A novel sensitive pathogen detection system based on Microbead Quantum Dot System

Author’s Accepted Manuscript A Novel Sensitive Pathogen Detection System Based on Microbead Quantum Dot System Tzong-Yuan Wu, Yi-Yu Su, Wei-Hsien Shu, Augustus T. Mercado, Shi-Kwun Wang, Ling-Yi Hsu, Yow-Fu Tsai, Chung-Yung Chen www.elsevier.com/locate/bios

PII: DOI: Reference:

S0956-5663(15)30569-8 http://dx.doi.org/10.1016/j.bios.2015.11.016 BIOS8146

To appear in: Biosensors and Bioelectronic Received date: 12 August 2015 Revised date: 4 November 2015 Accepted date: 6 November 2015 Cite this article as: Tzong-Yuan Wu, Yi-Yu Su, Wei-Hsien Shu, Augustus T. Mercado, Shi-Kwun Wang, Ling-Yi Hsu, Yow-Fu Tsai and Chung-Yung Chen, A Novel Sensitive Pathogen Detection System Based on Microbead Quantum Dot System, Biosensors and Bioelectronic, http://dx.doi.org/10.1016/j.bios.2015.11.016 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

A Novel Sensitive Pathogen Detection System Based on Microbead Quantum Dot System Tzong-Yuan Wua,b,c, Yi-Yu Sua, Wei-Hsien Shua, Augustus T. Mercadoa,e, Shi-Kwun Wangd, Ling-Yi Hsud, Yow-Fu Tsaie, Chung-Yung Chena,b,*

a

Department of Bioscience Technology, Chung Yuan Christian University, Chung-Li, 320, Taiwan

b

Center for Nanotechnology and Center for Biomedical Technology, Chung Yuan Christian University,

Chung-Li, 320, Taiwan c

R&D Center of Membrane Technology, Chung Yuan Christian University, Chung-Li, 320, Taiwan

d

Enviromental Analysis Laboratory, Environmental Protection Administration, Chung-Li, 320, Taiwan

e

Department of Chemistry, Chung Yuan Christian University, Chung-Li, 320, Taiwan

* To whom correspondence should be addressed: Chung-Yung Chen. Tel: 886 3 2653540; Fax: 886 3 2653599; E-mail addres:[email protected]

Present Address: Chung-Yung Chen, Department of Bioscience Technology, Chung Yuan Christian University, Chung-Li, 320, Taiwan E-mail:[email protected]

Abstract A fast and accurate detection system for pathogens can provide immediate measurements for the identification of infectious agents. Therefore, the Microbead Quantum-dots Detection System (MQDS) was developed to identify and measure target DNAs of pathogenic microorganisms and eliminated the need of PCR amplifications. This nanomaterial-based technique can detect different microorganisms by flow cytometry measurements. In MQDS, pathogen specific DNA probes were designed to form a hairpin structure and conjugated on microbeads. In the presence of the complementary target DNA sequence, the probes will compete for binding with the reporter probes but will not interfere with the binding between the probe and internal control DNA. To monitor the binding process by flow cytometry, both the reporter probes and internal control probes were conjugated with Quantum dots that fluoresce at different emission wavelengths using the click reaction. When MQDS was used to detect the pathogens in environmental samples, a high correlation coefficient (R=0.994) for Legionella spp., with a detection limit of 0.1 ng of the extracted DNAs and 10 CFU/test, can be achieved. Thus, this newly developed technique can also be applied to detect other pathogens, particularly viruses and other genetic diseases.

Key Words: Microbead Quantum-dots Detection System (MQDS); quantum dots; flow cytometry; pathogen; hairpin structure; the click reaction

1

1. Introduction Legionnaires' disease (also legionellosis or Legion fever) is a disease caused by Legionella pneumophila. Approximately 80–90% of the reported cases of this disease are attributed to L. pneumophila (Muder and Yu, 2002). Legionellosis is caused by sporadic and epidemic cases of atypical pneumonia. L. pneumophila belongs to Legionella spp., which are pathogenic gram-negative bacteria and are usually found in contaminated water supplies and air conditioning systems, particularly in hospitals. The mortality of patients with pneumonia exceeds 30% for elderly and immunocompromised patients. The time-consuming identification of the causative agent (Heath et al., 1996) has prognostic significance in delayed therapy of Legionnaires' disease. Therefore, the development of a validated and rapid diagnostic assay is very important in public health care. The standard laboratory detection of Legionella infections is diagnostic culture which is largely dependent on the physiological state of the cells. However, the identification of Legionella spp. Requires an approximately one week incubation period. Another drawback is that the viable but non-culturable bacteria, or legionellae, present within their protozoan hosts in the tested water sample cannot be detected by culture, leading to exposure to an undetected Legionella source (Delgado-Viscogliosi et al., 2009; Hay et al., 1995). Molecular assays based on 16S rRNA (Stolhaug and 2

Bergh 2006), 23S rRNA (Nazarian et al., 2008), defective organelle trafficking (dotA) (Yanez et al., 2005), gyrase subunit B (gyrB) (Zhou et al., 2011), LysR-type transcriptional regulator (LTTR) (Cho et al., 2015), and macrophage infectivity potentiator (mip) genes (Wilson et al., 2003) were used for the detection of L. pneumophilia strains, but there are critical defects in the diagnosis and identification of all Legionella isolates. For example, dotA has a high level of allelic diversity, which is a drawback for the detection of specific species. However, the efficacy of any DNA-based detection method significantly depends on the uniqueness of the sequence to the pathogen of interest. PCR amplification is needed prior to sequencing to identify the species. A previous study has successfully identified and specified the strains of Legionella using mip genes and the PCR method, but most of their primers were degenerate (Ratcliff et al., 1998). Another study by Stolhaug and Bergh used the mip gene sequence to discriminate between Legionella species more reliably than methods using the 16S rRNA gene sequence; however, this method required sequencing after RT-PCR targeting. DNA polymerase is prone to errors, which, in turn, causes mutations in the PCR fragments that are produced. Additionally, the specificity of the PCR fragments can mutate the template DNA, due to nonspecific binding of primers (Garibyan and Avashia, 2013). In another study, a sensitive and specific quantitative PCR assay has 3

