Occurrence of free-living amoebae (Acanthamoeba, Balamuthia, Naegleria) in water samples in Peninsular Malaysia

Pathogenic free-living amoebae (PFLA) including Acanthamoeba spp., Balamuthia mandrillaris and Naegleria fowleri are known to produce rare but serious human and animal infections (reviewed in Marciano-Cabral & Cabral 2003; Visvesvara et al. 2007; Siddiqui & Khan 2012). For example, pathogenic Acanthamoeba spp. cause blinding keratitis often connected with improper use of contact lenses as well as a fatal brain infection known as granulomatous amoebic encephalitis (GAE) in patients with weaker immune systems (Marciano-Cabral & Cabral 2003; Visvesvara et al. 2007). Out of the 20 different genotypes (T1–T20), several have been implicated in human and animal infections, albeit T4 genotype is more frequently associated with infections. In contrast, N. fowleri produces primary amoebic meningoencephalitis (PAM), typically affecting healthy children and young adults (Martinez & Visvesvara 1997; Tuppeny 2011). Comparable to pathogenic Acanthamoeba spp., Balamuthia mandrillaris is a causative agent of GAE that is known to affect both immunocompromised and healthy people (Matin et al. 2008; Lorenzo-Morales et al. 2013; Siddiqui & Khan 2015). Pathogenic amoebae enter the body via skin lesions and/or the nasal cavity and disseminate via haematogenous spread or travel along the neuroepithelial route to reach the central nervous system (CNS) to produce infection (Rodríguez-Zaragoza 1994; Martinez & Visvesvara 1997; Siddiqui & Khan 2012).

Infections due to free-living amoebae (FLA) are of particular concern to tropical countries such as Malaysia with a large Muslim population who practice ablutions and where recreational water activities are routine. Surprisingly, there has not been a single report of PAM or GAE infection due to pathogenic FLA in Malaysia although cases of Acanthamoeba spp. and N. fowleri infections have been reported in Thailand, a neighbouring country (Intalapaporn et al. 2004; Siripanth 2005). This could be due to lack of awareness and/or misdiagnosis of cases as bacterial meningitis. PAM and GAE patients usually have a history of coming in contact with warm water via activities such as swimming, bathing, ritual ablutions, the use of neti pots and working with soil (Siddiqui & Khan 2014). Moreover, there have been limited studies on the distribution of FLA in Malaysia (Ithoi et al. 2011; Onichandran et al. 2013; Richard et al. 2016; Majid et al. 2017). The aim of this study was to determine the occurrence of FLA and to compare different methodologies in their rapid detection in tap water, recreational places, water dispensers, filtered water (tap water with filter unit), drain water and water flowing in paddy fields from Peninsular Malaysia.

HeLa cells were cultured as food source for B. mandrillaris and N. fowleri in cell culture media. HeLa cervical cancer cells were obtained from the American Type Culture Collection (ATCC CCL-2) and cultured in RPMI-1640 medium supplemented with 10% foetal bovine serum, 1% L-glutamine, 1% minimal essential media non-essential amino acid (MEM NEAA) and 1% penicillin-streptomycin. The cells were maintained in 5% CO₂ at 37 °C as previously described (Rajendran et al. 2017).

A clinical isolate of Acanthamoeba castellanii (ATCC 50492) belonging to the T4 genotype, which had been isolated from a keratitis patient, was cultured without shaking in 10 mL of PYG medium [proteose peptone 0.75% (w/v), yeast extract 0.75% (w/v), glucose 1.5% (w/v)] in a T-75 tissue culture flask and incubated at 30 °C (Siddiqui et al. 2017). To obtain vegetative trophozoites, the media was refreshed 17–20 h prior to experimentation, which resulted in more than 95% amoebae in the trophozoite form. To obtain cysts, trophozoites were inoculated onto non-nutrient agar plates at 30 °C for 2 weeks. Following this incubation, the cysts were removed from the agar surface using 10 mL of distilled water and a cell scraper followed by centrifugation at 2,000 × g for 10 min. The pellet was resuspended in phosphate buffered saline (PBS) and cysts enumerated using a haemocytometer (Siddiqui et al. 2016).

