Identification of two aptamers binding to Legionella pneumophila with high affinity and specificity

Legionella pneumophila (Lp) is a water borne bacterium causing Legionnaires’ Disease (LD) in humans. Rapid detection of Lp in water system is essential to reduce the risk of LD outbreaks. The methods currently available require expert skills and are time intensive, thus delaying intervention. In situ detection of Lp by biosensor would allow rapid implementation of control strategies. To this end, a biorecognition element is required. Aptamers are considered promising biorecognition molecules for biosensing. Aptamers are short oligonucleotide sequence folding into a specific structure and are able to bind to specific molecules. Currently, no aptamer and thus no aptamer-based technology exists for the detection of Lp. In this study, Systemic Evolution of Ligands through EXponential enrichment (SELEX) was used to identify aptamers binding specifically to Lp. Ten rounds of positive selection and two rounds of counter-selection against two Pseudomonas species were performed. Two aptamers binding strongly to Lp were identified with KD of 116 and 135 nM. Binding specificity of these two aptamers to Lp was confirmed by flow cytometry and fluorescence microscopy. Therefore, these two aptamers are promising biorecognition molecules for the detection of Lp in water systems.

the application of corrective measures and increases the chances of an outbreak 15 . A pilot study performed in 2011 evaluated the consistency of the results obtained by this method between several different laboratories. Qualitative results did not differ drastically between laboratories, but quantitative results showed large variation, within and between laboratories 16 . Therefore, the culture method should be used with caution to precisely enumerate Lp. A second major limitation is the presence of viable but non-culturable (VBNC) Lp cells which leads to an underestimation of the true amount of infectious Lp in a system 17,18 . The qPCR method relies on the quantification of Legionella DNA. Its major advantages in comparison with conventional culture method is the rapid turn-around time, high sensitivity and specificity, low limit of detection, as well as the ability to detect VBNC cells. When used in conjunction with the culture method, qPCR can serve as a powerful tool. There are, however, several drawbacks: qPCR typically overestimates Lp burden because it detects dead cells and the presence of PCR inhibitors may limit the use of this method 19,20 . In addition, multiple processing steps are required which increases the overall cost of the qPCR method 21 . Unfortunately, it is impossible to develop these two methods into rapid, cost-effective, sensitive tests that would detect Lp in real-time, on-site, without any additional processing steps 22,23 .
Biosensors are attractive detection technology that could address the problems associated with culture-based bacterial detection methods. These analytical devices are commonly used to assess and quantify in real-time, with high sensitivity, the presence of an analyte such as a protein, peptide or cell in a fluid 24 . However, a biosensing approach to Lp detection would require a specific biorecognition element, which, when coupled with a transducer, translates its interaction with Lp cells into a meaningful readout 24 .
Various biorecognition elements, such as antibodies, lectins or aptamers, can be used. The latter are becoming the primary choice for biosensing strategies due to their easily modifiable nature and versatility 25,26 . Aptamers are antibody analogues. They are short single stranded DNA or RNA oligonucleotides that can be cost effectively synthesized in a high throughput manner. The aptamer folds into a specific, stable structure and can interact with its targets via shape complementarity, hydrogen bonding, electrostatic interactions and stacking interactions 27 . This allows aptamers to bind with high affinity and specificity to a wide variety of targets ranging from small molecules, peptides, proteins to whole cells 27 . A key characteristic of aptamers is the possibility to generate them in vitro in the same condition as those used for detecting the analyte. This is a clear advantage over antibodies which are produced under strict physiological conditions 28 . In addition, aptamers can be easily modified and therefore be optimized for various sensing platforms such as lateral flow assays, surface plasmon resonance sensors, flow cytometry or fluorescence microscopy 29 .
