Morphological and Molecular Diagnosis of Anisakid Nematode Larvae from Cutlassfish (Trichiurus lepturus) off the Coast of Rio de Janeiro, Brazil

Anisakid nematode larvae from Trichiurus lepturus off coast of Rio de Janeiro were studied using light, laser confocal and scanning electron microscopy, in addition to a molecular approach. Mitochondrial cytochrome c-oxidase subunit 2 (mtDNA cox-2), partial 28S (LSU) and internal transcribed spacers (ITS-1, 5.8S, ITS-2) of ribosomal DNA were amplified using the polymerase chain reaction and sequenced to evaluate the phylogenetic relationships between the nematode taxa. The morphological and genetic profiles confirmed that, of the 1,030 larvae collected from the 64 fish examined, 398 were analysed, of which 361 were Hysterothylacium sp. and 37 were Anisakis typica. Larvae of Hysterothylacium sp. were not identified to the species level due to the absence of similar sequences for adult parasites; however, the ITS sequence clustered in the phylogenetic tree with sequences of H. deardorffoverstreetorum, whereas an mtDNA cox-2 and LSU concatenated phylogenetic analysis demonstrated the presence of two clades, both of them under the same name as the larval H. deardorffoverstreetorum. Data on the occurrence of parasites during the winter and summer months were compared using the t-test. The greatest prevalence and intensity of infection were recorded for larval Hysterothylacium, with a prevalence of 51.56% and an intensity of up to 55 parasites per fish. The larval Anisakis exhibit a higher abundance and intensity of infection in the winter months, and those of Hysterothylacium during the summer. However, the t-test indicated no significant differences between the abundance and intensity of infection recorded during the months of collection for either of these larval nematodes. All sequences generated in this study were deposited in GenBank.


Introduction
Anisakid nematodes are parasites with an indirect life cycle, which utilizes hosts at different trophic levels in the food web. Aquatic vertebrates, such as piscivorous fishes, mammals and birds, are definitive hosts and aquatic invertebrates and fishes act as intermediate or paratenic hosts [1,2] [2,3].
Anisakid larva are usually very difficult to identify to species using morphology due to the lack of differential characters, but when adults are already described and genetically characterized, then such larva can be assigned to a species based on molecules [1,4]. The accurate identification of anisakid species is essential, because there are important pathogens within the group that can cause problems for human and animal health [2,5,6,7]. Molecular tools are therefore valuable for linking anisakid larva to known adults as well as for systematic, evolutionary and ecological studies of these parasites [1,4,5,8,9].
The cutlassfish Trichiurus lepturus L. (Trichiuridae) has a wide distribution, occurring throughout tropical and temperate waters of the world. Previous parasitological surveys on specimens from off the coast of Rio de Janeiro listed the occurrence of anisakid larva identified only to generic level by means of light microscopy [10,11]. In this study, the nematode parasites of T. lepturus from the same region are re-evaluated using light, laser confocal and scanning electron microscopy, and also by the determination of nucleotide sequences from the internal transcribed spacers of ribosomal DNA (ITS-1, 5.8S, ITS-2), the partial 28S (LSU) and mitochondrial cytochrome c-oxidase subunit 2 (mtDNA cox-2).

Results
A total of 1,030 nematode larva were collected from 64 fishes; 398 were analyzed for morphological data and 72 were used for genetic studies. The larvae identified by morphological and molecular approaches as Anisakis typica and Hysterothylacium sp. are characterized below.
Anisakis typica third-stage larva. Thirty seven specimens were collected from the body cavity and mesentery of T. lepturus; their measurements are presented in Table 1. They had the  following characteristics: cuticle smooth; lips poorly developed;  ventrolateral lips with single and double papilla, dorsal lip with two  double papillae; boring tooth present between ventral lips;  intestinal caecum absent; ventriculus elongate (Figures 1A-1C). Excretory pore present at the base of ventrolateral lips (Figures 1B, 1D-1E); tail short, round, with mucron ( Figure 1F).