been developed for L. pneumophila; nevertheless, different strains require individual detection, resulting in higher costs (Reischl et al., 2002). The microarray method can detect many species/strains in one chip; unfortunately, the accuracy of its quantitative determination is low (Su et al., 2009). To date, effective monitoring and a quick response to infectious diseases in the early stage of outbreaks have become an important issue. Thus, the development of efficient, sensitive, quick, qualitative and quantitative detection platforms for pathogens (such as Legionella spp.) from environmental and clinical samples is an important research issue and is of clinical interest. This study developed a microbead system (Gao et al., 2011; Han et al., 2001) that detects Legionella spp. using hairpin oligonucleotide templates and quantum dot-labelled probes (Janczewski et al., 2011; Long et al., 2011). The oligonucleotide contains species-specific sequences based on the mip gene, which is unique in every Legionella spp. (Su et al., 2009). The quantum dot-labelled probes make them attractive for in vitro fluorescence diagnostics because of their narrow and symmetric emission band, long luminescence lifetime, and good photostability (Mindt et al., 2008). However, the conjugation of quantum dots (QD) with DNA is a prerequisite for their use as biological probes in fluorescence imaging. Other studies used different methods to construct the QD-DNA nanosensor assembly, such as biotin-streptavidin 4

hybridization (Sharma et al., 2008; Zhang et al., 2005), EDC/NHS (Zhou et al., 2008), and ligand exchange (Zhou et al., 2005). For this study, Click Chemistry was used as a more novel conjugation approach for the quantum dots and DNA probes, as it triggers the covalent assembly of the triazole structure of two functional groups, azide and alkyne. These are introduced on the surface of the quantum dot and at the 3’-end of the oligonucleotide. Click Chemistry can provide efficient and stable quantum dot-labelled oligonucleotide probes (Orski et al., 2010). . Quantum dots have been used to detect nucleic acids (Enkin et al., 2014; Su et al., 2014; Zang et al., 2015; Zhang and Hu, 2010) and different pathogens, such as Escherichia coli (Carrillo-Carrion et al., 2011), Staphylococcus aureus (Shi et al., 2015), and Salmonella enteritis typhimurium (Nguyen et al., 2014). In this study, we exploited the broad excitation spectra of QDs (Bruchez, 2005) to excite a wide variety of QDs by the same laser source. Thus, a single laser flow cytometer can be used for multicolor experiments (Chattopadhyay et al., 2006) to identify the species of Legionella using unique oligonucleotide sequences conjugated with the QDs. This new microbead system has been optimized to ensure sensitivity and specificity and validated using environmental specimens. This technique speeds up the detection of Legionella spp., which significantly contributes to early diagnosis of Legionnaires' 5

disease. Moreover, this system can be developed to detect not only bacterial pathogens but also viruses and genetic diseases.

2. Materials and methods 2.1 Microbead preparation The microbeads (Supplementary Table S1) were precipitated and concentrated using a centrifuge at 14000 rpm for 10 min in room temperature. The beads were re-suspended in 680 μl 0.1 M MES Buffer (pH 5.5) and 100 μl of EDAC solution (20 mg/μl EDAC dissolving in 100 μl MES buffer). After 3 hours of mixing by rotation, 2 μl of 100 μM oligonucleotide probes were added and mixed again for 12 hours (Fig. 1A). The beads were collected by centrifuging the mixture at 14000 rpm for 10 min in room temperature. The microbead pellets were resuspended and stored using 100 μl Storage buffer (0.1 g BSA, 0.01 g sodium azide, 9.5 ml 0.1 M NaH2PO4 and 40.5 ml 0.1M Na2HPO4 in 1 L).

6

Fig. 1. The synthesis process of Microbead Quantum-dots Detection System. (A) The procedure of microbeads conjugated with oligonucleotide. (B) The procedure of quantum-dots coated with polymers to create alkyne functional groups. (C) The procedure of generation of internal/reporter probes with azide functional groups. (D) Click Chemistry Reaction to conjugate quantum-dots and internal/reporter probes by triazole linkage. 7

2.2 Synthesis of an amphiphilic polymer with quantum dots (modified from (Janczewski et al., 2011) A mixture of 0.04 g of poly-(isobutylene-alt-maleic anhydride), 10 ml of dry THF, 34 μl of DIPEA, and 16 μl of n-octylamine was heated at 60°C for 1 hour. After cooling to 30°C, 6 μl of propargylamine was added and heated for another 12 hours (Fig. 1B). The samples were placed in a Rotavap at 30°C for 20 mins to evaporate the THF and DIPEA. The solid residue was dissolved in 1600 μl of water and 520 μl of 1 M NaOH and evaporated to dryness. The residue was re-dissolved in 800 μl water and transferred into a dialysis tube to purify the amphiphilic polymers. The dialysis tube was soaked in 2 liters of water and 1 ml of 1 M NaOH for 12 hours and transferred into distilled water for another 12 hours. This step was repeated three times. The amphiphilic polymers were freeze-dried and re-suspended in water at a concentration of 0.002 g/ml. Quantum dots (Ocean Nano Tech, Arkansas, USA) with different wavelengths (525 nm, 605 nm, 655 nm and 705 nm) were used in this study. The QDs were suspended in 200 µl of THF and mixed with an equal volume of CH3OH. The mixture was centrifuged at 6500 rpm for 60 mins at room temperature. The quantum dots were washed with 1 ml of THF and dried in a Rotavap for 20 mins. Then, the residue was resuspended in 1 ml of water and dried again in a Rotavap for 4 hours. The samples were filtered through a 0.22 μm Millex PES membrane and 20 μl of 1 M 8

NaOH was added to the filtrate. Finally, the mixture was centrifuged at 25,000 rpm for 2 hours at room temperature and the supernatant was discarded. The resulting quantum dots were resuspended in 1 ml of water.

2.3 Oligonucleotide probes conjugated with Azide The oligonucleotide sequences of the internal probe (IP) were designed so that its complementary sequence had an extra “T” base on the 5’-end, designated as anti-IP (Fig. 1C). Additionally, the reporter probes (RP) were complementary, with an anti-RP with an extra “T” base on the 5’-end. One-hundred-micromolar solutions of each of the IP/anti-IP and RP/anti-RP mixtures, 10 μl each, were heated at 95°Cfor 5 minutes and cooled to room temperature. After cooling, 1 μl Taq polymerase (2U), 10x Taq buffer and 4 μl 10 mM of N6-(6-Azido) hexyl-3’-dATP (Azide) were added to the mixtures in a total volume of 10 μl and conjugated at 72°C for 30 seconds. Finally, the probes with azide functional groups were denatured at 95°C for 5 mins and immediately cooled on ice before undergoing the Click Chemistry reaction.

