New morphological and molecular data for Xystretrum solidum (Gorgoderidae, Gorgoderinae) from Sphoeroides testudineus (Tetraodontiformes, Tetraodontidae) in Mexican waters

Abstract Adults of trematodes in the genus Xystretrum Linton, 1910 (Gorgoderidae, Gorgoderinae) are parasites found exclusively in the urinary bladders of tetraodontiform fishes. However, limited and unclear morphological data were used to describe the type species, X. solidum Linton, 1910. Here, we present the first detailed morphological information for a member of Xystretrum. Morphological characters were described using light and scanning electron microscopy (SEM) of Xystretrum specimens from Sphoeroides testudineus (Linnaeus) (Tetraodontiformes, Tetraodontidae), collected at six localities off the northern Yucatan Peninsula coast of the Gulf of Mexico. We also compared sequence fragments of the 28S (region D1–D3) ribosomal DNA and mitochondrial Cytochrome c oxidase subunit 1 (COI) gene with those available for other gorgoderine taxa. We assigned these Xystretrum specimens to X. solidum, despite the incompleteness of published descriptions. The data provide a foundation for future work to validate the identities of X. solidum, X. papillosum Linton, 1910 and X. pulchrum (Travassos, 1920) with new collections from the type localities and hosts. Comparisons of 28S and COI regions described here also provide an opportunity to evaluate the monophyletic status of Xystretrum.


Introduction
proposed the genus Xystretrum Linton, 1910 (Gorgoderidae, Gorgoderinae) to include two new trematode species, X. solidum Linton, 1910 (type species) and X. papillosum , which, as adults, are parasites of tetraodontiform fishes of the families Balistidae (Balistes capriscus Gmelin [as B. carolinensis] from off Bermuda) and Ostraciidae (Lactophrys triqueter (Linnaeus) from Dry Tortugas, Florida, USA). Unfortunately, the original descriptions of both species are incomplete and unclear. This has resulted in taxonomic confusion when new species have been proposed. Several Xystretrum species have subsequently been reported, synonymized and later resurrected by some (but not all) authors, while the validity of others remain doubtful (Linton 1907;MacCallum 1917;Manter 1947;Yamaguti 1971;Siddiqi and Cable 1960;Nahhas and Cable 1964;Overstreet 1969).
The most reliable list of species of Xystretrum available is the public resource database of the World Register of Marine Species (WoRMS 2020), which lists 14 accepted species. Of these species, X. solidum and X. papillosum, along with X. pulchrum (Travassos, 1920) Manter 1947, are reported from the Northwest Atlantic Ocean and Gulf of Mexico Travassos 1922;Manter 1947;Mendoza 2016). However, X. pulchrum was also inadequately described from Sphoeroides testudineus (Linnaeus) (Tetraodontiformes, Tetraodontidae) collected in the southwestern Atlantic Ocean (Manquinhos State, southeastern Brazil) (Travassos 1922;Manter 1947) and its incomplete description has generated synonyms (e.g., Siddiqi and Cable 1960;Overstreet 1969). Xystretrum pulchrum was reported from the type locality of X. papillosum (i.e., Tortugas, Florida) and from the North Pacific Ocean (Hawaii), but the morphological data used to separate the two species are questionable (Travassos 1922;Manter 1947;Hanson 1955;Yamaguti 1970). Thus, the morphological descriptions of X. solidum, X. papillosum and X. pulchrum remain incomplete.
Despite the scarce taxonomic information from the western Atlantic, Overstreet et al. (2009) reported X. solidum parasitizing the kidneys and urinary bladder of five tetraodontiform fish species from four families (i.e., B. capriscus [Balistidae], L. triqueter [Ostraciidae], S. spengleri Bloch, S. testudineus [Tetraodontidae] and Stephanolepis hispidus (Linnaeus) [Monacanthidae]) distributed from Bermuda, the Caribbean Sea and the North Gulf of Mexico to the Atlantic coast of South America. Cutmore et al. (2013) provided the first genetic sequence data for an Atlantic species, tentatively identified as X. solidum, from S. testudineus collected in the Florida Keys near (200 km in an approximately straight line) Dry Tortugas, the type locality of X. papillosum. Recently, Pérez-Ponce de León and Hernández-Mena (2019) provided a second genetic sequence published as X. solidum from Balistes vetula (Linnaeus) (Balistidae) from Puerto Morelos, Quintana Roo, Mexico, from the Mexican Caribbean.
Several S. testudineus from the coasts and lagoons of the Northern Yucatan Peninsula, Mexico were examined for parasites between 1995 and 2013. Gorgoderids were recovered from the urinary bladder of these hosts and preliminarily identified as Phyllodistomum sp. (Tello 1999;Pech et al. 2009;Sosa-Medina et al. 2014. Following a study program to fully characterize parasite biodiversity (Pérez-Ponce de León and Nadler 2010; Nadler and Pérez-Ponce de León 2011), DNA sequences of the 28S gene were obtained from these "Phyllodistomum sp." and compared to GenBank sequences available for gorgoderines. A high similarity (BLAST scores) between the "Phyllodistomum sp." and trematodes tentatively identified as X. solidum by Cutmore et al. (2013) suggests that records for these Phyllodistomum should be reassigned to the genus Xystretrum.
As Nadler and Pérez-Ponce de León (2011) pointed out, phylogenetic analyses are essential for correctly characterizing a species (including cryptic species) and data from morphological approaches; e.g., morphology, morphometry and microphotographs of scanning electron microscopy (SEM) should be corroborated using molecular-based results. Both morphological and molecular information are lacking for three species of Xystretrum found in the Atlantic Ocean (i.e., X. solidum, X. papillosum and X. pulchrum). The intention here is to provide morphological descriptions to support the reassignment of trematodes previously identified as Phyllodistomum to Xystretrum and to provide new morphological and sequence data to facilitate future revisions of the genus Xystretrum.

