Sampling of E. diaphana was performed in a ditch (25°09′04.5′′N, 121°46′41.1′′ E) near the Aquatic Animal Center at National Taiwan Ocean University (NTOU), northern Taiwan ( Figures 1A-C ). Four large individuals (5~7 cm in height) were collected with a spoon from the ditch wall, and were transported to our laboratory at NTOU. The four sea anemones were rinsed several times with seawater, and cultured separately in four 40-L aquaria under T5 lights (6500k; light intensity, 20-40 μmol/m²/s, 12-hr light/dark cycle). Each aquarium was equipped with an external filtration system, and the water temperature was set at 28°C. Artemia naupli were provided as food, five times a week, and one-third of the seawater in the tanks was replaced every week. We then established two clonal strains (one male and one female) by allowing them to reproduce asexually. Developing testes/ovaries were examined in dissected polyps under a stereo microscope (EZ4, Leica Microsystems, Mannheim, Germany). To observe development of lacerated pieces (pedal fragment) into small anemones, collected pieces were cultured in 6-well plates containing 5 mL of 0.22 μm-filtered seawater per well. Micrographs were taken with a digital camera (RX100II, Sony, Tokyo, Japan) under the stereo microscope. All experiments were carried out in accordance with principles and procedures approved by the Institutional Animal Care and Use Committee, NTOU.

Three individuals were sampled from each male and female strain and homogenized with a homogenizer (IKA Ultra-Turrax, Sigma-Aldrich, St. Louis, USA) in lysis solution supplied with Wizard Genomic DNA Purification Kit (Promega, Madison, WI). Genomic DNA was extracted according to the manufacturer’s protocol. To compare phylogenetic relationships of the established E. diaphana strains and existing strains, a set of four E. diaphana-specific inter-simple sequence repeat (ISSR)-derived Sequence Characterized Amplified Region (SCAR) markers were used (Thornhill et al., 2013). PCR was performed with specific primers for each of the four SCAR markers ( Table 1 ). PCR products were resolved by electrophoresis, subcloned, and sequenced as previously described (Shikina et al., 2012). Sequences of SCAR markers of existing E. diaphana strains were retrieved from GenBank, and phylogenetically compared with those of E. diaphana strains established in this study. Sequence alignment was performed using Muscle software, and results were used to create a 10,000 bootstrap replicate phylogenetic tree (neighbor-joining method) using MEGA7 (Kumar et al., 2016).

Specimens of E. diaphana having different pedal disc-sizes (1-2 mm, 3-4 mm, 5-6 mm, 7-8 mm, and 9-10 mm) were sampled from male and female strains, and fixed with Zinc Formal-Fixx (Thermo Shandon, Pittsburgh, PA) diluted 1:5 in 0.22-μm-filtered seawater for 16 h. Pedal diameters were measured with a scale when sea anemones were attached to the sides of the glass aquaria. For both male and female strains, 3 individuals were collected in each size range. Samples were embedded in Paraplast plus (Sherwood Medical, St. Louis, MO) and 4-µm serial sections were prepared with a microtome (Thermo Shandon). Hydrated sections were stained with Hematoxylin and Eosin Y (Thermo Shandon), and micrographs were taken under a microscope (IX71SF1, Olympus, Tokyo, Japan). To compare the progress of oogenesis in E. diaphana of different pedal-sizes, oocyte diameters were measured and recorded as the average of two measurements (the largest dimension of an oocyte and the second-largest dimension at a right angle to the first). Approximately 50-300 oocytes were analyzed from each individual, and average values were calculated. Then, based on diameters of mature oocytes (Chen et al., 2008), oocyte development was arbitrarily classified into 3 stages (Stages I-III) ( Table 2 ). To compare progress of spermatogenesis among E. diaphana of different pedal disc-sizes, male germ cells comprising the spermary were observed. Their development was classified into 4 stages (Stages I-IV) ( Table 2 ). Approximately 100 spermaries were analyzed from each individual. All image analyses were performed with Image J software (Wayne Rasband, National Institutes of Health, Bethesda, MD, USA; https://imagej.nih.gov/ij).

Large E. diaphana (5~7 mm in pedal diameter) were cut horizontally into two pieces with scissors, and the resulting head and trunk were retained. Then, the heads and trunks were put separately in two 40-L aquaria (~20 heads or trunks per aquarium) and cultured for 3 months under conditions described above. Survivorship of heads and trunks was determined by dividing numbers at the end of culture by initial numbers. The experiment was conducted in triplicate, with a total of approximately 60 individuals (parts) cultured in each group. Presence or absence of gonads and gametogenesis was investigated in regenerated individuals (n=9 for each group). Histological sections were stained with Hematoxylin and Eosin Y (Thermo Shandon), and observed under a microscope (BX51, Olympus).

