Fig. 7 in Agarna malayi Tiwari 1952

© 2018 Academia Sinica, Taiwan Open Access Comprehensive Analysis of the Jellyfish Chrysaora pacifica (Goette, 1886) (Semaeostomeae: Pelagiidae) with Description of the Complete rDNA Sequence Jinho Chae1, Yoseph Seo2, Won Bae Yu2, Won Duk Yoon3, Hye Eun Lee4, Soo-Jung Chang5, and Jang-Seu Ki2,* 1Marine Environmental Research and Information Laboratory, Gunpo 15850, Korea. E-mail: [email protected] 2Department of Biotechnology, Sangmyung University, Seoul 03016, Korea. E-mail: [email protected]; (Seo) wonbae51[email protected] (Yu) 3Human and Marine Ecosystem Research Laboratory, Gunpo 15850, Korea. E-mail: [email protected] 4Ocean Climate and Ecology Research Division, National Institute of Fisheries Science, Busan 46083, Korea. E-mail: [email protected] 5Fisheries Resources and Environment Division, West Sea Fisheries Research Institute, National Institute of Fisheries Science, Incheon 22383, Korea. E-mail: [email protected] (Received 5 April 2018; Accepted 24 September 2018; Published 7 November 2018; Communicated by James D. Reimer) Citation: Chae J, Seo Y, Yu WB, Yoon WD, Lee HE, Chang SJ, Ki JS. 2018. Comprehensive analysis of the jellyfish Chrysaora pacifica (Goette, 1886) (Semaeostomeae: Pelagiidae) with description of the complete rDNA sequence. Zool Stud 57:51. doi:10.6620/ ZS.2018.57-51. Jinho Chae, Yoseph Seo, Won Bae Yu, Won Duk Yoon, Hye Eun Lee, Soo-Jung Chang, and Jang-Seu Ki (2018) The Scyphomedusae genus Chrysaora consists of highly diversified jellyfishes. Although morphological systematics of the genus has been documented over the past century, characterization of molecular taxonomy has been attempted only recently. In the present study, we sequenced an 8,167 bp region, encompassing a single ribosomal DNA (rDNA) repeat unit, from Chrysaora pacifica, and used it for phylogenetic analyses. The tandemly repeated rDNA units turned out to consist of both coding and noncoding regions, whose arrangement was found to be the same as that of a typical eukaryote. None of the 5S rRNA sequences were found among the repeat units. Comparative analyses of jellyfish rDNA sequences showed that the 28S locus is highly informative and divergent compared to the 18S locus. Phylogenetic analyses of the 18S and 28S loci revealed that the Semaeostomeae order of jellyfish is separated into taxonomic groups by families and genera, with a few exceptions. The family Pelagiidae was in a clade separate from other groups, thus forming a monophyletic lineage. All Chrysaora included here formed a strongly supported clade within the family Pelagiidae, and Pelagiidae manifested a sister relationship with Cyanea. Nonetheless, Chrysaora was found to be paraphyletic in both 18S and 28S phylogenies. Chrysaora pacifica was clearly distinct from close relatives C. melanaster and C. quinquecirrha. These results provide a special reference for the DNA taxonomy of Pelagiidae jellyfishes in terms of nuclear cistron rDNA sequences and improve our understanding of the molecular phylogenetic relationships among Semaeostomeae jellyfishes. Key words: Jellyfish, Pelagiidae, Chrysaora pacifica, Ribosomal DNA, Transcription repeat unit. *Correspondence: Tel: +82-2-2287-5449. Fax: +82-2-2287-0070. E-mail: [email protected] BACKGROUND The scyphozoan medusa genus Chrysaora Péron & Lesueur, 1810 (Semaeostomeae: Pelagiidae) is composed of many species, and frequently occurs in coastal waters around the Zoological Studies 57: 51 (2018) doi:10.6620/ZS.2018.57-51 1 © 2018 Academia Sinica, Taiwan world. Genera Chrysaora and Pelagia Péron & Lesueur, 1810 were both first described mainly on the basis of tentacle numbers. Studies on the developmental life cycle, including sexual maturity, showed that the genera are different from each other (Mayer 1910). Currently, species of Chrysaora are discriminated