Balamuthia mandrillaris (ATCC 50209) and Naegleria fowleri (ATCC 30894) were grown using HeLa monolayers as feeder cells in 10 mL of RPMI-1640 medium in a T-75 tissue culture flask at 37 °C incubator containing 5% CO₂. The amoebae consumed HeLa cells within 48 h, which resulted in more than 95% of amoebae in trophozoite form. Amoebae were centrifuged at 2,000 × g for 5 min and used for experiments (Siddiqui et al. 2007; Kulsoom et al. 2014; Rajendran et al. 2017). To obtain cysts, trophozoites were treated as for A. castellanii.

Water samples were collected (n = 250) randomly from religious, recreational, educational and health-care centres, filtered water, drain water and water flowing in paddy fields from all 11 states of Malaysia: Selangor, Perak, Penang, Pahang, Negeri Sembilan, Melaka, Johor, Kedah, Kelantan, Perlis and Terengganu from October 2016 to November 2017 (Figure 1). One litre water samples were collected in autoclaved Duran bottles and stored at 4 °C until subsequent analysis.

From each water sample, 500 mL was filtered through a sterilised 0.2 μm pore size cellulose filter under vacuum. The filters were cut into quarters and placed inverted onto 1.5% non-nutrient agar plates containing a lawn of heat-inactivated Escherichia coli K-12 laboratory strain HB101 as previously described (Yousuf et al. 2013). Non-nutrient agar was dissolved in Page's amoeba saline (PAS), (4.80 g NaCl, 5.44 g KH₂PO₄, 14.13 g Na₂HPO₄.12H₂O, 0.16 g MgSO₄.7H₂O and 0.133 g of 90% CaCl₂) with pH adjusted to 6.9 (Brindley et al. 2009). The plates were incubated at 30 °C and examined on a daily basis for the presence of amoebae for up to 3 weeks. For the positive control, 10,000 amoebae in 100 μl of PBS were added to 1 litre distilled water and processed as described previously (Yousuf et al. 2013). The remaining 500 mL of water was processed via the centrifugation method. Briefly, water samples were centrifuged at 2,500 × g for 10 min and the pellet was suspended in 100 μl of InstaGene matrix solution (BioRad) followed by DNA extraction. The identity of amoebae from both the plating and centrifugation method was confirmed using polymerase chain reaction (PCR).

PCR was carried out to confirm the identity of amoebae using gene-specific primers (Table 1). DNA was extracted using InstaGene matrix according to the manufacturer's protocol and as described previously (Khan et al. 2001). The supernatant containing the DNA was used as a template for PCR and then analysed for the presence of Acanthamoeba spp., B. mandrillaris and Naegleria spp. PCR was performed in a total volume of 25 μl containing 1 unit of Prime DNA Taq polymerase (GenetBio), 0.1–1.0 ng DNA, 2.0 mM dNTPs mixture, nuclease-free water, 4 mM MgCl₂ and 1 μM of forward and reverse primers. For Acanthamoeba spp. and B. mandrillaris PCR reaction involved initial denaturation at 94 °C for 3 min followed by 40 cycles at 94 °C for 30 s, 50 °C for 30 s, 72 °C for 30 s with a final elongation step at 72 °C for 10 min with the expected PCR-amplified product for Acanthamoeba of 950 bp and B. mandrillaris of 1,075 bp from 18S rRNA gene. PCR reaction for Naegleria spp. was carried out at initial denaturation at 94 °C for 3 min followed by 40 cycles at 94 °C for 30 s, 56 °C for 30 s, 72 °C for 30 s with a final elongation step at 72 °C for 10 min. The expected PCR amplified product for Naegleria spp. was 177 bp from ITS1 region of rRNA gene of Naegleria spp. Amplified DNA for all three protists was electrophoresed on a 1.5% agarose gel, stained and visualised under ultraviolet illumination (Yousuf et al. 2013).

To determine the specificity of the primers, each primer pair was tested against various DNA types including Acanthamoeba spp., B. mandrillaris and Naegleria spp. HeLa cells, E. coli K1 and methicillin-resistant Staphylococcus aureus (MRSA) were used as heterologous DNA (Le Calvez et al. 2012). To determine the sensitivity of the PCR assay, various numbers of Acanthamoeba spp., B. mandrillaris, Naegleria spp. ranging from 0 to 25,000 cells were centrifuged at 3,000 × g at 4 °C for 10 min. Next, DNA extraction was performed followed by PCR as described above (Lehmann et al. 1998).

