Investigation of Virulence Genes and Biolm Formation Among Legionella Pneumophila Isolated from Hospital Water Sources

Background: Legionella pneumophila as a ubiquitous bacterium is inherently resistant to chlorine in tap water. It can easily enter water piping systems and get transmitted to immunocompromised populations and cause severe pneumonia. Owing to the fact that its presence in water sources doesn’t necessarily lead to onset of disease; therefore, several factors such as inhaled bacteria dose, virulence factors and diversity of serogroups can be considered as contributing factors. The main aim of current project was to investigate the contamination rate of hospital water systems with Legionella by culture and evaluate presence of major virulence factor genes as well as the ability to form biolms among the Legionella isolates. Results: Twelve (12%) of the 100 water samples produced positive results in culture method. Additional conrmation was performed by PCR method with specic primers for Legionella genus (16SrRNA) and pneumophila species (mip). Fifty (5%) samples of 12 with positive culture have a colony forming unit higher than 1000cfu/100 ml. Legionella were isolated with a rate of 8%, 3% and 1% from shower heads, oxygen humidier bottle and water bath, respectively. PCR assay for the virulence genes showed that all 12 (100%) isolates were positive for mip genes, 9 (75%) were positive for dot gene, 8 (66.66%) were positive for hsp, 6(50%) were positive for lvh and 4(33.33%) for rtx. Two of the isolates displayed higher ability to form biolm in reference to the standard strain. Conclusion: Although the presence of Legionella pneumophila in hospital environment does not necessarily confer a threat to public health; continuous monitoring of water sources should be conducted in order to avoid elevated concentration of this bacterium and visible biolm formation.


Background
Legionella pneumophila is a facultative intracellular pathogen, which is the etiological agent of legionellosis, legionnar's disease [1][2][3][4]. Legionella pneumophila as a ubiquitous bacterium is resistant to chlorine in treated water and can easily enter piping systems and transmitted to human through contaminated arti cial and natural water systems including cooling systems, air conditioning, shower baths and drinking water and cause mild to severe pneumonia with high mortality among immunocompromised populations [5][6][7]. The bacterium enters respiratory tract by infected droplets and swallowed by macrophages. It prevents the fusion of lysosomes to contaminated phagosomes [8,9]. The bacterium is an opportunistic pathogen that can cause disease in compromised immune system individuals and a common cause of hospital acquired pneumonia [10]. Environmental surveillance in hospitals as performed by culturing of water supplies is useful for risk assessment and prevention of disease [11]. Owing to the fact that presence of Legionella in water source doesn't always leads to disease; therefore, several factors such as inhaled bacteria dose, virulence factors and diversity of serogroups can effect on this process [12,13]. Virulence factors of legionella are LPS, agella, T2SS secretion system and some other membrane proteins [14]. Many of these are found in the legionella pneumophila wall compositions, which include OMVs, peptidoglycans associated with lipoprotein (PLA), phospholipase A, external membrane proteins, HSP60, FeoB, and mip [15]. lvh is a locus derived from proteins of the type 4 secretion system that is involved in invasion and bacterial virulence [16]. The rtxa protein in Legionella pneumophila causes the bacterium bind to host cells and invades them. This bacterial super cial protein also plays a role in the binding and entry of bacteria to the Acanthamoeba castellanii protozoan and its survival in nature. In rtxa mutants, the ability to bind and entery of the bacterium into epithelial cells and monocytes is reduced [17]. The L. pneumophila major outer membrane protein (MOMP) is involved in the attachment to host cells [18]. The heat shock protein (Hsp60) is also important for attachment to and invasion of a HeLa cell model [19]. Mip, the macrophage infectivity potentiator, is a membrane-associated homodimeric protein that is mainly found on the bacterial surface [20]. One mechanism that has been proposed for establishment of the unique L. pneumophila replication vacuole involves the Dot/Icm secretion system [21].
The main aims of current project was to investigate the contamination frequency of hospital water systems with Legionella by cultural methods, evaluate the water condition for the presence of legionella and detection of major virulence factor genes and ability for bio lm formation among Legionella isolates for risk assessment and prevention.

Results
Collection and measurement of temperature, pH and residual chlorine of water samples During one year period, from December 2018 to December 2019, a total of 100 water samples were collected from nine hospitals in Tehran, Iran. Samples were from oxygen humidi er bottle (43%), shower head (46%) and water bath (11%). All samples had residual chlorine between 0.3 to 0.6 mg/ml (Fig. 1). The pH of the water samples were between 5.6 to 6 ( Fig. 1). Temperature of the samples was between 25-30 C in 43%, 31-36 C in 9% and 37-41 C in 48% of the cases (Fig. 1).