2.4 Click Reaction for QD conjugated with probes Click reaction (Mindt and Schibli, 2007; Ming et al., 2008) was derived from the manufacturer’s protocol (Invitrogen, USA). A reaction mixture containing 60 μl of 9

‘Click-iT

reaction

buffer’,

40

μl

of

azide-oligonucleotide,

50

μl

of

alkyne-polymer-QDs, 100 μl of ‘Step-1’ solution, and 10 μl ddH2O was mixed thoroughly. 10 μl of CuSO4, 10 μl ‘Click-iT reaction buffer additive 1 solution’ and 10 μl ‘Click-iT reaction buffer additive 2 solution’ was added in this chronological order. The QD conjugated with probes were purified using Sephadex G75 column and Sephadex G100 column (Fig. 1D). 2.5 DNA extraction from Legionella Culture Medium Legionella pneumophilia was cultured at 15 mL medium for one week. The culture medium was transferred to centrifuge tubes to collect the bacterial cells and spun at 3000 rpm for 10 mins. The supernatant was discarded and 475 µl of STE Buffer, 25 µl of 10% SDS, 2.5 µl of proteinase K, and 5 µl of RNase A was added to the tubes. After the mixture was incubated with shaking at 37°C for 1 hour, 90 µl of 5 M NaCl and 75 µl of 10% Cetyltrimethyl ammonium bromide (CTAB) in 0.7 M NaCl was added to the mixture. This was incubated for 20 mins at 65°C. Then, 500 µl of phenol:chloroform:isoamyl alcohol (25:24:1) was added and mixed with a 500 µl of the cell lysis. The resulting mixture was centrifuged at 4500 rpm at 4°C for 10 minutes. The upper layer of the resulting mixture was collected and added with isopropanol. The DNA was collected by centrifugation at 3500 rpm for 30 mins at

10

4°C. The extracted DNA was washed with 1 ml 70% ethanol twice and spun down at 13000 rpm for 5 mins. The DNA was resuspended with 200 µl STE buffer.

2.6 Hybridisation with DNA fragments and QD-probes The genomic DNAs of Legionella spp. were sonicated to produce 100-200 bp DNA fragments (Fig. S3) and added into a mixture of 20 μl of the microbeads-oligonucleotides and 100 μl hybridisation buffer (10% BSA, 7% SDS, 1 mM EDTA and 0.5 M sodium phosphate). The DNAs were denatured at 95°C for 10 minutes and re-natured at room temperature. QD-internal probes (QD-IP), QD-reporter probes (QD-RP), and 100 μl hybridisation buffer with 50% formaldehyde were added to the cooled mixture and incubated for 2 hours at 42°C. The resulting mixture was washed with 500 μl washing buffer (2X SSC). The QD fluorescence was measured using a flow cytometer.

2.7 Sensitivity, specificity, standard curve, detection limit and stability For the stability test, probes bonded with QDs from the Click reaction and microbeads conjugated with oligonucleotides were treated in various storage temperatures (-20°C, 4°C, 25°C, 37°C, and 60°C), pH values (5, 7, 9, 11, and 13) at different time periods (1 week, 2 weeks, 3 weeks, 4 weeks, and 2 months). The optical 11

density (OD) of the DNAs and QDs were measured using a photometer at different wavelengths (260, 525, 605, 655, and 705 nm) after usage, precipitation, and re-suspension. For the detection limit assay, different concentrations of the bacterial solution were prepared by serial dilution to determine the detection limit of the proposed technology. The cultured bacterial cells were lysed and directly hybridized with the microbead-oligonucleotide system. The QD fluorescence of the hybridized probes was measured by flow cytometry and standard curves were generated (Fig. S4). Finally, the sensitivity and specificity assay was performed using 500 ng of Legionella pneumophila DNA mixed with various amounts of DNA from other species, and the QD fluorescence was measured using a flow cytometer. The percentages of false positives and false negatives were calculated to evaluate the accuracy rate.

2.8 Water samplings A total of 26 water samples were collected from different sources (drinking water, tap water, bottled water, pool water, pond and waste water canal) during July to September, 2012. These samplings were granted by the Taiwan government institution, Enviromental

Analysis

Laboratory of

Environmental

Protection

Administration (EPA-101-U1U4-04-001). The samplings were collected within the 12

vicinity of Chung-Li City in Taoyuan County of Taiwan (See Supplementary Table S7 for sample types and GPSs of samples). Water samples were collected from public areas (pool, pond, and waste water canal) and inside Chung Yuan Christian University (drinking water, tap water, bottled water). The field studies did not involve any endangered or protected species. Two litres of the water sample were collected and centrifuged at 3000 rpm for 1 hour. Samples was resuspended in 100 μl ddH2O and 100 μl of HCl-KCl buffer (0.2 M KCl, pH=2.2 adjusted by HCl) was added (Water sampling protocol modified from EPA, Taiwan). Water samples were tested using the developed microbead-oligonucleotide system and the traditional assay for Legionella spp. The samples were sonicated using 10 × 10 sec pulses at 90% effect (Sonifier® B-12, Branson Sonic Power, Danbury, CT, USA) before direct hybridization (Fig. S3).

2.8 Statistic methods Statistical analysis was performed with SPSS Version 10.0 for Windows (SPSS, Inc., Chicago USA). The χ2 test was adopted to examine the difference of flow events. The Kolmogorov-Smirnov test was used to estimate the statistical significance of differences between the traditional and MQDS method. All of the P-value were two-tailed. 13

3. Results and Discussion 3.1 The design of Microbead Quantum-dots Detection System Different nanomaterials (poly-methyl-methacrylate/PMMA, polystyrene, and silica) and sizes (5mm, 8.5mm, 20mm) of the microbeads were used to examine the efficiency of the link between the oligonucleotide probes and the microbeads (see the Supporting Information Table S1). The oligonucleotide probe is functionally designed and composed of four fragments: poly-T linker, Tag sequence (Table 1, underlined letters and Fig. 2B), barcode region (also called “words”, Table 1, block letters and Table S2) and an anti-Tag sequence (Table 1, italicized letters and Fig. 2B). These oligonucleotide probes were conjugated with carboxylated microbeads through an amide bond. The anti-Tag sequence is complementary to the Tag sequence and is derived from the species-specific mip gene sequence of Legionella spp.(Su et al., 2009) (Table 1). Theoretically, the Tag and anti-Tag sequences will spontaneously form a hairpin structure without the presence of a target DNA.

14

15

Fig. 2. The principle of the Microbead Quantum-dot Detection System. (A) The structure of the microbead-oligonucleotide and the MQDS procedure. (B) The composition of the oligonucleotide hairpin includes: a poly T linker, a Tag sequence, the words and an anti-Tag sequence.