The procedure by which an aptamer is created is known as Systemic Evolution of Ligands through EXponential enrichment (SELEX). Developed in 1990 by the teams of Gold and Szostak 30,31 , SELEX is an iterative process which involves incubating a target with a large library of oligonucleotides, separating the target bound and unbound oligonucleotides and then amplifying the target bound sequences via PCR for the next round of selection. The selection rounds are repeated until the oligonucleotide pool is enriched with sequences that bind specifically and with high affinity to the target 30 . Over the years, many variations of the original SELEX methods were published. Among those, one is particularly useful for the present study. Cell-SELEX can be used to select aptamers binding to whole living cells and, thus, eliminate the need for prior knowledge of a target molecule 32,33 . Rounds of counter-selection are typically used to reduce aptamer cross-reactivity across targets by eliminating non-specific aptamers 29 . Cell-SELEX has been successfully employed to isolate aptamers against various bacterial species such as E. coli, Salmonella typhimurium, Campylobacter jejuni, Listeria monocytogenes, Staphylococcus aureus and Vibrio parahaemolyticus 28,[33][34][35][36][37][38] . Several of these aptamers are used in conjunction with optical, mechanical or electrical/electrochemical biosensors to mitigate the problems associated with traditional bacterial detection methods. Although numerous works have been done to detect Lp with the use of biosensors [39][40][41] no aptamers binding to Lp have been reported yet. Consequently, no aptamer and thus no aptamer-based technology currently exists for the detection of Lp.
In this work, the cell-SELEX procedure was employed to generate aptamers binding to Lp. Two Pseudomonas species were used for counter-selection to improve the specificity of the aptamers. These species were chosen because they are γ-proteobacteria like Lp and are routinely found in premise plumbing and water systems where Lp is prevalent 42,43 . Two aptamers were identified and their binding affinity and specificity for Lp were evaluated by flow cytometry and fluorescence microscopy.

Materials and Methods
Bacterial strains and culture conditions. The environmental Lp strain lp120292, isolated from a cooling tower implicated in the 2012 outbreak in Quebec City, was used as the target strain for aptamer generation 44 . The strain Lp*GFP is lp120292 transformed with plasmid pXDC31 expressing the green fluorescent protein (GFP) under the Ptac promoter 45 . The thymidine auxotroph Lp strain Lp02, derived from Lp Philadelphia-1 was used to confirm binding of the aptamers 46 . Lp was cultured on CYE (ACES-buffered charcoal yeast extract) agar plates supplemented with 0.25 mg/ml L-cysteine and 0.4 mg/ml ferric pyrophosphate, at 37 °C for 3 days. Lp*GFP strain was grown on CYE media supplemented with 5 µg/ml chloramphenicol and 1 mM isopropyl β-D-1-thiogalactopyranoside. For liquid culture, Lp was suspended in AYE (ACES-buffered yeast extract) broth supplemented with 0.25 mg/ml L-cysteine and 0.4 mg/ml ferric pyrophosphate until post-exponential phase (OD 600 of 2.5). Pseudomonas putida KT2440 and Pseudomonas fluorescens LMG1794, were first cultured on LB agar plates at 30 °C for 24 hours and then grown in LB medium (Difco) until the cultures reached post-exponential phase (OD 600 of 2.0). Pseudomonas sp., Brevundiomonas sp., Bacillus sp., Staphylococcus sp. Sphingomonas sp., Stenotrophomonas sp. and Cupriavidus sp. were isolated from cooling towers as part of a different study (Paranjape, in preparation). Briefly, the bacteria were isolated on nutrient agar incubated at 30 °C for 24 h. Isolates were then classified using 16 S rDNA sequencing. These strains were first cultured on nutrient agar plates (Difco) www.nature.com/scientificreports www.nature.com/scientificreports/ at 30 °C for 24 hours and then grown in nutrient broth medium (Difco) overnight until the cultures reached post-exponential phase (OD 600 of 2.0-2.5).

Oligonucleotides library and primers.