Genetic characterization of 22 larva enabled the species determination, with 13 being diagnosed by specific PCR as Anisakis typica (Diesing, 1860); 9 were submitted to PCR for family for each genetic region (ITS, mtDNA cox-2 and LSU) with subsequent sequencing reactions. Considering only those sequences of suitable quality for the genetic characterization, five sequences were obtained for the ITS region, one for mtDNA cox-2 region and three for the LSU region (accession n. JQ798962, JQ798968 and JQ798967, respectively). The alignment of the sequences from this study with sequences from GenBank resulted in 100% of similarity for sequences of the ITS region ( Figure 2) and 99% for sequences of the mtDNA cox-2 region. There were no previous sequences for the LSU region of Anisakis typica deposited on GenBank for comparison. Phylogenetic analysis for A. typica demonstrated a clear separation between different species of Anisakis with strong statistical support (Figures 3 and 4). This is the first identification of A. typica in T. lepturus in Brazilian waters, with the new LSU sequence being deposited in the GenBank.

Hysterothylacium
sp. third and fourth-stage larvae. Three hundred sixty one specimens were collected from the body cavity and mesentery of the cutlassfish. Measurements were taken from 28 L3 and 13 L4 individuals (Table 1). They had the following characteristics: small worms, with smooth cuticle and distinct lateral alae along each side of body between level just posterior to lips and pre-cloacal region ( Figure 5A). L3 with anterior region rounded and lacking defined lips; inconspicuous boring tooth present ( Figure 5B); L4 with developing lips and lacking boring tooth ( Figure 5C); esophagus claviform; ventriculus small and rounded (Figures 5D-5F); intestinal caecum smaller than ventricular appendix; excretory pore inconspicuous, located between nerve ring and anterior extremity of intestinal caecum, clearly visible in SEM and CLSM images (Figures 5A, 5F); tail long, digitiform, with terminal mucron ( Figures 5G-5H). Genetic characterization of 41 larva with PCR for family for each genetic region (ITS, mtDNA cox-2 and LSU) and DNA sequencing was carried out, and 17 good quality sequences were obtained for the ITS region, 30 for the mtDNA cox-2 region and 19 for the LSU region (accession nos. JQ798963, JQ798964, JQ798965 and JQ798966). The alignment of these sequences, with reference sequences (Table 2), resulted in a 100% of similarity for the closest sequence of the ITS region (JF730200), 99% for the closest sequence of the mtDNA cox-2 region (JF730213 and JF730211) ( Figure 6) and 96% for the closest sequence of the LSU region (AY821772). The mtDNA cox-2 sequences obtained presented a maximum pairwise distance of 8% when compared with reference sequences ( Table 2). The pairwise distance among our mtDNA cox-2 sequences had a maximum distance of 9%. An mtDNA cox-2 and LSU concatenated phylogenetic analysis (Figure 7) of sequences of Hysterothylacium specimens studied in the present paper demonstrated the presence of two clades, both of them including sequences under the name of H. deardorffoverstreetorum retrievable from GenBank ( Figure 7). The ITS phylogenetic tree was performed with a single sequence of Hysterothylacium; this sequence obtained in the present paper clustered in a single clade, well supported, including all the sequences of H. deardorffoverstreetorum deposited in GenBank ( Figure 8).

Ecological Data
The prevalence of Anisakis typica was 20.31% and the intensity varied from 1 to 10 specimens per fish. Hysterothylacium presented a prevalence of 51.56% and intensity of 1 to 55 per fish. The highest prevalences were found during November and December ( Figure 9). Larvae of Hysterothylacium sp. were the most abundant, with mean intensities between 2.5 and 20.5 (Figures 10 and 11). The t-test applied to verify the existence of variation between prevalence, intensity and abundance during winter (August) and summer (December and January) was not significant.

Discussion
The larvae of Hysterothylacium sp. are difficult to identify and their similarity with related genera has resulted in taxonomic confusion, with species of Hysterothylacium being identified as Contracaecum or Iheringascaris [12,13,14]. The position of the excretory pore, which has been reported as inconspicuous, is the main difference between larvae of Hysterothylacium (positioned at nerve ring level) and Contracaecum (situated at the base of lips). As mentioned above, the confocal microscopy and SEM were essential to ascertain its location. Likewise, species of Iheringascaris are separated from Hysterothylacium based only in the pattern of annulations of the cuticle [12]. In the present study, the SEM   micrographs showed the cuticle to lack annulations, as described for Hysterothylacium spp., although the phylogenetic analysis showed a close relationship with Iheringascaris. In the future, it is possible that species of Iheringascaris may be allocated within Hysterothylacium [15]. In this study, larvae of Hysterothylacium are reported at a high prevalence (51.56%), with an intensity of infection of up to 55 parasites per fish, but could not be identified to species level due to the absence of related adult sequences in the GenBank. Consequently, a specific identification could not be assigned.
