Detection of Naegleria Species in Environmental Samples from Peninsular Malaysia

Background In Malaysia, researchers and medical practitioners are unfamiliar with Naegleria infections. Thus little is known about the existence of pathogenic Naegleria fowleri, and the resultant primary amoebic meningoencephalitis (PAM) is seldom included in the differential diagnosis of central nervous system infections. This study was conducted to detect the presence of Naegleria species in various environmental samples. Methods/Findings A total of 41 Naegleria-like isolates were isolated from water and dust samples. All these isolates were subjected to PCR using two primer sets designed from the ITS1-ITS2 regions. The N. fowleri species-specific primer set failed to produce the expected amplicon. The Naegleria genus-specific primers produced amplicons of 408 bp (35), 450 bp (2), 457 bp (2) or 381 bp (2) from all 41 isolates isolated from aquatic (33) and dust (8) samples. Analysis of the sequences from 10 representative isolates revealed that amplicons with fragments 408, 450 and 457 bp showed homology with non-pathogenic Naegleria species, and 381 bp showed homology with Vahlkampfia species. These results concurred with the morphological observation that all 39 isolates which exhibited flagella were Naegleria, while 2 isolates (AC7, JN034055 and AC8, JN034056) that did not exhibit flagella were Vahlkampfia species. Conclusion To date, pathogenic species of N. fowleri have not been isolated from Malaysia. All 39 isolates that produced amplicons (408, 450 and 457 bp) from the genus-specific primers were identified as being similar to nonpathogenic Naegleria. Amplicon 408 bp from 5 representative isolates showed 100% and 99.7% identity to Naegleria philippinensis isolate RJTM (AM167890) and is thus believed to be the most common species in our environment. Amplicons 450 bp and 457 bp were respectively believed to be from 2 new species of Naegleria, since representative isolates showed lower homology and had a longer base pair length when compared to the reference species in the Genbank, Naegleria schusteri (AJ566626) and Naegleria laresi (AJ566630), respectively.


Introduction
Naegleria is a free-living amoeba that is ubiquitously distributed in the environment worldwide. Many species of Naegleria are recognized based on their small subunit ribosomal deoxyribonucleic acid (SSU rDNA), large subunit ribosomal DNA (LSU rDNA) and the internal transcribed spacer (ITS) regions, including the 5.8S rDNA [1]. However, only Naegleria fowleri has been shown to cause human disease, that result in primary amoebic meningoencephalitis (PAM), a rapid fatal infection of the central nervous system (CNS) [2]. PAM was most frequently reported in healthy young persons with a recent history of aquatic activities [3,4,5]. PAM was associated with pipe water and man-made storage ponds used by Muslims for routine ritual ablution, which involves sniffing water into the nostrils [6,7]. N. fowleri infects the CNS through inhalation through the nasal passages of the trophozoite, flagellate and cyst forms that contaminate water or dust. The flagellates and cysts after transformation to trophozoites then migrate to the olfactory nerves and invade the brain causing hemorrhagic necrosis and edema [8].
In Malaysia, Naegleria is unfamiliar to most physicians, pathologists and laboratory workers. Up to 2010, clinicians have seldom considerd N. fowleri in the differential diagnosis of CNS infections. Thus the question arises as to whether PAM is underdiagnosed or that pathogenic N. fowleri may be rare in the environment.
N. fowleri was reported to cause many PAM fatal cases and was easily isolated from aquatic environments of many provinces of our neighbor Thailand [9]. As both Malaysia and Thailand have similar environmental conditions, it is likely that N. fowleri is also present here. Other than aquatic activities, ritual ablutions using tap water by Malaysian Muslims would potentially expose them to infection if pathogenic N. fowleri is present in the environment. Therefore, the present study aimed to detect Naegleria species in the Malaysian environment by microscopy and PCR using esthablished primer sets designed from the ITS1-ITS2 regions.