Testes/ovaries were isolated with forceps under a stereo microscope (EZ4, Leica Microsystem) from sea anemones with pedal disc diameters >7 mm and pooled in a Petri dish (9-cm diameter) containing 50 mL of 0.22-μm-filtered seawater. After several washes with 0.22 μm-filtered seawater, isolated gonads were snap-frozen with liquid nitrogen and stored at -80°C until use. Total RNA of the 2 samples (pooled testis and ovary samples) was extracted with TRIzol reagent (Thermo Fisher Scientific, Waltham, MA). Then, DNase I-treated RNA samples were sent to the Beijing Genomics Institute (BGI, Shenzhen, China) and qualified using a Bioanalyzer 2100 (Agilent Technologies, Santa Clara, USA) with an RNA 6,000 Labchip kit (Agilent Technologies). cDNA libraries were constructed from high-quality RNAs (RIN > 9) using TruSeq RNA Sample Prep Kits (Illumina, San Diego, USA), according to the manufacturer’s protocol. Libraries were subjected to 150-bp paired-end Illumina Hiseq sequencing. Illumina adaptors and low-quality bases or reads (QV< 20) were trimmed from raw sequences. Only high-quality data were used for the subsequent transcriptome assembly. Trimmed reads obtained above were mapped to the genomic reference sequence of E. diaphana (Baumgarten et al., 2015). Unmappable trimmed reads of each sample were assembled using Trinity v2.0.6 software (Grabherr et al., 2011). CD-HIT-EST was used to remove contig sequence redundancy (Li and Godzik, 2006). Obtained unigenes were annotated to the NCBI non-redundant protein (nr) database using the alignment algorithm, RAPSearch2, with a cut-off E-value <= -3.

To identify genes that could be used as markers for germline cells or cells capable of producing germline cells, i.e., stem cell-like cells, genes that are reportedly expressed in cnidarian germline cells or multipotent stem cells were used as queries in the E. diaphana gonadal transcriptome, based on the literature (Mochizuki et al., 2001; Seipel et al., 2004; Extavour et al., 2005; Rebscher et al., 2008; Alié et al., 2011; Leclère et al., 2012; Nishimiya-Fujisawa and Kobayashi, 2012; Shikina et al., 2012; Shikina et al., 2015; Praher et al., 2017). We utilized the following two strategies: 1) Gene names or keyword searches were performed in annotation results; and 2) Full-length cDNA sequences of genes in other animals were retrieved from Genbank (NCBI), and local BLAST searches were conducted (BLASTP, cut-off -E value 1e^-5) against translated sequences of E. diaphana gonadal transcriptome.

Sequences of vasa and PL10 subfamily members from various animals were retrieved from GenBank, and phylogenetically compared with E. diaphana vasa and PL10 obtained in the present study. Sequence Alignment was performed using Muscle, and results were used to create a 10,000-bootstrap-replicate phylogenetic tree (neighbor-joining method) using MEGA7 (Kumar et al., 2016).

Different polyp tissues, i.e., tentacles, testes, ovaries, and mesenterial filaments, were microscopically isolated from E. diaphana. Total RNA was extracted with TRIzol reagent (Thermo Fisher Scientific) following the manufacturer’s protocol. First‐strand cDNA was synthesized from 2 μg of DNase‐treated RNA using SuperScript III reverse transcriptase (Thermo Fisher Scientific). Transcript levels were analyzed with quantitative real-time RT-PCR using a Bio-Rad CFX Connect™ Real-Time PCR detection system (Bio-Rad Laboratories, Hercules, CA). The reaction was performed with iQ™ SYBR Green Supermix (Bio-Rad Laboratories) under the following amplification conditions: 95°C for 5 min, 40 cycles of 95°C for 15 s and 60°C for 1 min. Exaiptasia diaphana β-actin was used as a reference gene. Calculations were performed with the 2−▵▵Ct method (Livak and Schmittgen, 2001). Primers ( Table 1 ) were designed using Primer Express 3.0 software (Applied Biosystems, USA). Primer efficiencies were determined with a 5-fold dilution series of template cDNA. Only primer pairs that worked at 90-100% efficiency were used for the analysis.