Previous genetic analysis of Anisakis simplex and Hysterothylacium aduncum from T. lepturus in Taiwanese waters [16] were described but not formally deposited in the GenBank. However, a comparison with these data showed these species are genetically distinct from the nucleotide sequences obtained in this study.
The similarity among our Hysterothylacium sequences for ITS and LSU regions was 100%; on the contrary, our mtDNA cox-2 sequences exhibited a high genetic heterogeneity. The presence of polymorphism in the mtDNA cox-2 region has likewise been reported before for other species of nematodes [17]. The K2P distances calculated among the sequences available in GenBank under the name of H. deardorffovertsreetorum and the Hysterothylacium sequenced here, showed a genetic differentiation ranging from K2P = 0.005 to K2P = 0.08. The present study indicates that the Hysterothylacium larvae analyzed were likely to correspond to the larva described as H. deardorffoverstreetorum; however, the marked genetic differentiation so far detected at the mtDNA cox-2 level seems to suggest a possible genetic heterogeneity. This needs to be further investigated by future genetic analysis, likely using other nuclear markers. Indeed, while a comparison with one of the sequences of H. deadorffoverstreetorum (accession no. JF730200) resulted in a 100% of similarity for the ITS region, the mtDNA cox-2 sequences deposited, under the same name, had, at the intraspecific level, a genetic differentiation value with K2P    distances ranging from 0.002 to 0.077. This value has been also found among the sequences of Hysterothylacium analysed here at the same gene (K2P up to 0.092). Interestingly, the value of K2P = 0.07 is generally observed at the mitochondrial level between sibling or cryptic species of other anisakid nematodes [1,18]. Therefore, future genetic studies will no doubt clarify the genetic heterogeneity indicated here using nuclear markers. On the other hand, the morphology of larvae, especially of sibling species, appears to be often overlapping and not fully diagnosed when not accompanied by the genetic methodological approaches.
There are about 60 species of Hysterothylacium which have been formally described based on the morphological features of the adult worm [19,20,21,22]. However, so far, scanty data are available for their molecular analysis. Hysterothylacium sequences determined in this work were not similar to those deposited in the GenBank based on adult characterization. The question remains: could it be a new species, as indicated by the phylogenetic analysis, or a known species based on the morphological features of an adult worm which has not yet been characterized by molecular means? Species descriptions should contain data from as many sources as possible, including morphological infomation from adult worms, molecular data and phylogenetic analyses, which can be used not only as tools for identifying an isolate specimen but also for understanding its biology and taxonomy.
Hysterothylacium sp. type MB larvae sensu Deardorff and Overstreet [12] were reported from T. lepturus in the Sea of Oman [23], but the authors refrained from naming it. Similarly, unknown anisakid larvae have been reported from fishes using a PCR-based approach as evidence for new species, but the new form was not formally described as adults were not available for morphological characterization and molecular comparison [24]. However, Hysterothylacium deardorfoverstreetorum has recently been proposed based only on morphological features of the larva and a comparison with sequences of the genus deposited in the GenBank, despite their small number [25]. It is possible that, in future when sequences of adults of all or most of the 60 nominal species of Hysterothylacium are deposited in the GenBank, this species will likely sink into synonymy, reinforcing the idea that molecular data should be accompanied by strong morphological evidence based on adult nematodes.
The genotyping of more species will enable GenBank to become a robust tool for identification and phylogenetic analyses.
However, at present, the number of sequences of Hysterothylacium deposited in this database represents less than 15% of the valid species. This limitation compromises any phylogenetic results when the objective is to identify a species. For this, it is necessary to characterize a larger number of valid species based on genotypic information and morphological analyses of adult worms in order to enable the genetic identification of Hysterothylacium larvae.
