Molecular Identification of Entamoeba spp. in Captive Nonhuman Primates

Abstract

This study describes the molecular identification of 520 Entamoeba-positive fecal samples from a large and diverse population of captive nonhuman primates (NHP). The results revealed the presence of Entamoeba histolytica (NHP variant only), E. dispar, E. moshkovskii, E. hartmanni, E. coli, and E. polecki-like organisms.

Various Entamoeba species are frequently found in the stools of both captive (15, 22) and wild (5, 7) nonhuman primates (NHP). Although the majority of these Entamoeba spp. are considered to be harmless, care should be taken when E. histolytica, the causative agents of amoebiasis, is involved. Infection with this gastrointestinal parasite in NHP may cause hemorrhagic dysentery (6, 28) and extraintestinal pathologies (e.g., liver abscesses) and death (12, 16). Moreover, amoebiasis is of major concern in public health, resulting in similar pathologies in humans and causing up to 100,000 deaths worldwide each year (19). Currently, little is known about the occurrence of E. histolytica in NHP, and the role of these animals as a potential reservoir for zoonotic transmission remains unclear. Most of the previous studies were based on the detection of cysts or trophozoites in stools by using light microscopy. However, differentiation between E. histolytica and other Entamoeba spp. (such as E. coli, E. hartmanni, and E. polecki-like organisms) based on morphological features is difficult (8, 27) and when E. dispar or E. moshkovskii is involved, it can even be impossible (4). For this purpose, molecular methods are more appropriate. Furthermore, recent molecular analyses of E. histolytica indicate genetic differences between human and NHP isolates (21, 23, 24). Although these differences may contribute to the elucidation of zoonotic transmission pathways, little is known about the distribution of these E. histolytica variants in both humans and NHP. Therefore, the objective of the present study was to identify the Entamoeba spp. in a large and diverse population of captive NHP, including differentiation between the human and NHP variants of E. histolytica.

A total of 520 stool samples containing Entamoeba cysts were selected for further molecular identification. These samples were obtained from previous epidemiological surveys (10; unpublished data) and were stored at −20°C. The animals were housed in nine zoological gardens and one sanctuary in Belgium and the Netherlands, representing 58 NHP groups belonging to 36 animal species (see Table S1 in the supplemental material). None of the animals showed clinical signs associated with gastrointestinal disorders. DNA was extracted using the QIAamp stool minikit according the instructions of the manufacturer (Qiagen) and the adaptations described previously (9). The identification of E. histolytica, E. dispar, E. moshkovskii, E. hartmanii, E. coli, and E. polecki-like organisms was based largely on a previously described PCR-reverse line hybridization blot (PCR-RLHB) protocol targeting the small-subunit rRNA gene (26). This assay was preferred since it allows the simultaneous detection of various Entamoeba species. The amplification reactions were performed in a volume of 25 μl containing 2.5 μl DNA, 0.5 μl of each primer (10 μM), 1 μl MgCl₂ (25 mM), 5 μl GoTaq Flexi buffer, 14.875 μl PCR-grade H₂O, and 0.125 μl GoTaq Flexi DNA polymerase. For E. histolytica, a novel probe (5′-YAT TRA ATR AAT TGG CCA TTT TGT A-3′) was designed based on the gene sequences of the human variant (GenBank accession number X64142) and the NHP variant (GenBank accession numbers AB197936 and AB282657) to ensure the detection of both variants. In each PCR-RLHB run, control DNA samples from E. histolytica (both variants), E. dispar, E. moshkovskii, E. hartmanni, E. coli, and E. polecki-like organisms were included. Samples showing hybridization with the E. histolytica probe were retained for additional differentiation between the human and the NHP variants by using novel variant-specific reverse primers (human variant primer, 5′-CAT TTC TAG AAA CTT TAC TTA CAT-3′; NHP variant primer, 5′-CAT TTC TAG AAA CTT TAC TTA TGC-3′) designed from sequences with the GenBank accession numbers mentioned above. The amplification conditions remained unchanged. In each PCR run, control DNA samples from both the human variant and the NHP variant of E. histolytica were included. PCR products were run on agarose gels, stained with ethidium bromide, and detected upon UV transillumination. Samples reacting only with the general Entamoeba probe and not with any of the species-specific probes were retained for further sequence analyses. To this end, the PCR preceeding the RLHB assay was repeated with unlabeled primers. The obtained PCR products were purified with QIAquick purification columns (Qiagen, Germany) and cloned into the pGEM-T Easy vector according to the instructions of the manufacturer (Promega, Madison, WI). Clones containing the expected amplicon of approximately 550 bp were sequenced using the BigDye Terminator kit (Applied Biosystems). Sequence reactions were analyzed with an ABI-3730xl sequencer (Applied Biosystems), and sequences were assembled using Seqman II (DNAstar, Madison, WI).