Collection of hosts and trematodes
Trematode specimens in this study were collected from the urinary bladder of Sphoeroides testudineus. Hosts were collected between 1998 and 2016 (collection permit PPF/DGOPA-070/16 issued by Comisión Nacional de Acuacultura y Pesca, Mexico) at six localities in and off the northern Yucatan Peninsula

Morphological data and morphometric analyses
Unflattened trematode specimens were stained with Mayer's paracarmine and mounted on permanent slides using Canada balsam. Specimens were measured, and drawings were made with the aid of a drawing tube attached to an Olympus BX50 microscope; measurements are presented in micrometres (μm) as ranges followed by the means in parentheses. For the SEM study, specimens were dehydrated through a graded series of ethyl alcohols and critical point dried with carbon dioxide. Specimens were mounted on metal stubs with silver paste, then coated with gold and examined in a Philips XL30 ESEM for variable pressure SEM at 10 kV. Trematodes were identified following/contrasting the taxonomic criteria of Linton (1907;, Travassos (1922), Manter (19471972), Winter (1959), Overstreet (1969), Yamaguti (1971), Campbell (2008) and Madhavi and Bray (2018). Holotype, labelled as X. papillosum (No. 1321174; now in the Smithsonian Institution National Museum of Natural History (NMNH) (ex USNM Helm. Coll. No. 8426) from Lactophrys triqueter from Dry Tortugas, Florida, USA, was studied to compare with the newly collected specimens. The X. solidum holotype described by Linton in 1907 was not found in the scientific collections in the USA, Europe or Australia, and should be considered lost. Several specimens collected for morphological analysis were deposited as voucher specimens in the Colección Helmintológica del CINVESTAV (CHCM), Departamento de Recursos del Mar, Centro de Investigación y de Estudios Avanzados del Instituto Politécnico Nacional, Unidad Mérida, Yucatán, Mexico (Tab. 1). Morphological measurements obtained in this study were compared with those of the 14 congeneric Xystretrum spp. (Suppl. material 1: Tab. S1).

DNA extraction, PCR amplification and sequencing
Deoxyribonucleic acid (DNA) was extracted from individual adult trematodes; DNA extraction was performed using the DNeasy blood and tissue extraction kit (Qiagen, Valencia, CA, USA) following the manufacturer's instructions. Partial sequences of the 28S (region D1-D3) ribosomal DNA were amplified by Polymerase Chain Reaction (PCR) (Saiki et al. 1988) using 28sl fwd (5´-AAC AGT GCG TGA AAC CGC TC-3´) (Palumbi 1996) and LO rev (5´-GCT ATC CTG AG(AG) GAA ACT TCG-3´) (Tkach et al. 2000). The primers JB3 fwd (5´-TTT TTT GGG CAT CCT GAG GTT TAT-3´) (Morgan and Blair 1998) and CO1R trema rev (5´ -CAA CAA ATC ATG ATG CAA AAG G-3´) (Miura et al. 2005) were used for the COI fragment. The reactions were prepared using the Green GoTaq Master Mix (Promega). This procedure was carried out using an Axygen Maxygen thermocycler. The PCR cycling conditions were as follows: for COI, an initial denaturing step of 3 min at 94 °C, followed by 35 cycles of 92 °C for 30 sec, 47 °C for 45 sec and 72 °C for 90 sec, and a final extension step at 72 °C for 10 min; for 28S, an initial denaturing step of 5 min at 94 °C, followed by 35 cycles of 92 °C for 30 sec, 50 °C for 45 sec and 72 °C for 90 sec, and a final extension step at 72 °C for 10 min. The PCR products were analyzed by electrophoresis in 1% agarose gel using the TAE 1X buffer and observed under UV light using the QIAxcel Advanced System. The purification and sequencing of the PCR products were carried out by Genewiz, South Plainfield, NJ, USA (https://www.genewiz.com/).

