Prevalence and Molecular Characterisation of Blastocystis sp. Infecting Free-Ranging Primates in Colombia

Infection with Blastocystis sp. has been reported in free-living and captive non-human primates (NHPs); however, surveys on Blastocystis sp. from north-western South America are scarce. This study aimed to identify Blastocystis sp. in free-ranging NHPs living in Colombia. A total of 212 faecal samples were collected from Ateles hybridus, Cebus versicolor, Alouatta seniculus, Aotus griseimembra, Sapajus apella, and Saimiri cassiquiarensis. Smears and flotation were used for morphological identification. For samples microscopically classified as positive for Blastocystis sp., we used conventional PCR to amplify and sequence two regions of the SSU rRNA gene and used Maximum Likelihood methods and Median Joining Network analyses for phylogenetic analyses. Via microscopy, 64 samples were Blastocystis sp. positive. Through molecular analyses, 18 sequences of Blastocystis sp. subtype 8 (ST8) were obtained. Strain and allele assignment together with a comparative phylogenetic approach confirmed that the sequences were ST8. Alleles 21, 156, and 157 were detected. Median Joining network analyses showed one highly frequent haplotype shared by specimens from Colombia and Peru and close relationships between haplotypes circulating in NHPs from Colombia, Ecuador, Brazil, and Mexico. This survey could support the elaboration of a more accurate epidemiological picture of the Blastocystis sp. infecting NHPs.


Introduction
Non-human primates (NHPs) have been found infected with a diverse array of intestinal parasites, including many protozoans and protists. Blastocystis is one of the most widespread enteric protists infecting animals such as reptiles, birds, and mammals (including humans). Its high occurrence in animal and human caeca and large intestine has raised a debate regarding its pathogenic role [1], as is it frequently found in asymptomatic individuals [2]. However, Blastocystis sp. may cause clinical signs including abdominal pain, constipation, and flatulence with diarrhoea [3]. Blastocystis has been found infecting NHPs, both free-living and captive platyrrhines and catarrhines, in areas where different subtypes (STs) have been identified: ST1-5, ST8, ST13, ST15, and ST39 [4][5][6][7]. Detailed information about Blastocystis sp. prevalence and STs occurrence in free-ranging and captive primates around the world has been recently reviewed by Hublin et al. [2].
As surveys on Blastocystis sp. infecting NHPs from Colombia are still scarce, this study aimed to increase the knowledge on the molecular epidemiology of Blastocystis sp. infecting free-ranging primates living in fragmented forests in Colombia.

Sampling
Fieldwork was carried out between December 2019 and February 2022 in seven locations in Colombia: San Juan in Santander Department; Cumaral, Cabuyaro, Guacavía, and Villavicencio in Meta Department; and Maní and Yopal in Casanare Department. Primates were followed from dawn to dusk, and one faecal sample per individual was collected from the soil immediately after defecation. For each sample, one aliquot was stored in 10% formalin solution, and another aliquot was stored in 96% ethanol solution. Overall, 212 samples were collected from free-ranging Ateles hybridus, Cebus versicolor, Alouatta seniculus, Aotus griseimembra, Sapajus apella, and Saimiri cassiquiarensis.

Microscopy Analyses
Samples stored in 10% formalin solution were used to perform smears with 1% iodine solution and 0.85% saline solution [25]. For each faecal sample, two microscope slides were mounted and systematically examined under a microscope using magnifications at 100×, 400×, and 1000×. Additionally, flotation with a salt-sugar solution was carried out. Each sample was placed in a 15 mL Falcon tubes with flotation solution and then centrifuged at 220 RCF for 5 min. Flotation solution was added until a slight positive meniscus was formed; then, a coverslip was placed on the top of the tube. After 10 min, the coverslip was removed and placed into a microscope slide. Thereafter, the meniscus was taken and placed into a new 15 mL Falcon tube, and flotation solution was added until the formation of a slight positive meniscus. A coverslip was placed on the top of the tube; after 10 min, the coverslip was removed and placed into a microscope slide with a drop of iodine solution. For each faecal sample, one microscope slide was examined using magnifications at 100×, 400×, and 1000× after the flotation procedure.

