1. Introduction
Legionella is gram-negative, ubiquitous bacteria that are most commonly found in man-made water systems.
Legionella causes Legionnaires’ disease, an acute pneumonia-like infection with high case-fatality rate and Pontiac fever, an influenza-like, self-limited illness that is not notifiable in the majority of countries and therefore, underestimated [
1,
2].
Legionella is one of the most important water-related pathogens, which can lead to both outbreaks and sporadic cases. Dissemination of
Legionella is carried out through water aerosols containing
Legionella, which occur in water supply systems, such as shower handsets, taps, cooling towers, fountains, and spas. In the form of aerosols, bacteria can be spread kilometers away from the primary source of infection while still maintaining viability [
3].
Inside the water supply systems and the biofilms,
Legionella come into contact with other bacteria and free-living amoeba (FLA). The most frequently isolated FLA are
Acanthamoeba,
Vermaoeba, and
Naegleria [
4]. Certain amoeba genotypes, such as
Acanthamoeba T4 and
Naegleria fowleri, can be human and animal pathogens.
Acanthamoeba can cause amoebic keratitis, which can lead to blindness if not treated, while brain infections can cause fatal Granulomatous Amoebic Encephalitis.
N. fowleri can cause primary amoebic meningoencephalitis, which is often fatal [
5].
Despite the fact that the majority of amoebas are predators of other microorganisms, especially bacteria, interaction between them has been observed. Several amoeba species may serve as hosts for bacteria in a natural environment and man-made water systems, such as water pipes, cooling towers, etc., as bacteria have developed adaptation mechanisms over time to escape predating [
6,
7]. As a result, bacteria may not only be amoeba resistant, they may be able to multiply in the amoebas and take advantage of the benefits they can offer, such as a nutrient source and additional protection against temperature, disinfectants or UV radiation [
8,
9]. Mechanisms that have helped bacteria escape predation can become factors of bacterial virulence [
10,
11,
12]. Parasitizing FLA and being a member of mixed species biofilm increases the capacity to produce polysaccharides and to form biofilms [
13,
14].
In Latvia, an average of 1.5 cases of Legionnaires’ disease are reported per 100,000 inhabitants, which is very similar to the EU average [
15]. However, the average age of residential buildings in Latvia is high, and only a small part of the buildings have renovated water supply systems [
16]. There are no previously published studies on the diversity of FLA and co-occurrence with
Legionella in water supply systems of residential and other buildings in Latvia. This study aimed to investigate the co-occurrence of
Legionella spp. and FLA in drinking water supply systems to understand the interaction between
Legionella and FLA and potential consequences to public health concerns.
3. Results
Overall,
Legionella spp. were detected in 114 of 268 water samples (
Table 1), including 11 of 44 cold water samples (25.0%) and 103 of 224 hot water samples (46.0%). At least one
Legionella spp. positive sample was detected in 50 of 92 buildings (54.3%).
The most commonly observed Legionella species were Legionella pneumophila, which was detected in 105 from 114 Legionella spp. positive samples (92.1%). Legionella rubrilucens was observed in seven samples (6.1%), and Legionella anisa was found in two samples (1.8%). The most predominant L. pneumophila serogroup (sg) was sg 3, which was identified in 60 of 105 cases (57.1%). L. pneumophila sg 2 was identified in 17 cases (16.2%) and sg 1 in 16 cases (15.2%), while serogroups 6 and 9 were observed in six cases each.
Level of Legionella spp. colonization ranged from 50 cfu/L to 13 × 103 cfu/L. For cold water samples, the average level of contamination was 4.6 × 103 cfu/L (min 100 cfu/L, max 12 × 103 cfu/L) and 1.5 × 103 cfu/L for hot water samples (min 5 cfu/L, max 13 × 103 cfu/L). Overall, higher levels of Legionella colonization were observed in samples from apartment buildings (average 3.3 × 103 cfu/L, min 50 cfu/L, max 13 × 103 cfu/L) than in public buildings (average 1.4 × 103 cfu/L, min 50 cfu/L, max 6.5 × 103 cfu/L), and hotels (average 8.9 × 102 cfu/L, min 5 cfu/L, max 7.5 × 103 cfu/L).
