Diversity of apostome ciliates, Chromidina spp. (Oligohymenophorea, Opalinopsidae), parasites of cephalopods of the Mediterranean Sea

Chromidina spp. are enigmatic apostome ciliates (Oligohymenophorea, Opalinopsidae) that parasitise the renal and pancreatic appendages of cephalopods. Only four species have been described, among which only three have been formally named. No DNA sequence has been reported so far. To investigate Chromidina spp. diversity, we sampled cephalopods in the Mediterranean Sea off Tunis, Tunisia, and identified two distinct Chromidina spp. in two different host species: Loligo vulgaris and Sepia officinalis. From haematoxylin-stained slides, we described morphological traits for these parasitic species and compared them to previous descriptions. We also re-described the morphology of Chromidina elegans (Foettinger, 1881) from Chatton and Lwoff’s original materials and designated a neohapantotype and paraneohapantotypes for this species. We describe a new species, Chromidina chattoni Souidenne, Florent and Grellier n. sp., found in L. vulgaris off Tunisia, and evidence for a probable novel species, found in S. officinalis off Tunisia, although this latter species presents similarities to some morphological stages previously described for Chromidina cortezi Hochberg, 1971. We amplified, for the first time, an 18S rDNA marker for these two Chromidina species. Phylogenetic analysis supports the association of Chromidina within apostome ciliates. Genetic distance analysis between 18S rDNA sequences of representative apostomes indicates Pseudocollinia as the most closely related genus to Chromidina.


Introduction
Species of Chromidina Gonder, 1905 [15] are enigmatic apostome ciliates being, after dicyemids, the most frequently encountered parasites inhabiting the reno-pancreatic appendages of cephalopods [18][19][20]30]. Initially in 1881, Foettinger observed infusoria parasites in Sepia elegans renal appendages. Recognising that these parasites were atypical protozoa, he erected the genus Benedenia [11], named after the embryologist Edouard van Beneden. In 1905, Gonder re-examined these parasites, highlighting their nuclear system organised as a network throughout the cell. Since this type of nuclear system arrangement was considered by Gonder as typical of chromidial systems, he changed the genus name from Benedenia to Chromidina [11]. Currently, 23 cephalopod species have been recorded as hosts of Chromidina spp., of which only three have been formally described: Chromidina elegans (Foettinger, 1881), C. coronata (Foettinger, 1881) and C. cortezi Hochberg, 1971. A fourth has been described morphologically but not named (Table 1) [25]. These ciliates infest mainly pelagic cephalopods, such as epi-meso pelagic squids, cuttlefishes and octopuses. Occasionally, they may be found in the kidneys of benthic hosts, if these hosts include a young nektonic stage during their development [13].
Chromidina spp. have a polymorphic dixenous life cycle, with two different budding processes which are monotomy and palintomy. The adult stage, the vermiform tropho-tomont, has a maximum body length varying from 400 lm to 2,000 lm [20]. Some rare adult stages can have an accelerated growth process. Their size increases so quickly that their length can measure up to 5,000 lm. Given their unusual extended size, these adult stages are called hypertrophonts. The tropho-tomont is uniformly ciliated and has no cytostome [4]. It is attached through its anterior end to the host kidney tissues with its body bathing in the renal fluids, and feeds by nutriment absorption from host cells and fluids [25]. Division by monotomy produces a single long bud from the posterior end, the apotomite, which is morphologically similar to its parent and develops into a second generation of trophotomonts after detachment and colonisation of the host kidney. Division by palintomy produces smaller buds that form a typical chain of individuals attached to the tropho-tomont, which differentiate into tomites. Budding occurs only from the posterior end. The tomite is a small ciliate form with a unique ciliature and a cytostome [4]. When detached, it is believed that the tomite leaves the renal appendages to be released with passage of urine into the sea. This stage is presumed to encyst, as a phoront, and to infest an intermediate host [17,18]. However, no intermediate host has been confirmed so far. Chatton and Lwoff [4] grouped Chromidina within the apostome ciliates even though their life cycle and their morphology show important differences from typical apostomes. The main argument for such an association is that Chromidina tomites share ciliature organisation similar to that of apostome tomites. Besides the reference work of Chatton and Lwoff [4], reports by Hochberg [17,18,20], and the recent description of a novel Chromidina sp. by Landers [25], little is known about this enigmatic genus and notably, no DNA sequence has been reported so far.
