Comparison of various diagnostic techniques for the detection of Blastocystis spp. and its molecular characterisation in farm animals in the United Arab Emirates

Blastocystis is a genus of common intestinal parasites of humans and a wide range of both wild and domestic animals (7, 26). A greater risk of Blastocystis infection was noted in people with close animal contact, suggesting its zoonotic potential (17, 25). Although infections are deemed to be associated with a variety of intestinal ailments such as irritable bowel syndrome, urticaria, and possibly gastrointestinal cancers (11, 27, 28), recent microbiome studies have shown that Blastocystis colonisation may be beneficial to human health (19, 21).

Molecular diagnostic tools based on small subunit ribosomal ribonucleic acid (SSU rRNA) subtyping have revealed the existence of extensive genetic diversity within Blastocystis (27). So far, 26 subtypes have been documented (13). Among them, nine (ST1 to ST9) have been identified in humans, all except the ninth being potentially zoonotic (7, 8).

Many studies conducted in different countries, including an earlier study performed by our group, revealed that subtype (ST)10 and ST14 were the most common subtypes detected in artiodactyls (sheep, cows and goats) (3). Unlike other researchers, we did not find any of the other subtypes such as ST1, ST3, ST4, ST5 or ST6 (4, 5). However, the artiodactyl sample size studied in our earlier report was limited and the information derived from it about the prevalence of Blastocystis infection would therefore have been inadequate to provide a detailed picture of the existence of potentially zoonotic subtypes in them and of their significance as carriers for human infections.

Microscopic investigation of wet-mount preparations and/or permanently stained slides is the main method of Blastocystis spp. detection (6). Nevertheless, the many different forms the parasite exhibits, the sporadicity of shedding of cysts and the parasite’s similarity to normal structures typically found in a stool sample present challenges to correct identification and lead to underreporting (9, 15). Numerous studies have therefore suggested the necessity for dependable diagnostic techniques such as in vitro cultivation and polymerase chain reaction (PCR) (15, 20). Scicluna et al. (22) have developed a PCR method based on the analysis of the partial SSU rRNA gene. This method is routinely used in most diagnostic laboratories in developed nations for diagnosing Blastocystis infection. Among the many advantages of this approach are its abilities to determine the parasite’s subtype and zoonotic potential (18, 24). However, because of financial constraints and the need for highly trained laboratory personnel if molecular detection is to be attempted, microscopy is often the only tool available for the screening of Blastocystis spp.

The aims of the present study were to determine by PCR the occurrence, subtype distribution and zoonotic potential of Blastocystis spp., particularly in farm animals of the Artiodactyla order, and to compare and evaluate the sensitivity and specificity of different diagnostic techniques in the detection of Blastocystis spp. in faecal samples compared with the PCR assay as the gold standard.

Sample collection and storage. Ninety-seven fresh faecal samples were obtained from sheep (69), cows (12) and camels (16) in Al-Ain, United Arab Emirates (UAE) and kept in leak-proof containers without preservatives. Samples were transported immediately to the University of Sharjah and submitted for DNA extraction and PCR. Of the 97 samples, 65 were separated into two containers (one aliquot stored at −20°C and the other used for microscopy and in vitro culture analysis). All of these 65 samples were screened using wet-mount, trichrome stain, acid-fast stain and in vitro culture techniques as described below. The remaining 32 samples were not screened using these four techniques as they were discarded prior to the microscopic examination and following the extraction of DNA.

Iodine stain. Lugol’s iodine was applied to non-fixed smears and the smears were examined microscopically using 10× and 40× objective lenses.

Trichrome stain. Smeared slides were allowed to air dry, stained with trichrome stain and mounted with distrene, plasticizer, and xylene (Sigma Aldrich; St. Louis, MO, USA) before they were examined microscopically under 100× magnification for the presence of Blastocystis spp. Slides were cross-checked by other members of the group.

Modified acid-fast stain. Smears were fixed with absolute methanol for 30 s, stained with Kinyoun’s carbolfuchsin for 1 min, decolourised using acid alcohol for 2 min and then counterstained using malachite green for 2 min. The stained slides were examined microscopically using 40× and 100× oil immersion objective lenses to confirm internal morphology.

In vitro culture. Fresh faecal samples were subjected to short-term xenic in vitro culture using modified Tanabe-Chiba medium with 10% horse serum at 37°C (14). Following three days of culture, sediments were examined microscopically for the presence of any of the four known morphologies. Samples negative for Blastocystis were re-examined after two more days of culture and marked negative when no parasites were observed on either of these occasions (2). A sample was reported as positive if any of the four morphologies of Blastocystis spp. was observed.