by morphological characteristics such as the tentacle number, shape of radial septa, order of tentacle development, coloration, and the form of nematocyst capsules (Morandini and Marques 2010). Nevertheless, these key features are obscure among preserved and live specimens owing to their fragile, transparent bodies and morphological differences depending on the environment. Thus, Chrysaora has been a subject of taxonomic confusion and species misidentification (Morandini and Marques 2010; Bayha at al. 2017). Recently, Morandini and Marques (2010) analyzed morphological features of various Chrysaora specimens and found 13 valid and two doubtful species. Although the taxonomic system was plausible, the species concepts and systematics of Morandini and Marques (2010) remain unstable due to inconsistency in morphological characteristics and the presence of undescribed and/or cryptic species. Molecular approaches can help resolve the ambiguities of morphological taxonomy (e.g., Collins et al. 2006; Ki et al. 2008; Lee at al. 2016). Molecular approaches include direct DNA sequencing, restriction fragment length polymorphism, and DNA chips (Ki et al. 2008 2010). Among these methods, DNA sequencing is quite effective in constructing phylogenies and elucidating evolutionary inferences in jellyfish. To date, DNA markers of many jellyfish species have been sequenced and are available in public databases. Species assignments and populationgenetic and phylogenetic analyses of jellyfishes have been conducted based on DNA sequences of nuclear and mitochondrial genes (Schroth et al. 2002; Dawson et al. 2005; Collins et al. 2006; Ki et al. 2009; Ramšak et al. 2012; Lee et al. 2013; Glynn et al. 2015; Dong et al. 2016; van Walraven et al. 2016). Until now, these phylogenetic studies have been mostly carried out at higher taxonomic levels (family or higher) and have not been attempted at lower taxonomic levels such as genera and species. Indeed, existing reports contain little information on Chrysaora (Bayha et al. 2010; Rizman-Idid et al. 2016) because of a dearth of DNA sequence data on this genus. Nonetheless, DNA sequence data on the family Pelagiidae have been obtained for some members of genera Chrysaora, Mawia Avian, Ramšak, Tirelli, D’Ambra & Malej, 2016; Sanderia Goette, 1886; and Pelagia and used in taxonomic redescriptions (Avian et al. 2016). More recently, Bayha at al. (2017) extensively analyzed U.S. Atlantic specimens of Chrysaora and tried to decipher the molecular relationships among pelagiid species, particularly focusing on the features distinguishing C. quinquecirrha (Desor, 1848) from C. chesapeakei (Papenfuss, 1936). Nuclear ribosomal DNA (rDNA) in eukaryotes is typically structured as tandem arrays of a basic unit that contains a transcription unit (18S, 5.8S, and 28S) and an intervening intergenic spacer (IGS) region (Hill et al. 1990). The different subunits and loci of rDNA, therefore, have different degrees of sequence variability and varying suitability for comparison at the intergenus level or inter-species level (Ki et al. 2009). In addition, these properties offer various options for data construction in phylogenetic analyses. Because the coding regions evolve at a slower rate, they allow phylogenies to be constructed for more distant divergences. Conversely, the high variability of the noncoding regions is useful for reconstructing relatively recent evolutionary events (Hillis and Dixon 1991). Previously, we (Ki et al. 2009) reported for the first time the complete rDNA sequence of the moon jelly Aurelia coerulea von Lendenfeld, 1884 (formerly Aurelia sp. 1). Except for this example, the complete rDNA region has not been sequenced in other jellyfish species, and thus, the rDNA sequences of jellyfishes are characterized insufficiently. In the present study, we determined the complete