To determine the specificity of the PCR-based molecular assay, the primers were tested against various DNA samples (i.e. parasite, human cell and bacterial DNA). Furthermore, sensitivity was determined using different numbers of cells (0 to 1,000 cells for Acanthamoeba spp. and B. mandrillaris and 0 to 25,000 cells for Naegleria spp. respectively). From the specificity assay, the results showed that Acanthamoeba spp., B. mandrillaris and Naegleria spp. primers were specific towards the respective DNA with a PCR amplification products of 950 bp, 1,075 bp and 177 bp respectively. Although there were some non-specific bands seen, the expected product was only observed with the relevant DNA (Figure 2). When tested for sensitivity, the results showed that the assay could detect fewer than 10 cells for all the amoebae tested (Figure 3).

The water samples collected from various regions in West Malaysia were tested for the presence of FLA. One litre water samples were collected randomly in Peninsular Malaysia. From each 1 litre sample, 500 mL was filtered and plates containing amoebae were observed on a daily basis for up to 3 weeks under an inverted light microscope to identify the presence of FLA. Of the 250 water samples collected, 142 (56.8%) water samples showed the presence of FLA. PCR carried out following the cultivation method using plating assays showed the presence of Acanthamoeba spp. in 91 (64%) water samples and Naegleria spp. in 99 (70%) water samples. Of note, all samples tested via the plating assays were negative for B. mandrillaris (Figure 4). The remaining 500 mL of water samples were centrifuged at 2,500 × g for 10 min. In contrast to the plating assay, only one of 180 water samples tested via the centrifugation showed the presence of Acanthamoeba spp. and 52 (29%) water samples showed the presence of Naegleria spp. Again, all samples tested for B. mandrillaris were reported negative (Figure 4). Table 2 shows the comparison of the detection of FLA using the cultivation method via plating and centrifugation method as well as the different sources of water. Overall, these finding show that Naegleria spp. were found in 29.8% (54/181), 66.7% (38/57), 33.3% (1/3), 75.0% (3/4), 100.0% (1) and 50.0% (2/4) of water samples from tap water, recreational places, water dispenser units, filtered water, drain and paddy fields respectively, while Acanthamoeba spp. was present in 24.9% (45/181), 70.2% (40/57), 66.7% (2/3) and 100.0% (4/4) of water samples from tap water, recreational places, water dispenser and paddy fields respectively (Table 2). Table 3 shows the wide distribution of FLA from all the 11 states. A detailed description of each water sample, source and method of isolation is given in the supplementary Table S1 (available with the online version of this paper).

Having a clean and safe water supply system is of major concern for public health throughout the world and it is estimated that 3.4% of all deaths are attributed to waterborne diseases (Berman 2009). Given the opportunistic nature of FLA, they represent a threat to human and animal health especially for countries in the Tropics where water-related activities are part of routine life. Malaysia is situated at the equatorial zone with a tropical climate throughout the year that promotes the growth of waterborne parasites. In addition to parasitic protists, free-living opportunistic amoebae (Acanthamoeba spp., B. mandrillaris, N. fowleri) are of particular concern as they can produce deadly infections with a fatality rate of more than 90% (Visvesvara et al. 2007; Trabelsi et al. 2012). Here, we demonstrated the presence of FLA and compared different methodologies in the detection of amoebae from public water supplies. Our results clearly demonstrated that cultivation via the plating assay is the preferred laboratory method for isolation and detection of FLA as compared with the centrifugation method, albeit it is time consuming and requires regular examination of plates for the presence of amoebae that can take up to several days. Although both methods have advantages and limitations, it is recommended to collect sufficient water samples to carry out both the plating assays and centrifugation for efficient and rapid detection of amoebae from environmental and clinical samples. Following amoebae isolation, PCR assays should be carried out to confirm the identity of amoebae.