Molecular data and phylogenetic reconstruction
To obtain the consensus sequences of specimens of Xystretrum, we assembled and edited the chromatograms of forward and reverse sequences using the Geneious Pro v.5.1.7 platform (Kearse et al. 2012). The 28S and COI partial sequences generated during this study were aligned with sequences of gorgoderids and representative outgroup sequences of members of the Allocreadiidae, Callodistomidae, Dicrocoeliidae and Encyclometridae (see GenBank accession numbers in Figs 2, 3) used previously Phyllodistomum lacustri (HQ325010-HQ325028) Phyllodistomum staffordi (EF032692, HQ325027, HQ325028)  Petkevičiūtė et al. (2015) and Urabe et al. (2015), using an interface available in MAFFT v.7.263 (Katoh and Standley 2016), an "auto" strategy and a gap-opening penalty of 1.53 with Geneious Pro, and a final edition by eye in the same platform. The Gblocks Website v.0.91b (Castresana 2000;Talavera and Castresana 2007) was used to remove ambiguously aligned regions of 28S. To evaluate the sequence and molecular marker utility for phylogenetic analyses at the intended taxonomic level (family level for the complete- outgroup dataset and genus level for the Xystretrum dataset), we tested the nucleotide composition homogeneity within each data alignment (28S and COI matrix data), using chi-squared metric provided in the program TreePuzzle v.5.3.rc (Schmidt et al. 2002). The software jModelTest v.2.1.3 (Darriba et al. 2012) was used to select evolution models through the Bayesian Information Criterion (BIC) (Schwarz 1978) for each dataset separately (28S and COI). The nucleotide substitution model that best fit the 28S dataset was TVM+I+G (Posada 2003). The COI dataset was partitioned into first-, second-and third-codon positions with the appropriate nucleotide substitution model implemented for each codon position (TrN+I for the first [Tamura and Nei 1993]; TPM3uf+I for the second [Kimura 1981]; and HKY+I for the third codon position [Hasegawa et al. 1985]). Furthermore, the net evolutionary distances between Xystretrum taxa, using p-value with variance estimation, with the Bootstrap method (500 replicates) and with a nucleotide substitution (transitions + transversions) uniform rate, were estimated for the 28S fragment in MEGA v.7.0 (Kumar et al. 2016). Phylogenetic trees were reconstructed for each gene separately (28S and COI), to test the monophyly of X. solidum analyzed in this study. Phylogenetic tree reconstructions were carried out using Bayesian Inference (BI) in MrBayes v.3.2.3 (Ronquist et al. 2012), with two parallel analyses of Metropolis-Coupled Markov Chain Monte Carlo (MCMCMC) for 20 × 10 6 generations each. Topologies were sampled every 1000 generations and the average standard deviation of split frequencies was observed to be less than 0.01, as suggested by Ronquist et al. (2012). A majority consensus tree with branch lengths was reconstructed for the two runs after discarding the first 5000 sampled trees. The robustness of the clades was assessed using Bayesian Posterior Probability (PP), where PP > 0.95 was considered strongly supported. The Bayesian phylogenetic reconstructions were run through the CIPRES Science Gateway v.3.3 (Miller et al. 2010).

DNA sequences and dataset analyses
In total, 12 bi-directional 28S and COI sequences were obtained from three individual adults of X. solidum. The final lengths (in number of base-pairs) of the 28S ribosomal sequence fragment were 892 (for two sequences) and 899 (for one sequence), with zero genetic variation, either among the new sequences or in the published 28S sequences of X. solidum (GenBank accession numbers MK648284 and KF013188). The total alignment length following the Gblocks exclusion was 814 bp. Nucleotide sequence variation in the 28S alignment from gorgoderids (excluding the outgroup taxon) in-cluded 330 conserved sites, 483 variable sites, 410 parsimony-informative sites and 73 singleton sites. The COI dataset consisted of 309 bp with a genetic distance of 0.3% between the three mitochondrial sequences. Nucleotide sequence variation (excluding the outgroup taxa) for each partition from COI (first, second and third codon positions) was 70/92/12 conserved, 33/11/91 variable, 29/7/79 parsimony-informative and 4/4/12 single sites, respectively.