nucleotide sequence of a single unit of tandemly repeated rDNA of Chrysaora pacifica (Goette, 1886) and characterized the molecular features of various rDNA components according to individual molecules. Comparative analyses of parsimony and dot plot analyses were performed on already known complete and/or partial rDNA sequences to gain a better understanding of jellyfish rDNA characteristics. Finally, we studied the phylogenetic relationships among members of the order Semaeostomeae, in particular focusing on the genus Chrysaora. page 2 of 15Zoological Studies 57: 51 (2018) © 2018 Academia Sinica, Taiwan MATERIALS AND METHODS Sample collection, morphology, and DNA extraction Chrysaora pacifica specimens were examined and collected from the waters of Namildae, Tongyeong, Jindong, Jangmok, Busan, and Yangyang (southern and south-eastern coasts of Korea) and showed no conspicuous morphological differences (Lee et al. 2016). Jellyfish specimens for this study were collected in Tongyeong Bay (34°55.59N, 128°5.79E), Korea, on 14 August 2013. Before sample collection, we photographed the jellyfish specimens (Olympus OM-D E-M5 with an underwater housing) in their natural habitat to record live morphology. For genetic analyses, oralarm and gonad tissues were preserved in 100% ethanol (Merck, Darmstadt, Germany). Prior to genomic-DNA extraction, the alcohol-preserved specimens were washed in distilled water to remove all ethanol, and this procedure was repeated several times during the night. Total genomic DNA was extracted by the cetyltrimethylammonium bromide (CTAB) method described in Ausubel et al. (1989). Polymerase chain reaction (PCR) Nuclear rDNA sequences were amplified by the long PCR technique with two sets of eukaryotic universal primers (forward 18F01, 5’-TAT CTG GTT GAT CCT GCC AGT AG-3’ and reverse 28R691, 5’-CTT GGT CCG TGT TTC AAG AC3’; forward 28F01, 5’-CCG CTG AAT TTA AGC ATA TAA GTA AGC-3’, reverse 18R, 5’-GCT ATT GGA GCT GGA ATT ACC-3’), according to our previous study (Ki et al. 2009). PCR was carried out in 20 μL reaction mixtures containing 12.9 μL of sterile distilled water, 2 μL of 10× Ex PCR buffer (TaKaRa, Shiga, Japan), 2 μL of a dNTP mix (4 mM each), 1 μL of each primer (10 pmoles), 0.1 μL Ex Taq polymerase (2.5 U), and 1 μL of a template. PCR cycling was performed on a BioRad iCycler via the following program: 94°C for 5 min; followed by 35 cycles of 94°C for 20 sec, 55°C for 30 sec, and 68°C for 5 min; with a final extension at 72°C for 10 min. Resulting PCR products were subjected to electrophoresis in a 1.0% agarose gel (Promega, USA), stained with ethidium bromide, and visualized under ultraviolet light on a transilluminator. The amplicons were then purified with the QIAquick PCR Purification Kit (Qiagen GmbH, Germany), and DNA sequencing reactions were run with the ABI PRISM® BigDye™ Terminator Cycle Sequencing Ready Reaction Kit (PE Biosystems, CA) using the PCR products and PCR primers. After that, the remaining DNA sequences were determined by primer walking. Labeled DNA fragments were analyzed on an automated DNA sequencer (Model 3700, Applied Biosystems, CA). Editing and contig assembly of the rDNA sequence fragments were carried out in Sequencher 4.7 software (Gene Codes, MI). The coding rDNA genes were identified with the help of the NCBI database and Aurelia coerulea sequence (GenBank No. EU276014). All the sequences determined here were deposited in the GenBank database. Data analyses General molecular features of the Chrysaora pacifica rDNA were calculated in Genetyx ver. 7.0 (Hitachi Engineering Co., Tokyo, Japan) and MEGA ver. 5.0 (Tamura et al. 2011). In addition, nucleic acid distribution, sequence complexity, and entropy across the entire rDNA nucleotides of C. pacifica were calculated by means of BioAnnotator in Vector NTI Advance ver. 10.3.0 (Invitrogen, San Diego, CA). The repeat sequence pattern in the rDNA IGS sequences was analyzed in Genetyx ver. 7.0 and Tandem Repeats Finder (http:// tandem.bu.edu/trf/trf.basic.submit.html). A comparison of molecular features was performed by comparing the rDNA sequence of Chrysaora pacifica with that of the moon jelly Aurelia coerulea (Scyphozoa, EU276014). Dot plot analysis of both rDNA regions was carried out in the MegAlign software, ver. 5.01 (DNAstar Inc Madison, WI). In addition, parsimony analyses were performed on 11 DNA sequences of jellyfish 18S and the corresponding 28S rDNA loci (Table S1). DNA sequences were aligned on the online server of MAFFT (https://mafft.cbrc.jp/alignment/ server/; Katoh and Standley 2013). Genetic distances were calculated using the aligned DNA sequences via the Kimura 2-parameter model in DNASIS ver. 3.5 (Hitachi Software Engineering America, San Bruno, CA). Molecular similarity was measured by means of BioEdit ver. 5.09. Further comparative analyses, such as determination of parsimony-informative sites and of the transition/ transversion ratio, were performed on the above data matrix in MEGA ver. 5.0. Putative signals of assumed termination (e.g., poly(T) tract) and promoter signals (bi-repeats) page 3 of 15 Zoological Studies 57: 51 (2018) © 2018 Academia Sinica, Taiwan within the IGS rDNA were analyzed in Genetyx ver. 7.0. Additionally, putative secondary structures of the assumed signals were estimated in DNA sequences of their flanking regions using webbased software Mfold ver. 3.2 (http://unafold.rna. albany.edu/?q=mfold) according to Zuker (2003). Phylogenetic analyses For phylogenetic reconstruction procedures, the 18S rDNA sequence was determined and the remaining sequences were obtained from the DDBJ/EMBL/GenBank database (Table S1). These sequences were aligned in MAFFT (Katoh and Standley 2013), and ambiguous regions were removed on the Gblocks server with the least stringent settings (http://molevol. cmima.csic.es/castresana/Gblocks_server.html; Castresana 2000). Maximum likelihood (ML) analysis was conducted on an 18S data matrix (1,483 alignment sites) in RAxML ver. 8.0 using the GTR+G nucleotide substitution model (Stamatakis 2014). Branch support was assessed with 5,000 bootstrap replicates. Additional Bayesian analysis of the same dataset was implemented in MrBayes ver. 3.1.2 (Huelsenbeck and Ronquist 2001) via the GTR+G nucleotide substitution model. The Markov chain Monte Carlo (MCMC) process was set to four chains, and 1,000,000 generations were carried out with the sampling frequency of one per 100 generations. After this analysis, the first 2,000 trees were deleted as burn-in, and a consensus tree was constructed. Bayesian posterior probabilities (> 0.50) are indicated at each branch node in the figures. Phylogenetic trees were visualized in TreeView ver. 1.6.6 (Page 1996). In addition, phylogenetic analyses of scyphozoan 28S rDNA sequences were performed on a 28S data matrix (843 alignment sites) via the GTR+G model, and an identical methodology was used for analysis of the 18S rDNA region. RESULTS The complete rDNA sequence of Chrysaora pacifica In this study, we performed morphological examination of the collected specimens before proceeding with alcohol preservation. Jellyfish specimens from the study site (off Tongyeong) had typical morphological characteristics of Korean Chrysaora pacifica. The bell diameter was approximately 15 cm, marginal lappets were rounded, and there were 32 brownish radiating strips and 40 tentacles. Live specimens (Fig. 1) possessed conspicuously extended oral arms and tentacles as compared to the formalin-fixed specimens. In addition, we determined the sequence of the full-length 8,167 bp single rDNA repeat unit of Korean Chrysaora pacifica (GenBank No. KY212123). It was found to be organized in the typical