Molecular diversity and high virulence of Legionella pneumophilastrains isolated from biofilms developed within a warm spring of a thermal spa

Phenotypic analyses and serotyping of environmental L. pneumophilaisolates

DNA analysis and molecular diversity of environmental L. pneumophilastrains

Molecular typing of the all environmental isolates allowed us to confirm the classification obtained by serotyping (Table 1). Actually, we used current standards in molecular diagnosis of the genus Legionella: mip gene (“Macrophage infectivity potentiator”), 16S rRNA genes [17]. Both genes were amplified by PCR from bacterial lysates of the 30 environmental isolates. Then, the discrimination of the specium pneumophila was performed by amplifying the gene lpg0774[18]. Finally, Lp1 typing of seven environmental Legionellae was obtained by independent gene amplifications of lpg1905 and wzm (a gene belonging to the cluster coding for the lipopolysaccharide biosynthesis) [11, 18]: LAXA21, LAXB6, LAXB8, LAXB12, LAXB22, LAB24 and LAXB25 (Table 1; Figure 1).

Thus, the LAXB environmental strains we isolated from the spring S mainly belong to Lp12 (15 isolates) and to a lesser extend to Lp1 (6 isolates) and Lp10 (3 isolates); it is interesting to underline that the isolate LAXB11 was classified as L. pneumophila only at the molecular level, and not by serotyping which could suggests a new serogroup. With regard to the LAXA strains, Lp10 (2 isolates) and Lp1 (1 isolate) were also identified, but Lp12 was not detected. Two isolates, LAXA53 and LAXA54, were classified as non Legionella species and were indeed further identified as Mycobacterium isolates on the basis of their 16S rRNA sequences using a different set of 16S rRNA primers (data not shown). The small number of Lp isolated in the LAXA campaign does not allow to draw any conclusion about the persistence of Lp between August and December 2010.

In order to assess the molecular diversity, DNA of 26 LAXA and LAXB strains (7 Lp1, 5 Lp10 and 14 Lp12; LAXB10 strain did not grow anymore after a long term freezing period) was analyzed by PFGE and led to the identification of five main patterns (PST1 to PST5). It is clear that these five patterns are different from those of other known L. pneumophila clinical isolates as Lp1 strains Lorraine, Biarritz and Paris (see Additional file 1; Figure 2) but also Lp1 Lens, Philadelphia and Corby (data not shown). It is interesting to stress that Lp10 and Lp12 strains were grouped in two independent specific patterns (PST4 and PST3, respectively). By contrast, the 7 Lp1 strains were differentiated in three distinct patterns (PST1, PST2 and PST5) which suggest a genomic diversity.

In order to assess more finely this molecular diversity, the mip sequences of 27 L. pneumophila strains were determined and compared. All mip sequences were performed on both strands and no mismatch was identified. The 27 sequences comparison the led us to identify three different types of mip sequence, so-called mip1, mip2 and mip3. These sequences exhibit a high identity (> 99%) and only differ by few substitutions (see Additional file 2): 5 substitutions between mip1 and mip2 sequences, 4 between mip1 and mip3 and a unique substitution between mip2 and mip3. It must also be underlined that these three mip sequences are very close to those of known clinical isolates (identity > 99, 6%), and the mip3 sequence is even completely identical to the mip sequence of the Lp1 clinical strain Corby (see Additional file 2). Actually, this sequence-based classification not only confirmed results obtained with other typing approaches (serotyping and molecular typing) but also allowed us to position the different environmental strains within the specium pneumophila (Table 2; Figure 3). Analyses of mip sequences confirmed the homogeneity of Lp12 strains belonging to the unique pulsotype PST3 and characterized by a unique mip sequence (mip2) (Table 2; Figure 2). Besides, this approach revealed a genetic diversity within the five Lp10 strains belonging to the pulsotype PST3 but differentiated by two mip sequences, mip2 and mip3. Finally, a high genetic diversity was also observed within PST1 and PST2 pulsotypes, where the environmental Lp1 strains could be discriminated according to the three mip sequences (Table 2).