An ITS sequence of the larvae of Contracaecum sp. found in Pagellus bogaraveo in Portuguese waters (accession no. JN005755) also presented 99% similarity with Hysterothylacium sequences from this study. Unfortunately, a formal publication with morphological characterization was not available for comparison.
Within the GenBank, the ITS sequence (accession no. EU828749) identified as Contracaecum muraenesoxi appeared to be very closely related to the sequences determined in this study. Nevertheless, this species was recently synonymized with Hysterothylacium amoyneze [21], explaining its phylogenetic position within the Hysterothylacium and proximity to our sequence. This highlights the fact that taxonomic changes of taxonomic names need, somehow, to be included in the GenBank in order to avoid phylogenetic misinterpretation. Similarly, the phylogenetic analysis showed an LSU sequence of Raphidascaris acus (accession no. AY821772) to be closely related to Hysterothylacium sp. from this study, which suggests that the morphological identification of that voucher of R. acus should be revised [15].
In this study, Anisakis typica was identified by molecular data, and our phylogenetic analysis for Anisakis species also indicated three distinct groups of species, agreeing with data from the literature [1].
The prevalence of Anisakis and Hysterothylacium larvae in this study were similar to those previously described in the cutlassfish off the coast of Rio de Janeiro [11]. Significant differences in prevalence were not observed between the winter and summer periods, although a moderate increase in prevalence and abundance was observed at the beginning of summer for Hysterothylacium. The prevalence of Anisakis simplex in fishes from Norway, for comparison, was most significant during spring, and the authors have suggested that a small variation in the occurrence of anisakids in tropical waters could be related to the low level of climatic variability typical for tropical weather [26]. The constant presence of definitive hosts along the Brazilian coast may also contribute to the presence of Anisakis and Hysterothylacium during both winter and summer, as observed in this study. Hysterothylacium  adults have been reported off the Brazilian coast parasitizing the following definitive hosts: Harengula clupeola, Scomberomorus cavalla, S. maculatus, Epinephelus guttatus [27]. These definitive hosts have a preference for coastal habitats, which may be related to the prevalence and abundance of Hysterothylacium in T. lepturus. Adults of Anisakis typica were described from the dolphins Sotalia guianensis and Stenella longirostris off the Brazilian coast. S. guianensis inhabits coastal waters, whereas S. longirostris prefers oceanic bays and island regions. A. typica larvae has been reported in Thunnus thynnus and Auxis thazard off Rio de Janeiro [28,29,30,31,32], indicating that the parasite is common in the area. During summer, there is an increase in whale-watching along the Rio de Janeiro coast, which is probably related to the seasonal upwelling in the region responsible for the addition of new elements to the food webs. At this time these food webs become more complex, thus promoting anisakid transmission [33,34]. This may explain the increasing abundance of these parasites in the summer. Furthermore, the increase in prevalence of anisakids off the coast during spring and summer could be due to the spawning period of T. lepturus, whose foraging behaviour increases in order to build resources for reproduction [35]. This is the first identification of A. typica in T. lepturus in Brazilian waters with LSU, ITS and mtDNA cox-2 sequences for larvae of both of A. typica and Hysterothylacium sp. This integrated study has shown the great need for a linkage between the analysis of morphological features supplemented by molecular data in order to enable the accurate identification of anisakid larva and provide robust taxonomic data.

Materials and Methods
A total of 64 fish were collected off Itaipu beach, Niterói, Rio de Janeiro (22u5391499S; 43u2294899O) from August 2010 to January 2011. Prevalence, abundance and mean intensity were calculated [36]. Data were transformed to attend the assumption of normality, and t-tests for independent samples were performed to verify differences between winter and summer months.