Two probes, an internal probe (IP) and a reporter probe (RP), conjugated with different emission spectra of quantum dots were prepared (see the Table S2 and Table S3). The sequence of the IP is complementary to the barcode region which is composed of four “words” (Brenner et al., 2000) (one word has four nucleotides). Four types of words were used in this research; thus, a combination of four words generated a total of sixteen nucleotides (Table 1 and Table S2). The nucleotide sequences of the generated four “words” has a unique a design for each species. In addition, the sequence of the RP is complementary to the anti-Tag sequence (Table 1 and Table S2). The detection strategy of this system has four steps: denaturation, annealing, hybridisation, and detection. After denaturation of the hairpin structure at 95°C, the linear oligonucleotide-microbead will hybridise with denatured DNA samples (Legionella spp. DNA was used as targets in this study); thus, it will block the formation of the hairpin structure of the DNA fragment attached to the microbead and will remain linear. The linear oligonucleotide-microbead will hybridise with both 16

internal probes and reporter probes (Fig. 2A). However, in the samples without the target DNAs, the hairpin structure will continue to form and prohibits its hybridisation with the IPs and RPs. The probes hybridized with oligonucleotide-microbeads present in two different emission spectra of quantum dots can provide both qualitative and quantitative information using a flow cytometer (BD, FACS Calibur). The fluorescence intensity of the quantum dots-labelled reporter probes can quantitatively measure the target DNA. If there is no target DNA during the denaturation and hybridization process, a fluorescent signal will not be detected because of the re-formation of the hairpin structure attached to the microbead. This system is designated as the Microbead Quantum-dots Detection System (MQDS).

3.2. The characteristic of Microbead Quantum-dots Detection System To verify the optimum size and material for the microbeads and to evaluate the conjugation efficiency of the microbead-oligonucleotides in this detection system, four different microbeads were used, including 20mm PMMA, 8.5mm and 20mm polystyrene and 5mm silica. Oligonucleotide probes with the same sequences were conjugated and hybridized with the same set of IPs and RPs. The number of oligonucleotide probes linked to one microbead was tested and showed that a 1:103 ratio of beads:probe had a better hybridization result in the MQDS. 17

The hybridization efficiency of the IC/RP from different materials and sizes of microbeads was determined by comparing the ratios of the fluorescence (I/R) of the IC/RP applied to the MQDS (Table S3 and S6). The data obtained would also determine the annealing ability of IPs and RPs due to structural interference of the long oligonucleotides. The I/R ratio of 20mm PMMA is closer to 1, which suggests that PMMA is a suitable material and provides stable IP/RP hybridization in the MQDS

(Table

S3).

For

storage

purposes,

the

stabilities

of

the

microbead-oligonucleotides and probe-quantum dots were examined at different temperatures and pH conditions over four weeks. The microbead-oligonucleotides and probe-quantum dots were stable in the pH range of 5-9 below 37°C for storage of over one

month

(Fig.

S1).

However,

the

oligonucleotide

probes

and

microbead-oligonucleotides degraded within one week when stored at a temperature of 60°C. In addition, the oligonucleotides were unstable at -20°C and 4°C at a pH of over 11. Knowing the storage conditions is beneficial in the preparation of labelled probes and its usage.

3.3. The detection limit, sensitivity and specificity of Microbead Quantum-dots Detection System

18

To estimate the detection limit of MQDS, DNA fragments from Legionella pneumophila underwent a series of dilutions (from 10 µg to 0.05 ng). These different concentrations of DNA were hybridised into MQDS using IP-QD525nm and RP-QD605nm and monitored in the flow cytometer. A linear correlation was found within the detection range between 0.1±0.01 ng and 1000±1.10 ng with a high correlation coefficient (R2=0.994) (a plot for the reporter probe-QD605nm is shown in Fig. 3A and Table S8). The detection limit (counted as the smallest value of linear correlation) was found to be as less as 0.1 ng of extracted L. pneumophila DNA in single test (Fig. 3A and Table S8). Seven other Legionella spp. (Table S2 and Table S8) were also examined and had a similar detection limit and range. Serial dilutions of L. pneumophila cells (‫أ‬10, 100, 1000, 10000, 100000 CFU/test, homogenized raw samples) were also tested in the MQDS and demonstrated a high correlation coefficient (R2=0.999, a plot for reporter probe-QD605nm is shown in Fig. 3B). In Figure 3B, the results indicated that the MQDS was able to detect >10 CFU/test of L. pneumophila in a single test. In addition, the other Legionella spp. were also detected at >10 CFU/test of bacterial cells in a single test. The detection limit using the MQDS is not comparable to the RT-PCR method used in a previous study, which can detect 10 fg of extracted DNA (Wilson et al., 2003). In a more recent study, RT-PCR method had a detection limit 5 fg/µl reaction mix and 7.4 CFU/ml reaction mix using 19

LTTR family gene (Cho et al., 2015). However, these studies only quantitatively detected and discriminated L. pneumophilia.

Fig. 3. The standard curves of MQDS. (A) To demonstrate the detection range between 0.1±0.01ng and 1000±1.10ng (n=3) with a high correlation coefficient (R2=0.994), L. pneumophila DNA was diluted into 5000 ng, 1000 ng, 500 ng, 100 ng, 50 ng, 10 ng, 1 ng, 0.1 ng and hybridized in the MQDS (see Table S8). (B) The serial dilutions of the crude extract of L. pneumophila cells (105, 104, 103, 102, 101 and 20

negative control) were hybridized in the MQDS. The standard curve generated a high correlation coefficient (R2=0.999). The flow events of the reporter probe-QD605nm and DNA concentrations were plotted (flow gated events: 50000) in Fig. 3A and 3B.

The sensitivity of the flow cytometry detection using an enzyme-linked immunosorbent assay is only adequate for symptomatic samples. Moreover, in some cases, cross-reactions with L. pneumophilia and other bacteria are also observed (Fuchslin et al., 2010; Keserue et al., 2012). In this study, eight common Legionella spp. (Table S5, item 1~8) and other eight gram-positive and -negative strains (Table S5, items 9-16) were tested for non-specific hybridization. Any non-specific bacterial DNAs that hybridized to the anti-tag and/or tag sequence in the MQDS will deform the hairpin structure and increase the fluorescence values of the internal control probe, giving a false positive result. In addition, any non-specific bacterial DNAs that interfere with the hybridization of the RP and/or the target bacterial DNA will decrease the RP values, resulting in a false negative result. In Table S6, an additional 0.1-10000 ng of non-specific DNAs from fifteen different bacteria were added into the MQDS. The flow events of the IP/RP probes of L. pneumophila did not show any significant variation (Table S6, Kruskal-Wallis test, p<0.05). The accuracy of other seven Legionella spp. was also tested and shown in Fig. S2. These results demonstrated that this system accurately ligates the correct and specific DNA without 21

the influence of ‫أ‬1000 ng of fifteen different bacterial DNAs present in the mixture

(Table S6 and Fig. S2). Moreover, the microbead-oligonucleotides for the eight Legionella spp. did not result in any false positive or false negative results in this detection range in the presence of non-specific internal and reporter probes. However, false positive results occurred and the correlation coefficient of standard curve was reduced (R2=0.707) in the presence of over 5 μg of background DNAs (other species DNAs); however, the sensitivity remained unchanged.