Chi-squared tests showed no association between water source (χ2 = 1.6, p = 0.25), type of building (χ2 = 1.1, p = 0.57) and occurrence of Legionella spp. No differences in the occurrence of Legionella spp. were observed between sampling season as well (χ2 = 0.5, p = 0.50). Significantly higher occurrence of Legionella spp. was observed in hot water samples (χ2 = 6.6, p = 0.01). The presence of Legionella spp. was significantly higher in hot water samples with temperature less than 50 °C (χ2 = 7.6, p = 0.049), while the temperature of the cold water did not show any impact on the occurrence of Legionella spp.
The average temperature of the cold water was 13.1 ± 0.9 °C. Temperature measurements showed that in five cases, the temperature of the cold water was equal to or above 20 °C.
The average temperature of the hot water was 46.6 ± 0.7 °C. Temperature measurements showed that only in 38 of 224 hot water samples’ temperature exceeded 55 °C (
Table 2).
No significant differences were observed between temperatures in apartment buildings, public buildings, and hotels in Riga and other cities (p = 0.43).
FLA were observed in 207 of 268 water samples (
Table 3), including 37 cold water samples (84.1%) and in 170 hot water samples (75.9%). At least one FLA positive sample was observed in 83 of 92 buildings (90.2%).
Chi-squared tests showed no association between water type (i.e., cold or hot) and the presence of FLA (χ2 = 1.4, p = 0.33). However, a higher diversity of FLA was observed in hot water samples (χ2 = 10.3, p = 0.0.35).
The presence of FLA was significantly higher in hot water samples with temperature less than 50 °C (χ2 = 21.3, p < 0.0001), while the temperature of the cold water did not show any impact on the occurrence of FLA.
The sampling season and type of building showed no association with presence of FLA (χ2 = 6.0, p = 0.11 and χ2 = 3.9, p = 0.14, respectively). A higher occurrence of FLA was observed in samples from buildings which received groundwater (χ2 = 5.8, p = 0.024). However, the water source had no impact on the diversity of FLA (χ2 = 7.5, p = 0.11).
Overall eight genus of FLA were identified in 207 samples, including Acanthamoeba spp. (146 samples), Vermamoeba spp. (77), Naegleria spp. (58), Flamella spp. (3), Centropyxis spp. (2), Vrihiamoeba spp. (2), Echinamoeba spp. (1), and Tetramitus spp. (1).
In 47.4% of samples, only one genus of amoeba was observed (127 of 268 cases). In 69 samples (25.7%), two different genera were detected, while in 10 cases (3.7%) there were three genera and in one sample (0.4%) four different amoeba species were detected. Most frequently observed combinations of amoeba species are shown in
Table 4.
Association between the presence of FLA and the occurrence of
Legionella spp. was also observed (χ
2 = 58.5,
p < 0.0001). No
Legionella spp. positive samples were observed in the absence of FLA (
Table 5) and the concordance rate between the presence of FLA and
Legionella spp. reached 55.1%. Correlation analysis revealed a moderate positive correlation (r = 0.467,
p < 0.01) between the presence of
Legionella and FLA in water samples.
The occurrence of
Legionella spp. was significantly higher in samples with lower diversity of FLA (χ
2 = 64.9,
p < 0.0001). The highest occurrence of
Legionella spp. was observed in samples with only one amoeba genera (
Table 6).
The diversity of FLA did not show any impact on Legionella species (χ2 = 1.9, p = 0.758), serogroups (χ2 = 4.6, p = 0.797) or on count of colony forming units (χ2 = 7.9, p = 0.247).
Only one amoeba genus showed significant association with the occurrence of Legionella spp.—Acanthamoeba spp., which was observed in 146 samples. Correlation analysis revealed moderate positive correlation (r = 0.404, p < 0.001) between the presence of Acanthamoeba spp. and Legionella spp. in the samples. Vermamoeba spp., which was observed in 77 samples (r = 0.131, p = 0.074), and other genus did not show a statistically significant impact on the occurrence of Legionella spp.
Overall, 14 species of FLA were identified. Only two genus—Acanthamoeba spp. and Naegleria spp. contained more than one species. From identified species, the most common were Vermamoeba vermiformis (54%), Acanthamoeba castellanii (12%) and Acanthamoeba polyphaga (8%).
4. Discussion
Our research showed that FLA and Legionella are common in drinking water in Latvia and water supply systems in multi-apartment buildings, hotels, sports clubs, and office buildings can become a potential source of Legionella infection.
