Potential Risk of Three Zoonotic Protozoa (Cryptosporidium spp., Giardia duodenalis, and Toxoplasma gondii) Transmission from Fish Consumption

In recent decades, worldwide fish consumption has increased notably worldwide. Despite the health benefits of fish consumption, it also can suppose a risk because of fishborne diseases, including parasitic infections. Global changes are leading to the emergence of parasites in new locations and to the appearance of new sources of transmission. That is the case of the zoonotic protozoa Cryptosporidium spp., Giardia duodenalis, and Toxoplasma gondii; all of them reach aquatic environments and have been found in shellfish. Similarly, these protozoa can be present in other aquatic animals, such as fish. The present review gives an overview on these three zoonotic protozoa in order to understand their potential presence in fish and to comprehensively revise all the evidences of fish as a new potential source of Cryptosporidium spp., Giardia duodenalis, and Toxoplasma gondii transmission. All of them have been found in both marine and freshwater fishes. Until now, it has not been possible to demonstrate that fish are natural hosts for these protozoa; otherwise, they would merely act as mechanical transporters. Nevertheless, even if fish only accumulate and transport these protozoa, they could be a “new” source of infection for people.


Introduction
Since 1961 fish consumption has increased at an average annual growth rate of 3.1%, overcoming average annual growth rates for all other animal proteins, except for that of poultry meat (2.7% per year). In 2017, per capita fish consumption reached a value of 20.3 kg, which represents about 17% of total animal protein. Annual per capita fish consumption varies widely between world regions, being lower in developing countries. Nevertheless, in these countries, fish represents a higher proportion of animal protein (29% in least developed countries), constituting a valuable component of their diet [1].
The rising concern about having a healthy diet to prevent diseases is one of the main reasons that has led to this increase in global fish consumption [2]. Fish composition includes different or C. meleagridis (IIIa-IIIf), between others. Currently, the only recognized zoonotic subtypes for C. parvum are the IIa and IId [11]. Table 1. Target genes used in Cryptosporidium spp. characterization. Adapted from [11,12].

Target Gene
Gene  1 The target gene is present in a single copy (Single) or in multiple copies (Multiple) in the haploid genome of the parasite; 2 In general, multiple copies genes have better amplification success in polymerase chain reaction (PCR) techniques.
Cryptosporidium oocysts are highly resistant in the environment and are also resistant to chlorinated compounds which are commonly used in drinking and recreational water treatment [13]. Transmission can be direct via fecal-oral route or by the ingestion of contaminated water or food. Cryptosporidium spp. is the main agent responsible for waterborne parasitic protozoan outbreaks, including a massive outbreak in Milwaukee in 1993 with approximately 403,000 cases [14] and counting with 239 outbreaks worldwide between 2011 and 2016 [15]. Concerning the foodborne route, vegetables and fruits, especially those which are consumed raw, probably suppose the highest risk. These foodstuffs get contaminated mainly by fecally contaminated irrigation water or due to poor hygiene measures during handling and processing [16]. As a result, Cryptosporidium has become becomes the fifth most important foodborne pathogen [17]. Cryptosporidium is not only present in vegetables and fruits, but also in other foods, like dairy products, meat, or shellfish (reviewed in [18]). Many studies have demonstrated the presence of this protozoan in shellfish (i.e., [19][20][21][22][23][24][25][26][27][28]). Shellfish filter and accumulate Cryptosporidium oocysts from water, which proves that Cryptosporidium is present in different water sources, including estuarine and coastal ecosystems. Therefore, fishes could also be infected or contaminated by Cryptosporidium spp.

Evidence of Fish as a Source of Zoonotic Cryptosporidium spp. Transmission
Various experimental studies took place in the 90s in order to ascertain if low vertebrates, namely fish, reptiles, and amphibians, could be infected by zoonotic Cryptosporidium parvum. In 1995, Arcay et al. [51] succeeded in infecting two fish species (Hypostomus plecostomus and Hoplosternum thoracatum), as well as different species of reptiles and amphibians, with C. parvum oocysts from human feces. Histological sections of experimentally infected animals revealed different phases of the Cryptosporidium cycle.
