Introduction

Foodborne diseases include a range of illnesses transmitted via the ingestion of foodstuffs contaminated with a variety of chemical compounds and pathogenic microorganisms, including parasites [1–3]. Whilst the global disease burden and costs linked to these illnesses are difficult to estimate, it has been calculated that foodborne infections have cost the Australian economy alone $1.249 billion in 2006 [4]. Amongst the parasites responsible for foodborne diseases, nematodes within the genus Anisakis (i.e. A. simplex and A. pegreffii, also known as herring worms) are the causative agents of the fish-borne gastrointestinal illness ‘anisakiasis’. Since the 1960s, over 20,000 cases have been reported worldwide [5]; however, this number is most likely severely underestimated. Indeed, in Japan, where the consumption of raw or undercooked fish is common, ~2000–3000 new cases of anisakiasis occur each year [5].

The life cycle of Anisakis is indirect, with cetaceans such as dolphins and whales harbouring the dioecious adult nematodes in their gastrointestinal tract; female Anisakis release unembryonated eggs that are excreted in the aquatic environment via the faeces. Following development into first-, second- and third-stage larvae (L1s, L2s and L3s, respectively), these hatch from the eggs and are ingested by crustacean hosts. When infected crustaceans are ingested by suitable paratenic hosts, such as fish or squid, L3s penetrate the intestine and encapsulate in tissues, particularly those of the liver and the mesentery. Following ingestion of L3-containing paratenic hosts by a suitable cetacean host, L3s develop to fourth-stage larvae (L4s) and subsequently to adult males and females [5]. Humans are accidental hosts for Anisakis, with the infection occurring via the ingestion of L3-containing raw or undercooked fish. In humans, the infection is usually self-limiting, and common symptoms range from epigastric pain, nausea, vomiting and low-grade fever (= gastric form), to intermittent or constant abdominal pain with possible complications such as peritonitis and/or ascites (= intestinal form). ‘Ectopic’ anisakiasis occurs when the ingested larvae penetrate the gut wall and undergo a somatic migration to other viscera [5]. However, individuals infected by Anisakis spp. can also become sensitised to parasite allergens, leading to the onset of allergic anisakiasis (the most significant form of disease), with symptoms ranging from urticaria, gastrointestinal and/or respiratory signs and/or anaphylaxis [5, 6]. Other than the ingestion of nematodes, allergic reactions can also be elicited by accidental exposure to hidden antigens in processed (including cooked) fish and fish products [7, 8], as well as inhalation of Anisakis allergens [9, 10].

To date, a total of 14 A. simplex allergens have been identified [see ref. 5]. Most of these allergens have been detected in the parasite excretory/secretory (ES) products, with Ani s 1, Ani s 5 and Ani s 7 being recognised by serum antibodies in the majority of individuals affected by allergic anisakiasis [11, 12]. A recent proteomic investigation of L3s of A. simplex led to the identification of 17 novel putative allergens, which included structural proteins (e.g. myosin-4), as well as a number of enzymes (e.g. one enolase and one endochitinase) [13]. These antigens, derived from both ES and somatic components (the latter released following death and disintegration of the larvae), act as triggers for the activation of complex immunological and cellular host defences, which often result in allergic sensitisation [14]. However, thus far, the exact number and nature of parasite molecules acting as potential allergens are currently unknown. Next-generation sequencing and bioinformatics now provide rapid and cost-effective opportunities to investigate the fundamental biology of parasites of medical significance [see 15], and to build curated molecular databases for in-depth analyses of specific sets of parasite genes and gene products [16, 17] of biological and/or medical relevance. For instance, the study of the transcriptome (= the complete set of mRNAs transcribed by a cell, tissue or organism at any one time) represents a powerful approach to identify and characterise thousands of parasite transcripts simultaneously, which can then be screened for one or more sequences of interest. In this study, we sequence and characterise the transcriptomes of A. simplex and A. pegreffii in order to build annotated datasets for fundamental studies of the biology of these parasites. Using these resources, we compile lists of putative parasite allergens based on comparative analyses with known allergen sequence data available in public databases.

