sp. n. capturing a bacterium.

Prokina, K. , Torruella G., Galindo L. J., Dudin O., López‐García P., and Moreira D.. 2025. “ Discocelia Plataet Sp. n., a Small Incertae Sedis Cercozoan Flagellate.” Journal of Eukaryotic Microbiology 72, no. 5: e70041. 10.1111/jeu.70041.

Data Availability Statement

The data that support the findings of this study are openly available in figshare at https://figshare.com/articles/dataset/Phylotranscriptomics_for_3_monadofilosa/27074848?file=49365286, reference number https://doi.org/10.6084/m9.figshare.27074848.