Moreover, oocysts recovered from these animals were infectious to laboratory mice. One year later, Graczyk et al. [52] tried to bear out these results. Concerning fish, they inoculated infecting oocysts of C. parvum in nine bluegill sunfish (Lepomis macrochirus). Fish excreted oocysts in feces for up to 12 days post inoculation, but any developmental stage of the parasite was detected in gastrointestinal tissue sections. Therefore, it is probable that oocysts came from the inoculum and that fish were not truly infected. A possible explanation for the conflicting results between both studies could be the use of different C. parvum strains, since no molecular characterization was carried out. On the other hand, in Arcay et al. [51], fish were tested prior to the inoculation by coprological test, but this is a technique of low sensitivity, so it cannot be ruled out that these fish were previously infected with other Cryptosporidium species.
Freire-Santos et al. [53] conducted an experimental infection in 50 rainbow trout (Oncorhynchus mykiss) from a fish farm in La Coruña, Spain. Infected animals were separated in two groups according to the housing conditions: normal environmental parameters and stressful parameters. Fish in both groups showed Cryptosporidium-like structures in sections from stomach and pyloric region, the number of these structures being higher in animals under stressful conditions. Structures were tested positive against anti-C. parvum antibodies by immunofluorescence antibody test (IFAT). Even though these results seem to be in accordance with the Arcay et al. study [51], IFAT can cross react with other Cryptosporidium species and, in this case, rainbow trout were also tested only by coprological test.
The question of whether fish could be infected by zoonotic C. parvum or they simply retain and disseminate the oocysts remained without a clear answer. Years later, field studies began to be carried out. Reid et al. [36] analyzed gastrointestinal tissue scrapings from 227 cultured freshwater fingerlings, 255 wild marine fish, and 227 wild freshwater fish from Australia. Only 0.8% of fish (6/709) were positive for Cryptosporidium spp., all of them marine fish (2.4%; 6/255). One positive isolate from a school whiting (Sillago vittata) was genetically identical to C. parvum at both 18S rRNA and actin loci, subtyping at the gp60 locus as the subtype IIaA18G3R1. Subfamily IIa is the most common in calves and humans in North America, Europe, and Australia [54]. This was the first description of C. parvum in a fish not experimentally infected. Nonetheless, no histopathological examination was performed, so it was not possible to know if this was a true infection or a mechanical transport. Furthermore, another isolate was identified as C. xiaoi and two others as pig genotype II (now C. scrofarum). Currently, both species are also considered as zoonotic (reviewed in [32]).
Gibson-Kueh et al. [55] carried out a nested polymerase chain reaction (PCR) of 18S rRNA gene on 20 previously fixed tissues of cultured barramundi (Lates calcarifer) from Vietnam. Partial sequences were obtained from two formalin fixed tissue samples, which were closely related with C. parvum type B.
Similarly to [36], Koinari et al. [39] worked with three fish groups: 133 cultured fish, 205 wild freshwater fish, and 276 wild marine fish from Papua New Guinea. The genetic analysis of intestinal and stomach scrapings revealed seven positives between the three groups. Five isolates were identified as C. parvum at the 18S rRNA locus. At the gp60 locus, four of them subtyped in the IIa subfamily and the fifth was identified as C. hominis (mixed infection). Histopathological examination failed because of autolysis, so it was not possible to discern between a natural infection and a mechanical transport.
Certad et al. [45] analyzed 41 adult fish from Lake Geneva in France. In total, 37% (15/41) and the prevalence of Cryptosporidium was detected. Sequences of positive animals correspond to C. parvum (n = 13), C. molnari (n = 1), and C. parvum-C. molnari mixed infection (n = 1). Three different subtypes were identified at gp60 locus, all of them pertaining to the IIa subfamily. Histological analysis suggests a natural infection for the first time; intracellular Cryptosporidium-like bodies were found in 10 of 13 C. parvum isolates.