Concentration And Treatment Of Water Samples
For this step, we examined two different concentrations (centrifugation and ltration) and then used two different treatments methods for each concentration (heat and acid). Water samples were concentrated with centrifuge at 3000 rpm for 10 minutes at 4º C. The deposits were resuspended in 5 ml of original water sample as concentrates. In concentration by ltration, at rst all water samples were passed through 0.45 µm pore size nitrocellulose membranes. Then the membranes were aseptically removed, put into sterilized 50 ml tube and resuspended in 10 ml of the original water samples. Each concentrated water samples was shaken for 30 min to get out bacterial cells from the membrane. To exclusion of nonlegionella bacteria, the concentrated water samples were diluted (1:10) in Kcl-Hcl solution (pH: 2.2), mixed and incubated at room temperature for 4 minutes. In an alternative treatment for concentrated water, heating at 50ºC for 30 min was used.

Identi cation of Legionella pneumophila colony
For veri cation of suspected colonies with the typical ground glass appearance on GVPC, these colonies were inoculated on BCYE with or without L-cysteine and onto non selective media such as blood agar. Identi cation of Legionella pneumophila were performed with gram stain and biochemical tests. Strains unable to grow on media without L-cycteine and blood agar were further analyzed by PCR with speci c primers for Legionella genus (16srRNA gene) and pneumophila specie (mip gene).

DNA extraction from Legionella pneumophila colonies
Freshly grown legionella colonies on GVPC medium were suspended in distilled water and 200 µl of the suspension was used for DNA extraction and puri cation according to a commercial kit manufacture instruction. 50 µl of lysis solution was added into each microcentrifuge tube containing legionella suspension for breakage of the cell membrane. The tubes were vortexed and then incubated at 95 °C for 10 min in a hot plate and then were left to equilibrate at room temperature for 5 minutes. The tubes were then vortexed and centrifuged at 6,000 rpm for 2 minutes and the eluted DNA was transferred to an eppendorf tube. Quality of the extracted DNA was assessed by optical density at 260 nm and electrophoresis on agarose gel. Extracted DNA from legionella colonies was stored at − 20 °C for a maximum of 2 days.

Biochemical Analysis, Culture And Pcr Methods On Water Samples
None of the 12 suspected colonies grown on GVPC with a negative reaction in gram staining when cultured on blood agar and BCYE without L-cystein showed any growth. Identi cation of these suspected colonies were performed by biochemical test such as positive oxidate test and weak reaction in catalase test. Additional molecular con rmation was performed by PCR method with speci c primers for Legionella genus (16srRNA) and pneumophila species (mip). All of the isolates were positive as Legionella pneumophila. Using culture and PCR methods for isolation and detection of Legionella in water samples shown that 12 and 42% of the hospital water samples were colonized by Legionella, respectively. All of the samples that were positive in culture methods (12%) were also positive by PCR. Legionella was isolated with a rate of 8%, 3% and 1% from shower heads; oxygen humidi er bottle and water bath respectively (Fig. 2). Fifty (5%) samples from 12 with positive culture have a colony forming unit higher than 1000 cfu/100 ml and the rest had values. A correlation was found between Legionella culture positivity rate and temperature of water samples in analysis by chi-square and likehood test (p value = 0.000, r = 0.493). No signi cant correlation was found between residual chlorine of water samples and the presence of Legionella with chi-square and Fisher test (p value = 0.313). A correlation was detected between the presence of Legionella and pH (p = 0.000; r = 0.546). Since sample collection continued for a period of one year, the isolation rate in summer and spring was 10 cases of 12 ( 83.33% ) as compared to winter and autumn which was 2 cases (16.66%).

Detection of virulence factors in the Legionella pneumophila isolates
To estimate whether Legionella isolates are pathogenic for human the presence of virulence genes including mip, dot, hsp, rtx and lvh were detected among the isolates. Findings showed that 12 (100%) isolates were positive for mip genes, 9 (75%) were positive for dot gene, 8 (66.66%) were positive for hsp, 6(50%) were positive for lvh and 4(33.33%) for rtx (Fig. 3, Fig. 4). Twelve cases showed eight virulence patterns that were reported in Table 2 and Fig. 5. All of the isolates had at least two of these virulence factors.