3.4. The application of Microbead Quantum-dots Detection System The combination of the different emission spectra of the quantum dots conjugated with the internal probe and reporter probe will serve as barcodes to detect the various Legionella species in a single reaction by distinguishing the emission spectrum patterns of the quantum dots. For this purpose, quantum dots with four different emission spectra (525 nm, 605 nm, 655nm and 705 nm) were used to label the individual species-specific IPs and/or RPs and mixed in one hybridization process (Table 2 shows the combination of quantum dots for each Legionella species). For example, the IP-QD of L. pneumophila emits at 525 nm and its RP-QD emits at 605 nm; the IP-QD of L. lansingensis emits at 605 nm and its reporter probe-QD emits at 22

705 nm. The flow event data demonstrated that the combination of quantum dots from unique IP/RP probes accurately detect eight Legionella species in one MQDS test (Table 2). These results indicate that the MQDS is capable of detecting various species in a single test. Table 2. The combination of three different spectrums of quantum dots to distinguish eight Legionella species 605 nm/655

IP

RP

525 nm1

525 nm

605 nm

0

6 ± 2.7

6 ± 0.6

525 nm

605 nm

4 ± 1.0

13 ± 1.0

60 ± 1.2

L. anisa L. birminghamenesis

605 nm 525 nm

605 nm 705 nm

52 ± 19.7 948 ± 16.6

2297 ± 95.8 0

40 ± 0.0 1082 ± 29.1

L. gormanii

525 nm

655 nm

1411 ± 41.6

1416 ± 28.4

27 ± 3.6

L. lansingensis

605 nm

705 nm

1 ± 0.6

1705 ± 32.7

167 ± 36.1

L. longbeachae L. oakridgensis

525 nm 655 nm

525 nm 705 nm

1519 ± 173.0 1 ± 0.6

66 ± 40.2 692 ± 10.8

10 ± 3.0 2105 ± 31.3

L. pneumophilia

525 nm

605 nm

1781 ± 17.0

2316 ± 347.6

0

L. wadsworthii

705 nm

705 nm

6 ± 2.1

1 ± 1.5

1750 ± 19.9

Blank Negative Control

1

3

nm 1,2

705 nm1

The data were shown as the average ± SD fluorescence events detected using a flow cytometer and

different filters. 2

Only three filters were installed in the flow cytometer. For L. gormanii and L. oakridgensis, 525 nm,

655 nm and 705 nm filters were applied, and 525 nm, 605 nm and 705 nm filters were used for the other six Legionella species and controls. 3

The negative control used 500 ng of DNA from E. coli strain DH5a. Both the blank and negative

control used the L. pneumophila internal control and reporter probes for detection. For the Legionella species’ DNA, 75 ng of DNA was applied for detection. Moreover, all of the internal/reporter probes of the eight Legionella species were mixed together into a single reaction to test the accuracy of detection.

The MQDS was applied to environmental testing using 26 collected water samples from different sources (drinking, tap, bottled, pool, pond, and waste water). The 2 L

23

of water samples were collected around Chung-Li City in Taiwan within July to September 2012. The samples were collected using the standard sampling protocols of the EPA (Environmental Protection Administration) of Taiwan (see Table S7). For each sample, half was cultured using the traditional method (BCYE selected medium) to determine the cell numbers and identify the Legionella 16S rRNA genes by PCR (Miyamoto et al. 1997), followed by sequence confirmation. The other half was directly detected by the MQDS. Ten water samples were positive for Legionella using both methods (Table 3). From the blind sample analysis, the results showed no difference in the typing results between the two methods. The same species of Legionella was found in the same samples. The cell numbers of the Legionella spp. that were calculated by traditional method and the MQDS are shown in Table 3. To validate the quantitative results, the cell numbers of the Legionella spp. in the water samples were not significantly different between the traditional method and the MQDS by the Kolmogorov-Smirnov test (p= 0.07) (Lilliefors 1967). This suggests that the proposed system can be used directly on water samples without prior preparations, cultures, and amplifications. In addition, the MQDS was able to identify the species present in the samples in one days, while the traditional method lasted for 1-2 weeks. Table 3. The accuracy of MQDS estimated from water samples. 24

1

Sample

1

525 nm

605 nm

705 nm

3

5

1

0

2

Bacterial count2,3

Species

1

MQDS

16srRNA

MQDS

BCYE

16

-

-

-

-

26258

38185

L. lansingensis

L. lansingensis

7224

6110

19534

23638

0

L. pneumophila

L. pneumophila

3843

3540

3

21308

0

25371

L. birminghamenesis

L. birminghamenesis

5922

5280

4

5570

6440

0

L. pneumophila

L. pneumophila

7699

5800

5

28402

32781

7

L. pneumophila

L. pneumophila

5897

5520

6

64

12960

34539

L. lansingensis

L. lansingensis

6279

6320

7

21925

33749

551

L. pneumophila

L. pneumophila

5922

5280

8

2002

215

35407

L. birminghamenesis

L. birminghamenesis

7699

5800

9

817

29510

876

L. anisa

L. anisa

5059

6320

10

543

16083

236

L. anisa

L. anisa

2551

1120

Control

4

1

The data are shown as the average fluorescence events detected from the flow cytometer using

different filters. 2

Two liter water samples were collected. One liter was tested using the MQDS and the other was

confirmed by BCYE culture and 16S rRNA PCR and sequencing. 3

To test the differences in the MQDS and BCYE data, the h = ks test2(x1, x2) was applied to perform a

two-sample Kolmogorov-Smirnov test to compare the distribution of the values in the two data vectors x1 and x2. After the two-sample Kolmogorov-Smirnov test, the KS test statistic = 0.2500, p=0.929, h=0. H0: The null hypothesis is that x1 and x2 are from the same continuous distribution. <==> h=0 H1: The alternative hypothesis is that they are from different continuous distributions. <==> h=1 4

As a control, 500 ng of DNA from E. coli strain DH5a was applied.

4. Conclusion The main purpose of this project is to adopt quantum dots and microbeads to generate a new system to detect pathogens, including bacteria and viruses. This new technology quantitatively and qualitatively measures the sample DNAs using stable fluorescence and a modified hybridization method, without prior DNA amplification.

