Morphological and molecular diagnosis of Pseudoterranova decipiens ( sensu stricto ) (Anisakidae) in imported cod sold in Brazil

An anisakid nematode larva found in cod sold in the state of Minas Gerais, Brazil was studied by light and scanning electron microscopy and by a molecular approach. Mitochondrial cytochrome c-oxidase subunit 2 (mtDNA cox -2), 28S rRNA and ITS1, 5.8S and ITS2 regions were amplified using the polymerase chain reaction and sequenced to evaluate the phylogenetic relationships of the larva. The genetic profile confirmed that this larva belongs to the species Pseudoterranova decipiens ( sensu stricto ) . This is the first molecular and ultrastructural study of Pseudoterranova decipiens ( sensu stricto ) in imported cod sold in Brazil. The health implications of these findings are discussed.


Introduction
Fishes known as cod actually comprise five different species.The term 'cod' originally referred to fishes from the Atlantic and Pacific Oceans that are processed by salting and drying (OLIVEIRA et al., 2012).Among these species, Gadus morhua Linnaeus, 1758 and Gadus macrocephalus Tilesius, 1810 (Gadidae) are known as Atlantic and Pacific cod that occur, respectively, in the North Atlantic and Pacific Oceans.Brazil imports salted cod mainly from Norway and Portugal, and according to Brazil's Ministry of Fisheries and Aquaculture, 45 million tons were sold in 2011 (BRASIL, 2013).These fish are known to host different helminth parasites, including anisakid nematodes (CHANDRA & KHAN, 1988;HEMMINGSEN & MACKENZIE, 2001;PERDIGUERO-ALONSO et al., 2008;CATALANO et al., 2014;MEHRDANA et al., 2014).In addition to G. morhua and G. macrocephalus, other species of salted fish are sold erroneously as cod in Brazil, namely saithe Pollachius virens (Linnaeus, 1758), ling Molva molva (Linnaeus, 1758) and tusk (zarbo) Brosme brosme (Ascanius, 1772) (LIMA & SANT 'ANA, 2011).Reports of the occurrence of anisakid nematodes in cod and other similar salted and dried fish sold in Brazil have begun to appear (PRADO & CAPUANO, 2006).
Due to their zoonotic potential and possibility of inducing allergic reactions (RAMOS, 2011;SUZUKI et al., 2010;NIEUWENHUIZEN & LOPATA, 2013), the occurrence of parasites in fish is considered a quality defect, according to Margeirsson et al. (2007), and as such, they are included in regulatory recommendations for the consumption of salted or dried salted fish in Brazil (BRASIL, 2007).With regard to the harmful effects of these parasites on human health, various authors have produced contradictory results/conclusions about the efficacy of salting to inactivate helminth larvae (RAMOS, 2011).Some authors claim that only live parasites can affect human health (ALONSO et al., 1997;RODRÍGUEZ et al., 2006;AUDICANA & KENNEDY, 2008), while others have demonstrated that even inactivated parasites can cause allergic reactions (FERNÁNDEZ DE CORRES et al., 1996;DEL REY MORENO et al., 2006;VIDACEK et al., 2009).

Materials and Methods
A single nematode larva removed from three fleshs (1Kg) of salted cod sold in the municipality of Viçosa, state of Minas Gerais, Brazil (20° 45'14" S, 42° 52' 55" W) was washed in distilled water, cut into three pieces and fixed in 70% ethanol.
The anterior and posterior regions were cleared in glycerine for examination by light microscopy.These fragments were then washed in 70% alcohol, dehydrated in an ethanol series, critical point dried with CO 2 , coated with 20 nm of gold, and examined in a JEOL JSM-6390 scanning electron microscope.
DNA cycle-sequencing reactions were performed using BigDye v.3.1 chemistry (Applied Biosystems, Foster City, CA, USA), while the sequencing reactions were performed in an ABI Prism 3100 genetic analyzer.The sequences were edited using DNASTAR SeqMan (DNASTAR, Inc., Madison, WI, USA) and compared for similarities with sequences from GenBank, using BLAST 2.0 (ALTSCHUL et al., 1990).To examine the phylogenetic relationships, the nucleotide sequences were analyzed using the CLUSTAL W algorithm of the BioEdit Package (THOMPSON et al., 1994;HALL, 1999).The Maximum Likelihood (ML) phylogenetic tree (FELSENSTEIN, 1981) was inferred with MEGA 5.0 software (TAMURA et al., 2011), using the Kimura 2-parameter (K2P) model, and gamma distribution was selected by MEGA 5.0.The Maximum Likelihood tree was resampled by 5,000 bootstrap replicates to evaluate the reliability of the groups.The model of evolution for the Bayesian inference (BI) tree was GTR+γ+I, and default priors were estimated using MrBayes 3.1.2software (HUELSENBECK & RONQUIST, 2001).After 2,000,000 generations for convergence, 25% of the trees were discarded as burn-in and the rest were used for topology and posterior probability reconstruction.

