Prevalence of Entamoeba species in captive primates in zoological gardens in the UK

Study areas and sampling design

A preliminary study was performed to establish the prevalence of Entamoeba infection in primates in a single zoological park in the United Kingdom, and to identify which families of primates required the most focus during the subsequent nationwide study. Hence, Twycross Zoo was chosen as the preliminary study site as it houses a wide variety of primate species and families, and amoebic infection had been identified historically and was suspected in their primates at the time of the study. Thirty-seven primates were available for inclusion within the study, including six species of primates from 23 enclosures (Table 1). To identify individual primates in group enclosures a feed item for each primate was impregnated with approximately 0.5 g of different coloured edible cake glitter (Rainbow Dust Colours Limited, Lockstock Hall Preston, England) and fed during the morning by the keeper. Each group individual was assigned and fed a different glitter colour from two days before sample collection until successful completion of sample collection; this was typically two to five days. All stool samples in each enclosure were collected separately on clean disposable paper plates during morning cleaning by keepers until three stools per primate were identified from the glitter colour allocated for each primate. Three samples per primate were collected to account for intermittently shed Entamoeba. A representative 3–5 g stool sample for each animal (identified by different glitter colours) was placed into a labeled clean 7 ml plastic bijoux, using a clean wooden swab stick, containing 10% formalin, and was stored at 5 ° C until further processing. Age, sex, species, treatment with amoebicidal medication and enclosure identity were recorded.

Subsequent to the preliminary study, a nationwide epidemiological study was conducted to identify the prevalence of Entamoeba infection, E. histolytica, E. dispar and uninucleate cyst-producing species, within Colobinae primates at six different zoos in the United Kingdom (Fig. 1). Primates from the Colobinae family (genus Semnopithecus, Trachypithecus or Presbytis) were selected as the sample population for the nationwide study based on two reasons. Firstly, the preliminary data from Twycross Zoo demonstrated the highest prevalence of amoebiasis in the Old World Colobinae monkeys. This is in agreement with results from other studies (Tachibana et al., 2001; Levecke et al., 2007). Secondly, primates from the Colobinae family have specialised sacculated stomachs, an adaption to their leaf-eating lifestyle, which provides favorable conditions for ingested Entamoeba cyst excystation, and trophozoite tissue invasion (Mätz-Rensing et al., 2004; Ulrich et al., 2010).

A total 350 samples were collected from 162 primates from six zoological parks within the United Kingdom (Table 2), between July 2010 and August 2011. This sample group included primarily primates from the Colobinae family, but also some New World monkeys. All zoological parks housed primates and non-primate species. Primates occupied indoor concrete enclosures with access to external grassed sections. Same species primates occupied mixed sex group enclosures. Some primates were housed alone for medical reasons, or due to social incompatibility with the rest of the group. It was possible to collect repeat samples from four primates sampled in the preliminary study; all other primates from Twycross Zoo were unavailable for sampling. The same stool collection technique was used as for the preliminary study with one modification to facilitate molecular examination of samples: stools were collected into 70% ethanol, not 10% formalin. The primate keepers at each facility administered the glitter and collected the samples. Two hundred and seventy-four stool samples could be associated with 89 individually identified primates; however, some stool samples collected could not be attributed to a specific primate from within a group enclosure. This was due to the limitations of deciphering different glitter colours when dealing with large number of primates, and hence glitter colours, in one enclosure. Hence, 76 stools from the remaining 73 primates had to be collated as samples from eleven groups of NHPs (Table 2). The entirety of each stool sample was examined grossly for the presence of blood as a possible indication of gastrointestinal illness and potential parasitism. Thirty two (82.1%) of primates had been treated with a vitamin D3 supplement and 10 day course of metronidazole (Flagyl) followed by 10 days of diloxinide furoate in the six months prior to sample collection.

The study was approved by The University of Nottingham (UK) School of Veterinary Medicine and Science (SVMS) Ethical Review Committee. The Committee reviews all research studies involving School personnel and is chaired by Professor David Haig. The committee passed this study as good to proceed, not requiring any further ethical review.

Parasite identification

All formalin preserved samples were analysed microscopically using a modified Ridley’s formol-ether concentration technique, which enhances microscopic sensitivity by producing ‘cleaner’ samples that are more efficient to examine. Following sedimentation, samples were then examined microscopically for the presence of Entamoeba species from the E. histolytica complex (E. histolytica, E. dispar and E. moshkovskii), E. coli and E. hartmanni. Data was analyzed using Minitab 15. Binary logistic regression was used to demonstrate statistical significance between prevalence of infection and primate demographics. All prevalence data is derived using the total number of primates as the denominator.

Molecular analyses

QIAamp DNA Stool Mini Kit (QIAgen, UK) was used according to the manufacturer’s instructions to extract parasite DNA directly from faeces. Technique modifications to improve the yield and purity of DNA extracts included increasing the lysis temperature to 95 ° C and adding an extra wash prior to sample elution with Buffer AE. Concentration and DNA purity in sample extracts was analyzed, using a Thermo Scientific NanoDrop™1000 Spectrophotometer, prior to PCR amplification. The strategy used for selection of PCR primers (Table 3) was based on the use of previously published diagnostic primers for the mononucleate Entamoeba species, E. histolytica and E. dispar (Ali, Zaki & Clark, 2005). Both species have been previously found in the faeces of NHPs. It is important to discriminate E. histolytica from other nonpathogenic amoebas because E. histolytica carries the risk of zoonosis (Rivera & Kanbara, 1999; Tachibana et al., 2000; Tachibana et al., 2001; Verweij et al., 2003; Rivera, Yason & Adao, 2010; Feng et al., 2011). Also, we used two species complex specific primers to amplify uninucleate cyst-producing species, but not tetra- or octonucleate cyst-producing species. All PCR products were subjected to DNA sequencing to identify the species/subtype of each amplicon including those amplified by the species diagnostic primers.

