Molecular detection of Blastocystis from animals in Italy: subtypes distribution and implications for the zoonotic transmission

Background: Blastocystis is a common intestinal protist distributed worldwide infecting humans and a wide range of domestic and wild animals. It exhibits an extensive genetic diversity and so far, 26 subtypes (STs) have been identied in animal hosts, ten of them (ST1-ST9 and ST12) reported in humans with varying prevalence. Since several STs are common to humans and animals it has been proposed that a proportion of human infections may have a zoonotic origin. Aims of the present study were to: 1) genetically detect Blastocystis in faecal samples of farmed animals and wild carnivores; 2) investigate the distribution of Blastocystis STs in different animal hosts; 3) provide a rst study on the Blastocystis STs circulating between animals and humans in Italy. Methods: Fresh faecal samples (N=269) were collected from carnivores and farmed animals in different Italian provinces and submitted to genomic DNA extraction and PCR amplication followed by both sequence and phylogenetic analysis (Neighbour Joining and Maximum Parsimony) Results: Blastocystis was detected in 50% of the farmed animals (42 out of 84), and 19 of them were successfully subtyped. Conversely, all the faecal samples (N=185) from domestic and wild carnivores (dogs, cats, foxes) tested in the present study, resulted negative. Phylogenetic analysis showed the nding of ST5, ST7, ST9 and ST10 in the samples from animals. The comparison with sequences of Blastocystis STs previously detected from humans in Italy showed the ST7, as a potential source of zoonotic transmission. Conclusions: The present study represents the widest epidemiological survey so far performed Italy. Further epidemiological studies using and to dene pathways Data curation, and


Background
Blastocystis is a common intestinal protozoon distributed worldwide, infecting humans and a wide range of domestic and wild animals. Molecular studies based on sequence analysis of the small subunit (SSU) ribosomal RNA evidenced an extensive genetic diversity allowing the identi cation in mammalian and avian hosts of at least 26 divergent lineages, termed subtypes (STs), which could be considered as separate species [1,2,3]. Ten of the 26 subtypes, ST1 to ST9, and ST12 have been identi ed in human samples and have also been reported in animals so far [4,5,6]. Therefore, it has been proposed that a portion of human infections may result from the zoonotic transmission of the protist. Interestingly, a recent molecular epidemiological survey carried out on commercially important sh species and marine mammals from North-East Atlantic waters, detected several STs occurring in both sh and marine mammals [7]. These ndings provided new insights into host range of Blastocystis STs including possible new STs [7]. However, the contribution of animal sources to human infection remains to be con rmed, since the direction of transmission routes to humans is uncertain.
In the recent years, the presence of Blastocystis has attracted attention also in Italy, where few epidemiological surveys have been published so far, demonstrating the occurrence in humans of seven STs, including those considered as zoonotic subtypes (i.e., ST1, ST2, ST3, ST4, ST6, ST7, ST8) [8,9,10,11]. Phylogenetic and genetic diversity analyses evidenced at the intra-ST level a high genetic homogeneity in ST4, while high values of nucleotide diversity and different haplotypes were observed in isolates identi ed as ST1 and ST2 [9]. This genetic variability seems to support the low host speci city of these two STs [12]. Indeed, the sequences identi ed as ST1 and ST2 showed 100% identity with isolates from a wide range of animals, including monkeys, cattle, pigs, dogs and non-human primates, highlighting that these subtypes of animal origin are zoonotic STs and able to infect humans at different frequencies [9,10].
Despite a prevalence rate of about 7% has been reported in Italian population [11,13] along with the identi cation of zoonotic STs in humans [8,9,10], studies aimed to identify the potential animal reservoirs of human Blastocystis infection in Italy are scant. So far, the detection of different STs from animals was carried out in few animal categories such as zoo mammals [14] and dogs [15].
Aims of the present study were to: 1) genetically detect Blastocystis in faecal samples of farmed animals and wild carnivores collected in Italy; 2) investigate the distribution of STs in different animal hosts, and 3) provide a rst comprehensive study on the STs circulating between animals and humans in Italy.
Faecal samples from wild foxes were gathered during necropsy from the large intestine, whereas faeces from pets and farmed animals were directly collected from rectal ampulla to avoid environmental contamination.
Animals included in the study were in a good state of health and did not receive pharmacological treatment in at least four weeks before the sample collection.