Materials and Methods

Results

The Anisakis transcriptomes

Paired-end Illumina sequencing of AS and AP cDNA libraries resulted in a total number of 64,065,430 and 65,508,456 raw reads, respectively (Table 1). Raw reads generated in the present study have been deposited in the Sequence Read Archive (SRA) database of NCBI (http://www.ncbi.nlm.nih.gov/sra) under study number SRP070744. Following pre-processing of raw reads, assembly and grouping of assembled Unigenes into sequence Clusters, a total of 34,746 (AS) and 33,747 (AP) transcripts were obtained. Conceptual translation of AS and AP transcripts resulted in 18,842 and 17,732 full-length predicted proteins, respectively (Table 1), of which 71.5% (AS) and 71.7% (AP) could be annotated via BLAST searches against the Nr, SwissProt, KEGG, COG and/or the GO databases (Table 1). Overall, 96.7%, 47.5% and 33.5% of AS transcripts and 88.0%, 90.5% and 73.8% of predicted proteins matched known nucleotide and amino acid sequences from A. simplex (http://parasite.wormbase.org/Anisakis_simplex_prjeb496/Info/Index/), as well as the phylogenetically related ascarid nematodes A. suum and T. canis, respectively; for AP, 96.6%, 46.1% and 33.2% of transcripts matched A. simplex, A. suum and T. canis sequences, respectively, while comparisons of amino acid sequence data resulted in 88.0%, 89.8% and 75.0% of AP predicted peptides, respectively, matching available peptide sequences from these two nematodes (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Transcriptome sequence data for third-stage larvae of Anisakis simplex and Anisakis pegreffii prior to and following assembly, and nucleotide and predicted peptide sequence data annotation (Nr = non-redundant; KEGG = Kyoto Encyclopedia of Genes and Genomes; COG = Clusters of Orthologous Groups of Proteins; GO = Gene Ontology).

https://doi.org/10.1371/journal.pntd.0004845.t001

A total of 5,561 (29.5%) and 5,357 (30.2%) AS and AP predicted proteins, respectively, could be assigned GO terms, while 8,945 (AS, 47.5%) and 8,629 (AP, 48.7%) matched homologous proteins in the KEGG database associated to 126 (AS) and 124 (AP) distinct biological pathways (Table 1). The 7,677 (AS, 40.7%) and 7,387 (AP, 41.7%) predicted proteins with matches in the COG database could be assigned to at least one of 25 functional categories, of which ‘general function prediction’ (AS = 15.7%; AP = 16.0%), ‘replication, recombination and repair’ (AS = 7.8%; AP = 7.7%), and ‘transcription’ (AS = 7.2%; AP = 7.4%), were the most represented (S1 Fig).

Putative novel Anisakis allergens

Comparative analyses of AS and AP predicted peptides with sequence data available in the AllergenOnline Database (http://www.allergenonline.com/) resulted in a total number of 38 (AS) and 31 (AP) matches, respectively, including 22 (AS) and 18 (AP) sequences matching previously known Anisakis allergens (S1 Table). Of the 16 (AS) and 13 (AP) remaining predicted peptides, 10 (AS) and 8 (AP) also matched protein sequences described as putative novel allergens by Arcos et al. [28] and Faeste et al. [13] (Table 2). BLAST comparisons between sets of putative allergens identified in AS and AP revealed six sequences unique to AS, which included a heat shock 70 kDa protein (ID: Unigene9825_AS1A), two fructose-bisphosphate aldolases A, a 60S ribosomal protein and two putative allergens matching sequences previously characterised in A. suum and Ascaris lumbricoides (Table 2); one sequence encoding a heat shock 70 kDa protein (ID: Unigene16173_AP1A) was unique to AP (Table 2). All of the putative novel allergens identified matched homologous allergenic proteins in arthropods (AS = 7; AP = 9), fungi (AS = 6; AP = 6), plants (AS = 3; AP = 3), other helminths (AS = 3; AP = 1) and fish (AS = 3; AP = 1) (Table 2). Of the putative allergens characterised in the present study, cyclophilins (AS and AP) and two predicted proteins of unknown function (AS) were identified in Anisakis for the first time (Table 2). Summaries of GO annotation and KEGG pathway analysis information for the whole transcriptomes of AS and AP are shown in S1 Fig, whilst the complete lists of AS and AP assembled transcripts, together with corresponding predicted peptides and annotation information, are given in S2 Table.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Putative novel allergens identified in the transcriptomes from third-stage larvae of Anisakis simplex and Anisakis pegreffii, based on homology of predicted amino acid sequences with known allergens in the AllergenOnline database (http://www.allergenonline.com/about.shtml).