In the present study, we examined the Tunisian teuthofauna for infections by Chromidina. We report here the description of two Chromidina species and compare their morphological traits to those of previously described parasites. We provide, for the first time, Chromidina 18S rDNA sequences. In the course of this study, we also re-described Chromidina elegans and designated a neohapantotype and paraneohapantotypes from the Chatton and Lwoff original slide collection [4]. A phylogenetic analysis was performed to evaluate the association of Chromidina with apostomes within the Oligohymenophorea.

Materials and methods
Host sampling and isolation of parasites Cephalopods were obtained from fishermen from the harbour of La Goulette, who collected them off Tunis, in the Mediterranean Sea (36°49 0 9.11 00 N, 10°18 0 22.49 00 E), in January 2013 and January 2014, using traditional earthenware jugs as fishing tools [31]. In total, 5 individuals of Loligo vulgaris and 38 individuals of Sepia officinalis were collected. The cephalopod species identification was based on morphological features relying on cited identification keys [23] and species descriptions (Marine Species Identification portal, http://species-identification.org). The hosts were rapidly dissected and their renal appendages collected. Small pieces of these renal appendages were smeared on glass microscope slides, which were immediately fixed in Bouin's fluid for 24 h then stored in 70% ethanol. The smears were then stained in Ehrlich's Haematoxylin and counterstained in eosin, then mounted with Entellan (Merck, Rahway, New Jersey) [12]. In parallel, additional renal appendage pieces from the same cephalopod host were put in individual Petri dishes in sterile distilled water and observed with a stereomicroscope to collect Chromidina specimens with Pasteur pipettes. These parasites were then transferred to new Petri dishes to be washed several times, in order to eliminate the host tissue possibly remaining attached to the parasites. Washed parasites were then transferred, one by one, to sterile Eppendorf tubes filled with 70% ethanol and were stored at 4°C until use for DNA isolation.

Morphological studies
Smears were observed by light microscope. Images were recorded using a Nikon DXM 1200C camera and processed using ImageJ software (http://imagej.nih.gov/ij/). Measurements were made with the aid of a micrometric slide by using the ImageJ Set Scale module. Target Chromidina spp. were observed on slides prepared from positive hosts; specifically, three infected L. vulgaris individuals (C21, C22 and C23) and one infected S. officinalis individual (S50). Smears of these specimens were deposited in the Protist collection of the Muséum National d'Histoire Naturelle, Paris, France (www.mnhn.fr/fr/collections/ensembles-collections/invertebresmarins/protistes). The morphologies of these Chromidina spp. specimens were compared to descriptions from the literature and to the original Chatton and Lwoff's smeared slides used to describe Chromidina elegans [4], which are deposited in the Protist collection of the Muséum National d'Histoire Naturelle. Since no type slide related to Foettinger's work [11] could be localised, we undertook to re-describe this species to which we associated a neohapantotype and paraneohapantotypes.

Genetic distance between apostome species
18S rDNA sequences from the four Chromidina specimens isolated from the infected hosts and 10 18S rDNA sequences representative of each Apostomatia species, retrieved from the EMBL or GenBank databases, were aligned using the online version of MAFFT, version 7 (http://mafft.cbrc.jp/ alignment/server/; [24]), using the secondary structure of RNA (Q-INS-I option). Evolutionary distances between the sequences were computed pairwise using p-distances and uniform rate analyses in MEGA 6.06 [33]. All ambiguous positions were removed for each sequence pair and there were a total of 1,477 positions in the final dataset. The numbers of base differences per site between sequences were expressed as percentages.