Proteins of tentacles, testes, ovaries, and mesenterial filaments were extracted with TRIzol reagent (Invitrogen) following the manufacturer’s protocol. Protein concentrations were determined using a bicinchoninic assay (BCA) protein assay kit (Pierce Biotechnology, Rockford, Il, USA). Protein samples (10 μg) were subjected to SDS-PAGE and Western blotting, as previously described (Shikina et al., 2012). For the primary antibody reaction, an antibody against Euphyllia ancora (a stony coral, recently renamed Fimbriaphyllia ancora Luzon et al., 2017) vasa was used (Shikina et al., 2012). The antibody was diluted 1:20,000 in Tris-buffered saline and 0.1% Tween 20 (TBST) with 1% skim milk and used for the reaction. For the secondary antibody reaction, alkaline phosphatase-conjugated goat anti-rabbit IgG antibody (AnaSpec, San Jose, California; 0.25 μg/mL in TBST with 1% skim milk) was used. Visualization of immunoreactive signals on membranes was performed using NBT/BCIP solution (Sigma-Aldrich).

Immunohistochemical staining was performed according to methodology described elsewhere (Shikina et al., 2012). Briefly, hydrated sections were incubated for 30 min with HistoVT ONE (Nacalai Tesque, Inc, Kyoto, Japan) for antigen retrieval. After washing with phosphate-buffered saline containing 0.1% Tween 20 (PBT), sections were incubated for 10 min in 3% H₂O₂, and for 1 h in in 5% skim milk for blocking. Sections were then incubated for 16 h in anti-E. ancora vasa antibody (1:4,000 in PBT with 2% skim milk) (Shikina et al., 2012) at 4°C. For the secondary antibody reaction, sections were incubated with a biotinylated goat anti-rat IgG antibody (Vector Laboratories, Burlingame, USA; diluted 1: 2,000 in PBT with 2% skim milk) for 30 min. Immunoreactive signals were visualized with avidin–biotin–peroxidase complex (ABC) solution (Vector Laboratories), and 3,30-diaminobenzidine (DAB; Sigma-Aldrich). Sections were counterstained with haematoxylin. Observations were performed under a microscope (BX51, Olympus).

All data are presented as means ± standard errors (SE). For comparisons among more than three groups, statistical significance was determined using one-way ANOVA followed by Tukey’s test, with statistical significance at P<0.05. All analyses were performed using SPSS software.

A number of E. diaphana were found in a ditch near the aquatic animal center of National Taiwan Ocean University in northern Taiwan ( Figures 1A, B ). We collected several specimens and reared them in aquaria in the laboratory ( Figure 1C ). Exaiptasia diaphana underwent active asexual reproduction following pedal laceration ( Figure 1D ). Pedal laceration began when the anemones formed small fragments (1-2 mm in diameter) by separating part of the peripheral margin of pedal disc tissue extended on the glass surface of aquaria ( Figure 1D ). The lacerated pieces (pedal fragments) transformed into small anemones within about 1-2 weeks ( Figures 1E, e’ ). Exaiptasia diaphana that were cultured for more than 2 years in the laboratory, continued to grow and produced many individuals asexually. Four to eight lacerated pieces were produced per week per individual. Eventually, we were able to establish two clonal strains (one male and one female) ( Figure 1F ). Phylogenetic analysis with SCAR markers showed that established clones were genetically different from known E. diaphana strains ( Figure 1G ).

To investigate the sexual reproductive capacity of these established E. diaphana strains, some large individuals (> 5-6 mm in pedal disc diameter) were dissected and examined anatomically and histologically. Ovaries and testes were formed in mesenterial tissues of some large females and males, respectively ( Figures 2A, B ). Microscopically, ovaries were faint yellow, whereas testes were white. Developing germ cells were confirmed histologically ( Figure 2C, D ). Subsequent investigation of the relationship between polyp (pedal disc) size and gametogenesis showed that larger individuals of both sexes tend to have more mature gametes ( Figures 2E, F ). Females with a pedal disc diameter > 5-6 mm had larger mature oocytes ( Figure 2E ). Mean diameters of oocytes in females with pedal sizes of 5-6, 7-8, and 9-10 mm were 37.2 ± 1.7 μm (n=137 oocytes), 45.0 ± 7.9 μm (n= 686 oocytes), and 59.8 ± 7.8 μm (n=935 oocytes), respectively. Male E. diaphana with pedal disc diameters > 3-4 mm had mature sperm ( Figure 2F ). Gametogenesis was not observed in females < 3-4 mm or males < 1-2 mm in pedal disc diameter ( Figures 2E, F ).