Class	Sg	PST	mip	Environmental isolates	Isolate number
1	sg1	PST1	mip1	LAXB8, LAXB12	2	7 Lp1
2	sg1	PST1	mip2	LAXB6	1
3	sg1	PST2	mip2	LAXA21	1
4	sg1	PST2	mip3	LAXB24, LAXB25	2
5	sg1	PST5	mip3	LAXB22	1
6	sg10	PST4	mip2	LAXA22, LAXA23	2	5 Lp10
7	sg10	PST4	mip3	LAXB1, LAXB3, LAXB20	3
8	sg12	PST3	mip2	LAXB2, LAXB4, LAXB5, LAXB7, LAXB9, LAXB13, LAXB14, LAXB15, LAXB16, LAXB17, LAXB18, LAXB19, LAXB21, LAXB23, LAXB10*	15	15 Lp12
					27	27

*LAXB10 was positioned into the class 8 according to serotyping and mip sequence.

Cytotoxicity to Acanthamoeba castellani

Alamar blue was used to quantify the viability of remaining amoeabae after Legionella infection (Figure 4a) and the cytotoxicity was assessed by percent of killed amoebae. The mean cytotoxicity of Lp1 clinical strains (Lens, Paris and Lorraine) was estimated to 40 and 73% after 24 h and 48 h post-infection, respectively. As expected, the avirulent mutant dotA, derived from the strain Lens [19] did not display any significant cytotoxicity (0 and 4% at 24 h and 48 h, respectively). Environmental strains isolated from the source S appeared much more cytotoxic than Lp1 clinical strains, especially at 48 post-infection: actually environmental Lp1, Lp10 and Lp12 are characterized by a cytotoxicity of 100% whatever their pulsotype (PST1, PST2 and PST5) or their mip sequence (mip1, mip2 or mip3) (Figures 4a and 4b).

Virulence towards Acanthamoeba castellani

Lp1 clinical strains involved in LD outbreaks (Lens, Lorraine) and the worldwide epidemic and endemic strain Paris were used as virulent references. 1 × 10⁵ and 4, 5 × 10⁵ extracellular clinical Lp1 cells were present in 3 μl samples taken after a 24 h and 48 h period of A. castellanii infection, respectively (Figure 4c). In the same periods, legionella cells released from amoeba cells infected with the dotA mutant were 100-fold less numerous. Interestingly, the number of extracellular Legionellae cells resulting from amoeba infections with environmental strains was very close to that of clinical Lp1 with the exception of extracellular Lp12 strains associated with a 10-fold increase after a 48 h-period of infection. No significant difference of virulence was observed between the different classes of environmental Lp1 at 48 h post-infection, even if some strains appeared to present a weak delay of virulence at 24 h post-infection (Figure 4d).

A co-infection experiment was also conducted in A. castellanii with two representative strains of Lp1 (LAXB24) and Lp12 (LAXB2) environmental isolates. Duplex PCR analysis (using wzm and lpg1905 primers) of extracellular bacteria revealed that 95% of 40 clones analyzed belonged to Lp12 strain (LAXB2), indicating the rapid and advantageous development of this Lp12 strain in competition to the Lp1 strain.

Environmental isolates

Glass slides were dipped into the contaminated spring S of a French Alpine thermal spa. After 15 days of incubation, the glass slides were covered with natural biofilms. These biofilms were harvested by scraping the glass slides and resuspended in 5 mL sterile water. Then, these suspensions were submitted to ultrasounds during 1 min in order to break up the aggregates formed by biofilms and to release bacterial cells. Bacterial suspensions were treated at 50°C during 30 min, and then submitted to an acidic treatment during 5 min by addition of 200 mM KCL/HCl pH 2.0. Aliquots (100 μL) were spread on agar GVPC medium (Oxoid, France) containing L-cysteine, iron pyrophosphate, ACES, charcoal and antibiotics (polymixin B, vancomycin, cicloheximide). After a 5 day-period incubation at 37°C, bacterial colonies with a fritted glass appearance were picked up and isolated again on GVPC. New independent colonies were picked up and suspended in cryotubes containing beads and bacterial preservers for storing at −20°C.