25

The backbone of this detection system is the formation of hairpins in the oligonucleotide probes attached to the PMMA microbeads. Two specific probes were conjugated with quantum dots with different emission spectra, which can be measured using a flow cytometer. We have demonstrated that the Click Reaction linkage between the oligonucleotide DNA and quantum dot surfaces improved the labelling efficiency and detection sensitivity. This system shows high specificity and sensitivity and a detection limit of 10 CFU/test of Legionella spp., with a high linear correlation coefficient of the standard curve (R2=0.994). Furthermore, the various Legionella species were detected in a single test. Similar to the traditional method, the MQDS accurately quantified the amount of bacteria present in the field water samples. The advantages of this system are the quick qualitative detection and quantitative determination from water samples, short reaction time, recognition of eight Legionella species from a single test, and lower cost due to its stable fluorescence and simple sample preparation.

Acknowledgements The authors would like thank Prof. Tsung-Chain Chang from National Cheng Kung University for providing Legionella spp. cultures. This work was supported by the Ministry of Science and Technology, Taiwan [NSC-102-2632-M-033-001-MY3 and 26

MOST-104-2321-B-033-001]; Environmental Protection Administration, Taiwan [EPA-100-U1U4-04-001 and EPA-101-U1U4-04-001] and Center for Nanotechnology and Center for Biomedical Technology, Chung Yuan Christian University, Chung-Li, Taiwan.

References Brenner, S., Johnson, M., Bridgham, J., Golda, G., Lloyd, D.H., Johnson, D., Luo, S., McCurdy, S., Foy, M., Ewan, M., Roth, R., George, D., Eletr, S., Albrecht, G., Vermaas, E., Williams, S.R., Moon, K., Burcham, T., Pallas, M., DuBridge, R.B., Kirchner, J., Fearon, K., Mao, J., Corcoran, K., 2000. Nat. Biotechnol. 18, 630-634. Bruchez, M.P., 2005. Curr. Opin. Chem. Biol. 9, 533-537. Carrillo-Carrion, C., Simonet, B.M., Valcarcel, M., 2011. Biosens. Bioelectron. 26, 4368-4374. Chattopadhyay, P.K., Price, D.A., Harper, T.F., Betts, M.R., Yu, J., Gostick, E., Perfetto, S.P., Goepfert, P., Koup, R.A., De Rosa, S.C., Bruchez, M.P., Roederer, M., 2006. Nat. Med. 12, 972-977. Cho, M.S., Ahn, T.Y., Joh, K., Lee, E.S., Park, D.S., 2015. Appl. Microbiol. Biotechnol. DOI: 10.1007/s00253-015-6759-y. Delgado-Viscogliosi, P., Solignac, L., Delattre, J.M., 2009. Appl. Environ. Microb. 75, 3502-3512. Enkin, N., Wang, F., Sharon, E., Albada, H.B., Willner, I., 2014. Acs. Nano. 8, 11666-11673. Fuchslin, H.P., Kotzsch, S., Keserue, H.A., Egli, T., 2010. Cytometry A 77, 264-274. Gao, Y., Stanford, W.L., Chan, W.C., 2011. Small 7, 137-146. Garibyan, L., Avashia, N., 2013. J. Invest. Dermatol. 133, e6. Han, M., Gao, X., Su, J.Z., Nie, S., 2001. Nat. Biotechnol. 19, 631-635. Hay, J., Seal, D.V., Billcliffe, B., Freer, J.H., 1995. J. Appl. Bacteriol. 78, 61-65. Heath, C.H., Grove, D.I., Looke, D.F., 1996. Eur. J. Clin. Microbiol. Infect. Dis. 15, 286-290. Janczewski, D., Tomczak, N., Han, M.Y., Vancso, G.J., 2011. Nat. protocols 6, 1546-1553. Keserue, H.A., Baumgartner, A., Felleisen, R., Egli, T., 2012. Microb. Biotechnol. 5, 753-763. Lilliefors, H.W., 1967. J. Am. Stat. Associ. 62, 399-402. Long, F., Wu, S., He, M., Tong, T., Shi, H., 2011. Biosens. Bioelectron. 26, 2390-2395. Mindt, T.L., Muller, C., Melis, M., de Jong, M., Schibli, R., 2008. Bioconjug. Chem. 19, 1689-1695. Mindt, T.L., Schibli, R., 2007. J. Org. Chem. 72, 10247-10250. Ming, X., Leonard, P., Heindl, D., Seela, F., Nucleic. Acids Symp. Ser. 52, 471-472. Miyamoto, H., Yamamoto, H., Arima, K., Fujii, J., Maruta, K., Izu, K., Shiomori, T., Yoshida, S., 1997. Appl. Environ. Microbiol. 63, 2489-2494. 27

Muder, R.R., Yu, V.L., 2002. Clin. Infect. Dis. 35, 990-998. Nazarian, E.J., Bopp, D.J., Saylors, A., Limberger, R.J., Musser, K.A., 2008. Diagn. Microbiol. Infect. Dis. 62, 125-132. Nguyen, P.D., Son, S.J., Min, J., 2014. J. Nanosci. Nanotechno. 14, 5646-5649. Orski, S.V., Poloukhtine, A.A., Arumugam, S., Mao, L., Popik, V.V., Locklin, J., 2010. J. Am. Chem. Soci. 132, 11024-11026. Ratcliff, R.M., Lanser, J.A., Manning, P.A., Heuzenroeder, M.W., 1998. J.Clinic. Microbiol. 36, 1560-1567. Reischl, U., Linde, H.J., Lehn, N., Landt, O., Barratt, K., Wellinghausen, N., 2002. J.Clinic. Microbiol. 40, 3814-3817. Sharma, J., Ke, Y.G., Lin, C.X., Chhabra, R., Wang, Q.B., Nangreave, J., Liu, Y., Yan, H., 2008. Angew. Chem. Int. Edit. 47, 5157-5159. Shi, J.Y., Chan, C.Y., Pang, Y.T., Ye, W.W., Tian, F., Lyu, J., Zhang, Y., Yang, M., 2015. Biosens. Bioelectron. 67, 595-600. Stolhaug, A., Bergh, K., 2006. Appl. Environ. Microbiol. 72, 6394-6398. Su, H.P., Tung, S.K., Tseng, L.R., Tsai, W.C., Chung, T.C., Chang, T.C., 2009. J.Clinic. Microbiol. 47, 1386-1392. Su, S., Fan, J.W., Xue, B., Yuwen, L.H., Liu, X.F., Pan, D., Fan, C.H., Wang, L.H., 2014. Acs. Appl. Mater. Inter. 6, 1152-1157. Wilson, D.A., Yen-Lieberman, B., Reischl, U., Gordon, S.M., Procop, G.W., 2003. J.Clinic. Microbiol. 41, 3327-3330. Yanez, M.A., Carrasco-Serrano, C., Barbera, V.M., Catalan, V., 2005. Appl. Environ. Microbiol. 71, 3433-3441. Zang, Y., Lei, J.P., Ling, P.H., Ju, H.X., 2015. Anal. Chem. 87, 5430-5436. Zhang, C.Y., Hu, J., 2010. Anal. Chem. 82, 1921-1927. Zhang, C.Y., Yeh, H.C., Kuroki, M.T., Wang, T.H., 2005. Nat. Mater. 4, 826-831. Zhou, D.J., Piper, J.D., Abell, C., Klenerman, D., Kang, D.J., Ying, L.M., 2005. Chem. Commun. 38, 4807-4809. Zhou, D.J., Ying, L.M., Hong, X., Hall, E.A., Abell, C., Klenerman, D., 2008. Langmuir. 24, 1659-1664. Zhou, G., Cao, B., Dou, Y., Liu, Y., Feng, L., Wang, L., 2011. Appl. Microbiol. Biotechnol. 91, 777-787.