DNA extraction and sequencing of Blastocystis isolates
Genomic DNA was extracted from each faecal sample, by using the Faecal DNA kit (Bioline, UK) according to the manufacturer's protocol. A fragment of about 500 bp from the SSU rDNA gene was ampli ed using the primers Blast 505-532 (5'-GGAGGTAGTGAC AATAAATC-3') and reverse Blast 998-1017 (5'-TGCTTTCGCACTTGTTCATC-3') following the protocol proposed by Santín et al [16]. Positive samples were further ampli ed using the primers RD5 (5'-ATCTGGTTGATCCTGCCAGT-3') and BhRDr (5'-GAGCTTTTTAACTGCAACAACG-3'), in order to compare sequences obtained in this study with human isolates from a previous survey [9], using the PCR-conditions described in Scicluna et al [17].
Amplicons from the barcode region of about 600 bp were sequenced using the forward ampli cation primer (Bio-Fab Research, Rome, Italy). The resulting chromatograms were analysed and edited in the computer software Chromas version 2.33 (Technelysium Pty Ltd, Australia). The sequences obtained were compared to the sequences of Blastocystis STs, previously deposited in GenBank™ and available at the website http://pubmlst.org/blastocystis/ [18] by using the BLAST application. They were aligned with SSU rDNA sequences representing all STs using Clustal X software version 2.1 for comparative analysis and an alignment gure was carried out by using Bioedit software [19]. The subtypes (STs) were identi ed by determining the exact match (100%) or closest identity (99%), according to the classi cation of the subtypes given by Stensvold et al [20]. In order to discriminate alleles of Blastocystis based on 18S rRNA gene, generated sequences were subjected to online software http://pubmlst.org/blastocystis/.

Phylogenetic and genetic diversity analyses
The phylogenetic analysis of the sequence dataset obtained in the present study was carried out by using MEGA6 software [21]. The phylogenetic elaborations from the animal sequences inferred from Neighbour Joining (NJ) and (MP) Maximum Parsimony analyses were carried out considering TVM+G (G=0.207) as the best substitution model, as implemented in JModeltest2 [22]. The phylogenetic analysis considering sequences from animals (present work) and those from humans obtained from previous studies [9,10,11] was carried out by performing a NJ by MEGA6 software [21], using the same substitution model previously selected.
Blastocystis lapemi (GenBank™ accession no. AY590115) was used as outgroup to root the trees. Bootstrap values >70 were considered as a good support to the nodes of the phylogenetic trees [23].

Results
Out of the total (N= 84) faecal samples from farmed animals (calf, chicken, cow, deer, donkey, duck, goat, horse, ostrich, peacock, pheasant, pig, sheep, turkey), 42 (50%) resulted positive at Blastocystis by molecular ampli cation. The number of positive samples varied among the animal host species, as detailed in Table 1. Conversely, all the faecal samples (N=185) from domestic (dogs, cats) and wild (foxes) carnivores tested, turned out to be negative.
A total of 19 samples from farmed animals showed a single infection with a speci c ST.
The sequence analysis showed a high identity (99-100%) to homologous sequences of Blastocystis previously reported in GenBank™, allowing the preliminary BLAST identi cation of four distinct STs (i.e. ST5, ST7, ST9, ST10). In the remaining 23 PCR products, the chromatograms revealed the presence of double signals, suggesting that mixed infections by different STs in the same host occurred. Therefore, subtypes from such samples were not assessed since the ST identi cation protocol followed in this study, i.e. PCR ampli cation coupled with sequencing, preferentially ampli ed and allowed to identify the predominant subtypes. Nevertheless, as previously demonstrated, the cloning of the PCR product or next generation amplicon sequencing is required to discriminate different subtypes in the same host [3,24,25,26].
The sequences alignment of the four STs found in the faecal samples from animals here analysed is reported in Fig. 1. The phylogenetic analysis showed the Blastocystis isolates of the present study clustering in four distinct clades each of them well supported at the bootstrap analysis by both MP and NJ phylogenetic analyses (Fig. 2), also including representative sequences retrieved from GenBank™ for those STs. Thus, the tree topologies of MP and NJ were congruent in showing the Blastocystis isolates as belonging to the subtypes ST5 (N= 14), ST7 (N= 1), ST9 (N= 1), ST10 (N= 3) with strong robustness (from 99% to 100%). In detail, the isolates from pigs (N= 9), goats (N= 4) and deer (N= 1) clustered with ST5 reference sequences (MG000956, MF541105). The sequence from chicken (N=1) collected in the site 1/ME clustered with ST7 (KP233733) sequence deposited in GenBank™, while that from pheasant (N=1), collected in the site 2/ME, showed high identity with ST9 sequences from GenBank™ (MK861942, MK861944). Finally sequences from sheep (N=3) formed a distinct clade with ST10 deposited sequence (MF974614) (Fig.2).