(e-value cut-off: <1e-05, identity cut-off: >70%). (Nr = non-redundant database; GO = Gene Ontology; KEGG = Kyoto Encyclopedia of Genes and Genomes; BP = Biological Process; CC = Cellular Component; MF = Molecular Function).

https://doi.org/10.1371/journal.pntd.0004845.t002

Discussion

In this study, we characterise the transcriptomes of two Anisakis species, and provide the scientific community with a resource to explore the biology of these parasites, as well as their allergenic properties, using proteomics and immunological tools. The caudal extremities of the larvae used in this study were removed in order to unequivocally confirm species identification by PCR-coupled RFLP of the of ITS1-5.8S-ITS2 region of the rDNA, thus potentially leading to biases in the sets of Anisakis transcripts identified and characterised. Nevertheless, we consider these datasets to represent a comprehensive snapshot of the complements of genes transcribed by the L3s of these parasites. Indeed, prior to this study, only 913 Anisakis transcripts were present in the EST database of NCBI (http://www.ncbi.nlm.nih.gov/nucest/?term=anisakis); however, a draft genome sequence for this parasite, accompanied by large-scale transcriptomic sequence data to support gene predictions, are currently available from the WormBase ParaSite database as part of the ‘50 Helminth Genome Initiative’ (http://www.sanger.ac.uk/science/collaboration/50hgp). Though, given that these resources are as yet unpublished, we opted to assemble the short reads generated in this study de novo, and compare the resulting full-length cDNAs (contigs) and, subsequently, the amino acid sequences predicted from these, to Anisakis transcripts and predicted proteins (respectively) on the WormBase ParaSite database. Overall, 96.7% and 96.6% of the transcripts assembled in the present study matched Anisakis nucleotide sequence data in the latter database, thus providing support to the reliability and robustness of our assembly (cf. Table 1). Conversely, BLAST comparisons of Anisakis nucleotide and predicted amino acid sequence data generated in this study with available transcripts and protein sets from other ascarids [23, 24] revealed higher sequence similarities between Anisakis and A. suum than between the former and T. canis (see Table 1). Currently, the relationships amongst Clade III nematodes (also referred to as suborder Spirurina; [28]) are defined according to phylogenetic analyses of the small subunit of the ribosomal RNA (SSU rRNA) [29, 30]. In one of these investigations, SSU rRNA sequences from Ascaris spp. and Toxocara group together to the exclusion of the Anisakis spp. counterparts [30], likely reflecting the similarities in morphology and fundamental biology between the former (‘terrestrial’) species. While the elucidation of the systematic and phylogenetic relationships between Anisakis, Ascaris and Toxocara species are beyond the scope of the present work, the availability of large-scale transcriptomic and genomic datasets for all of these parasites (http://www.sanger.ac.uk/science/collaboration/50hgp; [23, 24]) represents a useful resource for future, comprehensive phylogenomic studies of ascarid nematodes.