The Acanthamoeba castellani strain is an environmental isolate provided by F. Pernin (Institut des Sciences Pharmaceutiques et Biologiques - Faculté de pharmacie – Université Lyon 1, Lyon, France).

Reference bacterial strains

Reference strains obtained from the National Centre of Legionella (Bron, France) were used as controls in different assays: L. pneumophila serogroup 1 (Lens, Paris, Lorraine), L. pneumophila ATCC 35096 (sg 8) and ATCC 33155 (sg 3), L. anisa G12108, L. longbeachae ATCC 35096, L. micdadei ATCC 33218 and L. taurinensis ATCC 700508. The dotA mutant is derived from the strain Lens and shows a severe defect of virulence and cytotoxicity [19]. Routinely, Legionellae were grown on buffered charcoal yeast extract (BCYE) agar (Oxoid, France) or in BYE liquid medium. E. coli DH5α was cultivated on Lysogeny Broth (LB) agar medium at 37°C and Lactococcus lactis subsp. lactis IL1403 was grown at 30°C on M17 agar medium [24].

Serotyping of Legionellae

Legionella isolates were identified by polyclonal antisera coupled to latex-beads. Firstly, the Legionella latex test from Oxoid (DR0800M) allowed a separate identification of Legionella pneumophila serogroup 1 and serogroups 2–14, and the identification of seven non-pneumophila species: L. longbeachae 1 and 2, L. bozemanii 1 and 2, L. dumoffii, L. gormanii, L. jordanis, L. micdadei and L. anisa. Secondly, the 15 monovalent latex reagents prepared by bioMérieux allow the separate identification of 15 serogroups of L. pneumophila (bioMérieux, Craponne, France) [25].

In situassay of catalase activity

The presence of bacterial catalase activity was detected using H₂O₂ as the substrate. A bacterial colony was picked up with a sterile loop and diluted into a 15 μL drop of 10% (vol:vol) H₂O₂, loaded on an empty Petri dish. The rapid formation (in a few seconds) of oxygen bubbles indicates a positive result. E. coli DH5α was used as the positive control (Cat⁺) and Lactococcus lactis IL1403 as the negative one (Cat^-).

Molecular identification and DNA amplification by PCR

DNA amplification was performed with the 2 × PCR Master Mix DNAzyme II (Finnzymes) containing 0.04 U/μL DNAzyme™ II DNA polymerase, 400 μM of each dNTP, 3 mM MgCl₂, 100 mM KCl and 20 mM Tris–HCl pH 8.8 (and stabilizers). The PCR mixture (25 μL) contained the 2 × PCR Master Mix DNAzyme II (12.5 μL), 10 mM forward and reverse appropriate primers (1.0 μL each) (Table 1), and the bacterial lysate (8.0 μL). Amplification of DNA was performed in a Ep-gradient Mastercycler (Eppendorf) at initial denaturation of 94°C for 2 min, followed by 35 cycles of 94°C for 1 min, 55°C for 1 min and 72°C for 1 min with a final extension at 72°C for 8 min. Reactions mixtures were then held at 10°C. 8 μL of the PCR amplification mixture was analyzed by gel electrophoresis in a 0.8% agarose gel stained with ethidium bromide (1.0 μg/mL) and photographed under U.V. transillumination.