rDNA fashion of eukaryotes, i.e., 18S-ITS15.8S-ITS2-28S-IGS (Fig. 2A). The rDNA region of the species was found to have the following structure: 1,810 bp 18S, 246 bp ITS1, 158 bp 5.8S, 182 bp ITS2, 3,609 bp 28S, and 2,162 bp IGS. Intron-like sequences were not detected in 18S and 28S rDNA. Upon comparison, we found that each coding region was nearly identical in length between our study specimen and Aurelia coerulea (GenBank No. EU276014); however, noncoding regions (i.e., ITSs and IGS) considerably differed in length. Particularly, the IGS of C. pacifica (2,162 bp) is much longer than that of A. coerulea (1,603 bp). G+C content of the full-length region turned out to be 46.5% (A, 25.2%; T, 32.5%; G, 23.6%; and C, 18.7%). Cytosine (C) content of each rDNA sequence was the lowest. Nucleotide composition was found to differ between coding and noncoding regions of rDNA. Sequence complexity and nucleic-acid distribution were determined using sliding windows of 100 nucleotides along the entire rDNA sequence (Fig. 2B). The distribution of G+C content was ~50% across the complete rDNA sequence. Nevertheless, some sites in the IGS showed considerably lower G+C content and high fluctuation due to the presence of GCand AT-rich loci and poly(T) tracts (Fig. 3). In addition, sequence variability was analyzed via sequence complexity and entropy plotting. Overall, the two variables fluctuated against one another along the rDNA. These observations revealed a clear difference in profiles between the coding and noncoding regions (ITS and IGS). Sequence complexity was considerably higher in the noncoding regions, such as ITS and IGS, compared to the coding regions. The locus that manifested the lowest complexity corresponded to GCand AT-rich sequences within the IGS. Highly informative characteristics of IGS rDNA The IGS rDNA of Chrysaora pacifica contains 2,162 nucleotides (Fig. 3), and the nucleotide page 4 of 15Zoological Studies 57: 51 (2018) © 2018 Academia Sinica, Taiwan frequencies of A, T, G, and C were measured and turned out to be 26.5%, 31.5%, 21.3%, and 20.6%, respectively. Overall, A+T content was higher than G+C content owing to the presence of poly(T) tracts and an AT-rich sequence. In addition, we detected a GC-rich sequence in the IGS. After a comparison with the 5S rDNA database (http:// combio.pl/rrna/, accessed 20 Mar. 2018), none of the 5S rDNA sequences were found to be present within the IGS locus. Rather, the IGS consisted of transcription termination and bi-repeated sequences. We detected two poly(T) tracts (5’-TAT TTT TTT T-3’, 5’-TTA TTT TTT TCT TT-3’) in the 5’ external transcribed site (ETS) area adjacent to the end of 28S rDNA (Fig. 3), possibly serving as a termination signal. As a putative RNA polymerase I transcription initiation site, many promoters of RNA polymerase I are bipartite, consisting of a proximal promoter domain and an upstream control element (UCE) (Chen et al. 2000). In the present study, we identified a promoter candidate pattern of bi-repeated sequences (5’-CTG ATA TAG AG-TAG AGC ATG GCT TAG-CTG ATA TAG AG-3’ (underlined nucleotides likely form a hairpin structure) within the IGS. Fig. 1. Live Chrysaora pacifica in natural habitat: basolateral (A and B), lateral (C) and apical view (D). (A) (B) (C) (D) page 5 of 15Zoological Studies 57: 51 (2018) © 2018 Academia Sinica, Taiwan Intraand interspecific comparisons of complete rDNA We compared the rDNA of our study specimen, Chrysaora pacifica (collected in 2013 in Tongyeong, Korea), to other Korean specimens collected on the Namildae Beach (in Sacheon), Tongyeong, Jindong, and Jangmok in Korea during August 2014 (Lee et al. 2016) and found that all of them had the same DNA sequence, implying that Fig. 2. A schematic representation of the single unit of rDNA (A), and GC content (%), nucleic acid distribution (% thymine), sequence complexity, and entropy (dS) in 100-bp windows across the entire rDNA nucleotides of Chrysaora pacifica (B). In the full rDNA (A), solid boxes indicate the ribosomal RNA genes and thin lines represent ITS or IGS. Nucleotide sequences in length and GC composition of each locus are represented near a line by calculation from a single unit of rDNA. The putative transcription start site is represented by an arrow; solid inverted-triangles represent sub-repeats in IGS. 2,162 246 1,810 226 ITS1 ITS2 182 3,609 5.8S 18S 28S 28S 1 kb 49.6 47.6 49.1 48.2 53.0 47.1 ETS+NTS Length (bp) % GC Entropy (dS) (cal/K/mol) Sequence complexity Nucleic acid distribution (T%) Nucleic acid distribution (A%) Content (GC%) -2.07e +003 -2.74e +003 1 +0.95 +0.57 +70.3 +9.9 +52.5 +8.91 +67.3 +11.9 600 1,200 1,800 2,400 3,000 4,200 4,800 5,400 6,000 6,600 7,200 7,800 (bp) 3,600 18S rDNA 28S rDNA IGS ITS1 ITS2 5.8S Termination signal (A) (B) page 6 of 15Zoological Studies 57: 51 (2018) © 2018 Academia Sinica, Taiwan the species were identical. In addition, we compared the rDNA sequences of Chrysaora pacifica (GenBank No. KY212123) with the rDNA sequence of Aurelia coerulea (GenBank No. EU276014). High similarity and a small genetic distance were noted between the coding regions of these species; in contrast, the highest dissimilarity was identified in the rDNA IGS (67.3%), followed by ITS1 (45.6%). Of the coding regions, the 28S sequence (0.0651) manifested a greater genetic distance than did the others (18S, 0.0173; 5.8S, 0.0128) as calculated via the Kimura 2-parameter model. Dot plots of the complete rDNA regions of C. pacifica and A. coerulea graphically revealed sequence similarity (Fig. 4). It was determined by means of sliding Fig. 3. Nucleotide sequences of the Chrysaora pacifica IGS rDNA. Bi-repeats are indicated by asterisks, and microsatellite-like nucleotides are marked by lines. Putative termination signals (poly(T) tract) are represented as Ts-1, Ts-2 with box. Blocks of both ATand GC-rich regions are presented in separate boxes. -70 TATTGTACGAAGTAGAGTAGCCTTGTTGCTACGATCTTCTGAGATTAAGCCCTTCGTTCTATAGATTTGT 1 TAACACTTTGTTGTTAACACTATTTTTTTTCACAAAGTTTCATTCGACCTCCCACCTTTTTAAACTTACT 71 CCTCTACCTTATTTTTTTCTTTACTATATTTTATCTACTGACCACCTTTGTAAACTTGTATATTTCTTGT 141 ATACTGTATGCTTTCTACTGACCACCTTTGAATACTTATATATTTCTTCTGTATCTTGTCTACAAAATTT 211 CTATACTATATTTTATCTACTGACCACCTTTGGAAACTTGTATACTTCTTGTATACTGTATGCCCTCTAC 281 TGACCACCTTTGAATACTTACTGTATATATTTCTTCTGTATCTTGTCTACAAATGCCAGTTGTCTGAATC 351 TGCTGACCACCTTTGGAAACTTAGGTTAAATGTCTATATCAGAATATATACTTCCATGCTTTGATGATTA 421 TTATCACATTCGAAGTATCAACTATAATACATGAATTTTGCAGCAACAGTTGCAGCAGCTTTTGGTTGTA 491 CATTTAAATGCCTATCATTATATCTATATCTATATCTATTATTATCAATATCTATTATATCCATCTATAT 561 CTATCTACAATGTATCTATTATTATCTATCTATCTATCTACTGTATAGCTGTGTCATTATATCATTGCAT 631 ATATCTATATCATTGTACATACATACAGTATATACCTTTTGTCTATCAAAGACGCTGGAAAGAAATAGAA 701 ATAGAAAAGAAAATAAAGGAATTAATAATTAATAAATAATAATAAAATAATGAATAATAAAACAAATAGA 771 AGAGAATAAAAAAAAAAAGCAAAGGATGCCAGGAAGGAATTGTTGATCGAGGGAGTTGCGTCCGGGACGT 841 CGACGTCGCGCCGGGGGTTCGGGGGCGGAGCCCCCGAAAATTTTTTTGGGCCACGCCCTTTTTTATTCTT 911 GACCACGCCCTTTTTTAATGAGACCACGCCCTTTTTAAAGGCCACGCCCCTTTTCTAGATTTCTTACTAT 981 CTAATTGATAGACAAATGGATGCCATTTGACGAGCTGCAGACTGCAATACCAGATTTATCCATTGACTAG 1051 CAACTAAACGAGACTTCTTTCCACCACAGACGGCCCAGTCTGATATAGAGTAGAGCATGGCTTAGCTGAT 1121 ATAGAGCTCTCATCAATCAGCTAATATGGGGACAACAGCCGAGCTTAGTATCTTAGTATCGATAGGCAGG 1191 TAGATCAGTTGACTAGTTGTTGAGTGTCCAGAGCCAGCCGAACTGACTTCCGTTTACCTTTTCAAACTTT 1261 AGTTTTAGACAAGTTGTCGGGATACTATTAGAAAATGAAAAATATTGTTCGTTTTTTGACCTAGCTAACC 1331 TTCTGAACAATAATTCAATGTTTACGAAACGTTTCAGCTCAATGTGACTTGTCGGAGGCAGGAAACCTTT 1401 TGAAGTGCACAGTCGCCGACACGTACGTCTAGTTGGCCGGCTAGTTAATGTAGCGGGCATAGCAGCGATG 1471 ATGAAGCCTGGGAAGGGAGTGCCAGCTCTTCGACGTTGATCGAATAGTAGCGGCGCGACTCGTTACCTGT 1541 TGGATTCGTCGTGTTCAACTTTGCTCCGGTGTGTCGAGGTTCCCGTGTCTCATAACATGCTTGGCTTCCT 