Molecular Analyses
For samples microscopically classified as positive for Blastocystis sp., the respective aliquots stored in 96% ethanol solution were individually subjected to DNA extraction using the Isolate II Fecal DNA Kit (Meridian Bioscience, London, UK) according to the manufacturer's protocol. Two conventional PCRs were performed using the pair of primers BhRDr-RD5 and Blast505-532-Blast998-1017 in order to amplify two different regions of the small subunit ribosomal RNA (SSU rRNA or 18S) gene according to published methods [8,26]. All PCR products were visualized on a 1% agarose gel stained with SYBR Safe, and positive samples were purified using Sure Clean Plus (Bioline, London, UK) and shipped to an external company for bidirectional sequencing (Eurofins Genomics). Sequences were manually edited using Trace implemented in MEGA7 [27] to infer consensus sequences by retaining only high-quality electropherograms and then checked for multiple peaks to rule out potential mixed infections. Thereafter, consensus sequences were used as input for BLAST search and strain/allele assignment using the PubMLST.org website. Two datasets named according to primers used (Scicluna and Santin) were obtained and used for the following analyses.
Alignments of sequences were carried out per each partial region of the SSU rRNA gene studied using ClustalW in MEGA7. Initial comparisons with the aim of inferring the strain identity and phylogenetic relationships were performed, including reference sequences of all available strains and a proper outgroup (Supplementary Table S1). The best evolutionary model was obtained using ModelTest implemented in MEGA7, and the Maximum Likelihood (ML) method was used to infer phylogeny with the aim of confirming relationships between the Blastocystis sequences identified herein and the other strains, for which statistical support at nodes was provided according to bootstrap. Moreover, evolutionary relationships between sequences belonging to the same strain circulating in NHPs in Mexico and South America (Table 1) were explored using Median Joining Network analyses [28] carried out with PopART [29].

Data Analyses
Quantitative Parasitology Software was used to calculate the prevalence of Blastocystis per primate species and study site with 95% confidence intervals.

Microscopy Analyses
Sixty-four samples (30.2%) were classified as positive for Blastocystis infecting NHPs from five of the seven study locations and including all NHP species except A. griseimembra (

Molecular Analyses
Overall, we obtained ten high-quality sequences from the PCR carried out with the BhRDr-RD5 primers and eight high-quality sequences with the Blast 505-532-Blast 998-1017 primers ( Table 2). For seven samples, sequences were obtained with both pairs of primers. According to the best match in terms of BLAST, all sequences were identified as Blasto-cystis sp. ST8 with 97.6-99.8% identity. All sequences were deposited in GenBank (see Supplementary Table for detailed accession numbers). Strain and allele assignment determined using the PubMLST website confirmed matches for ST8 for the barcoding region and the assignment to allele 21, with the exception of two sequences assigned to alleles 156 (OP329405) and 157 (OP329407). The results obtained by ML phylogenetic inferences with the T92+G+I model (G = 0.63 I = 0.56) described the cluster affiliation of the material analysed herein with reference sequences of ST8 available for birds from Japan and captive NHPs for both partial 18S datasets, showing 99% and 100% bootstrap support. The best ML consensus tree obtained from the Santin dataset (primers Blast 505-532 and Blast 998-1017) is shown in Supplementary Figure S1, and the best ML tree from the Scicluna dataset (primers BhRDr-RD5) provided the same topology and is available in Supplementary  Figure S2.
The Blastocystis sp. ST8 MJ network built with the dataset of sequences obtained with the primers from the research of Scicluna et al., 2006 [8] showed five haplotypes, with a main haplotype shared by most of the sequences of A. seniculus from this study (four from San Juan, four from Yopal, and one from Maní) and Peruvian samples of Aotus nigriceps, separated by three or more SNPs from haplotypes circulating in Alouatta palliata equatorialis in Ecuador. For the dataset of sequences obtained with the primers from the research by Santin et al., 2011 [26], four haplotypes were obtained, each one separated from another by one SNP. Two haplotypes were characterised in the Colombian samples circulating in A. seniculus from the Yopal and San Juan regions, which were separated by a central haplotype shared by Brazilian and Mexican specimens reported in several NHP species (Ateles sp., L. lagotricha, A. palliata, A. pigra, A. fusciceps). Networks are available in Figure 1.
BhRDr-RD5 primers and eight high-quality sequences with the Blast 505-532-Blast 9 1017 primers ( Table 2). For seven samples, sequences were obtained with both pair primers. According to the best match in terms of BLAST, all sequences were identified Blastocystis sp. ST8 with 97.6-99.8% identity. All sequences were deposited in GenB (see Supplementary Table for detailed accession numbers). Strain and allele assignm determined using the PubMLST website confirmed matches for ST8 for the barcoding gion and the assignment to allele 21, with the exception of two sequences assigned to leles 156 (OP329405) and 157 (OP329407). The results obtained by ML phylogenetic in ences with the T92+G+I model (G = 0.63 I = 0.56) described the cluster affiliation of material analysed herein with reference sequences of ST8 available for birds from Ja and captive NHPs for both partial 18S datasets, showing 99% and 100% bootstrap supp The best ML consensus tree obtained from the Santin dataset (primers Blast 505-532 Blast 998-1017) is shown in Supplementary Figure S1, and the best ML tree from Scicluna dataset (primers BhRDr-RD5) provided the same topology and is available Supplementary Figure S2.
The Blastocystis sp. ST8 MJ network built with the dataset of sequences obtained w the primers from the research of Scicluna et al., 2006 [8] showed five haplotypes, wit main haplotype shared by most of the sequences of A. seniculus from this study (four fr San Juan, four from Yopal, and one from Maní) and Peruvian samples of Aotus nigric separated by three or more SNPs from haplotypes circulating in Alouatta palliata equat alis in Ecuador. For the dataset of sequences obtained with the primers from the resea by Santin et al., 2011 [26], four haplotypes were obtained, each one separated from anot by one SNP. Two haplotypes were characterised in the Colombian samples circulatin A. seniculus from the Yopal and San Juan regions, which were separated by a central h lotype shared by Brazilian and Mexican specimens reported in several NHP species (At sp., L. lagotricha, A. palliata, A. pigra, A. fusciceps). Networks are available in Figure 1.