Collection and cultivation of samples
All sampling sites and environmental samples (dust, debris and water) were selected based on the benefit, convenience and usefulness to the public health. They were swimming pools, recreational lakes, recreational streams, mosques and air-conditioners that located at Selangor and Federal Territory, representing the most developed region of Malaysia (Table 1). There was no permits were required to collect dust, debris and water samples from all of these public sampling sites. In order to collect these environmental samples, the representatives from our research group had shown the letter entitled 'request for permission to collect samples' from the Head of Department of Parasitology, University of Malaya, to the authorities or the security personnels concerned, following with the explanation of the study objectives.
The dust at the moist areas of air-conditioners and debris at the water margins of the wall of swimming pools and water tanks were collected by scrapping with sterile cotton wool (6 samples from each sampled site) and placed onto each of 6 non-nutrient agar (NNA) plates prepared from PAS (PAGE amoebae saline) overlaid with thin layers of live Escherichia coli. Water samples (2000 mL in sterile Schott glass-bottle) were first filtered through a nitrocellulose membrane (1.2 mm pore size, Millipore) and the trapped debris was carefully flushed in situ with 6.0 mL of sterile distilled water followed by spreading 1.0 mL evenly onto each of 6 NNA-E. coli plates as above. The culture plates were sealed with parafilm and each set of 2 plates were incubated at room temperature (2662uC), 37uC and 44uC, respectively, for up to 14 days. Three sets of 2 types of negative controls were carried out, each inoculated with sterile distilled water and sterile cotton wool respectively, onto NNA-E. coli plates and incubated as for the test samples.
Detection of Naegleria-like species All culture plates were examined daily up to 14 days using a light microscope (Olympus BX51) before being discarded. The morphological characterictics of Naegleria-like trophozoites, flagellas and cysts were photographed as shown in Figure 1. Initially, culture plates were colonized with mixed organisms and positive Naeglerialike cells could exist as mixed isolates. Cloning of each to produce a single-species cell lines was carried out. The trophozoites were induced to flagellate by adding 10 mL of sterile distilled water and then incubated at room temperature for 60 minutes. The swimming flagellates were pipetted out and dropped into all wells of a 96-well plate. Trophozoites from isolates that failed to form the flagellate stage were gently pipetted and dropped into all the 96well plate as described above. Wells with a single flagellate (or trophozoite) were selected, and amplified in xenic culture condition. The diameter of the cysts and the length of motile trophozoites were also determined. The cysts of the cloned cell lines were then subjected to axenisation with 3% (v/v) HCl overnight. Following the treatment, cells were seeded in protease-yeast extract-nucleic acid-folic acid-hemin (PYNFH) medium supplemented with 10% of heat inactivated fetal calf serum and 200 mg/mL gentamicin (Gibco) for cultivation. Naegleria isolates that proliferated well in axenic culture conditions were subsequently cultured in PYNFH medium alone.

PCR and DNA sequencing
Two sets of primer pairs, Naegleria fowleri species-specific (59-GTG AAA ACC TTT TTT CCA TTT ACA -39 and 59-AAA TAA AAG ATT GAC CAT TTG AAA -39), and Naegleria genus-specific (59-GAA CCT GCG TAG GGA TCA TTT -39 and 59-TTT CTT TTC CTC CCC TTA TTA -39) designed from ITS regions and used previously for Naegleria detection [10], were used in this study. Total genomic DNA was extracted from trophozoites of Naegleria using the QIAamp DNA mini kit (Qiagen, Hilden, Germany). Amplification reactions were performed according to Pélandakis et al. [10]. PCR amplicons were separated by a 1.5% agarose gel prepared in 0.56Tris-borate EDTA (TBE) buffer containing ethidium bromide. Electrophoresis was run at 100 V for 60 minutes using the 0.56TBE buffer as running buffer. After electrophoresis, gels were viewed under UV illumination in the chamber of a BioDoc-It TM Imaging System (UVP, Cambridge, United Kingdom). Amplicon sizes were estimated by comparison with the GeneRuler 100 bp DNA ladder Plus (Fermentas Life Science, Canada).
The amplicons from 10 selected isolates were gel-purified using the QIAquick gel extraction kit (Qiagen, Hilden, Germany) prior to cloning. The purified amplicons were cloned using the InsT/ Aclone TM PCR product cloning kit (Fermentas). The DNA fragments were inserted into vectors (plasmids, pTZ57R/T) to form a DNA recombinant. The recombinant molecules were transformed into Escherichia coli (strain JM109, Promega)) followed by selection of the white colonies carrying recombinant plasmids with a disrupted b-galactosidase gene.