Maximum likelihood and Bayesian rooted trees for Ciliophora
18S rDNA sequences from the four Chromidina specimens, and 25 sequences of representatives of Ciliophora and from the Dinoflagellate Biecheleriopsis adriatica (HG792067) taken as the outgroup, were aligned using the online version of MAFFT, version 7 (http://mafft.cbrc.jp/ alignment/server/; [24]), using the secondary structure of RNA (Q-INS-I option), and the alignment was filtered out using the online version of GBlock [2]. GBlock settings were set to allow moderately strict flanking positions (Maximum number of contiguous non-conserved positions: 8; minimum length of a block: 10; no gap position allowed), yielding a confident alignment of 1,157 positions. A general time-reversible (GTR) substitution model with gamma-distributed rate variation across sites was suggested by JModeltest V2.1.3 as the best-fit model for this alignment [8]. Accordingly, a Bayesian phylogenetic tree was constructed with MrBayes v.3.2.3 [32], using lset nst = 6 rates = Invgamma Ngammacat = 4 parameters. Four simultaneous Monte Carlo Markov chains were run from random trees for a total of 5,000,000 generations in two parallel runs. A tree was sampled every 1,000 generations and 25% of the trees were discarded as ''burn-in''. A consensus tree was constructed from the postburn-in trees using FigTree v1.3.1, and posterior probabilities were calculated in MrBayes. In parallel, Maximum Likelihood analyses were performed using the same alignment and the GTR+G+I model, with MEGA 6.06 [33]. Bootstraps were estimated from 1,000 replicates.  [4,17,29]. Wermel [34] observed C. elegans-like species in Russia on Loligo sp., and Jepps [22] and Clarke [6] in Spirula spirula Linnaeus, 1758 in the Atlantic Ocean.
Prevalence: No information available. Chromidina elegans was initially described by Foettinger in 1881 but under the genus Benedenia Foettinger, 1881 and not Chromidina [11]. In his pioneering work, Foettinger did not designate a type species despite detailed descriptions of C. elegans (formerly B. elegans). We were unable to trace the existence of a corresponding type-or hapantotype-slide in the literature, and the Ciliate Resource Archive database (http://www.uoguelph.ca/~ciliates/) indicates that these are absent for this genus. Later, Chatton and Lwoff provided a detailed description of C. elegans and clearly designated in their monograph this species as the type species for the genus Chromidina, but without indicating any type-or hapantotypeslide [4]. However, Chatton's slide collection was deposited by the French National Centre for Scientific Research in the Protist Collection of the Muséum National d'Histoire Naturelle, Paris, France, in the 1970s. We therefore used Chatton's C. elegans slides to re-examine this parasite. Particular attention was paid to the tropho-tomont and tomite stages found in the renal appendages of Sepia elegans, as these stages were considered representative of the species [11] by Chatton and Lwoff in 1935 [4]. This work enabled us to formally designate a neohapantotype and paraneohapantotypes for the C. elegans species, from this slide collection.
Materials examined: Forty-eight slides corresponding to smears prepared by Chatton and Lwoff from renal appendages of Sepia elegans collected in Banyuls-sur-Mer, France [4], were deposited in the Protist Collection of the Muséum National d'Histoire Naturelle.
Posterior end: Division by palintomy ( Fig. 1F) with chains of primary segments (Fig. 1H), generating up to 24 protomites by fission (Fig. 1I). Protomites with condensed chromatin network (Fig. 1F). Tomites with ellipsoidal form (Fig. 1J) and size of 27. Etymology: The species name was chosen in memory of the French biologist Edouard Chatton for his extensive and remarkable studies on apostomes.
Prevalence: Sixty percent (three out of five specimens of L. vulgaris examined from off Tunis, Tunisia).
Authorship: Note that the authors of the new taxon are different from the authors of this paper, Article 50.1 and Recommendation 50A of the International Code of Zoological Nomenclature [21].