Separate PCRs were performed with each primer pair in a reaction mixture of 40 µ l consisting of 4 µ l of extracted DNA, 20 µ l of Biomix (Bioline, UK), 15 µ l of sterile distilled water, and 0.5 µ l of each forward and reverse primer. The amplification reactions were performed using a Bioer Xp Cycler as described in Table 4. PCR products were separated by electrophoresis in 1.2% agarose gels run at 100 V on a Thermo Scientific Easycast B1 or D2 electrophoresis gel tank with a Thermo Scientific EC 1000 XL Power Pac for approximately 60 min. A mix of 7 µ l of PCR product and 3.5 µl of loading buffer (New England Biolabs Ltd., UK) were applied to each well. A 1-kbp molecular size ladder (New England Biolabs) was added to each gel for product size estimation. Gels were stained with 0.1 µ g/ml ethidium bromide solution. Amplified DNA was visualized under UV light.

Molecular phylogenetics

One positive amplicon per primate (a total of 16 amplicons) was selected for sequencing, based on visualization of PCR products (Table 4). Amplicons were purified, using a QIAquick PCR Purification Kit (QIAgen, UK), according to the manufacturer’s instructions and then subjected to sequencing on the Illumina platform by Source BioScience (Nottingham, UK) using the primers from the PCR. Nucleotide sequences were determined at least once on each DNA strand. Three representative Entamoeba nucleotide sequences obtained in this study were deposited in GenBank under accession numbers KJ149294, KJ149295, KJ149296.

Raw sequencing chromatograms were evaluated with Geneious (version 5.4) software. Newly obtained Entamoeba sequences were compared with similar sequences available at the GenBank database by using the Bl2Seq algorithm as implemented in BLASTn (Altschul et al., 1990). Multiple alignments of all nucleotide sequences were obtained by using the MUSCLE program (Edgar, 2004). The resulting alignments were adjusted manually when necessary using CLUSTALX (Larkin et al., 2007). The unmatched ends were deleted to obtain a homogeneous matrix of characters and thus increase the reliability of the tree obtained. Phylogenetic trees were inferred from the nucleotide sequence alignments by the maximum-likelihood (ML) method using the BIONJ algorithm (Gascuel, 1997) and distance method with HKY85 model (Hasegawa, Kishino & Yano, 1985) of nucleotide substitution implemented in PhyML-aLRT (Guindon & Gascuel, 2003). The reliability of the branching order was assessed by performing 1,000 bootstrap replicates.

Entamoeba prevalence at Twycross Zoo

One hundred and twenty-six stool samples were collected from 37 individual primates. No primate demonstrated ill health at the time of sample collection and no samples contained grossly visible blood. Microscopic examination demonstrated Entamoeba shedding in 81.1% of 37 primates sampled (Table 1). Entamoeba coli was the most prevalent Entamoeba species shed (62.2%), with three of six primate species shedding this Entamoeba species. Shedding of species from the E. histolytica complex was identified in 16.2% of primates (6 primates). Co-infection with two or more Entamoeba species was identified in 14 primates. Old World Colobinae primates showed the highest prevalence of Entamoeba infection. Entamoeba infection was significantly associated with species of primate (P < 0.05) and administration of metronidazole (P < 0.05). More specifically, infection with E. coli was significantly associated with both parameters (both P < 0.05). Primates previously treated with metronidazole showed greater infection with E. coli (76.9%) compared to those untreated (25.0%). No significant associations were identified between primate demographic characteristics and infection with Entamoeba from the E. histolytica complex or E. hartmanni. Eggs from Trichuris species were identified in samples from two primates.

Entamoeba prevalence at multiple zoos

Entamoeba was present in 101 (28.9%) samples (Table 5), indicating a notable prevalence of Entamoeba infection at the national level. No more than one species of Entamoeba was identified per sample. Three Entamoeba species were detected by species-specific PCR and confirmed with sequencing and BLAST: E. histolytica, E. dispar and E. polecki. E. histolytica was detected in one sample (2.9%), E. dispar in 14 samples (4.0%) and uninucleated cyst-producing Entamoeba species in 86 (24.6%) samples. E. histolytica and E. dispar were identified in samples from Colobinae primates only, whilst uninucleated-cyst producing Entamoeba species were identified in samples primarily from New World monkeys, but also in primates from the Colobinae family. Entamoeba infection was only detected in primates from three zoological parks (Table 5): E. histolytica was only identified at one park, E. dispar in three parks, and uninucleated cyst-producing Entamoeba species in three parks. No primate was found to harbor mixed Entamoeba species. All primates sampled appeared clinically healthy at the time of sample collection.

To infer the phylogenetic relationship of the isolates detected in the present study, E. histolytica, E. dispar and E. polecki, with previously characterized isolates, we used maximum likelihood method. PCR amplicons from sixteen samples were purified and submitted for sequencing (Table 4). The sequence from the sample that produced an amplicon with the E. histolytica-specific primers was identical to the corresponding region of the GenBank sequence for E. histolytica from monkey (AB197936) from a cynomolgus monkey. Likewise, sequences obtained from eight samples that produced amplicons with the E. dispar-specific primers were identical to the corresponding region of the GenBank sequence (AB282661) for E. dispar from a rhesus monkey. Seven sequences were obtained from uninucleate amplicons from Twycross Zoo and shared high sequence homology E. polecki. Two representative sequences were used to build a phylogenetic tree. As seen in Fig. 2, E. polecki sequences obtained in the present study clustered with and formed a monophyletic group with E. polecki subtype 4 isolates reported in Homo sapiens from Asia, Africa and Europe.