Couso-Pérez et al. [43] performed a study on 360 rainbow trout from a farm in Santiago de Compostela, Spain. In this case, aliquots of sediments from pyloric caeca and intestinal content were used in the analysis. A total of 33 samples (9.2%) were positives by IFAT. Eleven partial sequences of the SSU-rDNA were obtained and C. parvum was identified in seven of them. It was not possible to amplify gp60 and actin loci. The detection of a cluster of 10 oocysts in the pyloric caeca of one animal could indicate a true infection.
Another study by the same authors [47] analyzed the presence of Cryptosporidium in 613 brown trout (Salmo trutta) from Galician (NW Spain) rivers. Same samples (pyloric caeca and intestine) were used, resulting in 103 positive individuals by IFAT. Among these positives, C. parvum was molecularly identified in 47 fish (subtypes IIaA15G2R1 and IIaA18G3R1). Oocyst clusters were observed in five cases, suggesting a possible true infection.
Certad et al. [44] carried out two studies in France. A regional survey that was composed of 1508 fish from 31 different species showed a prevalence of 2.3% (35/1508). Eight of them were identified as C. parvum with four subtypes of the subfamily IIA represented. It was possible to identify compatible structures within the epithelial cells and in an apical position in just one case. The second study, at Boulogne-Sur-Mer, analyzed 345 fish. A prevalence of 3.2% (11/345) was detected, with two isolates identified as C. parvum (subtyping could not be attained). This is the first epidemiological and molecular data on fish cryptosporidiosis in European waters.
The latest evidence comes from Shahbazi et al. [46] who investigated the presence of zoonotic C. parvum in freshwater ornamental fish from different pet shops in Iranian Iran cities. Sixteen out of 100 animals were positive by histopathology; only two samples, corresponding to goldfish (Carassius auratus) were positive for C. parvum by nested PCR at the 18S rRNA gene [56]. One goldfish presented the intestine with emaciation, while the other was clinically healthy. Therefore, it was not possible to assess if it is a true or a mechanical infection. Molecular methods used in the different mentioned studies are summarized in Table 4. Gastrointestinal tissue scrapings are the most common sample used for the analysis of Cryptosporidium in fish. Nested PCR, and subsequent amplification products sequencing, has been the major technique used in all cases, 18S rRNA and gp60 being the most common target genes. Although some studies provide histopathological evidence of natural infection in fish hosts, this cannot be confirmed at present. However, in any case, fishes would be acting as mechanical vectors of zoonotic Cryptosporidium spp., contributing to the dissemination of oocysts in water and being themselves a potential source of human cryptosporidiosis. Currently, there is no evidence of fish to human transmission, but one study reports that urban anglers are at high risk of contracting C. parvum infection while fishing and consuming fish [57].

Giardia duodenalis
Giardiasis caused by Giardia duodenalis (Diplomonadida, Giardiidae) is a very prevalent disease worldwide, responsible for approximately 280 million cases of diarrhea every day [58]. Currently, eight species of Giardia exist, but only G. duodenalis can infect humans and, therefore, it is the only one considered as zoonotic [59]. G. duodenalis is a species complex with eight different assemblages (A-H) recognized [60]. Although assemblages A and B have been traditionally considered the only ones that can infect humans, assemblages C, D, E, and F have also been identified in people [61] (Table 3). Finally, assemblage A is subdivided into three different subtypes (AI-AIII), subtype AII being the most commonly detected in humans [62]. Recently, some authors have proposed to reconsider the different G. duodenalis assemblages as a different species based on traditional taxonomic descriptions and biological differences, mainly host specificity [12,63] (Table 5). Similarly to Cryptosporidium, different genes have been used to characterize Giardia species and genotypes, including the SSU rRNA, the ITS, the Giardia specific β-giardin, the triosephosphate isomerase (TPI), the glutamate dehydrogenase (GDH), or the elongation factor 1 (ef1) genes [12] (Table 6). Table 5. Recognized Giardia species, their host range, and zoonotic potential.