Phylogenetic reconstructions
We inferred the phylogenetic relationships from the 28S and COI sequence matrices separately. The 28S gene dataset contained 46 taxa (150 sequences) and the COI contained 18 taxa (63 sequences). Figures 2 and 3 show the phylogenetic topologies resulting from 28S and COI dataset analyses, respectively. The 28S tree shows that the sequences generated in this study form a clade with the sequences from the material tentatively identified as X. solidum (sequence KF013188) by Cutmore et al. (2013) and X. solidum (sequence MK648284) from B. vetula in the Gulf of Mexico (see below). Furthermore, in the 28S tree, species of Xystretrum form a monophyletic group, with high nodal support values (PP ≥ 0.95) (Fig. 2). The COI tree shows that all sequences of X. solidum form a clade (Fig. 3). Based on the phylogenetic trees constructed from the 28S dataset, the taxa most closely related to Xystretrum spp. are members of the genus Phyllodistomum (i.e., Phyllodistomum angulatum Linstow, 1907, P. macrocotyle Lühe, 1909 andPhyllodistomum sp.), whereas the relatives of the species X. solidum, based on the COI dataset, were Phyllodistomum centropomi Mendoza-Garfias & Pérez-Ponce de León, 2005 and Phyllodistomum sp. (Fig. 3). The differences in topology between the two trees are most likely due to the differences in the taxa included in the two datasets. The genetic distance values from the 28S dataset of X. solidum, when compared with Xystretrum spp., were 2.74 %, 3.50 %, 5.02 % and 5.02 % for X. caballeroi, Xystretrum sp., Xystretrum sp. 1 and Xystretrum sp. 2, respectively.