Nematodes were cut into three pieces and fixed in 70% ethanol. The anterior and posterior regions were cleared in glycerine and Figure 7. Maximum likelihood reconstruction between sequences of Hysterothylacium obtained in this study (*) and sequences of Hysterothylacium and Iheringascaris spp. from the GenBank, with the tree inferred from mtDNA cox-2 and LSU data sets. The numbers on the tree branches represent the percentage of bootstrap resampling. Heterocheilus tunicatus was used as an out group. doi:10.1371/journal.pone.0040447.g007  mounted as semi-permanent preparations on slides; the middle regions were used for molecular analyses. Specimens were examined using an Olympus CX3 microscope, and measurements were made with the aid of an ocular micrometer are given in micrometres as the mean, followed in parentheses by the range. High resolution confocal images were made using a confocal laser scanning microscope (Zeiss Axiovert 510, META). For scanning electron microscopical observations, some specimens were fixed for 24 hours at 4uC in a solution containing 2.5% glutaraldehyde and 4% paraformaldehyde in 0.1 M cacodylate buffer containing 3% sucrose at pH 7. The samples were washed in the same buffer and post-fixed overnight in 1% osmium tetroxide in 0.1 M cacodylate buffer at pH 7.2 in the dark. The specimens were dehydrated in an ethanol series, critical point dried with CO 2 , coated with 60 nm of gold and observed in a Jeol JSM 6390 SEM microscope.  The middle parts of parasites were prepared for total genomic DNA extraction using a ChargeSwitch gDNA Mini Tissue Kit (Invitrogen, Carlsbad, CA, EUA) according to the manufacturer's instructions. To amplify gene fragments of anisakid nematodes, a set of primers were used: NC5/NC2 [37] for ITS fragments, 211F/210R [15] for mtDNA cox-2 gene fragments and 391/390 [5] for 28S rDNA gene fragments. The primer ITSF/ITSR was used to amplify the ITS region of A. typica [32]. All PCR reactions were performed in a volume of 50 ml with 20 mM of Tris-HCl at pH 8.4; 50 mM of KCl; 250 mM of each deoxynucleoside triphosphate (dNTPs) and 2 ml of genomic DNA. The concentrations of MgCl 2 , primers and Taq Gold DNA polymerase (Promega Hot Taq Go Start Madison, WII -USA) were different for each reaction: NC5/NC2 (1.5 mM of MgCl 2 , 0.5 mM of each oligonucleotide primer and 1 U of Taq); ITSF/ITSR (2.5 mM of MgCl 2 , 0.4 mM of each oligonucleotide primer and 1 U of Taq); 211F/ 210R (0.5 mM of forward and 0.4 mM of reverse oligonucleotides, 2.5 mM of MgCl 2 and 1 U of Taq) and for 391/390 (0.4 mM of each oligonucleotide 3 mM of MgCl 2 and 1.5 U of Taq). PCR was carried out using a Mastercycler Personal/Eppendorf thermal cycler (Epperdorf, Hamburg, Germany) and cycling parameters as previously described [5,15,32,37].
PCR products were visualized with GELRED (Biotium Inc, Hayward, CA, USA) staining after electrophoresis on 1.5% agarose gels. All amplified PCR products generated were purified with Wizard SV gel and PCR clean up system kit (Promega) following the manufacturer's instructions and sequenced in both directions using the same primer sets as in the respective PCR assay. DNA cycle-sequencing reactions were performed using BigDye v.3.1 chemistry (Applied Biosystems, Foster City, CA, USA). Sequencing reactions were performed in the ABI Prism 3100 sequence analyzer. Sequences were assembled, edited, in DNASTAR SeqMan (DNASTAR, Inc., Madison, WI), and aligned with Bioedit Sequence Alignment Editor (version 7.0.4.1; http://www.mbio.ncsu.edu/Bio Edit/bioedit.html). The edited sequences were compared for similarities with sequences from GenBank using BLAST 2.0 (''Basic Local Alignment Search Tool'') ( Table 2) [38]. To examine the phylogenetic relationships, the nucleotide sequences were analyzed by CLUSTAL W algorithm of Bioedit Package [39,40]. The sequences of the two mitochondrial genes (mtDNA cox-2 and LSU) were joined using the software Concatenator [41]. Phylogenetic trees were inferred by using the software MEGA 5.0 [42] utilizing the General Time Reversible model (GTR) for ITS sequences and Hasegawa-Kishino-Yano model (HKY) for mtDNA cox-2 and LSU. These models were selected using the program jModelTest [43]. Kimura Two Parameters (K2P) values were calculated by the software MEGA 5.0 [42,44]. Maximum Likelihood method was used to construct trees [45] and were resampled by 100 bootstrap replicates to evaluate the reliability of the groups.