Giardia Species
Giardia Species According to [  Cysts of Giardia are environmentally resistant. Therefore, transmission occurs directly via fecal-oral or through consumption of contaminated water or food [62]. There are many documented outbreaks of waterborne giardiasis [74], however foodborne published outbreaks of giardiasis are few (probably under-reported). Even so, G. duodenalis is the 11th foodborne parasite according to the World Health Organization [17] (Cryptosporidium occupies the fifth position). The foodstuffs most commonly implicated in these foodborne outbreaks are fresh products [74], particularly those consumed raw or undercooked. Similar to Cryptosporidium, Giardia cysts arrived on fruits and vegetables by fecal-contaminated irrigation water or by handlers with poor hygienic habits [75].
G. duodenalis is present in different water sources, which become contaminated by human or farm animal discharges, or even by wildlife feces [8]. Like Cryptosporidium, G. duodenalis arrives at marine environments and bioaccumulates in shellfish [20,21,26,27,[76][77][78][79][80][81][82][83][84][85]. It would be possible that fish become infected by zoonotic G. duodenalis (due to the broad spectrum of hosts of this parasite) or, in other cases, they could act as a mechanical transmitter (like shellfish are). Therefore, potentially, the consumption of fish could be a new source of giardiasis in humans. In the next section, we review and discuss the potential role of fish as a source of G. duodenalis transmission.

Evidence of Fish as a Source of Zoonotic Giardia duodenalis Transmission
Lassek-Neselquist et al. [86] performed the first study on G. duodenalis molecular epidemiology in marine vertebrates. In the context of this study, G. duodenalis was reported for the first time in feces and gut content of a single analyzed fish, specifically in a thresher shark (Alopias vulpinus). The authors identified a human-infecting G. duodenalis haplotype of assemblage B in this individual; this haplotype corresponded to a GDH sequence.
Nonetheless, the first study aimed at the detection of different species or assemblages of Giardia in fish was conducted two years later in Australia by Yang et al. [87]. In this study three groups of fish were analyzed, cultured fingerlings (n = 227), wild freshwater fish (n = 227), and wild marine fish from coastal and estuarine areas (n = 255). Analysis was performed on intestinal and stomach tissue scrapings and consisted of four different loci analysis by PCR and a histological examination. In summary, 27 of the 709 fish (3.8%) were positive, 8.4% of the cultured fingerlings (19/227; Lates calcarifer, Acanthopagrus butcheri, Argyrosomus japonicus, and Pagrus auratus), 2.7% of the wild marine fish (7/255; Mugil cepahalus), and 0.4% of the wild freshwater fish (1/227; Galaxias occidentalis). G. duodenalis assemblages A, B, and E, in addition to G. microti were identified. It is important to remark that Giardia cysts and trophozoites were detected in large numbers in 10 positive samples. This finding, together with the fact that analyses were conducted on tissue scrapings (instead of feces) seems to demonstrate that fish are truly infected, and they are not mere mechanical vectors.