Discussion
The morphologies of the trematodes examined in this study are consistent with those of the genus Xystretrum provided by Campbell (2008); i.e., intestinal caeca forming a cyclocoel, presence of a pseudosinus-sac, and a corrugated hindbody demarcated by folds. This study adds detail to those descriptions by providing new morphological and morphometric data and revealing characters not previously described, such as the number of papillae on the tegument and oral sucker. However, the published descriptions for the species of Xystretrum are very basic, particularly from the American Atlantic, i.e., X. solidum, X. papillosum and X. pulchrum. Body size range (i.e., length) is the primary character used to distinguish these three species. Xystretrum solidum is the smallest (i.e., 1750), X. papillosum is intermediate (i.e., 2100) (a size that corre-sponds to the samples analyzed in this study) and X. pulchrum is the largest (i.e., 4500). However, since there are no data on intraspecific morphological variation for the three species, it is impossible to decide whether body size is sufficient for the correct identification of our specimens. There are several impediments to species-level identification, including: 1) scarce morphological data from congeners, and particularly the limited measurements for X. solidum and X. papillosum, 2) voucher material (holotype) apparently lost for X. solidum, and 3) incongruences in the host specificity patterns previously reported for Xystretrum spp. at family level. For these reasons and based on the genetic similarities and the phylogenetic relationships obtained in this study, we agree with the proposal of Cutmore et al. (2013) and identify our samples as X. solidum.
Based on the observation of material from this study, plus the holotype of X. papillosum (voucher 1321174), we found that X. solidum presents a fluted tegument on the hindbody and that along the dorsoventral margin there are short dense fringe papillae (only readily visible using SEM, see Suppl. material 2: Fig. S1I), which were referred to as "hair-like spines" by Manter (1947; page 330) but without mentioning their exact location. The material examined in this study shows a relatively broad range of polymorphism.
It is necessary to collect new X. solidum specimens from the original host (i.e., B. capriscus) and the type locality (i.e., off Bermuda) to compare their morphological measurements with our samples. Also, it is necessary to explore the possible presence of X. papillosum from L. triqueter co-distributed with B. capriscus off Bermuda, as part of a taxonomic revision of the genus Xystretrum, taking into consideration the morphological data presented here. In parallel, future revisions should seek to distinguish or synonymize X. solidum, X. papillosum and X. pulchrum, while being sensitive to the potential presence of cryptic species.
To date, 14 species of the genus Xystretrum are considered valid (WoRMS 2020). From a purely biogeographical standpoint, most of these species are non-conspecific with our material, as they have been reported from unique marine regions other than the Gulf of Mexico. This gives them a set of host-associations and biogeographical differences with respect to the remaining species. Thus, species such as X. chauhani Ahmad, 1982, X. manteri Ahmad, 1982, X. overstreeti Ahmad, 1982, X. srivastavai Ahmad, 1982and X. thapari Ahmad, 1982 appear to be confined to the Arabian Sea; X. abalistis Parukhin, 1964 occurs in the Gulf of Tonkin (South China Sea); X. triacanthi Ahmad & Gupta, 1985 occurs off the Indian coast of the Bay of Bengal (Indian Ocean) (Madhavi and Bray 2018); X. moretonense Manter, 1972 andX. plicoporatum Manter, 1972 inhabit Australian waters (Manter 1972);andX. caballeroi Bravo-Hollis, 1953 andX. hawaiiense Yamaguti, 1970 are distributed in various parts of the Pacific Ocean (Bravo-Hollis 1953;Winter 1959;Parukhin 1964;Yamaguti 1970;Arthur and Te 2006;Mendoza 2016). An additional marine region, extending through the Western Atlantic from Brazil to Bermuda and the Gulf of Mexico, harbors the species X. solidum, X. papillosum and X. pulchrum described from members of the Balistidae, Ostraciidae and Tetraodontidae, respectively. The fact that the latter species are geographically sympatric and were described some time ago, resulting in a confused taxo-nomic characterization, suggests that the published records may indicate incorrect host assignments and host localities.
In the phylogenetic tree obtained from the 28S dataset, all members of the genus Xystretrum included in the analysis formed a well-supported clade, but without nodal support with their sister clade. A similar result has been reported in previous phylogenetic analyses carried out for similar taxa using the same gene (e.g., Cutmore et al. 2013;Razo-Mendivil et al. 2013;Pérez-Ponce de León et al. 2015;Petkevičiūtė et al. 2015;Urabe et al. 2015;Stunžėnas et al. 2017). Cutmore et al. (2013) detected that the species of Xystretrum genus are not closely related to marine representatives of the family Gorgoderidae. At the present time, species of Xystretrum appear related to the freshwater Phyllodistomum spp. Even though the Xystretrum clade did not exhibit a high nodal support based on the 28S dataset in this study, a phylogenetic relationship with freshwater phyllodistomid trematodes was observed, e.g., Phyllodistomum sp., P. angulatum and P. macrocotyle (see also Stunžėnas et al. 2004;Petkevičiūtė et al. 2015). Because of the incomplete dataset of gene sequences for Xystretrum spp., the phylogenetic relationships of this genus remain unclear. Based on the 28S phylogenetic tree topology, X. solidum (found in Tetraodontidae and Balistidae) seems to be related to X. caballeroi (although lacking nodal support). Current phylogenetic information confirms another sub-clade, which includes samples of Xystretrum associated with the Balistidae from the Coral Sea, Australia and the Indian Ocean off Western Australia (as Xystretrum sp., Xystretrum sp. 1 and Xystretrum sp. 2 in Cutmore et al. [2013]).
The COI phylogenetic topology shows a clade with X. solidum from this study, and the clade formed by the freshwater taxa P. centropomi + Phyllodistomum sp. (from Urabe et al. [2015]) as sister, although this relationship does not have nodal support. However, based on COI phylogenetic topology plus 28S, it is possible to suggest a diversification of the most recent common ancestor of Xystretrum from freshwater to marine environments, and a subsequent diversification in tetraodontiforms via host-switching events. A similar evolutionary process of transition from freshwater to marine environments has also been suggested for other platyhelminth groups (e.g., Álvarez-Presas et al. 2008;Van Steenkiste et al. 2013;Martínez-Aquino et al. 2017).
As indicated by Cutmore et al. (2013), Xystretrum occurs only in marine fishes of the order Tetraodontiformes; however, patterns of host specificity at the family level for each species of Xystretrum are not currently well-defined. For example, in the reported cases of X. solidum from the American Atlantic, Overstreet et al. (2009) recorded X. solidum associated with five tetraodontiform fish species included in four families, including the host species from which X. papillosum and X. pulchrum were described, i.e., L. triqueter and S. testudineus, respectively. Furthermore, the recently published sequence by Pérez-Ponce de León and Hernández-Mena (2019) (sequences MK648284) suggests that X. solidum is associated with both the Tetraodontidae and Balistidae. Although there are efforts to build molecular libraries of parasite biodiversity (Pérez-Ponce de León and Nadler 2010; Nadler and Pérez-Ponce de León 2011), DNA sequences of parasites with little or no morphological support continue to be generated, primarily due to incomplete taxonomic descriptions. Here, we pro-vide molecular sequences supported by detailed morphological description, which can provide a foundation for future comparisons and revisions within Xystretrum and the Gorgoderidae.