A random ssDNA library of 10 15 sequences was chemically synthesized and purified by HPLC (Integrated DNA Technologies). The library consisted of a central random region of 45 nucleotides flanked by two different primer binding regions at the 5′ and 3′ ends: 5′-GCAATGGTACGGTACTTCC-45N-CAAAAGTGCACGCTACTTTGCTAA-3′. The forward and reverse primers were conjugated with fluorescein (FITC) and biotin, respectively. The forward primer (FP) sequence is 5′-fluorescein-GCAATGGTACGGTACTTCC-3′. The reverse primer (RP) sequence is 5′-biotin-TTAGCAAAGTAGCGTGCACTTTTG-3′ 28 . FITC was used to quantify ssDNA and monitor the SELEX procedure via flow cytometry. The biotin was used in conjunction with streptavidin coated magnetic beads (Promega) to generate ssDNA from the amplified double-stranded aptamer pool following PCR. For cloning and sequencing, PCR was performed with unmodified versions of the primers.
Bacterial cell-SELEX procedure. Cell-SELEX was performed as previously described 28 . Each round of SELEX consisted of three steps: Binding and elution, amplification, and recovery of ssDNA. Ten rounds of positive selection and two rounds of counter selection were performed (Fig. 1). The first round of counter selection was performed with P. putida KT2440; the second one with P. fluorescens LMG1794.
Binding and Elution. Cell-SELEX was performed with cells suspended in an artificial freshwater medium (Fraquil) to replicate the physiological state of nutrient-limited environmental conditions 47,48 . Fraquil was prepared as described previously 49 with a final iron concentration of 10 nM and filter-sterilized using a 0.2 µm filter (Sarstedt). Post-exponential phase cultures were rinsed twice with Fraquil (6000 g, 15 minutes) and suspended in Fraquil at an OD 600 of 1 corresponding to a concentration of 10 9 CFU/ml. The concentration of cells was confirmed by CFU counts for each round. The suspension was incubated at room temperature for 24 h. Fraquil exposed cells were washed three times in 1X binding buffer (phosphate buffered saline with 0.1 mg/ml salmon sperm DNA, 1% bovine serum albumin, and 0.05% Tween 20) at room temperature (25 °C) using 6,000 g for ten minutes. The cell pellets were then suspended in 330 µl of 1X binding buffer. The aptamer pool was denatured by heating at 95 °C for 10 minutes, cooled immediately on ice for 10 minutes, and added to the cell suspension. Finally, 1X binding buffer was added to a total volume of 1 ml. For the first round, 32 µg of the initial library was used. For the subsequent rounds, approximately 400 ng of aptamer pool was used. The final mixture was incubated at 25 °C for 1 hour with mild shaking using a tube rotator at 150 rpm. Following incubation, the mixture was centrifuged at 6000 g for 10 minutes and washed twice with wash buffer (phosphate buffered saline containing 0.05% Tween 20) to remove unbound sequences. To elute the bound sequences from the cells, the final cell pellet was resuspended in 100 µl nuclease free water (Ambion) and heated at 95 °C for 10 minutes and immediately placed on ice for 10 minutes. After centrifuging at 6,000 g for 10 minutes at 25 °C, the supernatant was collected and purified using overnight ethanol precipitation at −20 °C with 5 µg of glycogen as a carrier to recover the eluted ssDNA. The pellet was recovered, dried and suspended in nuclease free water (Ambion). The concentration and quality of the ssDNA was determined using a Nanodrop spectrophotometer (Thermofisher). www.nature.com/scientificreports www.nature.com/scientificreports/ For counter-selection the supernatant containing the unbound sequences was collected and purified via ethanol precipitation, as described above. To ensure there was no amplification or collection of unwanted bacterial DNA (instead of the desired amplification and collection of ssDNA oligonucleotides), a control sample consisting of bacterial cells without aptamer was included in each round.
PCR amplification. The purified aptamer pool was then amplified by PCR with One Taq DNA polymerase (NEB), according to the manufacturer's protocol. All primers were used at a final concentration of 0.5 µM. PCR conditions were as follows: initial heat activation at 95 °C for 5 min and 25 cycles of 95 °C for 30 s, 56.3 °C for 30 s, 72 °C for 10 s, and a final extension step of 10 min at 72 °C. After amplification, the concentration and size of the PCR product were confirmed by gel electrophoresis using a 2.0% agarose gel. PCR products were then purified using a MinElute PCR Purification Kit (Qiagen). As expected, no amplification was observed for the control samples, lacking aptamer template.