DNA extraction, PCR, and analysis of DNA sequencing. In this stage, first DNA was extracted from all stool samples using the Bioline stool DNA Kit (London, UK) as per the manufacturer’s recommendations and stored at −20°C until it was analysed. The 600-bp barcoding region of the SSU rRNA gene of Blastocystis was amplified using BhRDr and RD5 primers (22, 23). Reaction products were separated in 1.2% agarose gels stained with ethidium bromide and the size confirmed using a 100-bp ladder. The PCR products were gel-purified using a PureLink Quick PCR Purification Kit from Invitrogen (Carlsbad, CA, USA) according to the manufacturer’s instructions, and sent for sequencing in a commercial sequencing facility (McLab; South San Francisco, CA, U.S.A).

Subtyping and phylogenetic analysis of Blastocystis isolates. Products of the PCR were sequenced on both strands, and the resulting sequences assembled online using CAP3 Sequence Assembly Program online tool (available at https://doua prabi.fr/software/cap3/). The CLC Main Workbench 6 package (Qiagen, Redwood City, CA, USA) was also used for sequence end trimming and editing, in addition to contig assembly. Sequences were aligned with sequences covering the “barcoding region” of the SSU rRNA gene retrieved from GenBank to provide representatives of the known subtypes of Blastocystis. For phylogenetic analysis, SSU rRNA gene sequences were analysed together with an outgroup sequence (Proteromonas lacertae, GenBank accession no. U37108), and the evolutionary relationship between taxa was inferred using MEGA X software. To perform DNA sequence alignment, we used the Clustal W algorithm in MEGA X, with default parameters. Phylogenetic analysis was performed using the neighbour-joining method. Selected reference subtype sequences and Blastocystis sequences available from GenBank were included in the construction of the phylogenetic tree. We also used the public Blastocystis database at https://pubmlst.org/ blastocystis/ to check the STs and the alleles.

Ninety-seven stool samples were collected and screened for the presence of Blastocystis by PCR. Fifteen (15.5%) were determined positive by PCR (1 from a camel, 2 from cows and 12 from sheep). Twelve of the fifteen samples were confirmed by sequencing to be Blastocystis and were deposited in GenBank (accession numbers MT895705–MT895716). These were from sheep. The remaining three samples were not sequenced due to financial constraints. Subtype 10 (allele 152) was the only Blastocystis subtype detected. No zoonotic subtypes were detected, no mixed infections involving different STs were identified, and no intra-subtype genetic variation was observed within ST10. The isolates in the present study clustered in two areas within the clade comprising ST10 (Fig. 1). Isolates 4 (MT895715) and 5A (MT895716) clustered in a different sub-clade from the other isolates in the study. Isolates 2, 5B, 7 and 10 (MT895706, MT895708, MT895710 and MT895713) shared 80.72%, 84.74%, 78.36% and 80.53% identity with isolates from cattle in Malaysia (MG831493, MG831498 and MG831501), respectively. Isolate 6 (MT895709) shared 90.29% similarity with an isolate from deer in China (MK357785) and isolate 4 (MT895715) shared 97.03% identity with an isolate from an elk in the UK (MF186696). Interestingly, isolate 5A (MT895716) shared 99.11% similarity with one from a camel in Libya (KC148207). All twelve sequences were found to match Blastocystis spp. ST10 isolate CA (KC148207) when compared against the https://pubmlst.org/organisms/blastocystis-spp database. However, it was surprising that a few isolates (5 in total) did not return any matches when compared with Blastocystis spp. ST10 isolate CA (KC148207) using BLAST. These isolates had no score tags in Table 1. Table 1 shows the BLAST scores of our sequences compared to the previously mentioned KC148207 isolate in this database, wherein that isolate was the only match identified.