Bio lm Formation Assay Among Isolates
The ability of Legionella pneumophila for bio lm formation were estimated and results revealed that two isolates in rst day have a higher ability to form bio lm in reference to the standard strain and this ability increased to eight and ten isolates compare to standard strain in third and ninth days (Table 3).

Discussion
Water supplies in hospitals are the major infectious sources for Legionella. In many nosocomial outbreaks of Legionella, water supplies were the most frequent culprit [22]. Elderly individual, immunocompromised as well as patients having undergone surgery are more at risk and nosocomial Legionnar's disease showed a higher fatality than community ones [1]. For these individuals, aerosols generated from water supplies containing Legionella pneumophila can be a source of Legionella infection. Previous studies described that the isolation of Legionella SPP from shower heads, oxygen humidi er bottles and water baths are highly prevalent [5]. In the present study using culture method indicated that [12] 12% of the water sources including shower heads, water baths and oxygen humidi er bottle of nine hospitals were colonized with Legionella SPP. We could isolate Legionella with a rate of 8%, 3% and 1% from shower heads, oxygen humidi er bottle and water bath, respectively. The result of colony count of these 12 samples showed only 5 samples had a critical concentration of Legionella (more than1000cfu/100 ml) and rest had a moderate risk of Legionella concentration (lower than 1000 cfu/100 ml). Factors such as the present of viable but non culturable bacteria, loss of viability of bacteria after collection and low concentration of Legionella in the samples may in uence on unsuccessful isolation of Legionella in water samples [23]. The PCR method with 16SrRNA gene target of Legionella showed a rapid and sensitive test. Twelve samples in a total of 100 gave positive results for Legionella using the culture methods; whereas, 42 samples were positive when PCR was used. All samples that were found with the culture method were also positive with the PCR assay. Using primer speci c for mip gene con rmed that all the isolates in culture (100%) were Legionella pneumophila. There was a signi cant difference between PCR and culture results for detection of Legionella. Several studies described PCR to have a higher rate of detection than culture methods [24][25][26]. Molecular methods were suggested to be fundamental when an outbreak occurs. Molecular methods have several advantages such as rapidity, precision and are useful for analysis of samples usually contaminated with microorganisms other than Legionella. However, isolated DNA may not come from live microbial cells [27,28]. The physicochemical composition of water samples can affect the colonization and proliferation of Legionella in hospital water supply. The residual chlorine plays a very important role in clearance of the bacterium. The residual chlorine in our Legionella positive samples were between 0.3 or 0.6 mg/L. Several previous researches reported that there was no detection of Legionella SPP when residual chlorine was over 0.4 mg/L [29,30]. Concerning the physicochemical analysis of water samples in this study, we found Legionella pneumophila isolated from chlorinated water samples. Prior studies showed that Legionella has reduced viability at higher pH [31]. In this study, there was a signi cant association between Legionella isolation and pH of the water samples. The mean temperature of the samples positive for Legionella was lower (25-30ºC) compared to the mean temperature for those negative for Legionella pneumophila (31-41ºC). This result was consistent with a research in Ohio which showed that water temperature had an association with Legionella colonization [27]. However, is not consistent with a research in Italy that showed presence of Legionella was not affected by water temperature [32]. Sample collection in this study was carried for a full one year. All positive samples were collected in autumn, spring and summer with a high rate in summer (10%) and autumn (2%). Legionella not detected during winter. The main reason was high temperature and humidity during the summer months which is optimum for Legionella growth. Additionally, free living amoebae were also abundant mainly during autumn [33]. The existence of virulence genes in isolates from water sources is important to determine whether the isolates can have pathogenic potentials. When the virulence genes were studied, every isolate appeared to have different virulence pattern. The isolates Lp1, Lp3, Lp4 and Lp8 most the effective bacteria in 12 of experimented followed by the Lp12. Many studies explained the relationship between putative virulence genes and pathogenesis [1]. In this study, some isolates had least virulence genes.
Hung et al reported lvh and rtx genes as pathogenesis markers that can be used for determining the infection potential of an isolate [12]. In our study (33.333%) of the isolates showed a positive result for both of these genes. Previous studies revealed that rtx to be involved in the attachment and entry to Legionella pneumophila into acanthamobea, human epithelial and monocytes cells and pore formation in host membrane [34]. Hsp60 of L. pneumophila through a mechanism that involves surface interaction modulated macrophage function [1]. Legionella attach on surface and secret polysaccharide products for bio lm formation. The potential colonization of Legionella isolates could be analysed by the ability to produce bio lm on a polyethylene surface. Bio lm is an ideal microenvironment for proliferation, survival, dissemination and the likely the pathogenesis of L. pneumophila. Because Legionella in bio lm can become highly resistant to biocides, pH, excess temperature and chemical factors, outbreaks have been attributed to bio lm. In the present study, (71.35%) of the isolates had a strong potential for bio lm formation. This ability appeared in the rst day of the assay and increased on third and ninth days.