Purification and sequencing of PCR mipproducts

PCR mip products were analyzed by gel electrophoresis in a 0.8% agarose gel (50 mL) stained with 3 μL SYBR Safe DNA gel strain (Invitrogen). DNA products were visualized under blue U.V. transillumination and picked up with a band of agarose gel. Then PCR products were purified using GeneCleanR Turbo Kit (MP Biomedicals) according to the manufacturer’s instructions. Finally, the purified PCR products were suspended in 10 μL sterile water and then stored at −20°C. Sequencing was performed by GATC Biotech SARL (Mulhouse, France).

PFGE subtyping

Legionella isolates were subtyped by pulsed field gel electrophoresis (PFGE) method as described previously [26]. Briefly, legionellae were treated with proteinase K (50 mg/mL) in TE buffer (10 mM Tris–HCl and 1 mM EDTA, pH 8) for 24 h at 55°C, and DNA was digested with 20 IU of SfiI restriction enzyme (Boehringer Mannheim, Meylan, France) for 16 h at 50°C. Fragments of DNA were separated in a 0.8% agarose gel prepared and run in 0.5× Tris-borate-EDTA buffer (pH 8.3) in a contour-clamped homogeneous field apparatus (CHEF DRII system; Bio-Rad, Ivry sur Seine, France) with a constant voltage of 150 V. Runs were carried out with increasing pulse times (2 to 25 s) at 10°C for 11 h and increasing pulse times (35 to 60 s) at 10°C for 9 h.

Then, the gels were stained for 30 min with a ethidium bomide solution and PFGE patterns were analyzed with GelComparII software (Applied Maths, Saint-Martens-Latem, Belgium).

Quantification of Legionella virulence towards the amoeba Acanthamoeba castellanii

Legionellae were grown on BCYE agar and A. castellanii cells in PYG medium (Moffat and Tompkins, 1992) for five days at 30°C prior to infection. A. castellanii cells were first seeded in plates of 24 multiwell to a final concentration of 5 × 10⁶ cells per ml in PY medium (PYG without glucose. Plates were incubated during two hours at 30°C to allow amoeba adhesion. Then, Legionellae were added to an MOI (“multiplicity of infection”) of 5 (in duplicate). In order to induce the adhesion of bacterial cells to the monolayer of amoeba cells, plates were spun at 2000 × g for 10 min and incubated for 1 h at 30°C. Non-adherent bacteria were removed by four successive washings of PY medium. This point was considered as the initial point of infection (T0) and the plates were incubated at 30°C. Extracellular cultivable bacteria released from amoebae were quantified at 1 day and 2 days post-infection as follows. Aliquots (100 μL) of the supernatants were taken and diluted in sterile water to the final 10^-6 dilution. Aliquots (3 μL) of the serial dilutions (10^-1 to 10^-6) were immediately spotted to the surface of agar BCYE plates. Independent bacterial colonies of serial dilutions were numbered after 5 days at 30°C.

In the co-infection experiment, the same cells amount of each strain was added to achieve a final MOI of 5. Extracellular bacteria (10^-5 and 10^-6 dilutions) were plated on BCYE agar, 48 h post-infection. Independent bacterial colonies were picked-up after 3 days at 30°C to perform a PCR analysis.

Cytotoxicity to Acanthamoeba castellani

To quantify the viable A. castellanii cells remaining after infection with Legionellae (MOI 5), a monolayer of amoebae cells at the final concentration of 1 × 10⁶ cells per ml in a 96 multiwell plate was washed (fourfold) with PY and then treated with 10% Alamar blue (Invitrogen). Cytotoxicity of each Legionella strain was tested in triplicate. After an overnight incubation at 30°C, measurements were performed at the optical density (OD) of 570 nm and corrected for background at OD_{600 nm} with a μQuant microplate reader (Biotek Instruments Inc., Winooski, USA) The relative degree of amoeba mortality corresponds to the cytotoxicity and was expressed as the ratio of the OD value of infected monolayer to that of the uninfected one as following: [1-(mean OD value of infected/mean OD value of uninfected)] × 100%.