Recovery of ssDNA. Streptavidin coated magnetic beads (Promega Technology) were used, according to the manufacturer's recommendation. Briefly, 600 µg of magnetic beads were washed twice and then resuspended in 900 µl of washing buffer (phosphate buffered saline with 0.05% Tween 20). Next, approximately1 µg of PCR product was incubated with the magnetic beads for 10 min, mixing gently by inversion after every few minutes. The mixture was then washed in 1 ml of washing buffer. Finally, the beads were incubated with 500 µl of 200 mM NaOH for 5 minutes. The supernatant was then collected, and the FITC-labelled ssDNA was purified using ethanol precipitation as mentioned previously and quantified with a Nanodrop spectrophotometer (Thermofisher).
Monitoring of SELEX by flow cytometry. The binding of the FITC-labelled aptamer pools from rounds 1 (R1), 6 (R6), 7 (R7), 8 (R8) and 10 (R10) to Lp was assessed using flow cytometry. Briefly, 35 nM of aptamer pools from each of these rounds was incubated with 10 6 CFU/ml of Lp cells at 25 °C for 1 hour. Analysis was performed on a Guava easyCyte (Millipore) using the green fluorescence channel. A total of 5,000 events were recorded. Unlabeled cells were used as a control to measure autofluorescence. The Lp*GFP strain, producing strong green fluorescence from GFP, was used to adjust the gain of the green fluorescence channel. For analysis, a gate was first defined based on the forward and side scatters that included most of the cells. Then, a histogram of the number of cells vs the fluorescence intensity was used to define a region named Green_Lp where cells were considered positive for green fluorescence and therefore stained with aptamers. This region was setup to include very few cells of the unstained control and therefore represent fluorescence above the autofluorescence. Aptamer pools from R10 alone, without cells, was also analyzed to ensure that the aptamer alone was not forming aggregates that would be confused with cells.
Cloning and sequencing. To identify sequences binding to Lp, the aptamer pool from the 10 th round of SELEX was cloned with the pGEMT-easy Cloning and Ligation Kit (Promega). To investigate the effect of counter-selection on the aptamer pools, we also cloned and sequenced aptamers from the 6 th round of SELEX. Positive colonies, containing aptamer inserts, were determined via blue-white screening and confirmed by PCR. Plasmids were extracted and purified using a Miniprep Kit (Qiagen) and sequenced by Sanger Sequencing at the Plate-forme d' Analyse Génomique of Laval University. Secondary structures of the aptamer sequences were determined using the Mfold web server using default parameters 50 .

Characterization of aptamers R10C5 and R10C1. The binding of the aptamers R10C5 and R10C1 to
Lp and to the species used for counter-selection was further characterized. R10C5 and R10C1 were individually synthesized with FITC at the 5′ end (Integrated DNA Technology).

Determination of the disassociation constant (K D ).
To determine the K D of R10C5 and R10C1, varying concentrations of FITC-tagged aptamers (1000 nM, 100 nM, 10 nM, and 1 nM) were incubated with 10 6 CFU/ml of Lp cells suspended in Fraquil and the fluorescence obtained at each concentration was measured using flow cytometry, as described above, in triplicate. The number of bound cells (FITC-positive) were recorded and used to determine the K D by interpolating the logarithmic curve using GraphPad Prism 7.03.

Specificity assay.