The RLHB analysis revealed the presence of Entamoeba DNA in 372 (71.5%) of 520 samples. The distribution of the different Entamoeba spp. within these 372 samples is described in Table 1. E. hartmanni (present in 51.9% of samples) was the most prevalent species, followed by E. polecki-like organisms (in 42.7% of samples), E. histolytica (in 36.0% of samples), and E. coli (in 21.5% of samples). E. dispar (present in 2.4% of samples) and E. moskovskii (present in 1.9% of samples) were found in only a small number of samples. Most samples (51.9%) carried mixed infections. A large proportion of the samples (18.8%) hybridized with the general Entamoeba probe but could not be assigned to any of the known Entamoeba spp. The E. histolytica variant-specific PCR revealed solely the NHP variant in 124 of the 132 E. histolytica-positive samples. For the remaining 8 samples, no amplification was found in either PCR protocol.

TABLE 1.

Numbers of mono- and mixed infections with Entamoeba spp. in 372 samples based on a PCR-RLHB protocol targeting the small-subunit rRNA gene

Infecting organism(s)	No. (%) of samples
Organisms in monoinfections	109 (29.3)
E. histolytica	30 (8.1)
E. hartmanni	36 (9.7)
E. coli	12 (3.2)
E. polecki-like organisms	31 (8.3)
Organisms in mixed infections	193 (51.9)
E. histolytica + E. hartmanni	27 (7.3)
E. histolytica + E. coli	1 (0.3)
E. histolytica + E. polecki-like organisms	30 (8.1)
E. hartmanni + E. coli	28 (7.5)
E. hartmanni + E. polecki-like organisms	28 (7.5)
E. coli + E. polecki-like organisms	2 (0.5)
E. histolytica + E. coli + E. polecki-like organisms	3 (0.8)
E. histolytica + E. hartmanni + E. coli	3 (0.8)
E. histolytica + E. hartmanni + E. polecki-like organisms	36 (9.7)
E. dispar + E. hartmanni + E. coli	4 (1.1)
E. dispar + E. hartmanni + E. polecki-like organisms	3 (0.8)
E. moshkovskii + E. hartmanni + E. coli	2 (0.5)
E. moskovskii + E. hartmanni + E. polecki-like organisms	1 (0.3)
E. hartmanni + E. coli + E. polecki-like organisms	15 (4.0)
E. histolytica + E. hartmanni + E. coli + E. polecki-like organisms	4 (1.1)
E. dispar + E. hartmanni + E. coli + E. polecki-like organisms	2 (0.5)
E. moshkovskii + E. hartmanni + E. coli + E. polecki-like organisms	4 (1.1)
Unidentified Entamoeba organisms	70 (18.9)

From the 70 samples which could not be assigned to known Entamoeba spp., 20 samples originating from 20 different NHP groups were withheld for sequencing, resulting in 21 clones (for one sample, two clones were analyzed). Twelve clones could be assigned to one of the known Entamoeba spp., including the E. histolytica NHP variant (5), E. dispar (2), E. hartmanni (3), or E. coli (2). Four clones did not reveal homology with Entamoeba spp. Instead, homology was found to DNA sequences from Gregarina (GenBank accession no. FJ459742) and Saccharomyces (GenBank accession no. FN393078) species, humans (GenBank accession no. CT476837), and Galactomyces species (GenBank accession no. X69842). The remaining four clones showed homology to Entamoeba spp., but the sequences did not match completely with those from one of the known Entamoeba spp. (E. terrapinae [GenBank accession no. AF149910], E. insolita [GenBank accession no. AF149909], E. invadens [GenBank accession no. AF149905], E. ranarum [GenBank accession no. AF149908], E. equi [GenBank accession no. DQ286371], E. ecuadoriensis [GenBank accession no. DQ286373], and E. struthionis [GenBank accession no. AJ566411]). All four sequences showed the least homology (74.2% to 75.2%) to sequences from E. invadens. The highest level of homology found (87.4% to 88.7%) was to sequences from E. moshkovskii. The sequences of these four clones were submitted to GenBank under accession no. GU437823 (for clone JL70 from a Javan lutung), GU437824 (for clone JL2399 from a Javan lutung), GU437825 (for clone MG107 from a mantled guereza), and GU437826 (for clone NPGL93 from a Northern plains gray langur).

