Holocephalocotyle monstrosae n. gen. n. sp. (Monogenea, Monocotylidae) from the olfactory rosette of the rabbit fish, Chimaera monstrosa (Holocephali, Chimaeridae) in deep waters off Algeria

Based on a molecular and morphological study, a new monocotylid genus, Holocephalocotyle n. gen. is proposed to accommodate Holocephalocotyle monstrosae n. sp., found on the olfactory rosette of the rabbit fish, Chimaera monstrosa Linnaeus (Chondrichthyes, Chimaeridae), from the Mediterranean Sea off Algeria. Identification of fish hosts was confirmed by molecular barcoding of the COI gene. A partial 28S rDNA sequence (D1–D2 domain) of Holocephalocotyle monstrosae was obtained; it was distinct from all known monocotylid sequences (p-distance: 15.5–23%). A phylogenetic tree constructed from available monocotylid sequences showed that Holocephalocotyle monstrosae was included, and basal, in a robust group including species of Merizocotyle, Mycteronastes and Empruthotrema, confirming that the species is a member of the Merizocotylinae. The new genus is unique among the Merizocotylinae in having a distinctive pattern of haptoral loculi with one central, five peripheral and seven “interperipheral loculi” partially inserted between peripheral loculi and a compartmentalised sclerotised male copulatory organ. The diagnosis of the Merizocotylinae is amended to include this new genus. The new genus represents the second monocotylid genus recorded from holocephalans.


Introduction
The Holocephali or chimaerid fishes form a small and enigmatic subclass of Chondrichthyes, hypothetically diverged from a common ancestor with the elasmobranch fishes about 413 Ma ago [27]. Holocephalans are hosts of unique parasite groups including the Chimaericolidae (Monogenea) and the Gyrocotylidea (Cestoda). The rabbit fish, Chimaera monstrosa, occurring in the eastern Atlantic Ocean and Mediterranean Sea, is the most studied species among holocephalans [16,18,23,36,40]. Among metazoan parasites, only two monogeneans were reported from this fish namely, Chimaericola leptogaster (Leuckart, 1830) (Chimaericolidae) from the gills and Calicotyle affinis Scott, 1911 (Monocotylidae) from the cloaca. However, the olfactory organ has been neglected in previous parasitological examinations; this site could represent, as for numerous elasmobranchs, a potential site for the discovery of new species of monogeneans.
During a collaborative study of the helminth parasites of chondrichthyan hosts of Algeria (western Mediterranean), a species of the Merizocotylinae Johnston and Tiegs, 1922 was recovered from the olfactory chamber of the rabbit fish. In the present study, integrative taxonomy based on molecular and morphological analyses was used to provide a formal description of a new species and the proposal of a new genus for its accommodation. This paper belongs to a series of papers dealing with the morphological and molecular description of monogeneans from off the southern Mediterranean coast [2][3][4][5][6][7]22].
All fish caught were dissected shortly after capture and examined for parasites in the laboratory. The olfactory chamber was entirely removed and placed in Petri dishes filled with filtered sea water. The olfactory rosette was scraped out carefully, under a stereomicroscope and monogeneans were removed from the mucus with the help of a dissection needle and aspirated with a pipette. Some living specimens were narcotised with menthol crystal in a 5-10% solution [28], before fixation, to unfold the haptor.

Molecular barcoding of host fish
For two individual fish, total genomic DNA was isolated from absolute ethanol fixed olfactory lamellae using a QIAamp DNA Mini Kit, as per the manufacturer's instructions. The 5 0 region of the mitochondrial cytochrome c oxidase (COI) subunit I gene was amplified with the primers FishF1 (5 0 -TCAACCAACCACAAAGACATTGGCAC-3 0 ) and FishR1 (5 0 -TAGACTTCTGGGTGGCCAAAGAATCA-3 0 ) [39]. PCR reactions were performed in 20 lL, containing 1 ng of DNA, 1Â CoralLoad PCR buffer, 3 mM MgCl 2 , 66 lM of each dNTP, 0.15 lM of each primer, and 0.5 units of Taq DNA polymerase (Qiagen). The amplification protocol was 4 min at 94°C, followed by 40 cycles at 94°C for 30 s, 48°C for 40 s, and 72°C for 50 s, with a final extension at 72°C for 7 min. PCR products were purified (Ampure XP Kit, Beckman Coulter) and sequenced in both directions on a 3730xl DNA Analyzer 96-capillary sequencer (Applied Biosystems). We used CodonCode Aligner version 3.7.1 software (CodonCode Corporation, Dedham, MA, USA) to edit sequences, which were 651 bp in length, compared them to the GenBank database content with BLAST, and deposited them in GenBank under accession numbers MN397913 and MN397914. Species identification was confirmed with the BOLD identification engine [41].