The plasmid DNA in selected recombinant colonies were confirmed by PCR amplification and were gel-purified using QIAprepH Miniprep kit (Qiagen, Hilden, Germany) prior to sequencing. The purified plasmid DNA from purified amplicons were sequenced directly at both strands using the amplification primers (forward M13/pUC (220), 17-mer and reverse T7 promoter encoded on the pTZ57R/T vector) in an ABI PRISM TM Bigdye TM terminator cycle sequencing ready reaction kit V.3.1 (Korea).
DNA chromatograms were examined using the Chromas software version 1.45 [11]. The forward and reverse sequences were pair-wise aligned using ClustalW software [12] and were manually refined to obtain a better consensus sequence. For nucleotide similarities search analyses, each consensus sequence was blasted against all eukaryotic nucleotide sequences archived in the GenBank database [13]. Nucleotide residues corresponding to primer regions were excluded from the similarities search analyses. The new data on DNA sequencing of all these 10 selected isolates (SP9, L1, AC1, M1, SP1, M2, S1, M4, AC7 and AC8) has been deposited in GenBank with the accession numbers (JN034047, JN034048, JN034049, JN034050, JN034051, JN034052, JN034053, JN034054, JN034055 and JN034056), respectively.

Phylogenetic analyses
Phylogenetic analyses were based on the internal transcribed spacers (ITS) and the 5.8S sequences performed using the Molecular Evolutionary Genetics Analysis (MEGA) software [14]. The phylogenetic trees were subsequently constructed using the neighborjoining (NJ) [15] with molecular distances estimated by the Kimura two-parameter model [16]. The bootstrap procedure was then used to evaluate the robustness of each node [17].

Microscopic observation of Naegleria-like species
The trophozoite stage was normally observed after the second or third days of culture. The cyst stage could be seen after at least 5 days of cultivation. The flagellate stage was seen in the watery surface of the agar or could be induced by adding sterile distilled water to the surface agar. All the three stages are illustrated in Figure 1. Fourty-one (41) isolates from 41 locations were successfully grown in the laboratory. All these isolates showed good growth at 2662uC and 37uC and died at 44uC. Morphologically, the ranges of cyst diameter were between 8-10 mm for all the 41 isolates. Trophozoite lengths were 15-20 mm for 26 isolates, 15-25 mm for 13 isolates, and 20-35 mm for 2 isolates. All of these isolates also showed the flagellate stage after the test for enflagellation except for 2 isolates, AC7 and AC8 (Table 2).

PCR products
After PCR amplification, the N. fowleri species-specific primer set did not show any amplicon bands against all 41 isolates. In contrast, the Naegleria genus-specific primer set produced amplicons against all 41 isolates, These amplicons could be differentiated into 5 groups which were labeled as T1, T2, T3, T4 and T5 represented by 457 bp, 408 and 756 bp, 408 bp, 450 bp, and 381 bp, respectively [ Table 3 and Figure 2 (A, B, C)]. On the other hand, this primer set did not amplify the DNA template of Acanthamoeba isolates ( Figure 2B).

Discussion
Initially, identification of these 41 isolates of Naegleria-like species was based on the morphological characteristics of the trophozoite and cyst stages. The trophozoites of all isolates showed erupted pseudophodia or lobopodia ( Figure 1A) that actively moved in a unidirectional manner while the cysts had uniform smooth thick double walls. After cloning, all isolates showed a similar cyst diameter of 8-10 mm, while the motile trophozoite lengths were divided into three groups ranging from 15-20 mm (26 isolates), 15-25 mm (13 isolates) and 20-35 mm (2 isolates). All of these isolates exhibited flagella in distilled water except two from air conditioner dust samples, AC7 (JN034055) and AC8 (JN034056). The detection of flagellate (transient form trophozoite) stage was one of the important indicators for the rapid screening to differentiate the Naegleria from other free living amoebae especially Acanthamoeba species [18]. However, it was reported that some of the Naegleria species (eg: Naegleria chilensis and Naegleria indonesiensis) were unable to enflagellate [19], as seen with isolates AC7 (JN034055) and AC8 (JN034056) in this study. All the 41 isolates showed good growth at room temperature (2662uC) and 37uC, but died at 44uC. Pathogenic N. fowleri was reported to survive up to 45uC [20] and a temperature range of 30 to 40uC was suggested to be associated with increased occurrence of N. fowleri in thermally elevated environments [21,22]. In this study we used the PCRbased detection in order to confirm the Naegleria species that were present in our environment.