Discussion
The present study represents the widest epidemiological survey performed so far in animals from Italy. It also increases the current knowledge on Blastocystis STs distribution and their circulation in 17 animal species in Italy evidencing the protist in the 50% of the analysed livestock.
Mono-subtype infections were detected in 45.23% (19/42) of the animals while in 54.76% (23/42) of the positive samples mixed infection (more than one subtype present in a single sample) has been hypothesized, as evidenced in previous surveys on animals and humans [25,26,27].
All the samples from dogs, cats and foxes resulted negative in this study for Blastocystis STs. Similar results were previously obtained in other studies, which showed dogs not infected by Blastocystis [28,29,30,31,32] or infected with a low or moderate prevalence value [33,34,35,36,37]. Otherwise, the prevalence reported for stray dogs from India living mostly in areas of poor sanitation and hygiene was signi cantly higher (24%) [38]. Recently, Blastocystis ST3 was identi ed in 21.2% faecal samples from dogs housed in 6 different shelters located in Northern Italy, thus evidencing that sheltered dogs seem to be more at risk of harbouring Blastocystis than owned ones [15]. Similarly, in a survey carried out in the USA, Blastocystis was detected in 9.7% of examined shelter dogs while no positivity was found in owned dogs [34]. Negative results observed in this study could easily be explained by a sampling bias as more than half of the dogs (42 out of 71) were from the same kennel where probably Blastocystis did not circulate among dogs or in the environment, or by the daily cleaning of cages and the small number of animals per cage. The remaining 29 dogs were owned animals, frequently submitted to antiparasitic treatments. Anyway, this low or moderate prevalence of Blastocystis observed in numerous canine cohorts and con rmed also in this study, together with the absence of a dog-speci c/predominant ST, strongly suggests that dogs are unlikely to be natural reservoir of Blastocystis. Similarly, cats seem to play a minor role in the epidemiology of Blastocystis, although limited studied has been conducted so far on domestic and stray cats [25,32,37,40]. In the present study, only 3 feline faecal samples were examined corroborating these observations, but more specimens should be examined in order to assess the role of cat in the epidemiology of Blastocystis. Furthermore, Blastocystis was apparently absent in the investigated red fox wild population, although it was recently found in 2.2% of red foxes in Spain and about 2% of artic foxes in China [25,37].
Conversely, the results gathered from isolates of livestock evidenced a high percentage (50%) of farmed animals colonized by Blastocystis. This nding emphasizes the potential risk of the zoonotic transmission to humans. To corroborate this evidence, we compared the sequences from animals obtained in the present survey with human isolates identi ed as ST1, ST2, ST3, ST4, ST6, ST7 and ST8 from our previous studies [9,10,11].
Phylogenetic analysis evidenced a high similarity (100%) among the ST7 isolates from human patient (Hum190) with the chicken (isolate no.35) supported by high bootstrap value of this clade including also the ST7 reference sequences (Fig. 3). Accordingly, allele analysis showed the same allele (100) observed in both ST7 isolated from human and in chicken. Furthermore, the human ST6 isolates (N=7) previously detected in Mattiucci et al [9] clustered with reference sequences from chicken and humans available in GenBank™ database (Fig.3).
The identi cation of such zoonotic STs in edible animals as poultry emphasizes the potential risk of Blastocystis transmission directly from animals to humans. We cannot depict the precise mode of transmission, however, we could hypothesize that it may happen through the human handling of those farmed animals, as well as through the consumption of products of animal origin (e.g., eggs or cheese) [41]. Furthermore, the identi cation of the allele 100 from ST7 isolates in human and chicken represents a remarkable evidence of the direct transmission of this subtype from animals to humans.
In the recent years, several studies willing to address the issue of Blastocystis pathogenicity in humans, suggested that it could be related to genetic differences in the subtype-or strain-level of distinct STs [42,43], correlating the ST1, ST4, and ST7 with pathological alterations in humans, while ST2 and ST3 have been identi ed as non-pathogenic [43,44,45]. A recent multi-locus sequence typing analysis of Blastocystis ST3 and ST4 has provided valuable insight into genetic variation within and between the two subtypes, evidencing high or low level of genetic diversity for ST3 and ST4, respectively [46]. Similar results were obtained in our previous survey [9] where 3 haplotypes (H1, H3, H7) have been identi ed in ST3 isolates, while a single haplotype (H2) was observed in ST4 symptomatic patients. Thus, we suggested that intra-subtype diversity showed by ST3 and ST4 could be linked to the evolutionary history of Blastocystis subtypes and that ST3 may have co-evolved with human hosts over a longer period than ST4; this latter, instead, may has extended its range to humans more recently, and, therefore, shows a low genetic variability.
In this study, we may support this hypothesis as we evidenced the same allele within the ST7, which has been reported to be strongly associated with gastrointestinal symptoms in humans and showed pathogenic properties, not observed in other STs [43,47]. Indeed, features of gastro-intestinal symptoms were also found in human patients, here showing the same allele 100, at the ST7 (Mattiucci and Gabrielli, personal communication). Therefore, we could hypothesize that ST7, exhibiting such genetic low variability, was recently transmitted to humans and, accordingly, it shows also a higher pathogenetic role to humans.

