Isolation and identification of free-living amoebae isolated from well water in Istanbul

Free-living amoebae (FLA) are numerous and systematically heterogeneous protozoan groups with worldwide distribution including water, soil, dust, and air. They have been isolated from natural water environments such as rivers, lakes, springs, and also man-made water systems such as drinking water networks, poorly chlorinated swimming pools, and cooling waters (Tsvetkova et al. 2004; Edagawa et al. 2009; Gianinazzi et al. 2010; Caumo & Rott 2011; Scheikl et al. 2014).

Some genera of the FLA, such as Naegleria, Acanthamoeba, Balamuthia, Sappinia, and Vermamoeba, found in the environment are associated with human and animal diseases. Naegleria fowleri causes a rare, but an acute, fulminating, and rapidly developing fatal disease of the central nervous system, primary amoebic meningoencephalitis, in healthy children and young adults with a history of aquatic activities in freshwater. Acanthamoeba spp. are the causative agents of amoebic keratitis (AK), granulomatous amoebic encephalitis (GAE), and in some cases skin infections. Both encephalitis and skin lesions are seen in immunocompromised patients, whereas AK is commonly seen in contact lens wearers with insufficient hygiene (Visvesvara et al. 2007; Siddiqui & Khan 2012). The other FLA, Balamuthia mandrillaris and Sappinia diploidea (later re-identified as Sappinia pedata), cause central nervous system infections and Hartmannella vermiformis (later re-identified as Vermamoeba vermiformis) is rarely associated with keratitis (Kennedy et al. 1995; Gelman et al. 2001; Visvesvara et al. 2007).

Besides their own pathogenic ability, some FLA may serve as reservoirs of a large number of intracellular pathogenic bacteria (such as Legionella pneumophila, Mycobacterium spp., and Pseudomonas spp.) and viruses (Greub & Raoult 2004). Consequently, interest both in FLA and the association of microorganisms with FLA has increased significantly in the last decade in relation to the risk to public health (Scheid 2014).

It is important to mention that the distribution of FLA as well as their potential impact on public health remains unknown due to the limited studies in Turkey. Therefore, the aim of this study was to investigate the presence of potentially pathogenic FLA in well water samples in İstanbul. To the best of the author's knowledge, this is the first report on the identification of Acanthamoeba genotype T3 from well water samples in İstanbul.

The wells selected in this study are known as water supplies mainly used for irrigation. The water is extracted to the surface using a submersible pump from the wells with a depth of 100–250 m and a diameter of 6–7 in. After extraction, well water is stored in water tanks and distributed to all points of use by a hydrophore pump. The depth of the rectangular-shaped water storage tanks is approximately 2 m. Well yield is a maximum at 120 m³/day.

A total of 10 water samples were collected directly from the taps where the well water flows. From each sampling point, a total of 500 mL of water was collected in a sterile plastic bottle. Prior to sample collection, temperature, pH, and free chlorine level of water were measured. The water samples were transported to the laboratory at ambient temperature and processed within 24 h.

Water samples were filtered by a vacuum filtration system using a 0.45 μm pore size filter. The membrane filters were then inverted on Escherichia coli-seeded non-nutrient agar plates in triplicate. The plates were incubated at 30 °C and observed daily for amoebic growth up to 15 days under an inverted microscope (Üstüntürk & Zeybek 2012). Before applying the molecular methods, isolated FLA were identified at the genus level based on their morphology according to the criteria of Page (1988).

Osmotolerance and thermotolerance assays were performed to evaluate the potential pathogenicity of isolated Acanthamoeba strains. For osmotolerance assays, both trophozoites and cysts were inoculated on E. coli-seeded non-nutrient agar plates containing 1 M mannitol. Agar plates were incubated at 30 °C for up to 72 h. For thermotolerance assays, both trophozoites and cysts were inoculated on E. coli-seeded non-nutrient agar plates. Agar plates were incubated at 37 and 42 °C for up to 72 h (Khan et al. 2001). After incubation, the number of trophozoites or cysts observed around 20 mm from the middle of each plate was calculated and scored as 0 (−), 1–15 (+), 16–30 (++ ), and >30 (+++) (Mohd Hussain et al. 2019). All plating assays were performed in triplicate.

Two parallel plate cultures were set up for each isolated strain. Trophozoites were harvested from both plate cultures using a sterile cotton swab, transferred into sterile phosphate-buffered saline (PBS), and washed by centrifugation at 2,500 rpm for 10 min. The pellet was resuspended in 200 μL of PBS. Whole-cell DNA was isolated using a QiaAmp DNA mini kit (Qiagen, Germany).