28

Figure legend

Fig. 1. The process of synthesizing the Microbead Quantum-dot Detection System. (A) The procedure for conjugating the microbeads with oligonucleotides. (B) The procedure of coating the quantum-dots with polymers to create the alkyne functional groups. (C) The procedure of generating the internal/reporter probes with azide functional groups. (D) Click Chemistry Reaction to conjugate the quantum-dots and internal/reporter probes with a triazole linkage. 29

Fig. 2. The principle of Microbead Quantum-dots Detection System. (A) The structure of the microbead-oligonucleotide and the MQDS procedure.

(B) The

composition of the oligonucleotide hairpin includes: a poly-T linker, a Tag sequence, the words and an anti-Tag sequence.

30

Fig. 3. The standard curves of MQDS. (A) To demonstrate the detection range between 0.1±0.01 ng and 1000±1.10 ng (n=3) with a high correlation coefficient 31

(R2=0.994), the L. pneumophila DNA was diluted to 5000 ng, 1000 ng, 500 ng, 100 ng, 50 ng, 10 ng, 1 ng, 0.1 ng and hybridised in the MQDS (see Table S8). (B) The serial dilutions of the crude extract of L. pneumophila cells (105, 104, 103, 102, 101 and negative control) were hybridized in MQDS. The standard curve generated a high correlation coefficient (R2=0.999). The Flow events of the reporter probe-QD605nm and DNA concentration were plotted (flow gated events: 50000) in Fig. 3A and 3B.

32

TCAATTCTGGAAATGGTGTTAAACCCGGAAAATCGTCATACTAACTACTAA#CGATTTTCCGGGTTTAACACCATTTCCAGAATTGATTTTT

ACACTGATGTTCATTTGTTAGTCTCTTTTTACAATACTACTAATCATCATT#ATTGTAAAAAGAGACTAACAAATGAACATCAGTGTTTTTT

L. pneumophila

L. wadsworthii

The “words” sequences are shown in bold font, and the tag and anti-tag sequences were in underlined and italicized font, respectively.

GCTTAAGTCATAGCGTTTCCATTCTTTATTAAAGATCATCTAACATTCTAA#TCTTTAATAAAGAATGGAAACGCTATGACTTAAGCTTTTT

L. oakridgensis

The base at the 5’ end of the words is paired to the base at the 3’ end of the words to block the hybridization of the internal probes when the hairpin structure occurs.

#

*

CTCCGGATCCACCGTTTCAGGGGCTTTATTTTCTATCATCTAACATTACTA#TAGAAAATAAAGCCCCTGAAACGGTGGATCCGGAGTTTTT

L. longbeachae

AAACCAGGAGTAGGATTCTCTTTTAGAAAAGAATTTCATTCATACTACTAA#AATTCTTTTCTAAAAGAGAATCCTACTCCTGGTTTTTTTT

L. gormanii

CGCGTGCAGTCTTATCGAATTGTTCTTCCCGGGTATCATCTAACTAACTAA#TACCCGGGAAGAACAATTCGATAAGACTGCACGCGTTTTT

AAAGTCTGCTTAACGCCATTGCGAGCTATGTTCGCTCATCATTCATTACTA#GCGAACATAGCTCGCAATGGCGTTAAGCAGACTTTTTTTT

L. birminghamensis

L. lansingensis

GCTTGTCTTCGAGTAAGGTTAAAAACGCATTAGCGTCATCATTCTAAACTA#CGCTAATGCGTTTTTAACCTTACTCGAAGACAAGCTTTTT

Oligonucleotide probe sequences*

L. anisa

Species

Table 1. The nucleotide sequences of four words are species-specific to identify Legionella spp.

Table 2. The combination of three different spectrums of quantum dots to distinguish eight Legionella species IP

RP

525 nm1

605 nm/655 nm 1,2

705 nm1

525 nm

605 nm

0

6 ± 2.7

6 ± 0.6

525 nm

605 nm

4 ± 1.0

13 ± 1.0

60 ± 1.2

L. anisa L. birminghamenesis

605 nm 525 nm

605 nm 705 nm

52 ± 19.7 948 ± 16.6

2297 ± 95.8 0

40 ± 0.0 1082 ± 29.1

L. gormanii

525 nm

655 nm

1411 ±41.6

1416 ± 28.4

27 ± 3.6

L. lansingensis L. longbeachae

605 nm 525 nm

705 nm 525 nm

1 ± 0.6 1519 ± 173.0

1705 ± 32.7 66 ± 40.2

167 ± 36.1 10 ± 3.0

L. oakridgensis L. pneumophilia

655 nm 525 nm

705 nm 605 nm

1 ± 0.6 1781 ± 17.0

691 ±10.8 2316 ± 347.6

2105 ± 31.3 0

L. wadsworthii

705 nm

705 nm

6 ± 2.1

1 ± 1.5

1750 ± 19.9

Blank Negative Control

3

1

The data were shown as the average ± SD fluorescence events detected using a flow cytometer and

different filters. 2

Only three filters were installed in the flow cytometer. For L. gormanii and L. oakridgensis, 525 nm,

655 nm and 705 nm filters were applied, and 525 nm, 605 nm and 705 nm filters were used for the other six Legionella species and controls. 3

The negative control used 500 ng of DNA from E. coli strain DH5a. Both the blank and negative

control used the L. pneumophila internal control and reporter probes for detection. For the Legionella species’ DNA, 75 ng of DNA was applied for detection. Moreover, all of the internal/reporter probes of the eight Legionella species were mixed together into a single reaction to test the accuracy of detection

Table 3. The accuracy of MQDS estimated from water samples.