We excluded the use of pathogenic N. fowleri as a control because of restrictions in bringing the pathogenic organisms into the country. We used 2 primer sets, the N. fowleri species-specific and Naegleria genus-specific, designed from the ITS1-ITS2 region by Pélandakis et al. [10], which had previously provided successful results.
The ribosomal ITS sequence was reported to be a powerful tool for detecting inter-and intra-species differences of various organisms, including Cryptosporidium parvum [23], trichomonadid protozoa [24], Naegleria [25] and Vahlkampfia species [26]. As in Naegleria, the differences between genus (inter-species) and species (intra-species of N. fowleri) were due to the polymorphism sequence that occurred at the ITS2 and ITS1 regions, respectively [27]. The ITS region was also being used for the identification of the new Naegleria isolates [10,28].
In the current study, the species-specific primer set for N. fowleri failed to amplify and did not produce any amplicon against all the 41 isolates, while the Naegleria genus-specific primer sets successfully amplified the DNA template from all isolates, producing 5 DNA amplicons consisting of 381, 408, 450, 457 and 756 bp, respectively. The amplicon of 408 bp was the most common, produced by 35 isolates, whereas 381, 450 and 457 bp amplicons were produced by 2 isolates each, namely, AC7 (JN034055) and AC8 (JN034056), S1 (JN034053) and M4 (JN034054), and SP9  (JN034047) and SP14, respectively (Table 3). Subsequently, the 756 bp fragment was belived to be from organisms that contaminated the 3 isolates (SP12, L1 (JN034048), L5), although all possible causes of contamination in the Naegleria culture was considered prior to DNA extraction. These 3 isolates were originally from swimming pools and recreational lakes. Production of various amplicons by this Naegleria genus-specific primer set have been reported previously by Pélandakis and Pernin [29], with amplicons of 800, 600 and 500 bp showing homology with Hartmanella, Vahlkampfia and Willaertia, respectively. Neverthelless, this Naegleria genus-specific primer set did not amplify the Acanthamoebe-like species. were isolated from swimming pools frequently used by local and foreign students of the University Malaya. M1 (JN034050), M2 (JN034052), M4 (JN034054) were from water-tanks walls and the water used by thousands of Muslims to perform ritual ablutions. AC1 (JN034049) was isolated from dust at the air-conditioner in a frequently used lecture theatre at a Medical Faculty. L1 (JN034048) was isolated from water of the University's lake that actively used by the students at the evening as their water sports. S1 (JN034053) was from a favourite recreational stream (Sungai Congkak). AC7 (JN034055) and AC8 (JN034056) were from air-conditioners in the laboratories at another University in Kuala Lumpur. The 408 bp amplicon that was produced by isolates SP1 (JN034051) and M2 (JN034052) showed 100% identy in the entire lengths and sequences of ITS1(33 bp)-5.8S(175 bp)-ITS2(113 bp) against Naegleria philippinensis isolate RJTM (AM167890). Three isolates [M1 (JN034050), AC1 (JN034049) and L1 (JN034048) showed differences due to a single base substitution (TRC), at the position 120 (in the 5.8S), 252 and 227 (at the ITS2), respectively against Naegleria philippinensis isolate RJTM (AM167890). Isolates M1 (JN034050), Table 4. Length and position of base in the ITS1-5.8S-ITS2 sequence of the selected isolates (amplicon T1-T5) and their closest relative strain from GenBank.  