In this study, we aimed to sequence and annotate the transcriptomes of two Anisakis species, in order to build a molecular resource from which to draw information on the set(s) of molecules responsible for evoking allergic responses in the human host. Two recently published key studies [13, 27] utilised sera from human patients with IgE against A. simplex and positive to the skin prick test, coupled with mass spectrometry-based analyses of reactive Anisakis proteins to identify and characterise putative novel nematode allergens. Both studies revealed a broad array of potentially allergenic Anisakis molecules, with individual sera binding to multiple protein bands, which resulted in a substantial inter-individual variability of binding patterns [13]. This finding suggests that the whole complement of Anisakis allergens is yet to be fully defined, and supports the application of NGS technologies and bioinformatics to assist in this quest. While all of the putative novel allergens identified by Arcos et al. [27] and Fæste et al. [13] matched AS and AP amino acid sequences inferred from cDNAs generated in this study, only subsets of these also matched known allergens in the AllergenOnline database (cf. Table 2). The most likely explanation for this observation is technical and is related to the inevitable incompleteness of the AllergenOnline database which, while regularly updated and manually curated, relies on the collection of sequence data submitted by end-users to public sequence repositories and designated as ‘allerg*’, or extracted directly from peer-reviewed publications (www.allergenonline.com). On the other hand, the absence of some previously identified putative Anisakis allergens from the AllergenOnline database is justified by the fact that the allergenic properties of these molecules are yet to be fully elucidated [13, 27].

Amongst the putative allergens characterised in the present study, sequences encoding heat shock proteins 70 (HSPs 70) and enolases were identified in both AS and AP datasets, and had been previously detected using immune-proteomic approaches [13, 27]. Based on direct comparisons with sequence data in the Nr database of NCBI, these (and other) sequences displayed high sequence similarity with HSPs from Danio rerio (cf. Table 2). While this finding is likely to be linked to the overrepresentation of sequences from zebrafish in Nr compared with sequence data from parasitic nematodes, contamination of Anisakis mRNA with that from the fish host cannot be excluded. HSPs are a family of highly conserved proteins that play primary roles in maintaining cellular homeostasis but whose increased expression in the presence of conditions of stress such as sudden changes in temperature, injuries and infections, is responsible for the activation of a cascade of immune-molecular events that culminate in inflammatory responses [31]. In particular, in a recent study [32], the expression of an HSP 70 from A. pegreffii was increased in L4s compared with L3s, which may be linked to a response of the parasite to the heat stress that immediately follows infection [32]. HSPs 70 from mites, black flies, midges and cockroaches are known allergens and mediators of allergic contact hypersensitivity [33, 34]; in addition, levels of HSPs 70 are increased in the sputum and plasma of asthmatic patients compared with healthy controls [35], while antibodies against these proteins are associated with a number of immunological disorders, such as allergy to metals [36]. Based on this knowledge, it is therefore plausible that antibodies against Anisakis HSPs 70, whose expression is increased upon host infection [32], may play a key role in allergic responses to these parasites. Enolases are also recognised as major allergenic proteins in plants, fish, fungi, cockroaches and biting insects [37]. In a previous study, anti-enolase antibodies were detected in sera from mice experimentally infected with A. simplex L3s or exposed to parasite crude protein extracts, but not from mouse sera raised against the parasite excretory-secretory antigens [38]; in the same study, anti-enolase antibodies could not be detected in the sera of human patients infected with A. simplex, which led the authors to speculate that these molecules do not offer a sufficient antigenic stimulus to act as allergens [38]. Conversely, IgE against Anisakis enolases were detected in Anisakis-allergic patients by Fæste et al. [13], thus supporting the role of these molecules in allergic anisakiasis. However, it is worthwhile to note that, while sensitisation to Anisakis allergens usually follows the ingestion of fish and/or fish products, these products can also contain very similar allergens to those from the parasites ([e.g. 39]; cf. Table 2). Importantly, recent studies [39] have demonstrated the role of fish beta-enolase and fructose-bisphosphate aldolase as allergenic stimuli in patients sensitised to cod, salmon and tuna, thus further complicating the diagnosis of ‘true’ sensitisation to Anisakis allergens. In addition, tropomyosin (Ani s 3), one of the most immunogenic proteins known to man, was also first discovered in Anisakis and subsequently identified in over 150 invertebrate species (including shellfish and mites) [see ref. 1]. This muscle protein, which contains a coiled-coiled α-helical structure, is responsible for exacerbated immune responses in over 30% of the world’s population, and may be involved in the known cases of cross-allergenicity between mites and Anisakis [40].