Molecular analysis of monogeneans
DNA was extracted using a QiaAmp DNA Micro kit (Qiagen). A 28S rDNA fragment of 884 bp was amplified using the universals primers C1 0 (5 0 -ACCCGCTGAATTTAAG-CAT-3 0 ) and D2 (3 0 -TCCGTGTTTCAAGACGG-5 0 ) [19]. PCR reactions were performed in a final volume of 20 mL, containing: 1 ng of DNA, 16CoralLoad PCR buffer, 3 mM MgCl 2 , 66 mM of each dNTP, 0.15 mM of each primer, and 0.5 units of Taq DNA polymerase (Qiagen). Thermocycles consisted of an initial denaturation step at 94°C for 1 min, followed by 40 cycles of denaturation at 94°C for 30 s, annealing at 60°C, for 30 s, and extension at 72°C for 1 min. The final extension was conducted at 72°C for 7 min. PCR products were visualised on a 1.5% agarose gel, purified and directly sequenced in both directions on a 3730xl DNA Analyzer 96-capillary sequencers (Applied Biosystems) at Eurofins Genomics. Sequences were edited and assembled using CodonCode Aligner software (CodonCode Corporation, Dedham, MA, USA), and compared to the GenBank database content with BLAST. Sequences from three individual monogeneans were obtained and were found to be identical; they were deposited in GenBank under accession number MN412655-MN412657.

Trees of monogeneans and distances
A tree was constructed from our new sequence and 28S sequences of monocotylids (Table 1). We used almost all sequences available in GenBank, with the exception of sequences that were too short, such as those of Loimosina sp. (KF908848), Potamotrygonocotyle tsalickisi (JN379513), and Calicotyle affinis (AF382061), the latter of which did not align well (but five other species of Calicotyle were included in the analysis); a sequence of the microbothriid Leptocotyle minor was used as the outgroup. The dataset included 32 nucleotide sequences. Both extremities of sequences were trimmed to obtain a clean matrix and thus there were 744 positions in the dataset. After estimating the best model with MEGA7 [26], the tree was inferred using the Maximum Likelihood method based on the General Time Reversible model [32] with gamma distribution and invariant sites (GTR + G + I) in MEGA7 [26], with 200 replications. The neighbour-joining method was also used for comparison in MEGA7, with bootstrap calculated on 2000 replicates. Distances between sequences (p-distances) were computed from the same dataset with MEGA7 [26].

Morphological analysis
Five specimens were prepared as whole mounts in Malmberg solution (m) [29] to study the morphology of anchors and the male copulatory organ. Forty specimens were flattened under slight coverslip pressure and fixed in 70% ethanol and thereafter stained with acetic carmine (c) before being dehydrated in a graded ethanol series, cleared in clove oil, and permanently mounted in Canada balsam.
Measurements were obtained using drawing tube and ruler calibrated using a microscopic graduated slide. They represent straight-line distances between extreme points.
Measurements are in micrometers, and indicated as means ± standard deviation (if n > 30), and between parentheses the range and number (n) of structures measured. Because measured lengths may vary depending on how specimens are prepared and the degree to which they are flattened [2,21], they are given separately for specimens mounted respectively in Malmberg's medium (m) or stained in carmine and mounted in Canada balsam (c).
Drawings of the parasite were made using a Leitz microscope equipped with a drawing tube and redrawn afterwards on a computer with Adobe Illustrator. Terminology of the olfactory organ of Chimaera monstrosa follows that described by [17]. Haptoral terminology follows that presented and illustrated by [11,25].
The term uncinuli (singular uncinulus) is used for marginal hooklets after Pariselle & Euzet (1995) [37]. Uncinuli numbering follows that proposed by   [25]. We propose the term interperipheral loculi for the loculi partially inserted between peripheral loculi in the haptor. Type specimens of the new taxa were deposited in the Collections of the Muséum National d'Histoire Naturelle (MNHN), Paris.