Conclusion
Findings of this project showed that although the presence of Legionella pneumophila does not necessarily equate a threat to public health but continuous monitoring and checking of the water sources should be performed in order to avoid high concentration of this bacterium and visible bio lm formation.

Materials And Methods
Collection and measurement of temperature, pH and chlorine of water samples In this project, during a one year period between September 2018 and September 2019, a total of 100 water samples were taken from different hospital water supplies in Tehran, Iran (Table 1).
Five hundred millimeter or fty millimeters of water samples were collected in sterile bottles or small polypropylene containers respectively. Water temperature, chlorine and pH were measured at time of sample collection and then 0.1 M sodium thiosulphate was added to neutralize residual free chlorine in the samples. All samples were transported to the Microbiology laboratory of Iran University without delay.

Pcr Methods For Determination Of Virulence Genes
The virulence factor genes (rtx, dot, hsp, mip and lvh) were assessed by ampli cation with speci c primers ( Table 2) for all legionella isolates. PCR reactions were prepared in 25 µl consisting of 2.5 µl of 2X master mix, 0.2 Mm of each primers and 2 µl of DNA template. Initial denaturation of legionella chromosome DNA was carried out at 95ºC for 5 min then 35 cycles of PCR ampli cation followed with 95ºC for 30 s, 58ºC for 1 min and 70ºC for 90 s with a nal extension at 70ºC for 5 min. PCR products were analyzed on 1.5% agarose gel containing 0.5 µg/ml safe stain. Gels were visualized in a gel document device.

Determination of bio lm formation in Legionella pneumophila isolates
The legionella bio lm formation was performed based on Hinder et al protocol [35]. The legionella isolates freshly grown on BCYE agar and suspended to an OD of 0.3 at 600 nm. 200 µl of the legionella suspension were added to 96 wells microplates. After attachment of Legionella to the polysterene wells, 300 µl BYE broth was added for extended bio lm. The well contents were collected and washed gently three times. The wells were lled with 1% crystal violet at 15 min. After washing steps, the dye which had been absorbed by bio lm was revealed using ethanol: acetone (80:20) incubation for 15 min. Solutions were subjected to another microplate and quanti ed by spectrophotometer at 595 nm. The results were interpreted basis on Stepanovic et al. [36]  lvh1/lvhB3 ATTGGGAGCTTCTGGCAATA 186 [38] lvh2/lvhB3 GCTGGGGTGACCTTTGAATA 186 [38] rtx1/rtxA GCTGCAACCACCTCTTTGAT 181 [38] rtx2/rtxA CAGGGGCTGGTTATGTTGAT 181 [38] dot1/dotA CAAATCCGGCATTCAAAATC 174 [38] dot2/dotA CTATTGTCGCCTTGGGTGTT 174 [38] hsp1/hsp60 GCGAATCGTTGTTACCAAAGAAAAC 401 [39] hsp2/hsp60 CAATTTGACGCATTGGAGATTCAATAG 401 [39] mip1/mip GGTGACTGCGGCTGTTATGG 632 [39] mip2/mip GGCCAATAGGTCCGCCAACG 632 [39] Statistical analysis All data collected from each assay and analysis on hospital water samples were imported into an Exel le (Microsoft o ce, Microsoft Excel). Statistical analysis was performed using IBM SPP statistics software (v.22). Quantitative variables are expressed either mean values with standard deviation. Chi-square, Fisher's exact and likehood ratio tests were used to identify any association between each parameter and presence of Legionella in water by culture. A nding with a p value of less than 0.05 was considered to be statistically signi cant.
Abbreviations PLA peptidoglycans associated with lipoprotein GVPC Glcine, Vancomycin, Polymyxin and Cycloheximide MOMP major outer membrane protein Hsp60 heat shock protein Declarations Figure 1 Results of physicochemical assay (pH, Tempreture and residual chlorine) on water samples.   Electrophoresis of PCR product of virulence genes (mip, dot,hsp,rtx, lvh).

Figure 4
Frequency of different genes in isolated Legionella pneumophila isolates Electrophoresis of PCR products of different genes in isolated Legionella strains (16srRNA, mip, rtx, lvh, hsp and dot) on gele agarose.