In this study, of 250 water samples, 142 were positive for the presence of FLA, of which 91 (64%) showed the presence of genus Acanthamoeba spp. and 99 (70%) showed the presence of Naegleria spp. It was found that 54 out of 181 tap water samples contained Naegleria spp. and 45 out of 181 contained Acanthamoeba spp., suggesting the predominance of Naegleria spp. in tap water samples (religious centres, educational centres, public toilets, residential areas, restaurants, hotels and healthcare centres) in Malaysia. A high prevalence of Naegleria spp. and Acanthamoeba spp. in water samples can be attributed to the fact that some FLA (e.g. Naegleria spp.) can grow better at higher temperatures than others. Although cysts of both amoebae are resistant to harsh conditions, it has been shown that cysts of Acanthamoeba spp. can survive for long periods (up to 20 years) and are more resistant to environmental stressors than Naegleria spp. cysts (Visvesvara et al. 2007). In contrast to tap water, recreational places showed a higher presence of Acanthamoeba spp. and Naegleria spp., likely due to the fact that tap water is often treated with chlorine and/or filtered. Moreover, lake water samples were almost always positive for amoebae, unlike swimming pools, likely to due to chlorination of swimming pools. The greater prevalence of Naegleria spp. in tap water, despite filtration, chlorination and water treatment procedures was suggested to be due to their ability to colonise biofilms (Marciano-Cabral & Cabral 2003; Marciano-Cabral et al. 2010). Of the 57 recreational water samples (swimming pools, lakes, waterfalls, pond, well, stream and river) collected, 40 (71%) contained Acanthamoeba spp. and 38 (67%) contained Naegleria spp. Our observations are also in agreement with studies from Cairo, Egypt where 59 out of 120 (49%) swimming pools revealed the presence of Acanthamoeba spp.; in Poland, more than half of swimming pools were recorded as positive for amoebae (Górnik & Kuźna-Grygiel 2004; Ahmad et al. 2014); in Jamaica, West Indies, Acanthamoeba spp. was found in 42 out of 83 (51%) recreational water samples (Todd et al. 2015); and in Iran, several studies indicated that 30–60% of water samples contained Acanthamoeba spp. (Maghsoud et al. 2005; Rezaeian et al. 2008; Mahmoudi et al. 2015; Shokri et al. 2016). In Malaysia, of the 13 water samples collected from lakes, all were positive for Acanthamoeba spp. (Onichandran et al. 2013); and from two drinking water treatment plants in Sarawak, 90% (55/61) of samples showed the presence of Acanthamoeba spp. and Naegleria spp. (Richard et al. 2016). For Naegleria spp., water samples isolated from the Nile river, Egypt, showed the presence of Naegleria spp. in 3 out of 12 samples (25%) during the summer season (Al-Herrawy & Gad 2015), while Baquero et al. (2014) reported that 5 out of 13 samples (39%) were positive for the presence of Naegleria spp. from well water in Guinea-Bissau, Africa. Ithoi et al. (2011) isolated 41 Naegleria spp. from swimming pools, lakes, streams, mosques and air-conditioners, of which aquatic water samples had 80% and dust samples recorded 20% positive. Recently, in Myanmar (90%), Laos (82%) and Singapore (53%), untreated water samples were positive for Naegleria-like flagella and 9.7%, 4.5% and 6.7% from the respective countries showed the presence of Acanthamoeba spp. from untreated water samples (Majid et al. 2017). These findings suggest that our data on the presence of protist DNA of the genera Acanthamoeba and Naegleria in surface and tap water in Malaysia may be used to advise on possible population exposure to a potential serious risk for health, especially as Acanthamoeba pathogenic genotypes and N. fowleri spp. are present. This is of concern as ablutions with contaminated water have been shown to be a risk factor in contracting PAM (Siddiqui & Khan 2014). However, there is not a single case of GAE or PAM reported in Malaysia, despite the high prevalence of amoebae in water supplies and the fact that the majority of the population are Muslims who practice ablutions. This could be due to lack of awareness and/or misdiagnosis as bacterial meningitis. Notably, B. mandrillaris was not detected in water samples tested in this study. B. mandrillaris is generally isolated from organically rich soils (rarely from water samples) and its infection is often associated with exposure to soil (Schuster et al. 2003; Visvesvara et al. 2007; Lares-Jiménez et al. 2015; Latifi et al. 2016; Niyyati et al. 2016), which may explain our findings.

In summary, these results revealed the wide distribution of FLA from water samples tested in this study from Peninsular Malaysia. Acanthamoeba spp. and Naegleria spp. were more prevalent, while B. mandrillaris were not detected. Having demonstrated the presence of N. fowleri that are highly pathogenic for humans, and Acanthamoeba pathogenic strains in recreational and tap waters, it is necessary to perform another study to characterise DNA from the amoebae isolates in water specimens. Furthermore, future studies are needed for the routine monitoring of FLA and their seasonal distribution in water supplies to understand the prevalence of FLA in environmental settings in Malaysia. Water supplies should be monitored on a regular basis and also disinfected with suitable disinfectants, especially as a high percentage of the population uses this water for ablutions.

Authors declares that they have no conflict of interest.

Authors are grateful to Dr Reuben Clements for providing some of the water samples used in this study. This work is supported by University Research Award INT-DBS-2017-04 by Sunway University, Malaysia.