1611 AGCCTTCCATGGAGATGGGCACAGTGCGTGTGTGAGGGTTGGTTGTGCCAGCCATTTTCGAGTCATTTGG 1681 GACCGTACCCGATGATGATATACTATGAGAGGGAGGACTTGCAGGCCAATGAAGTCCGATCGGAAGTAGC 1751 CACGTCGTCGTCGGCCTCTAATGTCTCTCGATGGCAGGCCTCGGTGTCAGTTGAATGGGCCAGTCGTGTC 1821 CGCGGCTGGTGCATGAAGCATCGAAGCGTGAGGTGTTGTGTTGATGGTCGATGACCAACAACTCGAGTTG 1891 TGTAGCAAAATGGAGCTATATGCTTCACCATCGGTCGCTCTCGCCGCCCAGCTGAGCTCCTCGAGTTCAG 1961 CGTGGCCGTGTGTGTGAGTGCATCAGGCCGAGGAGGATGGAAAGATGTGGGAAAGACAAGGCTAGAGCCG 2031 AGCAAGTCGACCTAGAAGGCTAGCAGACTTTGTGTGCCCGGTCTAGTTAAGTGCGCGAGTGTGTGGAAAG 2101 TGACTCCGTCGTCGCATGCGCTTCGTATGCGTTCGTTCGGATGTTATCTGACTACTAGTCGTTATCTGGT +9 TGATCCTGCCAG IGS IGS *********** ****** ***** R1-1 R1-2 R1-3 R2-1 R2-2 R1-4 R1-5 Ts-1 Ts-2 AT-rich GC-rich page 7 of 15Zoological Studies 57: 51 (2018) © 2018 Academia Sinica, Taiwan windows of 60 nucleotides along the entire rDNA region. As expected, high similarities were detected in the rDNA coding loci (e.g., 18S, 5.8S, and 28S). By contrast, the dot plot showed no similarity between the loci within the noncoding rDNA regions. General features of jellyfish 18S and 28S rDNA sequences were analyzed among 11 cnidarian species (Table 1). The comparative analyses showed that the transition:transversion ratio (Ts/Tv) was slightly higher in the 28S rDNA sequences. Parsimony-informative sites, in contrast, had much greater differences in 28S rDNA (514 sites, 15.8%) than in 18S rDNA (156 sites, 9.1%). Phylogeny of Semaeostomeae jellyfish, including Chrysaora Phylogenetic relationships among the members of the order Semaeostomeae were investigated using ML trees inferred from the separate 18S and 28S rDNA sequences (Fig. 5). Additional Bayesian analyses generated nearly identical tree topologies as the ML trees. Hence, posterior probabilities (PP) from the Bayesian analyses were incorporated into the ML trees to support the strength of each branch. Comparisons of 18S and 28S trees indicated that branch topologies were generally identical, with a different position of the Cyaneidae family. According to expected changes per site, phylogenetic resolution of 18S rDNA was considerably lower (2.6-fold) than that of 28S rDNA. In a broad phylogenetic view, the order Semaeostomeae was clearly separated into four families (Cyaneidae, Drymonematidae, Pelagiidae, and Ulmaridae), forming unique monophyletic clades according to each taxonomic level. Among them, the family Drymonematidae diverged the earliest, and this finding was supported strongly by 100% bootstrap proportions (BP) and 1.00 PP. In addition, our 28S phylogeny suggested that Pelagiidae and Ulmaridae formed a sister clade (91% BP and 1.00 PP). As for the Pelagiidae relationships, our tree showed that the three tested genera-Chrysaora, Pelagia, and Sanderia-were not clearly separated into each genus group. For example, Pelagia and Sanderia ended up within clusters of Chrysaora in the 18S tree, but they were separated to form a unique clade in the 28S tree, with a sister relationship. On the other hand, Chrysaora formed a well-supported clade in the 18S tree (77% BP and 0.83 PP) and 28S tree (98% BP and 1.00 PP). After our analyses of these linages, Chrysaora formed a paraphyletic group according to both 18S and 28S phylogeny. Table 1. Sequence characteristics of nearly complete 18S and 28S rDNAs among cnidarians (MA in Supplementary Table 1), including hydrozoans and scyphozoans. p-distances were calculated with the Kimura 2-parameter model Locus Nn Nc Nv Ti Ts Tv Ts/Tv PI %PI 18S rDNA 1,729 1498 227 1,621 53 36 1.46 156 9.1 28S rDNA 3,251 2,544 690 2893 170 112 1.51 514 15.8 Nn, total number of sites; Nc, total number of conserved sites; Nv, total number of variable sites; Ti, Total numbers of identical pairs; Ts, Total numbers of transitional pairs; Tv, Total numbers of transversional pairs; PI, parsimony-informative site. Fig. 4. A dot