identical sequence with Naegleria laresi (AJ566630). S1 (JN034053) and M4 (JN034054) showed 16 and 15 base substitutions respectively, with a 2 base insertions in both isolates, resulting in 2 bp increase in length compared to Naegleria schusteri (AJ566626). Additionally, S1 (JN034053) and M4 (JN034054) were proposed to be strains of a new Naegleria species, because of their intra-species variation due to the differences of only a single base substitution (T 346 RC) in S1 (JN034053) at ITS2 region. De Jonckheere [1] reported previously that the polymorphism of ITS2 sequences could determine the species within the genus (inter-species) of Naegleria. All the strains from similar species (intra-species) of all Naegleria (except Naegleria byersi and Naegleria andersoni) showed identical lengths of the ITS2 sequences. In one strain each of N. byersi and N. andersoni, a difference in 1 bp has been detected. For the largest amplicon of 756 bp, it was initially suspected to be homologous with contaminated Hartmanella, since this fragment is nearest to fragment 800 bp that was reported to be homologous with Hartmanella [29]. However after sequencing this amplicon from isolate L1 (JN034048), it did not show homology to any of the sequences from the GenBank. The smallest amplicon of 381 bp produced by 2 isolates [AC8 (JN034056, AC7 (JN034055)] were 100 and 99.7% identical with Vahlkampfia avara (AJ698837), respectively. Only isolate AC7 (JN034055) showed a difference in a single base substitution, T 53 RC) in the 5.8S sequence. This finding was different from Pélandakis and Pernin [29], who obtained a PCR product of approximately 600 bp for Vahlkampfia spp.
Multiple sequence alignment by BLAST analyses using the ClustalW programme against rRNA sequences (SSU-ITS1 ciguana (AJ973126). Both, strains selected (from this study) and the reference isolates provided by GenBank, have similar length and sequence in the ITS1 (33 bp) -5.8S (175 bp) regions. Polymorphism in the ITS2 sequences is an important feature used for the detection of intra-and interspecies variation of non-pathogenic Naegleria as in this study. This finding supported the findings of De Jonckheere [1,30] who showed differences in the sequence of the ITS2 region in species of the genus Naegleria.
In this study, both results from PCR-and morphology-based detections revealed that all 39 isolates with flagella were Naegleria, while 2 isolates [AC7 (JN034055) and AC8 (JN034056)] with no flagella were Vahlkampfia species. For the Vahlkampfia species, besides being unable to enflagellate, its viable trophozoites were longer in length (up to 35 mm) and were pinkish in colour.
Subsequently, all the tests such as temperature tolerance, using an N. fowleri species-specific primer set and rDNA sequencing failed to detect N. fowleri from our environment. Therefore, up to this date we have not been able to isolate N. fowleri in the state of Selangor and the Federal Territory of Kuala Lumpur and Putra Jaya. This may explain why PAM fatal cases have not been reported from these areas. On the other hand, all three species of Naegleria detected by ITS1-ITS2 were present in both, chlorinated (from swimming pools and water tanks) and non-chlorinated water (from lakes and streams). Thus our chlorinated water did not affect the viability of Naegleria species. Therefore, N. fowleri contamination of chlorinated as well as non-chlorinated water, could pose a danger to public health.
Geographically, Malaysia has less aquatic environments compared to Thailand, thus explaining why Malaysia has less surface water related activities. This does not mean that our population is not exposed to N. fowleri. Most Malaysian Muslims frequently use water for their ritualistic washing before praying, which may espose them to the risk of amoebae entering the nostrils if domestic water is contaminated with N. fowleri. In addition the northen region of Peninsular Malaysia [Perlis (100.13uE longitude, 6.40uN latitude), Kedah (100.30uE longitude, 6.10uN latitude) and Kelantan (102.25uE longitude, 6.10uN latitude)] that borders with Thailand (which has reports of N. fowleri in their aquatic environment in several provinces), has floods due to heavy rainfall 2-3 times yearly. Most children would take the opputunity to play and engage in water-related activities during this time. Therefore, detection of N. fowleri should also be caried out in this region, especially during flooding. Clinicians should be be aware of PAM and include it in the differential diagnosis of meningoencephalitis. Preventive and control measures include health education for the public and awareness among medical practioners and adequate treatment of public water supplies.