Molecular analysis Molecular identification of fish
The provisional identification of fish species using morphological characteristics was reconfirmed by a DNA barcoding approach. BLAST analysis of the COI sequences of sequences of two host fish specimens with NCBI and BOLD database showed sequence similarity values of, respectively 99.38% and 100% with Chimaera monstrosa. The BOLD database [41] includes many sequences with published information for this species and we are thus confident that the identification is valid.

Phylogenetic analysis of monogeneans from 28S sequences
In our dataset, the genetic distance between our new sequences and all monocotylid sequences ranged from 15.5% to 23%, with the closest sequence being Empruthotrema dasyatidis (15.5%) and the most different sequences being Dendromonocotyle ardea (23%); this clearly showed that our species is different from all other sequenced monocotylid species.
A tree built from 32 sequences, with Leptocotyle minor (Microbothriidae) as the outgroup and 31 monocotylid sequences including the new sequence from our species (Table 1), provided the following results (Fig. 1). There was a total of 744 positions in the final dataset. The general topology of the tree, constructed with the Maximum Likelihood method, showed general agreement with the classification of species into genera, with most genera monophyletic; however, some nodes showed low support. A branch of the tree included species of Merizocotyle, Mycteronastes and Empruthotrema (three genera which are members of the Merizocotylinae) and our new sequence. An analysis based on the NJ method (not shown) showed a slightly different topology of the tree but the Merizocotylinae clade was also clearly defined. Support for the monophyly of the Merizocotylinae clade was high (78% of trees with ML, bootstrap 78 with NJ), strongly suggesting that our new species is a member of the subfamily Merizocotylinae. Within this clade, the new species is sistergroup to a clade including all other species (three members of Merizocotyle, one Mycteronastes and two members of Empruthotrema). Within this clade, members of Merizocotyle apparently do not constitute a monophylum, with Merizocotyle sinensis being a sister-group of Mycteronastes icopae. Etymology: The generic name was constructed based on the name of the subclass of the host (Holocephali) and the endingcotyle. Gender: feminine.
Molecular sequence data: The 884 bp ribosomal DNA sequences covering the D1-D2 domains of the 28S rDNA gene of three specimens were deposited in GenBank under accession numbers MN412655-MN412657. The individual monogenean barcoded as MN412656 was from the individual fish barcoded as MN397913.
Ovary loops right intestinal caecum dorsoventrally and narrows to form oviduct (Fig. 2B). Oviduct receives duct from vagina and common vitelline ducts in glandular junction.

Differential diagnosis
Holocephalocotyle monstrosae n. gen. n. sp. can be differentiated from the other species of the 29 genera within the Monocotylidae on the basis of the arrangement of the haptoral loculi. The haptoral loculi have a distinctive pattern in which the interperipheral loculi are partially inserted between peripheral loculi (Fig. 2). This feature cannot be accommodated into any of the established genera in the Monocotylidae.
According to the presence of three distinct apertures on each side of the anterior margin into which open ducts from the lateral glands containing a needle-like secretion together with oötype with a descending and ascending limb, we place the parasite of Chimaera monstrosa among the Merizocotylinae [13]. Within this subfamily this parasite has a very unique haptor, which allows it to be considered a new species, type of a new genus. This feature was included in the diagnosis of the subfamily to accommodate this new species in a new genus. The amendment of the subfamily was supported by the phylogenetic tree. With the erection of Holocephalocotyle n. gen., the number of genera belonging to the Monocotylidae increases to 30.