Ciliature: Consisting of 13 dextrally-spiralled kineties originating from the apex (Fig. 2G) and continuing uninterrupted on the entire cell body.
Posterior end: With typical Chromidina division segments that could generate either apotomites by monotony division or tomites by palintomy division (Figs. 2D and 2E).
Diagnosis: The C. chattoni n. sp. tropho-tomont has 13 kineties, which distinguishes it from other reported Chromidina species: 12 kineties for C. cortezi and the Chromidina sp. isolated from Illex coindetti in the Gulf of Mexico, and 14 for C. elegans [4] (Fig. 1E, Table 1). C. coronata has an unreported number of kineties but significantly differs from the other species by its bulbous anterior end covered by discernible superposed rows of dense and elongated cilia, presenting a crown-like aspect, which is the main identification feature for C. coronata (presence of a crown of cilia) [11,17]. In addition, C. chattoni n. sp. differs from C. elegans by a bulb-like anterior end (Figs. 2B, 2C, 2F) versus an inflated club-like anterior end with a distinguishable apical papillum (Figs. 1C and 1D; [4]), and a narrower head (40.7 ± 4.7 lm versus 53.2 ± 11.8 lm, respectively). C. cortezi was described with a rounded anterior end with widths ranging from 22 to 48 lm [17].
C. chattoni n. sp. has been observed in L. vulgaris off Tunis which differs from C. coronata that has been observed in Octopus vulgaris, Sepiola rondeleti, Illex coindetti, Eledone cirrhosa and Scaeurgus unicirrhus, in the Mediterranean Sea   and English Channel [4,7,10,20], C. cortezi that has been observed in Pterygioteuthis giardi in the Gulfs of California and Mexico by Hochberg [17,20], and Chromidina sp. described by Landers [25], that has been isolated from Illex coindetti in the Gulf of Mexico.
Chromidina sp. S50 from Sepia officinalis DNA sequence: Partial sequences of 18S rDNA amplified from parasites isolated from S. officinalis host S50 deposited in the EMBL/GenBank/DDBJ database (Accession number LT546663).
Prevalence: 2.6% (one out of 38 specimens of S. officinalis examined from off Tunis, Tunisia).
Description: Tropho-tomonts from only one S. officinalis out of 38 specimens examined; infection at low level; host co-infected by the dicyemid Pseudicyema truncatum (data not shown).
Anterior end: Majority of parasites with conical shape terminated by a pronounced apical papillum attaching the parasite to the renal epithelium; width of the anterior end similar to that of the body (Figs. 3A-3E). Slight inflation of the head observed in a few parasites.
Ciliature: Consisting of dextrally-spiralled kineties originating from the apex and continuing on the cell body (Fig. 3D). Number of kineties not determined due to unfavourable positioning of tropho-tomonts on smears. Posterior end: long tropho-tomonts with narrower posterior end than anterior part (Figs. 3A and 3C). No division (monotomy or palintomy) observed.
Remarks: The absence of dividing stages and the impossibility of determining the number of kineties limit the comparison with the other Chromidina species. The absence of a crown of cilia differentiates it from C. coronata and this type of tropho-tomont morphology with a conical head was not observed for C. elegans. Hochberg [17,20] described apotomites of C. cortezi with similar morphology on the renal appendages of the squid Pterigioteuthis giardi fished in the Gulfs of California and Mexico. Apotomites are large single buds produced by a monotonic process from trophotomonts. When detached, they differentiate into a second generation of tropho-tomonts that colonise the renal appendages. It is suggested that stress conditions such as a shortage of nutrients or of essential metabolites might result in the production of apotomites instead of tomites [17]. The presence of degenerative or cyst-like forms (Figs. 3F and 3G) in our preparations supports such stress conditions. Furthermore, co-infection with dicyemids indicates that the infection by this Chromidina sp. in S. officinalis S50 reaches its final stage. The Chromidina sp. population is progressively replaced by dicyemids, competing for the same ecological habitat, the renal appendages, when the host changes from a pelagic to a benthic life cycle [13]. Pterigioteuthis giardi was originally described from the Mediterranean Sea but is now known to be nearly cosmopolitan [17]. Whether the Chromidina sp. observed in our preparations corresponds to C. cortezi needs to be clarified and requires further investigation. In any case, this is the first description of a Chromidina sp. infecting Sepia officinalis. The numbers of base differences per site between sequences are shown as percentages. The analysis involved 14 nucleotide sequences. All ambiguous positions were removed for each sequence pair. There were a total of 1,477 positions in the final dataset.