Given the detection of zoonotic genotypes in fish [87], Ghoneim et al. [88] carried out another study in fish in Egypt. They analyzed a total of 92 animals, fish from farms and wild fish from the Nile River. In this case, they obtained fecal samples instead of intestinal and stomach tissue scrapings. Firstly, they used an enzyme-linked immunosorbent assay (ELISA) to screen for positive samples and secondly, they performed a specific duplex PCR on positive samples for identification of zoonotic assemblages A and B [89]. In this study, the only gene amplified was the TPI gene. Three fish were tested positive (3.3%), one farmed fish (Tilapia nilotica) and two wild sea mullets (Mugil cephalus). In all three cases, assemblage A was identified. Table 7 summarizes molecular approaches used in these three studies. Two of them have used feces (or feces and gut content) as choice sample for Giardia duodenalis detection. Similarly, to Cryptosporidium, nested PCR, and subsequent sequencing, has been the most common technique. The most frequent targeted genes have been β-giardin, GDH, and TPI ones. Since then, no further studies have been conducted on the occurrence of zoonotic G. duodenalis assemblages in fish. The only posterior reference was that of Nematollahi et al. [90]. These authors carried out a study on histopathological detection of endoparasites in freshwater ornamental fish in Iran. In this study, Giardia was observed in the intestine of one Oscar fish (Astronotus ocellatus). However, no molecular identification was performed.
Based on the study of Yang et al. [87], Tysnes et al. [91] decided to try an experimental model on the use of zebrafish (Danio rerio) as a G. duodenalis infection model. Except for assemblage A, other G. duodenalis assemblages seem to be refractory at being cultivated in vitro or to being established in laboratory animals [92]. The likely Giardia infection in wild and cultivated fish detected by Yang et al. [87] could be the solution, providing a novel animal model. Zebrafish were inoculated with viable cysts of G. duodenalis assemblages A and D and the assessment of the infection was made by different methods (direct microscopy, IFAT, direct cultivation, and histology). Giardia was not detected by direct microscopy or histology, and direct cultivation was also not possible. Despite this, IFAT demonstrated the retention of cysts in the intestine, even some days after the inoculation. However, the number of cysts was decreasing. Therefore, the results were inconclusive, and more research is needed to establish if there could be long term infection by G. duodenalis.

Toxoplasma gondii
Toxoplasma gondii (Apicomplexa, Sarcocystidae) is a well-known zoonotic parasite all over the world. It affects virtually all warm-blooded animals and both terrestrial and aquatic species. It is a widespread parasite of high concern for in public health which is the cause of severe illness and even death in the case of acquired toxoplasmosis in immunocompromised patients. It also causes spontaneous abortion and congenital birth defects when the infection is acquired during pregnancy [93]. It has been calculated that approximately 30% of the world's population is infected [94]. Toxoplasma gondii is a single species which mainly consists of three clonal lineages (type I, II, and III strains). Type I strains are known to cause acute mortality in murine models, while types II and III, lead to chronic infections. In Europe and North America, type II and, to a lesser extent, type III strains, are the most prevalent. Furthermore, in North America, an atypical sylvatic strain circulates in wildlife (haplogroup 12).
Conversely, in South America, the genetic diversity is much higher, with several atypical strains having been identified [95].
Intermediate hosts, including humans, can be infected by three routes: vertical transmission from mother to fetus (not discussed here), ingestion of tissue cysts from infected animals, and ingestion of oocysts through contaminated water, soil, or food [96]. It must be remarked upon that the only definitive host able to eliminate infecting oocysts are felids, with special attention given to domestic and feral cats (Felis silvestris catus).
Traditionally, most human infections have been considered to be acquired by the ingestion of raw or undercooked meat containing cysts [97], however, sanitary improvements in animal production and the food-processing industry have reduced this associated-risk [98]. Unlike G. duodenalis and Cryptosporidium spp., only two outbreaks have been associated with fresh products contaminated with oocysts [99,100].
Until recently, environmental transmission had been the less studied route, despite the high resistance of T. gondii oocyst [101]. Toxoplasma has come to be considered as a significant waterborne parasite, only in the last few decades when waterborne outbreaks by the consumption of water containing oocysts have been detected [102][103][104][105][106]. Several sea mammals have been found to be infected by T. gondii [107], setting up evidence that transmission between terrestrial and marine environments takes place and highlighting the potential significant role of waterborne transmission.