matrix comparison of rDNA sequences between Chrysaora pacifica (KY212123) and Aurelia coerulea (EU276014). Color scale bars represent consecutive sequence length of some regions detected similarly between the two sequence pairs. The open boxes in matrices indicate rDNA coding regions such as 18S, 5.8S, and 28S. Percentage: 50; Window: 30; Min Quality: 1 Filter Top: 0 -> 1588, Bottom: 611 -> 833 Seq1(1>8167) Seq2(1>7680) Total CJ-18S-Eumetazoa; MJ-18S-Eumetazoa; Diagonals (1>8167) (1>7680) 64518 1,588 0 0123456 7 (kb) 8 0 1 2 3 4 5 6 7 (kb) Chrysaora pacifica Aurelia coerulea 18S ITS2 ITS1 28S IGS 5.8S page 8 of 15Zoological Studies 57: 51 (2018) © 2018 Academia Sinica, Taiwan Fig. 5. Phylogenetic relationships between jellyfishes within the order Semaeostomeae inferred from nearly complete 18S rDNA (A) and partial 28S rDNA sequences (B) with maximum-likelihood (ML) algorithms. ML analyses of 18S and 28S were used as the nucleotide substitution model of GTR+G. Two hydrozoans (Hydractinia echinata and Podocoryne carnea for 18S rDNA; Astrohydra japonica and Melicertissa sp. for 28S) were included as the outgroups. Additional Bayesian analysis generated similar topology of the tree compared with the ML tree. Posterior probabilities (PP) from the analyses were incorporated into the ML tree to support the strength of each branch. The first and second numbers at the nodes display bootstrap proportions (BP) (> 50%) in ML and PP (> 0.50) in Bayesian, respectively. Branch lengths are proportional to the scale given. *Represents controversial species names, because they were suspected as different species by Bayha et al. (2017). Pelagiidae Drymonema dalmatinum (HQ234657) Drymonema larsoni (HQ234652) Outgroup (Hydrozoa) Hydractinia echinata ( AY920763) Podocoryne carnea ( AF358092) 0.01 100/1.00 100/1.00 73/0.97 79/ 0.93 100/1.00 100/1.00 100/1.00 81/0.70 78/0.92 61/- 51/0.72 96/0.98 53/0.90 50/0.61 70/0.98 77/0.83 53/0.90 84/ - 59/0.66 100/1.00 0.04 Drymonematidae Ulmaridae Pelagiidae Cyaneidae Desmonema sp. KMB-2010 (HM194804) Cyanea annaskala (HM194778) Cyanea sp. AGC-2001 (AF358097) Cyanea sp. LIS_4 (JX393285) Cyanea lamarckii (JX995325) Cyanea capillata (JX995327) Cyanea tzetlinii (KM279703) Deepstaria enigmatica (HM194791) Phacellophora camtschatica (HM194822) Phacellophora sp. B2 (JX393293) Poralia rufescens (HM194792) Poralia sp. NA (JX393294) Aurelia coerulea (EU276014) Aurelia labiata (JX393276) Aurelia aurita (AY039208) Aurelia limbata (JX393277) Aurelia sp. 2 KMB-2010 (HM194821) Aurelia sp. AGC-2005 (AY920770) Chrysaora colorata (AF358098) Chrysaora fuscescens (JX393279) Chrysaora helvola (JX393280) Chrysaora melanaster (AF358099) Pelagia noctiluca (HE591464) Chrysaora lactea (HM194810) Chrysaora quinquecirrha (JX393282)* Chrysaora sp. AGC-2005 (AY920769) Sanderia malayensis (HM194808) Chrysaora pacifica (KY212123) 77/ 0.99 0.1 100/1.00 85/0.82 98/1.00 100/1.00 98/1.00 100/1.00 100/1.00 89/ 1.00 100/ 0.90 83/100 66/0.90 98/1.00 91/1.00 97/1.00 80/0.99 100/1.00 Ulmaridae Cyaneidae Outgroup (Hydrozoa) Melicertissa sp. AGC-2001 (AY920798) Astrohydra japonica (AY920794) Desmonema sp. KMB-20 (HM194857) Cyanea annaskala (HM194831) Cyanea capillata (HM194873) Sanderia malayensis (HM194861) Pelagiidae gen. sp. AR-2014 (KM217201) Pelagia noctiluca (KJ573407) Pelagia benovici (KJ573399) Chrysaora sp. M0D14611Y (HM194864) Chrysaora pacifica (KY212123) Chrysaora hysoscella (KM651815) Chrysaora lactea (HM194863) Chrysaora melanaster (AY920779)* Chrysaora fuscescens (HM194868) Chrysaora melanaster (AY920780) Phacellophora camtschatica (AY920778) Poralia rufescens (HM194845) Aurelia solida (KX691634) Aurelia relicta (KX691623 Aurelia coerulea (KX691644) Aurelia sp. 1 Incheon-2006 (EU276014) Deepstaria enigmatica (HM194844) Tiburonia granrojo (AY149900) -/0.82 (A) (B) page 9 of 15Zoological Studies 57: 51 (2018)