To determine the specificity of R10C5 and R10C1 for Lp cells, the binding to counter-SELEX Pseudomonas strains as well as cooling tower isolates was tested using flow cytometry. All cells were suspended in Fraquil and prepared as described above for cell-SELEX. Briefly, 100 nM of R10C5 and R10C1 was incubated with 10 7 CFU/ml of the strain used for SELEX (lp120292), another Lp strain (Lp02), the strains used for counter-selection (P. putida KT2440 and P. fluorescens LMG1794), and the isolates from cooling towers (Pseudomonas sp., Brevundimonas sp., Bacillus sp., Staphlyococcus sp., Sphingomonas sp., Stenotrophomonas sp. and Cupriavidus sp.) for 1 hour at 25 °C with mild shaking. lp120292 was also incubated with 100 nM of a FITC-labeled scrambled sequence of aptamer R10C5 ( 5′-fl uo re sc ei n-AC AG AA TC AG TT CG AG TA CA TA CG CG CG AA GA CT CC TA AG GC CGTAGCGTTCTTCCCGGTAATACCATG) and R10C1(5′-fluorescein-TGTACTCCCGCGTCCCACCTGCTACCCGAAATAGAGTTTCCCTAGAAAGGCTTGCCCAAC). The suspension was centrifuged for 10 minutes at 6000 g to eliminate excess aptamer and resuspended in Fraquil. These suspensions were then analyzed using flow cytometry as described above. This experiment was done in triplicate. Cells suspended in Fraquil without any aptamer added were used as controls. The percentage of bound cells was determined as described above. Statistical differences were assessed using a one-way ANOVA and Dunnett correction for multiple comparison using GraphPad Prism 7.03. www.nature.com/scientificreports www.nature.com/scientificreports/ Confocal fluorescence microscopy assay. FITC-labelled R10C5 and R10C1 aptamer (100 nM) were incubated with 10 8 CFU/ml of target cell lp120292 or counter-selection strain P. fluorescens LMG1794 for 1 hour at 25 °C on a tube rotator at 150 rpm. Cells were suspended in Fraquil as mentioned previously. Negative controls included cells suspended in Fraquil without any aptamers. The suspensions (10 µl) were dropped on a glass slide (Fisherbrand), and a #1.5 cover slip (VWR) was used to make a thin layer. The slides with suspensions were then transferred onto a microscope chamber and imaged using a Zeiss LSM 710 confocal microscope (Carl Zeiss, Oberkochen, Germany), a 100 X oil objective (Plan-Apochromat 100×/1.40 Oil DIC M27a) and a 488 nm argon laser (25 mW,). A pre-set FITC filter was used with excitation and emission wavelengths of 488 and 564 nm respectively. Images were analyzed using Fiji 51 .

Results and Discussion
Selection of aptamers binding to Lp. Cell-SELEX was used to select aptamers binding specifically to lp120292. This strain was selected because it was involved in the Quebec City Outbreak in 2012 44 . To mimic the physiological state of Lp in a water system, Lp cells were grown to post-exponential phase and suspended in Fraquil for 24 h at 25 °C to induce starvation and the associated morphological and physiological changes 47,48 . Seven rounds of positive selection were performed, followed by one round of counter-selection, two rounds of positive selection, an additional round of counter-selection and a final round of positive selection (Fig. 1). Two Pseudomonas strains were used for counter-selection because they are also Gram-negative γ-proteobacteria frequently isolated from water systems where Lp is found as mentioned previously.
To monitor the progress of the SELEX procedure and ensure that the proportion of sequences binding to Lp was increasing, the binding of the FITC-labelled aptamer pools from rounds 1 (R1), 6 (R6), 7 (R7), 8 (R8) and 10 (R10) to Lp was examined using flow cytometry. Cells incubated with the initial aptamer library showed minimal fluorescence compared to the negative controls (Fig. 2d, Lp Lib). Cells incubated with aptamer from the first positive selection round showed a drastic increase in fluorescence (Fig. 2e, Lp R1). The saturation in the fluorescence intensity and the percentage of bound cells starting at R6 suggests that the pool is dominated by sequences binding to Lp (Fig. 2f-i, R6, R7, R8, R10). A small decrease in fluorescence and percentage of bound cells at R8 suggests that the first counter-selection step removed a few sequences. The fluorescence intensity remains similar between round 8 and 10 indicating that the second round of counter-selection did not remove Lp-specific sequences and that our strategy was successful in retrieving aptamers binding to Lp. www.nature.com/scientificreports www.nature.com/scientificreports/ Cloning and sequencing. Analyzing the sequences obtained from the 10 th round of positive selection allowed for identifying two different ssDNA aptamers, named R10C5 and R10C1 (Table 1 and Fig. 3). Of the 13 sequences that were retrieved, 12 of them were R10C5 whereas 1 was R10C1. In contrast, the survey of a non-exhaustive list of the sequences present in the R6 aptamer pool revealed eight different sequences out of 9 clones, but none similar to R10C1 and R10C5. This illustrates the directional evolution of the pool as a result of the additional positive selection rounds and counter-selection steps 52,53 . A strong bottleneck effect was likely caused by the last four positive selection rounds and the apparently stringent counter-selection rounds, which most likely led to the removal of several aptamers.