Of the molecules inferred as novel putative allergens, sequences encoding peptidyl-prolyl cis-trans isomerases (cyclophilins) were identified in both AS and AP (cf. Table 2). Cyclophilins belong to a family of conserved proteins present in both prokaryotes and eukaryotes, and thought to play key roles in a range of human inflammatory diseases, including rheumatoid arthritis and asthma [41]. Human cyclophilins also act as self-antigens, being recognised by serum IgE from individuals sensitised to environmental cyclophilins, such as those in pollens [see 42]. Cyclophilins have also been identified in a range of parasitic nematode species (e.g. Angiostrongylus cantonensis, Dirofilaria immitis and Haemonchus contortus; [43–45]) and are thought to operate as catalysts and chaperones in cuticle synthesis [46]; in particular, cyclophilins were identified in the excretory-secretory products of H. contortus and recognised by sheep hyper-immune sera [47]. Interestingly, cyclophilins were also detected by immunological screening of a cDNA library from the zoonotic cestode Echinococcus granulosus, causing cystic echinococcosis (CE), with sera from infected human subjects that had displayed allergic (skin) reactions [48]; sera from these patients did not recognise the homologous human cyclophilins, nor that from the yeast Malassezia furfur, thus supporting the hypothesis that the parasite cyclophilin was responsible for the allergic reactions observed in patients with CE [48]. Based on this information, we hypothesise a role of cyclophilins in the array of molecules responsible for allergic anisakiasis, a hypothesis that requires testing.

Amongst the novel putative allergens in AS were two sequences with high sequence similarity to ABA-1 proteins, members of the nematode polyprotein allergens (NPAs) from A. suum (cf. Table 2). These proteins are synthetised as repetitive polyproteins, which are subsequently cleaved during post-translational processing into multiple functional units with fatty-acid binding properties [49]. NPAs from Ascaris and other parasitic nematodes (e.g. Brugia malayi) are associated with hypersensivity responses in infected individuals [see 50]. In addition, no IgE cross-reactivity could be detected between human sera (from asthmatic patients from an area where ascariasis is endemic and exposure to mites is common) probed with recombinant Ascaris ABA-1 and mite fatty-acid binding proteins (FABPs), thus suggesting that parasites are solely responsible for allergic reactions against this protein [51]. Given the high level of sequence similarity (at both nucleotide and protein levels) between Anisakis sequences encoding for ABA-1 proteins identified in this study and the Ascaris counterparts, it is plausible that these molecules also contribute to the onset of allergic anisakiasis.

While our work represents a step forward in the application of NGS technologies towards building a molecular infrastructure for research on allergic anisakiasis, gaps still exist in our knowledge of the complex host-parasite relationships which culminate in the exacerbated immune reactions observed in individuals infected by Anisakis. The completion, curation and refinement of the whole genome sequence of A. simplex (http://www.sanger.ac.uk/science/collaboration/50hgp) will assist filling these gaps, by providing a solid resource for fundamental functional explorations of these relationships. Ultimately, the discovery of the whole array of parasite molecules responsible for immune hypersensitivity in Anisakis (and other parasites) will set the basis of future studies aimed at developing comprehensive, reliable and robust diagnostic tools, which will assist clinicians in choosing appropriate intervention strategies and effectively assessing their outcomes.

Supporting Information