Conclusion
The outcome of this study may be considered a starting point to de ne the distribution of Blastocystis subtypes in different animal hosts in Italy and to study pathways of the zoonotic transmission of the protist. However, further parasitological and molecular investigations need to be carried out with the aim to evaluate other potential animal reservoirs of Blastocystis in Italy. For instance, despite in this study several bird's species (chickens, pheasants, turkeys, ostriches, ducks and peacocks) have been analysed, we did not nd the Blastocystis ST6 in any of the tested birds. Conversely, in previous surveys, ST6 was detected in human subjects (3.2% of the examined), although it is considered an "avian subtype" rarely occurring in humans [1,6]. Also, as the sample size of some farmed animal species analysed in the current study was not large enough, further surveys should be planned including other avian species and more samples from the above tested animals as well as from other geographical areas of Italy.
Thus, it becomes evident that a greater knowledge of the evolutionary history of Blastocystis subtypes could, in future, explain any pathogenic aspects related to distinct subtypes. Studies on the genetic variation within and among identi ed subtypes may elucidate possible co-evolutionary aspects, and provide data for understanding Blastocystis biology, host-parasite interactions and pathogenicity to humans.
Lastly, as Blastocystis has been reported as a neglected waterborne protist and recently detected in tap water [45] and drinking water treatment plant [48], further investigations are needed to assess this source of transmission to human and animal hosts also in Italy.
In conclusion, these ndings represent only the tip of the iceberg about the epidemiology of Blastocystis in Italy and suggest the need to apply molecular and phylogenetic analyses to demonstrate the transmission dynamics among humans, animals and environment of this so far debated plastic protist.

Availability of data and materials
All data generated or analysed during this study are included in this published article. The raw datasets are available from the corresponding author.
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable. Multiple sequence alignment inter-and intra-subtype carried out with Bioedit software and performed by using sequences obtained from animals analysed in this survey. Dots represent identity and dashes correspond to gaps.