Fig. 1

The performance of the direct wet-mount modified acid-fast stain, trichrome stain, and in vitro culture methods in terms of sensitivity and specificity were compared to the PCR assay as the gold standard technique (Table 2). Calculations of sensitivity and specificity were made from the data of 65 samples. The sensitivity and specificity of the direct wet-mount and modified acid-fast stain techniques compared to PCR were 40% and 78.3% and 40% and 83.3%, respectively. However, the sensitivity and specificity of the trichrome stain and in vitro culture compared to PCR were 80% and 80% and 80% and 76.7%, respectively. The sensitivity of iodine stain (direct wet-mount), modified acid fast stain, trichrome stain and in vitro culture varied widely with 40%, 40%, 80% and 80% being seen, respectively, while the specificity of the four techniques was almost the same at 78.3%, 83.3%, 80.0%, and 76.7%, respectively. Table 2 shows the association of the iodine, acid-fast, trichrome, and culture tests with the PCR. Only the trichrome and culture tests were significantly associated with PCR (odds ratio (OR) = 16.00; 95% CI: 1.63–156.5; P = 0.007 and OR = 13.14; 95% CI: 1.35–127.4; P = 0.003, respectively) and it seems that the trichrome method was able to detect more positive cases than the culture method.

As in our previous study, all twelve sequenced sheep samples were confirmed to be the ST10 subtype (3). Some of the sequences obtained in this survey shared high identity with the Blastocystis spp. sequences in GenBank. Only allele 152 was detected in all of these cases. Similar ST10 alleles were detected in neighbouring Iran in dogs and cats (16). On the other hand, allele 43 of this subtype was detected in sheep and cows in a study in the Czech Republic, possibly evidencing intra-subtype variability in ST10 due to different geographical regions and diets (12).

Some Blastocystis subtypes have zoonotic potential; in a recent study Wang et al. (26) detected four subtypes in sheep (ST1, ST5, ST10 and ST14) with ST1, ST5 and ST14 identified in this animal species for the first time in Heilongjiang Province in northeastern China. Blastocystis ST10 was the only subtype found, with all isolates from the present study clustering in two sub-clades. None of the potentially zoonotic subtypes was detected. Thus, non-zoonotic subtypes seem to be more frequent in the UAE.

We have previously indicated the presence of two non-zoonotic subtypes (ST10 and ST14) in two sheep among individuals with mixed subtypes (3). Unlike our earlier study the present investigation did not find any mixed-subtype infections. The isolates in our previous study were cloned, allowing for the detection of mixed subtypes (3). The current report identified Blastocystis isolates using a fragment of the SSU-rDNA gene as a unique marker by PCR sequencing. This technical approach is not apt to identify co-infections with different STs in the same specimen because of the inherent nature of PCR, which preferentially amplifies the predominant subtype(s). This is a limitation of the investigation in respect of the variety of STs possible to detect per sample. A more accurate measure of the detection rate of particular subtypes, and therefore of potentially zoonotic types, in the Artiodactyla order in this country would be provided by screening large numbers of different member species.

The study also compared the efficacy of iodine stain, acid-fast stain, trichrome stain, in vitro culture and PCR for the detection of Blastocystis in 65 samples and indicated that in vitro culture and trichrome stain had comparable sensitivities, returning the same number of positive cases: 4/5 (80%) by culture and 4/5 (80%) by trichrome stain. Both tests detected positive cases that were false negative by iodine or modified acid-fast stain. Since the trichrome stain is rapid, provided a clear image of the organism and is less expensive than in vitro culture, it may be a better option to use in the examination of samples for Blastocystis spp. As it was in samples from other experimental groups, the vacuolar form was the most observed form in the samples from the sheep comprising this experimental group (1, 10, 15). The present study showed that when both microscopic techniques (trichrome and in vitro culture) were used together, the overall sensitivity and specificity of Blastocystis spp. detection was markedly improved (Table 2).

This study confirmed that trichrome staining has the highest sensitivity, is suitable for use in hospitals and clinics where Blastocystis testing is requested and is also more feasible to use in laboratories in developed and less developed countries because it is cheap, rapid, easy to perform and renders the definitive morphological details of the parasite clearly visible. Although comparable in sensitivity to the trichrome stain, the in vitro culture method is cumbersome, tedious and time-consuming, making it more suitable for use in research institutes. The results of this study concurred with what Mohammad et al. (15) reported on the relative utility of these techniques. Moreover, the finding of ST10 for the second time in artiodactyls corroborates previous studies which contended that the members of this order are natural hosts for these subtypes. No potentially zoonotic subtypes were detected in either this or our previous investigation. Further studies are required to ascertain the variety of subtypes present in farm animals and the zoonotic potential of the subtypes identified. Whether Blastocystis ST circulation is also linked to geographical areas remains to be elucidated. Therefore, sampling more hosts for Blastocystis from various geographical regions and analysing alleles within identified subtypes may offer an answer.