To our knowledge, this is the first study that describes the molecular identification of Entamoeba isolates from a large and diverse population of captive NHP based on a PCR-RLHB protocol and an E. histolytica variant-specific PCR approach. The results confirm the presence of the Entamoeba spp. described previously but also suggest NHP (the owl-faced monkey, Javan lutung, and Northern plains gray langur) as novel host species for the free-living E. moshkovskii. Until now, E. moshkovskii had been detected only in samples from sewage and humans (1, 3). E. hartmanni and E. polecki-like organisms were the most prevalent; approximately half of the samples contained one of these Entamoeba species. E. polecki-like organisms are considered to be harmless for NHP, but their presence warrants caution. The E. polecki-like organisms may cause postmortem pathologies mimicking those caused by E. histolytica (29), consequently hindering an accurate diagnosis.

E. histolytica DNA was detected in a large proportion (36.0%) of the samples, originating from 11 of 58 NHP species examined. Five of these species (the vervet monkey, crab-eating macaque, rhesus monkey, Hamadryas baboon, and chimpanzee) (12, 25) have been reported previously to be hosts of E. histolytica. New host species identified in the present study are the tantalus monkey, greater spot-nosed monkey, Sunda pig-tailed macaque, olive baboon, and Bornean orangutan. Although previously found in other studies, E. histolytica was not found in the patas monkey (2, 12), mandrill (13, 28), mantled guereza (11, 20), and Western gorilla (18). The occurrence of E. histolytica is surprisingly high and was limited mainly to Old World monkeys. At the time of sampling, clinical symptoms were absent in all these animals. However, due to the study design, the clinical importance of these infections could not be investigated in more depth. Among the isolates of E. histolytica, only the NHP variant was identified. This finding supports the hypothesis that this variant is restricted to NHP (20, 23, 24) but is in contrast with the results of a recent study in the Philippines, where the human variant was found in NHP (17). Although this would imply anthropozoonotic transmission, the distribution of both variants among humans and NHP remains largely unknown, underlining the importance of using variant-specific PCR approaches in future epidemiological surveys.

Sequence analyses revealed the presence of four novel Entamoeba-like sequences, which could not be assigned to any of the previously described Entamoeba spp. Studies analyzing additional genes, morphological features (trophozoites and cysts), and virulence are needed to determine their phylogenetic position within the genus Entamoeba.

Finally, this study underlines the need for improving detection techniques, as the presence of Entamoeba based on microscopic examination was not confirmed by PCR for a considerable proportion of the samples (38.5%). Although an initial microscopic misclassification cannot be ruled out, an interlaboratory comparison of a subset of the samples examined indicated moderate agreement in the microscopic findings (data not shown). Therefore, the most important factor contributing to this discrepancy for both techniques is probably the inhibition of the PCR by fecal components, a well-known problem in the detection of pathogens in feces (14). Moreover, the previously described primers for the detection of Entamoeba spp. are not as specific as previously assumed, which can be explained by small sequence differences in the primer regions.

In conclusion, this study confirms the presence of previously described Entamoeba spp. in NHP and is the first report of E. moshkovskii in these animals. The results also indicate that E. histolytica (exclusively the NHP variant) is prevalent in captive NHP, supporting the host specificity of this NHP variant of E. histolytica. However, studies using variant-specific PCRs are needed to elucidate the epidemiology of both variants of E. histolytica. Finally, this study reveals the presence of novel Entamoeba-like sequences, which warrants further attention.