For the identification of the Acanthamoeba and Vermamoeba strains, a fragment of the 18S rRNA gene was amplified using genus-specific primers as described previously (Schroeder et al. 2001; Scheikl et al. 2014). Acanthamoeba castellanii (ATCC 50373) and H. vermiformis (ATCC 30966) strains were used as positive controls for the Acanthamoeba and Vermamoeba polymerase chain reactions (PCRs), respectively. For the isolates other than Acanthamoeba and Vermamoeba, fragments of the 18S rRNA gene were amplified using the external universal SSU1 and SSU2 primers (Corsaro et al. 2014) and the universal internal P1–P3 and P1rev–P3rev primers (Walochnik et al. 2004). All PCRs were performed in 50 μL reactions. DNA-free water was used as a negative control. Amplified DNA was visualized by 2% agarose gel electrophoresis using GelRed (Biotium, USA).

After amplification, the quantified PCR products and appropriate primers were placed in sterile PCR tubes and sent to RefGen (Ankara, Turkey) for purification and sequencing. Sequences were obtained from both strands. A consensus sequence was compiled for each DNA fragment and compared with published sequences from GenBank using NCBI Nucleotide BLAST search. Multiple sequence alignments were performed using ClustalX (Thompson et al. 1997). Sequence data were processed with the GeneDoc sequence editor (Nicholas et al. 1997).

The physico-chemical parameters of well water are presented in Table 1. The maximum temperature was measured as 27 °C, while the minimum temperature was measured as 22 °C. The average pH value was 7.25 with a range of 7–8. Free chlorine was measured as 0 ppm for all the well water samples.

FLA were isolated from 8 out of the 10 (80%) well water samples investigated in this study. Samples coded S4 and S9 did not yield any FLA growth. According to the microscopic investigations, two different FLA genus morphologies were observed in the sample coded S8. Other positive samples yielded single FLA genus growth. The isolated strains were sub-cultured on E. coli-seeded non-nutrient agar plates. Unfortunately, several cultures of FLA (from the samples coded S1, S5, and S6) were lost due to fungal overgrowth. Six FLA isolates were successfully brought into clonal monoxenic cultures, i.e. two Acanthamoeba (from samples coded S7 and S8) and four Vermamoeba (from samples coded S2, S3, S8, and S10) strains. These strains were subjected to DNA sequencing for molecular analysis and confirmation of morphological identification.

The trophozoites and cysts of two Acanthamoeba isolates, isolated from samples coded S7 and S8, respectively, were subjected to osmotolerance and thermotolerance assays to evaluate their potential pathogenicity. The trophozoites and cysts of both of the two investigated Acanthamoeba isolates grew at 30 °C with 1 M mannitol and at 37 °C. Trophozoites and cysts of both strains did not grow at 42 °C. The results of the osmotolerance and thermotolerance assays are shown in Table 2.

Two of the isolated Acanthamoeba strains, isolated from samples coded S7 and S8, were molecularly identified as genotypes T4 and T3, respectively, showing more than 98% sequence identity to various reference strains. All representatives of the genus Vermamoeba (from samples coded S2, S3, S8, and S10) belonged to the species V. vermiformis, showing >99% sequence identity to all other sequences of V. vermiformis (or H. vermiformis) available in GenBank in the sequenced fragment of the 18S rDNA. NCBI Nucleotide BLAST search results are detailed in Table 3.

Water environments are well known to harbour pathogenic FLA, which can lead to severe infections both in humans and animals (Schuster & Visvesvara 2004). Therefore, investigating the presence of these microorganisms in such areas is very important. However, the studies screening the FLA in well water samples are very limited both in Turkey and worldwide. The present study reports, for the first time, the presence of potentially pathogenic FLA in well water samples (8 out of 10, 80%) in İstanbul. Acanthamoeba spp. and V. vermiformis were isolated in 20 and 40% of the investigated well water samples, respectively. A lower isolation rate of FLA was reported in a similar study carried out in Kayseri, Turkey (Kuk et al. 2013). In that study, the authors investigated only Acanthamoeba genus and detected Acanthamoeba genotype T4 in 5 out of 26 (19.23%) well water samples. Similar results were described in a study carried out in Guinea-Bissau where FLA were isolated in 15 out of 22 (68.2%) well water samples (Baquero et al. 2014). Surprisingly, no V. vermiformis was detected in that study. The authors reported the presence of Acanthamoeba genotypes T3 and T4, N. fowleri, Tetramitus sp., Willaertia sp., and B. mandrillaris in the investigated samples. Another study conducted in Italy showed 50% positivity (two out of four) of FLA in well water, describing Acanthamoeba genotype T15 in one sample (Montalbano Di Filippo et al. 2015). Consequently, the high variability observed in the prevalence of FLA around the world may be due to several factors, such as the application of different methodologies and the characteristics of both the water and the geographical region investigated. Since their potential to cause serious infections such as in the central nervous system, ocular keratitis, and other ailments, they have been extensively investigated throughout the world, understanding the factors affecting the occurrence of different genus/species is of paramount importance (Visvesvara et al. 2007).