Determination of K D . The calculated K D is 116 nM for R10C5 and 135 nM for R10C1 (Fig. 4). These values are comparable to high affinity antibodies that typically show nanomolar ranges of K D for small protein targets 54 . These values are also comparable to values of published aptamers created against whole bacterial pathogens. For example, aptamers isolated against Escherichia coli, Enterobacter aerogenes, Klebsiella pneumoniae, Citrobacter  www.nature.com/scientificreports www.nature.com/scientificreports/   strains (a,b) as well as to environmental isolates (c,d) was analyzed by flow cytometry. The binding of R10C5 to Lp strain Lp02 was also analyzed. The percentage of cells bound by R10C5 (a,c) and R10C1 (b,d) to counter-SELEX strains (a,b) and environmental isolates (c,d) are presented. The binding of a scrambled sequence of aptamer R10C5 and a scrambled sequence of aptamer R10C1 to Lp strain lp120292 was also investigated (scrambled). The values of three experiments are shown with the mean and standard deviation. A one-way ANOVA with a Dunnett correction for multiple comparisons was used to infer statistical significance compared to Lp strain lp120292: ***P < 0.001; ns, not significant. Figure 5a shows the binding of R10C5 and R10C1 to the strains used for counter-selection. Around 60% of lp120292 cells are stained by R10C5, consistent with previous results shown in Fig. 2, but only 20% of Pseudomonas strains are labelled. Similarly, R10C1 shows significantly more binding to Lp than to Pseudomonas (Fig. 5b). Of note, the scrambled sequences of aptamer R10C5 and R10C1 bind minimally to lp120292 (Fig. 5a,b). The aptamer R10C5 stained Lp02 similarly to lp120292, suggesting that the binding of this aptamer is not restricted to a particular strain of Lp (Fig. 5a). Moreover, both aptamers show very low binding to environmental isolates from cooling tower water. For the majority of the isolates, less than 10% of cells were labeled by the aptamers (Fig. 5c,d). The specificity of these aptamers for Lp was further analyzed by confocal fluorescence microscopy. Both aptamers strongly stained Lp (Fig. 6) but not P. fluorescens LMG1794, one of the strains used for counter-selection. These results support the notion that the aptamers R10C5 and R10C1 are highly specific to Lp.

Specificity of R10C5 and R10C1.
In conclusion, our cell-SELEX strategy was successful in identifying two aptamers binding to Lp with high affinity (K D = 116 nM for R10C5 and 135 nM for R10C1). Whereas R10C5 seems to stain Lp more strongly then R10C1, the latter seems more specific to Lp, showing minimal binding to the counter SELEX Pseudomonas strain. Both aptamers showed minimal binding to cooling towers isolates, indicating that the aptamers are suitable to detect Lp in complex water samples. Modification of these aptamers could be attempted to further increase their affinity and specificity to Lp. Based on the results presented here, these aptamers are promising candidates as biorecognition elements to develop a biosensor to detect Lp in real time and in situ. Figure 6. The specificity of FITC-labelled R10C5 and R10C1 aptamer was tested by measuring their ability to bind to Lp strains lp120292 (a,c and e) and to P. fluorescens LMG1794 (b,d and f) by confocal fluorescence microscopy. The No aptamer controls consists of cells alone (a,b). Cells were incubated with aptamer R10C5 (c,d) and aptamer R10C1 (e,f).