Discussion
In Latin America, even though Colombia is one of the countries wherein the majority of Blastocystis reports originate, surveys including free-ranging NHPs are still very scarce. In this study, we found positive samples for Blastocystis ST8, providing new information regarding free-ranging NHPs. Until this study was conducted, in Colombia, Blastocystis ST4 was the only identified ST infecting NHPs, which had been reported in two specimens of Alouatta sp. [11], while ST8 had only been found circulating in marsupials [11]. Accordingly, this is the first report of Blastocystis ST8 infection from Colombian wild NHPs, which was confirmed by phylogenetic analyses.
Blastocystis ST8 has already been found circulating in NHPs from South America, particularly in free-ranging mantled howler monkeys from Ecuador [12] and in captive A. seniculus, A. caraya, A. fusciceps, and L. lagotricha in Brazil [6]. Moreover, ST8 has been found in captive gibbons and in Varecia variegata from a zoo in Spain [4,30]. Human ST8 infections are rarely reported; however, a high prevalence of ST8 has been reported among primate handlers in the United Kingdom [31] and in symptomatic patients in Italy and Australia [32,33]. In Latin America, it has been identified in Brazil in patients with diabetes mellitus and in asymptomatic patients [34,35].
In this study, ST8 allele assignment was prevalently attributed to allele 21, which was previously reported in captive NHPs in Brazil [6] from sequences obtained with the primers used by Scicluna et al. [8], and in Colombia this allele has been found in samples from Didelphis marsupialis [11]. Network analyses allowed us to explore Blastocystis haplotype lineages and relationships among ST8 strains circulating in primates in Central and South America, for which a generally low level of variation was suggested and only slight differences in the 18S haplotypes were revealed in the Colombian specimens according to the sampling site (San Juan and Yopal). The Peruvian and Colombian sequences appeared to largely overlap, while the samples from Ecuador, Brazil, and Mexico displayed distinct haplotypes.
The analysis based on PCR amplification revealed a lower ability to detect positivity to Blastocystis sp. with respect to microscopy. Although unusual, this evidence could be related to factors that may jeopardize the efficiency of molecular characterisation, such as the presence of inhibitors in stool samples, low parasitic loads, and a lack of specificity of the primers employed, in which plant/ant sequences arising from items that are part of the host's diet are represented.
In this study, ST8 was identified in samples of A. seniculus, a species for which a preference for the upper strata of the forest has been observed [36]. A large Blastocystis sequence dataset including 30 genera of free-ranging and captive NHPs revealed a cryptic degree of host specificity for some STs; likewise, ST8 was primarily seen in arboreal NPHs native to South America and Asia [4]. Therefore, host specificity, behaviour, and ecological factors may be relevant in shaping the distribution and occurrence of different STs. It is worth underlining that the animals analysed herein did not show any symptoms related to gastrointestinal infections.
The present survey could support the elaboration of a more accurate epidemiological picture of the Blastocystis strains infecting NHP species. Surveys including other NHP species are strongly encouraged using both microscopy and molecular analyses, especially for NHP species listed as critically endangered, endangered, or vulnerable according to the IUCN Red List, as in the case of A. hybridus and C. versicolor included in the present survey. For a better understanding of Blastocystis sp. molecular epidemiology, new surveys including humans and other NHP species and aiming to explore the distribution, genetic variation, and host specificity are strongly encouraged.