In Turkey, very limited studies have investigated the presence of FLA in environmental samples up to date (Saygı 1979; Akın 2000; Saygı & Akın 2000; Kilic et al. 2004; Burak & Zeybek 2011; Coşkun et al. 2013; Üstüntürk-Onan & Walochnik 2018). Only a few had confirmed the morphological identification by molecular techniques among these studies. In addition, Acanthamoeba and Vermamoeba were reported as major contaminants in human-related habitats compared with other FLA, as in the studies conducted in Europe (Tsvetkova et al. 2004; Gianinazzi et al. 2010; Scheikl et al. 2014). The most effective factor in the absence/lower prevalence of other FLA genera in environmental samples could be due to the ecological characteristics of the sampling areas. The use of different methodologies for the isolation, such as the cultural temperature (30 °C), is not adequate for all the species, and also the identification (microscopic or molecular) could be considered as other factors.

Increased temperature and osmolarity have been indicated as physiological determinants to be used to differentiate and characterize the potentially pathogenic isolates of environmental and clinical Acanthamoeba strains. In vitro growth of an Acanthamoeba isolate under relatively high temperature or osmotic stress might be correlated with virulence, the ability of an isolate to adapt, and survive in the mammalian host tissues (Khan et al. 2001; Khan & Tareen 2003; Koehsler et al. 2009; Gianinazzi et al. 2010). In addition, previous studies have suggested that many Acanthamoeba isolates from tap water sources might possess some pathogenic ability (Kilvington et al. 2004; Lorenzo-Morales et al. 2005; Edagawa et al. 2009). Therefore, thermotolerance and osmotolerance assays were used in order to evaluate the potential pathogenicity of two Acanthamoeba strains isolated in this study. All isolates developed at 37 °C and in 1 M mannitol at 30 °C, whereas none was able to develop at 42 °C (Table 2). According to these results, these strains were considered to have low pathogenic potential.

To date, 22 different genotypes (T1–T22) of the genus Acanthamoeba have been identified based on 18S rRNA gene sequencing (Walochnik 2018). Among them, some genotypes are pathogenic causing infections in humans. The most common genotype in Acanthamoeba infections, both GAE and AK, is T4, which also is the most abundant genotype in the environment having a universal distribution (Schuster & Visvesvara 2004; Siddiqui & Khan 2012). T3, another genotype, is also of great clinical importance due to the increasing number of studies reporting AK infections caused by this genotype (Niyyati et al. 2010; Omana-Molina et al. 2016). In this study, one of the two isolated Acanthamoeba strains was identified as genotype T3 and the other as genotype T4. Based on the potential pathogenic abilities of these genotypes, a wide range of research is needed to detect the in vitro cytotoxicity and in vivo pathogenicity of the isolated strains in addition to determine the phylogenetic and pathogenic relations in Acanthamoeba infections.

In addition to their own potential pathogenicity, some species of FLA were considered as a ‘Trojan horse’, demonstrating the ability to promote a suitable environment for intracellular survival of many pathogenic bacteria of clinical relevance for humans and animals (Barker & Brown 1994). They allow the survival, the multiplication, and the dissemination of these pathogens in water systems, resulting in their better survival in the environment, resistance to antibacterial substances, and increased virulence (Greub & Raoult 2004). Therefore, effective disinfection methods that target not only bacteria but also FLA should be developed in man-made water systems.

Nevertheless, the present study reports, for the first time, the presence of potentially pathogenic FLA strains of Acanthamoeba and V. vermiformis in well water samples from İstanbul. Clinicians and public health professionals in this city should be aware of the risks of disease that these microorganisms may cause. Moreover, knowledge of the identification of FLA may help them to diagnose and treat either healthy or immunocompromised individuals. In addition, precautions should be taken into account in order to avoid potential health risks to individuals inhabiting these areas.

In summary, this study highlights that potentially pathogenic FLA are widely distributed in well water in İstanbul. The presence of these microorganisms in habitats related to human activities supports the relevance of FLA as a potential public health concern. Therefore, monitoring the presence of FLA in aquatic environments and assessing the pathogenicity of the isolates might be an approach to alert health professionals to improve the disinfection strategies and minimize the risks from pathogenic FLA.

The author wishes to thank Assoc. Prof. Zuhal Zeybek for providing the ATCC strains of A. castellanii and H. vermiformis used as positive controls in PCRs. This work was supported by the Scientific Research Project Coordination Unit of Istanbul University. Project number: FBA-2018-30882.

All relevant data are included in the paper or its Supplementary Information.