Aquatic environments, including estuarine and marine ones, get polluted by sewage and freshwater runoffs carrying T. gondii oocysts [108]. In coastal environments, oocysts can be accumulated by bivalves. This bioaccumulation has been demonstrated experimentally [109][110][111] and has been detected in wild and cultivated bivalves [24][25][26]85,[112][113][114][115][116][117][118][119][120][121]. Nonetheless, this cannot explain how offshore cetaceans that do not feed on bivalves come into contact with these parasitic oocysts and acquire the infection. Alternative pathways have been proposed, for instance, the marine macroaggregates, which are associations of micropolymers with bacterial biofilm, algae, and other microorganisms [84]. Another possibility is that fish, not only shellfish could act as mechanical carriers of T. gondii oocysts, disseminating these during migration [122]. This possibility implies a new potential source of T. gondii transmission to humans, by the consumption of raw or undercooked contaminated fish.

Evidence of Fish as a Source of Zoonotic Toxoplasma gondii Transmission
In 2005, Omata et al. [123] carried out the first attempt to test if T. gondii were be able to multiply and persist in cells of a cold-blooded host, the goldfish (Carassius auratus). Tachyzoites of the Beverley (type II) and RH (type I) strains penetrated and multiplied in oviduct epithelial cells of goldfish in vitro at 37 • C. However, at 33 • C no tachyzoite penetrated into cells. Additionally, an in vivo inoculation test was conducted at 37 • C by injecting tachyzoites intramuscularly in goldfish. Fish were euthanized on days three and seven post inoculation and T. gondii detection was performed by DNA detection (T. gondii B1 gene, [124]) and mice bioassay. In both cases, the parasite was detected on day three, but not on day seven. Based on these authors' results, T. gondii is not transmissible to freshwater fish.
Taghadosi et al. [125] used a different approach to check if fish, in this case, farmed salmonids in Iran, could be primarily infected by T. gondii. They obtained 50 sera samples and tested them for T. gondii specific IgM and IgG by ELISA. All samples were negative for IgG, but five of 50 samples were positive for IgM, indicating a possible recent infection in these fishes.
Similarly to the case of G. duodenalis, Sanders et al. [126] tried a new in vivo model of T. gondii infection using the zebrafish. Fish kept at 37 • C were intraperitoneally injected with tissue cysts bradyzoites from experimentally infected mice; two strains, corresponding to types II and III were used. T. gondii tachyzoites were observed up to seven days post injection in several fish tissues, which had clinical signs before they died or were euthanized. This experiment was repeated using cysts from mice brains and cultured tachyzoites and the same results were obtained. These authors were able to establish a successful experimental T. gondii acute infection in a cold-blooded host, the zebrafish, for the first time.
The opposite results obtained by Omata et al. [123] in goldfish and Sanders et al. [126] in zebrafish were able to rely on a different experimental approach: different host species, different inoculation site (intramuscular and intraperitoneal, respectively), and a different method of analysis (cytology, PCR and mouse assay vs. histopathology). Nonetheless, in both cases, high temperature (37 • C) seems to be an important limiting factor. In agreement with this finding, Yang et al. [127] showed that T. gondii invasion and replication in fish cells (in vitro study) is temperature-dependent: the minimum temperature for tachyzoites penetration is 27 • C, while a minimum of 30 • C is necessary for their replication. Furthermore, Yoshida et al. [128] have been successful in using the zebrafish as a novel model in vivo for studying T. gondii replication and interaction with macrophages. On the other hand, results from Sanders et al. [126] are in agreement with the findings of Taghadosi et al. [125], since both provide evidence of T. gondii primary infection in fish.
Previous studies seem to show contradictory or unclear results about the possibility of T. gondii infecting cold-blooded hosts, such as fish. For the moment, there is not enough evidence and further research is needed, especially concerning studies on wild fish populations instead of experimental infection studies.