Results and Discussion
The initial examination using light microscopy revealed that the worm was a third-stage anisakid larva with a boring tooth situated between the ventrolateral lips, the excretory pore at the base of the ventrolateral lips, a long ventriculus but no ventricular appendix, an intestinal caecum and a tail with a mucron.The scanning electron microscopy observations revealed three poorly developed lips, i.e., a dorsal lip bearing a pair of double papillae and two ventrolateral lips (Figure 1a).The tail bore a terminal mucron (Figure 1b-d).Based on the morphology of the larva, it was identified as a species of Pseudoterranova Mozgovoi, 1951.
The subsequent molecular analysis enabled a direct comparison of the ITS1, 5.8S, ITS2 and 28S rDNA and the mitochondrial cytochrome c-oxidase subunit 2 (cox-2) sequences with those of Pseudoterranova decipiens Krabbe, 1878 (sensu stricto) in GenBank (GenBank accession numbers KF806033, KF806034 and KF806035).For cox-2 (accession number KF806035), the BLAST results indicated 98% identity with 100% query cover and a maximum score of 399 for P. decipiens, and 99% identity with 74% of query cover and a maximum score of 309 for P. azarasi.For the 28S gene (accession number KF806033) in relation to P. decipiens, there was a 100% identity, 99% query cover and a maximum score of 1382; no match was found for P. azarasi.For the sequence KF806034 (ITS1, 5.8S, ITS2 and 28S), the results for P. decipiens were as follows: identity = 100%, query cover = 98% and maximum score = 1607, and for P. azarasi: identity = 100%, query cover = 95% and maximum score = 1559.While the cox-2 gene BLAST analysis resulted in a greater similarity to P. azarasi than to P. decipiens, the maximum score and query cover were greater for P. decipiens.For other genes, P. decipiens showed the higher values.The higher scores for P. decipiens can be explained by a longer query cover for P. decipiens than for P. azarasi, even though the similarity to P. azarasi was closer than to P. decipiens.Thus, the BLAST analysis pointed to the species P. decipiens.The finding of P. decipiens in salted G. morhua sold in Brazil is in accordance with a previous report (PEREIRA et al., 2000) and is now confirmed by a molecular approach.The alignment with Pseudoterranova species showed eight nucleotide differences between P. decipiens (this study) and P. azarasi (Figure 2).
Two different phylogenetic trees were generated (ML and BI) and provided similar results.Using the maximum likelihood tree (Figure 3), our sequence (KF806035) was closer to sequences of P. decipiens (accession numbers HM147278 and AF179920) than to the sequence of P. azarasi (accession number HM147281), despite the low bootstrap value (41%).However, compared to P. azarasi, the clade composed of these three P. decipiens sequences showed a considerably higher bootstrap value (84%).In the IB tree (Figure 4), the same group of sequences exhibited a reasonable posterior probability (58%) and a higher bootstrap value (89%).In accordance with Timi et al. (2014), our phylogenetic trees strongly support P. decipiens sp.E (accession number HM 147282-1) as a sister taxon to the remaining species of the P. decipiens complex.
Phylogenetic analyses are important, since the species P. decipiens and P. azarasi are very closely related (MATTIUCCI & NASCETTI, 2008).The identification of P. decipiens in commercial food fish such as cod can have both economic and public health impacts, with wild fisheries sustaining considerable losses due to possible recommendations by international or regional regulatory agencies to forbid the consumption of salted and dried fish.As many as 107 parasite species, including both protozoan and metazoan forms, have been identified in G. morhua alone (HEMMINGSEN & MACKENZIE, 2001).The finding of codworm P. decipiens (s.s.) in marketable cod in Brazil therefore rings an alarm bell in terms of a public health risk in this country.
As an emerging zoonosis first diagnosed in the Netherlands in 1955 ( VAN THIEL et al., 1960;SAKANARI & MCKERROW, 1989;ISHIKURA, 2003), anisakiasis, the disease caused by the ingestion and establishment of anisakid larvae in the human gastrointestinal tract, has received considerable attention in recent years.There are many reports of its occurrence in marketable cod and other dried, salted or fresh fish (e.g., TORRES et al., 2007;ARIZONO et al., 2011;MATTIUCCI et al., 2013;NIEUWENHUIZEN & LOPATA, 2013).Anisakiasis, including pseudoterranoviasis, represents an emerging public health problem in Brazil that requires greater attention by physicians and by sanitary and epidemiological services, especially considering the possibility of allergic reactions resulting from the ingestion of these larvae.In a recent recommendation, consultants of Brazil's Ministry of Health included P. decipiens on the Classification List of Biological Risks as a potential hazard for humans (BRASIL, 2010).This is the first molecular and ultrastructural study of Pseudoterranova decipiens (sensu stricto) in imported cod sold in Brazil.This report on the identification and occurrence of this parasite in imported cod underscores the need for a rational government program for the risk management of anisakid nematode larvae.

Figure 1 .
Figure 1.Scanning electron microscopy of L3 larva of Pseudoterranova decipiens (s.s.): a) anterior region with dorsal lip bearing two double papillae, a boring tooth situated between the two ventro-lateral lips and the oral aperture; b-d) tail with a mucron.

Figure 2 .
Figure 2. Sequence alignments of Pseudoterranova species.The numbers above the sequences represent nucleotide position.The dots indicate equal nucleotides compared with the sequence of P. decipiens found herein (in bold).A dash represents a gap.

Figure 3 .
Figure 3. Phylogenetic tree based on the Maximum Likelihood of the partial cytochrome c oxidase subunit II gene, showing the relationship of Pseudoterranova decipiens (bold) with others species of Pseudoterranova.Numbers (percentage) on the branches indicate 5,000 bootstrap replicates.GenBank accession numbers are shown.Scale bars indicate the nucleotide mutations per site.

Figure 4 .
Figure 4. Phylogenetic tree based on the Bayesian inference of the partial cytochrome c oxidase subunit II gene, showing the relationship of Pseudoterranova decipiens (bold) with others species of Pseudoterranova.Numbers (percentage) on the branches indicate the posterior probability.After 2,000,000 generations a convergence was achieved.GenBank accession numbers are shown.Scale bars indicate the nucleotide mutations per site.