Sample

525 nm1

605 nm1

705 nm1

Control4

3

5

1

0

2

Bacterial count2,3

Species MQDS

16srRNA

MQDS

BCYE

16

-

-

-

-

26258

38185

L. lansingensis

L. lansingensis

7224

6110

19534

23638

0

L. pneumophila

L. pneumophila

3843

3540

3

21308

0

25371

L. birminghamenesis

L. birminghamenesis

5922

5280

4

5570

6440

0

L. pneumophila

L. pneumophila

7699

5800

5

28402

32781

7

L. pneumophila

L. pneumophila

5897

5520

6

64

12960

34539

L. lansingensis

L. lansingensis

6279

6320

7

21925

33749

551

L. pneumophila

L. pneumophila

5922

5280

8

2002

215

35407

L. birminghamenesis

L. birminghamenesis

7699

5800

1

9 10

817

29510

876

L. anisa

L. anisa

5059

6320

543

16083

236

L. anisa

L. anisa

2551

1120

1

The data are shown as the average fluorescence events detected from the flow cytometer using

different filters. 2

Two-liter water samples were collected. One liter was tested using the MQDS and the other was

confirmed by BCYE culture and 16S rRNA PCR and sequencing. 3

To test the differences in the MQDS and BCYE data, the h = ks test2(x1, x2) was applied to perform a

two-sample Kolmogorov-Smirnov test to compare the distribution of the values in the two data vectors x1 and x2. After the two-sample Kolmogorov-Smirnov test, the KS test statistic = 0.2500, p=0.929, h=0. H0: The null hypothesis is that x1 and x2 are from the same continuous distribution. <==> h=0 H1: The alternative hypothesis is that they are from different continuous distributions. <==> h=1 4

As a control, 500 ng of DNA from E. coli strain DH5a was applied

2

Supporting Information Fig. S1. The stability test for the quantum dot-conjugated probes hybridized with the PMMA-conjugated oligonucleotide at different temperatures and pH values (each test, n=3). The quantum dot probe stability was examined at various temperatures (-20°C, 4°C, 25°C, 37°C, and 60°C), pH values (pH=5, 7, 9, 11, and 13) and storage times (1, 2, 3, and 4 weeks). Fig. S2. Each of the eight Legionella spp. was tested for their specificity by adding the other seven Legionella spp. to monitor the accuracy of the MQDS. The flow cytometer events were monitored to measure the specificity and sensitivity of the MQDS for each specific Legionella spp. probe. A high amount of Legionella spp. DNA was used to estimate the cross-contamination. The eight different Legionella spp. were tested individually and the flow events of reporter probes are shown on the y axis. For example, to test L. anisa, 500 ng of L. anisa DNA was applied to the L. anisa-specific MQDS at number 1 on the x axis. For number 2 on the x axis, 500 ng of L. anisa DNA plus one 500 ng of DNA from a second Legionella species were added to the L. anisa-specific MQDS. For number 8 on the x axis, 500 ng of the L. anisa DNA plus a total of 3500 ng of DNA from the other seven Legionella species were added to the L. anisa-specific MQDS. For each Legionella spp, there was no significant difference between the various bacterial mixtures that were added (χ2 test, p<0.05). Therefore, no cross-contamination from those eight Legionella spp occurred. The data (flow gated events: 50000) are shown as the means±SD, n=3. Fig. S3. The DNAs from the Legionella spp. were sonicated into approximately 200bp fragments before hybridization in the MQDS. Two percent agarose electrophoresis was used to validate the sonicated DNA fragments. Fig. S4. The histogram plots of the signals from the flow cytometry of the serial dilution test in the MQDS. A, Ten micrograms, 1 µg, 100 ng, and 10 ng of L. pneumophila DNA were tested in the MQDS. The histogram plots of the flow cytometry and fluorescence events were presented. B, The reporter probe with QD605 is detected using the 605 nm filter, but not the 525 nm or 706 nm filters. C, To detect multiple species, the internal probe and reporter probe were individually 3

conjugated with QD605 and QD525 and the flow results to demonstrate the events collected from the 605 nm and 525 nm filters, but not the 705 nm filter. Ňig. S5. The standard curves for other seven Legionella spp., L. anisa, L. birminghamensis, L. gormanii, L. lansingensis, L. longbeachae, L. oakridgensis, and L. wadsworthii. Each Legionella DNA was diluted into 5000 ng, 1000 ng, 500 ng, 100 ng, 50 ng, 10 ng, 1 ng, 0.1 ng and hybridized in MQDS. Table S1. Different sizes and materials of microbead used in this research. Description: A table that lists all of material, coating, diameter of microbeads. (DOC) Table S2. The sequences of the internal probes and reporter probes of eight Legionella spp. A table that lists the oligonucleotide sequences of internal probes and reporter probes using in this study. (DOC) Table S3. A comparison of the different microbead materials. The flow cytometry events of the different microbead materials using the same DNA concentration (200 μg) and probes. (DOC) Table S4. Conjugation of the quantum dots with oligonucleotides by the Click Reaction. A comparison of the linkage efficiency of the Click Reaction and streptavidin-biotin conjugation. The QD events detected by flow cytometer were counted under the identical hybridization process. (DOC) Table S5. Bacterial strains used in this research. The sixteen bacterial strains used in this research. (DOC) Table S6. The specificity and sensitivity of the MQDS for L. pneumophila. The flow cytometer events were monitored to measure the specificity and sensitivity of the MQDS. (DOC) Table S7. The sample types and locations of the collected samples used in this study. (DOC) Table S8. Raw data for the standard curves and detection limit tests of the eight Legionella spp. (DOC) 4

Table S9. Intra-assay and Inter-assay of MQDS for eight Legionella spp. (DOC) Table S10. The conjugation rate of the QD-probes and microbead-oligonucleotides. (DOC) Highlights l

We develop a new detection system to qualitative and quantitative determination for pathogens.

l

We labelled quantum dots into oligonucleotide probes by Click reaction.

l

The advantages of this new system are easy to operate, longer storage, qualitative detection as PCR and quantitative determination with a amplification-free approach.

l

This system shows high specificity and sensitivity with the detection limit as 10 CFU/test of candidate pathogens and a high linear correlation coefficient of the standard curve.

l

Using the combination of different quantum dots as barcode, the various pathogen species were able to be distinguished in a single test from dirty water samples directly.

l

This technology is not only for bacterial but also for virus, human chromosomal copies and others.

5