In spite of this, even if fish could not be naturally infected by T. gondii, they could still be a source of infection by accumulating and spreading oocysts, just as shellfish do [109][110][111]. To test the hypothesis of fish acting as biotic vectors for T. gondii, Massie et al. [122] exposed two migratory filter-feeding marine fish species, northern anchovies (Engraulies mordax) and Pacific sardines (Sardinops sagax), to T. gondii oocysts. Fish were euthanized at 2 and 6-8 h after exposure and their alimentary tracts were recovered and analyzed. Two hours post-exposure T. gondii was detected by PCR in 75% of northern anchovies and 60% of Pacific sardines, while 6-8 h after exposure 50% of the northern anchovies were positive. Concerning mice bioassay, 30% of mice fed samples from fish exposed to 100,000 oocysts/L manifest compatible clinical signs and were tested positive for T. gondii B1 gene. None of fish exposed to lower oocyst concentration became infected. Thus, both migratory filter-feeding species are able to filter T. gondii oocysts and to retain these oocysts infective, for at least 8 h, indicating that marine fish can act as mechanical vectors. Nonetheless, these results were obtained under experimental conditions and for a short-period duration.
In 2014, Zhang et al. [116] conducted a field study to detect T. gondii in the digestive tissues of cultured fish and shellfish in China. Four fish species were included: 309 Carassius auratus, 98 Momopterus albus, 456 Hypophthalmichthys molitrix, and 309 Cyprinus carpio. Only one H. molitrix (0.8%) tested positive by PCR for T. gondii ITS-1 region, which is markedly conserved [129]. This was the first study to find T. gondii in naturally exposed fish.
Aakool and Abidali [130] carried out a study to detect T. gondii in different tissues (muscle, liver, intestine, and gills) of freshwater Cyprinus carpio from the Tigris River (Iraq). Sixteen out of 24 samples were positive by real time PCR (rt-PCR; T. gondii B1 gene [131]) for T. gondii in the intestine, as well as two gill samples.
More recently, Marino et al. [98] carried out a large field study in Sicily, comprising 1293 edible marine fish species from extractive fisheries. These authors not only analyzed gastrointestinal tissues, but also gills and skin-muscle samples. For example, gills could retain T. gondii filtered from seawater and skin-muscle could be contaminated by contact with sea-beds, for example. Samples were pooled by species and sample type in 147 groups. Of these, 32 samples (21.8%) tested positive on rt-PCR [132]; positive samples were detected in 12 of 17 species and in all three-types of samples. Among the positive species, there was heterogeneity concerning the type of habitat and dietary habits. These findings provide more evidence of the possible role of marine fish in the transmission of T. gondii and, therefore, the emergence of a possible new seafood disease.
A summary on molecular methods used in Toxoplasma gondii detection in fish is shown in Table 8. Unlike Cryptosporidium and Giardia duodenalis, a broad range of samples have been analyzed in these studies, highlighting muscle and intestine. In this case, rtPCR and subsequent sequencing, has been the most applied technique. Genes used to molecularly identify Toxoplasma gondii are mainly two: T. gondii B1 gene and a 529-bp DNA repeat element from the parasite genome (GenBank accession no. AF146527).

Conclusions
The studies discussed above do not allow us to determine whether fish can be naturally infected by protozoa Crypstosporidium spp., Giardia duodenalis, and/or Toxoplasma gondii or they simply act as mechanical transporters. Nevertheless, even if fish are only mechanical carriers, they could be a "new" source of infection for people. Field studies demonstrate that all these three zoonotic protozoa are present both in freshwater and marine fishes, including species of commercial interest. Although fish seem to accumulate the parasitic forms mainly in the gastrointestinal tract, which is not consumed, there is still a risk of infection because of cross-contamination of the fish meat during evisceration and processing. This potential new risk could increase the list of fishborne parasitic infections, necessitating more studies to achieve a complete understanding of the possible role of fish in human cryptosporidiosis, giardiasis, and toxoplasmosis.

Conflicts of Interest:
The authors declare no conflict of interest.