Phylogenetic analysis
Positive PCRs were obtained for each of the four Chromidina specimens collected from the infected cephalopod hosts, using universal and specific primers designed to amplify a portion of the Chromidina 18S rDNA locus. Gene sequencing and assemblage enabled us to determine, for the first time, the partial 18S rDNA sequences for these parasite ciliates. Blast analyses revealed that the highest homology scores were obtained with known Apostomatia 18S rDNA sequences, in particular with species belonging to the genus Pseudocollinia [14,27]. A multiple alignment was built using these four Chromidina spp. sequences and a representative selection of Apostomatia to compute evolutionary distances ( Table 2). Bayesian and Maximum Likelihood phylogenetic trees were also constructed using a selection of species belonging to several Ciliophora classes (Fig. 4).
The three sequences obtained for Chromidina species infecting L. vulgaris hosts were highly similar to each other and very close to the sequence obtained for the Chromidina sp. S50 infecting S. officinalis, with a genetic divergence of less than 0.2% ( Table 2). The closest relatives of these Chromidina sp. sequences were those of Pseudocollinia brintoni (e.g. HQ591470, [14]), Pseudocollinia similis (e.g. HQ591485, [27]) Pseudocollinia beringensis (e.g. HQ591476, [27]) and Pseudocollinia oregonensis (e.g. HQ591473, [27]) with a genetic divergence of 3.9-4.2%, then Fusiforma themisticola (KF516511, [3]) with a genetic divergence of 4.7-4.8%. Gymnodinioides pitelkae (EU503534, [5]), Hyalophysa lwoffi (EU503538, [5]) and Vampyrophrya pelagica (EU503539, [5]), and Hyalophysa chattoni (e.g. EU503536.1, [5]) were found with a genetic divergence of 5.9-6.2% and Gymnodinioides sp. (EU503535.1, [5]) of 6.4% (Table 2). These values reveal the rather low divergence between these species, for this molecular marker, as previously observed by Lynn for other Apostomatia [27]. Pseudocollinia spp., parasitoid apostomes of krill, currently appear the species most closely related to Chromidina spp. The phylogenetic analysis using a dinoflagellate sequence as an outgroup revealed that, within the Ciliophora phylum, these Chromidina spp. sequences remained grouped with the Apostomatia (Chromidina, Gymnodinioides, Pseudocollinia, Fusiforma) within the Oligohymenophorea class, with strong bootstrap values (Maximum Likelihood analysis) and posterior probability values (Bayesian analysis) (Fig. 4). This phylogenetic tree, rooted on a dinoflagellate rDNA sequence, was inferred from 30 small subunit (SSU) 18S rDNA sequences corresponding to the four Chromidina specimens identified in this study, 25 representatives of the major classes of the phylum Ciliophora, and one Dinoflagellata sequence taken as the outgroup. There were a total of 1,157 positions in the final dataset. Both the Maximum Likelihood (ML) method and Bayesian inferences, based on the general timereversible +G +I model [28], were conducted and yielded similar topologies; the currently presented topology was obtained by the Bayesian inference. The tree is drawn to scale, with branch lengths measured in number of substitutions per site. Some branches were shortened by multiples of the length of substitutions/site scale bar (Plagyophylea, Phyllopharyngea). Numbers at the branches denote ML bootstrap percentage, from 1,000 resampling (first value) and Bayesian posterior probabilities (second value).