Contamination of drinking water with
Legionella was relatively high (42%). The occurrence of
Legionella varied from an average of 12.5% in cold water samples with the most frequent occurrence observed in hot water, in areas receiving untreated groundwater, where 54% of the samples were
Legionella positive. FLA and
Legionella occurrence and diversity did not differ significantly for samples taken within one building. Similar studies in Japan revealed
Legionella prevalence in 6.5% of water samples using a culture method [
23], while in the United Kingdom,
Legionella was found in 8% of water samples taken in household showers [
24]. In both studies,
Legionella was also analyzed using a second method—loop-mediated isothermal amplification in Japan and quantitative PCR in the United Kingdom, and the observed incidence of
Legionella changed to 47.8% and 31%, respectively. Given the nature of DNA-based techniques, it is possible that in our study, the incidence of
Legionella would be even higher if additional molecular techniques were used. However, the occurrence of
Legionella may vary significantly from country to country. Prevalence of
Legionella in potable water in Germany varied from 20.7% [
25] to 32.7% [
26], in Italian residential buildings different
Legionella species were detected in 26% of the hot water networks [
27].
In this study, close similarities with the study carried out in Hungary [
28] were observed. Between the years 2006 and 2013, a total of 1809 water samples were taken in Hungary in 168 different buildings, and 60% of buildings were colonized by
Legionella, 46% of hot water samples were positive for
Legionella [
28]. The main reasons for the high contamination are also similar in both countries. The Hungarian authors mention the low temperature of hot water and the lack of adequate monitoring and risk management. In our study, the average temperature of hot water was 46.6 °C, which is appropriate for maintaining the viability of
Legionella, with only 17% of samples of hot water above 55 °C, which would be the preferred temperature to avoid the proliferation of
Legionella [
29].
The low temperature in water supply systems can have several explanations. First, it relates to the overall economic situation and the public’s understanding of energy saving. Second, particularly in public buildings and hotels, it may be linked to a lack of staff competence, and third, in a large part of old buildings without the complete renovation of the water supply system, there is no technical solution to raise the temperature, since the old facilities are not suitable for maintaining temperatures above 55 °C. However, raising temperatures alone will not lead to significant improvements without the necessary dimensioning and recirculation [
30].
The occurrence of FLA was significantly higher. On average, 77.2% of samples contained at least one genus of FLA and, depending on the type of sample, the occurrence of FLA could reach 95%.
The most commonly identified were
Acanthamoeba (54.5% of all) and
Vermamoeba (28.7% of all), followed by vahlkampfiid amoebae, yet more than 20% of the samples contained more than one genus of amoebas, most commonly
Acanthamoeba and
Vermamoeba were found together. Other studies have also confirmed that FLA was found in drinking water and environmental samples [
31,
32,
33,
34,
35], biofilms [
36,
37], in industrial waters [
38], and cooling towers [
39].
In the samples collected during this study,
Legionella was always isolated along with FLA, no samples containing
Legionella in the absence of FLA were observed. Co-occurrence of
Legionella and FLA in the water systems may indicate an increased health risk in proximal areas of the system, where lower temperatures are commonly observed [
8]. In addition to protecting
Legionella, FLA are also able to maintain the viable but non-culturable strains [
40] and provide long-term persistence and transmission of
Legionella [
41].
Although Legionnaires disease is included in the list of notifiable diseases in Latvia, currently there are no regulations in Latvia for the introduction of risk management plans and regular environmental monitoring of
Legionella, and, in most cases, the risk of
Legionella is not taken into account in the management of buildings. Thus, the minimum requirements for hot water temperature at the points of consumption are not regulated in Latvia. Many countries have guidelines and regulations for the prevention of the Legionnaires disease. However, the regulatory framework for
Legionella control differs between countries [
42]. Most of the guidelines and regulations are focused on
Legionella control, with no regard to the presence of FLA. The current approach in
Legionella control involves monitoring of
Legionella sp. with culture-based methods and does not allow to assess true concentration of
Legionella cells [
43]. The data obtained in our study can help to focus on the extensive distribution, close interaction, and long-term persistence of
Legionella and FLA. However, further studies may be needed to identify whether FLA-targeted risk management plans will contribute to decreasing the risk to public health.