To further understand the sexual reproduction capacity of E. diaphana, we conducted an amputation experiment. Exaiptasia diaphana (5~7 mm in pedal diameter) were cut horizontally in two, and the resulting heads and trunks were cultured in separate tanks ( Figure 2G ). Heads (upper part of the polyp) were mainly composed of tentacles, the mouth, the upper part of the pharynx, and body wall, and contained no gonadal tissues. In contrast, trunks (lower part of the polyp) were composed of gonads, part of the pharynx, mesenterial filaments, body wall, and pedal disk ( Figure 2G ). After a few months of culture, all surviving individuals had regenerated their lost body parts. Survivorship of trunk-regenerated and head-regenerated polyps was 26.3 ± 6.1% and 61.5 ± 3.4%, respectively ( Figure 2H ). Particularly, in trunk-regenerated polyps, which originally had no gonads, formation of gonads and gametes was histologically confirmed in all individuals, as in head- regenerated polyps ( Figure 2I ).

To efficiently identify marker genes for germline cells of E. diaphana, we performed gonad transcriptome sequencing. Testes and ovaries were isolated from E. diaphana with pedal disc diameters 7-8 mm (one individual for each sex), and were subjected to RNA-seq ( Table S1 ). Approximately 234 million raw reads comprising 28 Gb of clean, transcriptomic, sequencing data were obtained by Illumina paired-end sequencing. De novo assembly of all clean reads produced 26,932 unigenes with an average size of 1,226 bp and an N50 of 1,660 bp ( Table S1 ). Sequences were deposited in NCBI BioProject under accession numbers PRJNA1004228.

After annotation, genes reportedly expressed in germline cells of sea anemones were searched and identified in the established gonadal transcriptomes ( Table S2 ). We then focused on vasa, and performed molecular identification and characterization. Full-length E. diaphana vasa cDNA was successfully determined by PCR, followed by RACE-PCR (GenBank No. ON375355). The deduced amino acid sequence was confirmed to contain conserved motifs characteristic of Vasa proteins, including the Q-motif, ATPase motifs, motifs involved in ATP binding and cleavage, RNA unwinding motifs, and the helicase C domain ( Figure 3A ). A phylogenetic analysis of DEAD-box protein family members showed that E. diaphana Vasa belonged to the Vasa family, and is not associated with PL10 proteins ( Figure 3B ). Tissue distribution analysis of E. diaphana vasa transcripts by quantitative RT-PCR showed that the transcripts were expressed significantly more abundantly in ovary than in other tissues ( Figure 3C ). Transcripts were also present in testis, although the statistical significance was not determined ( Figure 3C ). Subsequent Western blotting detected expression of E. diaphana Vasa protein primarily in testes and ovaries. The immunoreactive band was also faintly detected in tentacles and mesenterial filaments of both sexes ( Figure 3D ). The size of the immunoreactive band was approximately 82 kDa, close to the predicted molecular mass of 81 kDa of E. diaphana Vasa ( Figure 3D ).

Immunohistochemical analysis of gonads demonstrated that E. diaphana Vasa is expressed in oogonia and developing oocytes in the ovary ( Figures 4A, B ). Vasa was also expressed in spermatogonia ( Figures 4C, D ), spermatocytes, and spermatids in the testes, ( Figures 4E, F ). Immunoreactivity was faint or almost undetectable in mature sperm ( Figure 4F ). These results demonstrated that vasa can be used as a marker of germline cells.

Finally, we investigated germline cells or cells capable of producing germline cells in various somatic tissues in polyps using vasa expression as a marker. In tentacles, vasa-positive cells were detected in patchy clusters in the gastrodermis ( Figures 5A, a’ ). The morphology of vasa-positive cells in tentacles was different from that in gonads. Vasa-positive cells in tentacles exhibited irregular shapes and had small nuclei that stain well with hematoxylin ( Figure 5a’ ). In contrast, there are few vasa-positive cells in pharynx and mesenteries ( Figures 5B, b’ ), mesenterial filaments ( Figures 5C, c’ ), pedal disc ( Figures 5D, d’ ), or lacerated pieces ( Figures 5E, e’ ). Vasa-positive cells in those tissues were present singly, and were morphologically similar to early-stage germ cells that were observed in gonads of juvenile anemones ( Figures 5F, f’ ).