Discussion
Interperipheral loculi of Holocephalocotyle monstrosae may represent marginal loculi that have been shifted between the peripheral loculi. In the absence of a phylogenetic analysis conducted using morphological characters, homology of these characters cannot be assumed.
Based on a phylogeny generated from morphological data, Chisholm et al. [11] Burton, 1990 and Thaumatocotyle Odhner, 1910 so the three genera were synonymised with the senior Merizocotyle.
Chisholm et al. [10] provided a phylogenetic analysis of the Monocotylidae inferred from 28S rDNA and concluded that Merizocotyle was paraphyletic and suggested that Mycteronastes and Thaumatocotyle should perhaps be resurrected as valid genera. Neifar et al. [35], de Buron & Euzet [15], and Marie & Justine [30] recognised Thaumatocotyle as a valid genus. Kritsky et al. [24] resurrected Mycteronastes and amended the diagnosis of this genus.
The resurrection of Thaumatocotyle and Mycteronastes seems to be supported by host-parasite phylogenetic relationships. Species of Merizocotyle (sensu Cerfontaine) are usually parasites of Rajidae, members of Thaumatocotyle are all parasites of Dasyatidae, and members of Mycteronastes are parasites of Rhinopristiformes (as proposed by [1,31]). Merizocotyle urolophi, a parasite of Urolophus paucimaculatus (Urolophidae), may represent a separate genus, as suggested by Chisholm & Whittington [14], but closely related to Thaumatocotyle as shown by the current phylogeny.
The position of Merizocotyle sinensis is enigmatic. This species is a parasite of Platyrhina sinensis, a Rhinobatidae and not a Dasyatidae as indicated by Chisholm & Whittington [13]. It has a haptor with seven peripheral loculi like the other species of Merizocotyle (sensu Cerfontaine), but our phylogenetic analysis, based on a partial 28S rDNA unpublished sequence deposited in GenBank, placed this species as a sister-group to Mycteronastes icopae, and not with the two other species of Merizocotyle (Fig. 1); if Merizocotyle sinensis is kept in Merizocotyle, Merizocotyle is not monophyletic based on our analysis. Merizocotyle sinensis might represent a distinct genus as suggested by Kritsky et al. [24].
Our distance analysis also showed that the sequence of the partial 28S rDNA of Holocephalocotyle monstrosae is distinct from all other monocotylids by a significant distance (15.5-23%). Our phylogenetic analysis based on partial 28S sequences confirmed that it is a member of the Merizocotylinae, and that it was the sister-group of the other species of Merizocotylinae included in the analysis; however, available sequences of Merizocotylinae included only members of three genera, Mycteronastes, Merizocotyle and Empruthotrema, which constitute only a subset of the seven genera previously known in this subfamily.
The microhabitat seems to have an important impact on the evolution of monocotylids [12]. cavity. Locations of merizocotylines in other microhabitats than the nasal cavity have been widely discussed by Chisholm & Whittington [13] and might be erroneous. Species of Calicotylinae have been described from the cloaca, rectum, rectal gland, spiral valve, oviduct and inner wall of the body cavity [9]. Species of the other subfamilies, Dasybatotreminae, Decacotylinae, Heterocotylinae, Monocotylinae and Euzetiinae, were reported from different microhabitats such as the skin and gills. The haptor of Holocephalocotyle monstrosae attaches to the free edge of the primary nasal fold by folding longitudinally and embedding the blade of the anchors into secondary folds (Fig. 3). Holocephalocotyle monstrosae does not attach to the glass of the Petri dish by its haptor but rather the haptor remains folded with the blades of the anchors protruding. The same mode of attachment was observed for Monocotyle corali [8], a gill parasite of Pastinachus sephen. Presence of the loculi and attachment by folding ensure effective anchoring for H. monstrosae. The olfactory rosette is exposed to a strong water current. Howard [20] estimated that 20-50% of inspired flow of respiratory water passes through secondary folds of the olfactory rosette and then passes into the gills. It remains to be studied how the parasitism by Holocephalocotyle monstrosae may affect the respiratory efficiency of the host.